3EPA
            United States
            Environmental Protection
            Agency
            Municipal Environmental Research EPA-600/9-80-001
            Laboratory        May 1980
            Cincinnati OH 45268
            Research and Development
Proceedings of a
Symposium on
Economic
Approaches to Solid
Waste Management

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology, Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health  Effects Research
      2.  Environmental Protection Technology
      3,  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8  "Special" Reports
      9.  Miscellaneous Reports
This document is available to the public through the National Technical Informa*
tion Service, Springfield, Virginia  22161.

-------
                                                                   X)
                                        EPA-600/9-80-001
                                        May 1980
          PROCEEDINGS OF A SYMPOSIUM
            ON ECONOMIC APPROACHES
           TO SOLID WASTE MANAGEMENT
                         by

                Robert J. Anderson,  Jr.
                   MATHTECH, Inc.
              Princeton, New Jersey  08540
                Contract No. 68-03-2673
                   Project Officer

                  Oscar W. Albrecht
      Solid and Hazardous Waste Research Division
      Municipal Environmental Research Laboratory
                Cincinnati, Ohio  45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
      OFFICE OF RESEARCH AND DEVELOPMENT
     U, S. ENVIRONMENTAL PROTECTION AGENCY
               CINCINNATI, OHIO  45268

-------
                              DISCLAIMER
         This report has been reviewed by the Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency,  and approved
for publication.  Approval does not signify that the contents necessarily
reflect the views and  policies of the U. S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute endorse-
ment or recommendation for use.

-------
                                  FOREWORD
     The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people.  Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.

     Research and development is that necessary first step in problem solu-
tion and it involves defining the problem, measuring its impact, and search-
ing for solutions.  The Municipal Environmental Research Laboratory develops
new and improved technology and systems for the prevention, treatment, and
management of wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and treatment of
public drinking water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution.  This publication is one of the
products of that research; a most vital communications link between the re-
searcher and the user community.

     On September 19 and 20, 1978, a group of economists interested in
finding solutions to the solid waste problem met at Philadelphia, Pennsylvania.
The express purpose of the meeting was to exchange ideas and views on the
potential role of pricing in municipal solid waste management.  This report
contains the formal papers given at that Symposium.
                                       Francis T. Mayo
                                       Director
                                       Municipal Environmental
                                       Research Laboratory
                                     iii

-------
                                ABSTRACT
        This  report  contains the papers presented at a Symposium on Economic
Approaches to Solid Waste  Management at Philadelphia, Pennsylvania,  on
September 19 and  20,  1978.   The symposium was sponsored  by  the Municipal
Environmental Research  Laboratory of the U.S. Environmental Protection
Agency.  Its purpose  was to review the state of the art in  user charges and
product  charges for non-hazardous municipal solid waste, and  to provide a
forum for researchers to exchange  their views on methods and results in the
field.

        To aid  in  its  assessment  of  the  state of  the  art, the  symposium
participants considered four key questions:

        1.      Do incentive  systems designed to  alter waste
               generation behavior  actually improve the  economic
               efficiency of  solid  waste management and/or the
               equity  of its financing?

        2.      What  is  the  optimal  combination of incentive
               mechanisms?  For example, would  the  best system
               include  some  mix  of a user charge and a product
               charge?

        3.      How do  incentives relate to the choice of resource
               recovery  technology?

        4.      What are the  best  ways to pursue  the  testing of the
               effects of incentive mechanisms?   Should we use
               historical  data?   Should  we  try to  conduct
               experiments?

        The  report was submitted  to  MATHTECH,  Inc., as a subcontractor to
JACA Corp.,  JACA  Contract  68-03-2673 under  the sponsorship of  the U.S.
Environmental Protection  Agency.
                                   IV

-------
                                  CONTENTS


Foreword	   iii

Abs tract  	   iv

Section 1.  Introduction and Summary  	   1

            "Introduction and Summary" (Robert J. Anderson, Jr.) ....   2

            "Can Economic Incentives Help Resolve Solid Waste
            Management Problems? (Haynes C. Goddard)  	   14

Section 2.  Pricing and the Efficiency of Solid Waste Management  ...   17

            "The Economic Efficiency of User Fees:  Some
            Preliminary Empirical Results" (William N. Lanen)  	   18

            "Evaluating the Efficiency of the Solid Waste
            Charge" (Steve Buchanan)  	   47

            "Economic Approaches to Solid Waste Management:
            European Experience" (W. David Conn)  	   66

            "The Administrative Costs of User Charges"
            (Barbara J. Stevens)	   75

Section 3.  Pricing and the Equity of Solid Waste Financing  	   92

            "Equity Considerations in User Charges for Residential
            Refuse Collection and Disposal" (Roger Bolton)  	   93

            "Pricing Municipal Refuse Service:  Potential Effects
            on Municipal Budgets"  (Kenneth L. Wertz)  	   114

Section 4.  Comparison of User Charges and Product Charges  	   120

            "The Optimal Policy Mix for Solid Waste Pricing"
            (James F. Hudson)  	   121

            "Fundamental Comparisons of Alternative Solid Waste
            Management Policies" (Allen K. Miedema)  	   130

-------
Section 5.  Pricing and Resource Recovery Technology  	   155

            "Conditionally Predictive Estimates  of
            Secondary Materials Supply" (Tayler  H. Bingham,
            Curtis E. Youngblood, Philip C.  Cooley)  	   156

            "Secondary Material Demand and Supply Responses"
            (Robert C. Anderson)  	   189

Section 6.  Pricing and Solid Waste Management  	   205

            "The Seattle Solid Waste Management  Experiment:   User
            Charges and Separate Collection of Recyclables"
            (Bernard H. Booms)  	•	   206

Section 7.  Research Directions	   231

            "Research Activities in EPA" (Oscar  W. Albrecht)   	   232

            "Solid Waste Collection/Disposal Economic Research
            Strategies" (G.S. Tolley, V.S. Hastings)  	   240

Appendix A. Agenda  	«	•	• • • •   248

Appendix B. List of Participants  	   249
                                   vi

-------
                                 SECTION 1

                         INTRODUCTION AND SUMMARY


"Introduction and Summary" (Robert J. Anderson, Jr.)«

"Can Economic  Incentives Help  Resolve Solid Waste  Management Problems?
(Haynes C.  Goddard).

-------
                          INTRODUCTION AND SUMMARY



                                    by


                          ROBERT J. ANDERSON,  JR.
INTRODUCTION

        In late  spring  of  1978, Oscar  Al brecht and Haynes Goddard of  the
Municipal  Environental  Research Laboratory,  U.  S.  Environmental Protection
Agency began to  make  tentative plans for a workshop on  the role of pricing in
municipal  solid waste management.   Al brecht and Goddard, both of whom are
resource economists actively studying pricing policies  for residential  solid
waste  management, felt  that the pace of research on  the subject had so
quickened over the last few years that the time was  ripe  for researchers
working in the area  to  come together.   The  objectives of so doing were to
define the current  state of knowledge concerning the  role of pricing policies
in  solid waste management and to  identify research needs for the coming
years.

        To host the workshop, EPA contracted with the  JACA Corporation which
subcontracted  with MATHTECH to organize the workshop  sessions and to prepare
the  workshop  report.   The workshop was organized  into  five sessions at which
prepared papers  were delivered and discussed.  The agenda for the workshop is
contained in an  appendix to this volume.

        The workshop tasks, as noted above, were i)   to identify the current
state  of knowledge concerning the role of pricing in solid waste management,
and ii)  to identify research needs in this area.  In an effort to accomplish
this,  these tasks were narrowed in two important ways.  First, the workshop
considered only  non-hazardous wastes.  Economists  generally are in agreement
that  there is little or no role for pricing policies in management of wastes
which, if mismanaged, can result in catastrophy.

        Second, the  workshop participants organized  their papers  and
discussions around  two  of the most widely-discussed and intensely-researched
classes of pricing policies, "user  charges" and "product charges".  The
difference between  these two approaches lies in the point at which the  price
is levied.  Under the product charge,  the price is levied directly upon

-------
 producers of goods which, when consumed, generate solid wastes.   Under the
 user charge, the  price is levied directly upon consumers as they dispose
 their wastes.

        It  was clear  from the outset  of  the workshop that  Albrecht and
 Goddard were right about the need for  researchers to meet.   The workshop
 discussions were  pointed — at some times  heated — and researchers clearly
 needed the opportunity  afforded by this workshop  to exchange ideas and  argue
 out  points of difference.   Indeed, several  of  the discussions begun at the
 workshop have continued afterward.

        This report contains the papers and  a  summary of the discussions that
 took place at the  workshop.  Even a casual examination of its contents  will
 show that a great  deal  of progress has been  made  in understanding the effects
 of pricing policies on  the cost and effectiveness of solid waste management.
 In particular, the  papers and discussions  in the pages to follow contain
 evidence that pricing policies can improve the cost and  effectiveness of
 solid  waste management, although the gains  over  present policies that do not
 incorporate pricing approaches seem  to  be  relatively small given today's
 economics.  Tomorrow, however, the story may be quite different.  If landfill
 costs  and the prices  of virgin materials exhibit the  marked relative
 increases which many  observers forecast, pricing policies could result in a
 substantial  improvement in the efficiency of solid waste management.

        The following  pages will  also show that our knowledge  is relatively
 weak on what might be called the "political  economy" of pricing  approaches to
 solid  waste management.  Workshop participants  were unanimous  in their call
 for expanded effort in  the  study of  administrative, political, and  legal
 factors bearing upon  the workability of pricing approaches.

 SUMMARY OF PAPERS

 Charge to the Workshop

        The charge to  the  workshop  was delivered in an opening address by
 Haynes Goddard.    Pointing  out that economic approaches  to  solid waste
 management  have  failed  for the most part to grasp the attention of those
 responsible  for making  and  implementing policy, Goddard laid  much of the
 blame  squarely on economists.  Our faults,  Goddard noted, are an excessive
 preoccupation with theory  as  opposed to practice,  and a  failure to  deal
 effectively  and convincingly with difficult  (but  solvable) empirical  problems
 in our applied work.

        As a point of  departure for correcting  these  faults, Goddard
 recommended that  the  economics profession take stock of current and needed
knowledge.  To this end, Goddard asked workshop participants to  consider four
key questions:

-------
         1.  Do incentive  systems designed  to alter waste generation
            behavior actually improve the economic  efficiency of
            solid waste management  and/or  the equity of  its
            financing?

         2.  What  is the optimal combination  of incentive mechanisms?
            For example, would the best  system include some mix of a
            user  charge and a product charge?

         3.  How do incentives relate to the choice  of resource
            recovery  technol ogy?

        A.  What  are the best  ways to  pursue the testing  of the
            effects of incentive mechanisms?   Should we use
            historical data?  Should we try  to conduct  experiments?

        As will become clear below, the papers presented at the workshop  and
 the  discussions that surrounded them all related  to one or more of these
 questions.  Accordingly, I have chosen to  organize my summary of the  papers
 presented and  workshop discussions by the  questions to  which they pertain.

 Pricing and the Efficiency of Solid Waste  Management

         Four of the  papers presented pertained  directly to  the  question
 raised by Goddard  concerning the effects of  incentives  of the  efficiency of
 solid  waste management.  These papers included those by William Lanen, Steve
 Buchanan, W. David  Conn, and Barbara Stevens.

         In a paper on  the  effect of user charges on the efficiency of solid
 waste management, Lanen presented the results of case  studies of user charges
 in five  municipalities.  In general, Lanen's results  showed no statistically
 significant effect  of user charges on the quantity of waste  presented  for
 collection, the  usage of  special  services, or the degree of littering.   He
 did  not  examine  the effect  of  user charges on  costs  of collection  or
 administration of solid solid waste  systems.

        Lanen cautioned that his failure to  find significant  relationships
 between  user  charges and  household solid waste behavior does not mean that
 there  is no relationship.   Rather, he  concluded that, due to data  and
 specificaton problems, the historical  case study approach is not particularly
 well-suited to measuring any relationships which may exist. His results thus
leave  open the question concerning  the effect of user  charges on solid waste
management efficiency.

        Buchanan's paper  focused on efficiency  aspects of product charge
 systems.   Analyzing data compiled in studies carried  out for the Resource
 Conservation  Committee, Buchanan determined that the effect of a product
charge on national  solid waste  flows is likely  to be rather small, with
estimated  resource savings of between $25 million and $60 million per year.
Buchanan  points out, however,  that  the magnitude  of  net savings actually
realized may be  very sensitive to  picking the  correct product charge rate.
At rates  other than the correct  rate, resource savings may  fall  off very

-------
 rapidly,  and  even become negative.   When  the additional  costs  of
 administering  a  product charge are  considered, Buchanan concludes  that we
 don't know whether or not a  product charge would  contribute to more  efficient
 solid  waste management.  He is quick  to add,  however,  that  the relevant
 question is not how the  product charge approach compares with the theoretical
 ideal charge, but rather how it compares  with what we are doing now, and with
 other practical alternative  approaches to solid waste management.

        Conn's paper, like  Lanen's,  is based upon case  studies.  Conn
 examined the experience  of a number of European countries and municipalities
 with various  forms  of  user and/or product charge.  He  found that it is
 extremely difficult to get the kind of hard data that is  needed  to  make a
 fair appraisal of the  workings of the  various incentive systems  that have
 been tr.ied.  As a result,  Conn's data consist almost entirely of  anecdotes
 concerning the transition between systems that do not include incentives and
 systems that do.

        In general,  Conn found  that  user-charge systems, once introduced,
 have not taken hold.   Citing increases  in littering and/or  difficulty of
 administering  such systems,  communities that have had  them under
 consideration have not adopted them, and  communities that had instituted them
 have disconitinued them.

        European experience with the product charge, according  to  Conn, has
 been somewhat more hopeful.  Conn notes that a container charge in  Norway has
 been effective in preventing  any growth in the  use  of non-refill abl e
 containers for beer and  soft drinks.  In Sweden there has been  a  transition
 from non-refill abl es  to  refill abl es; somewhat mysteriously, use  of  cans has
 been on the increase.

        Virtually all  of the papers  summarized above  raised questions
 concerning the effect  of incentive approaches on the costs of administering
 solid waste management systems.  Barbara  Stevens'  paper presents  two  case
 studies and the results  of a statistical  analysis of the administrative costs
 associated  with user-charge systems.  In  her case studies, Stevens  shows that
 administrative costs depend heavily upon  the "skill" with which the  system is
 administered.  Indeed, cost differences due to differences in skill  probably
 are far more significant than are differences due to use or non-use  of a user
 charge  approach.

        Stevens finds that — on the average — administrative costs for user
charge  systems probably  range between 4 percent and 8 percent  of collection
 costs.  Because  she  has  no firm estimates of the responsiveness to
user-charges of quantities of waste disposed,  she is unable to make  an exact
 estimate of the net savings that result  from switch from tax-based financng
 to a flat-fee or quantity-based fee system.  Ranges of estimates  presented by
 her  suggest however that quantity reductions and/or savings of costs  through
change  in the collections  system would have to be  substantial  in  order  to
offset  increased administrative costs of  a fee  system.

-------
 Pricing and the Equity of Solid Waste Financing

        Efficiency  effects are not  the  sole grounds upon which economists
 have rested their case  for  economic incentives.   Arguing that polluters
 should pay the costs that their wastes impose upon society, economists have
 also stressed that pricing approaches to financing waste management are more
 equitable than are other  approaches.  The workshop heard two papers on the
 question of equity and public finance aspects of  pricing alternatives  for
 solid waste management.

        The first paper, by Roger Bolton, discussed considerations relevant
 to  study of the distributional  consequences  of financing solid waste
 management via user charges  in contrast to finance via  property taxes.
 Bolton argues that a complete analysis requires consideration both of effects
 of charges on household behavior and  on the costs of solid waste management
 --  the same evidence required  to understand  the effects of pricing
 alternatives on solid waste management efficiency.

        Noting  that a  complete analysis is  beyond reach given the current
 state  of  empirical  knowledge,  Bolton nonetheless was  able to make  some
 approximate calculations of the relationship between income level and burden
 of solid waste finance under a user charges policy, and to compare these with
 the burden that would result if solid waste management were financed through
 the property tax.  His finding  is that the distribution of the burden by
 income level is about  the  same under both alternatives. He concludes that
 there is "no reason to fear  any  noticeable increase  in  regressivity from
 substituting user charges  for  the  property tax, particularly if it is the
 residential property tax which is being replaced."

        Kenneth Wertz,  in  the  second paper on equity and government finance
 heard by the workshop, compared  the  effects of user charges with those of
 product charges.  Wertz argues that local governments and  their citizenry are
 likely to prefer  the product  charge approach  to approaches based on
 locally-levied user  charges.  While  there are,  Wertz  notes, several
 advantages that  accrue uniquely  to locally-levied user charges (e.g., placing
 the burden of finance directly on  the  waste generator, and development of
 data that  may improve the management  of solid waste collection and disposal),
 he also notes that  product charges avoid certain "liabilities" of a local
 pricing system.  "Under federal  pricing, administrative costs exist but  are
 essentially borne  elsewhere.   Responsibility for the  enactment of a new
 revenue program  can also be put  elsewhere.  The risk of increased littering
 that goes with local  pricing is removed by federal  pricing, and the risk of
 having  promoted a  bad  idea —  if  in fact pricing fails  to induce  much
 reduction in the  output of refuse  — is shifted  to other parties."  He
 concludes  that,  because  of these  budgetary impacts and risks, product charges
 are likely to gain  more widespread political backing  than are local  user
 charges.

 Comparison of User Charges and Product Charges

        Two papers heard by the workshop compared  user  charges and product
charges as alternative incentive  mechanisms.  One paper, by James Hudson,

-------
 argued  that product charges and user charges are  complementary, and  could be
 combined  to provide a more complete set of incentives than either taken alone
 would provide.  Hudson shows how a properly designed combination of the  two
 approaches can avoid double-pricing of wastes and can encompass a wider range
 of wastes than either alone could cover.

        The other paper,  by Allen Miedema, compared the economic efficiency
 of a product charge, user  charge, and recycling subsidy in a  simple general
 equilibrium framework.  Miedema finds that, in the illustrative framework he
 has constructed,  a product charges  policy yields the greatest  gain  in
 efficiency of  the three  pricing policies examined.  He notes, however, that
 the treatment of user charges in his model precludes recycling activity,  and
 thereby biases his results against user charges.

 Pricing and Resource Recovery Technology

        Two of  the papers  heard by  the workshop addressed aspects of the
 relationship between incentive mechanisms and resource recovery technology.
 The first of these papers,  by Tayler Bingham, presents the results of a study
 in which  a process model  was used to  predict  the quantity of secondary
 materials recovered from  the solid waste stream  under alternative prices for
 these materials.   Applying techniques  used elsewhere by  James  Griffin,
 Bingham estimates supply  elasticities that confirm earlier findings that the
 recovery of secondary material s form the  municipal  waste stream is  fairly
 insensitive to price changes.   This finding has important  implications for
 pricing  approaches.  In  particular, it implies  that charges levied  on
 disposal  of waste  (either product charges or user charges)  may not be very
 effective in encouraging resource recovery.

        The second paper  on the relationship between incentive systems and
 resource recovery,  by Robert C.  Anderson,  presents some new estimates of
 demand  and  supply  elasticities in the  markets for ferrous  scrap  and waste
 paper.   Anderson,  in contrast to  Bingham's approach, develops  his elasticity
 estimates  from econometric models of  the wastepaper and  ferrous  scrap
 markets.  While  neither Anderson's or Bingham's  paper contains all  of  the
 information required to  make a precise comparison,  it is  interesting that
 Anderson's estimate  of ferrous scrap supply elasticity is almost two  times
 Bingham's estimate.  Anderson  notes, moreover,  that  his estimate  should be
 adjusted upward  by  a factor of approximately two  to obtain an  estimate of the
 supply  elasticity  for ferrous  scrap other than  prompt.  Adjusted in this
manner,  Anderson's elasticity is about  3.5 times  higher than  Bingham's.
 Other things  being equal,  Anderson's results  thus  imply a  substantially
greater  shift  to ferrous scrap in the face  of disincentives  to  use virgin
materials.

        A  near infinitude of explanations could  be  offered for this
difference.   Unfortunately, little has yet  been  done  in reconciling process
modeling  approaches with  econometric approaches, as the  discussion of
 Bingham's paper makes clear. A tentative  explanation may be  that  Bingham's
process model  embodies more constrained production possibilities  than in fact
obtain, with  the result  that supply is  estimated  to be less responsive to
price than it  really is.   Or, of  course, it may be that supply really is  as

-------
 obtain, with the result that  supply  is  estimated to be less  responsive to
 price than it  really is.  Or, of course, it may be that supply really is as
 unresponsive as Bingham estimates, the econometric results being spurious.

 Pricing and Sol id Waste Management

        The  papers  discussed  above  for  the most part dealt with the
 relationship between  pricing  policies and the efficiency and equity of solid
 waste management.  Roughly  speaking, each of  the  authors of these papers took
 as given aspects of solid waste management other than pricing policy.  The
 fourth  question  raised  by  Goddard asks us to consider what happens in
 real-world policy-making and  service delivery situations where all  aspects of
 solid waste management may  be subject to change when pricing is instituted.
 Further, it asks us to consider what criteria are relevant  to evaluating
 alternatives in such  an environment.

        The workshop  heard  one paper on this  subject by Bernard Booms.   Booms
 described experience to date in Seattle, Washington, as that city has moved
 to experiment with user charges in combination with various other solid waste
 management alternatives.   He noted that the Seattle experience  provides a
 nearly unique opportunity for studying the implementation process.  In
 Seattle,  the entire solid  waste management system has been changed to
 accommodate pricing.  It  is possible,  therefore, to  observe both the
 political  and  administrative processes  that come into play as system
 management is changed. Importantly, Booms notes that the Seattle experiment
 does  provide for  monitoring and evaluation of effects on a wide  variety of
 dimensions of solid waste management.

 Research Directions

        Two papers were delivered  on  research.  One, by Oscar ALbrecht,
 outlined  the research activities of  the U.  S.  Environmental  Protection
 Agency.  Albrecht noted  that  although EPA has funded a great deal of research
 in the economics of solid waste, much more remains  to be done.   Citing the
 immense regulatory  responsibilities of the  Agency, he expressed the concern
 tha important research needs  may be overlooked.   He called  upon workshop
 participants to  articulate  clearly research needs.

        The second  paper heard on research offered George Tolley's and  Steven
 Basting's response  to  the request put by Goddard and Albrecht.  Tolley and
 Hastings divide their discussion into  two  parts,  one of which  discussed
 needed  research on the cost of solid waste management services and the other
 of which discusses  research needs relating  to the  demand for  solid  waste
 management services.   On the cost side, the  most pressing need identified by
 Tolley and Hastings is for more refined  studies of the relationship between
 services provided and  cost.  Noting that solid  waste management services  have
 a multitude of cost-causing attributes (including, for example, frequency of
 service,  location of collection, source separation requirements, quantity,
volume, container  type), Tolley and Hastings  called for detailed  cost studies
 that  explore the  relationships between each of the attributes and service
 cost.

-------
        Tolley and Hastings note that even  less is known about  the  demand
 side than the cost side.  Here they call  for  a  two-pronged approach.  First,
 they recommend that crude demand curves be constructed from information on
 the cost of alternatives  to collection  and  disposal.  Presumably what they
 have in mind is  something  similar to the analysis of resource recovery
 presented in Bingham's paper, with construction of household  process models
 describing alternatives to municipal collection  and disposal  of solid wastes.
 Second, they recommend that pricing experiments be conducted,  controlled and
 monitored so as to produce the data needed to  infer households'  demand  for
 solid  waste collection services.  To  guide such experiments,  they present a
 comprehensive list of questions that should  be provided for in  the design of
 each.

 SUMMARY OF DISCUSSION

        The papers summarized above  provided  the starting  point  for the
 workshop's deliberations on the various questions posed by Haynes  Goddard  at
 the beginning of  our sessions.   The  full  transcript of these deliberations
 runs to almost 600 pages.  It is impossible  for  me  to do more  here than  to
 try to draw out of  the  transcript some of  the key points of agreement and
 disagreement among workshop participants.  Complete copies of the  transcripts
 are on file with the Municipal Environmental  Research  Laboratory in
 Cincinnati.

 Pricing and the  Efficiency of Solid Waste  Management

        Discussion  of this  topic  touched on  several  aspects  of  the
 relationship between pricing  and efficiency.  The first aspect of  this
 relationship to be  discussed  was that between  pricing and households'
 generation of waste and selection of waste disposal  methods.   While several
 conference participants hypothesized that  changes of price  and/or pricing
 policies of the  magnitudes currently observed  around the nation have  only a
 small  effect on household behavior, the  consensus of the workshop seemed to
 be  that this hypothesis remains to be tested conclusively.  The  results  of
 quantitative studies that  have  been done all  are acknowledged by their
 authors to be  subject  to serious  data  and  specification problems.  While
 these  studies provide no basis  for thinking  that the effect of prices in
 ranges currently observed are large, conference participants concluded  that
 until more careful  research has been conducted, judgement should be reserved.

        A second aspect of the efficiency  question addressed  by the workshop
 was the cost of administering  a pricing  policy.   Here again the workshop
 participants felt that available data  are  inadequate to reach any firm
 conclusions.  However, most  participants  seemed to conclude  tentatively that
 any extra,  costs  associated with administration of a pricing  policy are small.

        A third aspect of the  efficiency question concerns  the balancing of
efficiency gains induced by changes in behavior  associated  with pricing
 policies  with the  costs of administering  them.  Because of  the  uncertainties
noted above, the workshop was unable to reach even  a  tentative conclusion
concerning the net  effect.

-------
Pricing and  the Equity of Solid Waste  Financing

        Workshop discussions centered  about two aspects of the relationship
between  pricing  policies  for solid  waste and  the  equity of solid  waste
financing.   The  first aspect is the  public perception of equity issues.  In
this regard, workshop participants noted  that there may be a growing  public
perception that finance alternatives which distribute the burden of  financing
solid waste  management according to  service  usage  — as do user charges
and/or product charges  — are more fair  than are alternatives  which
distribute the burden in a manner unrelated to  service usage.  If so,  we may
expect  to observe more and  more interest in solid waste service pricing
policies. Many workshop participants  also noted  that if measures  to  limit
governmental  tax revenues continue to  grow in popularity, public perception
of the unfairness of using general tax funds to provide free services which
otherwise could be priced will be heightened.

        The  second aspect of  the relationship between pricing and  equity
considered  by the workshop  is the effect of moving from a property tax
financed system.  While several participants cautioned that  the evidence is
far  from complete, most  seemed to agree that the distributional affects of
switching from tax based financing to  financing via revenue raised from
pricing of solid waste service would not be great.

Optimal Combination of Incentive Mechanisms

        This question  was the  subject of  a  great deal of discussion since
natiowide product charges are being considered  actively at the  federal  level.
Many workshop  participants expressed doubts that a  product charge approach to
pricing solid  waste collection and disposal services makes much sense.   Among
the  basic difficulties with product charges  cited by those expressing this
view are the following:

        1.   Product charges could not be levied  on  many household
            wastes.  For example, product charges would not (and,
            more  importantly,  could  not reasonably) apply to yard
            wastes.

        2.   Product charges set at a nationwide level would reflect
            national average costs of  assumed methods of collection
            and disposal rather than local marginal costs of actual
            methods of collection and  disposal.

        Most workshop participants who expressed  reservations  about product
charges argued that local  user charges, which  would apply to all  household
wastes presented  for collection and  would be  set according  to local  cost
conditions, may be a  preferable pricing approach.  However, the  workshop,
including even those most skeptical about product  charges, stopped  well short
of concluding that user charges are  a superior  pricing alternative.   It was
generally agreed  that user charges also are  subject to certain difficulties,
including (for example) the following:
                                    10

-------
        1.   User  charges,  given current technology,  cannot
            differentiate among different  types of waste.  Since
            collection and disposal  costs  probably vary widely by
            type, user charges also do not reflect costs accurately.

        2.   Institutional considerations of administrative cost,
            political feasibility, market structure  and so forth may
            weigh in favor of  a product charge approach as against a
            user charge approach.

        Workshop participants  concluded  that the only way  to  resolve the
question of what  sort of incentive  mechanism  is  best  is  to  obtain more
empirical  information on the political economy of pricing  approaches.  More
information is needed on  household  responsiveness to incentives, the
relationshp of incentives to other  aspects of solid waste management, the
costs of  administration,  and the  process  of  implementing incentive
approaches.

Pricing and Solid Waste Management

        As the above  paragraphs make clear, progress has been made with
regard to  study of particular  aspects of the relationship between incentives
and particular aspect of solid waste management.   While workshop participants
were  reluctant  to reach firm conclusions about the  responsiveness  of
households'  waste generation to user charge/product charge levels, evidence
to date does support some weak positive conclusions.

        This situation  contrasts sharply  with the state of our  knowledge
concerning  the relationship between  pricing policy and the entire process of
solid  waste management.  We  do not know, for example, what the relationship
between pricing policy and efficiency of collection and  disposal  may be.
Some  observers have hypothesized that the use of prices makes costs of solid
waste management more "visible", and hence promotes efficient management of
solid  waste collection  and  disposal.   We do  not  know  whether or  not  the
effect of  prices on household  generation behavior depends  upon the nature of
collection  and disposal option (e.g.,  self-hauling, recycling centers, source
separation  programs), although we hypothesize that it does.   Still  another
interesting and important aspect about which next to nothing is known is  the
relationship between information about waste content  of  goods and services,
pricing, and solid waste management.   Workshop participants felt that  the
scope of research on pricing and solid waste management must  be expanded to
cover these as yet unexplored  areas.

Research Directions

        Implicit in the papers we  heard and the  discussion we conducted is  a
long list of specific research topics.  These topics  certainly would include
the foil owing:
                                   11

-------
        1.   Effects  on  household behavior of changing  from tax-based
            finance to pricing.

        2.   Effects  on  household  behavior of availability  and/or
            prices of methods for solid waste handling  such as home
            compaction,  garbage  disposals, self-hauling, separate
            collecting  of  yard  wastes, and source separation and
            recycling programs.

        3.   Effects of alternative combinations of system design,
            operating procedures, and waste loads on  the  cost of
            solid waste  management.

        4.   Effects  of  pricing policy on cost  of solid  waste
            management.

        5.   Effects of  information on  waste content  and
            collection/disposal cost on household behavior.

        6.   Institutional considerations in the design  and operation
            of solid  waste management systems.

        Workshop participants also  spent a  substantial amount  of  time
discussing  research methods.  Noting that the data  that have been used  in
studies conducted  to date are  deficient in imporant  respects, workshop
participants seemed to be unanimous in recommending  improved research design.
Among the  specifics raised  by workshop participants  are the following:

        1.   An effort should  be made  to improve  and  expand  the
            collection  of  data  on current operations  and costs of
            solid waste  management systems.

        2.   Preparation  should  be made  to  capitalize on  the
            experimental  opportunity presented when  a community
            decides to change its solid waste management system,  or
            the  method  or level  of pricing of solid waste management
            services.   Case  studies of these natural  experiments
            should be conducted.

        3.   Case studies  of changes  in solid waste  management in
            municipalities  abroad should be conducted.

        4.   Serious consideration should be given to conducting  fully
            designed and controlled experiments.

CONCLUSIONS

        In general,  workshop participants  felt that the  time  for an expanded
research effort on pricing  in solid waste management has  arrived.  While
participants  did  differ  in  their views about  the  seriousness of
inefficiencies caused by failure to price  solid  waste management  services
correctly,  there was a consensus that pricing policy will become vitally

                                   12

-------
important in the  future.  This will come about  for  several reasons, including
increasing cost of landfill, and growing pressures  to reduce tax burdens.   If
we begin now to design  and  execute a  plan  of  research  designed to answer
conclusively the  practical questions raised  elsewhere in  this volume, we will
be ready to help communities manage  their  solid waste  services  more
efficiently.
                                    13

-------
   CAN ECONOMIC INCENTIVES HELP RESOLVE SOLID  WASTE MANAGEMENT PROBLEMS?



                                   by


                            HAYNES C. GODDARD
        The  purpose of this symposium  is to bring together  the principal  and
most active researchers in the country on the question of  the proper role of
economic incentives in a comprehensive approach to resolving  municipal  solid
waste management (SWM) problems.   Outside of  the investigations  of  the
comparative handful of researchers  assembled  here today, this  area  has
received  only a small fraction of the attention  that has been devoted to
economic  incentives for  the management  of air  and water  residuals.
Furthermore, except for the product  charge concept  (a manufacturer's excise
tax on packaging materials), the question of economic incentives  for SWM  has
tended to  be accorded the same degree  of hospitality  by governmental research
and program  administrators as for other economic incentives:  a  large yawn.

        None the less,  with the gradually increasing attention paid to the
question by  members of this group, our understanding of the potential  for
incentives  in this area  is  growing, and has  led  the organizers of  this
symoposium to conclude that that the time is propitious for a concentrated
review  of the major issues.   By bringing together for the time those with
research  interests in the  area, we  hope to enlarge our  collective  and
individual perspectives and understanding of the area.

        We are  assembled here with the common belief  that it  is  worth testing
the hypothesis that economic incentive systems in  SWM will raise  economic,
engineering, and administrative efficiency and perhaps also equity by making
the  costs  of the various  characteristics of  SWM explicit  to the  waste
generators.   Such characteristics are weight, volume, packaging  content,
location of  pick-up, frequency, and  any other locally unique  characteristics.
We hope that major outcomes of this  symposium will be a specification of  the
relevant  research and policy questions that need to be investigated, and a
set of relatively specific  recommendations  indicating how this  research
should be  conducted.

        Most of you are aware that,  in general,  the official  reception of
economic  approaches to residuals management has never been a very hospitable
                                    14

-------
one and at times has  been downright hostile.  Our colleagues  in  engineering
and  law have been much more successful  in convincing policy makers of the
applicability and validity of their proposed solutions to  the  residuals
management problems  and  while, of course, physical and regulatory measures
are necessary conditions for efficient and  effective pollution control,  they
are  not sufficient.   Undoubtedly part of the problem we  confront can be
explained by the low level of economic  literacy that characterizes the
population in general,  but  this notwithstanding, the major share  of the blame
for  the lack of success  which economists have  suffered  in obtaining
receptivity for  their  ideas must lie on the economists themselves.

       Two of the internal  impediments that some have observed  to diminish
the receptivity  of proposed economic solutions  are:

       •        a rather exclusive interest on  the part of academic
                researchers in the theoretical aspects of pollution
                control, with a corresponding  relative neglect of
                the empirical side of the problem (of which I myself
                am guilty);

       •        a tendency  to build models with  substantial
                specification error, either because of mathematical
                convenience  and/or because  data  on the correct
                dimension to be tested is unavailable.

        A major external  impediment on the research side is  that our
colleagues in engineering typically are given a limited exposure  to economic
analysis, and thus conclude that they can  perform what economic  analysis is
needed to resolve pollution problems.

        If we are to increase our effectiveness in making contributions to
the resolution of pollution problems, we must find ways to remove these and
other  impediments, and we must become more effective spokesmen for economic
understanding, but, of course, we must have an  understanding to convey—that
is why we are here.

        Research on economic approaches to  SWM  does not have a long history,
dating back only 4-5  years, but nonetheless, this research has been pursued
in several directions.  These directions have included user charges for  SWM,
manufacturing excise  taxes (product  charges), litter control  incentives,
relative efficiencies  of alternative institutional arrangements,  the impacts
of federal tax  policies on waste generation and materials recovery, and the
incremental costs  of collection and  disposal among  others.    Our
understanding in each  of the areas is growing,  but is still insufficient to
make reasonably  complete evaluations of the impacts of alternative  incentive
mechanisms.

        The speakers  today and tomorrow will be touching on a large  number
of current  research and policy questions, and I will not  attempt to
enumerate them here.   I do want to mention,  however, some of the  major
questions that  have  not  been answered,  and  hope that we will  remember to
reference them frequently throughout our proceedings.  These questions are:

                                    15

-------
        •      Do  incentive systems  that  are designed  to  alter
               waste generation behavior in fact raise the economic
               efficiency of SWM,  that  is,  produce  large  net
               benefits?

        •      What  is  the optimal combination of  alternative
               incentive mechanisms?   How should the product  charge
               and the user charge  be  related to one another "in the
               most  efficient configuration of complementary and
               substitute incentive systems?

        •      How do these incentive mechanisms  relate to the
               choice of resource recovery technologies?

        •      What  is  the optimal combination  of  all  SWM
               alternatives, and  what  criteria should  be  employed
               to make the choice?

        •      What is the best  way to pursue the  testing of the
               effects of the incentive mechanisms,  with historical
               or experimental  data?

        I hope that this  symposium will help  us much in our search for
answers to  these and other related questions.  Hopefully,  we can also  find
some ways  to  remove impediments  to the employment of economic knowledge in
the quest for  efficient and effective environmental management.
                                    16

-------
                                  SECTION 2

            PRICING AND THE EFFICIENCY OF SOLID WASTE MANAGEMENT


"The Economic Efficiency of User Fees:  Some Preliminary Empirical Results"
(William N. Lanen).

"Evaluating the Efficiency of the Solid Waste Charge" (Steve Buchanan).

"Economic Approaches to Solid Waste Management:  European Experience"
(W. David Conn).

"The Administrative Costs of User Charges" (Barbara J. Stevens).
                                     17

-------
                    THE  ECONOMIC EFFICIENCY OF USER FEES:
                     SOME PRELIMINARY EMPIRICAL RESULTS
                                    by


                             William N. Lanen
          One of the  most  fundamental propositions  of  economic theory is that
 the  demand for any  good  or service  declines as the price of that good or
 service Increases.  This is, undoubtedly, one of the major reasons economists
 traditionally look to pricing solutions rather than regulatory solutions for
 the control of environmental pollutants.  Today, economic solutions involving
 charges are being employed to deal with air,  water, noise, and land pollution
 problems.*

        Individuals charged with the management of  systems (for example solid
 waste management systems) have, on the other  hand,  tended to distrust pricing
 approaches.  The  reasons for this distrust include notions of equity (for
 example,  that charges  for  solid waste collection  are  regressive)  or
 practicality (e.g., that user charges are administratively burdensome).2   It
 then becomes important to determine whether user charge systems, especially
 incremental charge  systems, lead to improvements in economic efficiency.

        The approach  an economist would take  to make that determination  has
 been given the name "cost-benefit analysis."  Cost-benefit analysis is
 nothing more than  a  set of  procedures  for  performing an  analysis of  the
 merits   of alternative approaches to problem solution.  As such, it is done,
 perhaps only implicitly,  by  decision-making units at all  levels.    The
 difference between the  results of  the analysis (i.e.,  the reason  why
 different groups come up with different  "optimal" solutions)  is that  each
 group accrues different benefits and incurs different costs.  As we discuss
 below,  the conflict between system managers and economists about the optimal
 solution to the financing of solid waste management may often be a result of
 this difference in  the organizational unit considered.
1.    See,  for  example, the discussion in  Anderson, et. al.

2.    American  Public Works Association,  Solid Waste Collection Practice, 4th
     Edition,  Chicago, 1975.


                                    18

-------
        The purpose of this paper Is to  present some preliminary results of  a
 study conducted by MATHTECH, Inc. for the Municipal  Environmental Research
 Laboratories of  the U.S.  Environmental Protection Agency.  A case study
 approach was employed using five cities, each with a different structure for
 a  user  fee. (The term  "structure" will be defined below.)  The study was
 primarily empirical in nature and its purpose was to evaluate the impact of
 user fees residential solid waste behavior.

        In the first section we discuss  the concept  of economic efficiency
 and  its role in evaluating user fee systems.  We then describe the different
 types of fee structures.  Finally, we evaluate the qualitative effects which
 we would expect each of the fee structures to possess.

        In the  second section a selective review of the  literature is
 presented.   It  is  selective in that it discusses only work which has
 attempted to estimate price or income elasticities.  As we  will show, while
 there is some statistical evidence of  positive income elasticities of waste
 generation,  no study to  date  has provided convincing  evidence of  a
 statistically  significant non-zero price elasticity.  Non-zero price
 elasticities, while they cannot be used  to measure economic efficiency, are  a
 necessary condition for the employment of user fees to lead to an improvement
 in efficiency.

        In the  third section the results of our study are presented.^  Both
 the models and  the statistical results are discussed.  We will see that there
 is some additional evidence of positive income  elasticities.  While some
 evidence points to the existence of statistically significant price
 elasticities, it is not strong.

        Finally,  in the last section we  present our conclusions with  respect
 to the  effects of a user  charge system on residential solid waste behavior
 and  the consequent implications for their role  in increasing economic
 efficiency.

 USER FEE SYSTEMS AND ECONOMIC EFFICIENCY

        To an economist,  a system that is economically efficient is one in
which net social benefits are maximized.  Net social benefits are, of course,
 nothing more than the difference  between social benefits and social costs.
Under traditional  cost-benefit analysis, the measure  of social benefits is
 "consumers'  surplus."  In  other words,  it is  the  difference between what
consumers would  be willing-to-pay for a  particular  good or service (e.g.,
 carryout service)  and what  they are  required to  pay (if they take the
 service).   Social costs  are defined in  terms of opportunity costs.  Making
3.   At this time,  all of the  results of the study must  be considered
     tentative.

4.   The terms "benefits" and "costs" can, in practice, be misleading since,
     for example,  depending upon the baseline taken, benefits can be, and
     often  are, reductions in costs.

                                    19

-------
use of  these two concepts,  we  next  show how they might be  used to evaluate
the economic efficiency of a user fee  system.

An Illustration

        Consider the question of carryout  service.   If refuse collection
services are provided by the city,  it  has three basic alternatives:   provide
the service  to all residents;  provide  the service to no residents; or, charge
a fee for the  service and  let the  individual  resident choose.   The
comparative benefits can be assessed using  the  information presented in
Figure 1.
                Price
                                                   AC
                Figure 1.  Evaluating economic efficiency
        In Figure 1, the  demand for carryout  service is  drawn as a linear
function of price only for simplicity.  It is assumed that at  zero price, all
residents will demand the  service.  While this  may not  be true (if, for
example,  some residents feel there  are additional costs to having collectors
enter their  yards) nothing  in  the analysis is altered by this assumption.
Suppose  that, if the service is offered for a fee, the fee   (p)  will  be set
to equal the average cost (AC) of  providing the service. (Average costs are
drawn as constant In Figure 1, implying that average costs equal marginal
costs.  We have excluded from the figure any administrative  costs which might
be associated with the use of a fee.)

        Consider now the net social benefits  associated with the three
alternatives  outlined above.  The first alternative, providing the service to
everyone without an explicit change,  is equivalent to charging a zero price.
                                    20

-------
 Therefore, the demand  Is  100%.  The total willingness-to-pay (i.e., the  area
 under the  demand curve)  is equal  to the area OAD.   The  total cost  of
 providing the service  is  N x AC  where  N  is the  number  of residents.
 (There  are  no administrative costs since the service is provided to everyone
 and no monitoring is required.)  Therefore, the net benefits  to society are
 equal to the total benefits less net costs and depend on the relative size  of
 the two shaded areas in Figure 1..

         Now consider  the net social benefits of  not providing the service  at
 all.  In this case the net benefits are zero -- no costs and no benefits.5

         Finally consider charging a price  p  (assumed  to be equal to average
 cost).  The  percentage choosing the service will  be  W.  The  total benefits
 (i.e., w.illingness-to-pay) are  OABW.   The total  costs  are  AC x (W)(N).  The
 net social benefits are  the  vertically shaded area  pAB.   Note that the
 benefits  associated with this arrangement are greater than those associated
 with providing the service to everyone by the amount  WBD  (the horizontally
 shaded area)  neglecting the administrative costs.

         Similarly  the net benefits associated with the  charge system are
 greater than providing no service (again neglecting administrative costs)  by
 the amount  pAB.   The question of the  best choice depends  then on the level
 of  administrative  costs.  If they are  less than pAB, a charge  system is,  on
 net,  beneficial when compared to  providing no  service.  If they are less  than
 the area   BCD,  then the charge system  has greater net benefits  than the  free
 provision.  Of course, the  level  of the  administrative cost and the relative
 magnitude  of  the shaded area are  empirical questions.

        Suppose, however, that in the process of trying to evaluate  the net
 social benefits associated with a  particular  fee structure, that the analyst
 is  missing some of the  crucial information; e.g., the actual demand curve for
 the service.   Is there  some  necessary  condition that must hold  for a user fee
 to  improve economic efficiency (not equity)?   The answer is yes.  Basically,
 it  must be true that the  demand curve has a negative (i.e., non-zero)  slope.
 Consider what happens if the  slope  (or elasticity)  is zero in the relevant
 range.5  If  the  elasticity  is  zero, the waste presented for collection
 remains the same  and,  therefore,  total  costs remain, the same.   All that
 happens is that there is a transfer payment from residents  to the solid waste
 system, and under traditional  cost-benefit analysis  such a transfer has no
 effect on benefits.
5.   Again,  this  result  is dependent on the definition  of  the baseline.

6.   Obviously,  at  some price,  demand for  the service  will fall.   In
     discussions,  such as this, with policy implications,  it is of interest
     to determine whether  there are  significant price effects within  the
     range of  user fees normally encountered in  practice.
                                    21

-------
A Classification of Fee Structures

        As discussed  above, the approach  followed in this study has been a
case study approach of five cities,  each of  which has a different type of fee
structure.   By "different" we mean  merely that the resident in each  of  these
cities is faced with a different set of choices  concerning  the  disposal of
the solid waste generated  within the  household.  Before discussing the
findings in each of the five cities, however,  it is useful  first to  discuss
the classification systems used to develop the five different fee structures.
Once this has been done, we can analyze what the qualitative effect of a fee
increase would be.

        The nature of the fee systems used by the  roughly  1400  communities
identified in  the  study  that employ user  fees for the  collection of
residential solid waste vary greatly.  However, within that variation,  there
is one  characteristic which is   (a)  easy to  identify,  (b) useful in the
development of hypotheses  concerning  behavior,  and   (c) restricted  in
relevant possibilities.  This characteristic is, therefore, an obvious one to
use for the classification of user fee systems for  the purposes of selecting
the case study cities.   This characteristic is the set of decision choices
the resident has for the  disposal  of  solid waste. Although there may be
others,  we have selected five  different possibilities as including
(hopefully)  most of the fee structures currently used.  Below we identify and
define each type.

        The first, simplest,  and by  far the  most widely  used  fee system is
the uniform or flat fee  structure.  In this system, the resident (assuming
service is mandatory) has no choice  over his disposal options.7  The  goal of
this  type of fee structure is basically  one of  revenue raising.  In  other
words, the flat fee service is an alternative to  general  tax revenues for
supporting the solid waste system.

        A more complex system is one in which the resident has the option to
choose  the  number of containers (i.e., the capacity) for waste for pickup.
While  it is  not necessary to assume  that the resident must remain within that
constraint  once and for  all,  it is assumed that  there are some transaction
costs  associated with frequent changes from  one level of capacity to  another
that  are sufficiently large to discourage such  changes.  We will  refer to
this as a capacity-based structure.

        A third type of fee structure is one in which the resident can choose
alternative levels of service.  Service levels can  be defined in terms of the
point  of collection (curbside or backyard)  or frequency of collection (e.g.,
weekly or semi-weekly).  Most cities that have a service-based system provide
a choice in terms of location.  In this system, the number of containers that
can be  presented is specified (although  it may be unlimited).  This is
referred to as a service-based structure.
7.   Certainly,  the  household can choose illegal modes of disposal.
                                    22

-------
        Metered-bag systems are those  systems where the resident presents all
waste  to  be collected in  a  specially-marked  bag.   This system is  quite
similar to the  capacity-based system discussed earlier.  The difference,  and
this is an important one, is that  the resident can alter, at no cost (other
than the cost of the bag)  the number of containers  presented each  time.
While this system is not a true quantity-based  system,  where the resident
pays for collection according to the weight or volume of  the waste presented,
it is a step closer to that  ideal because of the  freedom offered at  each
collection period.

        Finally, some municipalities employ fee structures  that are
combinations of two or more of the  other systems.  For example, in Tacoma,
Washington, the resident chooses both  the number of containers and the  level
of  service (in terms of  pickup location).   This  type of system may  be
expected to be  more expensive ye.t more flexible than a system offering  only
one choice.  This will be referred to  as a combination-based structure.

Analyzing  the Effect of Different Fee  Structures

        In theory, one  could qualitatively analyze the impact upon resident
behavior  from the change  in the level of the fee for  any type  of fee
structure by first developing a utility maximization model with each of the
fee features.  Then, comparative statics could  be used to determine the
direction of change in the amount  of waste generated.  Such  an approach was
used, for  example, by Kenneth Wertz  (1976) when analyzing the  effect of price
and  service level upon waste generation behavior.  The problem with such an
approach is that, of necessity, most of the fee  systems that  exist  include in
their  rate structures such large discontinuities (e.g., the number of
containers) that applying such methods may not be particularly fruitful when
examining actual systems.   The results of the Wertz analysis are helpful,
however,  in that they provide a check on a more simple approach  to the
qualitative determination of the effects of a fee change.

        Consider, for example, an "ideal" fee structure where the resident
pays by the pound (or cubic foot, or  both) for  the amount of  waste disposed.
Further, suppose that 'he is allowed to choose the point of collection  (for a
fee).  A system of demand equations  that might be  expected to model  such a
decision process adequately would be like the two equations:
        3 - 8Cps, y)                                                   (2)


where  q  represents the quantity of waste  disposed,  pw  the price  per unit
of  waste,   y  is disposable income (after  payment of any flat fee for
disposal),  s  is a measure of service level  (e.g., the number of feet from
the curb), and  ps  is the unit price of  the  service level.  Noted above each
                                    23

-------
of the independent variables  in both equations  are the effects  we would
expect  changes to have.   For example,  following elementary theory we would
expect that  an  increase in  the price of either  quantity or  service would
decrease  the demand for  the  respective good, and similarly for income.  In
equation (1), we show that the effect of  an  Increase in service  level leads
to an increase  in  the amount of waste generated.  This is consistent with the
findings of  Wertz.  Intuitively, its justification is quite straightforward.
The disposal of residential solid waste requires the labor of two  different
units:  the  collector and the household.  Use  of  household  labor is  costly
just  as  is  the use of the  collectors'  labor.  Increasing the service level
while holding  all other variables constant, means that  the  cost  to the
household of presenting an additional unit of waste has fallen.  We would
therefore  expect to observe an increase  in the amount of waste presented.

        We can now use equations  (1) and  (2)  to determine the effect of a
change in  the price of either quantity or service level.   The  effect of a
change in  the price of quantity is known  directly from the demand curve.  The
effect of  a  change in the price of service  level  upon the  amount  of waste
generated  depends on the indirect effect  of  the service price on the level of
service.   Thus,


       9q/9p   =  9q/9s  .  9s/9p < 0                                 (3)
             8                   S

The effect of an increase  in  either price therefore is expected to be a
decrease in  the amount of waste presented for collection.  These results were
for the ideal system.  We  now evaluate the  five types of fee structures
described  above.

        In the  case of the uniform fee,  the  only equation is:


        q  =  f(y)                                                       <4>


Note that  the only variable influencing  the amount of waste presented for
collection is the  disposable income.  Suppose we define,


        y =  m - p                                                      (5)


where  m  is household income before the fee and   p   is the  flat fee.   Then,


        9q/3p  - dq/dy  •  9y/9p - - 9q/9m                              (6)


That  is,  the effect of  an increase in the flat fee is  equivalent to  the
effect of an equal decrease in  income.  In  terms of elasticities, however,
                                    24

-------
          "n = 3q/3p   .   p/q =  - 3q/3m   .    p/q » - 1m •   P/m       (7)


         This means that if the income elasticity of waste  generation is, say,
 on the  order  of .30, then  the elasticity  with respect to a change in the
 level  of the flat fee is insignificant.

         The capacity-based system differs from the ideal because the resident
 does not face an explicit charge each collection.  Rather, he  faces what  is
 essentially a capacity constraint.  Therefore,  his choice  can be modeled as:

            .,+  +.
         q - f(c, y)                                                    (8)

                  +
         c - 8Cpc, y)                                                   (9)


 Again, the signs above "the independent variables represent  our assumptions
 about  the effects  of changes in these  variables on the  amount of  waste
 generated.  In these equations, C represents  the number  of cans selected.
 Ignoring the discontinuity  of the number of  cans, we can analyze the effect
 of an  increase in the price per can on the amount of waste generated.   It is:


        3q/3pc -  3q/3c   .    3c/3pc  <  o                              (10)
The product of the two terms  in (10)  is  negative by virtue  of  the effect  of
price on the number of containers  selected.  However, if the first factor  is
less  than one, the resident can moderate any inhibiting effect  of the price
change.   This factor  3q/3c   measures the density change resulting from a
change In the number of containers.   That is, If it is less than bne, part  of
the effect of the change in price is  to  have residents packing each container
slightly more full.  This would be  a  perfectly rational thing to do since,  as
the price of an additional can  increases, the relative cost of the resident's
own labor in the preparation  of his waste for collection falls.

        This suggests that one  of  the "weaknesses"  of the capacity-based
system is that it provides a means for  the resident to avoid the full effect
of the price increase.   Of course, the quantitative importance  of this effect
is an empirical question.

        The model for a service-based fee would be
            .,     .
            f(s,  y)                                                    (11)


            g(? ,  y)                                                   (12)
                                     25

-------
 Here we are Interested not  so much in the effect of  changes in the level of
 the  fee on the choice of service level by the resident, which we know must be
 negative, but  the  effect  on the quantities of waste  generated.  But this is
 just,


        dq/dp   - dq/da   .   ds/dpa  <   0                          (13)
             S                     B


 Therefore, we see that  the effect is  dependent upon two  quantities;  the
 effect of price on  service  and the effect of service on quantity.

        The bag  system  differs from  the capacity-based  system  in two
 important respects.   First,  the resident chooses  the  number of containers to
 use  each  collection period.  The second  is the fact  that  each bag used
 includes  the cost of the  bag.  With  a  container,  the  container was  not
 "consumed" when filled.  With  the bag it is.   Therefore, the system of demand
 equations for this  system can be specified as:


        q = fCp , B,  y)                                               (14)
              B

        B - gCp , y)                                                  (15)
              B

where  pB  is  the price  of a bag.   Note  that  the price of bags  affects  the
 demand  for two separate goods,  the collection of waste and the number of
bags.  Therefore, we  would expect  the  effect  of  a  price change to be somewhat
more complicated.  In fact,  it is,


        dq/dp    =  dq/dp   +  dq/d'B 9B/3p   <   0                      (16)
            B         B                B


        The combination systems combine the effects of two or more of  the
types of fee structures.   Assuming  that  the combination is for the number of
containers  and the location  for pickup,  the model  for analyzing the effect of
fee changes would be:


        q - f(c, s,  y)                                                (17)


        c " 8(? , y, s)                                               (18)
              C

        s - M? , y, c)                                               (19)
              S
                                    26

-------
Measuring  the effect of  a change in the price  for  containers is now
complicated  by  the fact that the container fee will affect both the number  of
containers  selected and the level  of  service chosen.   The effect of a change
in the fee for  a container is
               =   dq/dc   dc/dp +  dq/ds  ds/dc dc/dp  < 0 ;      (20)
             c                   c                      c

the effect of the change in the fee for level of service is


        dq/dp  =»   dq/da   ds/dpo +  dq/dc  dd/ds da/dp  < 0 .      (21)
             S                   S                      8


We see that in  both cases  the  impact of an increase  in  either fee is to
reduce the amount of waste presented for collection,
REVIEW OF THE LITERATURE

        While there have  been innumerable studies  of  the cost of providing
solid waste  collection services (along with many prescriptions for decreasing
agency,  not necessarily  social,  costs with particular technologies), there
have been relatively few studies  that  empirically  address  the effects of
pricing  on  residential solid waste behavior.  Without such Information,
however,  the  analyses of the "optimal" type of fee  structure would be
impossible  to conduct in terms of  cost-benefit analysis.  The purpose of the
following is to provide a  brief,  and admittedly selective, review  of  the
empirical literature that has developed In this area. The  reason for focusing
on the following studies is that they form the basis  for much of the current
discussion  regarding solid waste  pricing and, because they provide numerical
estimates of elasticities and other economic data,  have come to be  relied
upon  in  policy-making discussions.  Therefore, it is important that they be
assessed  critically in order that the faith placed in the results be
commensurate with that due.

Wertz

        The Wertz article is primarily concerned  with the development of
qualitative implications for the effects of changes In  service fees or
quantity fees.   However, throughout  the paper are several brief examples
which provide some support for the theoretical propositions  developed.   The
ones  which  we  are primarily interested in are those relating to  income
elasticities, price effects,  and service level effects.

        With respect  to  income effects, Wertz used data from 10 suburbs of
Detroit which had similar financing and collection policies.  Using a  linear
functional form for the demand curve, Wertz found implied income elasticities
                                    27

-------
 of  .279 and .272 (depending  on  the actual sample used).   In a conclusion that
 does not seem  to  be  entirely  warranted, Wertz states "The foregoing mixture
 of  theory and  observation suggests  the expected:   residential  refuse
 quantities should decline  as  t  (price) increases."

        The "foregoing mixture" of theory appears to  be  the derivation of the
 usual classification of  effects into substitution and  income effects.  This
 is  perfectly straightforward following usual demand theory.   What  is not
 clear is how the observations about income elasticities  from cities employing
 no  incremental user charge can  support expectations about price elasticities.

        Following  this conclusion,  Wertz examines  some additional evidence
 bearing on the price effect.  He cites the fact that  per capita generation  in
 San Francisco,  which  employs an incremental user fee, was substantially less
 than  "for all urban areas  where general financing prevails."  As Wertz notes,
 there are too  many variables to place too much belief  in the numerical
 accuracy of the  implied price elasticity of .15.   In  addition to the lack  of
 data, however,  is  the fact that the  comparison was  between two different
 years made comparable  by applying "an average growth  rate."  In addition, the
 growth rate was applied  to a  figure which was composed of measured tons and
 estimated (from  volume)  tons.

        The estimates Wertz provides  on income elasticities appear to  be
 consistent with  the findings of others and are based  on  statistically  sound
 methods.   The  empirical  evidence of  a non-zero price elasticity is not,
 however.

 Tolley, Hastings, and  Rudzitis

        In an updated  version of an earlier study, Tolley,  et  al. , provide
 estimates of income  elasticities based on cross-sectional data from several
 wards in Chicago.  Because waste collection services  are  financed  out  of
 general  revenues in  Chicago, no estimates of price elasticities  were
 possible.

        The findings of Tolley, et al.,  were consistent with the earlier
 work.   Namely,   the estimated income elasticity in  Chicago appears to be  .3
 and  .7, depending on the season.

        Again, these findings are also consistent with those  of others  that
 there is an income  elasticity which is positive but less than one.

McFarland

        One of  the  most oft-cited studies concerning the existence  of a
significant price elasticity is the study by  the  University of California  on
solid waste practices  in that state.   Chapter IV, which formed the  economic
basis of the report, was authored by McFarland and has  come  to be known  by
that name.   We will  therefore continue to use  it.
                                    28

-------
         The approach McFarland  used to estimate the  price elasticity was to
 apply ordinary least squares  to the following equation:

                al   &2  a3
         Qd = a0Xl   X2  X3

 where  Qd  is  the  annual per capita  quantity of  waste generated,  Xi  was
 average  revenue,  X2  was  per capita income,  and X.  was population  density.
 The  results of the regression were:


         InQ  - 6.9 - .455  InX.  + .178 lnX? -  .212 InX,
                     (3.2)      (.36)    *   (1.5)


 where  t-statistics are noted in parentheses.   McFarland used this  result to
 state  "This indicates  that people will definitely respond to price  incentives
 or disincentives in their  use of the service."  There are, however,  several
 problems  with these results  -- problems  which  essentially vitiate  the
 results.

