EPA-600/4-76-045
September 1976
   Environmental Monitoring Series
             MONITORING  GROUNDWATER  QUALITY
         ECONOMIC FRAMEWORK AND PRINCIPLES
I
55
                                      .W.
                              Environmental Monitoring and Support Laboratory
                                     Office of Research and Development
                                     U.S. Environmental Protection Agency
                                           Las Vegas, Nevada 89114

-------
                RESEARCH REPORTING SERIES

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

     1.    Environmental Health Effects Research
     2.    Environmental Protection Technology
     3.    Ecological Research
     4.    Environmental Monitoring
     5.    Socioeconomic Environmental Studies

 This report has been assigned to the ENVIRONMENTAL MONITORING series.
 This series describes research conducted to develop new or improved methods
 and instrumentation for the identification and quantification  of environmental
 pollutants at the lowest conceivably significant concentrations. It also includes
 studies to determine the ambient concentrations of pollutants in the environment
 and/or the variance of pollutants as a function of time or meteorological factors.
   This document is available to the public through the National Technical Informa-
   tion Service. Springfield. Virginia 22161.

-------
                                       EPA-600/4-76-045
                                       September 1976
        MONITORING GROUNDWATER QUALITY:
       ECONOMIC FRAMEWORK AND PRINCIPLES
                           by

                     Robert L. Crouch
                     Ross D. Eckert
                     Donald D. Rugg
             General Electric Company—TEMPO
                Center for Advanced Studies
              Santa Barbara, California  93101
                       July 1976

                Contract No. 68-01-0759
                     Project Officer

                   George B. Morgan
     Monitoring Systems Research and Development Division
       Environmental Monitoring and Support Laboratory
                Las Vegas,  Nevada 89114
      U.S. ENVIRONMENTAL PROTECTION AGENCY
        OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
              LAS VEGAS, NEVADA 89114

-------
   This report has been reviewed by the Environmental Monitoring and Support Lab-
oratory—Las Vegas, U.S. Environmental Protection Agency, and approved for publi-
cation.  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 endorsement  or recommendation for use.

-------
                          TABLE OF CONTENTS
                                                                 Page

LI STOP ILLUSTRATIONS                                              v
ACKNOWLEDGMENTS                                               v?

SECTION I - INTRODUCTION                                         1
                 Why Groundwater May Become Excessively Polluted       1
                 Preventing Excessive Pollution of Groundwater           4
                 Groundwater Management Objectives                   4

SECTION II - INSTITUTIONAL AND LEGAL BACKGROUND               7
                 The Legal Situation                                   7
                 PL 92-500 (Federal Water Pollution Control Act) and
                    PL 93-523 (Safe Drinking Water Act)                10
                 EPA's Evolving Water Quality Strategy                 11

SECTION III - HYDROGEOLOGIC CASE STUDY AND ECONOMIC
              PRINCIPLES                                           13

                 Hydrogeologic Example                              13
                 Economic Principles                                 18

SECTION IV - MONITORING AND GROUNDWATER QUALITY
              STANDARDS                                          30
                 Definitions of Information and Compliance Monitoring    30
                 Information  Monitoring                              31
                 Compliance Monitoring                              34
                 The Pragmatic Alternative—Second-Best Solutions        61
SECTION V -WASTE RELOCATION RIGHTS: AN ALTERNATIVE
             SYSTEM OF GROUNDWATER MONITORING AND
             POLLUTION CONTROL                                 66
                 Administrative versus Nonadministrative Pollution
                    Monitoring and Control                            66
                 Waste Relocation Rights                              69
                 Court Enforcement and Strict  Liability                  71
                 The Zero Waste Relocation "Right"                    73
                                                        (continued)
                                   in

-------
CONTENTS (continued)
                                                                        Page

                   Setting the Initial Waste Levels                          74
                   Exchanging Waste Relocation Rights                      75
                   Restructuring Waste Relocation Rights                    76
                   Sales of Land with Waste Relocation Rights               78
                   Waste Relocation Rights and Nonlandowners              81
                   An Appraisal of Some Potential Problems                 82
                   Situations Inappropriate for Waste Relocation Rights       91
                   Conclusions                                            93

 REFERENCES                                                              94
 APPENDIX                                                               97
                                        IV

-------
                            LIST OF ILLUSTRATIONS
Figure No.                                                                 Page
     1         Geography of case study.                                      14
     2         Hydrogeology of case study:  elevation.                         15
     3         Hydrogeology of case study:  plan.                              17
     4         Alternative steady-state time profiles of chloride concentration.  18
     5         Farmer's total damages function.                                21
     6         Farmer's marginal damages function.                            21
     7         Oil company's total benefits function.                          22
     8         Oil company's marginal benefits function.                       23
     9         The efficient solution.                                         24
   10         First limiting case.                                             28
   11         Second limiting case.                                          29
   12         The probability of detection function.                           38
   13         The gain from a standards violation.                            39
   14         Detection and the size of the violation.                         39
   15         The probability of detection and the severity of the violation.     41
   16         The marginal cost of compliance monitoring.                     42
   17         The expected fine and the seventy of the violation.              45
   18         The expected fine and expenditure on compliance monitoring.     47
   19         The expected cost and the severity of the violation.              47
   20         The optimum violation.                                        49
   21         The preference function.                                       49
   22         The revised optimum violation.                                 50
   23         The benefits from violation reduction.                           52
   24         Diminishing returns to violation reduction.                       52
   25         Compliance monitoring and  the total social loss avoided.         53
   26         The total benefits from compliance monitoring.                   55
   27         The marginal benefits from compliance monitoring.               55
   28         The optimal  probability of detection.                           56
   29         Second-best alternatives.                                      61
   30         Land parcel  diagram,  chloride pollution example.                76
   31         Seawater intrusion example.                                    92

-------
                        ACKNOWLEDGMENTS
   Dr. Richard M.  Tinlin, Dr. Lome G. Everett,  and the late Dr.  Stephen
Enke of General Electric —TEMPO were responsible for management and
technical guidance of the project under which this report was prepared.

   The following officials were responsible for administration and techni-
cal guidance of the project for the U. S.  Environmental Protection Agency:

      Office  of Research and Development (Program Area Management)

         Mr. Albert C. Trakowski,  Jr.
         Mr. John D.  Koutsandreas

      Environmental Monitoring and Support Laboratory Las Vegas
      (Program Element Direction)
         Mr. George B. Morgan
         Mr. Edward A. Schuck
         Mr. Leslie G.  McMillion
         Mr. Donald B. Gilmore

    The following personnel of the U. S.  Environmental Protection Agency
 are to be thanked for their  review and constructive criticism of this  report:
 Mr.  George A. Garland, Deputy Director,  Systems Management Division,
 Office of Solid Waste Management Programs, Mr.  H. R. Reinhardt of the
 Office of Technical Analysis, and Dr. Richard Schaefer,  Office of Air,
 Land, and Water Use, Washington,  D. C.
                                   VI

-------
                             SECTION I
                           INTRODUCTION
   This report presents an analytical framework for evaluating the eco-
nomic issues raised by monitoring in support of the development and en-
forcement of groundwater quality standards.   The discussion that follows
incorporates certain legal and institutional aspects of various monitoring
philosophies and methodologies with economic principles and issues.
This analytical framework is presented in terms of case studies and theo-
retical situations,  some aspects of which may appear to be in conflict with
the mandates of the Federal Water Pollution  Control Act,  as amended by
Public Law 92-500, and the  Safe Drinking Water Act, Public Law 93-523.
Accordingly, it is to be understood that the concepts, as presented in this
discussion, do not necessarily reflect U.S.  Environmental Protection
Agency policy, although the  document has been reviewed by the EPA and
approved  for publication.

   Section II is a brief description of the existing institutional and legal
structure; Section III presents a hydrogeological case study and develops
the analytical framework in  which the economic issues  raised by ground-
water monitoring will be discussed; Section IV is an economic  analysis
of the principles governing groundwater monitoring in support  of (1) de-
veloping groundwater  quality standards,  and  (2) enforcing compliance
with those standards;  Section V examines an  alternative legal and insti-
tutional structure designed to ensure that an  optimum use of the waste-
assimilative capacity  of the  Nation's groundwater is achieved.


WHY GROUNDWATER MAY  BECOME
EXCESSIVELY POLLUTED
   If (1) all markets were  competitive, (2) market transactors had full
information, (3) resource-owners put the resources they own to their
highest valued use, and (4) enforceable property rights  existed in all re-
sources,  then society's economic resources would  be efficiently allocated.
"Efficiently allocated" is used to mean that no other allocation would be
possible which would make one transactor better off without making at
least one  other transactor worse off.  Such an allocation is also described
as pareto-optimal (Arrow  1951 and 1962, Arrow and Debreu 1954, Quirk
and Saposnik 1968).

-------
   Under such an allocation of society's resources, groundwater re-
sources would be used in .optimal amounts for irrigation,  cooling, drink-
ing,  etc. , as well as for the disposal of wastes.  Although the optimal
portion of groundwater resources used for waste disposal under such an
efficient allocation of resources  may be small, it cannot be stressed too
strongly that an abiding concern  for the environment will not change the
fact  that wastes are an unavoidable byproduct of all man's activities.
Recognizing that these wastes must be disposed of somewhere,  the  sub-
stantive  questions of the optimum production of wastes  and their optimum
disposal must be addressed.   Clearly, the Nation's aquifers, like every
other sector of the environment, are a candidate repository for the dis-
posal of some of society's wastes.   The question, then, is not "whether"
but  "where" and "how much. "

   The evidence indicates that some of the Nation's groundwater (like our
 other  environmental resources  such as the air and surface waters) has
 become, or is becoming,  excessively polluted.  This means that ground-
 water is not being efficiently allocated among its alternative uses.  On
 the  contrary, its potential for irrigation, drinking,  cooling, etc. ,  is being
 abused by the overutilization of groundwater resources as a repository for
 society's wastes.

    The  reason that groundwater becomes excessively polluted is that the
 fourth condition necessary for the efficient allocation of resources, en-
 forceable property rights, is not met with  respect to  groundwater.  In
 general, property rights  in groundwater are poorly specified.

     If the property right in a resource is well specified, the owner  has
 the maximum incentive to invest in the optimal amount of information on
 the alternative uses for his resource because he can  appropriate to him-
 self most of the rewards from  his investment in acquiring that informa-
 tion.  Indeed, the greater the appropriability of the benefits that can be
 derived from a resource, the greater the cost to an owner of no_t_allocat-
 ing that resource to its highest valued use and, therefore, the greater
 the incentive for economic efficiency.  Hence, a better specification of
 property rights in resources leads to the appropriation of greater bene-
 fits from those resources by owners and this in turn leads to more effi-
  cient conservation of society's resources.

     A burgeoning literature under the general rubric of "property rights"
  shows  that different configurations of legal rights and duties impact on
  the resource owner's cost-reward calculus  and generate different re-
  source allocations.   See, for example,  Alchian (1961),  Alchian and Kes-
  sel (1962), Coase (I960), Demsetz (1966 and 1967),  De Vany et al. (1969),
  and Furubotn  and Pejovich (1972).

-------
   The property rights to a resource are well specified when (1) the re-
source can be defined unambiguously (or, as in the case of land, where
there are well-defined procedures for delimiting its boundaries), (2) the
owner can capture all the benefits from the use of the resource to the ex-
clusion of all other persons,  (3) it can be protected at low cost against
unauthorized use,  and (4) title to the resource can be preserved and trans-
ferred among owners  (a tract index, for example, in the case of land).

   The importance of  specifications can be appreciated best when there
are no specifications.   An ambiguous description of the resource raises
negotiation or litigation costs and will, therefore, restrict exchanges;
nonexclusivity in the appropriability of the benefits from the resource
and high enforcement costs to prevent unauthorized use both attenuate
the incentive to  husband the resource,  while imperfect titles and high ex-
change costs diminish the incentive to transfer the  resource to higher
valued uses.

   The property rights in the Nation's groundwater are poorly specified
on all four counts given above.  Aquifers are poorly defined  "common
pools" with titles to their use and abuse not effectively preserved or trans-
ferred.  In particular,  lack of exclusivity in the appropriability of the
benefits from aquifers  and the high enforcement costs to prevent unauth-
orized use have given  rise to two classic problems in the economics of
hydrogeology.   The first (associated with the lack of exclusivity in the
appropriability  of the benefits) is the problem of overuse or too rapid
depletion of aquifers.   (For an authoritative discussion of the problem
see the appendix by Brown in Corker (1971).)  The second (associated
with the high enforcement costs to prevent unauthorized use) is the prob-
lem of excessive pollution of  aquifers.   When one party uses another
party's resource without authorization,  he thereby imposes uncompen-
sated damages on the resource owner.   An external cost is said to have
been generated.   (For the modern view of the connection between property
right specifications and the existence of externalities see Buchanan and
Stubblebine (1962), Cheung (1969), Coase (I960),  Demsetz (1964 and 1969),
Mishan (1971),  and Stigler (1961).  In general,  externalities  arise when
problems exist with respect to the definition, exchange, policing, and en-
forcement of property rights.)

   In the absence of any obligation to compensate those whose aquifer he
is using as a waste disposal site, such waste disposal services appear free
to the polluter.  Polluters will, therefore,  utilize such a service until
the marginal private benefit obtained by them from using that service is
driven down to zero —even though the marginal social cost may be positive
and large. Consequently, the aquifer's waste  receptor capabilities are
abused and it becomes excessively polluted.

-------
PREVENTING EXCESSIVE  POLLUTION
OF  GROUNDWATER
  In principle, there are two ways to prevent excessive pollution of
groundwater.  First, by government intervention in the effluent discharge
process.  Under this regime the government may either set, and enforce,
groundwater quality standards by some form of direct  regulations involv-
ing  discharge  licenses, or induce conformity to the standards by levying
a tax on effluents or paying a subsidy for effluent reduction.  (The moni-
toring  implications of direct government intervention in the effluent dis-
charge process are discussed in Section IV. )  Second,  the property rights
to groundwater can be respecified so that they approach more closely the
four requirements  of a well-specified property right.  (This possibility
and the monitoring that would be implied are discussed in Section V.  )
Under this response to the problem of excessive groundwater pollution,
the government's direct intervention is  kept to a minimum.  It merely
performs its function of assigner of property rights through legislation
and arbitration of disputes 'concerning property rights through its judicial
media.


 GROUNDWATER  MANAGEMENT OBJECTIVES

    Views similar to those outlined under this  subheading have been ex-
pressed by the U. S.  Council on Environmental Quality (1973).  Efficient
 environmental decision-making requires that consideration be given to
 four categories of costs.  First there are damage costs.  These are
 costs  which are generated directly by a polluting activity.  With respect
 to  groundwater resources, one example would be increased physiological
 damage caused by pollution of drinking water. Another example would
 be crop losses  resulting from pollution of an irrigation well.  Second,
 there are avoidance costs.  These are  costs which are incurred by soci-
 ety in order to  avoid,  or reduce, damage costs.  With respect to ground-
 water resources,  one example would be the importation of unpolluted
 water to replace that previously obtained from a well that has become
 polluted.  Third, there are abatement  costs.  These are costs associated
 with the reduction of pollution.  Such reduction of pollution can be achieved
 either by controlling the source or by treating the polluted water.  With
 respect to groundwater resources, one example would be the deep injec-
 tion into a safe geologic zone of noxious effluents previously disposed of
 in an aquifer.  Fourth, there are transactions costs.  Transactions costs
 include the cost of all those resources allocated to the establishment, and
 enforcement, of  environment-preserving policies  and regulations.   With
 respect to groundwater resources, the most important example of a trans-
 action cost,  and  of special relevance to this  study, would be the cost of
 monitoring groundwater pollution either to generate information on  quality
 and quantity or to detect violations of, and ensure compliance with, ground-
 water quality standards.

-------
    The essential principle to note is that these four cost categories are,
 in general,  interdependent.  Thus,  any new groundwater quality policy
 will probably affect all four categories.  For example,  if a policy is in-
 troduced" which is designed to reduce groundwater pollution, it will cer-
 tainly reduce damage costs but may very well increase abatement, avoid-
 ance, and monitoring costs.  Usually,  each feasible groundwater quality
 policy will affect these four categories of cost differently.   Obviously the
 only  groundwater quality policy alternative that should be seriously con-
 sidered for  implementation is that set of policies for which the reduction
 in damage costs exceeds the net increase, if any,  in avoidance, abatement,
 and monitoring costs.   To go even further than this, the most efficient, or
 optimal,  groundwater quality policy among the feasible set of alternatives
 is that policy which minimizes the sum of the damage costs, avoidance
 costs, abatement costs, and monitoring costs for a given groundwater
 pollution situation. Implicitly,  the minimization of these costs is equiva-
 lent to the maximization of society's income or gross national product
 (GNP).   Therefore, for this analysis the maximization of GNP has  been
 adopted as the environmental management objective. *
                                            r*
    To illustrate by example,  sewer leakage is known to pollute ground-
 water and may, therefore, impose certain damage costs and avoidance
 costs.  However, with present technology there is no way of controlling
 this source of pollution that does not impose abatement costs that are far
 in excess of the reduction in damage and avoidance costs which would be
 achieved.  It follows that the efficient policy is not to monitor and abate
 this pollution source,  but simply to accept the existing level of damage
 and avoidance costs.