        First,  the  price proxy used was  average  revenue.   Unfortunately,
 McFarland  did not  specify  the actual 13 cities  used  in the  analysis.
 However, since only  two  out of  the 58 in the  entire study appear to employ an
 incremental user fee,  the majority had to be  flat-fee cities.   In  that  case,
 there is no price elasticity.  A second problem is that many flat-fee cities
 impose quantity limits.  This means  that  service levels are  not held
 constant.

        Finally,  there is  undoubtedly a good deal  of simultaneity.  Most
 municipal  systems  are  designed  so  that the  flat fee is, at least somewhat,
 related to  system  costs.  But system  costs are related to the level of  waste
 collection.   Therefore, by not including a  second equation in the model and
 using simultaneous methods,  the  results do  not have the usual  desirable
 properties.

        These problems  may  also have  led  to  the  finding  of  an  income
 elasticity  not significantly different from  zero,  a finding which conflicts
with other  studies.

       McFarland goes  on in the  analysis to discuss the effect of incremental
 fees  on  litter  — the primary externality to  be expected from the imposition
of a  user  fee.   To do  this, she classifies cities as internalizing or
externalizing,  depending  upon  their mode of  financing and quantity
limitations.   Externalizing cities are those  cities not "providing unlimited
or generous  service  at zero marginal  costs."  She then found a significant
difference  in  the  costs associated with the  solid waste management system
between the two types of cities.  She attributed this difference to the extra
costs of  litter cleanup in the externalizing cities.
                                   29

-------
        The results of  this  analysis have been  criticized elsewhere (see;
e.g., Goddard).  However, an important point that we have not seen mentioned
is that flat-fee cities imposing quantity limits (while perhaps by definition
incremental fee cities; i.e.,  with an infinite incremental fee) are not what
is generally meant by incremental fee cities.

        The empirical results  of McFarland, therefore, do not appear to be
sufficient evidence of the  existence of non-zero price elasticities  either
directly from  demand equations or indirectly through litter effects.

Stevens

        In an unpublished research paper, Barbara  Stevens has presented  some
results, both  theoretical and  empirical, concerning service  level pricing.
In the  theoretical analysis  she extends the results of Wertz by explicitly
including frequency of collection and location of pickup simultaneously.  She
concludes the  theoretical section with:

         All mandatory  collection  systems,  whether the  fee  is
         explicit or implicit, cause the consumer to generate less
         wastes when prices are increased (provided that either  the
         price elasticity of demand for service or  that the relation
         between refuse generation and  income is  small) and  cause
         the consumer to generate more  wastes  in  response  to a
         costless increase in  service level.  Only  the service level
         fee pricing scheme  has the joint advantage of encouraging
         consumers to value  goods  implicitly according  to the
         disposal  cost  of  their  refuse component  and  of
         implementability.

        It is important  to note her qualifier "mandatory."  She goes  on to
assert:

         When  a  pricing scheme  does  not require mandatory
         participation of households, none of  the above conclusions,
         with respect to  any of  the  pricing models, can be stated
         with any  confidence.   In  any such  non mandatory
         arrangement, increases  in the price of  refuse collection
         services can lead to a  decrease1 in the proportion  of
         households selecting organized collection services.  The
         households opting out of the system may choose to self haul
         to legal disposal site (sic), to dump in illegal locations,
         to burn  refuse,  etc.  Any or all of these alternative
         disposal technologies may result in a perceived increase in
         disposable income and a  consequent increase  in  refuse
         generation.  In addition, some of these alternate disposal
         technologies may  result  in  increased  total  costs  of
         collection and  disposal of  refuse to  be borne  by  the
         society as a whole.

The reason for such a counterproductive generation  effect, at least in  terms
of economic theory, is unclear. For if the resident could make use of  the

                                   30

-------
 alternative  facilities at lower cost than at the new user  fees, the same
 alternatives were available at the lower cost.  That is, let  p(Q)   be the
 price of disposing of an amount  Qc through "conventional"  means.  Let  L(Q)
 be the  "price" of some alternative form of disposal (e.g.,  littering).  Then
 the  resident  will choose quantities   Q  ,   Q.   to  be disposed  of
 conventionally and littered so that           c
        TC - p(Q)Q   +  L(Q)QL
                  C
 is  the total cost of waste disposal.  Total waste generated is just   Q   +
 QL.  Now suppose  there is a  price increase  in  conventional disposal.
 Regardless of the functional forms of the price functions  p(Q) and L(Q) , the
 total cost cannot fall.   (It is  true, however, that there  will be more waste
 disposed through illegal means and, therefore, social costs may increase.)

        The  difference  between  mandatory and non-mandatory systems is that,
 under  the latter,  the  resident has more  substitution  possibilities'
 available. Thus, in such systems, we would expect the inhibiting effect of
 any price increase to be more moderate than with a mandatory system.

        With respect to  the  empirical portion of the paper, Stevens provides
 a useful test of  the efficiency of  service-based fee systems.  Stevens
 estimates three  demand curves:  one  for waste presented for collection;
 demand for service frequency; and demand for service location.  Using the
 latter two, we can estimate  consumers' surplus and,  hopefully, say something
 about the effect on  economic efficiency of service-based plans.

        For example, the demand  curve for location of pickup is


        BY = 271  +  . 003Y + 1. 4QH +  . 04FRE - 3. 3P   - . 028DEN    (22)


where   BY  = percent .of households selecting backyard service
        Y   = mean annual family income
        QH  - annual tons of refuse per household
        FRE = percent selecting  higher frequency of service
        P   - price  (per month)  for carryout service
        DEN = persons per square mile


Significant explanatory  variables were income and price.

        Substituting the mean values  for each of the variables except orice
into (22),  we  get
        BY  = 268 - 3.3p
                      BY
                                    31

-------
Equation (23) can be  used  to estimate consumers'  surplus.   First note that
(23) is linear and therefore consumers'  surplus  is  just the triangle bounded
by the demand curve,  the price, and the price such  that   BY  is just zero.
That is,
        CS = BY x (p     - E^ ) x .5
                   max  T3Y
(24)
where   Pmax is fc^e  Price such that  BY  is  just zero.  From (23)  p    =
$81. Using mean values for  BY  and  p    ,  we  find
                                     max

       CS = .28 x (81  - 2) x .5 = $11.06/year
        The $11 is  per household (since the quantity  term  is percentage  of
households).  To  determine whether the offering of  such a service leads  to
an  increase in efficiency,  we need to  compare  this benefit  to the
administrative costs of providing the optional service (we assume that the
price includes the  additional collection costs).  While there is no data,  we
can infer likely effects.  Stevens' sample generated  1.71 tons per household
annually.  Assuming average  collection and disposal costs of $30,  total
annual  costs per  household would be $51.  The benefits of $11 annually per
household represent 22  percent of  this.   Edwards  and  Stevens  found
administrative costs to be between  3%  and 18% of  collection and disposal
costs.  Thus, it appears that providing the optional  service increases
economic welfare.

        Because of  the assumptions necessary to estimate consumers' surplus
in this case, we would not want to say that the Stevens' results can be used
to infer that service-based plans increase economic  efficiency.  We  would
argue, however,  that they are highly suggestive.

        Based on the theory and previous studies  relatively little can  be
said  about the magnitudes  of the price and income elasticities.  There
exists no statistically valid evidence of a quantity  effect.  The Stevens'
result suggests that price is important to the resident in selecting service
levels.

        In the  next section, we present our analyses of the effects of five
different types of  fee structures.  As with the studies reviewed here,  the
results (or lack of results) depends on the amount of data available.


RESULTS

        In this  section, we present some preliminary  empirical results  of  our
study.  The emphasis in this section will be on results about  elasticities,
both income and  price, that can be derived from each  of the case studies.
                                    32

-------
Burbank, California

        The  City of  Burbank, California  employs a flat-fee system for
residential solid waste collection.  Service  from  the City  is  mandatory for
all  residents in single  family dwellings  and is  available to residents  in
multifamily housing.   Service is offered once-weekly and there is no limit  to
the  amount of waste  that may be presented for collection provided only that
it is properly packaged.  These two factors mean that City residents have  no
incentive for either reducing the amount of waste  generated or to engage  in
illegal disposal methods in response to a fee increase.

        As we will  show below, the income elasticity estimated for Burbank,
given the assumptions  and model specification discussed below, is consistent
with previous findings.   Recall that  with a flat fee there is no non-zero
price which the resident faces but rather a  one-time charge  that operates
through an income effect.   Therefore,  there is  no price elasticity to  be
estimated.

        The basic equation estimated for the  Burbank case study is:
                                         b c
                              Q  =  a(RS) s

where   RS  represents  retail sales, which was  taken to be a surrogate for
income.  Taking logs on both sides,


        ln(Q)   =  ln(a)  +  b ln(RS)  +  c ln(s)                       (25)
In this section, we present the results of the econometric  investigation  of
equation (25).

        The first  step is to select seasonal variables.   For the estimation
of the  equation,  we have used dummy variables to  represent Winter (December,
January, and February), Spring (March, April, and May), and Summer (June,
July, and August).

        Using monthly totals  on waste collected and including the 30%  of
"commercial" waste that is estimated by city personnel to be residential,  the
results of applying ordinary least squares to equation (25)  gives (standard
errors are in parentheses):
        InQ   -   6.18 + .1711nRS - .06 WINTER +  .08 SPR + .07 SUMMER
                       (.11)      (.025)      (.025)     (.024)

                                                      R2  -  .38
                                                      N   -  83
                                     33

-------
        As shown In  the results of the regression,  the coefficient in the
 income  surrogate  is not  significant at the 95%  confidence level (the
 t-statistic is 1.62).  Before failing to reject the  null  hypothesis of no
 income  effect, some additional investigation was  performed.  The most
 plausible reason for the finding of no effect is that  the effects of income
 are  felt on waste  generation only after some lag.   Therefore, we use lagged
 (by one month) retail sales in the next regression.   The results are:


        InQ = 5.73 + .221 In(RS-l)  - .058 WINTER +  .081  SPR
                  (.106)           (.024)       (.024)

            + .068 SUMMER                                          (26)
             (.023)

                                               R2 =  .36
                                               N  = 81


        The coefficient on  the  income variable implies an income elasticity
of .22 and  is  statistically significant  at  the 95% level.   This figure is
consistent  with previous  findings as discussed  in Section B above.

        To  check the sensitivity  of  the results to  the  particular quantity.
variable, the  results of  several  other regressions are summarized in Table 1.
As shown there, none of the Implied  income  elasticities are  significantly
different  from  one another.   (All regressions  Include the  same  set of
independent variables as  in equation (26)).


           TABLE  1.  ALTERNATIVE INCOME ELASTICITIES


Excluding 30% of
commercial waste
Daily totals including
30% commercial
Daily totals excluding
30% commercial
Per capita monthly
Estimated
income
elasticity
.221
.272
.273
.233

t-statistic
2.05
2.83
2.81
2.13

R2
.36
.28
.27
.35
     including  30%
   Note:  For all regressions, the number of observations was 81.  The
          period was February,  1972 to December, 1977 excluding
          December, 1973 and January, 1974.
                                   34

-------
        These results  lend support to previous findings In the range of .2-.4
for  the Income elasticity.  Because this analysis  Is a time-series analysis
we would expect the  elasticities estimated to be  short-run elasticities and,
therefore, somewhat  lower than those estimated by cross-sectional analyses.

Sacramento, California

        The  City  of Sacramento, California allows residents to choose, for  a
fee, the number of cans  to be picked up during regular collections.  We have
referred to such  a  system as a capacity-based fee.   Service from the city is
mandatory for all  residents.  The basic service is offered once per week with
special collections,  for  an additional fee, available upon request.  As we
have discussed above,  such a fee system may induce  residents  to dispose of
less waste by imposing an additional  fee.  In  addition,  it may  induce
additional use of  illegal disposal methods.

        Unfortunately,  the data in Sacramento are not detailed enough for us
to make a definitive statement about these hypotheses.  While good data on
the  amounts of waste disposed over time are available, other important data
items, such as the number of containers chosen, are  not.

        An important difference in the Sacramento system is the provision for
the pickup of lawn and garden refuse.  Lawn and garden refuse  in Sacramento
is not  collected  by refuse crews.  Residents are not required to place their
garden waste in containers  (except  those in areas  without  curbs) but  are
required to pile  the  waste in the street no sooner  than one day prior to the
scheduled pickup.  Thus  all  references  to the solid waste  system in  this
section exclude lawn and garden trash.

        The basic  equation estimated was:


                                      bed
                               Q  •  ap y s


Taking logs on both  sides gives:


        ln(Q)  -  ln(a)  +  bln(p)  +  cln(y)  +  dln(s)               (27)


Again, as with Burbank, we will use dummy variables  to represent three of the
four seasons.   Further,  a surrogate  is required again  for the  income
variable.  We again  will use retail  sales in the City of  Sacramento.   The
basic one-can fee is  not subtracted since it is  small relative to income and
no true income variable  is available.

        Using  the data described above,  the following results were obtained
after applying OLS  to  equation  (27)  (again standard errors  are in
parentheses):
                                     35

-------
              4.7  +  .09 ln(RS)  +   .23 ln(p)  +  .01 WINTER
                     (.07)          (.09)         (.01)

              -  .01 SPRING  -   .02  SUMMER
                (.02)          (.01)                                  (28)

                                                R2  =  .40
                                                N   =  26
 There are three important results  to  note.  First,  the  income elasticity,
 while positive,  is  insignificantly different from zero.  Second, the price
 elasticity is positive -- and  significant.  Third,  the  seasonal dummies  are
 not significant.

        Because the existence  of a positive price elasticity is theoretically
 implausible, the  results in  equation (28) were analyzed further.  The first
 step is to recognize  that,  as  with Burbank, the effect of  income and price
 changes may occur only after some lag.  The effect of lagging retail sales
 and incremental price by one period is:


        Q  =  3.6   +   .20 ln(RS)  +  .22 ln(p)  +  .01 WINTER
                     (.08)          (.10)         (.01)

              +  .02  SPRING  -  .01 SUMMER                            (29)
                (.02)          (.01)

                                                R2 =  .50
                                                N   -  22
Performing this regression results in  a  positive and  significant income
elasticity of .20.  However, the problem of  a positive and significant price
elasticity remains.

        Because the price  variable used is deflated by the  CPI and, for
several  months,  the nominal fee is unchanged, the effect is that the
incremental fee may be  acting  as a  trend  variable.  To check this
possibility, we substitute a trend variable for the seasonal dummies.  When
this is done, the results are:


        ln(Q)  =  1.95  +   .35 ln(RS)  +  .09 ln(p)  -  .002T
                         (.12)          (.09)         (.001)

                                               R   -  .48            (30)
                                               N   -  22
The effect  of this substitution is  to  retain the positive,  and  significant
income elasticity (although the magnitude of the elasticity has increased to

                                    36

-------
.35).  While the  price elasticity is again  positive,  it is insignificantly
different from zero.  Further, while  the R   is about the  same as before,
there are variables making the regression significant at a higher level.

Provo, Utah

        Among cities  with user  fees for  household solid waste collection,
relatively few employ a variable fee structure  based exclusively on  pickup
location,  charging one  rate for backyard service and another  rate for
curbside service.  Provo, Utah, is one  such city.   Residents of Provo may
choose  between paying $2.50 per month for curbside collection and paying
$5.00 per month for backyard collection.

        Above,  a  demand model for a service-based fee was presented.  In  that
model, the amount of waste generated was seen to be affected by the  price of
the  service through the  effect of service  levels on waste generation.
Unfortunately,  the lack of data in Provo makes  it impossible to estimate the
demand specifications presented.

        Other data that is available provides an  opportunity  to indirectly
test  for an income effect in the demand for increased service levels.   This
data  is  the breakdown by  billing tracts of  the  percentage of residents
choosing backyard service.   While these  tracts do not correspond  to Census
Tracts,  they can  be identified in terms  of "housing value" which  might be
taken as an (albeit imperfect) surrogate for income.

        We specify, therefore, the following equation:

               al
        BY = a.P     +  a-TRACT.  +  a-TRACT9   +  a^TRACTo            (31)
              0         6     1      J     "      "     J

                                                                    al
where  BY  is the fraction of residents  choosing backyard  service,   P    is
the  price  of backyard  service, and  the  TRACT^   are dummy variables
representing the  different areas of the  city.   There are 57 observations for
each  of  the years 1975 and 1977.  These represent years before and  after the
fee.

        Applying  ordinary least squares  to  (31), we obtain:


        BY -  -5.14 +  .13P   + 2.7TRACT. +  1.7TRACT, +  .82TRACT-
                    (.13)    (.33)    l (.32)   *    (.33)

                                                      R2  = .43
                                                      N   - 114
We see that backyard  service in each of the tracts, which are  in declining
order  of  housing value, is significantly and  positively related to P.   Note,
also, that both the magnitude and significance of  the estimated coefficients
decline with TRACT number.

                                    37

-------
         These  findings in Provo lend  some support to the existing evidence of
 positive  income elasticities for  solid waste services.  However, because of
 the nature of  the data, the support is not strong.

 Grand Rapids, Michigan

         Grand  Rapids,  Michigan  operates a  non-mandatory solid waste
 collection service  which participating households  pay for by purchasing
 plastic bags  or tag cards whose  price includes  payment for municipal
 collection and disposal. Specially marked city  garbage bags are sold  for
 $2.30 per  dozen; tags are sold for  $1.25 for 10.   The  city also sells special
 plastic  trash cans whose contents are collected upon payment  of an annual fee
 of $4.35  or $6.50  depending on the container  size.   In addition, several
 privately  owned collection firms operate in Grand Rapids,  and residential
 households may  arrange service with  these rather  than pay for municipal
 service.

         A  metered-bag system might  be expected to affect residential solid
 waste behavior in two ways that are different from capacity  based systems.
 First,  the capacity of the bag is generally less  than that of a waste  can.
 Therefore, a metered-bag system faces the resident with  a smaller increment
 of waste allowed between addition of a new bag. Second, because the bag is
 consumed when filled, the resident  faces a direct charge for  each additional
 bag used.  With a  capacity based system,  once the  number  of containers is
 selected,  there is no direct charge to the customer  for  using the number  of
 cans chosen.  Thus, the metered bag appears to fall in  between the "ideal"
 system outlined above and a capacity-based system such as  that in Sacramento.

         This hybrid nature of the metered  bag  system  makes the modeling of
 consumer demand for  waste collection services somewhat more tricky than that
 of the capacity-based system.  Fortunately,  it turns  out  that it is possible
 with a single equation model to  estimate  the  effect  of price on  waste
 generation whether  the  bag system is considered to be more  like the "ideal"
 system with a charge  varying with weight  or  more like  the capacity system
 with  the unit of capacity being the  bag rather than the  can.

        Recall  that with  an ideal system,  the resident  is charged for  the
 weight presented for collection each collection period.  If the metered-bag
 system were a reasonably  good approximation of the ideal system, the resident
 would not be able to "hide"  additional waste by filling each bag  with more
 waste.  Therefore,  the ratio  of bags to waste  would  be constant and  the
 "price per pound" could be calculated  as:
        P = PBAG/(q/BAGS)                                          (32)


where   p  is the price  per  pound,   P-RAQ is the price per bag,  q   is the
weight of the waste presented for collectTon, and BAGS is the number  of bags
used.
                                    38

-------
        If a  double log specification is assumed  for  the waste generation
function,  it  would look like:


              al *2 *3
        q  - a P  y  s                                                (33)


where   s   represents seasonal factors,  y  is income,  and q  and  p  are as
before.   Substituting (32) into (33)  gives:

              al         ~al  ~a2 ~&3
        q  - a p   (q/BAGS)   y   s                                   (34)


        If the assumption concerning  the constancy of the  waste to bags ratio
is correct, we can rewrite (34) as:
            a   p   y  s                                              (35)
            0  BAG
where:
        a   - a  (q/BAGS)


Thus, we needn't make use of the data we have on the number of bags sold in
Grand Rapids.  This is fortunate since the data  available are on wholesale
sales and,  therefore, incorporate inventory phenomena not on the part of the
household but on the part of the retail outlets.

        Suppose that  instead of being viewed  as an ideal system,  the
metered-bag  system were merely another capacity based system with a different
representation of  capacity.   Then,  from the discussion in above,  the
theoretical  representation of the demand model  would be (assuming  again a
double-log specification) :
        q =  b BAGS  y  s                                             (36)
            0
                 c      c  c
        BAGS = c p      y  q                                         (37)
               0 BACT

where the variables are defined as above.  Substituting (37) into  (36)  and
simplifying  gives:

              bi    b? b^
        « '  VBAG y  '                                             (38>
                                    39

-------
Note that the specification in (38) is the same  as  that in (35).  Therefore,
estimation of either  equation provides information about both possibilities.
The difference is that if  the raetered bag system is actually a capacity based
system,  then  the coefficients estimated for  (38) no longer represent
elasticities of price  and  income.  Given some reasonable restrictions on the
elasticities in  (36)  and (37), however,  we can determine the sign on the
price and income terms if  the actual  elasticities are to  be  negative and
positive, respectively.

        For example, if  it is assumed that:
        o >  b  >  -1,   1 + aj(l + b3) >  o,   ja^l < &2


then  the signs on  bl  and   b£  would  be as expected  from the economic
theory.  Note, however, that the numerical estimates of  bj  and  b^  in (38)
cannot be construed as elasticities since  they  are  complicated functions of
other parameters and are not identifiable.

        In order to estimate Equation (35) [or  equivalently, Equation (38)],
we must first transform it into the linear form:


        ln(q) = In(b') + b'ln(p   ) + b'ln(y) + b's                   (39)
                    0     1    BAG    <£         j

The next step  is  to identify the actual  data  series used in the estimation.
For total waste disposed, the series presented  above was employed.  Because
some  of  the data  is expressed in cubic yards and some  in tons, it was
necessary to  make the two comparable.  To  do this,  we used the formula:
                             TONS - .29 YARDS
The factor of .29  is  based on experience from Grand Rapids.   For the "price
of bags," we use the price of a bag.  For the  period during which bags were
not available and  tags were used, we use the implicit price of a bag based on
the price of a tag.  Therefore, when bags were not available but tags were
and they cost $.10, we use .167 as the price of  a bag.

        For income, a better substitute was available in Grand Rapids than in
any of the other cities.  Grand Rapids has a city income tax  whose rate  has
not  changed over  the  period  estimated.  Therefore, we use quarterly
observations on income tax collections as the surrogate for personal income
with the quarterly observation being used in all three months.

        For the seasonal  variable, we use a dummy variable for Spring.  This
dummy  serves two purposes.   First, there is a high generation point in the
months of May and  June.   Second, personal income tax collections peak in


                                    40

-------
April with the filing of returns.   Both  of  these factors are held constant by
the use of the Spring dummy.

        Applying ordinary least  squares  to  (39) gives the following result:


        ln(q) - 7.1  -  .331n(p    )  +   .021n(y)  +   .14SPR            (40)
                       (.46)   BAG   (.25)        (.10)

                                               R2  - .08
                                               N   =50
As we  can see,  the  coefficients  on  price and income have the expected signs
but are insignificantly different  from  zero.

        It is  unlikely, however,  that the effect of price or income changes
are felt immediately.  If a distributed lag model is specified and estimated
the results are (standard errors in parentheses):
        ln(q) = 5.4  -  1.21n(p   )  +   .501n(p   )    +  .291n(p   )
                        (.89)  BAG      (1.12)BAG-1            BAG-2

        - .0091n(y)  -  .431n(y)   +  ,701n(y)   +  .23s
          (.37)         (.42)   -1    (.31)   "Z    (.11)
                                               R2  =  .20              (41)
                                               N   =46
None of the coefficients on price  are  significant and only the second  lagged
income term is significant (and  of the proper  sign).

        The positive price elasticities are  somewhat troubling even  though
they  are  insignificant.  We,  therefore,  next specified and estimated the
model:


        ln(q) =4.4  - .501n(p  ) +   .401n(y)  +   .17SPR
                      (.49)  BAG    (.21)    "2   (.09)

                                                R2   - .15              (42)
                                                N    =46


The signs of the coefficients are all as  expected.  The coefficients on both
the price  and income terms are insignificant at  the 952 level.
                                    41

-------
 Tacoma,  Washington

         Tacoma, Washington  has  some fifty years'  experience in financing
 solid waste collection and disposal through user fees.  Over  the years the
 city has considered or experimented with most of  the various fee systems in
 current  use around the country.  The present fee structure  is  a combination
 of a capacity-based system  and a location- or service-based system.   Hence,
 it provides an example of the effects of combining  several fee structures.

         Residential households  living in single unit or duplex structures
 receive  mandatory weekly collection at  a  rate  determined  by  the number of
 containers presented for collection and the level of  carryout service
 provided.  Households presenting refuse within  25 feet of a legal collection
 point (usually curbside, but alleys in some districts) are charged $2.45 per
 month for the first can and $1.15 per month for each additional can.   Those
 presenting refuse between 25 and 75 feet from collection points are charged
 $3.90 for the first can and $2.65 for each subsequent can.  Between 75 and
 200 feet, and over 200 feet,  the base prices are $5.30 and $6.65 and marginal
 prices are $4.05 and $5.45, respectively.   A charge of $1.60 times the  number
 of cans is levied for each flight of up to  six stairs  between points of
 collection and presentation.  These  are monthly fees  paid along  with
 electricity, water,  and sewage bills on a monthly or bi-monthly basis.  In
 addition, occasional extra bags of refuse  left alongside  regular cans are
 collected for a  charge of $.75 each.   Residents are required  to pay for the
 minimum  service of one can at  less than 25 feet;  about 30% of residential
 customers demand  some level of optional service,  a proportion .that has
 steadily increased historically.  Records  from  1947 and 1958 indicate that in
 those years about 92% of  all households  received the minimum level of
 service; in recent years the  figure has been about 75%.  Households living in
 structures of  three  or more units  are charged jointly for  service at
 commercial rates.   Finally, Tacoma residents may haul extra refuse  to the
 city  landfill and  dispose  of it at no  charge;  about 40% of  all household
 refuse by weight is disposed  in this way.

         The first step in estimating  the model for a combination structure
 presented above is to  express each of the  equations in a linear form.   This
 results in:
        ln(q)  = Ina  + a InC 4- a InS + a  Iny -f a.lnZ
                  U    A       &*       J      4


        ln(C)  - Inb  + b Inp  + b InS 4- b Iny
                  01        2       3

        ln(S)  = Ind  + d Inp  + d InC + d Iny
                  01        2       3

For income,  sales tax receipts were used.  The  seasonal variable used is
precipitation  (primarily because rainfall is relatively great in Western
Washington).
                                    42

-------
 r*rma «7  -J    Varfables 1» c» and S,  we  have used the detail available in
 terms of  the numbers of households  selecting a particular  service
 Therefore,  the monthly quantity variable is  in terms of ?ons per  household'

   * "
 the*  c-f  h                              8vCe  eve  seeete   we
 we use the    d    f V0r? tha° the baSlC  leVel °ffered'  That  ls.  *«r cans,
 we use the  odds of choosing more than one can of service which is  the ratio
 of the  number  of households  choosing two or more  cans of service  to  the
Using two-stage  least squares, the resulting estimates
                                                             are:
        ln(q) - -5.5 + .85  InC  -  .34 InS + .23 Iny - .0002  PRECIP
                       (.93)    (.77)      (.22)      (.003)


        ln(C) - 3.2  - .12 Inp   -  .06 Iny + .41 ln(q)  + .61  InS
                       Ml)      (-13)     (.42)        (.12)


        ln(S) = -5.9 + .09  Inp  + 1.76 InC + .13  Iny - .87  Iny
                       (•15>      (.41)      (.30)      (.97)


                 the  only  significant explanatory variables are between the
                 chosen.
variableoresentin                     8i*«ific™* ««» because the dummy
                                m°nth8 aPPears to be important in many of
        ln(q) - -2.88 + 1.0 InC - .42  ln(S) - .02 Iny - .08 WINTER
                       (-86)      (.76)       (.21)     (.04)
                         -6               .9
               + 2.7 x 10 ,  EXT +  9.9 x 10  _  SELF
                (5.3 x 10-°)      (1.3) xlO-8           R2  .  .36    (43)


       ln(C) - 2.3 -  18 Inp   -  .004  ln(y) + .08 ln(q)  + .72  ln(S)
                    (-05)       (.06)        (.08>       (.05)

                                                      R2  =  .92      (44)
                                   43

-------
        ln(S) = -3.1 +  .15 ln(p ) - .0002  ln(y) - .14 ln(q) +  1.5 ln(C)
                       (.06)        (.10)         (.13)      (.13)

                                                      R2 =  .88      (45)
 The  results of these regressions are  Interesting.  Looking  at equation (44)
 we see that there  is a negative and significant  price elasticity of  .18.
 That  is,  for every 10% increase in incremental can price, the percentage of
 households choosing more than one can of service  falls by 1.8%.  However,
 from  Equation (43) it appears that waste generation per  household is not
 significantly related to the percentage of households choosing more than one
 can  of  service.  If true, this would  suggest that the way people respond to
 increases in the incremental  price of containers is by demanding  fewer
 containers and using those demanded more intensively.  The  price elasticity
 for location is, however positive and significant.

        A  second approach  to  testing for the existence of negative and
 statistically significant price  elasticities is to  derive the reduced  form
 equation  for waste collected  and to  use OLS.  The reduced form (excluding
 SELF and EXT) can  be shown to be:

                a'  a'    a'        a'
        q  = a*  p    p    y    WINTER                                 (46)
             0  L   c

 When OLS is applied to the logs  of (46), the result is:


        ln(q) =  -2.3 - .47 Inp   + .59  Inp   + .01 Iny - .10 WINTER
                      (.26)  L  (.49)    c  (.18)     (.03)

                                                      R2 = .43
                                                      N  =41
Neither of  the coefficients  on  the price of an  incremental can nor  on
location is  significant at the .95  level.

        Because the prices  on containers  and  location  are set by the same
authority and change at the  same time, it might  be  suspected  that the
coefficient  estimates are inefficient  (but not  biased) because  of
muliticollinearity.  However, the two  terms are not perfectly correlated.  To
regress quantity disposed on  one of the prices (either container or location)
would result in  specification  error.   Further, because  the exact form  of
specification  error would  be one of omitted variables and the omitted
variable would be positively  related to the included  variable(s).  Therefore,
the coefficient estimated for the included variable would  be biased away from
zero.
                                    44

-------
CONCLUSIONS

        The  empirical results  from the study are  somewhat mixed.   Income
elasticities  consistent with  those  found in previous  studies  have been
estimated for Burbank and Sacramento and additional, but indirect, evidence
of an income effect on service level was found in  Provo, Utah.  Insignificant
income elasticities were estimated in Grand Rapids and Tacoma.

        With respect  to  price elasticities,   the  only  indication of
significant  coefficients  is from  Tacoma.   This finding  is,  however,
complicated  by the fact that there are two price  elasticities  of  interest,
one on incremental containers  and one on location. The finding of  a positive
price elasticity for container price suggests that some type of specification
error might  be the cause.

        Additional, and indirect, evidence concerning  price  elasticities
comes  from  attempts to examine  the effect of user charge  on illicit disposal
methods.  With data from  the  waste  deposited  in  city parks (or  other
surrogates)  we regressed quantity of "litter" on user fee.  For no  city was  a
significant  relationship found.  This finding is consistent with a  finding of
no significant price elasticity.
                                    45

-------
                               REFERENCES


1.   Anderson,  F.,  et.  al.,  Environmental  Improvement Through  Economic
     Incentives, Resources for the Future.

2.   Edwards,  F.  and B. Stevens, "Testing  the Efficiency of Alternative
     Organizational Schemes," in Evaluating the Organization of Service
     Delivery; Solid Waste  Collection and Disposal,  E.S. Savas and  B.
     Stevens,  Principal Investigators, Report to the National Science
     Foundation (New York:   Center  for Government  Studies,  Columbia
     University, 1977).

3.   McFarland, J.M., et.  al., "Economics of  Solid Waste Management,"  from
     Comprehensive Studies of Solid Waste Management, Final Report,  Sanitary
     Engineering Research  Laboratory, College of Engineering and  School of
     Public Health, Report No. 72-3, University  of California,  Berkeley,  May
     1972.

4.   Stevens, B., Pricing  Schemes for Refuse Collection Services;  The Impact
     on Refuse Generation,  Research Paper No.  154,  Graduate  School  of
     Business, Columbia  University, New York,  January 1977.

5.   Tolley, G.S., et. al., Economics of Municipal Solid Waste  Collection and
     Disposal Services.

6.   Wertz, Kenneth L.,  "Economic Factors Influencing Households' Production
     of Refuse," Journal of Environmental Economics and Management, Vol.  2,
     pp.  263-272, April  1976.
                                   46

-------
             EVALUATING THE EFFICIENCY OF THE  SOLID WASTE CHARGE



                                    by


                              STEVE BUCHANAN
INTRODUCTION

        The  national  Solid  Waste  Charge  (SWC)  policy  has received
considerable attention  from the EPA over  the past few years.   The  proposal is
now being evaluated  by  the Resource Conservation Committee, and  one  specific
legislative version has  been introduced by  Senator Hart.   There is good
reason for interest on  the  part of policy makers because the  potential
advantages of the SWC  are quite striking, particularly in comparison to some
of the alternative policies.

        It is  simple enough to state  the logic or rationale of  the  SWC.  The
charge would act as  an  excise tax on that component of products  entering the
municipal waste stream.  The amount  of  the tax would reflect the associated
costs of collection  and disposal, causing producers and consumers  to  reckon
with  those costs as they are incurred.  So,  the charge could,  in theory,
correct  resource allocation relative  to the  present situation  in  which
incremental waste contributions are non-priced and waste flows  and  materials
consumption are correspondingly excessive.  In  addition, the  charge  would
further  reduce waste flows and virgin materials consumption by  incorporating
a recycling incentive'in  the  form of exemption  (in whole or  part)  of the
recycled  content of the  taxed products.  Since  the charge would, in most
cases, be levied on  producers who purchase secondary and virgin  inputs, the
exemption  would encourage  increased  secondary  materials use.   In most
versions of the policy, the charge revenues would be distributed back to
municipalities to  fund  solid waste management.   In that  case they would
effectively reduce local property taxes.  However,  they could  be channeled
into other public goods or used to reduce personal income taxation,  as well.

        Thus,  potentially, a solid waste  charge can confer benefits  both by:
1)   improving the efficiency  of  public  finance or pricing  of  waste
management,  and 2)   by reducing rates  of  environmental spoilage  and natural
resource  consumption.  The  first objective  (which we call  "PF")  can be
quantified in terms of the resource misallocation or "excess  burden"  of
various taxes  which  could be used to finance the needed services.  The second
objective, conservation of environment and productive resource stocks  ("CE")


                                    47

-------
is much more difficult  to evaluate, as everyone  is aware.  PF and  CE are not
entirely disparate  since because of existing market failures we are creating
too much solid  waste—to the detriment of  CE.   An efficient  level  of solid
waste flow,  even measured in purely market values, would improve CE.

        Now  we  have  reached the point of departure for this paper—namely the
"efficiency" of the SWC.   The first thing we  can say is that if the charge
policy aims  at  two different objectives there can be no unambiguous measure
or  index of efficiency,  particularly when  one of the objective  functions
("CE") has mostly subjective  variables  for  its domain.   Second, we must
recognize that  the SWC  proposal includes a range of design options   and that
the problem  of  optimizing the design of the charge is itself  endogenous to
assessing its  effciency  with reference  to either of the stated objectives.
The major variables in the  charge design  are:  1)  the  level(s) of the
charge(s);  2)   the relative  size of exemptions for secondary materials; 3)
the points at which  charges and exemptions are  assessed; and  4)  the use of
the resulting revenues.

        Two  approaches  to evaluating the efficiency of the solid waste charge
will  be  described here.  The first approach is based on a narow view of the
range of policy concerns:  that the charge is  intended only  to internalize
the attributable costs of  collection and  disposal in  the  unit  price of
products. "Attributable costs" here means some  estimate of short or long-run
marginal cost,  leaving  aside the social cost concerns reflected in CE.

        Using consumer  and producer surplus changes resulting from the direct
burden  of the  charge  and adding the cost savings brought about by  a smaller
total waste  flow a lower bound for potential  net benefit  of  the  swc can be
found.    Then  the  approximate  sensitivity of net benefits  to certain
unavoidable  sources  of  error in the level  of  the charge is analyzed.  On the
basis of current (but  still  relatively crude) empirical estimates of the
variables involved, we  conclude that the  SWC  is not at all  likely  to be an
efficient means of internalizing (a narrow measure of) solid waste management
costs.

        The   second  approach to efficiency  described here acknowledges the
second best  nature of all the policy proposals, including  the status quo.
Thus  the disadvantages in  terms of  resource  misallocation, equity and the
other environmental  and conservation dimensions  inherent in the status quo  of
public  finance of solid'waste management  and of materials  production
incentives would be  accommodated.  Enlargening  the basis of comparability in
this  way would clearly allow a more comprehensive statement of the social
opportunity  costs of the various policy alternatives.  However, this approach
is limited by our ability to determine the social costs of the  status quo and
by  the  prevalence of subjective variables.  In addition, a number of
difficult-to-access  empirical quantities come  into this larger picture—such
as  long-run elasticities of supply of secondary materials, responsiveness of
industry to charges in relative prices or  to subsidies,  the  impact of
environmental standards on long-run disposal costs and many others.

        Here we do  not attempt a  comprehensive analysis of the  various
policies, their possible  interactions  and  their probable effectiveness  in

                                    48

-------
forwarding  the  two objectives we have mentioned.   Rather we identify  some
considerations bearing on  such an  analysis which would  tend to improve our
initial assessment  of  the solid waste  charge.   In  light  of  those
considerations—particularly those relating to the infirmities of the  status
quo policy—it  seems that the SWC could  possibly be part of  a superior policy
prescription.   Some more  study would be necessary to confirm or deny that
proposition.

PERFORMANCE OF  THE SWC AS A PRICING INSTRUMENT

        Here  we develop a lower bound measure for the potential net benefits
conferred  by an ideal SWC without, a recycling exception.  This amounts to a
worst case  assumption in  that producers do not  respond to the  recycling
incentive nor take steps  to reduce the overall  waste content of  their
products, but simply pass the charge  through.  The purpose of the  following
analysis is  to establish  an acceptable measure  of  the sensitivity of net
benefits to error in the charge level; then we can make some rough  estimates
of how  well  the actual charge is likely to  perform given that there are
significant sources of error in the charge level.  The  main  sources are:  1)
substantial  variation among municipalities in solid waste  management costs;
2)   ambiguity  in allocating solid waste management costs among  waste
components;  and 3)   error  in estimating a national  average marginal cost,
thus creating bias in the charge level.

Net Benefits  of a Solid Waste Charge

        To  begin with,  the net welfare benefit of the solid waste disposal
charge can  be analyzed as the algebraic  sum of several  components of welfare
change.   Initially  we  will assume a homogeneous  waste stream,  having
geographically  uniform and  exactly measurable  per unit marginal costs of
collection and disposal.  Thus, the  level of the national solid waste charge
corresponds precisely to the real resource savings associated with  disposing
of one  less  unit of post-consumer waste.  For the time being, long and  short
run marginal  solid waste management costs are assumed  to be constant.   The
implications of large  incremental costs incurred in  switching to a new
disposal site will be discussed later.  We are  considering, then, an ideal
charge,  which perfectly internalizes  the relevant costs.

        The solid waste charge will raise prices  of the  waste component of
final products in accordance with the associated solid  waste  management
costs. Unless demand for the product is  perfectly inelastic or if the  supply
curve is a horizontal  line (reflecting constant costs) the new quantity of
product  traded  will be less than the  original quantity and the new  price will
be greater than the original by an amount less than the per-unit solid  waste
charge.  The increased price and  reduced  quantity imply a reduction in
welfare  for  consumers  and, in general, for producers of goods bearing the
charge—that  is, a loss in producers' and consumers' surplus.  Part  of the
loss  in  producers' and consumers' surplus is regained by society in the form
of the charge revenues.  Additionally, solid waste management resources are
saved by the reduced waste load corresponding  to the lower  quantities of
waste-producing goods consumed.  That resource  savings represents a  social
gain.  Finally, the amount of solid waste management services demanded may be

                                    49

-------
 further reduced by increased recycling,  if the charge  scheme includes  an
 exemption for recycled materials.   Recycling would  lessen municipal waste
 management costs to the  extent  that secondary materials are diverted from the
 collection and/or disposal  stages.

        dearly,  the resultant  net welfare changes  should be  positive  after
 correctly internalizing waste management costs.  However, the costs  of
 administering the charge and  of disbursing the revenues  must  be subtracted
 from  the  total  net benefit.  For the  time being we  will assume that these
 costs are a positive function of the charge revenues, denoted a(R).

        Now we can state the elements of net welfare change in the following
 general expression:

        AW = - (CS + PS)  +  WR + SR + R - a(R)                          (1)

        where

        CS,  PS    are the losses in consumers'  and  producers' sur-
                  plus,  respectively,

        SR,  SR    are the solid waste management cost savings attri-
                  butable to  reduced quantities of  post-consumer
                  product waste and increased usage of  secondary
                  materials,  respectively, and

        R, a(R)    are the charge revenues and administrative costs,
                  respectively.

        It should be noted  that  some actual  use  must be made of the charge
 revenues,  if they  are to be reckoned  as a benefit.  The  funds could  not
 produce benefit  if  they were, for  example,  placed in  escrow indefinitely,
 except for the possibly  salutory effects of reducing the  velocity of money.
 In  fact,  they  should  be  applied  to  the margins  of social or private
 expenditure prevailing at the new relative prices induced by the charge.

 Lower Bound of Potential Net  Benefits

        It would be useful  to determine what level of  potential benefit an
 ideal  solid  waste  charge-  would yield.   The analysis  is restricted  to
 "potential"  benefits because the administrative costs will be left aside.
 Other work has been,  and will be done to assess the size of  administrative
 costs.  Also,  we will initially ignore the savings in  solid waste management
costs arising  from increased  secondary materials use.   In other words,  firms
 pass  the  entire  amount of the charge  forward as increased product prices.
These assumptions characterize  a lower bound for potential net benefit  of  an
 ideal  charge  policy,  since  the  recycling benefits are not included.

        The  expression for  lower bound of potential net  welfare charge,  as
just defined,  is  the  following:
                                    50

-------
        AW
- (CS + PS)  + WR + R
                                                                      (2)
        It  should  be noted  that  the values  of  each of  the right-hand-side
terms in equation (2) would  in general be different  from  the corresponding
terms  of  equation  (1).   This is true because the price quantity equilibrium
in the case of (2)  is predicated  on  all of the  solid waste charge  being
shifted forward,  whereas equation (1) includes recycling effects, which would
reduce foward shifting.

        The potential lower bound of net welfare benefit can now be found.
Denote the per unit  charge level by  t.   Then the charge causes the  final
product  supply curve to  shift vertically, by the distance  t  in Figure 1,
from  s  to  s .  The resulting surplus loss is represented by the rectangle
p ade  plus the "deadweight loss" triangle  acd.   The consumers' surplus loss
is the area above the old  price line, pcb, and the producers'  loss is  that
below  the old price  line.  The area of rectangle   p  ade   equals  t .q , which
is also equal  to  the  total revenues yielded by the charge.  Therefore,  those
components of total  welfare change  "net out" as  a  transfer payment.  The
remaining  terms, then, are the  waste  reduction effect and  the deadweight
triangle,  acd.
            price
         Figure 1.  Effects of an excise tax on the
                    component of products.
                                                     Quantity
                                     solid waste
                                   51

-------
         Since  the  per unit charge,   t,  accurately  measures the per  unit
 savings in waste  management costs,  those savings  amount to  t. q.  That
 benefit  is offset by the  deadweight loss which is equal  to   (t. q)/2.  To see
 this, note that  the  area  of  the deadweight loss triangle   acd  can be found
 as

         acd = acb + bed

            = ApAq/2 + (t - Ap)Aq/2

            = tAq/2

Here  it  is assumed  that the demand and supply curve segments, ac and dc ,  are
 linear.   Given the small  relative price change brought  about by the charge,
 this  assumption should give rise to virtually no error.

         Summarizing  these results  in  the form  of  equation (2), the  net
welfare change, under  the assumptions detailed above,  can be expressed as
        AW = - (tqj +  tAq/2) + tAq  + tq j^                              (3)

           = tAq/2

This  holds true irrespective of the elasticities of  demand and supply, or if
the good is supplied under conditions of constant costs.   The elasticities
do, of course, enter in determining the size of  Aq, however.

        Therefore,  a lower bound of potential net benefit of the solid waste
charge may be calculated  as the per unit charge  times  one half  of  the
reduction in quantity  consumed.  For example, if a solid waste charge of  3
cents per container were passed  through  entirely by  the container
manufacturer, and  the resulting decrease  in annual  quantity traded were
300,000 units, then the potential annual net benefit  of charging direct solid
waste costs for those  containers would be $4,500.  It should  be pointed  out
that  the  benefits  of  the solid waste charge rise in  direct proportion to  the
size of the per-unit costs internalized.   To the extent that the relative
price  of  solid waste management continues  to rise,  the potential discounted
future net benefits of the policy will also  increase.

Sensitivity of Net  Benefits, to Error in the  Charge Level

        As has been mentioned, one prominent issue in the design of the solid
waste charge concerns the  preciseness of  the  charge level  in reflecting
actual  solid waste management costs.   Under somewhat limiting assumptions we
will now examine the welfare  losses that occur when  the  charge level  and
costs of  waste management differ.

        We will again  assume a homogeneous  solid waste  stream, having  per
unit collection and  disposal costs  denoted by c.   Also, we assume  for  the
time being that the taxed product is supplied under constant cost conditions
and that  competition prevails.  Thus the  supply curve is a horizontal line.
The per unit tax  shifts the  price up by  the  full  amount,  t, of the  tax

                                    52

-------
because  the  supply curve  is  horizontal.   Because of  error  in  the
specification of the charge  level, t and c are not  equal.  The expression for
net welfare change becomes

        AW = - (tq][ + l/2tAq(t)) + cAq (t)  +  t^                    (4)

           = Aq(t)(2c -  t)
                    2

        The change in quantity traded depends on the  charge level and so  must
be  written asAq(t). Whent  t = c, the welfare change  is the maximal value,
W* , equal to  Aqc/2  as  before.  Again  assuming  a  constant  elasticity  of
demand, e , Aq( t)  can be computed as

        Aq(t) =    qE                                               (5)
Let "a" denote a constant  of proportionality between  t and  c,  viz., t =  ac .
Then  we may examine  the  sensitivity  of  AW to deviation between the charge
level and waste management costs using the function AW(a).

         AW(a) = i|ai (2c  - ac)                                      (6)


               = (2c.2a - a2c2) fl                                    (6a)


        This function  behaves  as it should, having  its maximum of AW*
(=c'iqE/2p) at a =  1,  since

                                ;and                               (7)
A second important  and  obvious characteristic of W(a)  is that it is zero  for
a = 2 and negative  for  larger values.  Similarly, for  a - 0  W is zero, as we
would expect.   Thus,  if the solid waste charge overstates the actual relevant
costs by 100 percent  then it results in no net benefit and actually reduces
welfare if the overstatement exceeds 100 percent.

        However, under the present assumptions,  we find the net  benefit
calculus is quite forgiving  of relatively small values  of  error.  The
relevant part of  the  parabolic function  W(a) is graphed in Figure 2 (for
= 1) and pairwise  values of  W and charge error are set forth in Table  1.   A
10 percent error  in the charge level gives rise to only a one percent loss of
lower bound potential welfare, and a 20 percent error  corresponds to  a loss
of four  percent.  The rate  of benefit dropoff increases more rapidly with
greater deviation  between  the charge  level and  the relevant solid  waste
management costs.  While a 25 percent error in the charge results in only a
6.2 percent loss  of welfare, an additional 25 percent  error  leads to  a loss

                                     53

-------
 in  realized benefit of 25 percent,  or a fourfold loss resulting from a
 twofold greater  error.  A third  increment of 25 percent error—or 75 percent
 error—reduces benefits by about  56 perdent.
   TABLE 1.  SENSITIVITY OF LOWER BOUND POTENTIAL NET
              BENEFIT TO ERROR IN THE CHARGE LEVEL FOR
              e = 1

Value of
a (=t/c)
1.0
.9 or 1.1
.8 or 1.2
.75 or 1.25
.67 or 1.33
.5 or 1.5
.33 or 1.67
.25 or 1.75
.2 or 1.8
.1 or 1.9
0 or .2
Percent. error in
charge level
(|t - c|/c)100
0
10.0
20.0
25.0
33.0
50.0
67.0
75.0
80.0
90.0
100.0
Percent of
potential welfare lost
((AW* - W(a))/AW*) 100
0
1.0
4.0
6.2
11.6
25.0
44.9
56.2
64.0
81.0
100.0

        For the  present  purpose  we retain the  assumption of  unitary
elasticity  of  demand,  as a convenient  approximation  to the  range  of
elasticities  for  products  affected by the  solid waste charge.  It would not
be difficult, in a more  detailed analysis,  to utilize the elasticity and
consumption  shares data for major  product  categories used by Research
Triangle Institute in their  economic model of the  solid waste charge.   The
effect of higher elasticity  is to increase the  sensitivity of net benefits to
error.  In othe words, for products having price elasticities of demand  in
excess of one, the parabola  in Figure 2 would be more steeply shouldered.
                                    54

-------
    AW (a)
     AW*
     5AW*
                                   a =  t/c
            Figure 2.  Lower bound of potential net benefit as a

                       function of error in the charge level.
Applicability of the Analysis



        Now it  is appropriate  to inquire whether  the results just described

have much  relevance to estimating  the  net benefits of  the  charge under

Lnn   hi   T  U/?nS*  Preliminaryily.  we conclude that  the  method is
applicable  and useful as it  stands, and can be readily refined further.   The

fTctio'ns ° th8 °fhC°nSta"t elasticity  and  linearity of demand  and  supply

n^M %  ^8   TradiCt°ry' CSnn0t contr^ute  significant  error  except
possibly in those  few cases where the charge induces a very noticeable jump

aive riT  T    '     ^ ^ " iS unlikel? that  these assumptions would
give rise  to error  greater than that which would result in  an  attempt to
                                   55

-------
estimate entire supply and demand schedules.   Clearly there  is  no  payoff in
the kind of refinement.

        The major  assumption is that of the constant cost  supply  function.
It  seems reasonable that a great percentage  of consumer products are  in  fact
supplied under  conditions approximating constant costs, particularly since we
are considering aggregated  national markets.   Yet it would be  useful to
determine what  happens  to  our analysis  if  that  is not the case, and for
certain segments of the market it will  not be.

        With a  rising supply curve some of the tax burden  will be  borne by
producers.  That case,  which is of  course  the  general one,  is depicted in
Figure 1.  There the tax wedge raises the  consumer's price by  p  and  reduces
the suppliers' price by   ep.  Ultimately  the  producers' share of  the tax
would  be shifted  back  to owners of  those  inputs  having the  lowest
elasticities of supply  employed in  the industry, in the form of  reduced
rents.

        However, producers also have the  option of substituting  away from the
taxed  dimension of the  product, which  in  this  case is  essentially  the
packaging  component of  the  product,  or  the  virgin materials  content of the
packaging.   There  are two basic kinds  of  substitution possibilities which
allow producers to reduce their tax  burden:

        •       reducing the proportion of virgin  materials at  the basic
               materials level

        •       reducing the packaging  component of the product, or  at
               least the taxed dimensions of  the  packaging component.

        For the most products  these two  possibilities occur  at  different
stages of  the  production  process, usually  in  different  firms.   But the
observaton  still holds, regardless of  which  stage of production  is taxed,
that  producers having  relatively strongly  upward sloping supply  functions
have greater incentive to reduce the tax.  This effect makes the lower bound
measure of net benefit analyzed above more  applicable to the  upward  sloping
supply case, insofar as producers substitute  away  from the waste component as
a whole.   It makes no difference which side  of the market brings about waste
reduction effects  in measuring the net  benefit of  the charge.

        To  the extent  the  proportion of recycled materials is increased in
order  to reduce the tax, our  formulation understates the net benefit by
leaving out the recycling benefits,  viz.,  the  term "SR" in equation  (1).  Yet
the value  of such benefits,  incorporated  in  the level  of the  secondary
materials exemption, is  the  expenditure of solid  waste management
expenditures avoided by recycling.  This mode  of substitution amounts  to  what
is also basically a waste reduction  effect.

        Thus, on the whole,  the lower bound method just  described above
should be a useful and acceptable approximation for policy analysis.
                                     56

-------
        To  the  extent this analysis is actually representative  of  the net
benefit  behavior of the  solid waste charge,  the message  is  clear:   the
efficiency  benefits of the policy disappear  entirely if the degree of error
in measuring the relevant costs and in  implementing the charge  is comparable
to the size of  the charge itself.  It seems  that  this in fact is the  case.

Assessing Net Benefit Losses

Source of Error in the Charge/Exemption  Levels—

        Several sources of error or bias  in the level of a  national solid
waste charge can cause misstatement of the relevant costs which  the charge is
intended to internalize.  They are:

        •       Differences among municipalities  in the relevant costs
                of waste collection and  disposal.

        •       The existence of local user  fees  which partly or wholly
                recover waste management costs in  about 10 percent of
                local jurisdictions.

        •       Ambiguity in allocating  solid waste costs among  materials
                types and waste components.

        •       Determining the amount of the recycled materials exemption,
                based on estimates of the waste  management costs avoided
                by increased secondary materials  use.

        •       Determining the level of the charge based on estimates of
                the national average waste management costs imposed by
                increased disposal per household  at the margin.

        The first point  above is  probably the most  significant,   because
interlocal solid waste  management cost  variability  appears to  be very
substantial.  So, the charge overstates or understates  the relevant costs  in
accordance with  the variance of  observed  costs,  in  effect creating  a
component of random error in the charge from the point  of view of individual
communities.   We will concentrate  on  assessing the  impact  of the  error
component here, not only because it probably dominates  other  kinds of  error
or bias, but   also because there  is at least some of  the needed empirical
information available.

        The existence  of local user fees could create  some  significant bias
in the charge/exemption levels.  Local user  fees combined with the SWC  would
create  a situation of double-charging for waste collection in certain areas.
It seems reasonable  to expect  that  adjusting or discontinuing local user fees
would be part of the  start-up costs  of the SVJC, if,  in  fact, the charge
revenues were  to be redistributed  to  cover  local  solid waste  costs.
Otherwise, inhabitants  of those  areas would  be systematically overcharged,
relative to the rest of the nation.   This would  happen  to  the extent that
local fees were more  highly correlated with waste  management costs  per
household  than the property tax  or other sources of  funds  relied upon

                                    57

-------
elsewhere.  Evidence suggests that that  correlation is quite poor  in  the case
of  the  property tax and  noticeably imperfect in the case of  user fees.  We
will not attempt  to assess benefit losses  from imposing charges in  addition
to local fees.  The main points here are that evaluating those  losses depends
on  the  use  of the charge revenues and  on  the  second best  nature of the
alternatives,  which we discuss in more detail later on.

        The  problem of allocating costs  among waste materials  components
seems  to be a rather severe one, even at first glance.   First,  those costs
clearly depend  on the weight and volume  characteristics of waste components.
However, even in a given jurisdiction, the weight/volume cost relationships
can change if  packer trucks have different compaction ratios, or if differing
pickup (e.g.,  for apartments vs. houses) or disposal methods are used.  Also,
the weather can significantly alter  the relative cost  shares of waste
components.

        The volume/weight  ratios of cardboard  and  newspaper  change
drastically  when  they become wet, thus creating daily and  seasonal changes in
cost shares.  Finally,  the  compaction characteristics of any  given  item
depend  on the mix of waste surrounding it.  For example, cans will  compress
relatively more when surrounded only by  other cans than by  corrugated  boxes
and plastic  containers.

        EPA has initiated a project with the National Bureau of Standards to
measure compaction ratios  of waste components in a "typical"  mixed waste
environment  and to devise a cost allocation procedure, using representative
assumptions on pickup,  transport and disposal in a landfill.  No  results are
available as  yet, but  obviously there  will be some error  in  those  cost
estimates.   Here we will  simply make  some hopefully reasonable  assumptions
about  the size  of that error in order to  assess  the resulting net  benefit
losses.

        The  final two points mentioned above, determination of  the charge and
exemption levels, entail both conceptual  and  empirical  questions.   The
conceptual issues are the economic bases  for  the levels of the  charge and
exempion.  The  charge should reflect all costs attributable to  the additional
unit  of household post-consumer waste.  Those costs could be reckoned
relatively  narrowly or broadly as we discuss  in Section III, below.  At
minimum, a  more  inclusive  measure would also  reflect costs of  capacity
expansion and  environmental upgrading of  landfills.  The more  inclusive cost
bias notion  would be more applicable in  some areas than in others. Likewise,
the exemption  should reflect costs avoided by recycling.  Those avoided costs
may be collection costs, and/or disposal  costs,  depending  on locality and
type of material.

        In addition to bias and errors arising from Improper economic  basis
of the charge  and exemption, some further  inaccuracy will  occur in estimating
national average values  of the  cost components used in  setting  the
charge/exempion levels.  Some of the deficiencies of present  best estimates
of marginal  collection and disposal costs  are mentioned later.   The estimates
of  exemption  levels rely on 'the original cost estimates as well  as  on other
empirial information which has not yet been adequately sampled or analyzed.

                                    58

-------
The lack of precision in determining national average  charge and exemption
values will lead  to some degree  of  bias in the  impact of the charge  on
economic incentives.   Again, we must  rely on guesswork  in looking at the
economic effects of  that bias.  Presumably, such effects  would tend to  abate
if  the charge were  properly implemented, with some attention placed  on
improving data bases  and empirical estimates.

Probable Magnitude of Benefit Loss—

        As should be  clear by now, the best we can do at present in assessing
the efficiency of the SWC as a pricing instrument is to approximate a likely
range of potential benefit loss resulting from the  various  sources of error.
Ideally, we should treat each source of error as a stochastic variable with a
known or estimated distribution, and the total effect should be computed as a
sum of random variables.  Also, the welfare loss estimates should be based  on
estimated elasticities of demand  and  supply  for the  particular markets
affected.  For this purpose however, the assumptions of competitive supply
and unitary price elasticity of demand do not appear to be  unreasonable. The
consumption data used in the RTI model indicate  a demand  elasticity of .99
for  the non-food, non-beverage segment of the market—which bears by far the
greatest burden of the charge on paper and flexible packaging.  Also,  that
market segment  represents 75 percent of personal  consumption expenditures.
The major deviations  from unitary demand elasticities occur in beverage,
beer, wine and spirits categories comprising  2.7 percent of personal
consumption expenditures.  Demand elasticities range from .4 (beer) to 1.7
soft  drinks to  2.2  (wine) in those markets.  Canned foods, which could also
be relatively strongly affected by the charge, exhibit  low price elasticities
(.17)  and comprise  1.3 percent of  household  expenditures.  The consumer
surplus loss in inelastic  demand markets would be relatively less sensitive
to error in the charge level.

        Now it remains to  combine our scanty knowledge  of  the various sources
of  error and bias with the welfare  loss function of Figure 2 and Table  1.
The variance of  local  waste management costs  can be assesed only rather
indirectly. Data on  average cost per ton (collection only)  are available for
177 cities of greater than 2,500 population from the Columbia  Survey.   Table
2 shows the distribution of those  costs in 1974 dollars.  The national
average collection cost in 1974 was $24.91/ton.   Disposal costs averaged
$4/ton bringing the 1974 average total cost per ton  to  $29.  The distribution
of disposal costs (as measured by fees charged to private  sector collectors)
also showed considerable dispersion according to EPA staff.