   Of course, in many other groundwater pollution situations the reduc-
 tion in damage costs will exceed the increase in avoidance,  abatement,
 and monitoring costs that bring those reductions about.  The objective
 then becomes that of selecting the avoidance, abatement, and monitoring
 strategy which generates, at the margin,  decreases in damage costs  just
 equal to the increases in the avoidance, abatement, and monitoring costs
 required by the  strategy.

   The attainment of this objective will not simply involve the minimiza-
tion of monitoring costs by the responsible government agency.  For
 example,  in  any given groundwater pollution situation there may well be
 several different abatement strategies,  each with an associated monitor-
ing requirement, that would achieve the desired level of groundwater
*For a comprehensive discussion of other possible objectives, see Dorf-
 man and Dorfman (1972), Chapter  1.  They observe,  however, that "The
 GNP criterion is the most practicable one of all, in fact the only one
 that can be applied with much assurance. " (p xxvii).

-------
quality.  If the public agency responsible for selecting among the abate-
ment and monitoring strategies simply chooses that alternative which
minimizes its own monitoring costs,  this could imply higher private
abatement costs with the result that the combined monitoring and abate-
ment costs of that policy would be greater than the combined monitoring
and abatement costs of some other strategy.  Clearly, this would not be
efficient from society's point of view.

-------
                             SECTION II
               INSTITUTIONAL AND LEGAL BACKGROUND
                                                                 t
    Section I points out that the major reason why the Nation's ground-water
 may become excessively polluted is that property rights in groundwater
 are poorly  specified and so do not effectively protect aquifers against un-
 authorized  use.  Consequently, the capacity of aquifers to assimilate
 waste is being abused and uncompensated damages,  or external costs,
 are being imposed on the owners of the resource.  This section focuses
 first on the issue of unauthorized use by polluters, then on corrective
 efforts to date.


 THE LEGAL  SITUATION
    Views expressed under this subheading rely heavily on Corker's au-
 thoritative discussion of groundwater law (Corker 1971, Chapter III).  It
 may be stated as a general proposition that the law does grant a land-
 owner a property right in the groundwater underlying his land.  This
 title is granted under the doctrines of "absolute ownership, " "reasonable
 use, "  "correlative rights, " or "appropriative rights, " depending on the
 State in which the landowner resides.  In addition, a landowner whose
 groundwater is polluted  by  some other party may  seek redress through
 the courts.  Such redress may be sought under the traditional four con-
 cepts of negligence, nuisance, trespass, and, less frequently, strict
 liability.  Of these legal remedies, resort is had most often to the law
 of nuisance. Thus, theoretically, the law stands behind a  landowner who
 wishes to protect his groundwater against unauthorized pollution.   The
 problem is that,  in practice,  the actual application of the law of nuisance
 is fraught with uncertainty.  Corker states:

      Nuisance is an amorphous term which describes the  sub-
      stantial interference with the use of property for which a
      court  will grant relief.  To recover on a theory of nuisance,
     when pollution is intentional,  as is often "legally" the case*
     one must in addition to proving the existence of a nuisance
     also prove  that the activity is "unreasonable. "  What is
^Corker's footnote states, "One is said to intend the results of his con-
 duct if he continues to engage in an activity after recognition of its side
 effects. "

-------
     "unreasonable" depends upon a judicial weighing of an
     open ended list of factors, including the extent and charac-
     ter of the harm involved,  the social value which the law
     attaches  to the respective parties'  activities, the suitabil-
     ity of each party's conduct to the locality, and the burden
     that would be imposed by compelling an interfering party
     to cease  carrying on his activities. Because a particular
     activity will be characterized as a nuisance in one situa-
     tion, but not in another, a user of  land cannot always pre-
     dict,  in  advance of a court room battle, whether he will
     be held liable for the pollution caused by his activities.
     Yet, despite difficulties with the concept of nuisance in
     the pollution field and elsewhere,  legal critics have not
     found an acceptable substitute.   (Corker, 1971).

   The only legitimate inference that a landowner could make from this
authoritative interpretation of the law of nuisance as applied to ground-
water pollution is  that his property right in underlying groundwater is
decidedly shaky when it comes to its protection against unauthorized use
by polluters.  Given the uncertainty of retribution in the courts under the
law of nuisance, it should not be found surprising that polluters have
abused their neighbor's groundwater.

   When not dealt with under the legal rubric of nuisance (and in the ab-
sence of negligence, trespass,  or abnormally dangerous  conduct), lia-
bility for pollution of groundwater has,  instead,  been dealt with under
the doctrines of absolute ownership, reasonable use,  correlative rights,
or appropriative rights.

       Depending on the jurisdiction, the doctrines of absolute
       ownership,  reasonable use, and occasionally the doctrines
       of correlative rights or prior appropriation have tended to
       be mechanically applied ....   [Application  of] the doc-
       trines  of absolute  ownership,  correlative  rights, and prior
       appropriation to disputes among  competing water and land
       users can produce surprising and undesirable results.
       Under  the literal doctrine of absolute ownership, a water
       user will not be able to recover for injury to his water
        supply, even when the land user  can foresee  and avoid  such
        damage.  Under the doctrine of correlative rights a third
        party may be required to pay damages, even though injury
        is unforeseeable and his activity was a natural and legiti-
        mate use of his land.  Under a strict reading of prior
        appropriation, a mining or land  development corporation
        can be enjoined from use of land, even when the proposed
        operations will drain but one shallow neighboring well.
                                    8

-------
       The reasonable use doctrine in a third party setting is also
       unsatisfactory, although use of this doctrine has produced
       fewer disquieting results in the past.   Courts have utilized
       this doctrine to prevent recovery of damages when injury
       was unforeseeable,  a result consistent with tort doctrines
       of nuisance,  and strict liability.  This distinction  between
       percolating water and subsurface streams was developed
       and has been utilized to prevent imposition of liability for
       unexpected results.   It is uncertain, however,  whether
       under a reasonable use theory, damages can be recovered
       because injury to  "percolating" groundwater  was a fore-
       seen, but unavoidable,  consequence of land use and the
       land operation was pursued in the face of this knowledge.
       If in determining whether a land use is "reasonable" the
       extent of damage to the aquifer and other surrounding
       circumstances  are overlooked, the result will at times
       be clearly wrong.  If, however, in the determination of
       whether a use is reasonable the gravity of the harm done
       to the water user is weighed against the utility of the land
       use, the result will be the same as under nuisance theory
       .  . .  (Corker,  1971; emphasis added).

 It is clear that the attempted resolution of disputes over  the liability for
 groundwater pollution in water doctrinal terms is as unsatisfactory as
 resolutions sought through the law of nuisance. Resource owners will
 not  feel  secure in their property and protected against unauthorized users
 when the law produces "surprising and undesirable results. "

   Two cases quoted in the American  Law Reports,  Annotated, Second
Series (38ALR 2d),  "Liability  for Pollution of Substream  Waters, " illus-
trate by particular examples the uncertainty and arbitrariness  which sur-
round liability for groundwater pollution:

       Where a company drilling for and producing oil brought
       salt water to the surface . . . the fact that the company
       connected a pipeline conveying its salt water with a line
      of the sewer system of a city and the city permitted the salt
      water to flow into a canal, from where it seeped into the
      source of the water supply of the plaintiff, thereby  damaging
      it, did not excuse the company from the liability for damages
      for allowing the salt water to leave its  premises.  Berry v.
      Shell Petroleum Co.  (1934) 140  Kan 04, 33 P2d 953, reh den
      141 Kan 6, 40 P2d  359.
                                 Q   REGION III LIBRARY
                                 y   E:r/rnoN-is:iTAL PROTECTION AGENC?

-------
In this case it would seem reasonable to have held the city guilty and the
oil company innocent.  But such was not the verdict.   Conversely, consider:

      Phoenix v. Graham (1953) 349 111 App 326, 110 NE2d 669,
      in which the plaintiffs sought recovery for contamination of
      their water well by salt water from the pits into which the
      water was diverted upon separation from the oil.  The court,
      reversing a judgment for the plaintiffs, .  . .  pointed out that
      the  cases  in other jurisdictions are practically uniform that
      the  operator is not liable to his lessor for salt-water damage
      unless it was caused  by negligence in the operations.