        Assuming that the  charge was set at $25/ton, the  efficiency losses
resulting from variation of local costs are shown in Table 2.

        The overall  efficiency of  the  SWC computed  on  this basis is 65
percent.   In other words, only 65 percent of the welfare gains from properly
internalizing waste management costs are  realized.-  That figure neglects the
sizable  net losses  that  would occur in about  10 percent of the communities
                                    59

-------
TABLE 2.  DISTRIBUTION OF AVERAGE PER TON COLLECTION COSTS
           AMONG MUNICIPALITIES AND CORRESPONDING SWC
           EFFICIENCY LOSSES

Cost class
$/ton
0-10
10-15
15-20
20-25
25-30
• 30-35
35-40
40-45
45-50
50-70
70
Number
in
class
19
29
32
34
15
14
12
8
5
3
6
Percent
in
class
10.7
16.4
18.0
19.2
8.5
7.9
6.8
4.5
2.8
1.7
3.4
Deviation
from charge SWC
level % of efficiency
class median by class
233 (median = 7.5)
100
43
11
9
23
33
41
47
58
66 (medium = 75)

0
.84
.98
.98
.93
.88
.84
.76
.65
.55
Weighted
efficiency
(%)


15.12
18.81
8.82
7.35
5.98
3.78
2.13
1.11
1.87
65%
Based  on national average  collection costs  of $25/ton, 1974 dollars.
                                   60

-------
where the  cost are less than half  the  charge rate.  More  importantly, the
estimate  also  is based on  average  costs, not marginal costs.   Only if the
distribution of average costs is similar to that of marginal  costs does our
65 percent efficiency value have much significance.  Engineering analysis of
municipal  waste production functions  seems to be the only feasible way to
determine whether  the relevant marginal  cost would have a broader
distribution than average costs.   I  am  inclined to think that is not the
case, but  if it is, then it  appears that a big share of the potential benefit
of the charge  stands in jeopardy.   It should also be noted  that if the
relevant  cost  distribution  has  a mode in the neighborhood of  one-half the
charge rate or  less, then  net losses  occur  for inhabitants of those
communities.   That could  undermine the  efficacy  of the SWC policy
considerably.   The empirical data  sample used here does  not adequately
represent  major population centers, where costs are higher.

        Assuming  that local  cost variation does result in approximately a 35
percent loss  in net SWC benefits,  it  seems reasonable  to expect that the
other Implementation problems—estimating precise central  values for the
charge/exempion level and  allocating  costs among waste components—could
easily reduce the overall efficiency of  the SWC to below 50 percent.

        Some  empirical work is  necessary  to  guage  the degree of error in
assigning  cost shares among  various waste components.  In one sense, it  seems
to be misleading to think that those shares are fundamentally well-defined.
There is a problem akin to an index number problem in economics,  in assigning
those cost shares.  Beyond  that,  there will always  be some geographic
variation in cost shares,  owing  differences in relative importance of
collection,  transport and  disposal  costs, differences in  climate and
differences in mix of waste, among other things.  If error  in assigning cost
shares  averaged 33 percent  across  all goods and communities,  then the net
benefit of the charge could  be expected  to decline an additional  11 percent.
(It may be possible to minimize those losses if the cost  shares of relatively
elastically demanded goods,  e.g., wine,  soft drinks, prepared beverages  could
be determined relatively more precisely.)

        Finally,  if bias in  the charge  and exemption  estimates yielded  a
further loss of 5 percent of the benefits, we would find  that the SWC yielded
only about half of its potential benefits.  (Potential benefits here excludes
waste  reduction effects associated with increased recycling.  Those effects
do not  figure  importantly  in the  RTI  estimates of total net benefit.)
According to RTI's recent  estimates of total net benefit, a reduction by
one-half would place  the net benefit of  the charge in the  range of $13-$30
million  annually, depending  on  whether elasticities of supply of secondary
materials are  assumed  to  be relatively low or  high,  respectively.
Subtracting  $10 million for estimated  annual administrative costs,  we find
that the SWC appears likely to be  a rather marginal policy  alternative.  It
should  be noted, though, that the RTI estimates are preliminary.  Also, they
are based  on a  $30/ton charge rate,  which probably comes much nearer to
average cost pricing rather  than marginal cost pricing.

        Obviously, further work is necessary before we can  assess either the
efficiency or  the net benefit  of  the charge  as a  pricing instrument with

                                    61

-------
 sufficient reliability to make a policy determination.   However, if we
 consider the SWC  as  one of several second-best  public finance instruments,
 then much further refinement of data and estimating  procedures appears to be
 somewhat lower in priority.

THE SOLID WASTE  CHARGE IN A SECOND-BEST  SETTING

        The previous  discussion of  the  efficiency of the  SWC purely as a
pricing or public finance instrument assumes that  there are no  pre-existing
distortions elsewhere  in the economy.  Otherwise the demand and  supply curves
central  to the  analysis do  not fairly  represent the opportunity costs
associated with price  and quantity movements.   Because of the smallness of
the SWC relative to prices of most of the  affected goods, we are dealing with
truly marginal changes,  so our estimates of the efficiency performance of the
charge are fairly robust.  However,  what of the distortions inherent in the
status quo or in the other contemplated  policy  alternatives?   We would not
have  formulated our  two  objectives at  all  if  there were not pre-existing
market  failure  in the  waste  sector.  So, although  the charges  may  in
actuality yield  little of its theoretical  potential benefit or may even  yield
negative net benefits,  the  net benefits of  the status quo  and/or other
alternatives may be worse yet.  So, the less-than-ideal expected performance
of the SWC must  be matched  against that  of the  other policies.  Below we
discuss several considerations arising  out of the second-best approach  which
could conceivably place  the SWC proposal  in  ascendancy, all  things taken
together.

        We also  discuss  the level of charge.   The potential  net benefits
 increase with  the size of the cost internalized  in the charge  but the  usual
marginal cost pricing  logic combined with  current  econometric  estimates of
marginal cost may yield a charge level which falls substantially short of the
optimal one.  We take  up this issue first.

Cost Basis of the SWC

        Determining what the level  of the SWC should be raises two obvious
problems.  First is the problem of deciding which  elements of cost should be
 included in the charge.   Even if we  leave  aside extramarket social cost
(reflecting environmental  and  conservation objectives)  the answer is not
obvious.   Having decided on a cost  basis  there is, then,  the issue of
reliably estimating the costs.

        The  studies now being conducted by  the EPA  rely on  estimates of
long-run marginal  cost to set the charge level.  There  are several margins in
municipal waste management cost functions.   In collection the  margins  occur
in providing service to  an  additional household, providing a higher level of
 service, and in picking up a greater quantity per  household.  Additionally,
there are costs  of  transport, transfer and disposal  which  depend largely on
the location of the disposal site and  somewhat on  route patterns.  Finally,
there  are long-run costs  of  providing  new  disposal facilities.  In the
short-run the solid waste charge would  impinge directly only on the quantity
per household margin, and very slightly  on transport and disposal.


                                     62

-------
        Estimates of long-run marginal  costs using the best  available
(Columbia University)  data indicates  a  very low elasticity  of  cost  with
respect to  quantity per household—certainly no greater than 15  percent of
average  per  ton costs.   These  estimates are unreliable for several  reasons:

        •        The  sample of cities, while quite large  enough, excludes
                cities of greater  than  700,000 population, where the per
                capita costs would  be highest.  Also,  towns of less than
                2,500 were excluded.

        •        Costs of transfer  stations and incineration are excluded.

        •        Disposal costs are  measured by the disposal fees muni-
                cipalities charge.

EPA is  working  to overcome these  data  deficiencies and  to improve estimation
procedures, but even then  one would  have  to  question the  efficacy of
econometric techniques  in  producing  reliable estimates of the needed  kind.
The regressions produced  so far  show  inconsistencies  in  signs of
coefficients, spotty statistical significance  and gross differences  in
results on a ton vs  cubic yards measure of output.  Given the sensitivity  of
net benefits of the SWC to the relevant cost level,  it seems  inadvisable  to
rely on econometric estimates of cost elasticities. Engineering-based
estimates  would  surely be preferable  in this  instance.

        Returning  to the question  of  selecting the cost  basis of the  charge,
it  would  seem the  empirical estimates of long-run marginal costs could not
capture some of  the  relevant costs associated  with  higher  overall rates  of
disposal.   The costs of  acquiring  new landfills and  of operation with new
disposal sites are reported to be  increasing  rapidly.  Those costs  must  be
incurred  sooner without a SWC than  otherwise (unless implementing user fees
would  reduce wasteloads more than  the national  charge).  Perhaps  more
important, the  costs of  upgrading disposal  sites to  environmentally
acceptable standards, which may be on the order of $3  ton per year,  are not
adequately reflected  in  the 1974  data.   Some  portion  of both of  these
categories of incremental costs should  be  included in  the solid waste charge
level.

        The next question  concerns  the other  elements of cost mentioned
above.   Should  some of those be included  in  the SWC,  so that it approximates
more nearly an average rather than a  marginal  cost change?   This would amount
to  using  the SWC  as an  additional  taxation  instrument  to raise social
overhead  revenues  associated with  the  fixed  cost elements  of waste
management.

Optimal Taxation Considerations

        Using the SWC  as a means of  raising social  overhead  implies the
substitution of  an excise tax for  a property  tax (and  to a minor extent local
income  taxes).   In that  case, the  charge  could be  used to recover all the
cost components  except those related  to level  of service which  would probably


                                    63

-------
be more  efficiently and equitably  recovered through a  flat rate related to
service level.  An average cost recovery scheme would clearly result  in  a
much higher  charge  level,  (by  factor  of 3  to  6)  than the presently
contemplated marginal cost-based charge.  It would  create some excess  burden
effects,  which must be compared  to  those of the existing  property tax-based
financing arrangements.

        Research done to date indicates that the static,  and especially the
dynamic,  efficiency losses associated with the SWC  would  be noticably less
than those for  the property  tax.   The horizontal  equity of the SWC would
clearly  be greater, assuming that  redistribution of revenues could be kept
reasonably in line with local costs.   The incidence of the property  tax is
undoubtedly more poorly correlated with household waste generation rates than
the SWC.   In fact, the solid waste charge may  resemble the formulation for
optimal  commodity taxation which  prescribe relative small,  pervasive  excises
on commodities with relatively  low price and income  elasticity.    Some
additional research is needed to evaluate that resemblance  further.

Offsetting Pre-Existing Distortions

        Repeal  of tax laws which lead to excessive rates of mining and
harvest of productive resource  stocks does not seem to be in the offing.  The
SWC  probably  could be used to offset those tax subsidies.   Whether it would
be an effective  instrument for  that purpose remains to be seen.  Other policy
alternatives  which would  have similar effects  are subsidies to recycling
firms and standards for recycled content.  Those policies have other  serious
disadvantages,  however.

Conservation of Environment and Productive Resources

        The SWC may have some significant conservation advantages relative to
other material  policies and charging  schemes.   Since  it is a direct  tax on
producers, it constitutes a direct incentive for them  to  substitute away  from
waste components of products and particularly virgin material components.
Volume-related  user fees may do this  indirectly, but may also raise littering
and  unauthorized dumping,  both  of which  are very costly not  only in
environmental  but in fiscal terms.   The SWC  also directly induces consumers
to substitute away from waste intensive  products, which no other policy  does.
Finally,  it  encourages recycling  in accordance with market demand  through
price incentives.  One must  be skeptical of  lump sum handouts to  budding
industries  in order to  get them started.   Commercial  capital  and
entrepreneur ship should become  available when  the market is created.

        The advantages  just mentioned may not be particularly great.   Some
imponderable  notional values and  quite difficult-to-estimate empirical
factors  are  involved.   However, whatever positive  conservation and
environmental effect could be ascribed to the  SWC will certainly increase as
the  level of  the charge  increases.  Here  is  another argument, then, for
financing fixed-cost related municipal waste management overhead out  of SWC
revenues.  That approach  not only yields  excess  extra-market benefits, but
also reduces the waste load and overall  economic burden on households.


                                     64

-------
        The intention  here has not  been  to campaign for  the  SWC policy, but
rather to describe  two hopefully diagnostic points of view in analyzing  the
probable benefits of such a  policy.  Those benefits pertain to the public
finance  and  conservation  objectives implicit  in  the SWC proposal.   The
evaluation  method favored by the  RCC and  followed in Section II  above
confronts only the  PF objective.  And it does  that in an incomplete manner,
by leaving aside  the second-best nature of the  other PF instruments.   It
seems  that such an approach  is definitely  too narrow, given the  stated
concerns of the Resource Conservation Committee and of the EPA mandate,  for
that matter.   Hopefully  some of the lines of enquiry mentioned here can be
followed further  in order to produce a truer  assessment of the various
initiatives,  relative to both the PF  and CE objectives.

        This  argument gains force when the SWC and other policies, for  that
matter,  are  looked at  in  the longer view,  rather than merely in terms of
discounted net policy benefits for the next few years.  Are incentives in the
solid  waste  generation and management sector genuinely  askew?  Clearly the
attendant environmental and productive resource allocation costs are rising,
if not accellerating.  At this point, I  think debate should  focus on how to
restructure incentive patterns to create long-term  corrective influences.
Quite  possibly no single  proposal  or mix of  proposals could outweigh its
distortional  and  administrative costs,  but  such  a  judgment cannot  be
supported without examining more thoroughly the disadvantages of the  status
quo»  in terms  of both the PF and CE perspectives, than we  have as yet done.
                                    65

-------
               ECONOMIC APPROACHES TO SOLID WASTE MANAGEMENT:
                            EUROPEAN EXPERIENCE
                                    by


                               W. DAVID  CONN
ACKNOWLEDGEMENTS

        Information contained in this  paper was collected by the author under
contract to the U. S. Environmental Protection Agency.   However, any  views
expressed herein are those of the author alone.  Thanks are due to the many
individuals and organizations in Europe  and the United  States who assisted
with the research.

INTRODUCTION

       This  paper discusses the application  in  Western Europe of  (i)
continuously varying user charges for  household solid waste services  (that
is, charges that vary for each collection according to the quantity of  waste
collected), and (ii) product charges  (that is,  solid  waste-related charges
levied  on products at the  time of manufacture  or  distribution).   The
discussion of user charges  is based on research conducted between May -
September, 1978, while the  information on product  charges was collected
largely  between November,  1976 -  January, 1977  (with some subsequent
up-dating).

USER CHARGES

Earlier Studies

        A 1973  study by  the French consulting  firm CERU[1] compared the
method of financing solid waste services then generally employed in France
with those used in the U.K.,  Germany (Federal Republic),  and Sweden.  The
French  at that time  typically paid  for solid waste  services via  a tax
collected  at the  same time  as  property taxes.  The study identified  a
continuum of  alternative approaches ranging from:

        (i)   the use of local general  revenues (collected via property taxes,
etc-.),  with no direct  connection between the amount  paid by an individual
household and  the cost actually incurred (as in the U.K.);  through

                                   66

-------
       (ii)   the establishment of a charge  that varies according  to  the size
and number  of containers regularly collected from each household (as in many
parts  of  Germany); to

      (iii)  the  establishment of  a  charge that discriminates more  finely
between households according not only to  the  size  and number of containers
but also to the distance  from the storage  area to the street,  the distance
between pick-ups,  the frequency of collection, etc. (as in some metropolitan
areas  in  Sweden).

       The study pointed  out that the use of general  revenues may  be
justified on the grounds  of providing  a  public service advantageous  to all
members of the community, as  well as on the grounds of simplicity  (there
being  no need  to collect  an extra tax or  charge).  On the other  hand, the
study also mentioned three major problems,  resulting from the absence of a
direct link between the payments made  and  the costs incurred, the likelihood
of inadequate financing (due to competition for  scarce funds in the  municipal
budget), and the lack of any incentive  that can be used to encourage  "good"
solid  waste practices on the part of households.

        Solid waste charges were said to  be based on two principles,  namely
that the  total  sum collected should cover  the total cost  of  the service
provided and that the contribution of  each user should equal the cost  of the
service received by that user.  The study argued that the  use of charges
gives  flexibility in maximizing the  efficiency of the service, and  that  it
provides  an incentive for households to  cooperate  (or  be  charged
accordingly) .   A charging system  can  be made more or less elaborate, with
varying costs of administration, accounting,  etc.

       The possibility of establishing charges  that vary for each collection
according to the quantity of waste collected was one variation  of  the
charging approach considered in the report.   The system could employ trash
cans (or  some other kind of reusable receptacles) or disposable bags.

Trash Can System

        In  its most complicated form,  this would require the collection  crew
to make  detailed estimates of the quantities of waste placed out  for
collection, which would  then provide  the  basis for  billing.  The study
rejected  this method owing  to  the difficulties  of  estimating  the quantities,
of keeping a daily record, and of resolving  queries.  A simplification would
be to base the charges just on the number of  containers  (irrespective  of the
quantities contained);  another possibility,  which would avoid the need  for
recordkeeping  or billing, would be  to levy the charge at  the time  of
collection, e.g., using  tokens  attached to  the  containers.  The study  was
unenthusiastic about these methods; instead,  it presented the alternative  of
having households pay  a  regular  subscription for  the right to place out a
pre-specified number of  containers at each collection  (the charges no  longer
varying on a day-to-day basis  with the actual amount  of waste collected).
                                     67

-------
 Bag System

         The study  discussed the possibility of selling  bags  at a price that
 would include  the  costs  of collection and disposal.   The  advantage  would  be
 the avoidance of  record-keeping or billing, but several  disadvantages were
 also identified; these included the fact that the method  ignores differences
 in  the  costs  of collecting and disposing of different  bags (depending on
 their location,  the number picked up at any one location, etc.), the problems
 of requiring users  to go to buy the bags, and the incentives that users would
 have to  dump waste  improperly or to misuse the bags  (e.g.,  by overfilling),
 creating  difficulties for collection.  The study concluded that this method
 has to be  approached with caution.

         A  later (1976) study by the same firm examined solid  waste financing
 options  for the  community of Beaune[2], a small but  growing town (and  center
 for the production of burgundy wines)  in Cote d'Or,  near Dijon.  At the time
 of the study,  the municipal service was collecting waste  from  a population of
 about 19,000.   The existing  system  involved the use of  bags that  were
 distributed regularly (usually annually)  by the collection  service  to all
 houses,  apartment buildings,  etc., the  number of  bags depending on the number
 of people  resident.  The system was financed by a tax which in 1974 covered
 only part of  the cost incurred  but which was increased the following year  to
 cover  the  full cost.  It was felt that  the system was inequitable since the
 tax was the same for each household, while calculations showed that the cost
 varied significantly, depending  on the  number of  people in  the household.

         The study proposed either that  a charge should be levied depending  on
 the number of  people in the household,  or that bags  should be  sold at a price
 covering  the cost  of  the  service.   In connection with  this  second
 alternative, the study identified as problems the fact that arrangements
 would have to be  made for  selling the bags (and commercial  firms might not
 wish to  act, in effect, as revenue-collectors for  the municipality) and the
 fact that the system might  encourage  improper dumping or the placing out  of
 wastes without bags.  On the  other hand, the system would  be  equitable and
 would not require a register of  households  to  be maintained.  The study made
 no conclusive recommendation  as  to which system should be implemented.

         In reality, the community  decided  against using the sale of bags  as
 the means of charging for solid waste services, for fear that people would
 evade  the charge by improper  dumping[3].   Instead, the community is examining
 how a  register of households  might  be established,  so that charges  can  be
 based  on the number of people  living in  a  residence.

 Actual Experience;   Baden-Wurttemberg, Germany

        A system of charging based on containers was introduced  many years
 ago in a part  of the Lande  (State)  of  Baden-Wurttemberg, near  the city  of
 Stuttgart, on the  grounds that  it was an  equitable means  of  pricing[4].
Users indicated that they had  paid for the  solid waste service by  attaching
 tokens to  each (standardized, 35 or 50  litre)  container;  they  had  the  choice
 of  purchasing either  an annual  "stamp" or  individual "strips"  for  each
 collection (the strips  being broken when the  containers  were  emptied).  The

                                     68

-------
tokens were issued by the  local authorities and sold by local banks,  without
any surcharge.

        One problem was  that  the  strips were sometimes mutilated or  detached
prior  to  the collection, e.g., by children.  As a result,  an attachment  to
the strips was  introduced  which could be removed by the householder  and kept
as proof  that the  charge had been  paid.  In practice, containers  without
strips were generally emptied anyway, especially  when it was  apparent that
the strips had  been improperly detached.

        The prices charged for the  tokens were re-calculated  annually based
on an  estimate of likely  expenditures on solid waste services for the coming
year (including separate collection of bulky wastes,  provided at no direct
cost to.householders) and  an  estimate of the number of stamps/strips  expected
to be sold.  The stamp was sold at  a slight discount compared to 52  strips.

        The  system was  used  in the cities of Leonberg, Fellbach, Waiblingen,
Reutlingen, and Tuebingen  but it  was discontinued in all but  a part (about
one-quarter) of Waiblingen because  of two problems, namely:*

        (i)  difficulties  in  estimating the weekly amount of waste collected,
necessary for efficient  planning  of routes, vehicle capacities, etc.;  and

       (ii)  considerable  increases  in the quantities  of  bulky waste and
litter (although the quantities of  household waste placed in the standardized
containers decreased).  Specifically, the frequency of removal of bulky waste
had  to be increased 2  times per  year to 8-10 times per year, signifying  an
estimated increase in bulky waste volume of 400%-500%.  Although the ratio  of
household to bulky waste  generation is known to depend  on the  size  of
containers used  for the  household  waste (the  use of  smaller containers
typically causes an increase  in bulky waste), it  is reported that communities
using the token system generated  much more bulky  waste than other comparable
communities (using the same types and sizes of containers).

        Overall, costs were found to increase.  The proceeds from the sale  of
tokens were insufficient to cover  the costs,  and the  difference had to  be
made up from other sources, e.g., general revenues.  However,  users generally
liked  the system and some  applied  (unsuccessfully) to the courts for  its
continuance.

        It may  be noted that a 1973/74 study[5]  conducted in  the district  of
Kurten in the county of Kreis (a  rural/resort  area) revealed that where
residents were charged for the use  of bags for household  waste, the amount  of
waste collected from each  household was approximately half by  weight  of that
collected from an urban  district  with a normal container  system; however,  the
amount received  at the disposal  site (including household, bulky, and
commercial wastes) was about  the  same (roughly 350 kg per person per year).

1.    It has also been reported that the municipal finance officers  did not
     like  receiving revenue in 52 installments.
                                     69

-------
Netherlands

        A  new Wastes  Act  was passed in  the  Netherlands in  1977.   The
supporting text[6] mentioned that some communities paid for solid waste
services  via charges for bags, and stated (without supplying evidence) that
"such  a system encourages  illicit dumping of  normal and  bulky domestic
refuse. Moreover, the high price of the bags  is an inducement to  stuff  too
much into the bags so that they often  tear".  It is now reported[7] that due
to the  expense and problems caused by illegal  dumping (as experienced  by  the
solid  waste  departments),  less and less communities are using this  system.
Where it is still  in operation,  the price of the  bags does not  typically
cover  the full  costs of  collection and disposal but, rather, an arbitrary
fraction thereof.

Discussion

        It appears that  the use of systems in which charges vary for each
collection according to the amount of waste collected is not  extensive,  and
is currently on the decrease.   While the potential advantages of increased
equity  and efficiency (compared to a flat-rate  system) and  administrative
simplicity (where billing  can be avoided)  are widely recognized, they are
thought to be outweighted by the disadvantages, of which the likelihood of an
increase  in  improper dumping and litter seems  to be viewed as the  most
serious.  Such an  increase has been reported  in communities where  the  system
has been tried, although quantitative data unfortunately seem  to be lacking.

        Interestingly, there seems to have been little discussion  in  Europe
of the potential  use of the system specifically to encourage  waste reduction
or resource recovery; rather, the Europeans  have  been more  concerned  with
establishing an equitable and efficient  method of financing  than  with
reducing the  generation of waste.

        Furthermore, the concern for providing a satisfactory public  service
(with adequate protection of the public from  the hazards caused by improper
dumping and  littering)  is  evidently stronger  than any  concern for waste
reduction. For example,  in  many parts of Germany where  the municipality
decides the  size  of container to be used (and often supplies  it),  there is  a
trend toward  requiring larger containers to  remove any temptation  to dump
improperly[A] :   in addition, the public can take small quantities of waste
directly  to  the disposal  site without charge,  and there  is  always  the
possibility  of placing  wastes in the litter  bins on  street  corners.  In
Denmark, also, solid waste collection and disposal  are viewed primarily as  a
public service[8]; if it  were found that  the  container provided  to each
household were not large enough, its size would  be increased.^   Households
can also take wastes free of charge to  container sites throughout  the  cities.
2.   To accommodate  temporary surpluses of waste, households can  purchase
     bags at  nominal cost.   It  is pointed  out  that the marginal  cost of
     handling  these bags is probably very  low.
                                    70

-------
PRODUCT CHARGES[9]

        At the present  time, Sweden,  Norway, and Finland each have  taxes or
charges on certain beverage containers; the  French government has legislative
authorization to introduce  charges  and is proposing to do so initially on
beverage containers; and  legislation  enacted  in 1977 in  the Netherlands
authorizes the imposition of charges on certain goods.

Nature and Purpose of the Taxes/Charges

        Norway and  Finland introduced taxes for the  specific purpose of
discouraging the  use of non-refillable containers  for beer and soft  drinks.
The Norwegian tax on non-refillables is about  $.15 (plus 20% value-added  tax)
per container and  represents a significant  portion of the price of the
beverage; when the  tax was  first introduced  in 1974 this proportion was about
30%,  but beverage  prices have since risen,  and  it has now dropped  to about
15%.  The Finnish tax on non-ref illables, introduced in  1976, represents  a
smaller, though still significant proportion of  the beverage  price; the tax
is about $.08 on  beer containers (some 6% of the  price of the  beer) and  $.16
on soft drink containers (some 13% of the price of the drink).

        The  original  purpose of  the  Swedish tax on  containers  for
ready-to-serve  drinks between 20 cl and 3  litres, first introduced in 1973,
was to finance a  freeze on certain food prices rather  than to  discourage  the
use of any particular kinds of containers.  The  tax  is  small  (slightly more
than $.02) and it is added to the cost of non-refillables, whereas it becomes
part of the refundable deposit on refillables  (i.e.,  it  is paid on the latter
only when they are  not returned).

        In France a proposal  is being developed to impose a  charge (under the
existing legislation) on all  containers ranging from  10  ml  to  3 litres  for
wine,  beer,  soft  drinks, fruit juices, mineral water,  and  possibly milk
(although milk may be excluded for political reasons  since  it  is currently
subsidized).  The  charge would be 5-10 centimes (about $.01-$.02) per unit
and would be  intended  to reflect  the approximate  cost  of  handling and
disposing of the container once discarded.   The  tax  would probably be levied
when the containers are  manufactured  or assembled;  the revenues  would  be
collected by the Ministry of Finance  and then disbursed  to  local authorities
for solid waste projects  including  litter  cleanup,  demonstration  trials,
etc.,  as well as  to  industry (possibly)  to finance the development of low
waste technologies.  The Ministry preparing the  proposal  is  taking  a
pragmatic approach to  the development of  the charging  scheme, keeping it as
simple as possible; for  example, no  attempt would be  made to "fine tune" the
charge according  to  the   disposability or other characteristics of a given
container (except possibly that a distinction might be made  between  standard
and non-standard  containers).  No reduction or exemption would be granted for
the use  of secondary materials  in a  container; for while this  would  almost
certainly increase the level of resource  recovery from the waste stream, it
would  tend to lessen the effectiveness of the charges in achieving their
primary  objective of reducing the rate of generation of  wastes.
                                     71

-------
         In the Wastes Act enacted during 1977  by  the Netherlands parliament,
 there is  provision for  the  introduction  of  "levies"  both to limit the
 generation of waste and to provide revenue to carry out the provisions of the
 Act.   The  levies may be imposed on those  who manufacture,  import, hold  in
 stock for sale,  offer for sale, sell,  deliver or use certain goods that are
 difficult  to  reuse, difficult to handle,  cause  an excessive  increase  in
 waste, or  are likely to be littered.

         However, it is now reported[7]  that  implementation of this provision
 by the  present government  is most unlikely, especially while problems  of
 unemployment persist.  The  provision  is considered  too  expensive  to
 implement,  and  potentially too unpopular to  be  politically acceptable.

 Effects  of the  Taxes/Charges

         The Norwegian tax has been effective in preventing any growth in the
 use of non-refillable containers for  beer  and soft drinks.  Although domestic
 bottlers have traditionally used refillables, beer imported from Denmark and
 Sweden  was previously packaged in cans; now most imported beer is brought  to
 Norway for packaging in refillable bottles.  In Finland, also, there is very
 little canned beer.

         In Sweden, on the other hand, while  there  has been  a transition  in
 the  bottle sector from non-refillables  to refillables,  there was at first  no
 change and subsequently an increase in  the use  of cans for beer (despite the
 fact  that  consumers must pay  about twice  as much for the same amount of beer
 packed in  a can rather than a  bottle).   Other effects of the  tax in Sweden,
 as  reported by a  government-appointed  committee,  include a reduction in the
 production  of all bottles (due mainly to an  increase in the number  of  trips
 made by  each bottle, as well as to a shift to the use of larger bottle sizes)
 and various attempts to avoid  the tax  (e.g.,  by  introducing new  packages
 containing volumes outside the dutiable limits  and by substituting  undiluted
 drinks for  ready-to-serve drinks).  However, the  committee  also concluded
 that  the  tax has  had  no significant  effects on  costs,  employment,  or the
 environment.

 Discussion

         The taxes/charges have been applied  differently  according  to the
 objective(s) intended.   For example, it seems that  the primary objective of
 the current French proposal for product charges on beverage containers  is  to
 improve market efficiency by internalizing in the producer's decisions the
 subsequent costs  of handling/disposal (so that he has  an incentive to take
 these costs into account).  The monetary amounts involved  are  small  (relative
 to the beverage costs)  and so the  charges are not  expected to have   major
impact on the  physical  characteristics  of the  waste stream.   The charges
will,  however, raise  revenue  that  can be used to  finance  solid waste
services.  On  the other hand,  in Norway and Finland the  objective  has been
specifically to  discourage the  use of non-refillable beverage  containers, and
therefore the  latter have been  relatively highly taxed.
                                    72

-------
        In Sweden,  the present duty  on beverage containers was imposed not
for waste reduction  purposes but instead for  an  entirely unrelated reason (to
subsidize a price  freeze on certain foods);  the amount is relatively  small,
and while there  has  been some shift from non-refillable bottles, the overall
impact on  the economy  and  on the environment  has been  reported  to be
insignificant.   An increase in the duty is currently being proposed, with the
intention of  raising revenue to finance various  anti-litter activities.

        It is important  to  establish  with  what objective  taxes or  other
fiscal  mechanisms  might  be  introduced  in  the United States.  The product
charging scheme  that is being studied by the  Resource Conservation Committee
(as  specified in the Resource Conservation and  Recovery Act of 1976) appears
to have the primary  objective of internalizing solid waste management costs,
and  therefore seems most closely related  to  the French proposal. Careful
attention should therefore be  paid to  developments if/when the latter is
implemented.  It should  be  noted that  the charges currently proposed in
France would  apply only to beverage containers (whereas a U.S.  scheme  might
apply to all  consumer products), and that the French would grant no exemption
for the use of secondary materials (on the grounds that this might encourage
resource recovery at the  expense of efforts to reduce  the rate at  which
resources first  enter the waste stream).  One of the most  striking features
of the  French approach is that every effort  is  being made to keep the  scheme
as simple as  possible; rather than placing too much reliance on studies that
seek  to determine  the theoretically  "correct" level of charges, etc., the
authorities are  prepared to make reasonable assumptions and  in general are
adopting a pragmatic line.

        A problem encountered in several European countries,  and apparently
common  to the U.S.  also,  is that the use of  taxes or other fiscal mechanisms
for waste reduction purposes  tends to be viewed with great  suspicion by
finance  ministries  and policy-makers  in general.  It  seems that they
frequently have  difficulty in understanding that fiscal measures may be used
not  only  for raising revenue but also for incentive purposes.  Furthermore,
finance ministries on the  whole prefer to keep revenues  and  expenditures
separate; they generally dislike funds  to be "ear-marked" for particular
activities (such as  solid waste projects)  over  which they  may have little
control.   An educational effort is  required  to ensure  that those  making
decisions at  least have a  reasonable understanding of the  advantages and
disadvantages of fiscal approaches.
                                    73

-------
                                REFERENCES
1.   CERU,  Tarification du Service de Collecte et de Traitement des Ordures
     Menageres,  Prepared for Ministere de  la Protection de la Nature et de
     1'Environnement, Paris, 1973.

2.   CERU,  Elimination des Ordures Menageres, Prepared for Ministere de la
     Qualite de  la Vie - Environnement,  Paris, 1976.

3.   Montangerand, C., Le  Secretaire General,  Mairie de Beaune,  Personal
     Communication, 1978.

4.   Schmitt-Tegge, J., Goosman,  G.,  Orlich, J., and Wicke, Umweldbundesaint,
     Berlin, Personal Communication,  1978.

5.   I.G.T., Investigation of Waste Arisings in Rheinisch-Bergischer Kreis,
     1974/75 (quoted by Goosman,  G.,  Umweldbundesamt, Berlin).

6.   Netherlands, Regulations  Concerning Domestic Refuse, Derelict Motor
     Vehicles,  and Other  Categories of Wastes  (Wastes Act), Explanatory
     Memorandum,  Session 1974-1975.

7.   Erasmus, J. H. ,  Waste Management Policy Division,  Ministerie  van
     Volkgezondheid  en Milieuhygiene,  Leidschendam, Personal  Communication,
     1978.

8.   Welinder, A. S., Miljstyrelsen, Copenhagen,  Personal Communication,
     1978.

9.   Conn,  W. D., Waste Reduction Activities in Europe, Prepared  for Office
     of  Solid Waste, U.S. Environmental  Protection Agency,  Washington,  1977.
                                    74

-------
                 THE ADMINISTRATIVE COSTS OF USER CHARGES



                                   by


                            BARBARA J. STEVENS
INTRODUCTION

        Since  the passage of  the  resource Conservation  and Recovery Act
increasing attention has  been  focused on the problem of solid waste
management.! More and more officials and citizens are  becoming aware that
solid  waste  disposal is not a  simple matter, and that not only is the
availability of sites for safe disposal of wastes rapidly decreasing, but
that also  the costs of such  disposal are rapidly  increasing.  Responses  to
the problems of finite resources and finite numbers  of suitable locations  for
the disposal  of  the residuals from these resources vary.   For example,  some
groups  attempt to encourage the recycling of used goods  by  instituting source
separation programs for  paper.   Others attempt  to develop procedures  to
create  useful  products from municipal waste, such as  fuel or energy.  A third
group  agrees  that a market  failure is at least in part responsible for  the
solid waste problem.  Correction of this market  failure,  this group would
argue,  would do much to help solve  the  problem of increasing waste generation
coupled with rising costs of disposal.   It is the argument of this latter
group which is addressed here.

        Economists concerned about the market failure in solid waste
collection generally cite  the  fact that generators  of solid waste  —
especially  households — do not face a  price vector  for  the collection  of
such wastes which truly reflects  the  costs of solid  waste handling.  In  the
absence  of such  a price  vector  consumers' decisions are, in  general,
suboptimal. The  lack of a true price for solid waste  collection and disposal
signals that in some sense society  as a whole may not be  as well  off as  it
could be.

        Whether or not this loss  in social welfare can  be  remedied  through
the  introduction of a properly  specified price vector for solid waste
collection  and disposal services  is only partly addressed  here.  The entire
answer depends  upon the relative  magnitude of the  gain in. social welfare to
 1.   Resource Conservation and Recovery Act of 1976;  (IL 94-58)


                                    75

-------
be expected from institution of a price  vector reflecting  accurately  the
costs  of solid waste collection and disposal  services versus the opportunity
cost of the resources used to implement and administer such a program.  This
paper  attempts to  quantify the resources  which  would be  required  to
administer a  user  charge system for solid waste  disposal service.  The next
section describes  the economic  framework in which society as a whole  would
benefit  from  initiation of a user  charge system.  While social welfare is
most likely to increase when the  type  of user charge  implemented  is  a
marginal cost based system, practical considerations make implementation of
such a system most difficult. Thus, Section III presents empirical evidence
about  the costs  of administering various types of user charge systems.  The
final section of this paper presents some guidelines  about when and how user
charges  can and should be initiated and summarizes the empirical information
presented earlier.

ECONOMIC FRAMEWORK

        Goods prices do not include waste handling prices, and waste handling
prices do not,  in  general, vary with marginal  changes in waste generation.

        One of  the "problems" of solid waste management stems from the fact
that the costs of  waste disposal  are  not included in product prices when
consumers make  their purchasing decisions.  When households  are faced with
prices  which do  not vary  with decisions made at the margin, the optimizing
process is intercepted.

        Most  households in  the United  States receive refuse collection
services from a public agency, and the most common means of financing  those
services is from  general revenues.   Thus, the household is typically unaware
that it pays about $35 per year for refuse collection and disposal.  Even  if
the household is aware  of the costs of refuse collection, it is not able to
impact  its own charges  through individual  behavior;  there is  thus no
pecuniary reason for a household to cut back on refuse generation.

        Pricing schemes for users of refuse collection services fall into  two
broad  categories:  those where the marginal  cost is zero and those where it
is positive. All  methods of financing refuse  collection can  have an impact
on refuse generation; however, those  user  charge systems  with a marginal
price of  zero will affect production of waste  only through the income effect.
Such  systems  pother charge  the household  a flat fee or  finance refuse
collection from general  revenues.  Clearly, the higher the  fee (or cost
financed  from general revenues) the  smaller the disposable income of  the
2.   Savas and  Stevens, "The Cost of Residential Refuse  Collection and  the
     Effect of  Service Arrangement," Municipal  Year Book  1977.

3.   The  Cost  per ton  for  refuse collection average  about $20; using  a
     household  generation rate of 1.7 tons,  the household pays an implicit or
     explicit fee  of about $35 per year.   See Savas and Stevens, ibid.


                                     76

-------
household  and  (assuming that refuse  generation increases  with income) the
smaller is  the quantity of  refuse generated.4  Systems  where the marginal
price is positive should impact  the  volume of refuse generated through both a
substitution  and an income  effect.  Such pricing mechanisms have the very
desirable  attribute of causing households to consider  the  implicit cost  of
disposing  of the refuse component  of a good, allowing households to trade off
goods with respect to their total  cost—purchase price plus implicit disposal
cost.

       While  the theoretical  superiority of user charge  systems with a
positive  marginal price are quite  clear, they present problems in that they
are difficult  and costly to  implement.  In order to affect the marginal
consumption and disposal generating decision of the consumer it is necessary
for the. user charge plan to contain  a marginal  price for refuse generated.
In  fact, a  true marginal  pricing scheme does not exist.   The  closest
approximation  to  such a marginal price  system is the one in which households
are  charged  by the can or container of refuse  presented for collection.
Since the  contents of a container  of refuse  can vary widely,  even this  system
does  not  truly present the  consumer with a valid marginal price vector  to
impact on his  marginal consumption decision.

       In short, economic theory would indicate that  institution of  a
pricing scheme or user charge system for refuse collection which  truly
charged consumers for the disposal of each marginal item of refuse generated
would (a) cause a reduction in  waste generation due to both the income and
the substitution effects and (b) result in an increase in social welfare  to
the extent that  foregone use of  resources for waste disposal  and the value  of
goods not  disposed of exceeds the  value of resources used  in administering
the program.    Administrtion  costs are defined to include the  costs  of
billing users of the solid  waste  service, collecting  from delinquent
accounts,  enforcing the  system, and handling of complaints  arising  solely
from  the  user charge system.   To  be complete, one  should state  that
administrative costs should be  net of any changes in productivity arising
from the  imposition of the  system,  as, for example, might occur  were the
system that  of metered bags.5 In the following discussion, however,  no

4.   The value of resources used in  administering  the program can be  defined
     as equal  to not only  straightforward administrative cost  but  also costs
     associated  with other  activities  which may be reinforced via the
     imposition of the user charge system.  For example, littering may
     increase  when user charges are  unpaid.  As the cost of litter  collection
     exceeds  the  cost of  regular solid waste collection,  this may represent
     (if  the  increase in the propensity to  litter is great enough)  a real
     drain on  resources.   See McFarland, et.al.  Comprehensive Studies  on
     Solid Waste Management,  Final Report, Sanitary Engineering  Research
     Laboratory Report 72-3, University of California, Berkeley, 1972.

5.   Hudson, Issues  in Pricing  Solid Waste Management  Urban Systems Research
      and  Engineering, Inc., Cambridge,  Massachusetts 02138, #68-01-4362  for
     USEPA,  June, 1978 finds  significant savings  in costs  due to productivity
      increases associated with  initiation of  a metered bag or other user
     charge system.

                                     77

-------
 attempt will be made to  account for anything  other than the straightforward
 administrative costs of  a user charge system.

        From a pure,  maximize-social-welfare-point-of-view ,  then
 quantification of  several factors is an  essential first  step in evaluating
 whether a user charge system is desirable.  Assuming that one institutes a
 marginal based system  in  a market where prices  for refuse collection are
 already present, one first needs to know the price elasticity of  refuse
 generation.  From  this information,  together with known data on current
 refuse generation  patterns," it is possible  to calculate the quantity of
 resources which would  be  freed from entry into the disposal stream and thus
 saved for other, presumably superior, uses  by  society.  Second, the  quantity
 of refuse  going  into  recycling patterns  should be determined.   These
 quantities would already be included in the  first  category as refuse  not
 going to the final  disposal  site; they  should be considered as resources
 newly being  directed -to  an optimal use,  as a direct  result  of the institution
 of the  new  pricing  scheme.   These two factors represent the benefits to
 society of the user  charge  system.  No attempt  is made here to value  this
 latter  benefit stream.  In order to estimate net benefits to society, it is
 of course necessary to  know  the true marginal  social  cost of  refuse
 collection and disposal.  Net costs are  the costs of  administering the system
 less the net costs foregone for refuse  not generated  and not collected.

        While  this  discussion  is  straightforward  and presents  no  new
 theoretical  insight,  it does highlight the basis on which the user  charge
 system  can  best be evaluated.  First,  how large are the benefits to society
 from institution of  this system?  Where  the system is one in which a  flat or
 lumpy user  charge system  is  instituted instead of a true marginal  cost
 pricing  scheme, it is possible that there might be  no  reduction in  refuse
 generation  and,  thus,  no social  benefit.  Without  any  attempt being made to
 evaluate the benefits from reduction of  the consumption stream, we can  say
 that  a  user charge system is desirable where the  value  of resources freed
 from refuse  collection and disposal activities  is at least as  great  as  the
 cost  of  administering  the  program.  An'attempt at  this final calculation is
 presented in the next sections.


EMPIRICAL  EVIDENCE

        Two   approaches will be  used  to  evaluate the magnitude of
administrative costs for  user fees  relative to  the   cost  savings  resulting
 from  initiation of  such fees.  First, two cities' experience with  such fees
     Several studies have indicated  that refuse generation increases with
     income.   See  Center for Urban  Studies (A Condensation of Battelle
     Columbus  Laboratories)  Some Economic  Factors Affecting Demand  for
     Municipal Collection of Household Refuse, University of Chicago,  August,
     1973, 47 pp., J. McFarland, et. al. Comprehensive Studies of Solid Waste
     Management,  Final Report Sanitary Engineering Research Laboratory,
     College of Engineering and School of Public Health SERL Report No. 72-3.
     Berkeley,  California, University of California,  (May, 1972).

                                   78

-------
will be presented.  Each has  a user charge  system which is variable by
quantity  of  refuse collected.   The cities differ dramatically in how the
charges are administered and  consequently, in  costs.  The low cost  city
illustrates a "best case" for the administrative cost of a user charge system
while  the  contrast between  the two cities highlights procedures to  be
followed  in  instituting an efficiently  administered system.   Second, overall
data is presented for thirty-eight cities located throughout the United
States  that  have municipal refuse collection service and bill households for
this service.   These average  administrative  cost data are compared  with
refuse  generation and  the  costs of service delivery to obtain average cost
figures for administering a user charge  system.

        Having estimated  the administrative costs of user charges on a best
case and an average case basis, these data are related to best case and
average case estimates of the cost savings to be expected from imposition of
the user charge system.  These data are  drawn from  a 1974' survey as  well as
other  engineering and economic estimates.  Together, the probable impact of
user charges  can  then be evaluated.

Best Case-Worst Case

        The two cities  selected  to  illustrate the characteristics  of an
efficiently designed user charge system  are both  located in the Western part
of the  United States.8  Each was paired  with another city as part of a  series
of case studies comparing cities whose costs differed but which were  alike in
refuse  collection  service delivery, scale and location.  The matched cities'
respective residuals from  an  estimated cost  equation were at  least two
standard  deviations apart,  with one city's residual being positive and the
other's negative.   One city selected for discussion here was the  low cost
outlier of the matched pair located in the Mountatin Region while the  other
was the high cost outlier  from the matched  pair located in  Northern
California.  The  high cost outlier, Norcahigh, is,  not surprisingly,  here the
city with the high  costs for administration of the  user charge system.  The
low cost outlier, Mountainlow, is associated here with an extremely efficient
low cost system for administering a user charge  system.  The costs  of refuse
collection and administration are greater in Norcahigh than in Mountainlow
for several  reasons, one of  which is different ways responsibility for
billing and collection of fees is allocated in the  two cities.

        Norcahigh has a very complex way of administering their variable fee
user charge  system.  Norcahigh has a contract with WCF for the collection of
residential refuse.  WCF is a national firm that  has been in business in this
city since the mid-1970's.  Residential  service in Norcahigh is once  a week
curbside at the base rate of  $3.11 per month  per household for  two  cans.
Additional cans are collected at  $1.00 per month per can.  A  resident may
also pay a higher fee for backyard service.

7.   See Savas,  ed.,  The Organization and Efficiency  of Solid  Waste
     Collection,  Lexington Books, Massachusetts,  1977.

8.   Stevens,  Norcahigh-Norcalow available from Ecodata,  1700 Broadway, and
     Mountainhigh-Mountainlow.

                                    79

-------
         The Office of Property Code Enforcement  in the City  Government is
 specifically  empowered  to  enforce contract  provisions.   It  deals with
 delinquent accounts and unanswered complaints,  and is the liaison with the
 health department for matters such as vehicle inspection.  This office  is run
 by a deputy health officer; twenty-one of  its employees  spend  part of their
 time monitoring the refuse collection contract.  In the last  quarter of 1976,
 1,000 liens against property were issued  due  to  delinquency  in  payment of
 refuse bills.  The large numer is no doubt the result of a notably cumbersome
 billing procedure.

         WCF  must act  as a  bill  collection  agency as well as a refuse
 collection firm.  More than ten full-time  employees work at WCF to update the
 list of accounts to be billed  for refuse collection service.  Owners of
 property are supposed to institute services themselves, but,  in fact, the
 firm generally must itself  identify new residents in order  for billing to
 begin.   Once a resident has been  billed, the firm's collection problems have
 only  begun.   If a  resident  does not pay, very specific procedures are
 followed.  First,  WCF sends a delinquency  note.  If there is no response to
 the  delinquency note, after sixty days, WCF sends a second delinquency  notice
 and  simultaneously files a copy with the health  officer indicating who has
 not  paid and how much is due.   If the  bill remains delinquent  for seventy
 five days, WCF gives the bill to  the city for  lien and special  assessment
 proceedings.  After twenty  days, the health  officer can send the owner
 notificaion that he owes the  bill plus  ten percent charge and  also runs the
 risk of incurring a lien on  his  property  for the delinquent  amount.  At that
 time,  lien proceedings are begun  and the  health  officer notifies the city
 council  which in turn posts a time for a hearing on the proposed lien.   After
 the  hearing, if  the city council decides the  lien is justified,  it  is
 established as the  greater  of  $25 or ten percent of the refuse collection
 service charge.   The lien is  recorded,  the owner is notified  officially, and
 proceedings to have the sum of  service,  delinquency charges,  administrative
 and  assessment charges specially  assessed  against the owner's property  begun.
 This occurs if the  lien is  still not paid.   If the special assessment is
 approved upon hearing by the  city council, it becomes part of taxes payable.

         Two private firms,  Oaks & Sons and MoLo, service residential  accounts
 in Mountainlow.   Each has an  exclusive area and is franchised to service all
 households in the area.   MoLo collects mostly from recently  annexed  areas of
 MoLo county,  while Oakes & Sons picks up from the majority of  households in
 Mountainlow.   Oakes  & Sons has been  in business for thirty-one years.   For a
 fee, households  have  a choice of  location  (curb or carry out) and  quantity of
 refuse  to be  collected.   In Mountainlow, the base fees of $2.25 per month for
 one  can, $2.75 for two cans,  $3.10 for three cans,  and $3.45  for four cans
 also include  costs of billing  and  disposal.  Unlike in Norcahigh,  however, in
 Mountainlow these  two functions are performed by the city government,  which
 takes 21%  of gross  fees to pay for billing costs,  disposal expenses and
 franchise fees.  In both cities,  yard trash is collected along with  regular
 commercial customers on residential routes, although the fees  for commercial
 establishments generally differ from those for households.

        Mountainlow's City  Attorney and the Council take  responsibility for
writing franchise  agreements,  for  selecting the refuse collection  contractor,

                                     80

-------
and for determining  annual charges.   In  addition to these  functions, this
city has a separate  complaint and billing department, employing an equivalent
of four full time persons  to provide  the  complaint handling  and billing
services for refuse  collection.

        Mountainlow's  City Attorney and the  Council  take responsibility for
writing franchise  agreements, for selecting  the  refuse collection contractor,
and for determining  annual charges.   In  addition to these  functions, this
city has a separate  complaint and billing department, employing an equivalent
of four full time persons  to provide  the  complaint handling  and billing
services for refuse  collection.

        Mountainlow  operates a computerized  billing  system for Oakes & Sons,
employing two full time equivalent persons in  that function.  Residents are
divided into groups, each  of which receives quarterly bills on different
months.   The city  receives both weekly and monthly status reports on all
accounts.  Customers wishing additional service  on an occasional basis attach
a coupon (obtained from Oakes & Sons)  to each extra  can or bag of refuse; the
loader remits this coupon to the city,  and  the  resident  is billed for the
additional service on  the  next regular mailing.   The city provides Oakes  &
Sons, at no charge,  with a monthly printout  of the  status of its accounts.
Oakes  &  Sons must  pay  the  city for  additional computer  services, such as
mailing lists.

        The city remits a monthly check to Oakes & Sons.  The amount the  firm
receives is gross  receipts less bad checks,  a 3% franchise fee, a 6% billing
fee,  and a 12% landfill  fee.  The  city's concern with low cost service is
evidenced by a  current controversy about the amount  charged  the city by the
county for use  of  its  landfill.  Since 1973, Mountainlow has claimed that the
county is charging too high a landfill fee (i.e.,  12% of receipts), since  its
charge  to individual  residents is much lower.  The  city has withheld half of
the  12%  it owes,  holding the  6%  in a  trust  account.   Despite  the
disagreement,  Oakes  &  Sons in effect pays 12% of gross receipts for disposal,
and, netting out all charges, receives  (in 1976)  $1.78  per household per
month for refuse collection services.

        Mountainlow  has a city operated  complaint  handling facility.  Two
persons  work full  time  receiving complaints and monitoring their handling.
The city opened its  complaint and information center in 1974.  It is staffed
by a  former public relations consultant, and his assistant.  Residents
usually call Oakes & Sons with a complaint.   If  the  problem  is not rectified,
they  then call the  city.  When the complaint office receives a call, a staff
member writes it up,  trying to record  the complaint in  the caller's own
words.   The office personnel will  try by tactful, questioning to discover
whether the resident has observed all the "hard  card" rules printed up and
distributed by  Oakes &  Sons.  Complaints are  followed up by the city
sanitarian, or  the police department if the complaint involves a tip-over by
dogs.

        There is a regular  daily procedure for handling  complaints.  The
sanitarian comes to  the complaint office each  day  at 1:30  P.M.  He collects
all the complaint  slips and investigates by contacting either  the resident  or

                                     81

-------
 the private contractor.  Attempts  are made to resolve complaints the same day
 they are  received.  This is of  great concern to the city.

        The main complaint is missed collections or  litter around the  can.
 Complaints  increase during the  summer months,  as  corroborated by the
 complaint officer.  Complaints  are categorized by number and type.  Indeed,
 there is even a list of "chronic  complainers".  As is often the case,  missed
 collections are the most common problems, with lids  off the can the second
 most  frequent complaint.

        Mountainlow is involved  in  the service  on several  levels.   On  a
 day-to-day basis  the city monitors  the quality of the  service through its
 complaint system.  On a monthly basis it monitors the financial activity of
 the firm through its billing system.  Lastly, the city  has very specifically
 defined in its ordinances the parameters of the service  it  wants and the fair
 price that it will pay.  Approximately four full time equivalent municipal
 employees are devoted to the  refuse collection service.  These employees are
 paid  from the franchise and billing fees charged  to the  contractor.

        Table  1  presents a  summary of the  best-worst  case history of
 administrative  costs for these  variable fee user charge systems.   Both
 Norcahigh and  Mountainlow have  quite similar  collection costs per
 household—$25.29  and $23.40 respectively.  Due  to differences in generation
 rates in  the two cities, the  cost  per ton collected in Norcahigh exceeds the
 $13.47  amount achieved in Mountainlow by 35%.   The  different administrative
 procedures followed by the  two  cities  for billing results in an even  more
 striking  cost difference.  Mountainlow's  cost  per household per year for
 billing is but $1.64,  while that for Norcahigh is  $4.48—over four times more
 than its comparison city.

        While administration costs of a user  charge system  are defined  here
 to  refer mostly  to  the  expenses associated with  billing,  some extra
 complaints may arise due to the charge system, and  these expenses should also
 be  included.  For  example, households coulfl complain about inaccurate billing,
 or  about service so unsatisfactory as to  justify non-payment of invoices.
 While there is  no  evidence  to indicate  that the type of complaint or
 frequency of  complaint received in these  two  cities  is any different  from
 that received  in cities  without user charges,  the costs  of complaint handling
 are presented  in  Table 1.  Including these costs  as  part of administrative
 costs of  user charges  would result in  an  overestimate of the magnitude of
 that cost.  As  shown on  Table 1, the complaint handling costs in Norcahigh
 averaged  5.1% of  the annual per household  collection costs  while the  same
 function in Mountainlow  averaged at 8.6% of the annual collection costs per
 household. Together,  then, the administrative costs  of billing and complaint
 handling  amount to 22.9%  of collection  costs  in  Norcahigh—$5.97  per
 household per year—and 16.5% of collection  costs in Mountainlow—$3.86 per
 household  per  year.  Excluding  complaint  handling  reduces the  annual per
 household  costs  for  administration to $4.48 and $1.64 respectively.

        It is clear  that the cumbersome  procedures adopted  by  Norcahigh for
billing result directly in very high costs.   The smoother-working  procedures
in Mountainlow contrast  sharply with  those  in  Norcahigh.   Billing  by the  same

                                    82

-------
TABLE 1.  REFUSE COLLECTION COSTS: COMPARISON OF NORCAHIGH
	AND MOUNTAINLOW	

  Cost measure	Norcahigh	Mountainlow

  1976 costs

  Total cost                              $3,160,978             $468,000

  Number of households served               125,000               20,000

  Cost/ton                                  $18.25                $13.47

  Cost/cubic yard                           $ 4.44                $  3.36

  Cost/household                            $25.29                $23.40


  Billing data

  Billing personnel (full time
    equivalent)

    City                                      10                       2

    Firm                                      10                       2

  Billing costs as % of cost/household         17.8%                    7.9%

  Cost/bill                                  $ 1.12                    $.41


  Complaint data

  Complaint personnel (full time
    equivalent)

    City                                        4                       2

    Firm                                        2                       .5

  Complaint cost as % of cost/household       5.1%                    8.6%


  *It  should be  stressed  that  the fees charged by both these firms are at
  least 25% less than the expected costs for a municipal  agency delivering the
  same services  to its residents as  do these firms to the residents of these
  cities.  All costs are exclusive of disposal fees and percentages  are as
  percent  of collections costs only.
                                      83

-------
  agency which has responsibility for termination of service for non payment
  seem highly desirable, as does joint billing for  several  services.  Billing
  for  utilities and refuse collection at the  same  time makes record keeping  of
  new  arrivals in the city  much easier  than  when refuse collection is
  separately billed.  New arrivals must have  utilities officially connected  in
  order to start receiving service whereas this is  not necessarily required for
  waste  handling—collectors often service refuse containers set out  at the
  proper time and  place.  In short, joint billing  with utilities facilities
  accurate billings of all service recipients,  decreases the deadbeat  problem,
  and results in lower administrative costs.

  Survey Results

         In a 1974 survey of 101  cities providing municipal refuse collection
 services to  residents, cost  data was carefully  collected on site  using
 consistent cost categorizations.° One of  the categories for which  data was
 obtained  was  billing costs in cases where cities had  user charge systems.
 Thirty-eight  of  the 101 cities surveyed had  user charge systems  of one  type
 or another:   five  based  their  charge on  quantity collected  while the
 remainder  charged  a flat fee or  variable  fee  based on the  frequency or
 location of  pickup  service  selected  by the recipient households.   Figure 1
 and Table 2 display the data  collected  in the survey.

        The data show little relationship  between  billing costs and  city
 size, though the variance in billing cost as a percentage of the  total costs
 of  collection does decrease with  increases in city  size.  Several  other
 factors were  calculated  from these  survey data.  The  average  number of
 households  served in  the cities with user charges studied here  was  14,513;
 the market ranged from  under a thousand  to almost  ninety thousand.  The
 average cost per city  for  billing was  $12,000.  This  amounted to  about
 eighty-six cents per household per year for billing expenditures—the range
 was from  a disconcertingly low three cents per household  per  year to  $2.75
 per household per year.  Of the thirty-eight cities in the sample, 43% used
 joint billing.  Contrary to what one might believe, those  cities which  used
 joint billing had higher costs attributable to residental refuse collection
 bills than did those cities which billed refuse collection  separately.   This
 result  is  so counterintuitive that one must conclude only that  cities  where
more than one department pays for billing are more likely to  keep accurate
 records of the  true costs  of billing and  to  enable accurate separation of
 these  costs than  do cities where billing occurs  within a single  department.
Very  often, where billing  occurs  within a  single department,  in collecting
the cost data, interviewers would be willing to lump together  some of  the
 true  costs of billing (for  example,  the cost of  a clerk's time) in  other
categories rather than separating  them  out, so  long as the total  cost  figures
for refuse collection service delivery were  accurate.  This readily explains
why the  cost per  household for joint billing  might exceed  those  for single
billing.
9.   See Stevens, "The Cost of Residential Refuse Collection" in Savas,  ed.,
     op.  cit. for a description of the  survey.
                                     84

-------
Billing Cost as % of
Total Co  ection Cost
    6 —
        3  X
         O
     Key:

x = joint billing

o = single bill
                      10      15      20         25       30

                          Number of Households (Thousands)
             35      40      45 80
                   Figure  1.   Billing costs  by city  size.
                                          85

-------
TABLE 2.  USER CHARGES AND ADMINISTRATIVE COSTS:  1974 DATA

Households served
Service level
% of households with Ix/week service
% of households with curb or alley
service
Monthly charge per household
Billing' costs:
Average cost/city
% of cities with joint billing
Average billing cost per household
Average billing cost per household
for cities with joint billing (?8)
Billing cost/Total Collection Cost
Billing cost/Total Collection Cost
for city with joint billing
Mean
14,513

30%

67%
$2.70

$12,506
43%
$.86

$1.28
3.18%

3.80%
Minimum
762

0

0
$1.00

$50.00
-
.03

.11
1%

1%
Maximum
88,143

100

100
$6.00

$88,312
—
$2.75

$3.75
6.83%

8.33%
                                86

-------
        Overall,  the  average billing cost as a percentage  of  collection cost
was  3.18%.   In cities  with joint  billing an average o'f 3.8% of collection
costs were attributable  solely to the administrative expenses of billing for
refuse collection  services.  As this number seems more reliable than that for
single billing  of  refuse collection  services, we can take  the 3.8% figure as
a low estimate  of  the  administrative cost of user charges.

        Although the  3.8%  figure may  indicate  a  reasonable: level  for the
expenses of instituting  a user charge system, it should be remembered that
the administrative cost  of  instituting a quantity  based user charge  system
might be much greater  than  that figure.  This would be so  because charging on
the basis of quantity requires more  interaction  between the billing  system
and  the refuse collector himself.   When the refuse collector himself must be
involved with determining the eligibility of a household's presented  refuse
for  collection at each stop, work  productivity is probably  decreased,  which
may in effect increase the  administrative cost of a system more than  with a
non-quantity based user  charge system.

        While not  as  detailed as one might like, these data do  represent some
of the  best information  on the actual costs of administering a user  charge
system.  Further,  detailed  study on  these matters would  of course  tend to
increase the assurance with which policy decisions could be made.

CONCLUSIONS AND POLICY IMPLICATIONS

        The empirical  evidence has shown that the administrative cost  of user
charge system are  on  average likely  to be at least 3.8% of the  total  cost of
residential refuse  collection.  Indeed, for two cities  where billing costs
were collected  with  great care and detail,  it was  found  that the
administrative cost  of billing ranged from a low of 7.9% to a high of 17.8%
of collection and  disposal  cost per  household.

        The  3.8%  figure is representative of costs where billing is mostly
done on a non-quantity basis and by  public agencies billing services  jointly.
As such, it represents the low end  of the administrative  cost  scale. The
7.9% figure is  consistent with a quantity based user charge system  where the
billing is performed  by a  public agency billing jointly for  several  services
at one time.  The  17.8% figure, the  high end  of the administrative cost
scale,  is obtained  when  a private firm bills individually  for a quantity
based service.

        These three data points can  be used to derive estimates of  the likely
impact of imposition  of  user charge  systems on social welfare.  Two scenarios
for imposition  of  the  quantity based charge will be considered:

        1)  imposition of a quantity based charge over an  existing  flat fee
user charge system, and

        2)   imposition of the  quantity based  user charge system  over  a
service financed  through general tax revenues.
                                     87

-------
 Both of these systems are  hypothesized to impact household  refuse generation
 and the costs of refuse  collection in similar fashion.   Financing the service
 through property taxes as  compared  to  a flat fee user charge differently
 affects aggregate refuse  generation only to the extent  that the incidence of
 the property tax differs from equality across individuals.   If the property
 tax  incidence is sufficiently regressive to approximate the impact of a flat
 fee user charge  in  a heterogeneous community, then little  difference  in
 aggregate refuse  generation would be expected as a result of the financing
 scheme.  Many smaller communities are quite homogeneous  in  terms of  income,
 and  there the  probability is high that  the incidence of the flat fee user
 charge and the property  tax would not be too different.   It is assumed, then,
 that  household  refuse  generation  is  the same under  the flat fee  or  the
 property tax financing  scheme.  The  imposition of a  quantity based  user
 charge system  on  these financing schemes  will,  however, result in different
 increases in the administrative costs of the two systems.

        Figure  2 portrays graphically the marginal savings and costs to be
 expected from imposition of a quantity based user charge system on  the  two
 financing schemes.   In both cases,  households generate Q  refuse prior to
 imposition and Q  refuse post imposition.  The flat fee system is shown  as
 saving some of  its  pre imposition billing costs as refuse to be billed for
 decreases after imposition of the user charge system.

       Both  systems also reap whatever cost  savings  are attributable to
 collecting and disposing of a reduced quantity of wastes.  The quantity based
 user  charge system  is  showed as causing the equilibrim price to equal 1.079
 times the property-based cost of refuse collection.  There is,  obviously a
 greater  increase in billing costs for  the  property tax  financing scheme than
 for  the  flat fee user  charge system.  Clearly,  then imposition  of the
 quantity based system is desirable in either case if the extra administrative
 costs do not exceed  the  cost savings from reduced generation of refuse.

        Estimating  the  size  of  the cost  savings from reduced  refuse
 generation  is difficult. There is no  clear  cut evidence about households'
 reactions to  imposition  of such a charge.   Studies  of existing  cities with
 quantity  based user  charge  systems indicates  that the price elasticity of
refuse generation may  be very low, but  this  evidence is  by no  means
conclusive.   Rather  than estimate  these  refuse  generation responses directly,
 then, they will  be calculated  as a residual,  indicating  the minimum response
necessary to offset  increased administrative costs.  These estimates are
presented  in  Table 3.

        Table 3's  data  is based on the following general assumptions:  that
 the additional administrative costs imposed in moving  from  a property  tax
 financing scheme are $2.96 per household per year; that  the comparable figure
 for moving from the  flat fee user charge system is $1.53 per  household  per
                                    88

-------
                                                          , quantity based
                                                          fee schedule
                                                  1.038x=flat fee cost
                                                 x-cost of collection
                                   Q*          Q°      Q
                          Total Tons Collected
  Property tax system:
Collection
Cost Savings

Administrative
Cost Increases
Flat fee system:
        Figure 2.  Costs and savings from imposition of a
                   quantity based user charge system.
                                   89

-------
    TABLE 3.  REFUSE GENERATION RESPONSES REQUIRED
               FOR ACCEPTABILITY OF QUANTITY BASED
	USER  CHARGE SYSTEM	


    General assumptions:

       Refuse per household,  pre imposition       1.5 tons/year

       Cost per household, pre imposition         $37. 50/year

       Extra administrative costs household

           Property tax to quantity based user fee  $2.96

           Flat fee to quantity based user fee       $1.53


       COST  SAVINGS FROM REDUCED REFUSE GENERATION

                                                                  #
Alternatives:                   Required refuse generation reduction

                              Property tax to  Flat fee user charge
                               quantity based    to quantity based
                                user charge        user charge

Cost savings  = 90% of average       8.8%              4.5%
              collection costs

Cost savings  = 60% of average      13. 1%              6.8%
              collection costs

Cost savings  = 25% of average      31.6%             16.3%
              collection costs

* Reduction required such that cost savings will equal increase in
  administrative costs.

-------
year.10  Three  alternative assumptions are made regarding the collection cost
savings which would be experienced as a result of collecting  less  refuse at
each  stop:   savings will  be 30%, 60%, or  25%  of what the average cost of
collecting the  refuse no longer being generated  would have  been.   The 60%
figure  results from engineering models, while the 25%  figure  is a point
elasticity calculated from econometric work.  *

        If cities are efficient and can reroute and reschedule in response to
refuse generation reductions, then cost  savings can be high, and the required
refuse reduction to justify the quantity based user charge  system  is  quite
low—4.5  to 8.8%.   The converse also holds  true.  These simple calculations
highlight  the fact that for successful Imposition of a quantity based  user
charge  system, a streamlined administrative  system and  a  somewhat price
sensitive refuse generation behavior are  only necessary conditions.
Sufficient  conditions require also that the refuse collection organization
itself responds efficiently to the changed levels of demand for its  services.
Development of procedures  to aid managers in  effecting  these  efficient
responses  should be a cornerstone of any quantity based user charge  program.
10.  These costs are calculated 40 a percentage of baseline collection costs.
     No downward adjustment  it  made for reduced generation affecting thit
     product,  ae it is assumed that sdmnistrative coit will be more sensitive
     to number of sccounte billed than to tha  quantity of refuse generated*

11,  See  Stevens,  and  Hudson, Work  undar progress for  the Resource
     Conservation Committee, B«P.A,, 1978 and  op. cit. 1978.
                                     91

-------
                                  SECTION 3

               PRICING AND THE EQUITY OF SOLID WASTE FINANCING
"Equity Considerations in User Charges for Residential  Refuse Collection and
Disposal" (Roger Bolton).

"Pricing Municipal  Refuse Service:  Potential  Effects on Municipal  Budgets"
(Kenneth L. Wertz).
                                     92

-------
                 EQUITY CONSIDERATIONS IN USER CHARGES FOR
                RESIDENTIAL REFUSE COLLECTION AND DISPOSAL
                                   by
                              ROGER BOLTON
THE GENERAL  CONTEXT

        I first  want  to  set  the  context  for  considering  equity
considerations.   We are concerned with the distribution of effects of  a
change  from one system of public sector solid waste management (henceforth,
SWM), call it System A, to another system, call  it  B.   We are motivated
chiefly by the possibility that System  B has user charges,  System A does not.
Indeed,  in  many discussions of this  problem writers implicitly or explicitly
assume that  System A and System B differ only in that B has charges and A has
some  other  financing method.  More than one  author discusses the equity
effects  of  user charges  as  if the conditions of refuse management are
unchanged when changes are  imposed.  By conditions I mean to include the
aggregate quantity of refuse and the costs of handling it, as well  as  the
distribution of those aggregates over households.  Costs here must include
the cost incurred  by households as well as  the public sector.

       I think that is too restricted and too hypothetical  a comparison.   In
the real world,  if System B includes user charges  it will  probably also show
other changes from System A.  Those other changes  will affect the total costs
in the public sector.  There are the billing costs, for one thing—every one
mentions them.  But in addition there are changes in household behavior,
induced by the introduction of charges.  Those  changes in household behavior
are  precisely the changes  which are emphasized  in  discussions  of  the
efficiency aspects of charges.  If the price system "works" when user  charges
are  introduced,  then  the aggregate  quantities  of refuse and the aggregate
public sector costs of handling  it will  change.  And so  will the public
sector costs of  handling any one household's  refuse; one cannot assume the
costs imposed on the public sector by  a household's refuse under  System B
will be exactly  the  same as  they  were under  System A.   Nor can one assume
that the costs imposed  by an income class's refuse will be  the  same.
Finally, the changes  in household behavior will'change the aggregate costs
incurred by households,  and they  are  relevant for an analysis of  equity.
Again, one cannot assume  the  distribution  of household-incurred  costs is the
same  under System B as  it  was  under  System A.

                                    93

-------
         The upshot of this is that the analytical  and empirical problem  is
 one  of  "balanced budget  incidence," rather  than  one of  "differential tax
 incidence" (11,  ch.  10).  It is balanced budget  incidence because both the
 income and expenditure  sides  of the public budget will change:    the
 expenditure side changes because quantities of refuse and costs of handling
 change;  the income  side  changes if income  exactly covers costs.   The
 literature seems to  have sidestepped this problem,  and dealt with the problem
 of differential tax incidence,  instead.  Writers  have used differential tax
 incidence,  because  they have assumed total expenditure on SWM is constant,
 and  also  the expenditure required to handle each household's, or each income
 class's, waste.

         No doubt,   with the  uncertainties  and controversies over tax
 incidence, especially over property tax, analysts have felt  they had enough
 on their  plates if they  tackled the differential  incidence problem.   They
 have also  had the problem that most of  the data on public sector costs are
 for  cities where System A  is  in place,  and System B has  never been tried.
 Then, too, some  writers feel the  differences in costs between System A and
 System  B  are not large enough to  worry about — not  large  enough to require
 a balanced budget incidence model.   They may argue  that my initial point  is
 rather  academic.  That may be.   However,  I do find  it strange that there is
 so little  discussion of the point,  in analyses of a policy  change which  is
 recommended in  the first place on efficiency grounds.  Economists recommend
 charges  because  they will induce changes in the quantities of refuse offered
 by households for collection and  disposal, and, therefore,  changes in public
 sector costs.

 CHANGES  IN THE CONDITIONS OF SWM

 Changes  Initiated by the Public Sector

         First, we may have significant changes  in the  range  of choice offered
 by the  public SWM  system.  The  move to  user charges may  be accompanied  by
 changes in  the  frequency of collection, the regulations  on types of
 containers,  the locations  at which containers will be picked up (curb side.
 back  yard,  etc.), regulations  on separation of garbage  frm  other refuse or
 separation of newspapers from other refuse,  the availability of bulky  item
 pickups, etc.  Whether  the range of choice  will be expanded  or narrowed will
 depend on  specific  circumstances  in  the  municipality.  The nature of the
 previous system,  the  impetus to user  charges  in the  first place,  and the
 socioeconomic characteristics of the residents  are all relevant.   It's  hard
 to generalise, and one  should probably do no more than  point out the  need to
 keep  these things in  mind when analyzing  the  equity effects of a change in
 policy.

         Are  such changes important in practice?  Can't one assume them  away
 when  the financing method  is changed?  I think  it would be  wrong  to  do  so,
 for  some  very definite reasons.    For  one  thing,  SWM  is  a
 politically-sensitive function, and changes  in  it will produce "flak" for the
 managers.   That  is  especially true if changes include a move to an untested
 and basically unpopular—in the short run--pricing  system.   If  solid waste
managers  even thought about introducing any other changes in their  system,

                                    94

-------
they may decide to bite the bullet  and introduce  them along with charges.
They may prefer to endure one period  of intensive criticism  and then be done
with it, rather than endure several recurring periods of criticism.

        For another thing,  some changes in the conditions of collection and
disposal may be called for by the efficient and cost-saving administration of
the user charges scheme itself.

        But the most important reasons for expecting multiple changes by the
solid waste managers are different.   One of the reasons is  that a change to
user charges is likely to be stimulated by a crisis or quasi-crisis situation
in  the  municipality.   It may be a budgetary  crisis, or  a landfill-space
crisis.   In either case, user charges  are likely to be part  of a package tc
cope with the problem.   The municipality may eliminate some costly service
options, such as frequent collection,  or backyard pickup.  It may eliminate
these altogether rather than go to the trouble of pricing them correctly.  It
may also try to increase resource recovery revenue to the municipality  or to
reduce  landfill  space  requirements; for  example, it may require that
newspapers be separated out from other items.

         Solid waste managers may find it very difficult—very  costly,  that
is—to maintain the same environmental standards when they substitute System
B  for  System A.   We  often  worry, for example, that some households  will
respond  to the user charges  by increasing what is euphemistically called
"alternative disposal".   This may  include  illicit disposal by dumping the
rubbish in vacant lots  and parks or  open  burning.  It may also include
increased use of garbage  disposals  in the kitchen, which transfer  the
enironmental burden from the  SWM system to  the sewage  disposal  system,  and
increase the public  sector sewage  costs  and/or the pollution damages
accordingly.  Some analysts discuss user charges in solid  waste under  the
explicit assumption  that environmental standards are  unchanged.  This seems
wrong.   Either there will  be increased public sector costs to insure  the
maintenance of the standards, or else the standards will not be met.  In
either  case, the  usual limited approach of  differential  tax  incidence  is
incomplete, and misleadingly  so.   In the first case, public  expenditures will
be  affected.  Also, in most  jurisdictions  sewage disposal is already financed
by user charges, the equity aspects of  which may  be quite different from
those of the general property tax  which one usually  associates with the  SWM
function.   In the second case,  the distribution of  increased  environmental
costs is  itself an issue;  many  of these costs are likely to bear heavily on
the poor.

Changes Initiated  by Households

         There are  at  least  three  kinds of  changes here.  At the risk of dull
repetition, I point  out again that they are the very things  given the biggest
play in  the literature on  efficiency,  sometimes by authors  who go on to
ignore  them in their analysis of equity considerations.   Furthermore,  the
basic theory of household behavior inescapably concludes  that the  income
levels  of households may be  a prime determinant of the changes;  that  leaves
us with an  equally inescapable  conclusion that  the equity or  income
distribution Impacts  of added household costs needs some careful analysis.

                                     95

-------
        Households may  initiate these changes:

        1.   Change in  consumption habits in  general.  Different market goods
and services have different amounts of solid waste attached  to  them;  user
charges for solid  waste  cause effective prices of goods and  services to
change, theoretically leading to changes  in consumption.  Thus, solid  waste
charges might affect  the consumption  of newspapers, of frozen  food, of
beverages, neat lawns,  and other things.   The net  impact on the  welfare of
the household after it has  made its optimizing adjustments  will vary with
income and with  some  important  family characteristics, such as  the
opportunity cost and psychic cost of time spent in waste handling by various
members of the family.

        2.  Change in household time spent in handling waste.  Time may be
spent in alternative disposal, separating kinds of waste in order  to meet new
regulations or to earn  money from recycling,  carrying containers longer or
shorter distances for pickup, storing waste more carefully in order to reduce
the frequency of pickup service, etc.  Unlike in the previous paragraph, here
we are concerned with the impact on the household of the changes in its use
of time per se, rather  than the impact of the changes in its consumption of
market goods and services.  However, the same determinants mentioned in the
previous paragraph are  operating here—income and characteristics affecting
the opportunity and  psychic cost of using time in various ways.

        3.  Change in household money expenditures on some particular market
goods and services,  which are  used  in  solid waste management in  the
household.  Chief examples are items used in  alternative disposal, such as
automobile and kitchen garbage disposal costs,  and items used  in  waste
storage, such as  compactors and containers.   These are merely a subset of the
wider set of changes discussed under  Section  1,  consumption habits in
general, but it is useful to distinguish  them because user charges set up
definite  incentives for  alternative disposal and storage methods.  Again,
income and value  of  time are important determinants of the changes.

        My emphasis on household time  and  on expenditures for market goods
and services is inspired by a Becker-Lancaster household production model.
That  model is  fruitful in analyzing  the  household responses which need to be
taken account of;  in the household production approach, some broad good  like
"household environmental  quality" is an argument in the utility function.
The household chooses between  its own production  activities  and public
management services as substitutes  in  providing environmental quality; the
presence  or absence of user  charges for the public management services
obviously affects  that choice.   But user  charges also affect the price of
environmental quality per se,  no matter how  the household achieves it,  so
they  affect the consumption  of household environmental quality  relative to
other things which give utility.   Thus, they affect  consumption habits  in
general.

        There are  also production functions for the household activities  in
question,  the inputs in which  are time of  family members and market goods and
services.   User charges affect  the optimum level of  these  household


                                    96

-------
production activities, so they affect  the optimum allocation of family  time
and also  the purchase on  the market of inputs into production.  Thus,  when
all the household's reactions  to charges (and  accompanying public  sector
changes)  have been worked out and a new optimum allocation  of time and money
income achieved,  it is possible that all sorts of things have changed.   The
mix of environmental quality and other things in the utility function must be
achieved simultaneously with the mix of household production activity in
waste management and the use of public  SWM  services, and both must be
achieved  simultaneously with  the optimum utilization of  time  and
market-purchased inputs  in household production activities.  After all the
dust has settled,  the household may be better or worse off,  depending  on the
balance between its user  charges and the amount of taxes it saves, but  also
on the nature of  the accompanying changes  in the public system,  on  its own
household production functions, and, of course,  on its tastes.

        Clearly,  the task of equity analysis theoretically  should include a
prediction of how the net result of  these  household decisions varies across
households at the same income  level.  That  is essential for  the question of
whether "equals"  are affected  "equally,"  or  the  "horizontal equity" question.
It also must include a method  for expressing the net result of the decisions
in some kind of monetary equivalent form,  such as a "consumer surplus".   And,
finally,  it must include  a prediction on how the results of  household
decisions are correlated with income  levels—that  is essential  for the
question of "acceptable distribution of  income"  or  "vertical equity".

        One  way  to  make the  point I am trying  to make is to refer to Figure
1.  Here we have  three different households'  demand  functions  for public
collection and disposal.  Under  System A,  in which  there is a zero charge for
the service, the  three households all offer the  same quantity of waste,  4,000
units  per year.   However,  their demand curves  have very different
elasticities,  so that if  a user charge  of $.03 per  unit is  imposed  they
respond  in  very different ways.  At  the $.03 price,  for  example, C's
elasticity is unity,  D's is 1/3, and E's is zero.

         If for simplicity we accept  consumer's surplus analysis, we see that
 the households' loss of  surplus generally is not correctly estimated  by an
amount equal  to the new price,  $.03, times the quantities offered at the
 original  position.   The impact as measured in that  way is $120 for each
 household,  but  that is the  true  loss of  surplus only for Household E.  The
 $120  figure  overestimates  D's loss  of surplus,  which  is only $105,  and C's
 loss which  is only $90.   This is an obvious and simple point,  and needs no
 belaboring here. It is an example of  the familiar principle that  one must not
 be content  with an analysis based on "marginal" changes when the changes are
 actually  quite  non-marginal.   This point is especially relevant in the
 present context, for when user charges are  introduced for the first time they
 will  be decidedly non-marginal!

         In order to use this analytical point in equity analysis,  one must of
 course make  some conjectures on how the elasticities vary with income levels,
 and how  they vary across households  who are at the same income level,  or who
 are in the "same circumstances" from  the point  of view of horizontal equity
 analysis.   There are hardly any good studies  on the price elasticity of

                                     97

-------
0     1,000   3,000   3,000   4,000
                Figure 1
                  98

-------
demand at all,  let alone studies on how it  varies in those ways.   One would
expect, however,  that it does vary considerably across households  at  the same
income level.   How the elasticity varies with  income is a difficult question.
For example, higher income households probably have stronger preferences for
some market goods and services which generate  a lot of waste—packaged foods,
newspapers and magazines, and large well-kept lawns, for example.  And they
probably have stronger preferences for a tidy  household environment as well.
And they are  more likely to have higher opportunity costs of time,  at least
if we consider  the adult members of the family.  These all would seem to make
their  demand  less elastic.   On the other hand, they are more likely to buy
compactors and  kitchen garbage disposals, which  would work  in the opposite
direction.

THE MEANING OF  EQUITY

        It is customary to speak  of horizontal equity and vertical equity.
The first is said to require equal treatment of persons in equal  positions;
the second to require  equal taxes  for equal  abilities to pay, which most
politicians accept as requiring some progressivity  relative to income.

        If we  have a genuine public  good,  or at least a  good with such
pervasive positive externalities that the benefit theory  of  taxation can  be
disregarded, there is perhaps a simpler way of expressing the  two principles.
Then it is simpler to say there is only one principle:  equity requires the
equal  treatment  of persons in equal positions, but "equal positions" must  in
fact be determined with  reference  to equal ability  to pay and "equal
treatment" implies something like  equal sacrifice.   From the one basic
principle stems both the horizontal equity  and the  vertical equity principle.
Horizontal equity means requiring persons with the same index of ability  to
pay and in the  same general circumstances to pay the same tax.   Vertical
equity means  requiring persona, with different  indexes of ability to pay,  to
pay different  taxes, and a  generally accepted  implication of  this  is  that  as
income increases, taxes should increase as  a proportion of income.

        This is overly simplified, but seems to me  to  capture  the  essence  of
generally accepted notions of equity in  taxation for public goods.  But  in
passing it must be admitted that the notion of "equal circumstances," when
circumstances  include aspects other than income or other index of ability  to
pay, is a slippery one.  I  have suggested earlier,  for example, that  certain
family characteristics will affect the  responses  of  households  to user
charges.  Family  characteristics affect the opportunity cost  and psychic cost
of  the time required for aternative diposal,  and also the ability to finance
compactors and  garbage disposals.   While the incomes  of households are
important determinants of these  impacts, the impacts  are not perfectly
correlated with income.  Thus, different families may  feel widely different
effects of user charges,  even though they have identical incomes.  For
example, a family with several young children or a  nonworking  spouse, who  can
handle the rubbish, may feel the adverse  effects-of  user charges  much less
than another family with equal income but different circumstances.  How  is
that to be evaluated in equity analysis?
                                     99

-------
        This  discussion has been  for the case of  public goods.  When  the
public  service is a business-like one, and the benefits and public sector
costs vary considerably from household to another,  we might abandon  the
previous  notion of equity and  substitute the principle  of benefit taxation.
We are especially likely to do  so  for a  function like solid waste management,
where the service is so business-like that it is left  to business to provide
in many places!  And where the  price  needed to  cover costs  is  such  a
small—almost trivial—proportion  of total income.

        The benefit theory means that the recipients of SWM services,  should
pay  in  aggregate a sum equal  to the  total costs of  the services (this may
require  a  multi-part tariff pricing scheme in order to  preserve marginal-cost
pricing).   In addition to this  aggregate requirement,  it requires that  the
distribution  of payments across  households be the same as the distribution
either of  benefits received or  of  the public sector costs incurred.

        Now,  efficiency requires that the payments required of a household be
based on the  public sector  costs  incurred to  serve it.  However,  here as
elsewhere in economic policy, efficiency is not necessarily the only goal.
The distribution of benefits from  the public activity  is a relevant  equity
consideration; if the SWM benefits were larger, relative to total economic
welfare, we might pay a lot more attention to their distribution than we  do.
But  the  benefits  are not very large—not  nearly  on the same scale as
education  or  medical care, for  example—so we  are  much less interested in
sacrificing efficiency in order to get a more equitable distribution of them.

        The benefit principle  leaves no  room at all for vertical equity in
the  case  of  something as small in scale as solid waste management.   Or so it
seems to  me.  If we didn't have to worry about  the externalities from
alternative disposal, we wouldn't  need to talk  about benefit principles of
taxation at all.  We could simply  talk about business pricing.

        However, not every observer will accept the benefit principle.   The
principle tends to be most appealing to those who  believe firmly in the
desirability of the so-called Musgrave-Tiebout  "layer  cake" model  of
government  (8,  referring  to  11,17). .In  that model the functions of
governments  are sharply  divided into the allocative, distributive,  and
stabilizing  functions,  and  they are divided up among  the  levels  of
government,  in a federal or other multi-level political system, so that only
the  lowest level handles an allocation function like solid waste management.
In the model, SWM services are  properly  financed  by the lowest levels because
their externalities are not widespread geographically.  Furthermore,  those
levels  finance such services without regard for  distributional consequences.
They leave distribution to   the higher  levels of  government, which
redistribute income in  general  way,  rather  than through  price subsidies
which distort the efficient allocation of resources.

        Not everyone accepts the layer-cake theory.  Some critics would argue
that the distributional consequences of  every  government policy,  even one as
mundane as the management of  the garbage, must be considered explicitly by
the level  of  government to which history and political realities have  left
the  function.   The critics would argue that no local government can assume

                                    100

-------
that higher  levels will  in  fact redistribute  income  in the  desirable
direction,  let alone in the desirable,  non-distorting way.  They may argue
that Thurow's  notion of  "individual-societal preferences" (16), or  Tobin's
notion  of "specific egalitarianism" (18),  both  of which are originally put
forward  for  goods looming  as large as  medical care and education, can be
legitimately  extended to solid waste.   They may claim that the  externalities
of SWM are in fact pervasive enough  that it is  close to the public good end
of the  continuum.  A related argument  is this:  in our modern technologically
advanced society, the environmental  media  are so interconnected  and the
production of packaging  materials is so far  beyond the  control  of  the
individual consumer, that  the management of waste is in fact an  essential
function  for  government and it is no longer realistic to expect individuals
to feel  any  responsibility at all it.  These arguments, while they  are not
likely-to be  accepted by many of us, and while the last one in particular is
quite pernicious,  are not  totally ridiculous in the politicial  arena.

REVIEW OF EMPIRICAL STUDIES

        The sensitive and broad-minded solid waste manager will want  to  know
something about  all the changes in the distribution of benefits when he moves
from management system A  to management system  B, even if in the end he risks
a judgment that  some effects are not awfully  important.   And from what  I've
said,  it should  be clear that he will want  to know  about all  the changes  in
conditions of waste mangement, which accompany  or are induced by the move  to
user charges.

        The ideal empirical study for him would be a  "before-and-after" study
in some  community similar to his own.   Alas,  he is unlikely to get such a
quasi-controlled experimental  result*  At the moment, the best he can do  is
to look  at the few studies which relate quantities of  refuse offered for
public collection and disposal  to income and other household characteristics,
under one single existing  management policy.   In other words, he will have to
be content for now with analyses of the variations in so-called "demand" for
refuse services  under only one system, which corresponds  to  System A in our
hypothetical comparison.   He will find  a  few such analyses,  but unfortunately
they tend to be silent on  some  of the things which might  actually happen when
System B is substituted for  System A.   From them he will  have  to make his own
conjectures on households' elasticities of demand for public  service and how
the elasticities vary with income levels and other characteristics.

         I  shall review the  salient features of some  of the  major empirical
studies  here.   In this,  I am especially concerned,  of course, with the
estimates of the income elasticity of refuse generation,  as  that is very
important in inferring  vertical equity  effects.  There is a strong consistent
pattern to  the estimates of  income elasticity.  In addition,  for horizontal
equity,  we must  note  how well  any  statistical demand function fits the
behavior of households.   The more  poorly it  fits,  the more  variation in
refuse generation there  is,  at given  levels of  household income  and other
variables measuring  "equal  circumstances."   That implies  something for
horizontal equity.   However, rather than review  all the details of the
                                    101

-------
 studies in that  respect,  let me say here that the  statistical functions fit
 well enough for us to be  confident on the Income elasticity of demand,  but
 also poorly enough  to indicate that  horizontal  equity would be seriously
 violated under a financing  scheme which collected the  same  amount from  all
 households at the same income level, regardless of the  refuse they generate.

 Downing

        Perhaps  the most  useful  study  which confronts  equity aspects
 empirically is the one described in two papers  by Paul  Downing (5).  Downing
 analyzed variation  in refuse quantities, collection costs,  and estimated
 property tax burdens across 64 collection routes (henceforth CR's) in
 Riverside, California in 1971.  First, he estimated average volume of refuse
 per household in a CR, as a  function of average family  income, average family
 size,  and land area  per household in the CR.  Volume was based on one single
 pickup in each CR during February or March 1971.  We know that generation of
 refuse is highly seasonal, and the income  elasticity  varies considerably by
 season, at least in other parts of the country  (2, 14,  15,  19), and Downing
 in  fact recognizes  that  his results cannot easily be extrapolated to other
 seasons and/or localities.   The income elasticity in the equation was .39.

        That  income  elasticity  is  a "net" or "partial" elasticity, in that
 the equation holds constant  other household and CR characteristics.  As 1
 have  pointed out elsewhere (2, p. 3), in an analysis  of income distribution
 effects, the "total"  or  "gross" elasticity,  not  holding  other variables
 constant, is more important.  If  income  and refuse quantities/costs are
 correlated, that is relevant even if other variables,  correlated with both
 income  and refuse,  improve the  fit  of  a statistical  demand equation.
 Although Downing does  not explain this point very well, presumably the gross
 income elasticity in Riverside was .55, for that is the value he uses later
 in discussing income  distribution effects.

        Downing's average  income measure  for each CR had to be constructed
 from Census estimates  for Census Tracts for  1970.  CR's are  not coterminous
 with  Census Tracts,  so he  had to construct a  weighted average of the Tracts
 in each CR, weighting  each Tract by the number  of households living in  the
 part  of the Tract within the CR.  This is apparently the same method used in
 the Kemper-Quigley and Tolley studies reported  on later.  Such an estimate is
 valid, of course, only  if Tracts are geographically  homogeneous.  However,
 there is ample evidence that Tracts are not  very homogeneous as far as income
 is concerned.

        Downing  then analyzed the variation  in  collection cost  across CR's.
He estimated  cost per household per year on the basis of careful studies of
labor requirements and  user cost of equipment.  In  a  regression, he found
cost to  be a  function of  refuse per household (based, again, on the  one
observation per  CR in  February-March 1971),  truck travel times per household
 (to capture effects  of density),  and percentage of housing units which  are
single family. All these variables had positive and  significant  effects.
                                    102

-------
        Downing  added  to  collection costs estimates  for costs  of
transportation and disposal at landfill.  He then regressed total management
cost  incurred  to handle a  CR's  refuse on family income  in  the CR; the
estimated income elasticity was .44.  Thus, if a user charge were imposed,
and fell on households in proportion to the costs of handling their waste, it
would  have an income elasticity of  .44.

        An analysis of the impact of the alternative financing  scheme, the
property tax,  requires estimates  of the property tax payments by households
in each CR.  Downing first computed property values in  each CR from 1970
Census  block  data.  He assumed that the incremental property tax for  refuse
management would fall on each CR in proportion to the property values.   Fully
recognizing  the  shortcomings of many of his procedures, Downing then
regressed estimated property values  against family income, and found an
elasticity  coefficient  of  1.16;  that indicates that  the property  tax is
progressive, and that the elasticity is well above what the elasticity would
be for  user charges based on management costs.  Thus, Downing concluded that
user charges would be much more regressive than the property  tax.

        Remember that Downing found the income elasticity  of  volume of  refuse
collected was  .55.  Accepting this  as  a measure  of benefits, he concluded
that  the user charge would distribute costs in about the same proportion as
benefits, because  .44 is very near  .55.  The property tax method, however,
has  a net  effect which  is  progressive, in the sense  that tax costs to
households are more progressively distributed than benefits (1.16 is  higher
than  .55),  and also in  the  sense that tax costs  to households are more
progressively  distributed than would  be user charges (1.16 is higher  than
.44).

Kemper-Quigley

        One  of the most  prominent of studies is that by Kemper  and Quigley
(7).  They analyzed New Haven data  on weight of annual household  refuse and
family  income estimates  for 33 residential collection  routes  in New Haven,
Connecticut,  in  1972.  They tried a variety of econometric specifications,
using income and  other  variables  to explain the variation in  refuse
generation per household across collection routes.   The  estimates  of gross
income  elasticity of quantity varied with the specification, but  were
generally between  .65 and  .75 for  the better fitting specifications.

        For  example,  in a linear equation,  the gross  elasticity was .45
(income is the only independent variable), and the net  .35  (household  size
and density also  in the equation). However, log-linear  equations fitted the
New Haven data slightly better, and in  those the elasticity rose to  .82  when
income  alone was used,  and  to .67 when household  size (significant) and
density (not significant) were added.

        In alternative models for the  New Haven data,  they used refuse per
capita as the  dependent variable,  rather  than per household,  and personal per
capita  income as the  independent income, rather than household  income.  In
                                    103

-------
the log-linear versions of these,  gross income  elasticity was  .39  and net
.32.   The  estimates in linear forms  were almost exactly the same.  Household
size was not  significant in these  regressions; density was significant in the
linear form but not the log-linear form.

        Kemper and Quigley also analyzed a pooled sample, consisting of the
33 New Haven routes plus municipality-wide data for 35 other municipalities
in Connecticut.   There was  considerable measurement  error  for the  35
municipal  observations.   The degree of fit was  very much  smaller.   The
estimates of  income elasticity were  .65 (gross) and .51 (net)  in a linear
equation in  which household size  and density were not significant, but  had t
values above  one.  In the log-linear form, income was more  significant  than
in the linear form, and gross elasticity was  .67 and net .45.  Household size
was significant.

        On the equity  issue, Quigley and Kemper note that the total amounts
are small in  any event, perhaps only .3 percent  of median family income for
their  sample of Connecticut towns.  Their analysis, however,  does provide
"strong evidence that  lower  income households generate  less  refuse than
higher income households" (7, p.  89).  That  pattern can  be  compared to
existing estimates of the incidence of the property  tax nationwide in  order
to conclude  something about equity. As I  explain in the next  section,  the
authors accept  some traditional estimates of  property tax burden, which show
it as  very regressive;  Kemper and  Quigley note the older estimates give
elasticities  of property tax burden which are even lower than their estimates
of refuse  quantities.   They conclude, therefore, that  the property tax
financing  of residential  refuse collection is much more regressive than a
system of user  charges which is based on quantity collected.   They conclude
that  it would be "highly progressive" if cities went from a property tax to
user charges  based on tonnage.  Thus, they disagree  sharply with Downing.

         It should be  noted that  Kemper and  Quigley did  not attempt to
estimate the  income elasticity of the cost of collecting refuse.   This  is in
spite of their having  accumulated vast amounts  of data on collection cost by
collection route,  and analyzed it in other connections.   It would have been
most interesting  if they had estimated a relationship between cost and income
in CR's.

Chicago Study

         I note two other empirical studies here.  They are "demand" studies,
which  go into considerable  detail on the variation in refuse weight or volume
by  income and  demographic  characteristics of collection routes.   But they do
not match this  information  with estimates  of  the income distribution effects
of various financing  schemes in the sample cities.   Therefore,  they  are
relevant here only because they give solid additional  support  to the
conclusion that the income elasticity of  refuse generation is well  below
unity.

        The study  of Chicago's 50 CR's by  Tolley, Hastings, and Rudzitis (19;
this is an updated version  of  a previous study by Sheaffer  and Tolley, 14  and


                                    104

-------
15) Is especially important,  because it analyzes  generation at different
seasons.   It also includes some variables which other  authors have not, such
as race and variance of household  income in a CR.   In the specifications
which the authors feel are best,  net  income elasticities varied between  .38
and .66,  depending on the  season  and the exact  specification, and gross
income elasticities varied between .50 and .64.  The  elasticities tend to be
highest  in the fall and late  fall/early winter,  lowest in mid-winter
(February),  and lower in  the  spring and summer than in the spring and  the
fall.   From this seasonal  pattern, the authors  conclude  that there is  a
"basic,  year-round volume of  wastes, represented by the mid-winter volume,
that is relatively less sensitive to income than  'excess'  volumes  in other
parts of the  year, except possibly mid-summer" (19, pp.  88-89).  In this,
they reinforce  the earlier conclusion of  Sheaffer and Tolley (14, 15).

        This  is an  unusually rich and sophisticated demand study.  I cannot
go into more details here, however, because it does not confront the  equity
issues empirically.

Springfield Study

        Finally,  I  may note my own  unpublished study of  refuse generation
demand in 1972  in  Springfield, Massachusetts  (2; see also 3). My analysis of
the city's  20 collection  routes attempted to explain variation in  refuse by
variation in  several social and demographic variables:  property value  per
household;  percentages  non-white, younger than 18 and older than 62; and
percentage of  housing units occupied by their owners.  These functions fitted
rather well,  except in  the winter season.  I made no effort to  construct
income estimates for CR's,  having less faith in the homogeneity of  Census
Tracts than Downing, Kemper and Quigley,  and Tolley.  However,  I constructed
indirect  estimates of gross income  elasticity in the following way:   (a) I
estimated refuse demand equations using  CR's as units of observation, as
noted above;  (b) I then assumed that the same demand function would  apply if
Census Tracts  were the units of observation,  and used  the function  (based on
CR's)  and the observations on independent  variables  (by Tract) to estimate
the refuse generated in each Tract; and (c)  I then regressed those  estimates
of  refuse quantities  on income in Tracts  to get a gross income elasticity.
The results varied by  season:  .17 in winter;  .44 in spring and also in
summer; and .72 in fall (the fall elasticity  is  high obviously because of all
the New  England  leaves falling on the  large lots of higher income
neighborhoods).

        I had no direct data on costs, and did not engage in the same useful
exercise of estimating  them carefully, as did Downing  and Kemper and Quigley.
I resorted to very rough  estimates of relative collection costs in each area.
To  do  this,  I assumed  that in a  CR,  the  collection  cost per household
relative to the city average, would  be the  same as the relative  number of
standard collection  crews required to service the CR in each season.   Each
crew consisted of  the  same  number of men and truckst  so  I felt able  to assume
the  collection cost in CR's varied in proportion to the number of  crews*  By
assuming this, I estimated  an elasticity of collection cost per household
with  respect  to  refuse  per household:  .79 in  winter,  .75  in spring,  .81  in


                                    105

-------
 summer,  and  .63 in fall.  When  I  multiplied those  by the elasticities  of
 refuse  quantity  with respect  to income, previously estimated, I got these
 estimates  of elasticity of cost  per household with respect  to  family  income:
 winter, .13;  spring, .33; summer,  .36; and fall,  .46.   While my cost data
 fell  far short of the quality of  Downing's, they are consistent with his
 results, and extend them to other  seasons:  the cost  of serving households  on
 a route  rises much less than proportionately to family income  on the route.

         I  did  not attempt  to estimate the pattern of property tax incidence
 in Springfield.  But I concluded that the property tax elasticities are  at
 least a bit higher than the refuse  elasticities just given.  Therefore, I
 concluded  that if Springfield  refuse and costs data were typical, one  could
 say that  "the use of property  taxes to finance refuse management may have a
 very  slight equalizing effect  on the income distribution"  (2,  p. 2).

 IMPORTANCE OF ASSUMPTION ON INCIDENCE OF PROPERTY TAX

         The most likely alternative to user charges for financing SWM is the
 local property tax.  Thus, what  one assumes is the  incidence of that tax  is
 crucial  for his analysis of both horizontal and vertical  equity effects  of
 substituting user charges for  it.  That was shown by the  conclusions of the
 Downing  and Kemper-Quigley studies referred to in my  previous section.  The
 authors reached  different conclusions  on the distributional impact, not
 because  they assumed very different income elasticities of  refuse collected,
 but because they  assumed radically  different income elasticities of the
 property tax burden.  Therefore, I think it useful to delve into the theory
 and literature on property tax  incidence as part  of a  discussion of the
 equity considerations in SWM.

         I  have dipped  into  the voluminous literature on property tax
 incidence  several  times,  and I tried to refresh my knowledge of it in writing
 this  paper.  We all know there  is a  "New View" on this, which argues that
 even  the residential property part of the tax is much  less regressive, and
 may even be slightly progressive, than the conventional wisdom of 10-15 years
 ago.  I think it is generally agreed by now that the  New View  is sound enough
 to force a major  revision of  thinking,  although not all are so bold  as  to
 insist that the tax is regressive or is^ progressive.   Note  that by "New View"
 here  I am not referring to the abstract  theoretical  literature which argues
 that  the property  tax is highly progressive under  certain artificial
 assumptions.  There is such  literature,  but  by New  View I mean rather the
 Aaron-Netzer-Mieszowski theoretical and empirical studies  which amend the
 abstract theory in  applying  it to the real world (1,  13, 9, 10).

        Such a  champion of  the older conventional wisdom as  Dick Netzer has
 partially  recanted  his earlier position,  and now agrees that  the residential
 property tax is  not  as  regressive as once thought,  and  that the  total
 property tax is probably  slightly progressive in a typical jurisdiction  (13).
 I may add that Kemper and Quigley appear to have relied much more on  Netzer's
 earlier  position, and on other writers  using  the same assumptions,  than  on
his later  change-of-mind  and other representatives of the New View.
 (Netzer's earlier study was  (12)).


                                    106

-------
        Netzer's partial recantation is  such a useful review  of  all the
 arguments that I  have  relied heavily on it  in structuring my owa discussion
 here.  What he is  most convincing on is that  the whole subject is  incredibly
 complicated, and  that it is rather misleading  to speak of "the" property tax
 in a "typical" jurisdiction.  However, 1  feel the following remarks represent
 fairly the opinion of the majority of public  finance economists at this  time.

        1.      First, under either  the  older view or the  newer one, the
 residential property tax is more regressive or less progressive  than the
 nonresidential property tax.   Netzer  takes pains  to argue that  the
 residential part is still probably regressive in a typical jurisdiction, with
 an elasticity of property tax burdens with respect to income  of about .7,
 even though the  other parts  of  the  tax  and  the  tax as a whole are
 proportional or slightly progressive.  Thus,  we have at the start a  problem
 of context:  is it the residential tax  or the general  tax which is the
 alternative to user charges?  Clearly, it depends on the jurisdiction we are
 studying.   Most  writers assume it is  the residential  tax  which is the
 alternative, because SWM services are provided  to persons living  in  housing
 units.   However, I am sure that many  jurisdictions cannot  legally or
 politically change the residential tax rate without also changing  the rate on
 other  property.   Over  a period of  time, of  course, different assessment
 ratios for different kinds of property appear, whether legal  or not.   But
 ratios  would not  change merely because  user charges are subsituted in one
 public function; namely,  SWM.  Thus, at the margin, it appears  that  in  some
 jurisdictions the  decrease in tax rate would  have to apply to  all  property.

        2.      Neither the residential part nor the  total  tax is  as
 regressive as the conventional view of  10-15 years ago suggested it was.
 This point needs elaboration in depth, and I  return to it later.

        3.      Strictly speaking, there  is no such thing as "the"  property
 tax.  The tax is levied  at widely different rates, and the differences per se
 are relevant for incidence theory.  Even  within a jurisdiction, some  property
 is exempt, and in  addition the rates differ  greatly across  jurisdictions.
 The  jurisdictions differ in the possibilities property owners have for
 shifting the tax.   This  means the best analysis is one which  is tailor-made
 to the  jurisdiction.   It must allow for  characteristics such as assessment
 practices, possibilities  for tax exporting  (shifting to other  regions) by the
 industries whose  capital is taxed, the  elasticity of substitution between
 capital and other  factors  in those industries,  the ratios of market value to
 annual rent in each housing market,  the competitiveness of the rental housing
 market and the kinds of  individuals  who own rental  housing units, and  many
more.                                                                    J

        4.      The conclusion on incidence depends greatly  on whether one
 adopts  annual income or a  longer-run,  "permanent" income  as  the income
measure.  The tax is less progressive or more  regressive, the shorter the
 period  for which  income is  measured.   The  New View contains as  one crucial
element  a conviction that a longer-run measure  is more relevant.
                                    107

-------
        However, if one  adopts so permanent a concept  as  lifetime income, the
question  becomes  complicated by the  fact that many  families move  from
jurisdiction to jurisdiction several times in a lifetime,  and some of  them
move  each time from a lower tax base base (per capita) community to a higher
tax base (per capita) community.  Thus, even if the tax  is progressive  for
all  the families  living in a jurisdiction at any one  time, it is possible
that over a whole lifetime  the tax price of local services rises much  less
slowly than income.  Some who push this argument point  out that many of the
poorest families are confined for their entire lifetimes to central cities
where  tax bases are low,  tax rates are high, and public  services are poor.
However,  even if multiple jurisdictions in a lifetime  are important  in
analyzing the property tax as an institution,  I don't think they are very
important in analyzing the  changes in tax rates in any one  jurisdiction.

        Along the same line as the multiple jurisdiction  point, Hamilton  has
suggested  that  the  property tax ultimately is far more regressive than any
cross-section evidence on incomes and estimates of tax burdens  can suggest
(6).   That is because the property tax is part and parcel  of a comprehensive
system of local finance  in which restrictive zoning is used to limit high tax
base communities  to high  income individuals.  He  hypothesizes that  the
restrictive zoning,  which  is employed  by affluent  communities precisely
because they must  rely  heavily on local property taxes  to provide services,
raises the price of housing in poor communities. This adds significantly to
the impact of the property tax payments per se.  This argument probably
should be  ignored  in our present context, which is the financing of a single
function which does not  absorb much of  the local budget.   In other words,
changing the financing of SWM will not  in itself have  much impact on  the
system Hamilton is  describing.

        With these  summary conclusions as background,  allow me to sketch  the
basic  theoretical  arguments.   Under  the old view, the  tax decreases the
returns to capital  and land in the short run.  In the  long  run, the supply of
land is inelastic,  so the rate cannot be restored by any  shifting of land out
of taxed uses, if the tax is general.   Therefore, the land tax  is borne  by
landowners, which means it is  quite progressive.  But  the supply  of
reproducible capital has  some elasticity in the long run, and  the owners  of
taxed capital will  shift  the tax at least partly on to consumers, and perhaps
other factors, by reducing the quantity supplied until the  gross return rises
enough to  restore  the net return to equality with other  investments.  In the
end there is some fall in the.net return  to capital, but  the lion's  share
falls  on consumers  of goods, and services, and disproportionately on housing,
a very capital-intensive  good.  As the elasticity of general consumption with
respect to current income  is  below one,  and the  elasticity of housing
expenditures well below one,  the tax on reproducible  capital is seen  as
regressive  and the  residential property tax especially so.

        The New View comprises two parts.  First, the  theoretical derivation
of the incidence of  a general tax at  a uniform rate on all capital and land;
second, the adjustment of the theoretical results to allow  for the fact  that
the tax is not actually levied on all capital, and not  at a uniform rate on
the capital which is  taxed.
                                     108

-------
        A key assumption In the New View is that the supply of reproducible
capital to the economy is inelastic, even  in the long run,  just  as is the
supply of land.   In that case, a general  and uniform tax on capital must fall
on the  owners of capital.   The inelasticity of supply eliminates  the
possibility of  shifting  the burden to  consumers or to  other factors of
production.   As  the earnings from capital are concentrated  in the higher
income groups, this means the tax is quite progressive.

        In the real world, the tax is not uniform.  Some capital is taxed at
much higher  rates  than other capital, partly because of intra-jurisdietional
variation and partly because of inter-jurisdictional variation.   The owners
of capital which is taxed at more than average  rates can — and will  —  shift
some of the  burden by moving their capital out  of the high tax sectors and
into the low or  zero tax sectors; that will raise the prices  of  goods and
services  produced using the heavily taxed  capital.  The more uses  of  capital
are taxed, and the less the variation in  tax rates,  the less  of the burden
can owners shift off themselves.  Of course, the process will result  in lower
prices for goods using less heavily taxed capital.  To the  extent  there are
such  shifts forward on  to  consumers,  the incidence of  those excise-type
effects will depend on how  the consumption of various goods  and services
varies with income.    Some  backward  shifting is  also  possible, as the
restriction  of capital in some industries and some localities may reduce the
demand  for  that labor which is relatively immobile and must accept reduced
wages to gain employment.   However, these backward shifting  effects are
probably  quantitatively minor compared  to the forward shifting or excise
effect.

        Housing is very  capital intensive,  and the  property tax  rates on
capital in housing appear to be higher than average.  That  means the excise
effect  will raise the price of housing  to a significant  degree. From a
practical standpoint, housing is probably the most important  good the price
of which  is raised or lowered by the  excise effects of differential tax
rates.  The  fact  that the price of housing does rise, therefore, is a central
fact  under  the  New View,  just as it was under  the older view; however, the
New View stresses  that the rise will be less than  formerly  thought,  because
the  fairly  widespread taxation of capital and  the presumed inelasticity of
supply reduces shifting.  Furthermore, the same process of  forward shifting
which increases  the price of housing will reduce  the prices of lightly  taxed
goods, the consumption of which may be  roughly proportional to permanent
Income.  For consumers as a whole, this may offset the housing effect.

        Break has  concluded:

         "Property tax burdens  on  consumers  ... are probably less
         important than commonly supposed. They are a  function only
         of  interregional and intraurban tax  differentials and not
         of  the  total property tax rate; the excise tax effects
         generated by  those tax differentials contain  labor and
         landlord burdens  which reduce  the pressures making  for
         consumer burdens; and such national consumer burdens as do
                                    109

-------
          exist  fall not on consumers  In general but  only on those
          with  relatively strong tastes for the  output of the
          industries that are more heavily  taxed than  others."  (4,
          p.  165.)

         The New View  not only has a different conclusion  on the extent of
  forward shifting, it  also has a different  concept of income.  It  favors a
  "permanent" notion  of  income, rather than  a single year's  income.  The
  empirical studies of housing demand  show  that  the elasticity with respect to
  permanent income is considerably higher than the elasticity with  respect to
  annual  income  — it  is perhaps unity or even  slightly higher.  Thus,  even if
  housing consumers paid all  the residential property tax, it  might not be
  regressive!

         Somewhat  strengthening this conclusion  is  other empirical  evidence
  that  the  housing market produces ratios of market value to annual rent which
  rise with income.  If the annual rental is proportional to permanent  income,
  the value  then rises more  than proportionately.   If the  property  tax is
 proportional to value, then the tax rises more than proportionately as  well.

         Thus,  the New View has several strings to its bow.  The widespread
 taxation of capital  limits  the shifting by capital owners; the  income
 elasticity  of  housing expenditure  with respect to permanent income is  around
 unity; and the  value of property, which is the tax  base,  rises with  income
 even  faster than  housing rental expenditures.  And even if  there is  a
 regressive housing effect,  there are some offsetting effects for the  average
 consumer from lower prices  for  other goods and  services.

         Against these  there  are  some counterarguments,  a  few of  which have
 already  been  mentioned.   The housing occupied by lower income  families may be
 systematically  overassessed, which offsets the  value-rent  ratio phenomenon.
 Many feel the  permanent income  is  not  a solid  enough empirical concept on
 which  to rest a policy decision.  Others Accept the  permanent  income concept,
 but feel that if  properly  extended  to lifetime  income  it  points  to
 regressivity; this is  the multiple-jurisdictions-in-a-lifetime phenomenon
 mentioned earlier. Another consideration  is  the possibility that much rental
 residential property is owned by  low and middle income families, who rent out
 parts  of their own homes  or who  rent out  at most  one  or two houses.  If
 rental property  owners are lower  income people, and their  tenants are also
 lower  income people,  it  doesn't matter  how much of  the property  tax on the
 property is shifted forward and how much reduces the net returns to owners.
 Finally, the deducibility of property  tax for Federal income tax purposes
 means  that the net impact on higher income  owner-occupiers is reduced.  One's
 feeling  on  how  important this  is  depends on whether  he thinks the
 deductibility feature  should be analyzed as a feature  of the property tax or
 as  a feature of  the  Federal  income  tax.  It  is true that a change  from the
 property tax  toward user charges will  have  a blunted favorable effect  on high
 income families because part of the reduction  in property tax will be offset
by higher Federal  income tax.  In an empirical  analysis, one  must remember
 the very recent trend toward raising  the standard deduction and  reducing the
number  of Federal  taxpayers who itemize deductions.


                                    VI0

-------
        By now,  if you have never gotten  into  this  literature before,  your
head may  be  swimming.  If you have gotten  into it, your head may  be  swimming
anyway,  for you  may be nagged by still additional considerations which I have
not had time to mention.   The safest  conclusion, I think, is Netzer's:
"There  seems no substitute for empirical work  that  recognizes local
differentials and perhaps, no case for the kind of nationwide estimates of
incidence  by  income class that I and others have made in the past."  (13,  p.
519.)

        However, if one does insist on generalizing  from national  studies to
a particular jurisdiction, say a "typical" jurisdiction, one could  conclude
with Netzer that the property tax on the  whole has an elasticity with respect
to income  of  about 1.1 and the part on residential housing and consumer-owned
durables an elasticity of about  .7.  Netzer makes these ball-park estimates
on  the  assumption that  income is measured  as annual income, but with some
modification at  the lower end of the scale to  raise  incomes  there above the
ones  reported as annual incomes.  This modification is to allow for  the fact
that many poor people are elderly homeowners  who have a high asset-income
ratio, for the fact that some other poor  people are  poor only temporarily  and
also have comfortable asset  positions or  at least other  reasons  for higher
income  prospects, and  for the underreporting by  some poor people receiving
public assistance.

CONCLUSION

        As noted earlier,  the empirical studies of the income elasticity of
solid waste volume and collection  cost  produce estimates  which are about
equal or  a bit  lower than the Netzer  guess on residential property  tax
burdens.  It is  unlikely that the income  elasticity of  solid waste  is above
unity.   This suggests  that the vertical equity  of  user charges and  the
residential property  tax are not much different  from each other.   Indeed,  it
hardly  matters  anyway,  because  the dollar amounts  are very small to start
with.  Therefore, there  is no  reason to fear any noticeable  increase  in
regressivity from substituting  user charges for  the property  tax, especially
if  it is the residential property tax which is being replaced.

        Nor are there valid  arguments, in my opinion,  against  user charges on
grounds of "merit goods"  or  on  grounds  that  society  should  want  to
redistribute income  in a way  dependent on poor persons'  use of  solid waste
management services.   As  I have  said  earlier, I do not think we can  apply to
solid waste  the notions which justify specific price  subsidies on medical
care, housing, and education.   In  fact, I would  be  prepared to argue  the
opposite,  that persons  should be encouraged, even  at  some  cost  in  terms of
efficiency, to  take  care  of their  own waste.  Society  as  a  whole  should  want
to  encourage habit  of  resource  conservation,  for one  thing, and  the price
system  in  solid waste may be useful in this respect.
                                     Ill

-------
                                 REFERENCES


 1.   Aaron,  Henry,  Who Pays the Property Tax.  Brookings, Washington:  1975.

 2.   Bolton,  Roger, "Household Income and Refuse Generation," unpublished.

 3.   Bolton,  Roger, with P. M.  Meier and  J.  Kuhner,  An Assessment of Wet
      Systems for  Residential Refuse Collection:   A Case  Study  for
      Springfield, Massachusetts, for  U.S. Environmental Protection  Agency,
      1974  (NTIS).

 4.   Break, George, "The Incidence  and Economic Effects of Taxation," in Alan
      Blinder  and  Robert Solow,  £t al,   The Economics of Public Finance.
      Brookings, Washington:  1974,  pp. 119-240.~~

 5.   Downing, Paul, "Intra-Municipal Variations in the Cost of Residential
      Refuse Removal,"  and "The  Distributional Impact of User Charges  for
      Refuse Collection,"  unpublished, mimeo.

 6.   Hamilton, Bruce,  "Local Government,  the Property Tax,  and the Quality of
      Life:  Some Findings on Progressivity," in Lowden  Wingo  and Alan Evans,
      eds., Public Economics and  the Quality of Life,  Johns Hopkins Press  for
      Resources for the  Future, Baltimore:  1978, pp.  111-122.

 7.    Kemper,  Peter and  John Quigley,  The Economics of  Refuse Collection.
      Ballinger, Cambridge, Massachusetts!1976.

 8.    McClure,  Charles, "Revenue Sharing: Alternative  to  Rational Fiscal
      Federalism?"   Public Policy, Summer 1971,  pp.  457-478.

 9.   Mieszkowski,  Peter, "Tax Incidence Theory:  The  Effects of Taxes on  the
     Distribution  of Income," Journal of  Economic Literature. December 1969,
     pp. 1103-1124.                                    ~~

 10.  Mieszkowski,  Peter,  "The  Property  Tax:  An Excise  or  a Profits Tax?"
     Journal  of Public Economics, April 1972, pp. 73-96.

 11.  Musgrave,  The Theory of Public Finance. McGraw-Hill, New York:   1959.

 12.  Netzer,  Dick, Economics of  the Property  Tax. Brookings,  Washington:
     1966*

13.  Netzer,  Dick, "The Incidence of the  Property  Tax Revisited," National
     Tax Journal,  December 1973,  pp.  515-535.                      	

                                    112

-------
14.  Sheaffer,  John R.  and George S. Tolley, Decision Making and Solid Waste
     Disposal,  Center  for Urban  Studies, University of Chicago,  1971.

15.  Sheaffer,  John R.  and George Tolley, Socio-Economic Factors  Affecting
     Demand for Municipal Collection of Household Refuse (condensation),  U.S.
     Environmental  Protection Agency, Cincinnati, Ohio:   1973.

16.  Thurow,  Lester,  "Cash Versus  In-Kind Transfers," American Economic
     Review,  May 1974,  pp. 190-195.

17.  Tiebout, Charles,  "An Economic Theory of Fiscal  Decentralization," in
     Public Finances;   Needs, Sources, and Utilization,  Princeton University
     Press for  National Bureau of Economic Research, 1961,  pp.  79-122.

18.  Tobin, James,  "On Limiting  the Domain of Inequality,"  Journal  of Law and
     Economics, October 1970, pp. 263-78.

19.  Tolley,  George, V.S. Hastings, and G. Rudzitis, Economics  of  Municipal
     Solid Waste Management;  The Chicago Case, U.S. Environmental  Protection
     Agency,  Cincinnati, Ohio, 1978.
                                     113

-------
   PRICING MUNICIPAL REFUSE  SERVICE:  POTENTIAL  EFFECTS ON MUNICIPAL BUDGETS
                                    by
                             KENNETH L.  WERTZ
        I wish to  examine  how  two approaches  to pricing  household's
generation of refuse  could affect municipal  budgets.  The  local  pricing
approach contemplates  a municipal charge  system for residential service under
which a property owner  would pay more  as he  presented more  standard
containers of refuse  for collection and  disposal.  The  federal  pricing
approach would create  specific excises against certain products  used in  the
packaging of consumer goods and distribute  the revenue to local governments
to assist their financing  of refuse service.