In this case, it  might seem reasonable to  hold the oil company liable for
its nuisance.  However, such was not to be the appeal court's verdict.
As might  be expected, it does  seem easier to obtain redress in the courts
if negligence on the part of the polluter can be shown. However,  this is
little comfort to an owner of groundwater  because,  in practice, negligence
is often not present even though pollution of his property occurs.

   Even from this brief analysis of the application of the law of nuisance
and various water doctrines to the problem of liability for groundwater
pollution, it is  apparent that aquifer owners are not secure in the protec-
tion of their property against unauthorized users.  Consequently, it should
be no surprise  that such unauthorized use occurs and that the Nation's
groundwater is  threatened with excessive pollution.   It is against this
background of legal uncertainty that Corker felt compelled to recommend
statutory, and publicly administered,  water quality controls which would
require permits for the execution of activities which are potentially harm-
ful to the Nation's groundwater (Corker,  1971).

PL 92-500  (FEDERAL WATER POLLUTION CONTROL ACT) AND
PL 93-523  (SAFE DRINKING WATER ACT)
   Statutory,  publicly administered, and permit-oriented legislation di-
rected towards  the protection of the Nation's water supplies in general,
including its groundwater  supplies, was enacted by Congress in October
1972 over the President's veto in the form of the Federal Water Pollution
Control Act, PL, 92-500.  Analogous legislation designed to protect the
Nation's drinking water supplies became  law in December 1974 in the
form of the Safe Drinking Water Act, PL, 93-523.

   PL 92-500's objective "is to restore and maintain the chemical, physi-
cal, and biological integrity of the Nation's waters" (Section 101(a)).  In
pursuit of these objectives, "The [EPA] Administrator shall establish
                                  10

-------
 national programs for the prevention, reduction,  and elimination of pol-
 lution and as part of such programs shall .  .  . establish,  equip,  and
 maintain a water quality surveillance system for the purpose of monitor-
 ing the quality of ... groundwaters" (Section 104(a) (5)).   Similarly,
 PL 93-523 obliges the EPA Administrator to  ". .  .  establish recom-
 mended maximum  contaminant levels for each contaminant which .  . .
 may have  any adverse effect on the health of persons (Section 1412  (b)
 (1) (B)).  Before the Administrator may approve a State's  regulatory
 program,  he must be assured that the State's program includes "inspec-
 tion,  monitoring, recordkeeping, and reporting requirements"  (Section
 1421(b) (1) (C)).

   In the larger  sense,  this report  explores the issues involved in estab-
 lishing, equipping,  and  maintaining alternative water quality surveillance
 systems in order to reveal the principles on which the  selection of a pre-
 ferred surveillance, or monitoring, strategy should be based.


 EPA's  EVOLVING WATER QUALITY STRATEGY
   The obligation to implement PL  92-500 and PL 93-523 falls on the
 EPA.  In execution of that responsibility, the EPA issued  the second edi-
 tion of its  Water Quality Strategy Paper in March  1974 (USEPA, 1974).
 The EPA's strategy is not cast in concrete and a conscious effort has
 been made to ensure that "there is flexibility  to adjust to ...  circum-
 stances through periodic revisions" (USEPA,  1974).  However,  certain
 policies appear to be well-enough established to rely on them as guide-
 posts  to direct this inquiry.  The most important of these  is the policy
 to control  pollution by a groundwater quality  standards  mechanism im-
 plemented through  a permitting scheme.  Consequently, this analysis  is
 limited to monitoring in the context of such a permitting scheme and
 does not pursue the issues involved in monitoring if groundwater quality
 standards were to be achieved through an effluent charges  scheme,  a
pollution reduction subsidy scheme, etc.*  The only exception to this
limitation occurs in Section V,  which explores an alternative legal and
institutional  structure designed to ensure an optimum use of the waste-
assimilative  capacity of the Nation's aquifers.
*It cannot be stressed too  strongly that this  restriction does NOT imply
 an opinion that direct regulation by a permit scheme is the most effi-
 cient form of groundwater pollution control.  An elementary  discussion
 of the pros and cons of different pollution control schemes is contained
 in the Fourth Annual Report of the U.S.  Council on Environmental Qual-
 ity (USCEQ,  1973); a more sophisticated treatment is contained in Maler
 (1974).
                                  11

-------
   In March 1974 the EPA concluded that its "review of FY 1974 monitor-
ing activities indicates that . .  .  there is still a need to establish a basic
national direction of effort"  (USEPA,  1974).  The remaining sections in
this report  are intended to help direct the Nation's groundwater monitor-
ing effort along fruitful paths.
                                   12

-------
                              SECTION III
                     HYDROGEOLOGIC CASE STUDY
                      AND ECONOMIC PRINCIPLES
    This section and Section IV examine the monitoring implications of a
 government permitting scheme to prevent excessive groundwater pollu-
 tion from effluent discharge.

    Since groundwater is a relatively unfamiliar resource, a hypothetical
 case study  is used to  convey the minimum essential hydrogeologic infor-
 mation that is required to obtain an appreciation of the complications in-
 volved in monitoring. This hydrogeologic example is  then used to illus-
 trate the economic issues involved in groundwater quality control and its
 monitoring.


 HYDROGEOLOGIC EXAMPLE
    The hydrogeologic example involves a wide variety of factors which
 include the  basic situation (parties involved,  land use,  etc. ), geography,
 hydrogeology, existing groundwater pollution information, and the con-
 figuration of the pollution plume.


 Pollution Situation*

    The illustrative example is a simple one involving an oil company
 (Party A) and a farmer (Party ]3). Party 13,  the pollutee, uses a  100-
 gallon per minute (gpm) well located on his property for supplemental
 irrigation of 1000 acres  of level land. ^ Although Party B_ is  not aware
 of it, his well is  becoming polluted by an unlined oil-field brine disposal
 pit operated by Party  A.  The farmer uses his land to  grow rice —a crop
 for which irrigation is~mandatory, given the average local rainfall of
 48 inches per year.  Since the rain is fairly evenly distributed through-
 out the year,  the well is used for irrigation only during the months of
August and  September.  The farmland is easily converted to  produce
 cotton,  soy  beans, and Bermuda grass —lower valued crops which are
less sensitive to  salt pollution and do  not require irrigation,  although
irrigation greatly increases yields.
*For an actual case study  see USEPA (1972).

 An English-to-metric unit conversion table is appended to this report.
                                 13

-------
Geography

   The geographic relationship between the disposal pit and the irrigation
well is presented in Figure 1.   The farmer's well is located 750 feet from
the unlined disposal pit, and 250 feet from the property line.  The eastern
border of the farmer's  land is an ocean coastline into which the ground-
water  ultimately flows and from which seawater may conceivably intrude.

   A 500-milligram per liter (mg/1) isopollution contour is shown at time
to when pollution is detected at B_'s well.   (An isopollution contour is  de-
fined by the locus of all points of equal pollution intensity. ) This concen-
tration level is of critical importance since rice is  assumed to be insen-
sitive  to lower chloride concentration levels.  The dimensions of the
plume are not known to either  party at this time;  all that is known by B is
that his well is polluted.

Hydrogeology

   An  unsealed elevation of the hydrogeological situation at time (t(j) is
shown in Figure 2.   The geology of the strata underlying the subject area
is assumed to be known to the  polluting oil company from its drilling
        A's PIT
                                 PROPERTY
                                 BOUNDARY
                                                     COASTLINE
                                     500 mg/l
                                     ISOPOLLUTION CONTOUR
                                        B's WELL
                                GROUNDWATER
                                GRADIENT
OCEAN
                       Figure 1.  Geography of case study.
                                   14

-------
   INDIRECTION OF FLOw||
NOTE: NOT TO SCALE =
                               Figure 2.  Hydrogeology of case study: elevation.

-------
activities but not to the farmer.  The farmer's well draws from an un-
confined alluvial aquifer near the surface.  This aquifer is separated
from a deeper unpolluted sand aquifer by a layer of confining clay.   Thus,
in principle, the farmer can avoid the damage he is sustaining by deep-
ening his well into the unpolluted sand aquifer.