        In the opening sentence  I emphasized "could" because it  connotes
potential and uncertainty of  result.  The  budgetary impact  of  a  pricing
program depends not  only on the responses  of  households but also on decisions
made by municipal  officers.  So long as the  dollar amount is not trivial, the
introduction of a new component of revenue  creates the leeway for  managerial
discretion and may quicken  the propensity  to  use  it.  I cannot  predict  the
specific  responses (and  their budgetary  images) of municipal officers who
operate under heterogeneous conditions in- the nation's 18,517 municipalities,
but  I do  endeavor  to  describe what the opportunities might be.  I utilize
many approximations  to  suggest the order  of magnitude of certain effects, and
indicate  the method of  calculation in the event others  care  to employ
alternative  estimates.

        The first section looks  at  the  revenue side of the budget, and the
second at  the expenditure  side.  A final  section concludes with a bit of
speculation.

REVENUE SIDE

        The  revenue  yield of a pricing  system  depends on the  charge
structure,  the height of the charge,  the base to  which it applies,  and  the
reactions  of consumers and  producers.   With a  federal program there is
additionally  the matter of how revenue is  to  be returned to municipalities.
For  concreteness  assume  that  a municipality's total revenue from a pricing
system, whether locally or  federally administered, just equals what  the
municipality was  previously  spending for  residential refuse  service.  The


                                   114

-------
 basic question for municipal officers is whether  these receipts will  be
 allowed to change the  total amount of local governmental revenue from own
 sources that had  previously been planned for  in  preliminary budgets.  The
 polar  possibilities  are  that charge receipts are treated as an addition  to
 own revenue or are used  to fund reductions  (or to defer planned increases)  in
 local tax rates.

        If own tax reduction is chosen with the  intent of holding constant
 total  local governmental  revenue from all sources, then municipal officers
 must plan on reducing  own  tax revenue by an  amount that  is less than
 anticipated charge receipts.  This is so because some intergovernmental
 transfers, of which local governments are the net beneficiaries, are allotted
 according to formulae  that  reward a local  government  for collecting tax
 revenue but not charge revenue.  A universal  case in point  is the federal
 general revenue-sharing  program.  To a first approximation, the grant to a
 unit of local government varies directly and  proportionately  with the
 magnitude of its "adjusted" tax collections.  Thus if a municipality were  to
 replace X% of its own  tax  collections with charge revenue, its federal
 revenue-sharing   grant  would decline by X%, other things  being equal and
 assuming  that the municipality's allotment is not determined  by the
 provisions governing maximal and minimal allotments.  A rough estimate  is
 possible for the  aggregate of municipalities.  In 1975 all  municipalities
 combined  raised $21.1  billions in own  taxes  and  $3.6  billions  in
 miscellaneous own general  revenue, of  which  $3.6 billions  were  spent for
 education (municipal expenditures for education were actually $7.2 billions
 in 1975, but only 49% of local governments' expenditure for  education  came
 from own  sources).  Thus "adjusted" taxes — the adjustment excludes revenue
 raised  for education expenditures — were about $21.1 billions. If we  take
 municipal  outlays for  residential refuse service and  financed from local
 general revenue to have been $1.2 billions in 1975 (actually,  $1.8 billions
 was spent  for "sanitation other than sewerage,"  but  some  of this was not for
 residential service and  some  was not financed by own taxation), then its
 replacement by  an equal amount of charge revenue would  have implied  a
 reduction in revenue-sharing grants to all municipalities  of  about 6%.  The
 dollar  loss would have been $125 millions, since all municipalities received
 $2.2 billions in federal revenue-sharing payments in that year.  The
 calculation indicates  that  on average a municipality could enact local tax
 relief  of  $.94 and  lose $.06 in intergovernmental transfers  for each  dollar
 collected  in charges, if total revenue from all sources  is to be unchanged.
 These  figures will differ from place to  place, depending primarily  on the
budget  share of public residential refuse service in  a given municipality and
 on higher  order interactions in the workings of the revenue-sharing formula.

        For municipalities in some states even less local  tax relief could  be-
afforded because  of state-local intergovernmental flows. For example,  in
 1973 the  Michigan  state government  distributed  $54  millions to local
governments according to formulae that increased  the  share of any particular
1.    Outlined at Advisory  Commission on Intergovernmental  Relations,
     Federal-State-Local Finances, 1974, pp.79-81.
                                    115

-------
 local government  as  the  latter's own tax collections increased.   Similarly,
 the North Carolina state government distributed $26 millions  according to
 local  property  tax  levies.   Such programs*  were also  to be found in
 California, Ohio,  and  Wisconsin in 1973,  and it  is  quite possible  that their
 adoption has since spread  in  imitation of the federal general revenue-sharing
 program.

        The discussion has so far regarded charge revenue  as a source of
 local tax reduction, but it applies as well  if  charge revenue  is meant to
 enlarge local revenue  from all sources.  When an extra dollar  of  revenue is
 raised through charges rather than taxes, the  local government foregoes the
 bonus  that higher  level governments will  pay on  the latter.

 EXPENDITURE SIDE

        One obvious difference between a  local and  a federal  pricing system
 is  the incidence  of  the  costs of running the  system.  S. 1281 (95th Cong.,
 1st sess., 1977) to my reading would erect no condition that should occasion
 signifcant  municipal expenditure solely for the purpose of  making the
 municipality eligible for the return of charge  revenue  from the federal
 government.   A local pricing system,  however, is sure to cause additional
 local  expenditure  for  administration.  It is necessary to have  an account
 with  each property owner; to publicize  rates and  information about standard
 containers;  to bill,  collect, and perhaps issue  receipts; and to adjust
 disagreements about the correct charge.   Although municipalities have  already
 opened accounts  for purposes of property  taxation  and although  they could
 economize by  combining billing with tax or public utility billing, extra work
 and extra expenditure  must  still be expected. Furthermore, it is necessary
 to  observe  and record the number of containers  emptied at each property.
 This task involves  additional expenditure too, for  such data are not compiled
 where general revenue  financing is employed.

        There is  nothing incorrect about this fatiguing recital  of
 conventional wisdom.  But it takes too narrow a view of  "administrative
 costs" by ignoring  what they could buy.   One return on the costs of running a
 local  pricing system is  the regular  collection of data which might  be used
 for the more  efficient supply of service.   The  quantitative importance of
 this  potential is unknown,  but  the state of  municipal recordkeeping that
 researchers  have encountered is at the very  least  suggestive that  that
 potential  remains largely unrealized.   Microdata on refuse generation would
make it possible to  search thoroughly for  cost  reducing route designs and
 factor assignments.   Aggregated and  kept  over time  the data could be  used to
 sharpen planning for capital expenditures,  manpower adjustments,  disposal
 site  acquisition,  and  adjustments  to seasonal  variations in  refuse
generation. Efficiency savings from these  sources, assuming that  they are
 realized  through managerial action, are  properly considered  as offsets
against the direct  cost of  operating a local charge  system.  The federal
 product  charge  system generates no comparable information,  and no comparable
efficiency savings  should be expected.

        A second type of return on administative costs is  an  Intangible that
appears in no governmental budget,  even  though much public spending and

                                    116

-------
revenue raising  are  undertaken in its name.  The intangible is equity.  The
content of the term changes  from one application to the  next, and unanimity
about  its proper  content in any one  application is rare.  These realities
understood, we may note  that  some of the enthusiasm for  pricing proposals  is
based  on  the view that a person  ought to pay for service in proportion  to the
costs  which his  refuse generation visits upon  the  public treasury.  The
regular  satisfaction of that norm under local  tax  financing,  which  is
essentially  property tax  financing,  is  just  not in the cards:  The
connections between determinants of  refuse generation and property value  and
the connection between assessed value and property value are too irregular.
How the burden of local  taxation for refuse service might be distributed  is
scrambled another turn  by the deductability of local property taxes (but not
charges and then only  for itemizers) from  the  federal (and many  state)
individual income tax base(s).  A local charge system,  on the other hand, is
predicated on a clear and direct connection between payments made and  costs
imposed,  from one person to  the next, provided that the logic of the  system
is  carried out.  Administrative costs for running a local charge system  are
then costs of producing  horizontal equity.  "Local" is  emphasized because the
same results are unlikely to  follow  from a federal product charge system that
returns revenue  to local governments on a capitation basis.2  No systematic
relief from current inequities can be expected if municipal officers utilize
their  federal grants for improved  refuse service (as contemplated by S.
1281), because neither the magnitude  nor the distribution of local  taxes
would be affected, and they are the putative sources of  the inequities.  Some
systematic relief  would  occur  if municipal officers were instead to utilize
federal grants to  fund reductions in local taxes.  But  in either case federal
pricing would create new variances between payments made and costs imposed,
because product  charge collections are based on refuse generation while the
return of revenue  is based on  population, and considerable variance appears
to  exist in per capita rates of generation from one place to  the next.

        In summary,  the potential  for  efficiency  savings from data
availability and  an  improvement in  horizontal  equity are direct offsets
against administrative costs of managing a local pricing system.  It  is  not
possible to arrive at  a  monetary figure for net administrative costs, because
one of the products is an intangible and will be valued  differently even  by
persons in the same locality.

        The collection and disposal of refuse is the second major area  where
pricing is likely  to have a noticeable  effect  on the expenditure side of
municipal  budgets.  Federal  pricing would create a private marginal cost
against one's generating  extra refuse.   Local  pricing would create a private
marginal cost against  one's presenting extra refuse  for public  collection.
Local  tax financing of  service does neither.   Hence  pricing, relative to  the
status  quo, would  reduce the amount of refuse which  the municipality  would
have to treat, and  service with constant features (frequency, place  of
2.   This position  is elaborated at K. L. Wertz,  "Equity Impacts of Proposals
     for Federal  Product  Charges," Journal  of Environmental  Economics and
     Management,  Dec. 1977.
                                    117

-------
 collection, etc.) could be supplied  for  a  smaller municipal outlay. 3  Since
 the charge-induced reduction in operating costs is to be realized  year after
 year,  it could function as an endowment  for programs whose budgetary impact
 would  last over many periods.  The cost  savings could be returned to  citizens
 through reduced local tax rates, distributed to municipal employees as higher
 effective wages, applied to  providing refuse service of a higher quality, or
 applied to augment other municipal programs.

        An order of magnitude estimate  of the charge-induced reduction in
 operating costs is possible through  a succession of approximations.   I  have
 elsewhere  instanced the fact that  in 1970 per  capita collections  of
 residential refuse in San  Francisco,  where pricing is practiced,  were 25%
 less by weight than the per capita figure  for all urban areas  combined, where
 pricing is very seldom practiced.  This is a gross comparison; it  does not
 adjust, up or down, for 'other possible influences on refuse generation.   Let
 us stipulate a 20% reduction to follow  from the introduction  of pricing, the
 downward adjustment made because the height of the  charge in San  Francisco
 probably exceeds  the average level  that would  be contemplated in other
 municipalities.  The second figure that  is needed is the elasticity of total
 cost  of refuse  service with respect to refuse tonnage.  Barbara Stevens has
 produced estimates for collection costs  that range from 0.8 to 1.0, depending
 on city size.    Let us use  an intermediate value, 0.9, and assume  that it
 applies to disposal costs as  well.   The  conclusion, then, is that  pricing
 could  result in an 18% decrease in annual operating costs.

        The significance of this figure  in the aggregate can  be better  seen
 by carrying the  approximation through a few  more steps.  Assuming  that
 municipalities as a group are now spending about $1.8 billions per  annum for
 residential service, an 18%  decrease equals annual savings of $0.32  billions
 in 1978 dollars.  Assume finally that the current municipal borrowing rate is
 6.67%; that savings in operating costs are realized in perpetuity;  and that
 both the nominal savings and  nominal borrowing rate grow at the  same  rate as
 inflation generally.  Then the present value of the charge-induced reduction
 in operating costs to all municipalities combined would be $4.8 billions in
 1978  dollars.   This is a large amount, but not an imposing one.   It  is, for
3.   The  metering method which is  employed under  local pricing could
     contribute  to savings in  operating costs,  independently  of any
     charge-induced effects on the output of refuse.  For example, if the use
     of standard  disposable  bags were required  in lieu of  conventional
     containers  so that quantity for billing  purposes is determined by the
     number of  bags set out, then the costly return of empty  containers  from
     street  to dwelling would  be eliminated.   Such  possibilities are not
     considered in  the estimates in the text.

4.   K. L. Wertz, "Economic Factors Influencing Households'  Production  of
     Refuse," Journal of Environmental Economics and Management." Apr. 1976.

5-   B« •!•  Stevens, "Scale,  Market  Structure, and the  Cost of Refuse
     Collection," Review of Economics and Statistics. Aug. 1978.

                                    118

-------
example,  only double the amount that all municipalities are now receiving in
federal general revenue-sharing grants  in one year.

       The  use of the foregoing analytical  structure will probably  lead to
underestimates of the charge-induced  savings in operating costs  experienced
by any one municipality in the event that a  majority of municipalities adopts
local  pricing  or that federal pricing is installed.  It would  then become
reasonable  to  think of secondhand markets becoming more active and of
producers taking the existence of charges into account when making decisions
about packaging,  two changes that would reduce the amount of refuse  which a
municipality collects and disposes of.

CONCLUSION:  A  RACE OF ONE?

        A municipal officer who might have earlier considered the  adoption of
a local charge  system could have framed the  matter this way.   "I  get charge
revenue,  a  modest reduction in operating costs,  the likelihood of  cost
savings due  to  data generation, and much better  satisfaction of a  particular
distributional norm.  In return I incur a certain increase in administrative
costs and the equally certain unpopularity  of  having proposed a new % tax'
(even  if  existing taxes  were adjusted so  that revenue collections from all
sources were held constant)."  However  attractive this package  might  once
have been, its  attractiveness probably evaporated  with the recent  legislative
advances of  the federal proposals  for  product charges.   From a municipal
officer's perspective, or so I should think, the federal plan would have none
of  the liabilities of a  local pricing  system.   Under federal  pricing
administrative  costs  exist  but  are   essentially borne elsewhere.
responsibility  for  the enactment of  a  new  revenue program can also be put
elsewhere.   The risk of increased littering that goes  with local pricing is
removed by federal  pricing, and the risk of  having promoted a bad idea—if in
fact pricing fails  to  induce much  reduction in the  output of refuse—is
shifted to other  parties.  Yet  the  federal  plan would  provide revenue and
stimulate whatever  cost-reducing decreases in waste  generation there are to
be  realized.  Only the potential  efficiencies that arise from  data generation
and the  increased  equity  in  finance that are unique  to  local  pricing need be
given  up.  As witnessed  in other  intergovernmental aid programs, the prospect
that  the federal government may in the  future bear the costs, broadly
conceived, of providing  many of  the  benefits that  could be had  by  local
action is probably  enough to retard local action  now.   I suspect that the
contest of pricing propoals  is now a race of one.
                                    119

-------
                                 SECTION 4

               COMPARISON OF USER CHARGES AND PRODUCT CHARGES


"The Optimal  Policy Mix  for Solid Waste Pricing" (James F.  Hudson).

"Fundamental Comparisons  of Alternative Solid  Waste Management Policies"
(Allen K.  Miedema) .
                                    120

-------
               THE OPTIMAL POLICY MIX FOR SOLID WASTE  PRICING
                                    by
                              JAMES F. HUDSON
INTRODUCTION

        It would be difficult to claim a connection between the concept  of
"optimality" and  the  field of solid waste management, despite the title of
this paper.  The handout  from our recent work for the  Resource Conservation
Committee, and other  papers in this  seminar, summarize the state of the
theoretical and practical knowledge  about solid waste  management and
potential pricing  policies.  According to these papers, there is little
likelihood that we  will  approach a mix  of policies  which could be  called
"optimal", as defined by  economists.

        Still, it may be  possible to  locally optimize  our policy choices,
given  the set of options which appear politically  feasible.  It is this more
limited sort of optimization which this paper addresses.   It describes steps
which can be taken  towards  optimization, and it defines  reasonable goals  in
this regard but stops  short of determining  some form  of globally optimal
policy mix.

        This  paper will ignore  issues such  as the  assumptions behind the
optimality of marginal  cost  pricing, the requirement for constant returns  to
scale,  and the declining marginal benefit curve for pollution.  Further, the
paper will not repeat the descriptions of research  contained in the handout,
but rather will focus on  the conclusions arising  from the  research, and the
potential for policy improvements in  the field  of  solid  waste management.
Although the conclusions  reached in this paper are  somewhat  speculative, they
are based on the author's experience in the field of solid waste pricing and
management.

APPROACHES

        There are two basic  economic approaches which  have been suggested for
reducing the quantity  of solid waste  reaching•final disposal; these are
product charges and local user charges.  Both act through  reducing the amount
of product-related  wastes,  and through increasing the  amounts recycled.  This
paper  will argue  that,  under a reasonable—at least to the author—set of


                                   121

-------
 assumptions,  the  interactions can be generally beneficial  and  conducive to
 achieving  efficiency and  equity goals.   It will make arguments related
 particularly  to the incentive structures  of  the various actors in the system,
 and the directions of their probable responses.

         The effects of pricing policies are  described schematically in Figure
 1, along with  the impact points of the charges.  While this  schema represents
 a simplification, it leads us to suggest  a number of actors  in the process:

         •      extraction firms;

         •      production firms;

         •      fabrication firms;

         •      wholesaling/retailing;

         •      households and other  consumers;

         •      solid waste management firms and agencies;

         •      secondary materials  retailers/wholesalers;

         •      state and federal  governments;

         •      equipment suppliers  for waste management.

Product  Charges

         Product charges act  relatively early in the cycle,  as  described in
the handout.   They  will reduce  the level of output and  revenues  for
extraction firms.   This is  intended to correct the inefficient  or  excessive
output levels  prevalent prior  to the  imposition of charges.  There will be a
direct  incentive to  increase the  production and  fabrication of  recycled
materials and/or  to  decrease the use of materials  for  packaging.  Price
increases will be passed on to  consumers,  leading to  a  loss in  consumer
surplus, which could  be  counterbalanced somewhat  by  reduced  solid waste
management costs.   In addition,  the product charge  revenues  will be
distributed to local  governments, to offset solid  waste  management costs, or
increase the   spending for other social  services,  leading to  (it is  hoped) a
net benefit consumer  surplus.  The local  solid waste mainagernent  agency will
be  able to lower cost,  and may receive revenue transfers from the system.
Recycling firms at the wholesale and retail levels will  see  increased demand,
and probably  higher prices;  current users of the wastes, however, will also
have to pay higher prices for  their goods*

         In  essence, the product  charge  provides  revenues to local
governments, increases the demand for post-consumer wastes,  and provides an
incentive for firms  to modify their use  of packaging, with bonuses for forma
which minimize waste-handling  costs.
                                    122

-------
                                                    Extraction
                                                    Production
                  Key:

           — Original material flows

           ~ Revenue flows
                                                     Fabrication
CNJ
OJ
                          Recycling
                                                       Use
                                                    Collection/
                                                    Disposal
Litter
                            Other
                        Self-Disposal
                                       Figure  1.   Impact points of pricing options.

-------
 User  Charges

        User charges related  to  the  quantity of waste may provide  some
 incentive for consumers or firms to change their consumption patterns and  to
 select products  with relatively  small contributions  to  the wasteload.  The
 basic  incentive is  to reduce  the  quantity  of  waste placed  for
 collection/disposal: there are a number of methods to achieve this.  One
 method  is to increase product lifetimes:  this is likely to result in reduced
 consumption, with potentially significant effects on waste quantities.  Other
 techniques include at-home recycling (such as composting yard waste); product
 recycling; littering (including disposal at workplace,  etc.) and other forms
 of  disposal such  as direct hauling  to the  disposal  site, or  outside
 contracting.   Inventory increases, for example of  newspapers, are not going
 to  affect total waste quantities for the long-term,  though they may lead  to
 change in the  amount of collected waste.1  The handout  suggests in Section
 4.3 that it is relatively easy  to achieve 30-50% reductions in wasteloads  if
 yard  wastes can be used  on-site, and if recycling programs are available for
 beverage containers and newspaper.  With user charges,  recycling firms should
 see an increased  supply of waste.   Collection/disposal costs may go down;
 litter collection  costs, on the  other hand, might  rise,  depending  on
 household behavior.   The direct  incentive of the charge is to reduce the
 quantity of wastes placed for collection/disposal.

 MARKETS

       There are two submarkets  of interest in analyzing  solid waste:  the
 traditional production/consumption  submarket, and  the disposal/recycling
 submarket.  We are primarily concerned with the operation  of the latter, and
 the interactions between  the markets,  which occur through:

        •       transfer  of revenues (e.g. product charges);

        •       changes in consumption caused by disposal  practices;

        •       supply and demand for post-consumer  wastes for re-
                cycling.

        The disposal submarket consists  of  the household engaged  in
 post-consumption  waste disposal and recycling,  the  firms  or agencies charged
 with  solid  waste collection and disposal, litter management,  and the
 retailing of recycled materials.

 ANALYSIS OF  SYSTEM COMBINATIONS

        To identify  a  desirable mix of  policies, the effects of combinations
of product charges  and  incremental user charges  should be  considered.  These
 include  effects on  consumption, household  income, levels of recycling (and
prices for recycled products), operations  of the  extraction/production/
 fabrication firms,  operations of the  solid  waste collection and disposal
agencies or  firms,  and  requirements for other forms  of disposal.
                                   124

-------
Effects on Consumption

        As mentioned above, the product charge is intended to  have an effect
on consumption,  shifting the purchases  of  cost-conscious households away from
heavily packaged products.   The result should  be a shift  in consumption
towards efficiency of at least the materials affected by product charges.

        Because  of the averaging that takes place in creating a =an or bag of
solid  waste,  and the low  cost per item, I do  not expect any change  in
consumption to result from imposition of incremental user charges.  Product
lifetimes may  be increased somewhat, but  I would  expect this effect  to be
relatively small.

        With respect  to  consumption, then, there may  be  some  adverse
interaction of the product  charge and the deposit system, but this will be
slight.

Effects on Household Income

        With the product charge  households will  experience increases in
prices with an effective loss in income somewhat less than the  product  charge
payments (because of shifts in consumption).  The distribution  of this  burden
across households depends upon their consumption patterns.

        Product  charge  revenues  will be used to  defray the  administrative
costs of the program,  and to provide funds  to  state and  local governments
which will either be used for reducing  solid  waste disposal charges,  property
tax relief, or other  social  programs.   If  we  assume that consumption of
product-related waste rises with income,  i.e. the price increase  burden is
progressive, and that the  property tax  and  waste charges are  relatively
regressive, while the social programs  are  progressive, then  the  net  result
should be a progressive shift in the distribution  of income.  In addition, if
the administrative costs of the product charge are minor, then  adjustments in
consumption would probably  lead to a net  welfare gain, resulting from the
improvement in allocative efficiency.

Levels of Recycling

        The product  charge will  increase the demand for product-related
waste, because of the  lack  of charges for reused materials.  The incremental
user charge can be expected to increase the  supply of  such wastes  provided by
households, by giving  them  an incentive to dispose of  their  wastes by methods
other  than collection.   If the product  charge is on  the order of 1/2
cent/pound, then the  price increase  should initially be on the order of
$107 ton.   This will have little effect  on recycling  of metals such as
aluminum, but should  affect other  materials,  such as  glass, steel cans, or
newspapers.  The price  support  may  also  lead  to  use of these materials as
base-load sources,  rather than  marginal  ones,, and  smooth  out  the price
variations described  in the literature.

        On  the  other side of  the system, recycling cans,  bottles, and
newspapers will result in greatly  reduced wasteloads,  and  reduced solid waste

                                   125

-------
 collection costs.   The  easy availability of mechanisms  for recycling, such as
 separate collection,  will  reduce  the  chance of diversion to other  less
 desirable modes of  disposal, such as littering, and lead to benefits from the
 incremental user  charge scheme.  Therefore,  the product charge will aid in
 the diversion of waste  from  collection  to recycling.   The imposition  of
 incremental user charges should reinforce this effect, particularly in areas
 where recycling is  available.   Reductions in  waste collection costs,  and
 increased availability of revenues for municipal recycling programs, may also
 result.

         If the incremental user charge is set at $.50/bag, or about  $30/ton,
 then  the total incentive for recycling will be $40/ton for  the household if
 both systems are implemented.  However, unless  the  recycling  firms can charge
 households for the materials collected, they will see a price increase of
 $10/ton,  and, probably, an  overabundance of materials  available  for
 recycling.  The interaction of the  two systems will mean that sufficient
 supply will be available, with demand  likely to remain the  constraint.   A
 related  effect is that, without the product charge, it will be difficult to
 recycle all the materials removed from the collection/disposal stream.   This
 is a major problem with the incremental user charge system, implemented
 without the product charge.

 Effects on  Production Firms

        The incremental  user  charge  is  not expected to affect  these firms,
 except  slightly  through the purchase  of disposable  containers.   Therefore,
 there  is no major interaction  here.

 Effects on  Solid Waste Agencies and Firms

        This  is the area where  most  of  our research has  been directed.   At
 present, we believe  that most  systems  operate fairly inefficiently,  and that
 the elasticity of cost  with respect to  waste generation would  be:

        •       0.2  for inefficient firms;

        •       0.5  to  0.8 for efficient firms;

        •       0.7  to  0.9 for firms unconstrained by  the real world;

 The  latter are a special  case, similar to the frictionless plane  of.physics:
 firms and agencies are  not constrained by actual vehicle  sizes,  the slack
 related to the  need for  fixed routes, or the variation in load  from week  to
week;  instead,  they can operate efficiently  under  any  condition,  and
 therefore can make the best possible adjustments to  reductions  in the level
of waste to be collected.  The "efficient" firms go  as  far  in  this direction
as is  practicable, while the inefficient firms make no  major adjustments in
the face of changes  in wasteload, except reductions  in overtime and equipment
wear.

        For major reductions in wasteload, as might  occur with introduction
of incremental user charges,  an elasticity of 0.5 would appear to us to be

                                   126

-------
reasonable.   Changes  such as this are usually simultaneous  with other major
system  revisions, hopefully in the direction of  efficiency.  Therefore,  the
result should he  an initial capturing of the potential  savings in many cases.
If the wasteload  reductions are large, for example 40%,  then the cost savings
will be on the order  of 20%.  System revenues, though,  would drop 40% if  the
charge was set to cover the before-introduction average  cost, and there would
be a net deficit.

        In order to  eliminate this deficit, an increase in charges of 33% is
required, given the elasticity of 0.5.

        If the product charge  revenues were distributed according  to
population (or the number of households),  then communities  reducing their
waste  loads  through incremental user charges would  benefit.  Not only would
their waste management costs drop, but their product charge  revenues would
remain constant,  reducing the necessary local contribution.

        Let us consider a simple example, with initial  collection costs  of
$30/ton, and 1 ton generated per household annually.   The  elasticity is 0.5,
and the product charge receipts amount to half the waste  disposal costs  at
$30 per ton.   The community decides to use the product  charge  to defray waste
collection and disposal costs, and there are no administrative costs.

        Therefore,  in the  initial case,  the product charge provides
$15/household annually, and another $15 has to be found to  cover the costs of
collection.   Now- introduce an incremental user charge set at marginal cost,
which is $15/ton  or $.25/bag.  Assume that use of metered bags breaks even in
this case, and that wasteload drops 40%.

        We now have the household generating 0.6 tons,  and  paying  $9 annually
for  collection  and  disposal.  But  the product  charge adds  $15/household,
resulting in  a total  revenue of $24/household, just equal  to the  collection
cost which is reduced  by  20%.  The product charge  revenues  cover the fixed
costs, as long as they are disbursed on the basis of  population.

        The product  charge itself  will have some  effect on collection and
disposal costs through changing the  character of  the  wastes, and  reducing
total  quantity.   This may be significant in some cases, but appears to us to
be small in comparison  to  the effects  expected  through  incremental  user
charges.

        One major issue for solid waste management systems, however,  is  the
ability to find  sufficient  funds to operate effectively.  Many  suggestions
have been made that these systems are under-funded; this claim, though,  has
not  been sufficiently documented.  Both incremental  user charges  and product
charges have  been proposed as potential remedies for  these  problems.

        However, product charge  revenues  will-not necessarily  be spent on
solid waste management only.  Because of the number of  firms engaged in solid
waste  collection and disposal,  and because of  the  fragmentation  of  the
industry, it would  be  difficult to distribute  the  revenues to  all  such
entities.  Assuming that  local  governments attempt to  maximize  welfare

                                    127

-------
through their budgeting  procedures, there is no need  to do so.  Tax relief of
any form will have similar effects to tax relief for  solid waste.  Additional
services will, it is  hoped, have a net  social benefit.  Thus,  there is  no
strong need to  limit  the use of the  product charge revenues to waste
management, given the administrative problems Which may be expected.

        User  charges are likely  to  represent an attractive revenue source.
It is very possible that  local governments will view  charge levies as an  easy
method  for increasing  their budgets for "under-funded" services, leading to
gold-plated operations,  particularly since they are  often monopolies.   The
imposition of  incremental user charges may accentuate  the problem  of
inefficiency by making  the system appear equitable to observers and  by
reducing  waste  volumes.  Such charges,  then, may have more chance of adverse
taxation effects than flat-rate charges, though this  is speculative.

Other Modes of Disposal

        The product  charges are  unlikely  to  have  an  effect on the use of
littering, garbage grinders, or other modes of disposal; therefore, there  is
no interaction of  concern.

RESULTS AND CONCLUSIONS

        Figure  2  shows the relative  effects of product  charges  and
incremental user charges  based on the above discussion.  In general, most  of
the interactions  are  beneficial, and the two systems would appear to be  able
to operate together  effectively.  The  major interactions, in  system
financing, consumption, and recycling,  all appear  to be positive, based on
certain design assumptions.

        However,  these  design assumptions are fairly critical and should be
restated.   We  have assumed an elasticity of cost with waste generation of 0.5
throughout, which is effectively an assumption about system efficiency.  We
have assumed that  the  product charge revenues are disbursed on the basis  of
population, and that local governments use them to  maximize social welfare,
either for tax relief  or  for the provision of services, which  may include
solid waste management.  We have assumed  that waste  disposal charges have no
effect on  consumption  behavior, or at least that such effects are minimal,
and we have assumed that  recycling centers will not charge for the waste  they
receive.

        In addition,  we have assumed  that the incremental user charge  will
use the marginal cost  of  collection as  the marginal price, with  a flat  rate
for service or  another  revenue source  used for the  first block on the rate.
This would apparently  result in an overly large amount of recycling.  Given
assumptions  such as these,  however, it should  be possible  to design a
combination of the incremental user charge and the product charge which  does
not  double-count any  costs,  and does provide  a more complete set  of
incentives than  either form individually.
                                   128

-------
                     Effect on
     Product charge
  Incremental user charge
      Combination
               Consumption.


               Income
                                     Shift towards less waste
Small effect, probably
  net increase
                              Minimal
Small effect, probably
  net increase
                              Like product charge
Small effect,  probably
  net increase
                Recycling
Increased demand for
  base load of re-
  cycled product waste
Incentive for households
  to supply recycled
  products
Increased recycling
  with a probable
  imbalance towards
  exce a s supply
NO
                Production firms
Shift towards reduced waste
  in packaging and in-
  creased use of recycled
  materials
                                                                    No effect
                              Like Product Charge
                Waste management
                  firms
Slightly reduced costs
Reduced costs
Increased availability
  of revenues
Reduced costs
Increased availability of
  revenues
                Other modes
                  of disposal
None
Incentive for increased
  use
Incentive for  Increased
  use
                                  Figure  2.  Relative effects  of user and product  charges.

-------
                FUNDAMENTAL COMPARISONS OF ALTERNATIVE SOLID
                          WASTE  MANAGEMENT POLICIES
                                    by


                              ALLEN K. MIEDEMA
 ACKNOWLEDGEMENTS

       The motivation for  this  paper was the  absence of a fairly rigorous
 case for the conceptual  superiority of the product charge.   The need arose
 because our empirical analyses of the product charge and recycling subsidies
 could not be easily and  intuitively  related  to a paradigm  structure that
 could disentangle  otherwise puzzling  quantitative results.  The crude
 outlines of the analysis  in  this paper were suggested by Fred L. Smith, Jr.
 in  a  briefing session on a related project during May 1977.  Other thoughts
 in  this paper, especially the consideration of practical complications, are
 indebted  to my  discussions with  both Tayler H. Bingham and Curtis A.
 Youngblood at RTI.

 INTRODUCTION

        The "solid  waste  problem" has  received considerable attention for
 many years and particularly during the last decade.  It has been argued  that,
 like air and water  pollution, solid wastes "need" to be "controlled."  Yet
 the reasoned case for  the  control  of effluents to common property air and
 water resources hardly extends in any direct manner to solid waste flows.

        The consumption and production residuals  that are disposed to the air
 and water inflict costs on society in the  form of health deterioration or, at
 the very  least,  an unsightly, uncomfortable,  or unenjoyable physical
 environment.   Given the institutions that  have emerged in our  society, the
 residuals, aside  from litter,  that  are  disposed to the land as solid waste
 inflict  only a financial cost,  a cost  that is borne ultimately  by  all
 individuals  who directly or indirectly pay those costs.  In the U.S.  a large
number of state and  local ordinances and  the efforts of the  U.S. Public
Health  Service have virtually  eliminated the  egregious damages  to human
health and  physical environments  caused by "unabated" solid wastes.   (This is
certainly  not true  in less  developed countries where fetid open dumps and
sewers  constantly threaten the lives and well-being of their citizens.)   The
"abatement"  of solid waste  has, by and  large, taken the  form either of

                                   130

-------
government provision of the resources necessary to collect and dispose of all
wastes  in a sanitary manner or of the  imposition of ordinances  that require
the self-provision  of those services by  private  agents.  Where  then  is the
externality?

        Superficially it might appear that  the  current methods  of  handling
solid waste in the U.S. could be cited  as  an exemplary method for curing the
air and water pollution externalities:  simply  have the government pay for
all  the abatement costs or have the government pass ordinances  that require
those expenditures by private agents.   Indeed  that has apparently  been
considered a sterling example.  With  rare exception the chosen "cures" for
air and water pollution have been both regulations and government subsidies.
The latter include  a full complement of  government assisted abatement schemes
like pollution equipment loan guarantees, extension of preferential  municipal
bond rates to abating industries, rapid  abatement cost write-offs, federal
development of abatement  technology,  government  supported technical
assistance for abatement methods, etc.  The main point is that, regardless of
the institutional mechanisms used, the achievement of the "same level" of air
and water pollution abatement as is already realized for solid waste  might
well be viewed by many individuals as considerable, if not fully  adequate,
progress in the battle against all three externalities.

        Thus if the analyst assumes  that current  solid waste "control" is
provided  efficiently,  it  is  not surprising  that he would defend the
materials-, packaging-, and  disposable goods-intensive society.  He can
argue,  e.g., Johnson (I960),  Becker  (1965),  and Arthur D. Little,  Inc.
(1966),  that materialism in the form of  relatively high consumption  of waste
intensive goods  in both  industrial  and household production  functions
enhances  social welfare  because it conserves  other apparently scarcer
resources like labor time and energy.

        What solid  waste imposed externality  do  we cite  in  rebutting  this
argument?  The critical  first step in the debate is to recognize  that "the
externality" in solid waste flows is not unlike "the externality" that would
still prevail if all air  pollution were  "controlled"  through the  use of
regulations.  Again aside from littering,  an externality from solid waste no
longer  exists in the sense that individual  A is directly physically  harmed or
interfered with by  individual B's production or consumption activity.  Rather
most of the externality  in  current solid waste management practices is
attributable to the maze  of market distortions imposed by inefficient,
regulation-like pricing, institutional,  and legal  structures that combine to
produce solid waste "control."

        A number of those  inefficient  structures have been enumerated in
other sources; e.g., Anderson (1977), Page  (1976), and Fiekowsky  (1976).
Each is mentioned here again  for completeness:

Virgin Materials-Biased Tax-Policies

        Current  tax  laws provide for  virgin  material depletion allowances,
capital gains tax advantages  on  standing timber,  and  favorable treatment of
expenses for  exploration and development.   All these  policies allegedly

                                    131

-------
 reduce  the  relative price of virgin  materials and hence cause  both the
 underutilization of recyclables and  the  underpricing of waste  intensive
 goods.

 Virgin Materials-Biased Regulations

         A number of administrative policies  also distort relative  prices of
 virgin and recycled materials.  First, many regulated freight rates are
 systematically  lower  for virgin than for  recyclable materials.  Second, an
 archaic  federal law grants  free mineral rights to those who  make discoveries
 on open federal  land.  Third,  under the pretext of quality-assurance,
 governments enforce virgin materials-biased labelling and procurement
 requirements.

 Free Rider Problem

         The  predominant method of  payment for post-consumer solid  waste
 collection and disposal costs  is flat  assessments; i.e.,  the
 "all-you-can-eat-for-$X-buffet"  pricing scheme.  The most common method of
 administering flat assessments is  through the  application of property taxes,
 local  income taxes, or,  given their  fungibility, federal revenue sharing
 funds  in payment for  municipally provided  services.  The  second  most
 prevalent method is the  direct administration of flat assessments by  waste
 collection and disposal contractors.   It is argued (Wertz, 1976, p. 269) that
 this financing method fails to make individuals aware of the waste collection
 and  disposal component of the marginal social  costs of  the goods  they
 purchase  --  like   eating another  slice  of  apple  pie  at  the
 "all-you-can-eat-buffet," discarding another  unit of waste costs the
 individual nothing  on the margin.  Thus, only  the fairly weak,  negative
 income effect from increased flat assessmnt rates, will reduce  the purchase
 of waste intensive  goods.  Society forfeits the potentially more powerful
 stimulus for waste reduction in the form of  the  substitution effect  that
 might  be achieved by  full marginal cost goods  prices that  include  waste
 handling costs.   This is the classic  free rider problem.

 Indirect Subsidization  of Virgin Materials

        Another  less commonly  mentioned distortion (Summers, 1973,  Chap.  3;
 Smith,  1974,  p. 3; Train, 1976,  p.l)  is  attributed to the market's  failure  to
 incorporate the eventual  collection and disposal  costs  of some virgin
 materials in their prices.   By  comparison, the use  of recyclables avoids
 disposal costs while  collection  costs are included in their market prices.
 Therefore, even apart from the distortions in the relative  price  of virgin
 and recyclable materials  mentioned in 1 and 2 above,  this indirect subsidy
 alone is  allegedly accountable for  underpriced virgin materials; supposedly
 this simultaneously induces both excessive use of underpriced materials, and
hence excessive generation  of wastes, and uneconomically inhibits the use of
 recyclables.

        Besides  these   four core arguments for  the existence of  "an
 externality" in  materials  markets two other arguments  have been propounded.
The first is  that solid  waste flows do entail a genuine  externality in the

                                   132

-------
form of littering through which individual A's production and consumption
activity  adversely affects  individual B.^  The second is  that, compared to
recyling,  virgin material extraction and processing  causes a significant
amount of  other externalities such as  air and water pollution and spoilage of
scenic natural environments.  However, as Anderson (1977, p. 355) points out,
the potency of  this argument is diminished by the partial removal of these
externalities through the implementation of the National Environmental Policy
Act, the Federal Water Pollution Control Act, and the Clean  Air Act.

        A  number of federal policies have  been advocated  or are under
consideration to correct these market distortions; e.g., see especially Sec.
8002 (j) of  U.S. Congress (1976).  Three major generic alternatives have been
suggested:   (1)  product disposal charges, (2) recycling subsidies, and (3)
user fees.  Product charges would provide for a tax on virgin materials  that
are eventually  to be disposed as post consumer solid waste; the tax would
equal the  marginal social cost of solid waste handling.  Recycling  subsidies
would  provide  for direct  payments to users of recycled materials on a unit
basis.  User  fees would be  direct, volume sensitive charges for waste
handling  services in proportion to the extent that households and  businesses
use them.

        To date little  analytical rigor has been applied  to define clearly
the comparative  strengths  and weaknesses  of  each  of  these policies.
     Since it may  cost as much as  19  times more  to  collect  a  littered
     container than to collect that container from a curbside (Goddard, 1975,
     p. 39), it  is sometimes argued  that a corrective "litter tax"  is
     warranted.   Although this argument initially  seems appropriate and
     although a  significant amount of littering  still  does  occur,  the
     magnitude of the "litter tax"  that  appears to follow logically from this
     argument is  infinitesimal.  The logic would be to  set  the litter tax
     equal  to the cost of collecting a unit of litter times the proportion of
     all  units that are actually 1 ittered.  Under such a scheme the proceeds
     from the litter tax would just cover litter collection costs.  For
     example, an  official of the National  Soft Drink Association, obviously a
     vested  interest group in view of recent legislative activities on behalf
     of mandatory deposits, recently proposed  a  litter  tax  of  .1 cent per
     beverage container (see "NSDA's Reed Proposes Litter Law; Bottlers Would
     be Taxed for Program Litter Control Funding," Beverage Industry.  1977).
     The flaw in this logic derives from waste disposal patterns that result
     from flat assessment methods for funding  solid  waste handling services
     (mentioned  above in the discussion of  the free rider problem).   Since
     the marginal cost  of  disposal  under  this  payment method  is zero, it
     undoubtedly accounts for  the sanitary processing of what  would be a
     significant  volume of litter if point-of-dtsposal charges were levied.
     Therefore  the collective  decisionmaking  that  leads to  "free" waste
     handling  services  also undoubtedly accounts  for litter volumes far below
     those that would be observed if those services  were prices at  their  full
     marginal  social  costs.

                                   133

-------
 Generally, comparisons of alternative solid waste management  policies; e.g.,
 Conn (1975),  consist of  fairly  casual  inferences about the  general
 directional  effect each  policy  would  have on the waste stream.  Evidently
 previous work fails to distill  the essential elements  of  "the externality"
 from solid waste  flows into  a simple model that both precisely defines the
 market distortions that  characterize  materials  markets and provides  a
 rigorous theoretical basis for  inferences about the conceptual superiority of
 one  solid waste management policy  over another.

        The purpose of this paper is to examine the  real income and market
 effects of each of the three major  alternative  solid  waste management
 policies and of the status quo  and  to determine whether these  policies can be
 ranked in terms of  their ability to increase social welfare.

        The analysis begins with the development of a  paradigm model of a
 simple economy in the next section.  The model assumes  the existence  of  only
 the  free rider problem and the  indirect subsidization of virgin materials;
 i.e., not only all  three alleged  sources of downward biases in relative
 virgin materials  prices, but also  the littering problem are assumed to be
 nonexistent.  Also, zero transactions costs are assumed.  In a later  section
 this paradigm is  used to analyze each of the  four  policy settings.  The
 policy simulations  presented there  are used to draw tentative conclusions
 about  the relative  real income, net  waste, waste  generation, resource
 recovery and recycling rate effects of the alternative  policies.

 THE MODEL

        Since  the  objective of  this analysis is to make statements about  the
 welfare characteristics  of  alternative solid waste management  policies, it is
 necessary to define  a very rudimentary general  equilibrium model of  a
 paradigm economy.   The model development  parallels that of neoclassical price
 theory, with appropriate departures to account for the  materials flows  that
 cause the solid waste  problem.

        The macroeconomic  setting is assumed to be identical to that  of  the
 standard (timeless)  general equilibrium model,  except that a constitution and
 government are assumed.  The government represents  all individuals  through
 its constitutional  authority to protect resource  ownership and  to expropriate
 (tax) and expend  those resources.  As usual the economy is  assumed  to  be
 autarkic  and  both  exchange  and  (in  this model)  government functions are
 undertaken costlessly.

        All production  activities occur  within profit  maximizing competitive
 firms that are fully owned by  individuals.   The firms themselves  are  not
 endowed with  resources and  do not consume.   Rather all  resources are owned
and provided by individuals who are paid  the value of the  output (including
 rents) attributable  to the  resources they provide.  Thus the macro accounting
identity is met:   the  total  value of output (GNP) is identical to total
national income.

        A major departure of this model from the  standard  neoclassical model
is that all resources other than a single natural resource are subsumed into

                                   134

-------
 a  single factor of production.  This  resource endowment  is a composite  of
 human and physical  capital  from which a fixed total,  k  units, of services
 are available.  The resource endowment  is owned in aliquot shares by  the
 individuals and  is  completely (and costlessly)  mobile among all production
 activities in the economy.  The society is also endowed with a "mine" from
 which the natural  resource, "clay,"  is extracted.  Once extracted,  this
 material flows through the  economy and can either  be  recycled or disposed  to
 a  sanitary landfill.

        On the demand side  all individuals have identical  but independent
 preference functions.   Therefore,  envy and benevolence  are absent  and a
 single indifference curve can be used to  represent  society's real income.
 Also  all resource  expropriation (taxation) by the  government, whether by
 excise taxation or direct property (income) taxation, is borne in aliquot
 shares  by the individuals in the  economy.  Similarly,  the benefits  of
 government expenditures  are enjoyed equally because  individuals' preference
 functions and resource endowments are identical.

        With this general setting it is now possible  to define more fully the
 specific characteristics of  production and consumption activities within the
 paradigm economy.   Only  two final goods  are produced:  a "beverage" and a
 residual (numeraire) good.  The beverage is packaged in a "clay" container
 which, when disposed, constitutes the only solid waste  generated within  the
 economy.  A single  unit  of "contents," f, is assumed to  be required along
 with  a single container  in producing  a unit of beverage.   This  fixed
 coefficients beverage production activity simply represents  an operation that
 (costlessly) blends the  beverage contents and containers; i.e.,
                        q = min(j.f)                                   (1)


where       q   »   beverage output,
            j   »   containers, and
            f   -   contents.

Because of the specialized form of(l), q - j - f.

        The  two intermediate goods used in (1),  containers and contents,  are
produced by applying  the capital output.  The production process for contents
is assumed to be a  simple constant-returns process,
                       f - Pk                                         (2)


where        P  • parameter of the beverage contents production process.

        A "used" container is —  like modern  non-refillable beverage
containers — unsuitable  for reuse; it must  either be disposed or recycled


                                    135

-------
 into a new container production process.   The  latter is described by a simple
 Cobb-Douglas production function,


                        r = «0kal                                     (3)

 where         r = number of containers  produced from recycled containers,
                  and

          «0, «. « parameters of the recycling  and container reconstitution
                  process.

 The activities  represented  by the production  process of  (3)  include the
 collection, transfer, and reconstitution of "used" containers  into  new
 containers.
         Alternatively a new  container can be  produced from virgin  "clay"
 extracted  from  the  "mine."  The  entire process  of mining, materials
 processing,  and container  manufacturing from  virgin material  is also
 described  as  a simple Cobb-Douglas  production process.


                        v =  |9lA                                     (4)
 where         v = number of containers produced from virgin material, and

          /?   /?  • parameters of the mining and container manufacturing
                  process.

        The total  number of  new containers (and of  beverage units produced)
 is  identically the  sum of  those produced from recycled containers and  those
produced from virgin  materials; i.e.,


                       q =  j = v + r.                                 (5)
The quality and appearance of containers produced  from recycled material  is
assumed identical to  those produced from virgin materials.


       Both the basic production processes of  (3)  and (4) and the (costless)
blending  operations  of  (1)  and (5) are assumed to be completely free  of
materials  losses.  Also it is assumed that none  of the used  containers are
lost.   Thus the total flow, h, of potential  solid waste is exactly equal  to
the number, j,  of containers used or, equivalently, the number  of units,  q,
of beverages consumed.

        The numeraire good, which  represents  all other  goods  besides
beverages, is also  assumed  to be produced  by a simple  constant returns
production operation,

                                    136

-------
                        X - Trk                                        (6)


where         x - numeraire good,  and

              if = parameter of the numeraire good production  process.

        The  utility  function shared by  all individuals is  a simple
exponential function


                            A. 1-A.
                        u = q  x                                       (7)


where         u - utils  enjoyed by each individual, and

              A - parameter of the utility function.

It is also assumed that  if even a  single container is littered; i.e., neither
recycled nor disposed, all  individuals'  utility becomes negative.2
Therefore, all  individuals will unanimously vote to have the government
collect and dispose all  used containers if the government is  as efficient  as
individuals in  completing that  activity.  Finally individuals' utility is
assumed to be completely unaffected by whether a used container is  disposed
as solid waste  or  recycled;  i.e.,  utility is independent  of the  recycling
rate.

        If prices in the paradigm  economy are denominated in  units of capital
services, and if disposable incomes are regarded by individuals as completely
exogenous, then individual utility maximization requires that for the society
as a whole the following expression is maximized


                           A  1-A
                      0 - q x    +  A(E-p*q-wx)                       (8)


where         E - society's aggregate disposable income,

             p* - price  of beverages including excise taxes,  if any, and

             A  - Lagrangian operator.

Since E is actually the  sum of  individuals' disposable incomes and since  the
latter are  assumed to be perceived  as exogenously given, the partial
differentials of E with  respect to q and x are zero irrespective of  the solid
2.   If L is  the number of units littered, a more  precise  definition of  the
     utility  function of (7) might be  u - (l-^L)qAcl"A  where l> 1.
                                   137

-------
 waste management policy that  is  applied.  This is assumed  even though E may
 be substantially affected at the macro level by choices of q and x,  since
 individuals  always require their  government to  tax them for the cost.of
 collecting and disposing the used  containers that  are not recycled.

        Under these assumptions  the  solutions to  the first order conditions
 yield the aggregate demand functions for q and  x in terms of aggregate
 disposable income, the price  of  beverages, and the parameter of the utility
 function.  Specificaly, the aggregate  demand functions are


                        q - AE/p*  and                                 (9)

                        x = (l-A)E/7r


        The production process for the collection and  disposal activity is
 also  assumed to be a constant returns  process and is assumed to be identical
 whether the service is  provided  by the government or by individuals,


                        d - /xk                                         (10)

 where         d = number of used containers collected and  disposed to a
                  sanitary landfill, and

              M - parameter of the collection and disposal process.

        Figure  1  summarizes resource and product  flows in the paradigm
 economy.   Only beverages and  the numeraire good  are consumed directly as
 final products.  All other products in the economy  are  intermediate goods
 used in or associated with the waste intensive  good.   Three of the five basic
 industries in the  economy, i.e., those industries  that directly use  the
 capital  input, are  intermediate goods production processes for f,  r,  and v;
 the other two are  the  numeraire good production  activity,  x,  and waste
 collection and disposal, d.   Economic forces together with solid waste
 management policies determine the extent to which wastes are either recycled
 or disposed  to  a  sanitary landfill,  the  extent to  which virgin
 materials-based and recycled materials-based containers  are  substituted in
 container production,  and the extent to which beverages and the numeraire
 good are substituted in  consumption.

 ANALYSIS

        The paradigm of  the  previous section can now be used  to examine  the
 materials and resource flows  as well as the price  and real income  effects
 associated with four  alternative states of  the  world  implied  by four policy
 settings.   The four settings to be analyzed are:  (1)  non-intervention  of  the
government in  either  final  product or input markets;  i.e.,  the status quo,
 (2) government  payment  of a subsidy  for recycling,  (3) government  imposition
of a collection and disposal or user  fee, or equivalently  in  this  model, of
 an excise tax on  beverages,  and (4)  government  imposition  of  a  product

                                   138

-------
 virgin
material-
 based
container
 'composite1
   capital
   input, IT
beverage
container
                        beverage
                        contents
                       production
                         numeraire
                           good
 beverage
production
              on-
           sumption/
           ide prod-
            uct  gen-
            eration
  container
  recycling
     and
reconstitution
                          used  container
                            collection
                               and
                             disposal
            Figure 1.  Paradigm economy resource and product flows.

-------
disposal charge on the virgin material  content of eventual side products.
These  four policies  will be referred  to as the  status quo, user  fee,
recycling subsidy and  product charge policy options.

       The  analysis of each policy develops  the equations whose solutions
characterize market  equilibria and real income levels  in  the paradigm economy
under alternative policy choices.  Specifically, a total  of  six equations and
six endogenous variables define market equilibria.  These variables are:  (1)
capital services used  to produce virgin materials-based containers, k ; (2)
capital services used  to produce recycled materials-based containers, k ; (3)
aggregate  consumption of beverages, q;  (A) the  f.o.b. price of beverages, p;
(5) rents from container production, R;  and (6)  aggregate disposable income
level,  E.   From  the  six  endogenous variables others  such as the level of
numeraire goods consumption, x, and the real income level, u, can be computed
using  (9)  and (7), respectively.  As  would be expected, some of the six
equations differ depending upon the particular policy  choice that is assumed.
Therefore,  the next section first defines  the particular general equilibrium
system that  obtains  under each of the four  policy alternatives.  Then, since
all  systems are  fully  nonlinear, we  present some  policy simulations for
assumed values of the  various parameters in each general  equilibrium system.
The  results of these  simulations are then  used  to draw conclusions about the
comparative  desirability  of the four  alternative solid waste management
policies.   Table 1 defines the parameters and unknowns that appear in the
alternative  six equation systems.
       TABLE 1.  GENERAL EQUILIBRIUM SYSTEM VARIABLES

             Classification                  Variables

          Function parameters          a, ft, y

          System parameters            w, k

          Policy parameters             t,  a

          System unknowns              k , k , q, p,  R, E
                                   140

-------
 Policy-Specific General Equilibrium Equations Systems

        Each of the  four  policies considered here implies a different method
 of government  Intervention to manage solid wastes.  Each method  in turn may
 imply a unique profit function for beverage producers.  Therefore  the profit
 equation is  the first specified below in the  general equilibrium systems
 associated with each policy.  Then, using the profit maximization assumption,
 the second equation expresses the optimum quantity of capital  to be used in
 recycled materials-based  container production (determined  from the first
 order conditions) as a function of that to be used in virgin materials-based
 container production.  The third equation, also derived from the first order
 conditions tor profit maximization, expresses  the  price of beverages  as a
 function of  the optimum quantity of  capital  services in  recycled
 materials-based container production.  The fourth equation defines  disposable
 national income;  i.e., national income net of government expenditures for
 waste handling services.  The fifth equation simply specializes the aggregate
 demand function  of  (9) for each policy option.  Finally the identity of (5)
 completely identifies the system for each policy.

        To  simplify the algebra the production functions of  (2), (3), (4),
 and (6) are  further specialized by assuming that  P-«o=/3o!a7r=l  and by
 suppressing  the subscript  on «j and  /Jj.  While  they are notationally
 convenient,  these assumptions do not substantively  affect  the  conclusions
 that  derive  from the  policy simulations of the next  section.

 Status Quo —
        The  status quo policy is intended to reflect the basics  of current
 waste handling practices.   Specifically,  government  (or  a  group  of
 competitive private  firms)  is  assumed to provide  waste collection  and
 disposal services under  a  flat  assessment payment scheme.  The  property  tax,
 for example, is  one method by  which this is  accomplished in  practice.  A
 similar pricing technique  used  by private waste collection and  disposal firms
 is  to charge a flat  rate per period for waste removal  services regardless of
 the volume generated.

        Under the  status quo beverage industry profits, R,  will simply be the
 difference between beverage  sales  and total input costs.  Since  one unit  of
 contents costing one  unit of capital services (P - 1) is required per unit of
 beverage output,  the  cost  of contents will be q.  The cost  of containers will
 just be the sum of all  units of  capital  services used in  container
 production.   Thus  industry  profits  are
                       R
where         R -  beverage industry profits.
                                    141

-------
When  the  specialized forms  of  the container production functions, (3) and
(4), are substituted into (5)  and when the latter  replaces q in  the  above
expression,  beverage industry  profits  can be re-expressed as
                                                                     (ID
        When  (11) is  differentiated with respect  to k  and k   and  the
resulting  expressions are set equal to  zero, these two  first  order conditions
can be  solved to express the profit maximizing quantity of  k   as a  function
of the production  function parameters and k  ,; i.e.,
        The first order  condition in which the  expression  of  (11)  is
differentiated with respect  to kr can also be solved to express the f .o.b.
price of beverages as a function of k  ,


                       p = (l~a/«) + 1.                             (13)
        Disposable national income under  the_status quo is the value  of its
fixed  endowment of  capital services,  k,  plus  profits from  beverage
production,  R,  less income taxes required to pay  for  waste handling;  i.e.,
less the  product of the  unit cost of waste  handling, w, and  the flow of
virgin  materials  used (and discarded)  or


                       E = k + R - wk ^                                (14)
                                     v

        Finally since no excise taxes  are imposed under the status  quo, the
f .o.b.  price is  the retail price, p  = p*, and the aggregate  demand function
of (9)  becomes


                       q = XE/p.                                       (15)


        Table 2 summarizes  the six equations  of  the general equilibrium
system  that is defined by a status quo  solid waste management policy.  This
system  can be solved tor k , k , q, p,  R, and E, although not  in closed  form.
                                   142

-------
    TABLE 2.  GENERAL EQUILIBRIUM EQUATIONS SYSTEMS
Policy choice
Status quo
Recycling subsidy
User fee
Product charge
Equation System
(H),
(16),
(11),
(22),
(12),
(17),
(12),
(23),
(13),
(18),
(13),
(13),
(14),
(19),
(20),
(24),
(15),
(15),
(21),
(15),
(5)
(5)
(5)
(5)

Recycling Subsidies—
        Among  the .proposed legislative remedies  for the solid waste problem  a
recycling subsidy policy is one of the most frequently mentioned.  Anderson
(1977)  reviews in some detail two major federal bills, H.R. 148 and H.R.
10612, that would grant alternative torms of recycling  subsidies.  Those
documents and others referenced by the National Commission on Supplies and
Shortages (1976, pp. 155-72)  provide details  on the nuances of specific
subsidy plans.

        For this analysis the simplest possible  form of a  recycling subsidy
is assumed.  Specifically it is assumed that a  flat subsidy rate, a, is paid
directly to beverage producers for  each recycled  container  they use.
Therefore the beverage industry profit function under  a recycling subsidy
differs from that under the status quo policy in that  profits  are augmented
by subsidy payments; i.e,


                 R = (p-lHkr* + k^f) - kr - kv + a kra.               (16)


        Differentiating (16) and solving the first order conditions yields an
alternative expression for the  optimum usage of  capital services  in recycled
materials-based container production,

                                    I-/?        !/(«-!)
                       k  -  [/3/(a(k    +]8a))]        .               (17)
                        r           v
                                    143

-------
        Also,  one  of  the  first order conditions for a  maximum of (16) can be
solved to develop a price  equation analogous to (13),


                        p  -  (klr~a/«) - a + 1.                         (18)


        Disposable  national  Income under the recycling policy differs  from
that  under the  status  quo in that incomes must be further  taxed to pay the
recycling subsidies.   Therefore, disposable income will be


                        E  -  k + R - wk /8- ak ".                       (19)


        Of course the  absence of an excise tax on beverages implies  that  the
aggregate demand function under a recycling subsidy is the same as  under the
status quo; i.e., equation (15).  The six equation general equilibrium system
under the recycling subsidy  is also summarized in Table 3.

User Fee—
        The user fee  policy in the paradigm economy represents either of two
situations.   First,  the policy can represent  the imposition of  a
volume-sensitive  waste tax at the point of disposal.  For example, a tax
equal to the cost of used  container collection and disposal  could be  levied
on  each  container that is  thrown away.  Alternatively,  given the assumption
that transactions costs  are  zero, the policy is  Identical,  in the  paradigm
economy,  to a litter tax,  since government can costlessly impose a charge
either at the point of disposal (user fee) or point of  purchase (litter tax).

        Under a user fee policy the  profit function and  the associated
solutions for k  and p from  the first order conditions  are identical to those
under  the status quo  — equations (11), (12), and (13) — because the policy
does not directly affect beverage producers.

        However, when a user fee  is  applied  at the rate  of t per unit of
beverages consumed (and  per unit of  container  discarded  by consumers),
disposable national income is increased over that under the status quo by the
amount of the user  fee proceeds.  Therefore disposable  national income  will
be


                        E-k + R+tq- wkv^                          (20)


        Of  course,  the  specification of the  aggregate  beverage  demand
function will also  be  affected since the effective price that consumers  face
will be the sum of  the f.o.b. price of beverages and the user fee; i.e., p* -
p + t so
                                    144

-------
                        q - XE/Cp+t).                                  (21)


 The  general equilibrium equation system under  the user fee' policy will,
 therefore, consist of equations (11),  (12), (13), (20), (21),  and  (5).

 Product  Charge— •
         A product charge policy would, as usually conceived (see Smith, 1974;
 Page,  1976; and U.S. Senate, 1976),  impose a unit tax on all virgin materials
 that are eventually disposed.   In the paradigm economy this implies that
 every  virgin materials-based container will be  taxed at, say, the  rate t.
 Therefore, beverage producer profits will be reduced by the amount  of these
 taxes  so the modified profit function  is
                                           r - Ky - t kv.             (22)


        The corresponding solution of  the  first order conditions for  k  is


                                           (a"l)  .                    (23)
        The solution for  p  is  identical to that under the  status quo and user
fee policies; i.e.,  to  (13).

        Aggregate disposable income will  be increased  over that under  the
status quo by the amount  of the product charge proceeds  so
                       E - k + R - (w-t)ky  .                         (24)


        Finally the  specification  of  the aggregate  demand function  for
beverages will be the same  under the product charge  policy as under  the
status  quo and recycling  subsidies.   The resulting six equation system is
summarized in Table  2.

Solid Waste Management Policy Simulations

        Since closed-form  solutions  of the  four  alternative general
equilibrium systems  are not  attainable,  fixed values  were chosen for  the
parameters identified in  Table 2 and  the resulting nonlinear systems were
then solved numerically using the simple half interval method  (Southworth and
Deleeuw, 1966).   For all policy simulations the following  parameters are held
constant at the  indicated values:  A - .2, w - 10,. k- 10,400.  Then in three
different sets of simulations the production function parameters a and /? were
varied to analyze the  effects of differences In the degree of diminishing
returns (diseconomies of scale) in the virgin materials-based as opposed to
the recycled materials-based container manufacturing industries.  The assumed


                                    145

-------
alternative parameter  values are given  in Table 3  and,  as indicated,
correspond to three possible situations:  (I)  both the virgin and recycled
materials-based  container manufacturing industries have equal diseconomies  of
scale, (II) the  recycled materials-based industry has greater  diseconomies,
and  (III) the virgin  materials-based industry has greater diseconomies.
Diseconomies of  scale obviously imply that increasing amounts of inputs must
be used  to produce additional containers.   In the recycling materials-based
industry this could be  caused by diminishing  quality of recyclable materials
or by increasing  unit collection costs as the recycling rate increases.   In
the virgin materials-based industry it  could  be caused  by  diminishing ore
quality and capital fixities in mining industries.
  TABLE 3.  CONTAINER PRODUCTION FUNCTION PARAMETERS FOR
              THREE ALTERNATIVE SETS OF POLICY SIMULATIONS
Policy
simulation
I
II
III
Production function
parameter values
ft
.5 .5
.4 .6
.6 .4
Containe r manufacturing
industry with greater
diseconomies of scale
Neither
Recycled materials-based
Virgin materials-based

        For each of  the three sets of policy simulations the single solution
for  the  status quo was  obtained.  Since the  status quo is always a special
case of all three  alternative policies — i.e.,  all  four equation systems  of
Table  3  are identical when t = a = 0 — that solution is also a solution for
the zero value  of  the policy parameter associated with each.  Then the policy
parameter  was  incremented  by a value  of  2  up to a value of 20 (twice the
marginal cost of collection and disposal).  The  resulting solution values  of
interest  are  displayed  in  the three  panels (labeled I,  II,  and III  to
correspond  with policy simulations  I,  II,  and III) of Figures 1-4.  The
figures  display, respectively, the real income, net waste, waste generation
and  resource recovery,  and  recycling  rate  effects  of  the  alternative
policies.   Each of those effects is discussed  in turn in the following four
sections which  compare the four alternative  policies.

                                    146

-------
Real Income Effects—
        Figure 2 displays the most important  results because  it  shows the
fate of  individuals' real  income under alternative solid waste  management
policies.  Several critical conclusions are  suggested by these results:

The product charge maximizes social welfare  when the charge rate is set  equal
to the marginal  cost  of waste collection and disposal.  No  other  policy at
any  setting of their policy parameters can generate as high a level  of real
income.

Unless  the product  charge  rate is well in excess of the marginal  cost of
waste handling,  w, society will be better off with a product charge  set at
any level than  under  the status quo.

At any identical charge, subsidy, or fee rate the  product charge is superior
to  both the user fee and  recycling subsidy policies and the user fee  is in
turn superior to the  recycling subsidy.

When the  diseconomies  of scale are substantially  greater  in  virgin
materials-based container  manufacturing than in recycled  materials-based
manufacturing none  of  the  innovative solid waste management policies  are
markedly superior to  the status quo.

Only the product charge is always superior  to the status quo when the policy
rate is set equal to  the marginal cost of waste collection and  disposal.

All three innovative policies are superior to the  status  quo for values of
the  appropriate policy rate that are close to zero.

 Net Waste  Effects—
         Figure 3 shows the  associated  net waste that is generated. All three
 alternative  policies lead to a lower net  flow  of  waste  that needs  to be
 collected  and disposed  by the government.   At identical charge, subsidy, and
 fee rates  the  product  charge  causes  the lowest  net  waste flows.   If
 diseconomies  of scale are greater in the  virgin material-based container
 manufacturing industry user fees are associated  with the highest net  waste
 flows.   Otherwise recycling subsidies are.

 Waste Generation and Resource Recovery Effects—
         The reasons  for the preceding results are displayed in Figure 4 which
 shows both the  gross waste generation and recycling levels.  The upper  set of
 curves  in each panel  represents beverage consumption (waste  generation)
 levels and the lower set represents recycling levels.  The vertical distance
 between  the two curves for  each policy  are  the net waste effects  shown in
 Figure 3.   The  superiority of the product  charge is also  suggested by the
 fact  that it causes both a positive waste  reduction effect (the upper  dotted
 line has a negative  slope) and a positive resource recovery effect (the lower
 dotted  line has a  positive slope).  These reinforcing effects contrast  with
 the recycling  subsidy which  causes a  perverse (negative)  waste reduction
 effect  with a positive resource recovery  effect and with the user fee which
 causes  a  positive waste reduction  effect with a comparatively perverse
 resource recovery effect.

                                     147

-------
 (I)
(II)
(III)
                             8    12     16
                            Policy Rate
                                                     ©
	 Product Charge
 — User Fee
—> Recycling Subsidy
    Status Quo
                       4      8     12     16
                           Policy Rate
          o>
          at
             3.42
             3.38
                       4      8     12     16
                            Policy Rate
                  Figure 2.   Real income effects.
                                     148

-------
  (I)
 (II)
(III)
                      4     8     12     16
                           Policy Race
                                                    ©
	 Product Charge
	 User Fee
— Recycling Subsidy
   Status Ouo
                            8     12     16
                           Policy  Race
                     4     8     12     16
                          Policy Rate
                   Figure 3.  Net waste effects.
                                149

-------


50
1 (
*H / C
£ 45
nJ
u
V
C ? ,3 jU
CO Pi
o 25
0 /
V
01
00
dl
 Waste Generation .
A A A""\ ' A
*--•*--*! Recycling *' .'
* - •*- -*J ,•' •
,.'' y '
' f'
'•' *
s /
f •*' j'
t' y'
,'' ,'
lf" *'
.' '
N' '_ »'

». -• ' ^ . • 	 *
^7 v/
^*-~ '">»,
""*-^ "^-.
•*., """-i
S* >»
*"^ ^"~~.^
*""•-•
     8   12   16
    Policy Rate
        (I)
48    12   16
   Policy Rate
      (ID
 8    12    16
Policy Rate
   (III)
Figure 4.  Waste generation and resource recovery effects.

-------
Recycling Rate Effects--
        Figure 5 shows  the recycling rates (recycling levels divided by gross
waste generation levels) associated with each policy.   The recycling  rates
are  greater under the product charge at any particular charge/sub sidy rate
than under the recycling policy, except under policy simulation III when  the
comparatively greater  returns to scale in recycled  materials-based container
manufacture lead to higher recycling rates under a recycling subsidy.   Also,
for  all  positive charge and subsidy rates, the recycling  rates under either
of those two policies exceeds that associated with the  status  quo.  That  is
not  the case for the  user  fee under  which the recycling rate may actually
decline (simulation III),

        This  figure combined with Figure 2 makes a  point that should be, but
apparently isn't always, obvious:  policy-induced increases in the recycling
rate are very poor indicators of associated changes  in  social  welfare.  For
example, even though  the  user fee leads to much lower recycling rates than
the recycling subsidy (as shown by Figure A), it is nonetheless uniformly
superior to the recycling  subsidy  in a social welfare  sense (as shown by
Figure 1).  Also, a very high product charge rate would actually make society
worse  off than  under  the  status quo, but it would at the same time further
increase the recycling  rate.
                                    151

-------
(I)
(II)
(III)
          c
          0)
          
-------
                                REFERENCES

1.   Anderson, Robert C.   "Public  Policies toward the  Use of  Scrap
     Materials."   American Economic Review.  56,  No.  1  (Feb.,  1977):
     355-358.

2.   Arthur  D. Little, Inc.  The Role of Packaging in the U.S. Economy.  New
     York:   American Foundation for  Management Research,  Inc., 1966.

3.   Becker, G.  S.   "A Theory of the  Allocation  of Time."   The Economic
     Journal. 75, No. 299 (Sept., 1965):  493-517.

4.   Conn, W. D.  "Analysis of Source Reduction Policies."  Unpublished draft
     for the California Solid Waste  Management Board.  (December, 1975).

5.   Fiekowsky,  S.   "Tax Subsidies for Recycling Solid  Waste Materials:  An
     Overview."   Treasury Department, Unpublished Staff Memorandum BSM-76-15.
     (1975).

6.   Johnson, H.  G.  "The Political  Economy of Opulence."  Canadian Journal
     of Economics and Political Science.  26, No. 4 (Nov.,  1960):  552-64.

7.   National  Commission on  Supplies  and Shortages.  Government and the
     Nations Resurces.  Washington:  National  Commission on Supplies and
     Shortages, 1976.

8.   "NSDA's Reed Proposes Litter Law:  Bottlers, Brewers,  Would be Taxed for
     Program Litter  Control Funding," Beverage Industry,  Vol.  62, No.
     (February 18, 1977), p. 17 ff.

9,   Page, R.  T.   Conservation and Economic Efficiency;  An Approach  to
     Materials Policy.  Baltimore:  Johns Hopkins Press for Resources for the
     Future,  1977.

10.  Smith,  F.L.   "The  Disposal Charge Concept."  Environmental Protection
     Agency, Unpublished Staff Paper.  (Sept., 1974).

11.  Summers,  Wall en M.   "Externalities of  Paper and  Paperboard."
     Unpublished  dissertation, Harvard University, 1973.

12.  Train,  R. E.  Official letter  to  Senator  Russell  B. Long, Chairman,
     Committee on Finance on a revised Senate Recycling Tax Credit proposal.
     February 27, 1976.
                                   153

-------
13.  U.S.  Congress.  Resource Conservation and Recovery Act of 1976.  Public
     Law 94-580, 94th Cong.,  2nd  Sess., 1976.

14.  U.S.  Senate,  Committee  on  Public Works.  Hearings,  to Consider  the
     Effects of Product Disposal  Charges on Municipal  Waste Recovery  and
     Reuse.  94th Cong., 2nd  Sess., May 20, 1976.

15.  Wertz, K. L.  "Economic  Factors  Influencing Households'  Production of
     Re fuse."  Journal  of Environmental  Economics and Management 2, No.  4
     (April 1976):  263-72.
                                   154

-------
                                  SECTION 5

                  PRICING AND RESOURCE RECOVERY TECHNOLOGY


"Conditionally Predictive Estimates of Secondary Materials Supply" (Tayler H.
Bingham, Curtis E. Youngblood, Philip C. Cooley).

"Secondary Material Demand and Supply Responses" (Robert C. Anderson).
                                     155

-------
      CONDITIONALLY PREDICTIVE ESTIMATES OF SECONDARY MATERIALS SUPPLY
                                    by
                             TAYLER H.  BINGHAM
                           CURTIS E. YOUNGBLOOD
                             PHILIP C.  COOLEY
INTRODUCTION
        The effectiveness of a number of solid  waste policies,  most notably
the disposal charge,  is  influenced by  the potential  supply of secondary
materials.   The  initial efforts in this  area were,  necessarily, descriptive
studies  that presented materials  flows,  described solid waste management
practices  and discussed  the  institutional and economic  impediments to
additional resource  recovery.  Quantitative  studies of secondary materials
supply from the  post-consumer (municipal)  waste stream have been hampered by
a number  of obstacles—most  notably the lack of  an adequate data base,  and
for many materials,  the lack of any significant level of  recycling.

        There are basically two approaches to supply estimation:  econometric
and process modeling.  The econometric method utilizes statistical techniques
and empirical observations  to develop  positive estimates of behavioral
responses to exogenous  events.  These estimates relate to what  has happened
in  the  past and can provide the basis for good predictions if there has been
no change in the economic or institutional relationships.  The process  model
approach  to supply estimation is normative—it describes what  should happen,
given various underlying  factors  and  the  objective function  of the
institution.  The quality  of the  predictions developed from the use of  a
process  model is only  as good as the set of choices that  are modeled and  the
relevance of the assumed objective function to  that of decision makers.

        A small  number  of econometric studies of secondary materials supply
have  been  undertaken.  Most  of the estimates indicate an inelastic supply
function for the secondary  mater)als  studied.  Frequently,  however,  the
materials  studied have not  been post-consumer wastes but rather obsolete
scrap  from  industrial  establishments.

        This paper  presents the  results of  utilizing  a  process model  to
predict  the quantity of secondary materials recovered from  the solid  waste
stream under alternative prices for these  materials.  The quantity-price data


                                    156

-------
 are  then  used to estimate  a  conditionally predictive  set of  own- and
 cross-price  supply elasticities for  secondary materials.

         The  general approach has been employed by Griffin (1977) in another
 setting where  the joint product nature of production was  important—petroleum
 refining.  The  joint product problem, which arises when the production of  one
 product (e.g., beef) results in the concomitant production of another (e.g.,
 hides)  is  not  handled well  by current  econometric methods.  Many of  the
 existing  resource  recovery systems  have important joint product features.
 For example, aluminum recovery  requires the recovery of ferrous materials.

         Material  flows are  identified in the  following section.  This  is
 followed by  a discussion of  the solid waste management options, their costs,
 and the process  model.  Finally,  statistical  estimates  of the supply  of
 secondary  ferrous, glass and aluminum  are presented and  some  conclusions
 drawn.

 MATERIALS  FLOWS

         The  process model presented  later in  this paper uses a materials
 balance  framework to account  for  all materials flows.  In  this  section the
 materials  and their  major  paths are identified.

        In 1976, individuals, commercial  establishments,  and institutions
 generated  almost 145 million  tons of solid waste  (about 3.7 pounds per  person
 per day)   of which nine  million tons, or  six percent, were  recycled.  Thus,
 approximately  136  million  tons were disposed of by dumping,  littering
 landfilling,  and incinerating.   Table 1  shows  the breakdown of municipal
 solid wastes used  in  this  study.  The  category "all  other waste"  is  the
 residual which is  included  for completeness.  It  does  not  include
 agricultural  and mining  wastes, construction and  demolition debris,  abandoned
 vehicles, or industrial  wastes.

        Current levels  of recycling are  presented  in table 1  for each  of the
 five components of the municipal waste  stream.  Paper and aluminum have the
 highest recycling  rates.   The relative  availability of aluminum recycling
 centers and the price paid for  aluminum cans  ($340 per ton  in  1976) make
 aluminum  an attractive material to recover.  Newspapers,  corrugated
 containers, and office paper  constituted  over  three-fourths of  all  paper
 recycled  in 1976.   When  separated from  other wastes at the  point  of
 generation (commercial and  business establishments),  old  corrugated
 containers and sorted  white ledger (office paper) have  the  characteristics
 cited by Page (1976) that make a material more amenable  to recycling:   mass,
 concentration of  mass, known contaminants, and homogeneity,  or  consistency.
Newspapers, while generated in sufficient quantities,  are dispersed over a
much  wider area than are corrugated container and office  paper.  A paper
drive accumulates  these dispersed quantities  into a  mass sufficient for
recycling.

        Glass and ferrous metals from postconsumer wastes  are not  recovered
in large quantities.  The raw materials used to make glass are  so cheap and
abundant that there is little incentive to recover glass  from postconsumer

                                    157

-------
  TABLE 1.  COMPOSITION AND DISPOSITION OF POSTCONSUMER SOLID
             WASTES,  1976
Disposition
Composition

Paper
Glass
Ferrous
Aluminum
All other waste
Total (Average)
Amount
(106 tons)
50.4
13.9
11.2
1.2
68.0
144.7
Proportion
.348
.096
.077
.008
.470
1.000
Recycled
Share
(%)
16.1
2.7
0.9
8.8
0.0
6.0
Amount
(106 tons)
3.1
.4
.1
.1
0.0
8.7
Disposed
(106 tons)
42.3
13.5
11.1
1.1
68.0
136.0
Source:   Adapted from estimates by Franklin Associates, Ltd. for the Resource
         Recovery Division, Office of Solid Waste, U.S. Environmental
         Protection Agency, January 1978.
                                    158

-------
wastes.  Gullet of known quality needed  in the glassmaking process  is  easily
generated by the producer.   Gullet  from municipal wastes, on the other hand,
must be cleaned of contaminants like metal and  ceramics and color  sorted!
Its weight-to-value ratio  is  so large that transporting it more  than a few
miles  is uneconomical.  Ferrous metals, too,  suffer from contaminants,
chiefly tin,  that  must be  removed prior to reuse in steelmaking.  Ferrous
scrap  is also much less valuable than  aluminum.

        As  shown  in  figure  1,  the  generated  solid waste may  either go
directly to the municipal or  private collector,  or  some of  the more  useful
materials may  be  source separated and recycled.   Source separation can be
either voluntary  or mandatory.  The majority  of  the recycling activity
presented in table 1 represents  the  voluntary action of the waste generators.

        Voluntary source separation  alters not only  the composition  of  the
waste  stream but also its energy content.  This affects the economics of the
various waste management options facing municipalities,  as will  be  seen
below.  Counterbalancing this effect, however, voluntary source  separation
decreases the volume of municipal wastes that must be collected and disposed.
One difference between mandatory and  voluntary source separation is that the
revenues from the sale of the  secondary materials  accrue  directly  to  the
municipality under mandatory  separation, thus offsetting some or  all of the
incremental costs of separate collection.

        The actual  waste collected by the  private firm or municipality is
usually unseparated.  However,  in some areas,  separation of certain fractions
is  required as a condition of service—mandatory source separation.  The city
(or the city's contracted hauler) collects these source-separated wastes  for
delivery to a  processor.  In voluntary activities,  this service is provided
by  the private sector.

        A survey of  resource  recovery  and waste reduction programs  throughout
the U.S.  (McEwen,  1977) showed that as of  August 1974,  174 cities  had
implemented separate collection programs.  Of these, 156 collected newsprint
only; the remaining  18 collected various combinations of mixed paper,  glass,
and cans.   Whether all  of these programs are mandatory in the sense that we
use the word is doubtful.

        Unlike  some forms of  voluntary source  separation, there  are not
direct  financial incentives for generators to  comply;  since enforcement of
the municipal  ordinance   requiring  source  separation is practically
impossible,  compliance has a  strong  voluntary  element.   If the benefits  are
to outweigh  the  costs,  it requires the cooperation of waste  generators.

        Mandatory source  separation  reduces  the volume of waste  requiring
disposal,  just as  voluntary recycling does.   This  is  increasingly important
as landfill  space becomes more scarce.  Thus,  avoided disposal costs are  one
of  the  benefits of  mandatory source  separation.  It also  alters the volume,
composition,  and energy content of the solid  waste stream, which  naturally
affects the  economics of  processing  and disposing of  the remaining  wastes.
                                    159

-------
Haste
Generation
                  Voluntary
                    Source
                 ' Separation
Waste
Collection
                   Mandatory
                     Source
                  "Separation
                      T
Pre-processing
 Materials/
   Energy
^Recovery
                                                                                                                     Sanitary
                                                                                                                     Landfill
  Paper
  Glass
 Ferrous
Muminum
 Energy
                                           Figure  1.   Solid waste material flows.

-------
        The unseparated wastes  next  follow one  of  three paths leading  to
 final disposition:   1)  disposal,  2)  preprocessing and disposal, or 3)
 preprocessing, resource recovery,  and disposal.

        Neglecting open dumping,  the sanitary landfill  is  the most widespread
 method of waste disposal currently used in the U.S.   Although this  method
 does require  expenditures  on heavy equipment, such as compactors, scrapers,
 and graders,  the  major cost for most communities  is the cost  of  land.
 Because  of recent  restrictions on  and requirements for landfills,  the
 availability of suitable sites has shrunk.  The reluctance of property owners
 to sell  acreage  for use as a  landfill and  their  resistance to having a
 landfill adjacent to their properties further constricts supply.  Additional
 processing to achieve more size reduction and more  resource recovery will
 extend the life of sanitary  landfills.

        Processing and disposal, the second  major alternative,  involves
 shredding the waste stream prior  to disposal.  The size reduction achieved by
 shredding is substantial.  For instance, a sanitary landfill with an expected
 life  of 20 years designed to handle  1,000 tons  per day  requires about  31
 percent  more land  than does a  landfill of  the  same capacity at which the
 wastes are shredded prior to disposal.  Additional  size reduction  can  be
 achieved  by incineration,  after the  waste has  passed  through a one-stage
 shredding operation.   The residue remaining is  landfilled;  it represents
 about 10  percent  by volume and  about 25 to  35  percent by weight  of  the
 incoming waste stream (Levy  and Rigo, 1976, p.  26).

        The third  major  alternative available to municipalities is to process
 the  waste stream, recover  certain  of the resources,  and dispose  of  the
 remainder.  The basic  processing  step is to shred  the wastes.  Resource
 recovery connotes  both retrieval of the glass,  ferrous,  and  aluminum fraction
 for  recycling and  the  recovery  of  the energy  content  of the combustible
 fraction.   The two  methods used to recover energy from  the waste stream  are:
 1)   air  classification which separates the  "light"  (combustible) fraction
 from  the "heavy"  (inorganic) fraction to produce a refuse-derived fuel  (fluff
 RDF), and 2)   incineration to generate steam.  The  steam is a more valuable
 product than fluff  RDF,  but  requires more processing  to  make it so.

        Materials recovery takes place at several  points  following  the
 shredding  operation.   In plants that produce a  fluff  RDF,  the ferrous portion
 is magnetically separated  prior to and just  after  air classification while
 the glass  and  aluminum fraction is separated and recovered  after the waste
 stream has been air  classified.  For the incineration processes, we  assume
 that the  ferrous, glass, and aluminum fractions are separated  and recovered
 prior to  incineration.  Disposal of the wastes remaining following resource
recovery  operations is usually by  sanitary landiflling.

        Table 2 shows the  resource recovery facilities in operation in 1976
across the nation.  Quite obviously, most solid wastes  do  not enter resource
recovery facilities but  are directly disposed of.
                                    161

-------
   TABLE 2*  MIXED-WASTE RESOURCE RECOVERY FACILITIES IN OPERATION (SUMMER 1976)
Location
AHoona, Pennsylvania
Ames, Iowa
Blythevflle, Arkansas
Braintree, Massachusetts
Chicago, Illinois (Southwest)
E. Brldgewater, Massachusetts
Franklin, Ohio
Grove ton, New Hampshire
Harrlsburg, Pennsylvania
Merrlck, Mew York
Miami, Florida
Nashville, Tennessee
Norfolk, Virginia
Oceans Ide, New York
Palos Verdes, California
St. Louis, Missouri
Saugus, Massachusetts
SI loam Springs, Arkansas
S. Charleston, West Virginia
Type
Compost
RDF
Incineration
Incineration
Incineration
RDF
Wet pulp
Incineration
Incineration
Incineration
Incineration
Incineration
Incineration
Incineration
Methane recovery
RDF
Incineration
Incineration
Pyrolysis
Capacity
(tons per day)
200
400
50
240
1,200
160
150
30
720
600
900
720
360
750

300
1,200
20
200
Products /markets
Humus
RDF, Fe, A1
Steam/process
Steam/process
Steam/Industrial park
RDF/utility
Fiber, Fe, glass, Al
Steam/process
Steam/ sludge drying
Electricity
Steam
Steam heating & cooling
Steam/navy base
Steam
Gas utility & Fe
RDF coal-fired utility
Steam/process
Steam/canning Industry
Gas, Fe
Startup.
date
1963
9/75
11/75
1971
1963
1974
1971
1975
1972
1952
1956
7/74
1967
1965/74
6/75
1972
4/76
9/75
1974
Source:   L.  B.  McEwen, Jr., Waste Reduction and Resource Recovery Activities; A Nationwide Survey.
         Environmental Protection Publication SW-142, Washington, 1977, p.  2.

-------
SOLID WASTE  MANAGEMENT OPTIONS

        The solid waste  management  options represented in  the process model
include  the major alternatives  available  to  municipalities.   These
alternatives can,  as shown in figure  1, be divided into six  broad solid waste
management strategies:

        1)  sanitary landfill,

        2)  mandatory source separation-sanitary landfill,

        3)  mandatory source separation-preprocessing-sanitary landfill,

        4)  mandatory source separation-preprocessing-resource recovery-
            sanitary landfill,

        5)  preprocessing-sanitary landfill, and

        6)  preprocessing-resource recovery-sanitary landfill.

        The six strategies represent combinations  of four options  (santiary
landfill* mandatory  source  separation,  preprocessing,  and resource recovery.
Table 3  identifies  the  choices that  the  decision maker  has  within each
option.  These options  represent the majority  of  those  currently available  to
municipalities although,  as  shown in  table  2, the experience with most  is
limited.


                   TABLE  3.  SOLID WASTE MANAGE-
                               MENT OPTIONS
                     Sanitary landfill
                     Mandatory source separation

                          Paper recovery
                          Glass recovery
                          Ferrous recovery
                          Aluminum recovery
                      Preprocessing

                          Shredding
                          Incineration

                      Resource recovery

                          Glass recovery
                          Ferrous recovery
                          Aluminum recovery
                          Refuse-derived fuel
                          Steam
                                     163

-------
Sanitary Landfill

        The  sanitary landfill  is currently the most common  method of waste
management.  The location of  the landfill is critical—an attempt  is  usually
made  to locate  it  so that the externalities  imposed on residents of the
municipality are minimized:  the  heavy truck traffic, the noise  of  heavy
machinery, the dust,  and the noxious odors all contribute to  a  desire on the
part of residents to have the landfill located as far away from  themselves as
possible.  The geology  of  the landfill site must also be selected or altered
so that the leachate (runoff water) does not contaminate water  supplies.

        The  landfill in our model has a lifetime of 20 years.  The site has
twice as many acres as are  actually needed for disposal; this  helps  minimize
the  externalities  mentioned above  and provides  room for access  roads for
garbage trucks and other vehicles.  Each acre used for disposal  is dug out to
a depth of 25 feet and can  hold 20,000 tons of waste, when properly  processed
by compactors, scrapers,  graders,  and truck-type tractors that  constitute the
landfill's capital  equipment.   In  the usual procedure, a load  of waste is
deposited in the pit,  compacted, leveled and then covered with  two  or  three
feet  of earth.   The  layers of dirt help prevent odors, rodent  infestations,
and unsightliness—major complaints leveled against the open dumps  still in
use in  some areas of the country.

        Properly managed, the site can be reclaimed for other uses—usually
recreational — at  the end of its useful life as a landfill.   Once  a site is
closed, it sits idle for a  period  of time to allow decomposition and  settling
to take place.   Since  the site  does  have alternative uses, the original
investment in the land can  be recouped.  Thus, the true cost of the  land to
the municipality  is  the cost of the money used to purchase  the  land, i.e.,
the interest rate.

Mandatory Source Separation

        Once the most  valuable wastes have been skimmed, the remaining wastes
are collected from their points of generation:  households, businesses,  and
other  commercial  establishments.  It is at this point that mandatory source
separation can take place.  There  are two major types of separate  collection
programs, one  in  which newspapers  only are picked up, the  other in which
newspapers and mixed cans and glass are collected.

        Newspapers are  the easiest  component of residential  solid waste to
separate and to maintain in a relatively clean  condition, i.e., free  from
contaminants such as  moist food wastes.  The newspapers/mixed  cans  and glass
program requires  each  participant  to separate  his  wastes into three
categories:   newspapers, cans and glass, and all other wastes.  The time and
storage requirements are greater than those for  the  newspaper  program,  but
the possibility  of  diverting the large can and glass fraction of municipal
solid wastes could  prove economically attractive  enough for the  city to
mandate such a  program.  The attractiveness must be tempered,  however, with
the realization  that the more time  and effort  the  program imposes  on
participants, the greater the likelihood that people will not  cooperate.


                                   164

-------
        The  success of mandatory  source separation programs depends on four
 factors  (EPA,  1977):  1)  the  availability of long-term markets for  the
 recovered materials, 2)  an ongoing  publicity campaign to maintain  the
 public s awareness  of the  program,  3)  careful  planning,  and A)  an
 antiscavenger ordinance that prohibits pickup of  the separated materials by
 any party other than the municipality or the municipality's contract hauler.
 Success can  be partially measured by the participation  rate—the  percentage
 of residents in  the collection area  who place separated materials out  for
 collection.  Because many of the incremental collection  costs  aresfixed,
 increasing participation in the program means lower  costs and usually greater
 revenues  to  offset those costs.

        This is plainly  shown when comparing  the  Marblehead and Somerville
 separate  collection programs.   During  the first nine months of 1976,  the
 Marblehead  program with a participation rate of  75  to  80 percent  resulted in
 net  savings* of  around $2,857 per month.   Approximately 26  percent of  the
 total waste  stream (by weight) was diverted from  disposal.  In the Somerville
 program,  on  the  other hand,  participation was  much lower during the same
 period of time so  that net savings averaged about $333  per month.  Only 8
 percent of Somerville's waste stream was diverted.   Eliciting the  community's
 support and cooperation,  then, is probably the key  factor  in determining  the
 success or  failure  of a separate  collection  program.   Details of these
 programs  can be  found in several  publications (EPA,  1977;  McEwen  1977-
 Resource Planning  Associates, Inc.,  1976).                            '     '

        The  future  of such source  separation  programs appears promising.
 They involve smaller  capital expenditures than do other methods of processing
 municipal  wastes.   There  are fewer uncertainties  associated with this
 approach than there  are with other more  capital-intensive methods.  Also
 with disposal  sites  becoming  increasingly  scarce and more  expensive'
 mandatory  source separation programs offer  a relatively  inexpensive way to
 extend the life  of  sanitary landfills.

 Refuse-Derived Fuel Plants

        While shredding can reduce landfill costs to a certain  extent,  the
municipality is  still  disposing of a large  proportion of  its  solid wastes
 (betwen 80 and 90  percent, even if the glass,  ferrous, and aluminum fractions
 are recovered).  In cases  where landfill  costs  are prohibitive,  or where
 energy  prices are favorable,  the  municipality may find  that some form of
energy recovery  is necessary to keep disposal  costs  at a manageable level.
There are  two routes a  municipality  can  take to recover energy  from its solid
wastes:  1)   produce  a refuse-derived fuel  (RDF), or 2)  incinerate the
wastes  to generate steam.  This section  considers the former alternative-
incineration is  discussed in a later section.
     incremental collection costs.
                                   165

-------
        The RDF  option's produce a  fluff  RDF consisting of  combustible
 particles with an  average size of one  inch  and with most  of  the  dense
 organics and inorganics  removed.  It is primarily used as a  supplementary
 fuel  in coal-burning facilities.  Typical  processing operations are depicted
 in the flowchart  of  figure 2.

        The waste  stream is first  shredded to an average particle size of
 four  to eight inches.   A  portion of the  ferrous metals is  recovered by
 magnetic separation and the remaining wastes are air classified.  The light
 fraction is passed  through a trommel to  separate glass fines and  any
 remaining heavies.   The wastes that  do not  fall through the trommel are  then
 shredded again to produce particles ranging  from I/A inch to 2  inches in  size
 and with a Btu content of about 5,000 Btu per pound.  The fluff is  then piped
 pneumatically  to  a storage location and then to a coal-fired boiler, where it
 constitutes between  10 and 20 percent of the charge.

        If no  materials  recovery is being undertaken,  the heavy  fraction,
 including the materials  that fell through  the trommel, is  landfilled.
 However, if any materials recovery is to take place, the heavies from the air
 classifier and  the  trommel undergo a second stage of magnetic separation.
 Aluminum is recovered (if desired) by eddy  current separation  and  glass (if
 desired) by froth flotation.  Any remaining  residue is landfilled.

        This procedure can  recover approximately 90 percent  of  the  light
 fraction  in   the  form  of  fluff  RDF.   This would,  of course,  mean a
 considerable savings  in  disposal costs, since  the lights  constitute  about
 three-fourths of municipal solid wastes.   Adding materials recovery modules
 to the basic RDF  system  will also reduce disposal costs.

 Shredding and  Materials  Recovery

        Under  certain circumstances, such as high land prices or favorable
 secondary materials prices,   it may be more economical for a municipality to
 process  its waste stream  prior to  disposal or to undertake  the recovery of
 certain materials, rather than simply landfill all of it wastes.   The options
 shown in  figure  1  give  the municipality  this sort of flexibility in
 determining its least-cost waste management  program.

        The flowchart  of figure 3  illustrates the shredder options.  The
 first stage of this  set  of  alternatives is  to  pass all wastes  through a
 shredder.   The  size  reduction achieved by  shredding is considerable; the
wastes leaving the shredder have an increased  density, which means that
 larger  quantities of waste can be landfilled  in the same  amount of land
 compared with  unprocessed wastes.   After shredding, the density of  the  waste
                                    166

-------
Mixed
Hastes
Primary
Shredder
Key:
     mixed solid wastes

r._ 	 ,'-\ ___

Drum I^^J Air Classifier
Magnet ^^^^
i
L_ _ _ _ _^
. organlcs, paper

Trammel

A
*
Drum
Magnet





'~A '
i
i
i
i
i
..J



Secondar
Shredder
y


Eddy
Current
i
*
0 «
Froth
Flotation
1
1
1
1
1
._ 	 *__








Fluff
RDF

*• Ferrous

J Aluminum

J Glass
1

Sanitary
Landfill
           Figure 2.  Solid waste processing in refuse-derived fuel plant.

-------
oo
                                                Drum
                                               Magnet
Air Classifier
                                                  L	f
                    Key:
                         mixed solid wastes

                         "light" fraction, e.g.. paper.  Inorganics

                         "heavy" fraction, e.g.. Inorganics
                                                                                 i	J
                                                                     Drum
                                                                    Magnet
Trommel I	1->
                                                                                    Eddy Current
                                           Aluminum
                                    Figure 3.  Solid waste processing by shredding.

-------
stream is about 30 percent greater  than that of unprocessed waste, i.e.,
about  26,000 tons of shredded  waste can now be landfilled on each acre  of
land.  The result  is to extend the life of an existing landfill or to reduce
the  number of acres required  to dispose of a  fixed amount of waste.  The
shred-dispose option, one  of  two  process-dispose  options noted  in  the
overview above, is represented in the  flowchart by the heavy arrow connecting
the boxes labeled  "shredder"  and "sanitary landfill."

        If any type of materials recovery is to be undertaken, the shredded
wastes first move on a conveyor belt  under a drum magnet,  the first stage  of
ferrous metals separation.  If no other materials  are being recovered, the
shredded  wastes are passed under a second drum magnet; the wastes are
landfilled and  the recovered  ferrous scrap undergoes cleaning and  some
additional  processing  prior  to  sale.  Magnetic  separation can  recover
approximately 90 percent  of the ferrous metals in the waste stream.   It  is a
proven tehnology.  The  recovered  materials can be sold to detitmers  or
directly to some  steel producers.

        The  processing  sequence is slightly altered if glass or aluminum is
to  be  recovered in addition to the ferrous metals.   As shown in the
flowchart, after passing under the first  drum magnet  the shredded waste
stream is air classified  by dumping  it into  an upward-moving  column  of  air.
Basically,  the air classifier divides the waste  stream  into  a  "light"
fraction consisting of organics, such as  paper, food, and yard wastes,  and a
"heavy" fraction composed  of inorganics,  such as metals, glass, dirt,  and
rocks, and dense  organics,  such as wet food  and yard wastes.   The  light
fraction is landfilled. The "heavies"  pass under a second drum  magnet  for
additional recovery of ferrous metals, then into a trommel, a  cylindrical
revolving screen that  allows "heavy"  particles  smaller than four inches to
drop through for  further  processing.  Particles remaining in the trommel are
landfilled.

        The feed  to  the  glass or aluminum  separation modules  consists  of
glass, aluminum,  rocks,  dirt, bones, some  ferrous metals, heavy organics  and
some  other  inorganics (Levy and Rigo, 1976).    These modules are
complementary—separation  of  aluminum leaves a fraction  that  is rich in
glass,  and vice versa.   However, as shown by the  flowchart, we allow for  the
recovery of one, the  other,  or both.   If glass and  aluminum  are  to be
recovered, the remaining heavies are passed  "on  a  conveyor belt over a linear
motor.  Electric current in the motor creates a magnetic field that induces
an eddy  current in any conductor (metallics) passing  through the field.  The
 induced current in  the  conductors opposes that being  generated by  the linear
motor  strongly enough to knock  the  conductors off the  belt.   Eddy current
 separation,  also known as the "aluminum magnet,"  can  recover about 70 percent
of the aluminum fraction.
                                    169

-------
        Glass cullet is separated  from the remaining stream of  heavies  using
 froth flotation.   In this process  the stream is treated  with  a reagent that
 gives  glass "hydrophobia surface characteristics," i.e.,  that  tends  to make
 it  float.  This stream is then  fed into the flotation chamber,  which contains
 a liquid undergoing constant agitation by rotors.  The air bubbles produced
 by  the agitation  attach themselves to the glass particles,  which then float
 to  the surface.  The resulting  froth is skimmed and cleaned.  The product  is
 a  clean  mixed cullet, or "sand,"  that can be  used  in small amounts  in
 glassmaking.  Compared to color-sorted cullet, however,  the product  of this
 model  is a low quality aggregate mainly suitable for use  in  making brpcks or
 "glassphalt."

        This  system can recover  about 66 percent of the glass in the solid
 waste  stream.  The  residue  remaining in the flotation chamber (and which does
 not float to the surface)  is removed and landfilled.

 Incineration

        Incineration is the second major form of  energy  recovery available to
 municipalities and  included in the  model.   Two  types  of incinerators are
 available to municipalities, depending on their capacity requirements.  We
 assume that a municipality  processing less than 500 tons per  day will use the
 small  modular, or  package,  incinerators.  These are available in capacities
 of 25 and 50 tons per day and can  be  connected  in series to achieve the
 desired capacity.   For those  cities requiring processing capacity between 500
 and 3,000 tons per  day,  the larger waterwall incinerators  are available.  The
 combustion chambers  of these incinerators are  enclosed by  closely-spaced
 water-filled tubes  that recover heat  from the  burning waste.  Steam  is
 generated by built-in  recovery boilers,  which also reduce  the temperature and
 volume of the exhaust  gases.

        Solid waste processing by incineration  is depicted  in the flowchart
 of figure 4.  One feature of  this flowchart that  bears noting is that  any
 materials recovery  that  occurs takes place on the "front end,"  i.e.,  prior to
 incineration.  While  materials recovery from  the  incinerated residue  is
 possible,  the recovered materials,  especially  the ferrous scrap,  are
 fundamentally different  from  the product  recovered  prior to incineration.
 Incineration of tin-plated  steel causes the tin  to fuse with the steel, thus
making detinning impossible.  Incinerator bundles,  as they are called,  are
 only  suitable for  use  as  scrap inputs to  foundries.  To  make the materials
 recovered by a municipal waste mangement program  homogeneous, regardless  of
 the option selected,  we  employed "front end" processing  of wastes  destined
 for  incineration.

        In the processing sequence depicted in  the flowchart,  bulky wastes,
 e.g.,  appliances,  are  manually separated  from  the  waste stream prior  to
 shredding.   This  applies only to municipalities  using the modular
 incinerators,  since  the  shredders feeding these incinerators  are too  small  to
 process  large, bulky  wastes.   After shredding,  the waste stream is  air
classified  and divided  into light and heavy  fractions.  The light fraction is
 fed into  the furnace  by  spreader stokers that move the wastes across the
combustion  chamber to  a  traveling grate.  This  is  a  semi-suspension

                                    170

-------
Mixed
Wastes
Key:







Bulky Shredder . Air Classifier '
Waste ••} •••* 1

1 i
« ^
' i
i
i
I 	 £
solid wastes
	 > Drum 	 .j, Trommel _ -A, Sanitary
Magnet ^ 7k ^ Landfill
„ _ i
' i

	 ^, 	 j

.... _ |~ I
Currpnt 1 	 1

i
I
i
      "light"  fraction, e.g., organics, paper

      "heavy"  fraction, e.g.. Inorganics
Froth
Flotation
	->I Glass
                   Figure 4.   Solid waste  processing by incineration.

-------
incineration process in which the waste is  partially burned  while falling
through  the chamber and  finishes burning  while lying on the grate.   Air is
introduced  below (underfire  air) and above  (overfire air)  the grate to
promote  the mixing of gases and to ensure  complete combustion.  The  ash and
any incinerated  inorganics are landfilled.   The  ferrous, aluminum, and glass
in the heavy fraction are  removed (if at  all) by magnetic separation, eddy
current separation, and  froth flotation,  respectively.  All unrecovered
residue is landfilled.

        Incineration reduces raw waste to  approximately 25  to  35 percent of
its original weight, including the glass,  ferrous, and aluminum fractions.
The steam generated is about three times more valuable than fluff RDF; this
occurs because the  incineration option includes  an additional  processing  step
(combustion), which increases  the cost.  Approximately 59  percent of the
energy content  of the original waste stream is recovered  in the  form of
marketable  steam (Levy and Rigo, 1976).

SOLID  WASTE  MANAGEMENT COSTS

        The  solid waste management costs for the options identified in the
previous section are discussed below. The  costs were primarily derived  from
data developed for  the Kansas City Metropolitan region  in a 1977 study by
Franklin Associates and Black & Veatch, Consulting Engineers.  In that study
costs  were  determined from contacts with equipment suppliers  and from
operating RDF plants for 500 and 1,000 tons per day with materials recovery,
and shredding plants with materials recovery.  Extrapolations were made for
other  plant capacities.   Because of the  limited data,  mandatory  source
separation is not covered.

        Although it is conceivable that in a few areas of the country, plants
larger than  3,000 tons per day may be the  best  option, those plants were not
considered  for this analysis.  Plants larger than 3,000 tons/day have special
siting problems  primarily due to the heavy truck traffic in  and out of the
plants,  and high  transportation costs.   They  cannot generally be treated as
simple extensions of the lower capacity plants.

        Three basic types of mixed solid waste  (MSW) plants are considered in
this analysis.   One is a plant which genrates a "fluff" RDF to  be burned  as  a
supplementary  fuel in a  coal-burning facility and which also recovers  some
combination  of steel, aluminum, and glass  for recycling.  Another  type of
plant  simply shreds the solid waste for landfilling and has the capacity for
adding material  recovery  facilities.  The third type  is an incineration
facility that generates steam.

        The  RDF  plants are generally similar to the  facilities  in Ames, Iowa,
Southwest  Supplementary  Fuel Processi  g facility in  Chicago,  and the
Americology plant in Milwaukee.  The shredding plant  is  similar  to the
Recovery I facility in New Orleans.

        The  RDF  and shredding plants  were  evaluated at  capacities  of  500,
1,000,  2,000,   and 3,000 tons of  solid  waste per day.   The  incineration
facilities were  evaluated at capacities of  50, 200, 500,  1,000,  and 2,000

                                    172

-------
 tons  per day.  For each of the capacities,  costs  were developed  for plants
 with no material recovery,  steel  recovery  only, steel and aluminum recovery,
 steel and glass recovery, and steel,  aluminum,  and glass recovery.  Plants
 with aluminum  and/or glass recovery  without steel recovery were not
 considered since it is essential  to remove  the  steel from the waste stream  to
 successfully remove the aluminum  and  glass.

 Sanitary Landfill

        The  initial per  acre cost of  the land depends  on several  things:
 alternative uses of surrounding land,  distance  of  the site  from  population
 centers,  population density  of  the  region, etc.  They are also  undoubtedly
 influenced by a host of region-specific  factors, such as the growth rates  of
 industry  and  population  in the  region, climate,  employment, etc.  A
 preliminary estimate of landfill  costs was  developed using data on landfill
 practices in 17 communities  (table 4).  The 1975 populations of  these cities
 ranged from 21,000 to 368,000 and  represented five of the 10 size  categories
 used  in the process model  (table 5).   Data  on  the cost  and  size of the
 landfill sites were available for  15  of  the 17  cities (two  municipalities
 were leasing the sites).  From these  data,  the per acre price of  the land was
 calculated; it ranged from $163 to $10,000 per  acre.  A simple  functional
 form relating price to the population  and to the square of the population was
 fitted to these data using ordinary least squares:

        LPRICE = 3.1461(POP)  + 0.0492(POP)  where
                 (0.47)          (2.01)

        LPRICE = price per acre

           POP = population  (in thousands).

 The  t-values for  the coefficient estimates are in parentheses.   The R  for
 this function was .76.   The predicted land values for  each  representative
 city  size  in the  model are  given in  table 6.  The estimated land  prices for
 the larger and smaller city  sizes tend to be extremely large  and small,
 respectively.  This occurs because  we extrapolated beyond  the range of our
data—a risky procedure at best, but one that is superior to  the  arbitrary
assignment  of "reasonable" values  to those out-of-range  cities.

Refuse-Derived Fuel Plants

        The RDF  plants for  which costs were developed  are  complete systems.
The items,  not all  itemized in the  analysis, but  which are included are
listed below.

        Truck scales
        Tipping  floor conveyors
        Primary  and secondary shredders with necessary input  and output
               conveyors.  (Only  50 tons per hour or larger  shredders
               were considered.)
        Trommel  screen—used  only  when there is metal or glass recovery
        Air  classifier  subsystem

                                    173

-------
   TABLE 4.  DATA USED TO ESTIMATE PRICE OF LAND FUNCTION
Landfill data
City-
Cape Glrardeau, Ho.*
Fairfax County, Va
Fresno, Ca.
High Point, N.C.
Independence, Mo.
JopUn, Mo.
KenosJra, Wfs.
Lancaster, Va.
Madison, Wis.
Madison, Wfs.
Nashua, N.H.
Norfolk, Va.
Odessa, Tex.*
Raleigh, N.C.
Richardson, Tex.*
Toledo, Oh.
Topeka, Kans.
Total land
expenditures
(10*$)
___
85
190
60
94
165
40
108
56
50
295
83
245
233
	
680
65
Size
(acres)
60
81
200
60
44.7
130
33
215-
H
24
295
83
1500
137
227
63
160
Price
($/acre)
_*_
1049
950
1000
2103
917
1212
502
4000
2083
1000
1000
163
2066
	
10000
406
Population, 1975
(10-3 persons)
31
21
177
67
111
39
80
ISO
168
168
62
287
78
111-
57
368
119
* Landfill site was leased.
Source: EPA, Office of Solid Waste Management Programs.
                                  174

-------
    TABLE 5.  DISTRIBUTION OF CITIES, BY SIZE CATEGORY,  1976
Size categories
U
500 - 1.000
250 - 500
100 - 250
50 - 100
25 - 50
10 - 25
5 - 10
2.5 - 5
Less than 2.5
Total (Average)
Number
of
places*
s
20
30
100
240
520
1385
1839
2295
627
7062

Total
(103)
19,748.6
13,738.2
10,947.6
15,026.1
17,602.0
13,390.0
22,539.2
13,738.2
8,586.4
853.6
141,674.0
Population
Percent of
U.S. population*
9.2
6.4
5.1
7.0
3.2
8.3
10.5
6.4
4.0
0.4
66.0

Average city
size (103)
3,291.4
686.9
364.9
150.3
73.3
36.3
16.3
7.5
3.7
1.4
20.1
*  1970 valued.  The 1976 population was used In calculating the total
   population and average city size for each category.

Source: U.S. Bureau of the Census, Statistical Abstract of the United
        States; 1977.  (98th edition) Washington, D. C.,  1977,  Table 20.
                                 175

-------
TABLE 6.  ESTIMATED PRICE OF LAND USED FOR SANITARY LANDFILL
           FOR REPRESENTATIVE CITIES
Size
Category
1,000 +
500
250
100
50
25
10
5
,2.5

- 1,000
- 500*
- 250*
- 100*
- 50*
- 25*
- 10
- 5
< 2.5
Average
population
(103 person)
3291.4
686.9
364.9
150.3
73.3
36.3
15.3
7.5
3.7
1.4
Land
price
(S/acre)
543354. 1
25375.2
7699.1
1111.4
495.0
179.0
64.4
26.4
12.3
4.5
 Estimates for these city sizes fall within the range of the data from which the
 coefficients were estimated.
                                  176

-------
        Pneumatic  RDF delivery system to storage and to  boiler
        Dust  collection  system
        RDF  storage  system
        Control  panels
        Residue  conveyors and bins
        Coal  burner  modifications
        Truck loading facilities
        Ferrous  metal recovery subsystem consisting of  two  stages of
                magnetic separation for each processing  line  and the
                associated conveyors and bins
        Aluminum recovery subsystem consisting of one aluminum magnet
                per  processing line and the associated  conveyors and
                bins
        Glass recovery subsystem (froth flotation)
        Delivery and electrical and mechanical installation of all
                equipment
        Mobile equipment
        Supplies and furniture

Some of the  assumptions  made for this cost analysis are:

         1.   The fuel generated is "fluff" RDF.

         2.   Costs  for  plants as  a function of  capacity  were
         determined  by the
         function:
               Ca    /Sa\n
               Cb  " Vsb/
         where C is plant  costs, S is plant capacity, and  n  is 0.8
         for all costs except  labor  and  land.  For  labor and  land
         the  exponent is 0.6.   These exponents were chosen on the
         basis of costs for  the 500  and 1,000 TPD plants.

         3.     Land  costs  for  RDF plants  are  calculated at
         $50,000/acre.

         4.  The buildings (reinforced concrete) are large enough to
         permit the addition of material recovery subsystems.

         5.   Interest during construction is equal to 6 months of
         interest on the total capital investment.   (It is  assumed
         that  construction time  is 2 years, during which time part
         of the capital may  be reinvested.)

         6.    Startup costs are equal to  the  operating  and
         maintenance costs for one year.  (This  is  to operate the
         plant during shakedown,  when there is little or no revenue
         expected.)
                                    177

-------
          7.  Labor costs are based on Kansas City,  Missouri  rates
          and include 40 percent for fringe benefits, overtime, etc.

          8.  ML RDF plants are  assumed to process solid waste 2
          shifts  per day,  300  days per year  with an 85 percent
          availability.   The availability factor  is  to allow  for
          routine maintenance and breakdowns.  System costs were  not
          developed on the premise of providing standby process lines
          for use in the event  of breakdown.  However,  diverters  and
          RDF trailers are included to be  used in the event of  power
          plant breakdown.

Shredding and Materials Recovery

        The  analysis for the  shredding  and shredding with material  recovery
plants is similar to the RDF plants.  However, only one stage of  shredding  is
needed and facilities and equipment for energy recovery are not included.   In
the plant with no  material recovery or with only steel recovery  the air
classifier system is not needed.

Incinerators

        Cost  for the modular  incinerator are given for 50, 200,  and  500 tons
of solid waste per  day.  The  data for the 50  and 200 TPD  plants  without
materials recovery came  from the  Franklin Associates Ltd. report; the 500 TPD
data were obtained by extrapolation.   The costs for recovering materials were
estimated,  based  on the previously submitted RDF data.  There  are  no  known
municipal modular incinerators  in operation where materials are mechanically
separated prior to incineration.

        The mechanical  processing consists of single  stage  shredding, air
classification, magnetic  separation, aluminum magnet separation, and  glass
separation by screening  and  froth flotation.  Because of the smaller size
requirement for  the shredders, large  bulky items would need to be handpicked
from the waste.  (In order to shred bulky wastes, 50 ton per  hour or larger
shredders would  be  required.)   Included are cost estimates for truck scales,
mobile equipment for feeding  the  solid waste  into the incinerators, and
delivery and installation of all equipment.  The systems  for which cost
estimates are made include single  lines of mechanical processing  and two
parallel  incinerator units for  the  50  TPD plants, four  for the 200 TPD plant,
and eight for the 500 TPD plant.  Each incinerator  is equipped with  a steam
boiler  which generates about 3 pounds of low pressure steam  per  pound  of
solid waste.  Steam  lines are provided to a user approximately 500 feet  away
from the  boiler.

        It is assumed that incinerators would operate continuously, with the
mechanical  processing  being done eight hours per day,  five days per  week.
Amortization costs are calculated at 7 percent and  20 years.
                                   178

-------
        It  is  assumed that Incinerators would  operate continuously,  with the
mechanical  processing being done eight  hours per day, five days  per  week.
Amortization costs  are calculated at 7  percent and 20 years.

        The  semi-suspension cost data were  derived from actual contract  costs
of the 1,000 TPD recycle energy system  under  construction  in Akron, Ohio/
The  process In Akron consists of coarse single stage shredding of  the  solid
waste; combustion  of  the light fraction in  semi-suspension  fired waterwall
boilers;  recovery of ferrous metals; and  future recovery of glass and
aluminum.  The aluminum magnet and glass screening and froth flotation system
costs were  added from the Franklin Associates  Ltd. data.

        In  the  semi-suspension  process,  the shredded light fraction  is
approximately 50 percent  burned while  falling through  the flame,  thus
improving the  boiler  efficiency over mass burning.  The  remainder is burned
on  a grate.  The  energy  product  is  steam which is distributed  to  steam
customers which must  be located within a mile  or two of  the plant.  In the
Akron plant,  steam  generation  is  estimated at  4 pounds per  pound of solid
waste.  After  the  steam for  in-house use and distribution system losses  is
deducted,   it  is expected that 3.2 pounds steam per pound  of  solid waste will
be available for sale.

        Amortization costs  for the equipment  and  buildings  are calculated at
7 percent and 20 years.

        Average availability of  the modular plant was  assumed to be the same
as  for  the  semi-suspension plant,  that  is, 85 percent,  350 days per  year.  It
is  assumed that  the incinerator will produce  saturated  steam at 560 psig
continuously and that the mechanical processing equipment will operate  5 days
per  week, 8 hours per day.

PROCESS MODEL

         The sol id waste  management  options and their costs are included in a
solid waste  management process model (SWMPtt)  along with  relevant  waste
generation rates and materials  prices.   The decision maker, confronted with
an  exogenously  established  flow  of solid wastes  is assumed  to  select  the
 solid waste  management option  that  minimizes the cost of waste management.
The analysis  is done for the average city  sizes shown in table 5.   The flow
of  operations is SWMPM is presented in figure 5 and discussed below.

Enter Secondary Materials Prices

         The first step in the operation of the model is to determine the  set
 of  secondary materials prices  that will confront each representative
 2.    Hileman, G. P., and F, B. Pyle,  "Recycle Energy for Central Heating  and
      Process  Steam," presented to International Heating Association, 1977.
                                     179

-------
C    START    )
0~

1
i
Select secondary
material price vector


/Average, high,
/ each material


and /

 Select cost minimizing
 solid waste management
 option
 Weight results
                                         Population,  by
                                           municipal ity
                                         Per capita waste
                                           generation rate
                                         Masta component



Estimate voluntary
recycling, by
material
< 	 1

* •
Estimate recycl ing
through mandatory
source separation
programs
	 *,
Adjust volume and
composition of
waste stream
4-



/ proper
\
Compute vo
stream, by


/Cost of m<


tions /
r
lume and
n of waste
municipality

/ . Supply elasticities, /
Y hy mflf«"l»l /
/ . Baseline recycling /
/ rates, by material/


mdatory source /
on programs /


 Cost  of  sol td waste
   management options :
 Energy prices
• Interest rates
 Outputs:

  . Price  and quantity supplied,
     by material
  . Energy price and quantity
     recovered as energy,  by
     material
  • Summary of waste management
     options selected, by
     municipality         	
      Figure 5.   Flow of operations  in the  supply model,
                                180

-------
   ow'    v- Elem*ntS  °l thlS Set are drawn from the universe of "high" and
  low   deviations from  the average  price of each material.  A total of 16
 unique price vectors^ is  selected by the model;  each drives the model  and
 the resulting price-quantity pairs constitute the  data set.

         Selection  of  a price  vector  has  two major effects:   1)  for  each
 material  the percentage  change of  the  high or low price from the  average
 price  elicits some  percentage change  in the level of  voluntary  source
 separation through  an elasticity calculation, 2)  secondary material prices
 obviously impact the  conomics of municipal solid waste management by altering
 the potential revenues available from recovering and selling the materials?
 The rationale for  selecting the range of  price variation and the exact method
 used to vary the prices are discussed in  the  section on Price Variation.

 Calculate  Amount and Composition of  Solid  Waste

         For  each representative municipality,  the total waste generated each
 year is the product of the  population  and  the per capita waste generation
 rate.   We  apply a vector of fixed material  proportions  to  decompose this
 total   into  the  five material  components used in the model:  paper,  glass
 ferrous materials,  aluminum, and all  other wastes.

 Estimate Extent of Voluntary Source  Separation

         At  this  point we estimate the amount of each secondary material  that
 is voluntary  separated at its  source  of generation, because the level  of  this
 activity directly affects the  amount of  waste collected and disposed  by  the
 municipality.  Voluntary  source  separation  activities are  influenced  by the
 prices of secondary materials.  Thus,  each material  is  represented by a
 baseline recycling  rate  that is  a  function of the average material  price; any
 deviations from this  price  alter the baseline recycling rate through a  supply
 elasticity.  The resulting  percentage  change in quantity  supplied  is applied
 to  the amount of  that  waste material  to determine  the absolute level of
voluntary source separation taking place.

Estimate Effects  of Mandatory Source Separation

        As explained  above,  two programs are available  to municipalities, one
involving  separate  collection of  newspapers, the other for collecting
newspapers and mixed  cans and bottles.  The costs  of both programs depend on
 their  efficiency in  recovering the relevant materials.   In this and the next
step Uhe  cost minimization  routine)  the model  investigates  three
alternatives:  1)   all  municipalities opt for separate collection of
     The  number  of possible unique combinations of high  and  low prices  for
     four materials is 2 .  The price of  the  fifth material is held constant
     (at  zero),  since no materials are recovered from  "all  other" waste.
     Energy is  recovered, and we account  for  it  in the model, but we did  not
     need to know the own- and cross-price  elasticities  of  the supply  of
     energy from solid wastes, so its  price was also held constant.
                                   181

-------
newspapers,  2)  all municipalities opt  for  separate collection of  newspapers
and cans  and bottles, 3)   no  municipality engages in separate collections.
waste management;  the least-cost alternative is  then  chosen,  ^he
municipality in effect has 48 waste management strategies to choose  from..

Revise Waste Stream Composition

        The  amount of each material in  the waste stream  is adjusted  to
reflect the  quantities diverted by voluntary and  mandatory  (if any)  source
separation activities.