   The alluvial aquifer is about 40 feet thick and the water table is  about
10 feet below the surface of the ground.  The polluted aquifer has a per-
meability of 1500 gallons per day per foot (gpd/ft) and a groundwater
gradient of  1.4 feet per mile in an easterly  direction (i. e. , directly from
the brine disposal pit and past the water well to the ocean).  With this
permeability and gradient the natural rate of groundwater flow is about
20 feet per year.  It therefore takes about 38 years for the pollutants to
travel from the  brine pit to the irrigation well.   Since groundwater gener-
ally has  an  extremely slow rate of movement the rate hypothesized here
is not at all unusual.  These slow rates of groundwater movement and
pollutant transport complicate analysis of pollution problems.

   Note that seawater intrusion represents a second possible pollution
source and  that the groundwater polluted by the disposal pit eventually
flows into the sea, which is assumed to be a zero cost dump for the
contaminated water.

Existing Pollution Information

   The types of information required in an analysis of groundwater pollu-
tion situations include the type, source, distribution, concentration, and
movement of pollutants.  It is assumed, as indicated above, that the
farmer has recently identified  the type of pollutant by appropriate chem-
ical testing, but that he is entirely  ignorant of all the other information
and hence not able to predict probable future damages to his crop.  It
is further assumed that the oil company is completely unaware of the
problem.


Pollution  Plume
   The three-dimensional pollution plume changes  shape over time due
to the normal diffusion process occurring within the aquifer.  For  pur-
poses of case specification the chloride  concentration level at the farm-
er's well is of particular interest between the time when the pollution is
first discovered by the farmer (IQ)  and the time at which the plume con-
figuration reaches a steady-state configuration (t«).  The steady-state,
or long-run equilibrium, position in the absence of any abatement activ-
ities, is shown  in Figure 3.  By time tg  all of B_'s groundwater would  be
tainted by chlorides in excess of 500 mg/1 and the chloride concentration
at his well would be 5000 mg/1.  Such levels of pollution are assumed to
                                  16

-------
                     GROUNDWATER
                     GRADIENT
                                                        COASTLINE
                                                                 OCEAN
                Figure 3.  Hydrogeology of case study: plan.
 he lethal to all feasible crops.  The time-profile without abatement of the
 chloride concentration at B's well between t0 and tg is shown by the solid
 curve labeled Clj in Figure 4.  If the polluting activity is abated com-
 pletely at to by some pollution control device such as pit-lining and rely-
 ing on evaporation to dispose of the water, the time-profile of chloride
 concentration at B's well would be as shown by the dashed curve labeled
 Cl£.  When effective abatement is undertaken the chloride concentration
 gradually decays back to the ambient level as a result of dilution and flush-
 ing processes.  Because of the  slow rate of pollutant transport by ground-
 water,  the time taken for the maximum  chloride concentration to occur
 at the farmer's well if no corrective  action is undertaken, or the time
 taken for the chloride concentration to return to the ambient level if effec-
 tive abatement is undertaken, can both be considerable—easily 30 to  100
 years.  Because of these sizable time lags in the movement of polluted
 groundwater, one strong inference may  be drawn immediately.  If it pays
 at all, it will probably only pay  to treat  existing polluted groundwater at
 the point of use because abatement at the point of source may take decades,
 or even centuries, to have any impact.  Consequently,  with any reasonable
discount factor, the present discounted value of the net benefits derived
from source abatement will almost certainly be zero or negative.
                                 17

-------
CHLORIDE AT
FARMER'S WELL, mg/l
                                                    Cl, , STEADY-STATE
                                            CL, CORRECTIVE ACTION
                                               TAKEN AT tQ
                                            CI0, AMBIENT LEVEL
                                                                    TIME, YEARS
       Figure 4.  Alternative steady-state time profiles of chloride concentration.

 ECONOMIC PRINCIPLES
    Since the optimal allocation of resources to groundwater quality moni-
 toring depends on the optimal  level of aquifer  quality and especially on the
 costs and benefits  to  society of deviations from that optimum level of
 aquifer quality, it  is  necessary to discuss how the optimal level of aquifer
 quality is, itself, determined.

 The Damages (Cost) Functions

    In the illustrative  case the  costs of pollution are incurred by the farmer
 in the form of crop damage; increases in brine pollution will first reduce
 rice crop yields and eventually force the farmer  to  employ his land in the
 production of irrigated crops with a lower net value.  Ultimately, the
 brine pollution will force the farmer to cease irrigation from his well
 entirely and rely on natural precipitation, drill a new well, import water,
 etc.  The  total damage (D^) to the farmer (B) attributable to the oil com-
 pany (A) pollution  of  his well can be represented as follows:
                          t=<»

                      = N  1   JH - K[Cl(t)]j

                          t=0
dt
(1)
                                    18

-------
  where
        N        is the number of acres under cultivation

        t         is the time period (years)

        H        is the net annual dollar return from an acre when
                 employed in its highest valued use and the farmer's
                 well is not polluted*

        K[C!(t)]  is the net annual dollar return to an acre of land
                 when employed in its highest valued use, given that
                 Cl(t) is the average chloride concentration at the
                farmer's well during time t

        r        is the discount rate.

    In words,  the damage per acre at time (t) is the difference between the
 net annual return from the land in its highest valued use given no pollution
 (H) minus the value of the net annual return from the best crop that can be
 grown given the chloride concentration at time (t),  K [Cl(t)] .  The magni-
 tude of this net  loss per acre is then discounted into present value terms
 by the discount  factor (e~rt) summed over all time periods (here assumed
 to be infinity) and multiplied by N, the number of acres under cultivation.
 The  result is  the total loss (D^) incurred by the farmer due to the pollu-
 tion  activities of the oil company.

    The variable K[Cl(t)], i. e. , the net annual dollar returns to an acre
 when employed in its highest valued use, given that the average chloride
 concentration  in the farmer's well during  t  is Cl,  will decrease as Cl
 increases.   There are two basic ways for this to occur.  First, as Cl
 increases either the yields of the farmer's rice crops will decline or he
 will be forced to substitute crops with a lower market value; either alter-
 native means that K[Cl(t)] will decrease as Cl(t) increases.  Second, the
 farmer can continue to grow rice crops and maintain their yields by drill-
 ing a new well, importing water, desalinizing his existing well, etc.
 Each of these alternatives, however, will increase his costs,  and conse-
 quently K[Cl(t)]  will still decrease as Cl increases.

   Assuming that the farmer wishes to maximize the present value of his
farm, he will always minimize the  damage that  a given chloride concen-
 tration in his well would cause him. He  could do this by substituting
'^Conceivably  H could also be made a function of t by forecasting future
 changes in land use and commodity prices but this would require many
 additional assumptions and projections and is independent of the main
 problem at hand.

                                 19

-------
crops, importing water, or some other best (least costly) available alter-
native.  Thus,  the farmer's self-interest dictates that he take that com-
bination of avoidance and abatement actions which minimizes the sum of
his avoidance,  abatement, and damage costs.

   The total damages sustained by the farmer for various  chloride con-
centrations at his well are plotted in Figure 5.  These increase at an in-
creasing rate as the chloride concentration increases because increases
in chloride concentration levels when those levels are still low will merely
cause crop yields to decline slightly while subsequent increases at higher
concentration levels will require  high cost responses such as drilling a
new well, importing water, etc.  Thus, the farmer's marginal damage
curve as a function of various chloride concentrations at his well will
also slope upwards as shown in 'Figure 6.  Neither total nor marginal
damages is sustained, however, until a certain minimum threshold con-
centration of chlorides at the farmer's well is reached.

   Most of the  variables and parameters in Equation 1,  from which the
total and marginal damages sustained by the farmer  are calculated, would
not be known to either the farmer or the oil company when the farmer's
well becomes polluted.  Their discovery would require  appropriate mon-
itoring, groundwater quality modeling,  and other information collection.

The Benefits Functions
   While the farmer sustains damages as a result of the oil company's
polluting activity, the oil company derives an advantage, or benefit,  from
engaging in that activity.