Minimize Processing and Disposal Costs

        The  municipality  must process  and  dispose of  all remaining  solid
wastes by  using the least-cost mix of options available to  it.  The cost of
each  option is a nonlinear function of  the amount of waste to be processed.
Revenues  are derived from the sale of secondary materials and energy.
Unrecovered wastes are disposed of at a cost  that also  varies with the
amount. This is  the general form of the  net cost function that is minimized
by the nonlinear  optimization algorithm:
        min   rnin
         k   L i
                                2i
   a  [X(l-Rk)     + a3.[X(l-RkMl-E.)
                                      I4i
              -X(l-Rk)E.PJ
                                             for k = 1,2,3
                                                i = 1, .. . , 16
subject to


        X(l-Rk)Ei
         - Rk)(l  -E.)
                                            = X*-V
where
X

R
E.
 1
                      total waste to be processed by the municipality
                      proportion of waste stream  diverted by the k
                      tory source separation program
                                                                 th
                                                                    manda-
                      proportion of waste stream  recovered by the i
                      basic option

                      price of recovered materials
                                                                  •fch
4.   The  48  alternatives  come  from the  16  basic options, each applied in
     conjunction with the  three separate collection options.
                                    182

-------
                 ^k =  cost of the k   mandatory  source separation program

                 X*  »  total waste generated  in the municipality

                 V   «*  amount of waste stream diverted by voluntary
                       source separation

                 ai  »  estimated cost coefficients.

         The  first term is the cost of processing all wastes  (net of those
 diverted  by a mandatory source separation  program)  by the i   basic  option;
 the second  term is the cost  of disposing  of  the wastes  remaining after
 processing.   These  costs have three components:  land,  operating and
 maintenance, including labor, and capital.  The  third term  is  the revenues
 generated by  processing wastes using  the i   basic option.  The fourth and
 fifth terms represent the  costs and revenues,  respectively, of  the  k
 separate  collection program.   The constraint is that the amount recovered
 (first term) plus the amount disposed (second term) plus the amount diverted
 by mandatory  source separation (third  term) must equal  the total  amount of
 waste generated  in a municipality less  the  amount that is  voluntary recycled.

 Outputs

         Once the above steps  have been performed  for  each  representative
 city, the  results are weighted  by the number  of cities in  the U.  S.  in each
 size category.  The outputs of  primary  interest are the price-quantity pairs
 constituting the data set.   Other outputs include  the amount  of materials
 recovered  as  energy  and  a  summary  of  the  options  selected by  the
 municipalities for handling  their solid wastes.

 SUPPLY FUNCTIONS

         Supply functions were estimated for three secondary materials—glass,
 ferrous, and aluminum,  using  the  data generated  by the process model
 presented above.   Since  paper  is only  recovered as a fiber under voluntary
 and mandatory source separation,  its  supply was estimated  outside the  model
 but incorporated in the model,  as its recovery affects the energy  content of
 the waste stream and the economics of resource recovery systems.

 Price Variation

        The model requires three  prices for each material:   the  average, low,
 and high prices.   Deviations  above and below the average  price are used to
 determine the  amount  of "front end" recylcing (i.e.,  source  separation) that
 occurs.   The municipality is  confronted with combinations  of  the high and low
 secondary materials  prices—16 in all.

        This system  of price  variation is a 24 factorial design  In which  two
 prices  for paper,  two  for  glass,  two  for steel,  and two  for aluminum are
 tested in all combinations.   The  16 resulting price  vectors  represent  points
 at the  16 corners  of the  experimental  region,  which is a  four-dimensional
hypercube.  Thus, the range of price  variation determines the size  of  the

                                   183

-------
experimental region over which  the mathematical  model of secondary materials
supply  is  assumed to be valid.  However,  as  the  size of  this region
increases,  the  likelihood  that  the  model is valid over the entire  region
decreases.   Hence, the range over  which the prices  are varied  is  a critical
factor  in  the  estimation  procedure.   In determining  the  range of  price
variation,  we started with average 1976 prices and  considered the  historical
variation in secondary material  prices along with the probable effects  of  the
disposal charge.  The values are shown in table 7.
     TABLE 7.  SECONDARY MATERIALS PRICES AND SUPPLY
                ELASTICITIES
Price
(dollars per ton)

Paper
Glass §
Ferrous metals §
Aluminum
Low
5.0
0.1
9.0
200.0
Average
25.4
10.0
39.0
300.0
High
60.0
40.0
69.0
380.0
Supply
elasticity*
Low
.4
.1
.1
1.1
High
1.7
.71
3.9
4.3
     Energy ($/thousand Btu)

       RDF                    0.5

       Steam                  1. 2
1.5

3.6
   *  Estimates supplied by the Resource Conservation Committee.
   +  Prices developed from Official Board Markets.   Since OMB quotes
      are delivered prices,  they were multiplied by . 5 to approximate
      prices paid to generators.
   §  Franklin Associates Ltd. , Memoranda to EPA and ICF, February
      17, 1978.
                                   184

-------
 Supply Estimates

         The data  set  generated by the process model  consisted of  16 sets of
 secondary  material quantities and prices.  A function  that  was linear  in the
 logarithms  of its variables was fitted to these data using ordinary least
 squares.   The  functional form is given by
       InO.  =
where           Q.
                a.
                 i

      > V P3>P4
= the quantity of the ith  material  recovered, i=2-4,

= the regression coefficients,

= prices the municipality  receives  for the seondary
  materials, and
                P5    ~ price the  municipality receives for recovered energy.


The  resulting  elasticity  of  supply estimates are shown  below.  They
incorporate voluntary source separation  elasticities (Table 7) along with  the
estimates of the response  of municipalities to the price changes.
;- — —
Paper
Glass
Steel
Aluminum
Energy

Paper
1.7
0.022*
(0.85)
0.068
(2.56)
0.050
(1.97)
-0.123
(-28.23)
Coefficients
(t-statistics)
Glass
0.0
0.190
(17.63)
0.049
(4.43)
0.044
(4.16)
0.003
(1.76)
Steel
0.0
0.136
(4.27)
0.396
(12.13)
0.246
(7.96)
0.016
(2.93)
Aluminum
0.0
0.120*
(1.19)
0.207
(2.01)
1.102
(11.27)
0.010*
(0.59)
Energy
0.0
0.014*
(0.70)
0.065
(3.05)
0.051
(2.56)
0.236
(68.01)
R2

.928
.882
.896
.995
    *Not significant at a =  .10  level.
                                   185

-------
        The estimated supply elasticities confirm  the belief that the supply
of secondary materials  from the municipal  solid waste  stream is fairly
unresponsive to  price changes.  The data confirm the interdependence of  the
supply of secondary materials from the municipal waste stream.

CONCLUSIONS

        The combined use of process and econometric models provide a  means
for estimating the supply elasticities  for products for which there  is no
adequate historical record or data base for use in a more positive approach.
It can be used to make predictions of quantities supplied for sets of prices
not used in the estimation (the new  price  vectors must  fall within  the
experimental region, however).  Because  we are using a process model,  the
resulting data depict ponts on a well-defined  production surface.

        The data are subject to two  types of errors (Griffin, 1977):  1)
errors  in measurement  that result  from differences between the observed
technological  matrix and  the technically correct matrix, and 2)  behavioral
errors  that result because the frictionless long run cost minimization
subsumed by the  process  analysis/statistical reduction approach does  not
obtain  in the short run.   Another problem involves determining the size of
the experimental region.  If it is too small,  then the process model tends to
generate an "all or nothing" response which can render statistical estimates
useless.  Over too large  a region, the mathematical model may not accurately
represent the  responses.
                                    186

-------
                                REFERENCES


1.   Baseline Forecasts of Resource Recovery, 1972 to 1990.  Midwest Research
     Institute. Final Report for Office  of Solid Waste Management. U.S.  EPA,
     March 19757                      ~		

2.   Darnay,  Arsen and William E. Franklin, The Role of Packaging in Solid
     Waste Management  1966 to  1976. Prepared for Public Health Service
     (U.S.H.E.W.) by Midwest Research Institute, 1969.

3.   Griffin, James  M.   "The Econometrics of Joint  Production:  Another
     Approach," The Review of Economics  and Statistics. LIX, November 1977,
     pp.  389-397T     ~~~

4.   "Long-Run Production Modeling  with Pseudo Data:   Electric Power
     Generation," The  Bell Journal  of Economics  8(1), Spring 1977,  pp.
     112-127.        ~~

5.   Hunt, Robert G., and William L. Bider, Analysis of Environmental and
     Economic Impacts of Waste  Reduction Procedures and Policies. Final
     Report prepared for Office of Solid Waste Management, U.S. EPA, December
     1977.

6.   McEwen,  L. B., Jr., Waste Reduction and Resource Recovery Activities;  A
     Nationwide Survey.   Environmental  Protection Publication SW-142.
     Washington, D.C., U.S.  Government Printing Office, 1977, 78  p.

7.   Quiraby,  Thomas H. E., Recycling;  The Alternative to Disposal. The Johns
     Hopkins  University Press,  Baltimore, 1975.

8.   Savas, E. S., The Organization and  Efficiency of Solid  Waste Collection.
     D. C. Heath and Company, Lexington, Mass., 1977.

9.   Sawyer,  James W., Automotive Scrap  Recycling;   Processes,  Prices, and
     Prospects. Resources for the Future, the Johns Hopkins  University Press:
     Baltimore, 1974.

10.  Shumway,  C.  Richard  and  Anne A.  Chang, "Linear Programming versus
     Positvely Estimated Supply Functions:  An Empirical and Methodological
     Critique," American Journal  of Agricultural Economics,  Vol.  59 (2), May
     1977, pp. 344-357.

11.  Stevens, Barbara  J.,  "Pricing of  Solid Waste Collection and Disposal,"
     Memorandum, date unknown.


                                   187

-------
12.  "Cost of Refuse  Collection and Disposal," Memorandum to Fred  Smith,
     March 1, 1978.

13.  U. S. Department  of Commerce, Bureau of the Census,  Statistical Abstract
     of the United  States;  1977, (98th edition),  Washington, D. C., 1977.

14.  U. S.  Environmental Protection Agency,  Office  of  Solid Waste Management
     Programs, Resource Recovery  and Waste Reduction, Fourth  report  to
     Congress,  Environmental Protection Publication  SW-600, Washington, U. S.
     Government Printing Office, 1977,   142 p.
                                   188

-------
               SECONDARY MATERIAL  DEMAND AND SUPPLY RESPONSES



                                    by


                             ROBERT C. ANDERSON
 INTRODUCTION
        The desirability of  either  the solid waste user fee or the consumer
 goods  product  charge can be evaluated according  to several criteria:
 administrative feasibility,  effectiveness, efficiency, equity, legal  and
 political feasibility.   Other participants in this conference have addressed
 the issues of efficiency,  equity and  administrative  feasibility.   In  this
 paper I will address the issue of effectiveness with special reference to the
 consumer goods product charge.  The principal analytical  tools are a set  of
 econometric models of  the  ferrous  scrap and wastepaper  markets, developed
 over  the  past few  years under grants  from the  Environmental Protection
 Agency.

        As currently envisioned, the consumer goods product charge would  be
 structured as an excise tax on inputs to the production of consumer goods
 which ultimately  enter the solid waste  stream (e.g.,  beverage containers,
 newspapers, packaging materials, etc.).  Of two principal  production inputs'
 virgin materials  and secondary materials,  only the virgin inputs would  be
 subject to the product charge.  This would create a  change in the relative
 prices of  virgin  and secondary materials and, theoretically,  produce  shifts
 in resource use patterns and in  resource recovery  as the  effects filter
 through the economic system.

        In order to project accurately the probable impacts of regulatory
 actions such as  the  product charge,  one must have  a detailed knowledge of the
 market  parameters  of  interest, the demand and  supply elasticities for  both
virgin and secondary  materials.  This  paper is principally  devoted  to  a
 discussion of our  most recent empirical- work in this area.  Quantitative
 assessments of  the likely impacts of the product  charge based upon estimates
presented  here  are left  to others.
                                   189

-------
        As  noted the econometric models presented here are  extensions of  the
ferrous  scrap and wastepaper demand and supply models developed for EPA in a
previous study. *•  Three major modifications have  been made to the earlier
published  models.   (1)  The  original ferrous scrap model was based on the
years 1962-1972; this model has been updated to include data through October
1977.   The wastepaper model was previously estimated over  the period January
1962-December  1974 and has been extended to August  1977.   This additional
period  is  especially interesting because of the large price changes in both
materials which occurred between 1973 and 1975.  (2)   The earlier ELI models
for wastepaper and ferrous scrap were two equation demand and supply systems.
Additional equations are  added here  to represent market behavior  more
accurately.   The  ferrous  scrap model now contains supply, demand, export
demand,  and mill inventory  equations.   (3)  In  the wastepaper model,  the
supply of ferrous scrap was initially hypothesized to be a  function of price,
activity in the  prompt producing sectors (automobile production),  scrap
exports, lagged prices and a trend variable.  Some  alterations were made to
this equation  to improve the fit (exclusion of  the  trend  variable) and to
accommodate  the export  equation (scrap exports were added to the left-hand
side of the equation).  A few other minor changes in the demand equation  for
ferrous  scrap and the wastepaper supply and demand equations are discussed
below.

THE WASTEPAPER MODEL

        The wastepaper model,  as  it  now  stands, is composed  of  four
equations.


        Demand :                                                       ( 1 )


        QtD =  a0 + alPt + a2X(t) + a3X3(t) + ^4(t)  + Ut


        Supply:                                                       (2)
        Inventory:                                                    (3)


        lt  = cO + clI(t-l)+c2Q°+c3AQf+c4p(t-l) + Vt
     Environmental Law Institute,  Impact of the Federal  Tax Code on Resource
     Recovery.   Prepared  for  the  Environmental  Protection  Agency.
     Washington,  D.G.:   PB-264  886,  1976.  Also,  Anderson, R. C.  and
     Spiegelman, R. D., "Tax Policy  and Secondary Material Use."  Journal  of
     Environmental Economics and Management (JEEM), A (1977):  68-82.

                                    190

-------
Where:
        Identity:                                                       (4)


        QSt=QtDH-dt-I(t-1))
        Q.       is the consumption of wastapaper in thousands  of
                tons
         o
        Q,       is the purchases of wastepaper in thousands  of  tons

        P,       is the wholesale price index of wastepaper deflated
                by the wholesale price index for all commodities
                (1967 = 199) and multiplied by 1000

        X.       is the Federal Reserve Board Index of paperboard
                        container production (1967 = 100)  multiplied by  10

        X-      is the Federal Reserve Board Index of converted
                paper product production (1967 = 100) multiplied
                by 10

        X,      is the output of construction paper and board  in
                thousands of tons

        X^      is the output of combination paperboard in
                thousands of tons

        W       is the wholesale price index for market wood pulp
                (all grades) deflated by the wholesale price index
                for all commodities and multiplied by 1000

        P,.  ..   is an Almon lag over wastepaper prices lagged  one
         *  "^    to four months

        I,       is inventories of wastepaper held by major
                consuming mills at the beginning of the month
                in 1000's of tons

        Estimates of the structural coefficients  (the a's, b's  and c's)  of
equations  (1)  through  (3)  were obtained by  two stage  least squares. All
equations were  corrected  for first  order serial  correlation.   The  final
estimated equations are reported below.  T-ratios appear in parenthesis  under
the respective variable.^
2.   The properties of T-ratios and other test  statistics  in  simultaneous
     equation systems have not been fully developed.  The T-ratios should be
     interpreted as guidelines to the significance of the variables.


                                    191

-------
        Demand:                                                       (5)


        Q?  = 41.9  - 0.13P. - 0.03W. + 0.37X,. + 1.51X4.
          *          (-4.3)  * (-1.0)  *  (4.8)     (15.9)

        Supply:                                                       (6)


        Q? = 51.9 + 0. 36P, + 0.08X. ,.  ., + O.Z5X ,  ?.  -  0. 004P    .
          t         (11.4)  *  (1.5)  I(t"1}  (4.8)  *(t-*}   (0.09)  lt>J)

          P = .45

        Inventory:                                                     (7)


        I, =  -39.6 + 0.951,, .. + 0.04QJ3 - 0.08 AQ?5 + 0.03P,  ..
         t          (21.8) (t"1}  (0.7)  *  (-1.2)    t  (1.7)  ^-J'

          P =  -0.05

        All  the variables but one have the  proper sign as suggested by
 economic  theory.  Demand responds negatively to price.   The  estimated
 elasticity of demand,  evaluated  at  the means,  is 0.19.   Consumption of
 wastepaper is a positive function of  activity in construction  paper and
 board,  two sectors of  the  paper industry that  are  large users of  wastepaper.
 While one would expect  the  price of woodpulp to have a positive  impact on
 wastepaper demand,  the estimates presented here show  a negative sign,
 although the variable is  not significant.

        The  supply  of  wastepaper  is  positively  related  to price,  with an
 estimated short-run elasticity of 0.5.   The negative  sign for lagged prices
 is  also expected since  the level of  prices in past time periods  provides a
 good index to the  extent  of depletion to the reservoir of  available scrap.
 No  medium term elasticity is reported because the lagged price  variable is
 insignificant.  The generation of prompt industrial scrap, represented by the
 lagged  values for paper  board container and converted paper production, acts
 to  increase the supply  of  scrap as indicated by the positive sign  for these
 variables.

        Comparison of the current wastepaper estimates  with  estimates from
 the earlier  ELI models, reproduced in Table 1, must be done with some care.
In order to take advantage of the publication of a more  consistent  series,
 the price  variable  for  wastepaper was  changed in the  updated model.  The
previous model  used a wholesale price index  for  wastepaper  that was based  on
 two different base years, 1957-1959 and 1967 = 100, the  new series has been
weighted to a single base year,  1967 =  100.  It  does not  seem likely,  though,
 that this  modification  can account for  the  over twofold  increase in the
demand  elasticity, 0.08 in the  first study to .19 in the  current work.   The
                                   192

-------
 same sort of change was made to the woodpulp  price index, but cannot  be taken
 as  the cause of the negative sign for that variable.  The estimates of  the
 supply elasticity are  similar in both studies.
       TABLE 1.  COMPARISON OF ESTIMATED ELASTICITIES
       FROM PAST AND CURRENT ELI WASTEPAPER MODELS

Elasticity
Demand
Gross price
Supply (short -run)
Supply (medium- run
EPA
0.16
0.13
0.40
0.15
*
Estimate
JEEM
0.08
0.17
0.53
0.40
Current
0.19
>!<*
0. 12
0.50
0.50

     **
Reported elasticities refer to the EPA study cited earlier based
on the years 1962-1972, then the results published in JEEM
(based on the years 1962-1974).
Estimates based on the years 1962-1972.
        One cause of the difference in the  price coefficients of  the  demand
equations  may be found in the market convulsions of 1973-75.   The beginning
of the  period was marked  by  a world wide  increase in demand  for paper
products,  as  well as most other commodities.  This was shortly  followed by a
more  than  proportional  increase in the demand for  wastepaper  as
non-traditional users  entered the market.  Mills that normally  relied upon
woodpulp  inputs were buying  wastepaper  to compensate for a developing
shortage of pulp for paper production.

        In  response to increased demand,  the  price  of wastepaper more  than
doubled  from  early 1973 to the first quarter of 1974.  The  price  of woodpulp
also rose during this time period.   During  the period from October  1973  to
November  1974 consuming mills increased their  inventories of  wastepaper,
possibly in anticipation of even further  price increases.   (Time series  data
for these variables are plotted in  Figure 1).

        By  mid-1974, the commodity  boom had begun to slip  into the recession
period of  1975.  As demand for  paper products declined,  demand for wastepaper
inputs fell  even more as mills  used up accumulated  inventories  instead  of
buying  in  the market.  The wholesale price  index for wastepaper  fell from a

                                    193

-------
                    II LIONS OF SHORT TONS
O
i-
                       wastepaper p ice
                            index      A.
                                                                             purchases or
                                                                             wastepaper  /
                      stocks or wastepaper

                        - A/^
                                                            TIME
              Figure 1.  Time  series plot of wastepaper data, from January 1962 through August 1977

-------
high of 222.4 in March  1974 to 56.1 in March 1975.  The price of woodpulp  did
not  experience such  a  severe drop.  Actually,  the higher prices of woodpulp,
established during the  1973-74 period, held fairly constant  throughout  the
recession period of  1975.   As shown  in Figure 2, these  price patterns
signalled a change in the historical relationship between woodpulp price  and
the consumption of wastepaper.

        More formally, the inability of the wastepaper  demand model  to
capture the substitution effect of woodpulp for wastepaper can be formulated
as an error-in-variables problem.  It is likely that while  posted woodpulp
prices  remained  high  in the aftermath of the  shortage period, actual  prices
were somewhat lower due to discounts being given buyers.  The wholesale price
index  for woodpulp  would not account for such discounting.  The use of this
index would generate  biased estimates of the woodpulp price  coefficient  and
unreliable elasticity estimates.

        To demonstrate the effect  of the different price structure  for
woodpulp on the  demand for wastepaper,  the demand equation was re-estimated
using two woodpulp price variables, one representing prices  before 1973  and
another representing  prices after 1973.  It was anticipated that the price of
woodpulp prior  to 1973 would have  a positive sign and the post-1973 price
would capture the  change  in  the woodpulp price  - wastepaper consumption
relationship.   The  pre-1973 woodpulp  price had a marginally significant
positive sign and an  elasticity of  0.12.  The  post-1973  price  was  not
significantly different from zero.

        The higher price  elasticity of  demand, although still  quite
inelastic, can  be explained, at least in part, by the large increases  and
ensuing decreases  in the  consumption and price  of wastepaper.  It  seems
likely that industry  decisionmakers behave differently at different points in
the business cycle.   Mills may be more responsive  to prices that are rapidly
increasing than  to  prices  that are  falling. The fact that the elasticity
estimates presented in  Table 1 vary depending on  the time  periods on  which
the models were estimated lends some credence to this hypothesis.

        Although there  were major price  and therefore demand changes from
late 1972 to early 1975,  an analysis of the  residuals  from the demand
equation suggests that  the estimated equation was able to account for most of
the shifts in demand  that occurred during this  period.

        All of  the variables in the inventory equation have  the proper sign
as suggested by  the theory  outlined earlier, although only the once  lagged
price ad inventory  variables were  significant.3  The extremely  small
coefficient of  adjustment  (8), .05, suggests that inventories are adjusted
only a fraction each  month towards the desired  level.  This is understandable
3.   The specification  of  the inventory equation is discussed in detail  in
     "Analysis  of Scrap  Material Supply and Demand."   Draft  Interior Report
     No. 2 to the EPA.  Prepared by E.L.I., September  1978.
                                    195

-------
                 MILLIONS OF SHORT TONS
C
~
                                                        TIME
                  Figure 2.  Time series plot of woodpulp price and wastepaper consumption data.

-------
 given  the  variable  nature of  scrap  prices  and  the high  cost of holding
 wastepaper  in storage.  Past rates of change in consumption appear to be more
 important in  determining  inventory adjustment plans than current consumption.
 This would  be the  case  if many  of the mills that  consume wastepaper do not
 produce strictly to order.

         Several different forms  of lagged prices were used in  the inventory
 equation and all  generated similar results.  The once lagged  price variable
 was  chosen  based on maximum explanatory power of the equation.   The fact that
 stocks  of  wastepaper  are  adjusted  in response  to expected  price is not
 surprising in light of  inventory behavior during the 1973-1974 period
 described above.

 THE  FERROUS SCRAP MODEL

         The four structural equations making up the ferrous  scrap model  are:
                                                                        (8)
Qs
where:
         t = aO  + alPt  + a2 X3t + a3P(t-j)
           = C0  +ClPt  + C2X4t
             dO + Vft-l)  + d2X5t + d3AX5t
        Q       is net receipts of  ferrous scrap  (all  grades)  in  thousands
                of tons
         s
        Qt      is total  receipts of  ferrous  scrap  (net  receipts  plus
                exports)  in thousands of  tons

        Pt      is the wholesale price index  of ferrous  scrap  (all grades)
                deflated  by the wholesale price index  for all
                commodities (1967 =  100)  and  multiplied  by 1000

        X It     is total  u*s>  steel production in thousands of tons

        X2t     iS an ^mon la§ of  the wholesale price index of market pig
                iron deflated  by the  wholesale price index for all
                commodities (1967 - 100)  multiplied by 1000, lagged
                five to eight  months

        X 3t     is the Federal  Reserve Board  Index  of automobile  production
                (1967=100)  multiplied by  10
                                    197

-------
         X        is  an  index  of  total  foreign  steel  production
          •p>
         Q        is  net U.S.  exports  (exports  - imports)  of  ferrous  scrap
          *       all grades)  in  thousands of tons

         I        is  beginning of month stocks  of ferrous  scrap  (all  grades)
                  in  thousands of tons

         X_       is  the U.S.  consumption of ferrous  scrap in thousands of
                  tons
           (t-j)
         is an Almon lag on the ferrous scrap wholesale price index
         lagged one to four months
        The  structural  coefficients  of equations 8-11  were estimated by
 two-stage least squares  and corrected for first order  serial correlation.
The estimated equations  were:


        QD = 353.58 - 0.90P.+ 0.26X7. + 0.65X,/f  -x                    (12)
         *           (-1.8) * (9.5) "   (1.2) *{t~3)
              .53
Q^  = 350.85 + 2.16P.  + 0.67X.. - 0.04P                         (13)
 1           (18.9) T  (11.9)      (.24) lt>J)


 P  = .77


QE  = 402.49 - 0.05P.+ 0.076X4-                                 (14)
 t           r-n.^^   r-^.Ti
                     (-0.6)   (3.3)
              .65
            = 270.12 + 0.961,.  ..  - 0.02XC,.  - 0.02AXC. + 0.13P,.  .x     (15)
                     (46.4) (t-1)  (-0.9) bt  (-0.6)  5t (2.2)  (t-j)
              0.1
        The demand for ferrous scrap is negatively related to price although
the coefficient is only marginally significant.   The elasticity of demand,
estimated at the means, is 0.34.  The elasticity of scrap demand with respect
                                     198

-------
 to total production,  0.92,  indicates  that  demand for ferrous  scrap rises
 almost  proportionately  with increases  in steel production.   A one ton
 increase in the production of  steel increases the demand  for  scrap by .26
 tons.   The lagged pig  iron price has the  expected positive sign but  is not
 significantly different  from zero.

        All  of the  variables  entering  the  supply  equation  have the
 appropriate sign,  although  lagged scrap prices are not significant.  The
 short-run supply  elasticity at the means is 0.71.  The production of
 automobiles,  which in the process generates  prompt  scrap, augments the  supply
 of scrap forthcoming  at  any given price.

        Although  the U.S.  scrap prices  are shown,  in  the export demand
 equation,  to have an insignificant  effect on the foreign demand for U.S.
 scrap,  activity in foreign  steel industry  is an important determinant of
 demand.   An  increase  in total foreign steel  production has a positive  impact
 on the quantity of scrap exported.  There are several possible  reasons for
 the  lack of  an estimated  price effect.   (1)  It may  be that  exports are
 exogenously determined from the  rest  of  the model.  If this  is the case,
 scrap  exports  should appear as an explanatory variable  in  the supply
 equation.  (2)  While  foreign consumers may  actually respond to  the price of
 U.S. scrap,  the price variable used in the  equation may not reflect the  true
 U.S. export price  for scrap.  The price index used  in  the analysis is highly
 aggregated and may  give  more weight to  prices  in non-exporting cities or
 grades  of scrap that  are not normally exported in large quantities.  (3)  The
 equation may omit other  variables  that  are important in a  foreign scrap
 consumer's decision to import U.S. scrap.   The most obvious example of this
 is  the  lack  of a variable  defining  the  costs  of transporting scrap  to the
 importing country. It seems reasonable  that the level of transportation
 costs would be  a significant factor in explaining demand  for U.S.  scrap.  (4)
 While current prices  may not affect current  export demand,  an  expected or
 lagged  price  variable may play a more prominent  role in determining the level
 of scrap exports at any  given point in time.   This  same argument  applies to
 the  activity variable.  More attention  must be  paid  to the specification of
 the export demand  equation before firm conclusions  concerning  the existence
 of a price effect  can i»e made.

        Initially  it  was felt that the imposition of export controls during
 part of  the  data  period may have been one  reason for  the insignificant price
 effect  in the export  demand equation.   Export controls on ferrous  scrap were
 imposed  by the U.S.  Department of Commerce in July 1973 in an attempt to
 assure  domestic consumers adequate supplies  of ferrous scrap during a period
 of unusually strong  demand.   The first  form of export control  required all
 export  shipments of   over 500  tons to be -licensed, beginning with orders
 accepted after July  1.  Regulations for orders  of 500 tons or  more received
 after July 1  were  not  issued until later.  All orders of less than 500 tons
 would be granted licenses.   The effect of  these  control regulations  was to
 reduce  the size  of individual orders but appears to have  little immediate
 effect  on  the  total  amount  of exports.  In August 1973,  scrap  exports were
actually higher  than  they had been in July.
                                   199

-------
         In  September, the  Commerce Department  acted on the  problem with
 unlimited small orders and  stopped consideration of any small orders received
 after September  11,  although  orders  over 500 tons  which had  been received
 prior  to July 1 could  still  be  filled.  In addition, Canada  and Mexico  were
 allowed  to import up to 75,000  tons of U.S. scrap  per  month.

        The licensing  system was replaced by a quota system  in the  first
 quarter of 1974.   The quotas  were  based on historical export patterns and
 licenses were granted based on the  country to which the  scrap was  to  be
 exported.  The  quota was  set  at 2.1  million tons  for  each  quarter until
 export controls expired on December 31, 1974.

        An  attempt was made to estimate the  impact of  export controls on
 export demand.  A dummy variable was included in the export  demand equation
 that  had a value  of  0 for  the  period before and  after controls, and a value
 of 1 for the period August  1973 - December  1974.  The coefficient  of  the
 dummy variable was insignificant indicating that export controls may have had
 little impact, perhaps because  the demand for exports  was  already declining
 by the time controls were imposed.

        The inventory  equation  suggests  that the level  of  stocks held  by
 scrap consuming  mills is  predominantly a positive function of stocks and a
 negative function of scrap prices in the previous  month. The extremely small
 coefficient of  adjustment  ( ),  0.04,  indicates  that  like wastepaper
 inventories, actual stocks of ferrous scrap adjust slowly to desired levels.
 During a one-month period stocks  are  adjusted  only 4%  of  the amount  of
 adjustment required to reach the desired level of  inventories.   The absence
 of a  strong short-run buffer  stock motivation for holding stocks, evidenced
 by the insignificant consumption and change  in consumption  variables,  may
 indicate  the consuming mills react  to  longer-run  considerations.   The
 positive sign for the  lagged price variable seems  consistent  with a  visual
 analysis of the  price and  stocks data  plotted  in  Figure 3.  For the most
 part,  the level  of stocks rises  with increases in  the  ferrous   scrap  price.
 Two periods are  exceptions to this rule.   From early  1972   to late 1973,
 stocks  of  scrap  at consuming mills fell  quite  regularly  in the face  of
 sharply  increasing prices.  Mills were  depleting inventory levels to hold
 down the cost of  raw materials.  This trend was reversed  in  late 1974.   As
 scrap  prices began to  fall, mills replenished their inventories.  When prices
 returned to  more  normal levels,  stocks on hand continued  to  increase  above
 the historical average.   Further research  is required before  the inventory
 behavior in  these  two  periods can be  adequately specified in the model.

        Table  2 reports summary statistics  of elasticities for this model and
 our earlier  model.   The  fact that  the current model   shows  the demand and
 supply of  ferrous  scrap  to  be  less responsive to  price  signals  (more
 inelastic)  than in the earlier model  may be the result  of several factors.
These  will be  discussed in turn.
                                   200

-------
:-

O
               MILLIONS OF SHORT TONS

               M.» -I	
- user mills stocks of scrap    A.           J(    .•' \  ls

^v/^t^
^--
                                                                 .
                                                                       / '-'

                           by user mills
                                                                         '^
                                              TIME
            Figure 3.  Time series plot of ferrous scrap data, January 1962 through October 1977,

-------
            TABLE 2.  COMPARISON OF ESTIMATED DEMAND
            AND SUPPLY ELASTICITIES FOR FERROUS SCRAP

                                          Estimates
          Elasticity                 EPA*             Current

        Demand                     0.64'              0.34

        Supply:
Short- run
Medium- run
1.43
1.12
0.71
0.71**
      *  Environmental Law Institute.

      ** No significant medium term effect.
        (1)   Like wastepaper  demand, the  demand for ferrous  scrap inputs
increased during the commodity boom of  1973  more than proportionately to  the
demand for  final  steel outputs.  This was due to  a relative shortage of coke
for pig iron production.  To compensate  for the lack of pig iron, mill demand
for scrap increased.  In  response, the,price  for  scrap  rose to all-time
highs,  as  shown in Figure 3, between  the beginning of 1973 and early 1974.
Unlike wastepaper,  though,  the supply of scrap did not increase significantly
in  response to the new higher prices.   Several  structural changes in  the
market  acted in concert  to  limit scrap  supply and create a shortage of
ferrous scrap. 4   (a)  There was a decrease in  the availability of home  and
prompt scrap.  Prompt scrap generators returned scrap back to the originating
mills  instead of  selling  scrap to brokers  or dealers.  In addition, higher
yield  steel products were being produced, cutting  down  on home  scrap
production.  (b)  A. scarcity of gondola cars  constrained the movement of
scrap to areas with the greatest demand.  (c)   In  anticipation of higher
prices  and  further shortages, mills  began to  build up  inventories above
levels required by  current  consumption needs. The lower estimated elasticity
of  supply  from the current  model can,  in  large  part, be  ascribed to  the
shortage conditions described above.
4.    The Commodity Shortages of 1973-1974;  Case Studies.  Prepared for the
     National Commission on Supplies and  Shortages.  Washington,  B.C.:  U.S.
     Government Printing Office, 1976.
                                    202

-------
        (2)  The price controls,  that were in effect from 1972-1974, may have
 had  some effect  on the responsiveness of demand to price.   Analysis of the
 theoretical impact is clouded by  the different price control  regulations that
 were enacted during this period.   Generally,  controls were placed on all
 unprocessed industrial scrap.  Small  scrap dealers were exempt  from price
 controls while dealers with over 60 employees were not.  In December 1973 a
 one-time cost pass through was allowed on a dollar-for-dollar basis  for  both
 buyers and sellers.  Prices  were  set at the average price during the last
 quarter of 1973.   By early 1974, all ferrous scrap was exempted  from price
 controls.  Our attempts to estimate the impacts of price controls on  ferrous
 scrap  supply and demand were  unsuccessful.

        (3)   The  price indexes used here are highly aggregated,  fixed weight
 averages of the published market  prices.  These prices may  not  be  accurate
 representations of the true  prices to which scrap consuming  mills react.  To
 the  extent that they do not reflect the actual market prices, both the demand
 and  the supply elasticities may be understated and biased.

        Furthermore, the nature of the prices making up  the  pig  iron price
 index  may provide some insight into  the insignificance of the  variable in
 determining demand for scrap.   Since  the market price for  pig iron  is a
 posted price, and then only  accounts for a small portion of total pig iron
 consumed, it may not accurately reflect the relative scarcity or abundance of
 the  input.  The  impact of  the  availability  of  pig iron on the demand for
 scrap might be better captured by a variable depicting the production  of pig
 iron and not its  price.  A  pig iron  production index was included in the
 scrap demand equation.  Although  it did not generate entirely satisfactory
 results,  pig iron production was a significant and positive determinant of
 the  demand for scrap.

        (4)   Finally,  the definition of the  dependent variables for the
 supply and demand  equations may not be specific enough to display all  of the
 subtle changes that occur  in the  ferrous scrap markets during periods of
 rapidly increasing (or decreasing) prices.  For example, the  average grade of
 purchased scrap  is known to change with changes in the price  of scrap.

 CONCLUSIONS

        This  paper demonstrates  that  econometric models of secondary
materials markets  yield results which are consistent with known institutional
 relationships and with economic theory.  It is important  to recognize the
limitations of the econometric approach, however,  before directly  applying
 the  parameter estimated developed  here to policy problems.  The estimated
demand elasticities are believed  to understate seriously the  actual  long-run
 parameters of the scrap consumers.  For policy purposes it may be reasonable
 to employ much higher direct  price elasticities of demand and infinite cross
elasticities.

5.   This  is  a highly simplified view of the complex system of price control
     policies and regulations enacted during 1972-1974.   A more detailed
     outline  of  the price  regulatory events  can be  found in  The Commodity
     Shortages of  1973-1974;   Case Studies.                    	


                                    203

-------
        On the supply side,  one important  adjustment must be made.  If one is
interested in the impact of product  charges  or other  similar forms of market
intervention on the recovery of post-consumer secondary materials, one  must
adjust the elasticity estimates.   This  adjustment is necessitated by the fact

        On the supply side,  one important  adjustment must be made.  If one is
interested in the impact of product  charges  or other  similar forms of market
intervention on the recovery of post-consumer secondary materials, one  must
adjust the elasticity estimates.   This  adjustment is necessitated by the fact
that the market models are estimated  for all  scrap  (prompt  industrial  scrap
plus  post-consumer  scrap)  as dictated  by  the manner in which the data are
reported.  If one assumes that  prompt scrap recovery is unaffected by price,
the fact that prompt ferrous scrap comprises  approximately 50% of  total
supply would dictate  that the reported  supply  elasticity  be doubled.   For
wastepaper, approximately  40%  is  prompt,  indicating the wastepaper supply
elasticity should be  increased  to  1.67  times  its reported value.
                                   204

-------
                                 SECTION 6

                     PRICING  AND  SOLID WASTE MANAGEMENT


"The Seattle Solid  Waste Management Experiment:   User Charges  and Separate
Collection of Recyclables"  (Bernard H. Booms).
                                    Z05

-------
               THE SEATTLE  SOLID WASTE MANAGEMENT EXPERIMENT:
             USER CHARGES AND SEPARATE COLLECTION OF RECYCLABLES
                                    by


                             BERNARD H.  BOOMS
 ACKNOWLEDGEMENT

        The author  gratefully acknowledges  the comments and suggestions of
 Dr. Haynes Goddard,  EPA, Cincinnati, Ohio;  and Mr. Ed Steyh,  Manager  of the
 Seattle Solid  Waste  Experiment, City of Seattle.

 THE SERVICE AREA - THE  SETTING

        This  section  provides a general  profile of the  Seattle area and
 describes the  current residential solid waste  management system  operated by
 the  City  of  Seattle.   The  purpose of  this  descriptive material  is to
 familiarize the reader with the setting for the Seattle experiment  and to
 highlight  situational  factors  which  could conceivably  influence the
 development and outcomes of the experiment.

Description of the Area

Population and Geography—
        The City of  Seattle is located in Washington State on  the  shores of
Puget  Sound.   The  current  1978 population is estimated to be 490,000.  The
boundaries of  the city encompass about 88.5 square land miles (91.5 sq. miles
of area  including lakes and waterways).  The  area is bounded on  two sides by
water, with Puget Sound to  the west and Lake Washington to  the  east.   An
aerial view of the land mass reveals an hourglass shape.   (Perhaps because of
her "figure"  Seattle is  called  the Queen  City!)  The city  is thus
geographically divided into north  and south  sections with  the downtown area
at the narrowest  part of the hourglass being the  usually accepted dividing
line  between the  north and  south sections (See Map 1).

Topography—
        Seattle was once a  city  of seven hills,  but  the  marvels of
engineering reduced the  number  to six  hills and one  regarded  bump!
Nevertheless, the general terrain  is  still far from flat with many areas of
rolling hills  and numerous areas  with steep hills.   Obviously, the  hill


                                   206

-------
N
                                                                                                                                                    --

-------
nature  of  the area makes  the  operation of collection vehicles  and the
collection  of solid waste more  difficult than  in  flat areas.  This  is
especially  true during the winter months when the ground  is occasionally snow
or ice covered.

Climate—
        The climate  is  generally  mild.  Summer daytime  temperatures usually
reach the upper 60's  to middle  70's.   Summer days with temperatures over  80
are  unusual.   Winter daytime  temperatures range in the high 30's to 40's.
Below freezing readings occur only  about 15 days in an average year.

        Seattle has  a reputation  for being rainy.  This reputation is based
less on absolute amounts (average 36  inches/year)  and  more on consistency.
The Seattle area is often cloud-covered, and rain, when it falls, occurs at a
light rate  for long periods.

        Winds  are seldom strong.   Normal  wind speeds  are 3 to 7 miles per
hour.   Winds  are rarely  strong  enough to blow debris about during the
collection  process and cause collection difficulties.

Population  and Housing Characteristics—
        Table  1 depicts  population and housing statistics  for Seattle and
offers comparative statistics in some instances  with selected cities over
250,000 population  surveyed  by  R.  L. Polk and Company.  The comparative
statistics  were calculated by Polk  based on  1975 data.

        Population—Seattle has a population mix quite different  in  several
important dimensions from most  other cities  of similar size.   The age
distribution  of Seattle's population  tends to be bimodal to the younger and
older  ends of the scale.  Seattle has become predominately a city of  young,
single, often  highly  educated,  high income persons, and of retired,  older,
lower  income persons.   The degree of  Seattle's  singleness obviously
influences the proportion of  households  with children, with  a higher
proportion being childless  in comparison  to other cities.  It appears that
national trends of urban population composition are exaggerated in  Seattle.
The  high proportion of  young, single or  young married  persons with high
education and  income  levels in Seattle  suggests a great potential  interest in
and willingness to participate in  social experiments.  The attitudes found in
Seattle can  be a strong  determinant in  the  observed outcomes of  the
experiment,  and any attempt  to  transfer the findings from Seattle to other
areas should  be done  with caution.

        Housing—Approximately 65%  of the population live  in single  family
housing units.  There  is  a  trend for more single people to occupy  single
family housing. Solid waste collection for  the 35% of the population  living
in multiple family units  is most  often carried on using large collector
containers  with resultant economy of  collection.  But approximately 5%  of  the
structures  in  the city have 5  or  more housing  units, so  that large  scale
collector  systems are not extensively used.  Despite a high proportion of
single persons in the population,  Seattle continues to retain a  relatively
low population density.


                                   208

-------
TABLE I.  SUMMARY OF POPULATION AND HOUSING DATA IN SEATTLE AND
           OTHER CITIES OVER 250, 000

Other cities over
Seattle 250, 000, 1975 data
Population
% White
% Non-White
% Black
% Other
Households
Average per sons /HH
% One person HH
490,000
85%
15%
9%
6%
203,237
2.4
35% Average 26%
% HH with children



% Retired heads of HH



Housing units

% Single family units

% Multiple family units
     24%



     27%



216.058

     65%

     35%
Range 20% -- 31%

Average 38%

Range 30% -- 40%

Average 20%

Range 16% — 26%
Source: 1977 Polk Data and City of Seattle.
                                   209

-------
 Collection and Disposal

         Solid waste operations for the  City of Seattle are run as  a  utility.
 Utility  status means that the solid waste  operation must "stand on  its own
 bottom," i.e.,  operate without  subsidy  from the City's general fund.  Fees
 are charged for the service and these  fees must cover all costs of  operation.

         For  solid waste management purposes  the city  is  divided  into two
 service  areas.  These areas  coincide  with the north and south geographic
 sections of the city described  earlier.

         The city operates a transfer station in  each service  area.  These
 transfer stations are the  points  where  the two contracted commercial haulers
 deliver  their collections.   The city then transports the solid  waste  from the
 transfer station to the city-operated disposal sites.  The  transfer  stations
 are  also open for residents (and for non-residents at a fee) for individual
 dropoff of solid waste.

         The Solid Waste  Utility controls and manages only residential  solid
 waste.  Disposal  of  commercial  solid  waste is  left to the private sector,
 without  city involvement.   One  major commercial firm dominates  the  commercial
 collection and disposal business.  This firm  operates a marine  transfer
 station  in the downtown  area.  This  station is used  to transfer commercial
 solid waste to a barge. The waste  is then hauled out of the county  area and
 disposed of on a private landfill  site north of Seattle.

        Map 2  shows the boundaries of the service areas and the location of
 the  transfer stations.

        The city has recently passed an ordinance  that could  require  that all
material be disposed of at city-operated landfills and run through the
 city-operated  transfer stations.

        Charges for the  collection and disposal of  residential solid  waste
 are  collected on  a flat fee basis by  the city.  Charges per  single family
household  are  currently $5.20  a month for  an  allowable limit  of  4
 conventional garbage  cans plus  two  ordinance units (boxes or  bags  of
non-putrescible substances) per  can  (total allowable ordinance  units, 8).

Collection—
        The collection service  is operated on  a contract basis by the  city
with the two  service areas being  separate  bidding areas.  Award rules for
service contracts  prohibit any single  contractor  from being awarded the
contract  for both areas.

        Collection  frequency  is  once a week.   Backyard pick-up  is provided.
Though curbside placement  is purely voluntary,  it is estimated that  from 15%
to 20% of the  households do set  their  cans at curbside on  collection  day.
Setting  cans  at curbside is  one way residents  can obtain  special favors from
collection  crews, like accepting occasional oversized or overweight items.
                                   210

-------
                         Map 2
                        NORTH
                        TRANSFER
                        STATION
                 SOUTH
                 TRANSFER
                 STATION
 SERVICE AREA BOUNDARIES:

 '/////////.-HQi\t\\ South

     O 3 -Marine (Some Solid
     Waste from East ol City)
Source: City of Seattle, Departments of Engineering and
         Lighting, Seattle's Solid Waste ... An Untapped
         Resource, May 1974,

                            211

-------
        Approximately 30% of  households are on alleys;  in these areas,
collection is  via the alley.   Alley collection usually involves shorter
distances for  the collection  crews  to walk in picking up  the solid waste.   A
recent  survey  showed that 52% of  the housing units locate their refuse
containers  only in the  backyard,  while the remaining 48% utilized  a
combination of  backyard, curb  or alley locations for  their containers.

        The current  distribution of can usage per household within the City
is estimated to be:

                            Zero cans        1%

                            1 can           20%

                            2 cans          60%

                            3 cans          15%

                            4 cans           4%


A recent survey of container usage and volume of refuse indicated that  the
average  number of units per  stop was 1.9 cans and 0.1 ordinance units.  The
average weight  of material per stop  was 53 Ibs., ranging from 40 to 69 Ibs.

Disposal—
        Two disposal  sites are currently leased and  operated by  the  Solid
Waste  Utility.  The two landfill sites are both outside  of the city limits.
The  close-in  Midway site  is nearly up  to grade and  is limited to
non-putrescible  wastes.   The  Kent-Highlands site  is estimated to have only
three years of  capacity remaining at current rates  of solid  waste generation.
The  city  pays a disposal  royalty  of 65 cents/ton for the use of  the
landfills.

Waste Flow

        Approximately 550,000 tons  of solid waste  are collected each year in
Seattle.   This flow is composed  of approximately 218,200 tons/year of
residential waste and 331,800  tons/year of commercial waste.

        A study of percent composition by weight of  solid waste in the area
that  includes  Seattle estimated the following percentage breakdown by waste
type.
                                   212

-------
     TABLE 2.  PERCENTAGE COMPOSITION BY WEIGHT OF COLLECTED SOLID WASTE
               IN COUNTY AREA THAT INCLUDES SEATTLE
Waste type
Paper
Newspaper
Cardboard
Mixed Paper
1971 data
residential
41%
15%
6%
20%
1973 data
industrial/commercial
22%
0
7%
15%
      Plastic                      5%                     1%

      Ferrous Metals                6%                     3%

      Aluminum                    1%                     1%

      Glass                       7%                    15%

      Garbage (putrescible)          13%                     5%

      Other combustibles            13%                    34%

           Wood                       1%                           19%

           Garden                     10%                            0

           Other                       2%                           15%

      Inerts                      14%                    19%
      Source: Market Analyses of Recovered Materials and Energy from Solid Waste,
            CH2M Hill Study, 1977.
        A follow-up study using  these percentage  estimates  attempted  to
determine  the  proportions of paper, ferrous metal, aluminum and  glass
currently not entering the  collected  solid waste stream and  being recycled
instead.  Table 3 illustrates the  estimated level of  recycling activity prior
to the beginning of the Seattle  experiment and compares these levels  with
"maximum possible" recovery rates  published by EPA.1
1.   Maximum possible effort estimates are based  on use of  all presently
     accepted technologies and  favorable market prices for recyclables.
                                      213

-------
               TABLE 3.  ESTIMATES OF RECYCLED MATERIAL IN SEATTLE AND EPA MAXIMUM EFFORT ESTIMATES
ro
Waste type (tons)
Paper
Newsprint
Cardboard
Mixed paper
Plastic
Ferrous metals
Aluminum
Glass
Garbage
Other combustibles
Wood
Garden
Other
Inerts
Totals
(1) total recycled/
collected material
225,791
53,922
60,268
111,601
14,228
30,973
9,820
88,929
44,956
141,178
65,224
21,820
54,134
93,590
649, 465
(2) recycled (3) collected
material material
63,333 162,458
21,192 32,730
23,950 36,318
18,191 93,410
14,228
7,927 23,046
4,320 5,500
23,885 65,044
44,956
141,178
65,224
21,820
54, 1 34
93,590
99,465 550,000
(4) percentage (5) EPA/s percent
re cycled /total "maximum effort"
28 46
39.3 55
39.7 55
16.3 35

25.5 63
43.9 46
26.8 50






15.3
Source: City of Seattle.

-------
        It  is  clear that Seattle is an active recycling area.   Approximately
15 percent  of  all material is being recycled.  This compares with a  recycling
rate of 7 percent on the national level.

        Twenty-eight  percent of  the  paper,  44 percent of the aluminum, 27
percent of  the glass and 26 percent of the  ferrous metal are being  recycled.
This high recycling  rate may be  explained by  the attitudes  and
characteristics of the population which were  described earlier.   A large
number of  groups and individuals are involved in recycling operations
throughout  the city.  One estimate places the number of recycling operations
(church and school paper  drives, etc.)  around  120.  At least 18  recycling
locations pay  cash for one or more recyclable materials.

        Figures 1 and 2 give an indication of  the monthly and daily averages
of solid waste material flow from the Seattle area.

BACKGROUND

        In  this section the  factors leading  up to the  experiment will be
outlined,  and the  nature of  the  citizen  and governmental  initiative
described.

Factors Leading to the Experiment

Citizen Concerns—-
        During 1976 the City of Seattle renegotiated the refuse  collection
contracts.   As a part of this process, the  City Council held public hearings.
Citizen input  at these hearings revealed that some  individuals were not
satisfied  with a flat fee  charge.  These vocal  individuals  argued that the
flat fee was unequitable because the fee was  fixed regardless  of  the amount
of waste generated (up to four containers).   Thus the driving factor behind
the variable rate portion of the experiment seems  to be a concern  with equity.

Landfill Capacity—
        For a  number of years the City has realized that at current rates of
solid waste generation it will soon exhaust its present  landfill  capacity.
This concern has led to  the study  of a number of  alternative disposal
methods, including using the waste for fuel or for making  ammonia.  Source
separation and recycling  have also been  discussed as ways of extending the
life of the landfill.  Landfill capacity and  imminent  decision  on locating
another landfill or using other technology  were  also  factors leading to the
experiment.

Thrust of the  Experiment

        Several broad areas of concern could  be said to be  factors  behind the
experiment.
                                   215

-------

January
February
March
April
May
June
July
August
September
October
November
December
Total (TPY)
Monthly Avg.
T/Mo.
41,818
41, 150
45,770
48,086
49,908
49,217
48,621
47,549
46,036
45,256
43,057
42,568
549,036
45,753
Average
Av. T/Day
1,349
1,470
1,476
1,603
1,610
1,641
1,568
1, 534
1,535
1,460
1,435
1, 373


        NOTE:  Data include the estimated Marine Transfer
               Facility quantities of 220, 800 T//Yr. , or
               18,400 T./Mo. , assumed constant year-to-
               year and month-to-month.
             50
          o  40
          o
          o
          £  30
          c
          o

          "?  20
             10
          0
          H
50


40


30


20


10


0
                JFMAMJJASOND

        Source: City of Seattle
Figure  10   Five-year annual average of solid waste.
                            216

-------

Out
1
2
3
4
s
6
7
3
9
10
11
12
13
14
IS
16
17
18
19
JO
21
22
23
24
25
26
27
28
29
30
31

D.y
Su
Mo
Tu
W«
Th
Fr
Si
Sv
Mo
Tu
W«
Th
fr
Sa
Su
Mo
Tu
W«
Th
Fr
Si
Su
Mo
Tu
We
Th
Fr
Si
Su
Mo
Tu

NTS • STS
171
1.116
1.094
99O
(.021
951
157
122
1.099
1.120
1.075
1.074
964
279
116
1.122
1.072
1.040
1.003
956
229
212
1.123
1.105
1.04fi
993
947
220
178
I.10S
1.149
STS
Cm
110
132
140
96
102
89
118
133
1O1
106
109
108
83
110
62
89
101
106
77
97
179
112
114
104
92
86
82
123
S3
107
115

•MTF
0
767
767
767
767
767
383
0
767
767
767
767
767
383
0
767
767
767
767
767
383
0
767
767
767
767
767
383
0
767
765

Tot.l
281
2.015
2,00 1
1.853
1.89O
1.807
658
2S4
1567
1.993
1.951
1349
1.8)4
722
178
1.978
1.940
1513
1.847
1,815
791
324
2.004
1.976
1.904
1.846
1.796
726
266
1.979
2.031

X
19
134
133
124
126
121 ,
44
17
131
133
130
130
121
48
12
132
129
128
123
121
53
22
134
132
127
123
120
48
18
132
135
Oiily Avaraga • 1.499 tpd
    * Based on an estimated rate of receipt of 18,400
      tons per month, received at an assumed equal
      daily rate 5-1/2 days per week.

    Source:  City of Seattle.

    Panel A.  Solid waste received  at North (NTS) and
              South Transfer Stations (STS), and the
              Marine Transfer Facility (MTF), in tons
              per day.
 SMTWTFSSMTWTFSSMTWTFSSMTWTFSSMT
                        AUGUST 1976

Source:  City of Seattle.


Panel B.  Seattle solid waste handled daily, August 1976
                        Figure 2


                          217

-------
 Equity—
        It has already been suggested that equity among  households seems  to
 be the most  often stated factor motivating the experiment.  This concern is
 readily understandable by the  average citizen and has  broad appeal.

        Ironically, if the service level of the collection system, i.e.,  the
 frequency and location of pick-up, account for a large  proportion of costs
 and these costs  are not appreciably  influenced by the amount of material
 collected, then a flat fee charge may be more equitable  than a variable
 charge based on the quantity of refuse!   (It should be noted that the  fee
 structure can vary according to service level.  Some cities are currently
 using such a rate structure.)

 Efficiency—
        Concerns relating to the allocation of resources  focus on social
 costs  and benefits being equal to individual costs and  benefits and the  true
 costs  and benefits being borne by  the individual  making  the  resource
 allocation decisions.  In this way  the consumer can  correctly compare  the
 consequences of alternative resource allocation decisions.  The variable  rate
 is one way of trying to ensure that the costs of solid waste generation  are
 borne by the generator, thus facilitating informed decision  making by  the
 consumer in  trading off between more or less waste generation activities.
 However,  little  mention is made about efficiency in discussions of  the
 experiment.

 Implementation--
        Another  area  of concern that  could relate to  the  experiment  is
 implementation questions.   The experiment is viewed, in  part, as a method  for
 identifying implementation  problems  and  for trying alternative implementation
 schemes.  The experiment  is  an opportunity to learn —  on a  trial basis  —
 how to operate  alternative solid  waste systems.   City  government  is
 especially interested in  discovering problems without  having to risk  the
 possibility of a  big failure (see below).  So information potentially useful
 to implementation  concerns  is also a partial basis for the experiment.

 Initiative in .the  Experiment

        It is important  to realize  that the initiative for the experiment
 started with the  citizens and was picked up by the  City  Council.   The
 Council,  not the  City Utility (city  department) initiated the call for  an
 experiment.   The  city department  has  been reactive  to the idea of   an
 experiment,  rather than proactive.

        The environment  in local government may help  to explain why the
 initiative  lies where  it does.  As  a  result, error  is terror for  city
bureaucrats.   The  constituents are close  at hand to complain  if things  go
wrong and  there are often individuals  willing to  turn mistakes  into political
advantage.   Most often  the  facts of  life for bureaucrats in this setting are
 such that  if  you do the same old thing and something goes wrong, it is an  act
 of God,  but,  if you try something  new and It goes wrong,  it is your fault!
Therefore,  organizational resistance.to  experiment and change is increased.


                                   218

-------
        It is important to understand the environment  in which the experiment
is  being  conducted when  considering  the  experiment and  thinking about
suggestions for modification to the experiment.

THE EXPERIMENT

        In this section the objectives and design of the Seattle Solid Waste
Experiment are discussed.  The experiment was designed under contract with a
consulting engineering firm.  The program is  currently being  managed by  the
City's  Solid Waste Utility.   The actual running of the experiment and the
evaluation have been contracted out  by  the  City to  private firms.  These
firms are responsible for the regular collection  of solid waste under a
variable rate structure, a recycling program based on home  collection,  and
evaluation — as well as design — of the experiment.

        The experiment consists essentially of three test  treatments in  the
mangement of solid  waste.  The treatments are:

Variable Can Rate

        In selected  areas of the  City, residents are charged  for refuse
collection and disposal on the basis of the number of  refuse  containers  they
use.  Household participation in the test areas  is on  a voluntary basis.

Source Separation with Separate Truck Collection - Recycling

        In selected areas,  residents have the  opportunity to voluntarily
separate recyclables from their refuse for separate collection on a  schedule
different from the  regular  solid waste pick-up  schedule.  The resident is
asked to separate waste into three recyclable categories:  glass, metals, and
newspaper.

Variable Rate Plus  Source Separation

        In selected  areas, both of  the above described treatments are being
tried simultaneously.  The variable rate applies only  to containers used for
refuse, and not to  containers used for the recyclable materials.  Again,
participation in both treatments is voluntary, and the resident can volunteer
for one or both of  the elements of the treatment.

        An evaluation of  these various treatments is being conducted  as a
part of the Seattle Solid Waste Experiment.

The Objectives of the Experiment—
        There  are  several written statements  outlining the objectives or
goals of the experiment.  The main objectives are as  follows:

        To  test the  feasibility  of a variable can rate structure and
        determine its impact on refuse generation and  collection.

        To  test the  viability  and  economic feasibility of  separate
        collection  of household sorted recyclable materials.

                                    219

-------
         To test the  impact of separate  collection  of  household  sorted
         recyclables  on refuse generation and collection.

         To test the impact of a variable can rate structure on household
         participation  rates  for separate collection  of  sorted recyclables.

         The  research design developed for the City  under contract  was
 prepared to meet these stated objectives.   The research design contains
 little  discussion or  analytical framework  relating  to possible hypothesized
 impacts  (both wanted and unwanted)   of the  various experimental treatments.
 Without  the benefit  of such a model  or analysis  it is difficult to determine
 a methodology for evaluating the experiment and,  in  fact,  to design  the
 experiment  and identify data collection needs.

 The  Research  Design

         The research design was constrained by  several factors.  First,  was
 a concern and a requirement that the experiment be as  self-supporting  as
 possible,  and  that  in no  case should the experiment result in more than
 $250,000 of utility  funds being dedicated to the project, or result in more
 than $55,000 of  revenue loss  to the  utility.  In  addition,  the experiment
 had  to meet all tests of  legality.

         The test  of  legality  raised the most  difficult  complication  in
 research design.  The City Attorney  rendered an opinion that  variable rates
 exceeding existing standard rates would  be illegal under State law requiring
 that all utility  customers, of the  same class be charged at  the same rate.
 Thus, the possibilities for designing a rate  structure that  would reward
 residents  who generate  little waste and penalize  those who  generate large
 amounts  of  waste, while at  the  same time generating the revenue needed  to
 cover the  total costs of the service, were  greatly  limited.  As a result, a
 proposed variable rate that  had  considerable differentiation between numbers
 of containers had to  be abandoned.  The  variable rate is discussed in detail
 below.

 Three Treatments—
         Three treatments are  involved in  the  experiment:  variable rates,
 source separation and .separate  collection of recyclables, and  a combination
 of these two treatments.  Participation  in all cases is  voluntary.  A part of
 the experiment is to  determine  the response  of the residents  to the various
 treatments, given various promotional and educational efforts.

Treatment Areas—
        Thirty  areas in  the  City  were  identified  for inclusion in the
experiment.  In  addition, five control areas similar to  the test  areas were
 designated for inclusion in the study so  as to provide baseline data which
allow for measurement of  seasonal and weather factors  that influence solid
waste collection.

        The thirty  treatment  areas represent regular collection routes or
combinations of regular collection routes.
                                   220

-------
        Each  of the three treatments is tried  in  10 of the 30 test  areas.
Each set  of  10 test areas was picked  so as to be representative of the City
as a whole.   The areas were selected on the basis of income,  housing density,
topography,  and  location  within the City — with income  levels being the
dominant selection factor.   For each  treatment,  there are two low  income
areas,  two high income areas, and six middle income areas.

        The  treatment  or test areas  are  not contiguous,  and  are spread
throughout the  City.

        Test  areas 1 through 10 are the variable rate areas;  areas  11  through
20  receive  a combination  treatment  of variables rates and  recycling, and
areas 21 through 30 are  the recycling only  areas (Maps detailing  these areas
are available from the City of Seattle.).

Variable Rate—
        The  collection  and disposal of solid  waste  is being conducted by the
regular contractors.  The level of service  is  identified by special stickers
on the containers.  Cans  without stickers are  not  collected.  This  system was
chosen over  route books  because it was believed to be more efficient, and
also  because the high  proportion  of alley  pick-up  in the City  limits the
suitability  of the address-based route book system.

        The  variable  rate treatment areas  are  being monitored by  five City
employees hired throught the CETA program.  (The CETA contribution to the
experiment  amounts  to  approximately $75,000.)   The monitors follow the
collection  crews, recording numbers of  containers  collected and the time
required  for the  various components of the collection process.  The monitors
also make  general observations about the collection  process and the
experiment.   The  monitors are  responsible for attaching the  stickers  to the
containers and for making sure the sticker system is kept operational.

        The billing  for the variable  rate areas  is conducted through the
regular City billing  system.

        Early in the design phase  of the experiment there was  much discussion
concerning how to  determine  the "appropriate"  variable  rate structure.
Ideally, the  rates  should reflect the  marginal  costs  of collection and
disposal.   In the  absence of  any data  on marginal  costs (on a container
basis)  the discussion focused  on the requirement to generate revenues  needed
to cover the  cost  of  the experiment and on guesses  as to what rate spread
would provide the  desired incentives to residents.  The following rate
structure was proposed.
                                     221

-------
              Number of Containers                 Rate

                   Zero cans                      $1.00

                     1 can                         4.60

                     2 cans                        5. 60

                     3 cans                        7. 00

                     4 cans                        8. 50

       When  the  City Attorney was asked  to  review  this structure he
suggested that it would be illegal  on the basis  that  several of  the proposed
rates exceeded existing rates for residents outside  of the  test area
receiving similar  service.

       Two alternate  rate structures were considered  (see below),  but  these
were judged inappropriate either because they generated insufficient revenue
or because they  did  not appear to provide a large enough incentive  for
residents to modify their behavior.


                    Alternative Rate Structures Considered

No. of Cans           Rate               No. of Cans              Rate

   Zero              $1.00                  Zero                 $1.00

     1                2.80                   I                     4.00

     2                4.00                   2                     4.95

   3 or 4              5.20                 3 or 4                   5.20


                       Rate Structure Finally Adopted

              No. of Cans                           Rate

                  Zero                             $1.00

                   1                                4.00

                2, 3 or 4                            5.20



       This  rate was judged to minimize  revenue  loss  and to provide  a
"reasonable"  incentive  to switch  from two can  to one can service.   (It waa
                                  222

-------
felt few residents would be willing to move  to  a  zero can service level.)  It
was  noted in the discussion that thir rate  structure has the drawback of not
providing a rate break between the 3 and 4 can  users.  It remains to be  seen
if  the experimental rate  structure provides incentives for the big waste
generators to reduce waste output.

        It  is  important to make clear the fact that the one can service does
not include collection of ordinance units.   What  doesn't fit  in the one  can
is not collected, and the resident must dispose of any extra himself or store
it for collection at the time of the next pick-up. No provision is made  for
occasional increases in the amount of solid waste generated, e.g., from yard
cuttings or special events such as the holiday  seasons.  This  fact may  have
cut  down on the number of  residents willing to adopt the one can service
level.  Note also that in treatment areas where both the variable rate  and
recycling tests are in operation, containers for the recyclable material are
not counted in the variable rate.

        Generally, the  discussion  about what  rate  structure to adopt was
based on guesses, hunches, common sense, with the greatest emphasis being on
revenue needs.  Words like "ideal," "appropriate," and "reasonable" were used
to describe various alternative rates, without the words  themselves  being
given precise meaning.  There is a need for  cost  information on which to base
such rate making decisions.

Source Separation and Separate Collection of Recyclables—
        The sorting and recycling part  of  the experiment is being conducted
under contract by a private recycler — Seattle Recycling, Inc.  The research
design  specified  certain requirements  that the recycling contractor had to
satisfy in his operation of the experiment.  These requirements seem  to be
aimed  at  ensuring  that the  technology and procedures used in the recycling
process would represent efficient equipment and  processes  for large  scale
operations.   In this way  experimental operating experience would be
indicative of operating results that might be obtained in a citywide program.
The aim of the experiment is to simulate a citywide program.  To further this
end —  publicity requirements, processing requirements and requirements  for
the marketing of collected recyclables are specified in the research design.

        To assist the household in the storing  of source separated material,
each  participating household is provided with two cloth mesh bags (potato
bags) for holding metal and glass and one plastic bag for newspapers — at
the time  of each  collection.  The bags are cleaned and recycled back to the
households.

Combination of Variable Rate and Recycling—
        The combination of the two treatments is  carried on in 10 test areas.
The contractors  each conduct their separate portions of the experiment.  No
modifications to the two treatments or special  arrangements  are used  in the
combination test areas.
                                     223

-------
Current Evaluation Design—
        Considerable  mention is made in the research  design of the data to be
collected.  The data  monitoring is  summarized in Figure  3,  taken directly
from the Final Research Design.

        The current evaluation effort appears to be  heavily  focused  on  the
"time and motion" aspects of the collection and recycling  process.

Experiment Modification—

        The managers of the experiment have indicated  from the conception of
the  experiment  that the  program is  and  will continue  to be  open  to
modification.   Thus, the possibility exists to make additions or changes in
both the research design and the evaluation procedures.   As early results
become available, modifications will begin to be considered.

SOME EARLY RESULTS AND OBSERVATIONS

        Some results  are available based on the first month  of operation of
the Seattle experiment.

Response and Participation Rates

        Figure  4 and Tables 4 through 7 summarize the response rates and
actual participation  rates for the various treatments and  test areas.

Variable Rate—

        The participation in the variable (one can  rate) rate test averaged
12%.  There was no difference in the participation rate between those areas
wit and without separate collection of recyclables.   This  finding is contrary
to the expected outcome.  One might assume that collection  of recyclables
would make movement to  the  one can  level of service  feasible for most
households.

        The 11% participation  rate  was  obtained  using only a mail
solicitation.   In addition,  media coverage described  the variable rate
program.

        Approximately 30 households chose the zero rate.   The households  have
been  given composting information and are being monitored by the city  health
department.

        Income  levels did not seem to systematically influence the observed
participation rates among the various test areas.

Recycling1—
        A total  of 58%  of the households responded  favorably  to  being
involved  in the  recycling experiment.  A higher proportion of households in
the combination treatment areas (61%) expressed a willingness to be involved
than in the recycling-only treatment areas (55%).


                                    224

-------
              Variable Rate Program

              Monitoring In the variable rate area! (and control areai a< well) will Include recording of the following Informa-
tion Lolh prior to implementation and once every Iwo months thereafter:

              •      Frequency distribution! for the number of refule container! and ordinance units per atop:

              •      Average quantity of refute per atop;

              e      Frequency Dlmrlbutlon for the collection time per ttop;

              e      Average Travel time between stops;

              •      Total stops serviced per day;

              •      Total ton>  collected per crtw per day;

              •      Total lime to service pilot and control areas;

              .      MUcellanemn travel limes and distancea (yard and route,  route to landfill, landfill to yard, etc.):

              •      Number of trip* lo landfill;

              •      Unavoidable delay! and breakdown!;

              •      Scheduled break*.

              Source Separation and Separate Truck Collection

              Monitoring In the pilot area! performing source ieparation of reeyclablea la more exteniive than in the variable
 rale only areas.

              Prior to implementation, monitoring'tasks are identical to those summarised In the laat section on the variable
 can  rate teat.  Once lource aeparatlon  beglne. It will be necetaary to monitor not only ongoing reCuie collection activities but
 alao the source separate  collection ae well.

               Below U a Hit of the data required for (valuation of the tource separate collection of glass, metal,  and newa-
 pnper.

               •      Addresses of homea participating by aeltlng out recyclable! (percent participation);

               e      Identification of Tecyclalilea let our (newspaper, glass, metal);

               e      Average quantity of  reeyclablea by type per atop;

               e      Frequency distribution of time per atop for recyclable* collection;

               •      Average travel time between atopt:

               a      Total time for separate truck to  service pilot area;

               •     Miscellaneous; time*  and distances (yard to route, rout* to yeard, etc.);

               •     Unavoidable delays and breakdowns;

               •     Scheduled breaka.

               With this data, the source separate collection  method can be evaluated for performance and cost.  A sample form
 for collection of this data follows.   Naturally, the control areaa will not be involved nor monitored far source separation.  ,

               At the yard area, monitoring will again  be performed to Identify the manpower required and time associated with
 unloading, processing and storage of the (lass, metals, and newspaper collected.   This data will allow cost trade-offs to be
 nude,  and other sorting/storage techniques to be evaluated.

               Monitoring will be relatively simple consisting basically of recording times and labor  requirements for the various
 processing and sorting operations previously described.  Typical activities lo b* timed Include]

             ,  e     • Unhook trailer;

               a     Paper unload from pick-up to roll-off bin;

               a     Position and dump metal  bin, overaee convey, magnet, crushing system;

               •      Position,  unload glass bin.

               «      Sort glass.

               Variable  Can Bats Plus  Source Separation

               Tlie monitoring requirements for the combined variable can and source separation pilot  program are virtually
 Identical lo those for source separation above.  Afstn, pilot areas and control areas will be  selected.
 Source: Final Research Desljm  Seattle Variable Rale Source Separation Pilot Programs, City of Seattle. 1978.




                                     Figure 3.   Monitoring Design


                                                           225

-------
rtKltHI 1UU
90
80
KEY:
oorbell 70
hone
rn

ar. Rate 50
Irst
ol lect ion nn

30

20
10


TEST AREAS

TREATMENT
DATA
Doorbell

Mai 1
Variable rate
Total HHH
SE£
1st collection













.


1














••••


2
H














•§•

3














•M


4














_


5














••M


6
H
















7
L












_



o
0

















9
L
















10

VARIABLE RATE ONLY

.

83
786
M



.

62
H


.

48
328
M


.

55
469
H




71
539
M


.

45
451
H


.

30
361
L



.

17
203
M

_

73
522
L



„

65
457
M










%
m
































K









WM









Wk









•,;-'-'::










IP









"•'•.:*.









•


















mfc^?
p^^^p^








P^
m









:::::










^*-
m
w*



f -••••^jiiiiii •;•;•: f*"*1 •'"•'- U™j
! iW-x^tiXWfcxWvlxffiwliiiii.J '

11
II

12

VAR
101
160

66
448
H
1S4
118
l?2

61
643
M
184
x.WS
13

^-i-:'
M


15
H
16


17
L












^











iff











wn

' :











^















•//"//

•:•:-:•:•:-:

18



19

'"""> :;:$: SxvfcSJSr" ''• 1 K^*S?%

20
L
ABLE RATE AND RECYCLING
81
ns
1 71
64
480
M
170
175
121
1 22
71
713
H
179
95
S4

445
H
142
77
62
46
54
291
M
63
92
85
72
68
475
L
67
'!'
76
23
327
M
98
138
112
82
70
H
133
146
2(|
47
409
L
90

21


22
H

23
H




25


26
L
RECYCLING
122
132
4R
^36
M
146
81
64
399
H
SI
94
40
346
H
122
109
20
423
M
137 il34 124J123
109
168
92
658
H
220
iSI
100
14
537
L

27



28

mm

29

mst

30
L
ni.'LY
119
127
19
595
H
75 125
135
136
31
546
M
L56
125
122
1\
564

75
114
85
17
522
L
63
Source:  Seattle Recycling, Inc.  and City of Seattle.
                      Figure 4.  Percent response Seattle experiment.

-------
                    TABLE 4. PARTICIPATION SUMMARY VARIABLE RATE TREATMENT
Area*
1 - 10
11 -20
Totml
Treatment
Variable only
Combination


543
572
1115
#HH total on map
4605
4827
9432
%
12%
12%
12%

Source:  City of Seattle
                        TABLE 5.  RESPONSE SUMMARY RECYCLING TREATMENT
Area*
11 - 20
21 - 30
Total
Treatment
Combo
Recycle only

Mail
768
368
1136
Phone
1030
1164
2194
Doorbell
1138
1254
2392
Total f HH
responding
2936
2786
5722
Total # HH
on mapt
4827
5076
9903
%
response
61%
. 55%
58%
Source: Seattle Recycling, Inc.
              TABLE 6..  FOLLOW-THROUGH SUMMARY (1ST COLLECTION) RECYCLING TREATMENT
Area*
11 - 10
21 - iO
Total
Treatment
Combo
Recycle only

Total HH
collected
1280
1272
2552
Total # HH
responding
2936
2786
5722
%
follow-through
44%
46%
45%
 Source:  Seattle Recycling, Inc.
           TABLE 7.  ACTUAL PARTICIPATION SUMMARY (1ST COLLECTION) RECYCLING TREATMENT
Areas
11 - 20
21 - 30
Total
Treatment
Combo
Recycle only

Total HH
collected
1280
1272
2552
Total I HH
on mapa
4827
5076
9903
%
participation
27%
Z5%
26%
 Source: Seattle Recycling, Inc.
                                                  227

-------
        The follow-through percentage,  i.e.,  the proportion o±  those
households who said  they  would participate — and actually did,  overall  was
45%.   In other words,  over half  of  the households who  said  they  would
participate did not put  out recyclable materials during the first  collection
period.   Time will tell if this "drop out" rate reflects actual  rejection of
the program or just forgetfulness in putting out materials for collection.

        The recycle-only  treatment areas had a higher follow-through rate
(46%) than the combination  treatment areas (44%), thus offsetting  the greater
initial response rates for  the  combination treatment areas.

        Actual participation rates  for all recycling areas  averaged out  to
26%.   The combination  treatment  areas averaged  a 27% participation rate,
whereas the recycle-only treatment  areas had a 25% response  rate.  How much
of  the 2% point difference can be  attributed to incentives created by the
existence of a one  can rate is  difficult to judge.  In any case,  2% points do
not seem like  a very large  difference between the two treatments.

        In the case of recycling treatments, the  income levels  in the  test
areas  did seem to  have  a systematic influence on participation.   High income
areas tended to have higher participation  rates and lower income areas  lower
rates.   (This finding parallels results from previous studies.)   The highest
participation rate  for an area  was  46% (test area 22, a high income area);
the lowest rate was 12%  (test area  30, a low income area).

Material Collected  and Revenues

Trash—
        Information on the  amount of non-recycled material  is being recorded,
but is not yet analyzed.

Recycled Material—
        The data  on  the amount of recycled material collected are presented
in Table 8.

Revenues--
        Information on materials collected, market prices  for the materials
and the resultant revenues  are  given in  Table 9.

Other Data

        Time  and motion type  information and  costs data are being recorded,
but these statistics are not yet available.

Some Observations

        Several implementation difficulties have been encountered during the
early operations of the  experiment.
                                    228

-------
  TABLE 8.  AMOUNT OF RECYCLABLE MATERIAL COLLECTED (1ST COLLECTION) IN POUNDS
            (June 19. 1978 -- July 14, 1978)
toute
'll
12
13
14
15
16
17
18
19
20
'21
22
23
24
25
26
27
28
29
,30
Metal
677
885
861
929
600
353
255
250
640
374
623
805
458
567
1,521
310
620
585
260
280
Giles
3,494
4,065
3,042
3,409
3.359
1,200
1,345
1,650
2,480
1,889
2,310
3,816
2,880
1,826
5,367
1,250
1,760
3,060
1,422
1,325
Newspaper
4,503
3,864
4, 139
3,926
3,332
1,068
581
2,380
2,010
2.877
3, 100
6,799
4,180
3,448
6, 289
2,036
4,391
3.610
1,950
1,006
Total
8,674
8,804
8,042
8,264
6,291
2,621
2. 181
4,280
5,130
5,140
6,033
11,420
7,518
5,841
13,177
3,596
6.771
7,255
3,632
2,611
Percent
participation Average
34%
29%
35%
25%
32%
22%
14%
30%
22%
22%
28%
46%
36%
29%
33%
14%
21%
28%
13%
10%





1







N
£





Source:  Seattle Recycling, Inc.
     TABLE 9.  AMOUNT OF MATERIAL COLLECTED, MARKET PRICES AND REVENUES
                (1ST COLLECTION) (June 19,  1978  -- July 14, 1978)
Material
Paper
Glass (cullet)
Tin/steel
Aluminum
Beer bottles

Amount Collected in Tons Unless Indicated
Market Amount collected
$36. 00 /ton
20. 00/ton
60. 00/long ton
.22/lb.
, 60/caae

33
21
4 long tons
2331 Ibs.
496 cases
Total
Revenues
$ 1,188
420
240
513
298
$ 2,651
     Source:  Seattle Recycling, Inc.
                                            229

-------
Variable Rate—
        The main problem encountered with implementing  the variable rate  has
been the use  of  stickers for differentiating  service  levels.  Trouble first
arose when residents objected to the City monitors placing  the stickers on
the  cans.

        Several phone calls were received from  residents reporting strangers
walking around their yards and garbage  cans.   Some householders have removed
the  stickers.

        In addition, the different  colored  stickers have faded so that it is
difficult to tell the colors apart.   The problem is further complicated by
the  fact that the stickers do not adhere well to some cans.  Some effort will
have to be made to improve sticker design and technology.

Recycling—
        The recycled materials  are  collected  at  curbside.  A significant
number of scavenging incidents have  been reported.  Efforts  are being made to
record the extent of this theft problem.

CONCLUSION

        The Seattle solid waste management experiment involves the test .of a
variable  user charge for residential refuse service and separate collection
of household sorted recyclables.   The Seattle program  is probably the most
extensive residential refuse  experiment  conducted  to date in the United
States.   It is hoped that the  experiment will shed light  on the economic
feasibility of, and the impact of these  proposed innovations, and at the same
time provide insights into important  implementation questions related  to  the
possible adoption of these suggested  innovations.

        The Seattle experiment deserves  close watching by all those concerned
with the  study or  management of urban  refuse  systems.  The results from the
experiment presented here are  only preliminary.  As of this writing  the
experiment is still on-going.
                                    230

-------
                                  SECTION 7

                             RESEARCH DIRECTIONS


"Research Activities in EPA" (Oscar W. Albrecht).

"Solid  Waste Collection/Disposal Economic Research Strategies"  (G.S.  Tolley,
V.S. Hastings).
                                     231

-------
                         RESEARCH ACTIVITIES  IN EPA




                                     by


                              OSCAR W. ALBRECHT
 INTRODUCTION

         I have the privilege of discussing an area  that is probably  less
 controversial  than some of the  areas we've been discussing up to now.   The
 papers have all been interesting and have generated some  lively discussion.
 My assignment  is to discuss the  research programs of  the U.S. Environmental
 Protection Agency (EPA).   I  will briefly mention the research and activities
 of the major offices,  and  then discuss in a little more detail the research
 conducted by the Office  of Research and Development, and  particularly at  the
 Cincinnati Laboratory which  is  sponsoring this workshop.

 RESEARCH BY OFFICES,  EXCLUDING  THE OFFICE OF RESEARCH AND DEVELOPMENT

        The U.S.  Environmental Protection Agency  (EPA) was established as an
 independent agency  in  the  executive  branch of  the  federal  government,
 effective  December  2,  1970.   Many  of  EPA's  activities  are
 enforcement-oriented,  but  the Agency is  involved in  numerous research
 activities throughout  the  various offices,  divisions, and branches of the
 organization.

        The EPA  is directed by an Administrator, presently Douglas M.  Costle,
 and reporting  to  him  are various offices headed by assistant administrators.
 Within  these offices  are  various divisions and branches which conduct
 research; for  example, the Office of Planning and  Management has an economic
 analysis division which  evaluates impacts of Agency programs  and policies.
 These  impacts include  potential plant  closings, effects  on  employment,
 prices,  and Gross National Product (GNP).   Estimates of abatement costs  and
 cost/benefit analyses are made for selected  industries, principally those
 pertaining  to air  and  water regulations.   The Office of Planning  and
Management (0PM) frequently performs the Impact analyses when  program  offices
 do not have  sufficient capability or  resources  to conduct  such analyses on
 their own.   The 0PM also performs the  impact  analyses on selected industries
where more  than one regulation is involved.   The investigations by  the Office
of Planning  and Management do not preclude similar research by  the  various
program offices within their  own areas of responsibility.


                                    232

-------
        The  Office of Water and Hazardous  Material performs economic  analyses
and prepares  inflationary impact statements  to support regulations developed
by the Office  of Water Planning and Standards.   The Office  also  performs
cost/benefit  analyses, develops and analyzes alternative control  options and
prepares issue papers describing the costs, risks, benefits, legal, and  other
aspects of alternative control options.

        The  Office of  Solid Waste  (OSW)  within the Office  of  Water and
Hazardous Materials has three divisions, each doing research related  to  their
own particular areas of concern.  The three divisions in  the  OSW are:  the
Hazardous Waste Management Division (HWMD), the Systems Management  Division
(SMD), and  the Resource Recovery Division  (RRD).  The HWMD conducts  economic
and policy  analyses related to the handling and disposal of  industrial and
hazardous wastes.   It is  at present involved in developing criteria and
procedures  for implementing the regulations  mandated under Subtitle  C of the
Resource Conservation and  Recovery  Act (RCRA) of 1976.   The HWMD is also
conducting economic impact  analyses of the proposed  hazardous  waste
management regulations for selected industries.

        The  SMD in OSW has responsibility  for a national program to improve
the collection, processing,  storing, and disposal of municipal and  other
nonhazardous   solid waste.   Its major  areas of responsibility are  providing
technical assistance to cities in improving the efficiency  and safety  of
solid  waste management, developing a better  understanding of the  solid waste
disposal to  land practices,  and demonstrating system management plans.

        The  RRD within OSW performs  studies  related  to the  recovery of energy
and materials  from  solid waste.  It  is also  responsible  for  implementation  of
a national  program  to reduce the generation of solid waste.  A major focus  of
RRD has been on large scale  resource recovery demonstration projects. The
Division has   also studied  the use of  incentives to  stimulate demand  for
seconday materials and waste reduction.   Other economic studies by RRD
include estimates of  consumer demand  elasticities for  selected products,
product charges for solid waste, study of the demand  and  supply for secondary
fiber  in the   U.S. paper  board industry,  and the economics of waste oil
recovery.

        Other  offices  which conduct economic research  include the Office  of
Air and Water  Management which  has an Environmental Analysis  Division; and
the  Office of Toxic  Substances which  has a Benefits and  Field  Studies
Division. Without  giving  further detail on these, in the remaining  time I
would like  to  discuss research  by the Office of Research and Development.

RESEARCH CONDUCTED  BY OFFICE OF RESEARCH AND DEVELOPMENT

        The Office of  Research and Development  (ORD) is responsible for the
national research program  for control of all forms of  pollution.   The  ORD's
funtions are  Agency-wide  in satisfying the research and development needs of
the Agency's  operating  programs,  and  it is responsible for  conduct  of
integrated  research within the Agency.  Within ORD are four main offices;
                                     233

-------
 these are:  the Office of Monitoring and  Technical Support;  the Office of
 Energy,  Minerals, and Industry; the Office of  Air, Land, and Water Use;  and
 the Office of Health and Ecological Effects.

         The Office of Monitoring and Technical Support, as the name suggests,
 is engaged primarily in activities related  to monitoring  and enforcement.
 The Office of Energy, Minerals, and  Industry (EMI) has responsibility  for
 assessment of environmental and socio-economic Impacts resulting from energy
 and mineral resource extraction, processing, and  utilization.

         The Office of Air,  Land, and Water Use (ALWU)  is  responsible for
 assessment of the environmental and  socio-economic impacts of land,  water,
 and air  pollution control and management activities.  This office is also  the
 focal point within ORD for  providing liaison  with the Office of Water  and
 Hazardous Materials and the Office of Air and Waste Management (which are
 outside  the ORD organization).  Within the Office  of  ALWU is a Waste
 Management Division which has  responsibility  for  assessment of  the
 environmental  and socio-economic impacts of  methods to control and manage  the
 discharge of  wastes from municipal, recreational  and other domestic sources.
 There is  also the  Agriculture and Non-Point  Sources  Management Division
 within the ALWU, which assesses the environmental and socio-economic impacts
 of alternative control methods for pollution from agriculture and other
 non-point sources.

         The ORD organization also includes the research laboratories at seven
 locations over  the country.   These are  at  Cincinnati, Ohio;  Research Triangle
 Park,  North  Carolina;  Athens,  Georgia;  Corvallis, Oregon; Duluth,  Minnesota;
 Naragansett,  Rhode Island; and  Gulf Breeze,  Florida.

         Before discussing  the research  at Cincinnati  laboratory, let me
 briefly mention the  other  two offices within  the ORD.  The Office  of Health
 and Ecological Effects (OHEE)  has a criteria development and special studies
 division  which is developing economic models and  methodologies for  conducting
 socio-economic  assessments relating to the benefits of  pollution  control on
 an  individual and.broad program basis.   The  Office of Toxic  Substances  has a
 benefits  and  field  studies division which conducts benefit-cost analyses on
 the impacts  of  cancellations  or suspensions  and  the  emergency use  of
 pesticides.

 RESEARCH CONDUCTED BY MERL,  CINCINNATI

        Research activities at Cincinnati, by the Municipal Environmental
 Research Laboratory  (MERL),  are  directed at  three principal  areas of concern:
 solid and hazardous waste, wastewater treatment,  and  water supply.

        The Solid and Hazardous  Waste Research Division  (SHWRD) within  the
 MERL, which is  sponsoring  this  workshop,  has  to  an  increasing degree been
directing its research  to  meet  the program needs  of  the OSW  which  is
 responding to the RCRA  legislation.   In our disposal  branch,  considerable
effort is being expended on  waste  characterization  in order  to learn more
 about pollutant transport,  including movement through  soils  by leachates and
gases,  in order to develop more  effective treatment  and control  methods.  The

                                     234

-------
problems associated with co-disposal of municipal, industrial and  hazardous
waste  are  also areas Investigated and economic and environmental  effects of
the disposal alternatives are being analyzed.

        Waste characterization is also an  area of concern for the  processing
branch because of the  implications for  resource recovery systems.   Our
current projects in the  area of resource  recovery include research on
technologies associated with waste-as-f uels,  preprocessing techniques for
recovery of various fractions from the solid  waste stream, and  the design of
resource recovery  systems.  Some of our research characterized as economic is
more appropriately described as operations  research, but it is mentioned here
mainly because it  does not involve technological hardware.

        The  first few years after  formation of the  EPA,  several  research
projects involved  studies on the collection,  storing, and disposal  activities
of municipal solid waste.  The feasibility of wet systems for collection and
transport of residential refuse was investigated,  as well as the  potential
for wage incentives to  increase the efficiency of solid waste collection and
disposal.  Institutional arrangements for regional and  solid waste  management
were   studied, and  sample  and test procedures developed for  materials
recovered from municipal solid waste.  Management systems for  hospital and
institutional solid  waste were examined in detail.  A survey  was made of
housewives' knowledge  and attitudes towards  solid  waste and  resource
recovery;  there  was  also  a study on citizens' attitudes toward  disposal of
hazardous waste.   The  manner in which federal government procurement policies
and practices affected solid waste recycling was also studied.

        On the issue  of  externalities, we funded a  study to measure the
external effects of  solid waste management. This  study, by the Institute of
Policy Analysis (IPA), used regression analysis  to determine the  effects on
property values.  Using  several  landfill operations in Los Angeles County
(California), it was found  that proximity to and view of a sanitary landfill
did not significantly  affect surrounding  property values.  This  study pointed
out, however, that additional work should be done on the  specification and
exercising of the model as  there was  inadequate  information on the
relationship between property values and  the non-environmental variables.

        In the recent  past, there was considerable concern over the abandoned
automobile problem.   We suported  a study to determine  the  extent of the
problem and  to evaluate strategies  for  dealing with  it.  One suggested
strategy was a mandatory deposit  fee, refundable  upon certification of proper
disposal.    A case  study approach  to analyzing  the  salvage  industry was also
conducted.  This  study showed  that many.salvage  firms are  family  owned, and
that  salvage firms tended to vary widely in the  extent of their operations
and capital investments.  We also  supported a study of  the beverage container
problem by  Research  Triangle Institute.  The results suggested that
government intervention could not  be justified solely  on the  basis of  a
misallocation of resource  inputs.   And that  to justify  government
intervention  solely on the  basis  of  energy savings would  ignore the  utility
that consumers derive  from  the  convenience of nonrefillable  containers.
                                     235

-------
        An important component of the solid waste  stream  is  the  combustible
 portion.   This component has been of increasing  interest recently because  of
 the concern over diminishing energy sources and  high consumption.  A study  by
 the  International  Research and Technology for EPA estimated the  magnitude  of
 combustible solid waste; in a subsequent study they developed an  input-output
 model  for predicting  the  quantity  and  composition of the municipal  solid
 waste. Additional efforts  are currently  underway to update  the estimates
 using EPA's SEAS model  and forecasts of economic  expenditure patterns.

        Wastepaper  is the most important component in municipal  solid waste
 in that it comprises about 35 percent of the total residential and commercial
 solid waste.   Several EPA sponsored studies have  investigated this component.
 A  study by Resource Planning Institute,  using  a pro forma income analysis
 approach, investigated  the  potential role of incentives  for increasing
 recycling.  It  suggested  that high  subsidies would be needed to obtain a
 significant increase in recyclng of wastepaper.

        Another area  of concern  has  been  the  issue or  allegation  of
 descriminaton in regulated transportation  rates, particularly  rail freight,
 and  its affect  on  the recycling of secondary materials.  Two studies funded
 by the EPA arrived  at somewhat different conclusions.  The Resource Planning
 Institute  study,  funded  by  our  office, concluded  that on   a
 chemical-equivalent  basis ferrous scrap moved at a lower rate than the virgin
 material  while  the  Moshman Associates study sponsored by the OSW arrived  at
 somewhat opposite results.   These differences resulted  from  variations  in
 the  manner  in  which the  data  were used, and varying  opinions on the
 appropriate chemical  formulae and points of substitution between secondary
 and competing virgin materials.  A shortcoming of  these studies,  particularly
 the Moshman study, was  the short-run  time  perspective in which the issues
 were addressed.  At present,  we are  supporting work at the University  of
 Cincinnati on demand  modeling  which we hope will ultimately shed further
 light on this issue.

        Another area of concern is the discarded passenger  tire disposal
 problem.   Although  some discarded tires are recycled as door mats,  highway
 crash barriers, artificial reefs, and the  like,  these uses constitute only  a
 small solution to the overall scrap tire disposal  problem.  One recent study,
 initiated  by  our  office, investigated  the  net benefits derived from using
 ground  scrap rubber from  tires as  an additive  to asphalt roadways.  These
 potential  benefits are currently undergoing further study with actual field
 tests on  several highways.    Another study just being implemented  will
 investigate  the state-of-the-art, trends, and impediments  to various
 utilizations  of the  discarded tires.

        Resource  conservation  and recycling have been and continue to be
 proclaimed as desirable goals which this  Nation ought to pursue.  A study  by
 the Environmental Law Institute addressed  the question as to whether market
 prices  determine resource  conservation  and recycling in a socially optimal
 fashion.  While the  study provided no quantitative estimates of the benefits
 of conservation, it pointed out that numerous external forces act to create
divergencies  between  the socially optimal  prices and the prevailing prices  in
 the  market  for natural resources.  These forces,  which act in varying

                                    236

-------
directions and with varying intensities over time,  are difficult to  measure.
Moreover,  from  an intergenerational viewpoint,  a  method for determining the
socially optimal prices has yet  to be devised.

        One of  the external  forces  affecting recycling  is taxes.   The
Institute found  that elimination of  subsidies to extractors through the
federal  tax  code could cause  an upward shift  of at least 1 percent in the
virgin material supply curve for paper; for steel,  2 percent;  and  copper 5
percent.   If  taxes were  equalized  between secondary and virgin materials,
considering the supply and demand elasticities,  the percentage  increases in
recycling  were  estimated  as 0.63  percent for  wastepaper; 0.42 percent for
scrap steel; 0.35 percent for copper; and 1.00 percent for aluminum.

        .The Environmental Law Institute also conducted  a  study on the
feasibility of  using scrap futures  markets for  encouraging the use  of
secondary  materials.  They concluded that futures trading for ferrous  scrap
and wastepaper was technically feasible,  and could result  in  net  gains to
society by reducing uncertainty  and  stabilizing prices  and incomes of
suppliers and users of secondary materials.  Successful  operation of such
markets  however,  would depend upon  adequate participation by the private
sector.  A participating role by the  government was argued  because of the
public good aspect of the information generated  by  the market.

CURRENT RESEARCH AT MERL (SHWRD), CINCINNATI, OHIO

        In addition to the projects just mentioned, several new projects have
been  initiated  recently  in  response  to  provisions of  the  Resource
Conservation  and  Recovery  Act.   The National Academy of Science is
establishing for us priorities  for government supported research in   the area
of resource  recovery.  This study will establish  priorities for recovery of
both materials and energy from  solid  waste, taking into  consideration the
lack  of  incentives in the private sector to do the needed research, and the
net benefits expected to be gained by society.

        Another  study  will investigate the impediments to successful
operation of resource recovery facilities.  Substantial research efforts in
the past have been directed at the technological "know-how" of  facilities and
techniques for recovering materials from municipal  solid waste, but yet the
overall  recycling rate achieved to date is disappointedly low.  This  study
will investigate in particular  how  initial plant  design  and  expectations
compared  with the actual  market  experiences once the  facilities became
operational.

        In response to  the recent  legislation  (RCRA),  we have  initiated
several new studies dealing with hazardous waste.   One study will determine
the cost-effectiveness of various treatment and  disposal alternatives and the
relevance of the various options to  management of hazardous  waste in the
municipal  waste  streams.   Another  study will evaluate  existing federal
programs for risk management, with particular emphasis on- the methodologies
used  by different  government agencies  in assessing  risk,  and  their
applicability to risk assessment for management of  hazardous waste.


                                    237

-------
        We  are continuing  to  have an interest  in  issues related  to  efficient
management  of  solid  waste.   Several  early studies investigated  the
theoretical basis for pricing solid waste management and you have just  heard
the report on the  five case studies  of  user charges conducted for us by
Mathtech (the Contractor for  this study).  Several other studies related to
pricing  have been  mentioned.  A list of the past research on pricing and
related  research was included in the  bibliography mailed  to each of  the
participants.   The substance of these  reports generally emphasize  the
difficulties encountered in acquiring adequate and meaningful data  and
indicate the need for further research before  any conclusions can be reached.
We are,  of  course, interested in your recommendations for further research in
this area.

CONCLUSION

        I hope  this discussion has given  you a general  overview of  the
research activities by the EPA.  You have,  no  doubt, noted that  my  discussion
was mainly on economic research — the  list of research  projects  in the
technological area would be much longer. As  stated earlier, the EPA's  total
research effort is not confined to the ORD,  but is spread out  among various
offices of the  Agency.   Since  the  ORD  is  separate from the
enforcement-oriented program  offices,  perceptions of  research  needs
frequently differ  as the research laboratories are inclined  towards  a
longer-run view of research programs.  Maintaining an effective degree of
continuity  in these research  projects is sometimes difficult.

        Another  problem  is  the coordination of research within the Agency.
Up to now,  the research efforts in the area of solid waste have  been minimal
when compared with those for air and water  pollution.  I think we all agree
that the issues on solid waste we've been  discussing are important  and
research support and funding are needed if they  are to be resolved.  You can
help in  this effort by articulating  these  needs whenever you  have  the
opportunity to do so.
                                   238

-------
                               REFERENCES
1.    Richard Schmalensee,  et al.  Institute for Policy Analysis.  Measuring
     External Effects of  Solid Waste Management.  EPA-600/5-75-011, March
     1975.

2.    Booz-Allen Applied Research.  An Analysis of the Abandoned  Automobile
     Problem. NTIS PB-221879,  1973.

3.    Taylor H. Bingham  and  Paul E. Mulligan.   Research Triangle Institute.
     The  Beverage Container Problem:   Analysis and Recommendations.  NTIS
     PB-213341, 1972.

4.    Frederick A.  Cardin.  Resource Planning  Institute.  Secondary  Fiber
     Recovery  Incentive Analysis.  NTIS PB-241-082,  October, 1974.

5.    John F. Foran, et al.  Raw Materials Transportation Costs and Their
     Influence on the  Use  of Wastepaper and  Scrap  Iron and  Steel.
     NTISPB-229816, April  1974: David G. Abraham, et  al.  Transportation
     Rates and Costs  for Selected Virgin and  Secondary Materials.   NTIS
     PB-233871, 1974.

6.    Haynes C. Goddard.   An Economic  Evaluation of  Technical Systems  for
     Scrap Tire Recycling.  NTIS PB-249197/5WP.

7.    Robert  C.  Anderson.   Evaluation of  Economic Benefits of  Resource
     Conservation.  (In Print).

8.   Barbara  J.  Stevens.  Impact  of the Federal Tax  Code on  Resource
     Recovery: A Condensation.  EPA-600/8-77-012,  August 1977.

9.   Robert C. Anderson and Roger C. Dower.   Environmental Law  Institute.  An
     Analysis of  Scrap Futures Markets for Stimulating Resource Recovery (In
     Print).
                                    239

-------
        SOLID WASTE COLLECTION/DISPOSAL ECONOMIC RESEARCH  STRATEGIES
                                    by
                               G. S. TOLLEY
                              V. S. HASTINGS
      A question of major  concern in municipal solid waste management  deals
with  the rising costs  of solid waste  collection/disposal.  Can there be a
payoff in holding down these costs in ways that take  into account economic
considerations  of  more than a strictly engineering or  technical nature in
providing solid waste  management services?   The rising  costs are largely
associated with rising input costs,  particularly  labor, the real cost of
which continues to  rise, and rising waste volumes.   Concerns associated  with
rising  volumes include  environmental degradation  and  increased use of scarce
land resources.  These concerns are all worth economic investigation if costs
can be  significantly reduced by economic approaches  that would increase net
social benefits from solid waste collection/disposal.   This paper addresses
the subject of  research  strategies  related to  the  use  of incremental user
charges  as an economic approach for  this  purpose.   Cost-side and  then
demand-side  strategies are discussed, and finally a list of research
questions that waste managers might ask is offered.

COST-SIDE STRATEGIES

        It is a basic economic proposition that  net  social benefits are
maximized when services  are charged for at the marginal  cost of providing the
services.  Under existing methods of financing solid waste  services,  usually
flat  user charges  or  general fund financing,  there  are  no  charges related to
the amount or to the marginal cost of  collection/disposal.   This tends  to
lead  to  overuse of services in terms of quantity, to  the point where costs
exceed benefits.  However,  this does  not mean that  all  charges for  solid
waste  co11ection/disposal  services   should  be based  on  amounts
collected/disposed.  Some  services provided are not  waste  quantity related.
Charges  for these services should be based on their  marginal costs.

        This point  is worth emphasizing not only because it can  make  a
considerable difference, but because there has been confusion on this point.
For example, Ernst  (1975) assumed that a choice must be  made between  either a
flat  charge or a quantity  charge.   He related all costs  to quantity.  This
led him  to attribute certain empirical findings to scale factors related  to


                                    240

-------
population size, whereas in fact more plausible  and reasonable  explanations
were possible  based on true cost relationships.  Ernst's cost relationships,
for example,  would  result in no costs for traveling a route  or  going to the
back door to look when no wastes were collected.  Costs are associated  with
these  actions  regardless  of  quantities  collected.  Ernst's conclusions
concerning the  efficacy of marginal cost pricing as a general proposition are
thus cast in doubt, even  though his  conclusions concerning strictly
quantity-based incremental charges  (for  example,  his  conclusion  that
incremental charges would not cover costs)  are probably correct.

        These  considerations lead to the research strategy proposition that,
for properly setting marginal cost prices,  different services associated  with
solid  waste collection/disposal must be  identified  and their costs
determined.   In addition  to  amounts  collected,  these  services  include
customer  services, frequency of collection services, location of collection
services,  and kind  of waste services.

Customer Services

        Certain costs are associated strictly with adding a customer.  These
include billing and adding a stop to the route.   This means  that part  of the
total  charge should be  a  flat charge.   As an example  in another  field,
electric  utilities now  often estimate  and sometimes  include a  specific
customer  charge in their billing, based on billing costs  and amortization  of
customer hookup costs.  This avoids the erroneous inference from declining
block  charge structures that there is a scale factor involved  in the cost  of
electricity to  a customer.  Declining block charges can be replaced  with two
marginal  cost pricing components that actually apply marginal customer cost
and marginal electricity cost components.  A summer cottage customer would
then be given a true picture of his hookup service costs, and could decide
whether to take service based on these costs.  He would not  impose  costs  on
the  system greater than the value of the service to him.  In the solid waste
field,  a customer  two miles off  the normal route could decide whether  he was
willing to pay  the  additional customer service cost  to him.  Even more to the
point,  no customer's use of the  service in terms of quantities offered for
collection would be distorted by an erroneous inclusion of a customer cost  in
the incremental quantity charge, nor would the solid  waste collection  agency
have the problem of marginal cost prices not covering costs.

Other Services

        Other important services include frequency  of collection, location of
collection such as  curb or  backdoor, and kinds of waste collected.   There may
be more, such as service  on Mondays as opposed to other days of the  week, and
quieter service.  Each of these  other  services  would have  a  cost  function
separate from  the quantity of service cost  function.  The  frequency  of
collection cost function would depend on the cost of  traveling  the  route  an
additional  time per base  unit of  time, such  as per  week,  the back-door
service on distances  to back door  and time costs for labor and  route travel
time extension, and  the kind  of  waste on  the salvage or special disposal
costs of particular kinds of waste.  Service  on  Mondays, if Monday service
were in heavy demand  for  disposing of weekend  collected yard wastes, for

                                    241

-------
example,  would  involve costs  for peaking  on a Monday  (equipment for that
day's  collection alone).   A demand  for  particularly  quiet service on  a
particular route  would involve the added  cost of providing equipment and
training  personnel for quieter service.

       The literature has indicated that greater  frequency of service (Quon
1965)  and  backdoor service  (Wertz  1976)  lead to  greater  volumes for
collection/disposal.  However, while these interrelationships are important,
they do  not affect  the  basic cost functions for the  different services.
Thus,  the cost functions  can be estimated independently  of  what  these
relationships might be.

Interrelationships Affecting Costs

       Switching to  quantity based charges would  involve additional customer
costs due to the  need .for measurement of quantitites of  various services and
of quantity billing.  There may also be added littering  costs associated  with
quantity-based  charges.  Furthermore, switching from general  fund financing
would  involve  a cost due to loss of income tax advantages.  (Such financing
is federal income tax free, whereas individually borne user -charges are not.)
These  considerations lead  to the need for a benefit-cost assessment  to
determine if there would  in fact be net advantages in switching  to marginal
cost pricing for  the  services.

Cost Variation

        Basic cost  functions will vary from place to place.  In some areas,
least cost disposal  may  be  land disposal at close-in low  price per  acre
sites, in other areas the opposite.  In some areas, back doors may be close
to the street,  in others  far removed from the street.  In some areas, housing
may be high density,  others low density, leading to different costs for route
coverage.  Some areas may be more susceptible to littering  with incremental
quantity charges imposed  than others, possibly depending  on socioeconomic
characteristics.  Because of these various differences,  any general study  of
the  efficacy of  marginal cost user charges must consider the possible ranges
in variations of  conditions and associated least costs for providing various
services.

DEMAND-SIDE STRATEGIES

        To estimate  whether  the  potential gains from  marginal cost pricing
are substantial and  if such pricing is worthwhile, various demand curves  for
services must  be estimated.  However, estimation has been limited by a lack
of sufficient data.   Application of marginal cost pricing  or  any form  of
incremental pricing that  would generate the needed data has been limited.
Given this situation, a possible approach for at least partial estimation  of
demand  curves  is  one  based on estimating the cost of  alternatives  to
community collection/disposal  services.
                                    242

-------
Cost of  Production Approach

        Volumes of wastes collected  and associated waste  management services
can only be substantially decreased  if  there are lower cost alternatives  to
waste generation, to community collection, and to the  associated services
such as  frequency  of service and location of service.  The shape of  demand
functions  for the various  collection/disposal  services depends on these
alternatives and their  costs  to users.  These  alternatives include  sink
disposals, carrying wastes to curb, storing wastes over  a  longer period,
finding  alternative collectors for certain wastes and possibly  storing those
particular wastes  longer (e.g., paper  to Boy Scouts), and  changing
consumption patterns to  generate less wastes  (e.g., use of  returnable
bottles).   Lower cost  alternatives to individuals, but not  necessarily to
society, may also include littering and on-site burning.   Some  progress
toward  estimation of demand  curves can be made through research into these
costs.

        The cost  of  production  approach would cover total costs  to  society
rather  than to the individual alone for determining  optimum demand  for waste
services functions.  The  costs added to individual  costs would include those
of littering and on-site burning.  The point  can be  made  that even  if  greater
experience in  the application of marginal  cost pricing, and data  from these
applications,  were available, it would still  be necessary  to  modify demand
curves  so generated by  social cost considerations  for determining  net social
gains from marginal  cost pricing.

Creation of More Experiments

        Actual  pricing experiments would still  be necessary.  Some costs,
such  as the bother costs of  returnable bottles, and of  carrying wastes to  the
curb, would  be difficult  to estimate in the absence  of  such experiments.
Indeed, the research strategem  most needed for  investigation of economic
 approaches to  solid  waste collection/disposal services  is  the creation of
more  pricing  experiments and the generation of demand side information from
 more experiments.   The depth  of  the concerns over the costs of increasing
waste volumes and evidence  to date  on potential  gains  from marginal cost
 pricing appear  to justify this  strategem.

Two Basic  Approaches to the Creation of More Experiments

        Two  basic approaches are  possible  for  the creation  of more
 experiments and generation of  more  experimental data.  One is  the creation of
 a few  well designed controlled experiments, with certain control  areas
 remaining  on  flat charges for cross  section comparison.  Time  series  analysis
 would also be possible if historical information were available  on quantities
 collected  by  subareas.  The advantages  of this approach lie in the  care  that
 can be applied for obtaining needed and reliable comparative results.  A
 second  approach is the promotion of  more widespread adoption of pricing of
 solid  waste collection/disposal  service.  In this case, cross  section
 comparison of results with  communities not adopting pricing  would be  used.
 Multiple  regression techniques would be used  to control for other  variables.
 Again, time  series analysis  would also be  possible  given  appropriate

                                     243

-------
 historical information.  The advantages of this approach  lie  in the fact that
 greater  variation in  other variables can be considered with  a greater output
 of data to cover various conditions for the same budget and time constraints.

 A  LIST OF  RESEARCH QUESTIONS  TO ASK IN SOLID WASTE  SERVICES PRICING
 EXPERIMENTS

         The focus  of  solid waste  services pricing experiments should be on
 ultimate  applicability, providing answers to solid waste  systems managers'
 questions.   Their questions  would include what  form financing should take,
 what specific  charges  should be set and  how the level should  be  determined,
 and how  to  measure the services provided.  They would  also  include what
 services  to  provide, and what  degree of  environmental  amenities, such  as
 litter control and clean  up,  to provide.   As a research  approach, it is
 useful  to  consider  the various questions that might be asked.  The following
 outline of possible question is offered:

 1.   What collection services  and level  of  collection  services should  be
 provided?

        a.  What frequency  of  service should be provided?   Should this  be
 decided on a systemwide or  individual  collection area basis?

        b.  Should curb or back  door service  be  provided?   If  back door,
 should  it be  optional,  and  if  optional,  should  it  be optional at each
 collection?

        c.   Should unlimited-quantity  service  be  provided,  or  should
 consideration be given to approaches  that  might limit or reduce quantities  to
 be .collected?

        d.  Should combined, ,same  frequency, , and location of  collection  be
 offered for all  residential  solid wastes,  or should separate collection  be
 offered?

 2.  What degree of community cleanliness,  in  terms of  freedom  from  litter and
 other  pollution  from  solid wastes,  is demanded,  and  how  is  this  best
 achieved?                                          ,

        a.  What are the marginal benefits (what is the  demand function for)
 different  degrees of cleanliness?

        b.   What are the various means for achieving  cleanliness and what are
 the  marginal cost  functions (cost per unit of cleanliness achieved) of the
 various means?

        c.  What is the optimum level of cleanliness and what  are the ,optimum
levels of  input of  the  various means, given these demand and cost functions?

 3.   What disposal  services and level of disposal services  should be provided?
                                    244

-------
        a.   Should  there be processing of wastes prior  to  ultimate disposal
(e.g.,  incineration)?

                (1)  What are the benefits of processing in terms of reducing
transportation,  ultimate disposal, or other costs?

                (2j)  What are the costs  of processing,  including economic
costs  (net of any energy  values obtained) and environmental costs  (air
pollution)?

        b.   What are  the options  for ultimate disposal?  If the least-cost
alternative is by land disposal, what sanitary land-disposal  standards should
be set  for  this disposal?

                (1)  What are the marginal environmental  benefits of  added
efforts to  protect  against environmental  damages from the  disposal operation?
These might include benefits from:

                       (a)  More frequent covering  or  burying of wastes.

                       (b)  More careful site  selection,  for example,  to
find less permeable soils.

                       (c)  More remote  sites,  for  example, to remove the
site further from population centers, or to remove further  from groundwater
sources.

                       (d)   More care  to separate  off  and neutralize the
possible health and safety effects of potentially hazardous wastes  that  might
be introduced into  residential solid wastes.

                (2)  What are the marginal costs associated with  achieving
the marginal benefits?

4.  How should collection/disposal services be financed?

        a.  What  are  the alternative ways of  financing collection/disposal
services?  These may  include:

                (1)   Flat  charges, based on average costs of  services
actually rendered over a  period  of time.

                (2)  Incremental  user charges, based on the  incremental  costs
of the various collection/disposal services rendered.

                (3)   Strictly  marginal cost  pricing, which may  be some
combination of flat and incremental  charges.

        b.   What  are the advantages/disadvantages (benefits-costs) of
alternative financing methods compared to existing financing  methods, for
example incremental compared to  flat charge methods?
                                    245

-------
                (1)  What are the estimated  changes in the  amounts of  the
various services in shifting to  incremental user charges?

                (2)  What are the direct benefits (estimated  direct cost
savings) from reducing these amounts?

                (3)  What are the direct  costs to households (estimated loss
in benefits) from reducing these amounts?

                (4)  What are the indirect costs from shifting to incremental
charges?

                        (a)   What  are the  added  (or  reduced) costs of
administering such charges?

                        (b)   What are  the added  environmental  cleanliness
costs?

                        (c)   What are  the tax/revenue sharing costs?

5.   How can service requirements be best projected for planning and  decision
purposes?

        a.  What  are the various projection methodologies and how would they
be implemented?

        b.  Which projection methodologies are the more reliable in  terms of
accuracy?  Which in terms of  greater flexibility for different conditions  or
different modes  of operation?

        c.  Which projection  methodology, or methodologies,  should  be used
for  the particular  system?  If  more  than one, could there be irreconcilable
differences in results,  and  how  would  this be handled?

6.   What  specific-system research is needed  to make all  of the above
determinations?

        a.   What are the  specific questions to be asked and answered?

        b.   What are the  research procedures?

        c.   What are the  data needs?
                                    246

-------
                               REFERENCES


1.   Ernst, Ulrich.  1975.  Evaluation of the Feasibility and Economic
     Implications  of  Pricing Mechanisms in Solid  Waste Management.  Abt
     Associates, Inc., Cambridge, Massachusetts.  US  Environmental Protection
     Agency, January 1975.

2.   Quon,  Jimmie E.,  Masaru Tanaka, and Abraham  Charnes.  1968.   "Refuse
     Quantities and Frequency of Service,"  Journal  of  the Sanitary
     Engineering Division,  Proceedings of the American Society of Civil
     Engineers, 94, 403-420, April 1968.

3.   Wertz, Kenneth  L.  1976.   "Economic Factors  influencing  Households'
     duction Production  of Refuse,"  Journal of Environmental Economics and
     Management, 2,263-272 April 1976.
                                   Z47

-------
                                               APPENDIX  A

                                                 AGENDA
                TUESDAY, SEPTEMBER 19, 1978
                                                                     WEDNESDAY, SEPTEMBER 20, 1978
8:30-
  9)00
           WELCOME:  Robert J . Andereon
   KEYNOTE ADDRESS!  Elaynee Coddard
  ADMINISTRATIVE COSTS OF USER FEES
            Barbara Stevena
9:00-
  9.JO
        EFFICIENCY OF USER CHARGES
               William Lanen
    COMPARISONS OF ALTERNATIVE SOLID
       WASTE MANAGEMENT POLICIES
            Allen Mledatnft
9:30-
  lOiOO
     EVALUATING THE EFFICIENCY OF THE
            SOLID WASTE CHARGE
               Steve Buchanan
TAX POLICY AND SECONDARY MATERIAL USE
           Robert C. Anderaon
10:00-
  10:20
                           BREAK
                                                                                BREAK
10:20-
  10:50
    EQUITY CONSIDERATIONS IN USER CHARGES
                  Roger Bolton
             DISCUSSION
  Frank Smith            Seymour Flefcowaky
10:50-
  11:20
             OPTIONAL, POLICY MIX
                  Jamea Hudaon
 11:20-
  12:00
                  DISCUSSION
Robert C, Anderson   Peter Kemper   Kernel Goddard
       SUMMARY GENERAL DISCUSSIONi

 PRICING AND SOLID WASTE MANAGEMENT:
       FACT. FAITH, AND FICTION
12:00-
 1:30
                           LUNCH
                                                                                LUNCH
Ii30-
 2:00
           THE SEATTLE EXPERIMENT
                 Bernard Boome
        RESEARCH PROGRAMS AT EPA
            Oacar Albrecht
2:00-
 2t30
             EXPERIENCE IN EUROPE
                W. David Conn
        RESEARCH STRATEGIES
            Steve Haattaga
2:30-
 3.00
   ESTIMATES OF SECONDARY MATERIALS SUPPLY
                 Taylor Blngham
  WHAT THE SOLID WASTE MANAGER NEEDS
              Jack Dunn
3:00-
 3:20
JiZO-
 J.50
3:50-
 4:30
 7:20-
 8|30
                           BREAK
      PRICINQ MUNICIPAL REFUSE SERVICE
                 Kenneth WertB
                                                                       SUMMARY GENERAL DISCUSSION:
                                                                            RESEARCH PRIORITIES
                  DISCUSSION
    John Tbompaon            M0rrlle« Bonney
    Jochen Kuhner             Seymour Flekowaky
                  ADDRESS BY:
              David Tundarman, CEQ
                                                     248

-------
                           APPENDIX B

                     LIST OF PARTICIPANTS
Oscar W. Albrecht
U. S.  Environmental Protection
  Agency
Cincinnati,  OH

Robert C. Anderson
Environmental Law Institute
Washington, D. C.

Robert J. Anderson,  Jr.
MATHTECH, Inc.
Princeton, NJ

Tayler Bingham
Research Triangle Institute
Research Triangle Park,  NC

Roger Bolton
Williams College
Williamstown, PA

Bernard H. Booms
University of Washington
Seattle, WA

Steve Buchanan
Public Interest Economics Center
Washington,  D. C.

W.  David Conn
University of California at
  Los Angeles
Los Angeles,  CA

Philip C.  Cooley
Research Triangle Institute
Research Triangle Park,  NC
Haynes C. Goddard
University of Cincinnati
Cincinnati,  OH

V.  Stevens Hastings
WAPORA, Inc.
Chicago,  IL

James F. Hudson
Urban Systems Research
  & Engineering, Inc.
Cambridge, MA

William N.  Lanen
MATHTECH, Inc.
Princeton, NJ

Allen K. Miedema
Research Triangle Institute
Research Triangle Park, NC

Barbara J. Stevens
ECODATA,  Inc.
New York, NY

Curtis E.  Youngblood
Research Triangle Institute
Research Triangle Park, NC

Kenneth L. Wertz
University of North Carolina
Chapel Hill, NC

George  S. Tolley
University of Chicago
Chicago, IL
                               249

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 1. REPORT NO.
   EPA-600/9-80-001
                              2.
                                                            3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
   Proceedings of a Symposium on Economic Approaches
   to  Solid Waste Management
                                                            5. REPORT DATE
                May 1980 (Issuing Date)
                6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
   Robert J. Anderson, Jr.
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
   Mathtech,  Inc.
   Box 2392
   Princeton,  New Jersey 08540
                                                            10. PROGRAM ELEMENT NO.
                 1DC818 - SOS #5, Task  06
                11. CONTRACT/GRANT NO.
                 68-03-2673
 12. SPONSORING AGENCY NAME AND ADDRESS
   Municipal Environmental Research Laboratory--Gin., OH
   Office of Research and Development
   U.S. Environmental Protection Agency
   Cincinnati,  Ohio 45268
                                                            13. TYPE OF REPORT AND PERIOD COVERED
                   Final
                14. SPONSORING AGENCY CODE


                EPA 600/14
15. SUPPLEMENTARY NOTES
   Project Officer - Oscar W. Albrecht
    513/684-7881
16. ABSTRACT
   Incremental  user charges or prices as a means of assuring equity and efficiency in
   solid waste  management have interested economists  for some time.  The ultimate
   effects of pricing are not well-known, however, and most communities continue to
   rely on revenues from property  taxes and flat charges to finance the collection and
   disposal of  residential solid waste.  The Symposium sponsored by the Municipal
   Environmental  Research Laboratory (Cincinnati, Ohio)  provided an opportunity for
   economists interested in solid  waste problems to exchange ideas and views on the
   potential  role of pricing in municipal solid waste management.  This report
   contains the formal papers presented at the Symposium held at Philadelphia,
   Pennsylvania on September 19 and 20, 1978.
 7.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
  b.lDENTIFIERS/OPEN ENDED TERMS  C.  COS AT I Field/Group
   Demand (economics)
   Mathematical models
   Econometrics
    User charges
    User fees
    Solid waste
    5c
   68c
   91a
 8. DISTRIBUTION STATEMENT
   Unlimited
  19. SECURITY CLASS (This Report)
     Unclassified
21. NO. OF PAGES
   256
                                               20. SECURITY CLASS (THitpage)
                                                                          22. PRICE
EPA Form 2220-1 19-73)
250
                                                                       ftUSQPO: 1980 — 657-146/5695

-------