   The oil company is assumed to conduct its operations so as  to maxi-
mize its present discounted value (Ga) where

                          t=oo

                    Ga =   /   [R(t) - C(t)] e-rt dt                   (2)
                         t=0

and R(t) is the  revenue obtained in time period  t while  C(t) is  the  inter-
nal costs it incurs  in that time period (i. e. , exclusive of any external
costs it is imposing on the farmer).  In maximizing  its  present discounted
value, the oil company is  generating, however,  a certain concentration
of chlorides  at the farmer's well.  This concentration is labeled R (not
to be confused with R(t) —-the oil company's revenues) in Figure 7.  The
chloride concentration at the farmer's well can be reduced below R by
placing controls on the oil company's activities.  It could, for exam-
ple, line its  brine disposal pit with a nonporous material,  deep-inject its


                                  20

-------
  TOTAL DAMAGES
  TO FARMER, $
                                   TOTAL DAMAGES CURVE
                                                    Cl CONCENTRATION, mg/l
               Figure 5.  Farmer's total damages function.
MARGINAL DAMAGES
TO FARMER, $
                                     MARGINAL DAMAGES CURVE,
                                                  Cl CONCENTRATION, mg/l
           Figure 6.  Farmer's marginal damages function.
                                 21

-------
PRESENT VALUE OF
OIL COMPANY, $
                                                           B, TOTAL BENEFITS
                                                           FUNCTION
                                                       Cl CONCENTRATION AT
                                                       FARMER'S WELL,
               Figure 7.  Oil company's total benefits function.

brine, or reduce the scale of its oil pumping operations. In any event,
reducing the chloride concentration at the farmer's well will increase the
oil company's costs or, possibly,  decrease its revenues.  In either case,
as shown in Figure  7, the company's present value will decrease as the
chloride concentration it generates at the farmer's well decreases.  Since
the oil company is still assumed to maximize its present value subject to
the controls placed  on it,  the oil company will achieve any given level of
pollution at the farmer's well in the least-cost, most efficient way.  If
the cost increases for successive  unit increments  of pollution reduction,
as is likely,  the oil company's  present value curve as a function of the
pollution it generates at the farmer's well will be  shaped as shown in
Figure  7.  If it is very costly for  the oil company  to reduce the pollution
it generates to low levels, in principle there is no reason why the total
benefits curve should not  intersect the horizontal axis rather than, as
shown,  the vertical axis.   The associated marginal benefits curve is  il-
lustrated in Figure  8.  When the oil company takes no account of the pol-
lution it generates,  it will generate pollution until the marginal benefit
it derives  from increasing pollution has fallen to zero.
                                  22

-------
                                                         Cl CONCENTRATION AT
                                                         FARMER'S WELL,
              Figure 8. Oil company's marginal benefits function.

 The Efficient Solution
    The marginal benefits to the oil company and the marginal damages to
 the farmer associated with the oil company generating various different
 steady-state pollution concentrations at the farmer's well are combined
 in Figure 9.  Assuming that resources are efficiently allocated when so-
 ciety's collective wealth is maximized, and ignoring the distribution of
 wealth,  the optimal level of pollution at the farmer's well would be OS,
 where marginal benefits and marginal damages are equal.  This is equiv-
 alent to  maximizing the net wealth of the oil company and the farmer who
 jointly comprise "society" for the purposes of this analysis.   OS is the
 optimal  level of pollution because any increase in  pollution beyond that
 level would impose greater additional damages on the farmer than it would
 generate additional benefits to the oil company, and any decrease in pollu-
 tion below OS would decrease the  benefit to the oil company by more than
 it would decrease the damages to the farmer.  Since the current level of
 pollution (OR) exceeds the optimal level of pollution (OS) then pollution
 is excessive.  If pollution is reduced from  OR to OS,  the gain to the farm-
er in terms of damage avoided is equal to area A plus B while the loss
                                 23

-------
 MARGINAL BENEFITS
 AND DAMAGES, $
             MARGINAL BENEFITS
             FUNCTION
                                                  MARGINAL DAMAGES
                                                  FUNCTION
                                                         CONCENTRATION
                                                         OF Cl, mg/l
 0       T                     S                   R
                       Figure 9.  The efficient sofution.

to the oil  company in terms of benefits foregone is area B.  Thus,  the
net gain to society would be (A + B) - B = A.  The area under the mar-
ginal curve between any two chloride concentration levels (i. e. , the in-
tegral)  is equal to the difference in the total damages and benefits (which
corresponds to that particular marginal) between those same  chloride
concentration levels.


Achieving the Efficient Solution

   Whether the  efficient level of pollution (OS) could  be achieved by the
farmer and the oil company negotiating a voluntary agreement depends
on two factors:  (1) whether or not the two parties have well-defined prop-
 erty rights in the groundwater underlying their land,  and (2) their costs
 of transacting an  agreement.  The transactions costs associated with ne-
 gotiating  an agreement are of two types.  First, there are contract nego-
 tiation  costs proper.   These include a variety of costs such as discovering
 the extent of the harm or benefit which is going to be done as •well as how
 much harm or benefit is to be exchanged.   This can  be defined as moni-
 toring for information in order  to discover the extent of  the harm or bene-
 fit.  Second, there are costs for policing property rights because once
 an agreement is reached, the parties must monitor the groundwater for
 compliance, i. e.  , keep track of what pollutants are  being emitted and  in
                                  24

-------
 what quantities, to ensure that the agreed upon amount of harm or ben-
 efit is exchanged.  These, information and compliance monitoring costs
 are perhaps the most important component of transaction costs.

    To illustrate the influence exerted by these two factors on whether or
 not the optimum level of pollution will be achieved requires examination
 of the case study using  different specifications for property rights and
 transaction costs.  Assume that (1) both the oil company and the farmer
 have well-established property rights in their groundwater,  (2) their costs
 of transacting an agreement are zero, and (3) they pursue a  solution to
 the pollution problem by negotiation.  These assumptions are,  of course,
 unrealistic.  Property rights  to groundwater are often not well defined
 and the costs of transacting (monitoring) a groundwater pollution agree-
 ment would certainly not be zero.   They can, however, be used to dem-
 onstrate that the oil company  and the farmer would come to a negotiated
 agreement specifying OS as the level of pollution. *

    The argument runs as follows.   The area under the marginal benefits
 curve (Figure  9), for a  given  level of pollution,  represents the total gain
 to the oil company.  Similarly, for a given level of pollution  the area un-
 der the marginal damage curve represents the total damage incurred by
 the farmer.  If there is no negotiation the  oil company will operate so  as
 to generate OR because  at that level it will maximize the value of its ben-
 efits in an amount equal to (B  + C 4- D).  However, at  this level of pollu-
 tion the damage to the farmer is (A + B + C).  If the oil company reduces
 the pollution it generates to  OS the damage to the farmer will be reduced
 by  (A + B).  The farmer would, therefore, be prepared to pay up to that
 amount to  reduce the pollution to OS.  At this lower level of pollution the
 loss to the oil  company  (in terms of gain foregone) would be B.   Conse-
 quently, the company would be willing to cut back to OS for payment of
 slightly more than B. Since the farmer  is willing to pay up to (A + B)  for
 a reduction to  OS, and the oil company would be willing to cut to this level
for slightly more than B, the two parties will negotiate to cut the pollution
 generated by the oil company to this optimal level.  The precise amount
the farmer pays to the oil company in excess of the required minimum
 (B) depends on the respective  negotiating abilities of the two parties.   Pol-
lution will  not be reduced below OS because at lower levels the damages
avoided by the  farmer would be less than the gain foregone by the oil com-
pany.   Thus, the farmer's maximum offer for a further reduction in pol-
lution would be rejected by the oil firm.


*This solution  through negotiation was first demonstrated by Coase (I960).
 See Turvey (1963) as well for another informative discussion.
                                 25

-------
   Notwithstanding the existence of an externality, as long as the interact-
ing parties are willing to negotiate and can do so without cost, intervention
by the State is unnecessary to secure an optimum resource allocation.
However, the outcome when the  State does impose on the oil company a
legal obligation to compensate the farmer for any damages it causes him
happens to be precisely the same.   The total gain to the oil company for
being able to generate OS pollution is (D + C), whereas the total damage
to the farmer of OS pollution is only C.  Thus,  if the oil company is le-
gally liable to compensate the farmer, the company would be willing to
pay up to D + C for the privilege of generating OS pollution.  The farmer,
on the other hand, would be willing to accept just C.  The precise pay-
ment in excess of the minimum payment necessary (C) would again de-
pend  on the negotiating skills of  each party.  The oil company will not pay
to increase the level of pollution beyond OS because the marginal damages
inflicted on the farmer that the oil company would become liable for would
exceed the marginal gains to the oil company.

   The assignment of a legal duty to the oil company to compensate the
farmer does not affect the outcome,  but only the distribution of wealth.
With no legal liability for the oil company to compensate the farmer, 'the
farmer will pay the oil company up to (A + B) to reduce the pollution it
generates from OR to  OS; when the oil company does have a legal duty to
compensate the farmer,  the oil  company will pay up to (D + C) to gener-
ate OS pollution rather than no pollution.

   Although the outcome under zero transactions costs is interesting to
consider, it is not realistic because transactions costs are not zero.
These costs may be quite substantial,  not only because of legal fees and
negotiating costs, but because of the limited state of present knowledge.
In the present  example the only  information now known by the farmer is
that his  well is polluted with brine water.  He does not know whether the
cause is  seawater intrusion, the polluting activity of the oil company,  or
some other cause,  and neither party knows the shape of the damage func-
tion.  Faced with these uncertainties, and the reality that the well is con-
taminated, the farmer must gather information to  (1) determine the cause
of such damages, and (2) determine the probable extent of such damages.
 To accomplish each task the farmer must gather information through some
process  of monitoring,  modeling,  and information collection—a process
 that is certain to be very costly.  Once this preliminary information is
 collected the farmer can confront the oil company with a claim for esti-
mated damages plus a cease and desist order. The probable result is  a
 costly legal battle involving additional monitoring expenses to substantiate
 conflicting legal claims.
                                  26

-------
     Since legal, negotiation, and monitoring costs may be substantial,
  transactions costs will not be zero.   In such circumstances,  the above
  conclusions cannot be guaranteed to hold  and the solution when transac-
  tions costs are positive must be examined. Recall that the oil company
  is now generating pollution at the level OR and the farmer is being dam-
  aged in an amount (A + B + C).  The net gain to the farmer (and society)
  of reducing the pollution to OS is equal to  A.   However,  if the transac-
  tions costs to the farmer of arriving at such an agreement (Tg) exceeds
  A,  he  will not attempt to negotiate and in  the absence of any further ini-
  tiative on his  part, the oil company will continue to pollute at level OR.
  However,  if there is an enforceable  legal burden on the  oil company,  the
  farmer can be expected to take action by filing suit against the oil com-
 pany for damages and an injunction.   Obviously,  the oil  company will
 prefer to avoid the injunction and negotiate an agreement with the farmer
 to pollute at level OS. However, if the transactions costs to the oil com-
 pany of arriving at such an agreement (T^) exceeds D, the firm will not
 attempt to  negotiate and must therefore suffer the injunction.

    Thus, if the oil company is liable for damages, the firm will generate
 zero pollution whenever the cost of reaching agreement (T^) exceeds its
 net  gain (D) from reaching the agreement; similarly, if the company  is
 not  liable for the damage it causes, it will generate OR pollution whenever
 the  farmer's cost of reaching agreement (Tg)  exceeds his net gain (A)
 from reaching  such an agreement.  Hence, large transactions costs are
 likely to prevent negotiating the socially optimal level of pollution,  and
 lead to an inefficient  allocation of resources.  When,  however, these
 transactions costs do not exceed the potential net gain, an agreement
 will be reached to operate at OS.  If the oil company is liable its gain
 will be (D - TA) instead of D, and if the company is not liable the farm-
 er's  gain will be (A - Tg) instead of A.  In this situation, positive trans-
 actions costs do not change the optimal level of pollution from OS—they
 simply reduce the gains which result from polluting at this level.

   If transactions costs are large enough to cause a suboptimal level of
 pollution other than OS,  the situation may be remedied by government
 intervention.  For example, government can induce a shift to the optimal
 level by imposing an effluent charge on the oil  company,  equal to SU per
 unit  of pollution, or invoke other policies such as permits to achieve the
 same result.  The effluent charge will induce the  oil company to cut back
 its pollution from OR to  OS because, for all levels of pollution greater
than OS, the marginal gain is less than the marginal cost per unit of
pollution.
                                 27

-------
   Thus,  government has the power to induce an efficient allocation c
resources.  However, government must estimate the marginal gain <.
damage curves before it can know the optimum, level of pollution and
termine the appropriate charge,  again involving a great deal of costl;
monitoring and modeling.  The cost to government of ascertaining th:
information and enforcing the implementation of any such policy will
probably be at least as large as it would be to the principals  involved
there seems to be no  a priori reason that  government could reduce rt
toring costs.


Two Limiting Cases

   The analysis presented in this section  so far has dealt with the ge:
eral case. However, it is equally  applicable  to two limiting  cases th
are of practical interest.  First, it is quite possible that the efficien
solution is zero pollution.  That is the result  when the marginal dam
function exceeds the marginal benefits function at all levels of polluti
as in Figure 10.   In practice, it  is possible that such might be  the  si
tion where the polluter is emitting  carcinogenic materials or heavy r
als into a community's drinking water supply.  Second,  it is  also pos
ble that the efficient solution is to impose no standard at all.  That ii
implication if the marginal damage function is everywhere zero and t
      MARGINAL BENEFITS
      AND DAMAGES, S
                                  MARGINAL DAMAGES
                                  FUNCTION
                                   MARGINAL BENEFITS
                                  .FUNCTION
                                                       CONCENTRATION
                                                       OF Cl, mg/l
    0=S
                     Figure 10.  First limiting case.
                                  28

-------
contiguous with the horizontal axis as in Figure 11.  In practice, it is
possible that such might be the situation where previous pollution has
already degraded the quality of an aquifer to such an extent that it is in
effect a "sink" and,  as a result,  additional effluents will not produce
additional damages.

   With these principles concerning the establishment of the optimal
level of groundwater quality as background, Section IV considers the
issues  specifically related to monitoring.
    MARGINAL BENEFITS
    AND DAMAGES, $
                 MARGINAL BENEFITS
                 FUNCTION
                                                   •MARGINAL DAMAGES
                                                    FUNCTION
                                                       CONCENTRATION
                                                       OF Cl, mg/l
                                              R=S
                    Figure 11.  Second limiting case,
                                 29

-------
                           SECTION IV
       MONITORING AND  GROUNDWATER QUALITY STANDARDS
   The primary purpose of this  section is to analyze in the context of the
framework developed in Section III the problem of selecting a preferred,
or optimal,  monitoring strategy under the assumption that the govern-
ment intends to  combat the excessive pollution of ground-water by (1) set-
ting, and (2) enforcing groundwater quality standards by some form of
direct  regulations involving discharge licenses  or permits, although a
discharge license or permit procedure is not the only (or even necessan y
the best) way in which groundwater quality  standards can be attained.  I*1
principle, they  can also be achieved by an effluent charges  scheme or by
paying subsidies for effluent reductions.  The problem is analyzed in the
context of a discharge licensing scheme because that is the option which
has been adopted by EPA (USEPA, 1974).

   A secondary purpose is to discuss important institutional considera-
tions associated with implementing the preferred strategy selected, such
as who should do the monitoring and who should pay for it.  In general,
these institutional considerations are embedded in the analytical discus-
sion although some are discussed  separately.


DEFINITIONS OF INFORMATION AND
COMPLIANCE MONITORING
    From an analytical point of view, it is convenient to divide monitoring
into two components—monitoring for information and monitoring for com-
pliance.  In essence, monitoring for information is monitoring which is
undertaken for  the express purpose of estimating the marginal  gain and,
 especially, the marginal damage function.  Monitoring for this informa-
tion is prerequisite to  setting an optimal groundwater quality standard
 although groundwater quality standards may legitimately be set without
 any assurance  that they are optimal for reasons that are explained sub-
 sequently.

    Monitoring for compliance is that monitoring which is undertaken to
 insure that the groundwater quality standard that has been established
 is,  in fact,  adhered to.  Since  monitoring for information and monitor-
 ing for compliance  are analytically distinct, they are discussed separately*
                                  30

-------
  INFORMATION MONITORING

    Information monitoring is designed to specify the damage function and
  inform policymakers of the damages which will be avoided if the standard
  is set at different levels.

    The damage function (Equation 1) contains two terms; namely, H,  the
  highest valued use of the farmer's land, and K[Cl(t)],  the value of the
  farmer's land at time (tg) given some time profile of the chloride concen-
  tration at his well over the period tg to t<».  Since variable (H) is assumed
  to be known,  only the determination of K[Cl(t)]is related to monitoring.
  To specify K.[d(t)]it is necessary to (1) estimate the reduction in crop
 yields associated with different chloride concentrations and (2) estimate
 the chloride  concentration profile through time for different pollution
 avoidance —abatement strategies.  The function relating crop yields to
 chloride concentration levels is a problem in agronomics, not monitor-
 ing,  so it need not be dealt with here.  Of interest is the estimation of
 the chloride  concentration profile through time for different pollution
 avoidance—abatement strategies which involves three activities:  data
 collection, groundwater flow modeling,  and monitoring proper.

    The place to start will always  be with the estimation of the chloride
 concentration profile in the absence of any pollution avoidance—abatement
 action—i. e, , with the estimation  of the solid curve labeled Clj in Figure 4,

    To predict Cli (i. e. , the time-profile of chloride  concentrations at
 the farmer's  well between discovery (t0) and the steady state (tg)in the
 absence of any pollution abatement—avoidance steps being taken) it is
 essential to employ some  form of  groundwater quality modeling technique.
 In principle,  this modeling process  involves specifying  the parameters
 of a simulation model which is then  used  to estimate the properties of
 the pollution plume over time.  However,  in practice, groundwater qual-
 ity models vary from simple extrapolations based on  experience  and ex-
 pertise to  sophisticated computerized simulation models.  Naturally,  the
 degree of effort expended on the flow model will depend  on the serious-
 ness of the threat posed by the pollution situation,  but any technique of
 quality modeling used will require input data.  These data must either
 be collected from existing sources or obtained by monitoring.  Apart
from being a  source of input data  for the  model, monitoring must also
be employed to check the accuracy of the  groundwater flow model.  The
sort of data that would have to be collected or generated in order to

-------
simulate Clj with any degree of accuracy using a sophisticated ground-
water quality model could include:*
      • Data describing the geological column at both the oil com-
        pany brine disposal pit and  at the farmer's water well

      • Brine pumping rates and brine concentration levels at
        the pit
      • Net groundwater recharge and water table levels during
        the period in which the pit has been in existence
      • Chloride concentration levels at the water well

      • Past hydraulic head pressures
      • Past rates of water withdrawal at the farmer's well
      • The porosity, hydraulic  conductivity, and saturated
        thickness of the aquifer
      • The mass per unit volume of the solution in the aquifer
        under ambient conditions

      • Relevant dispersion and diffusion coefficients

      • Geographical data such as contour maps or simple topo-
        graphical maps
      • Statistics describing land use in the immediate vicinity.

   To the extent that these data are not already available, it would be
necessary to determine exactly what needed data are missing, how to
procure them, and their acquisition cost.  Since passage of PL 92-500
much of this information will be available at a lower cost than it would
have been prior to the passage of that legislation because Sec. 308(a)
states that "The Administrator .  .  .  may at reasonable times have ac-
cess to and copy any records. "  Thus,  under the new law many of the
data required are more easily accessible.  In fact, prior to the passage
of PL 92-500 the  polluting agent almost certainly would have made infor-
mation to which it was privy available only after legal proceedings had
been implemented and the process of discovery instituted.

   Having predicted Clj through a process of data collection, modeling,
 and monitoring and, consequently,  being able to estimate the level of
 damages associated with the oil company's activities in the absence of
 any abatement— avoidance actions,  the next step would be to predict the
 time-profiles of chloride concentrations associated with different abate-
 ment—avoidance  strategies using the groundwater flow model.  The first
 *See Finder (1973).  In this article Finder simulates the areal distribu-
  tion through time of groundwater contaminated by cadmium originating
  at an  aircraft plant on Long Island.
                                  32

-------
   obvious prediction would be that of C12,  the time-profile of chloride con-
   centration at the farmer's well if a "perfect" abatement technique such
   as deep-injection of its brine is adopted  by the oil company at to (see
   Figure 4).  Using this  information, it would be possible to assess  the
   damages which would be experienced by  the farmer even though a perfect
   abatement technique were adopted immediately.

     Analogously, other time-profiles of chloride concentrations at the
  farmer's well between  Clj and C12 would have to be predicted  using the
  groundwater quality model for less  than perfect abatement—avoidance
  strategies to obtain a reasonably complete picture of the overall marginal
  damage function.  However,  the damages associated with Clj (no abate-
  ment—avoidance) and C\2 (perfect abatement—avoidance) are most cru-
  cial.  If the difference between these two damage estimates is  greater
  than the cost of a perfect abatement—avoidance technique, then the per-
  fect abatement—avoidance technique should be adopted because its cost
  would be less  than the damages which would be avoided (if the marginal
  gain and marginal damage functions are not extreme).  All other informa-
  tion concerning the marginal damage function would be redundant.   This
  possibility corresponds to the first special  case discussed at the end  of
  Section III,  where the marginal damage curve  exceeds  the marginal curve
  for all pollution levels.   (As an aside, most discussions of pollution dam-
  age occur in this context. That is, pollution damages are estimated in
  terms of the existing level of pollution with the implication that these
  damages could be avoided by reducing pollution to zero.  (See for exam-
  ple,  U.S. Council on Environmental Quality (1971),  Chapter 4.)

    "Monitoring for information" is, then,  an amalgam of data collection,
 quality modeling, and monitoring used to predict the properties of the
 pollution plume and estimate the damage function associated with the
 source of pollution.

    Although groundwater quality modeling  is in its infancy, the discipline
 is continuously advancing and such models are proving increasingly  use-
 ful for  making predictions, especially in situations involving point sources
 of pollution.   However,  as of now, such modeling is extremely costly be-
 cause to be reliable it requires the procurement of a great deal  of data
 and the services of highly skilled experts in  the areas of hydrology,  geol-
 ogy* programming, etc.   Consequently, the  possibility arises that the
 cost of acquiring the information on the damage function (Mj) may ex-
 ceed the cost of the damages which can be  avoided by having that infor-
mation.   This,  of course, poses a real dilemma.  However, through the
accumulation  of experience concerning recurrent pollution situations and
 reductions in  the real cost of modeling that can be expected with the pas-
sage of time,  such uncertainty can be expected to diminish.
                                 33

-------
   The problems associated with the estimation of damage functions of
groundwater pollution are not peculiar to that particular resource.   Nu-
merous references attest to the  difficulty of obtaining damage functions
for air pollution, ozone,  and surface water discharges.  See, for exam-
ple, Barrett and Waddell (1973), pp vii, 4,  24,  and 28; Blower (1973);
Ridker (1967), p 153; and U.S.  Council on Environmental Quality (1971),
pp 107 and 135.  These references  illustrate an official,  and semiofficial,
awareness of the difficulty of estimating damage functions associated
with other environmental resources analogous to the difficulty of estimat-
ing damage functions associated with groundwater pollution.


COMPLIANCE  MONITORING
Permit Terms
   This section assumes that the data  collection,  quality modeling,  and
information monitoring prerequisite to the estimation of the marginal
benefits and marginal damage functions illustrated in Figure 9 have been
accomplished and that the standard-setting authority has  established an
optimal level of pollution of OS mg/1 of Cl at the farmer's well.  This
level of pollution is associated with adoption of a given abatement tech-
nique by the oil company which, since OS is positive, is not perfect.
Some residual amount of chloride  still finds its way into the  groundwater
at the source, and chloride concentration at the farmer's well is a phys-
ical relationship established through the groundwater quality model.
Similarly, the model implies that  if the residual percolation of brine at
the source is, say, X gallons  per  year (gpy) and that this will generate
OS'mg/1 of Cl at the  farmer's  well, it will also generate OS'  mg/1 at some
point between the source and the farmer's well—for example, at the bound-
ary of the oil company's property.

    From this it follows that a discharge permit issued to the oil company
that is designed to prevent violation of the OS level of pollution at the
farmer's well may have several dimensions.  It may require that (1) the
oil company initiate  the appropriate abatement technique, (2) the oil com-
pany not  allow more than X gpy of  residual brine to percolate to the ground-
water,  and (3) the oil company not allow the chloride at its property line
to  exceed OS'.  In fact NPDES permits are frequently structured in this
way.  See, for example,  Application for NPDES Permit to Discharge
 Treated Waste-Water to U. S. Waters. U. S.  Environmental  Protection
Agency,  Region IV Application No. AL 074 OYM 2  000570.   When a per-
mit is so structured, the permit dictates where monitoring may be under-
taken in order  to insure compliance with its terms.  This may involve
monitoring of all three of the restrictions listed  above.   If these permit
terms are not violated,  then presumably the optimal pollution level of OS
at the farmer's well is not violated either.
                                   34

-------
 The Basic Principles

    The example discussed in Section III illustrates the principles involved,
 the essential elements of which are shown in Figure 9 (see page 24).  It
 will be recalled that if the oil company is not induced to recognize the
 harm done by the chloride emissions it generates, it will operate so that
 the chloride concentration its activities produce at the farmer's well is
 equal to OR in Figure 9.

    If a permitting agency issues a permit to the oil company to limit the
 chloride concentration level at the farmer's well to  the optimal level of
 OS by specifying given abatement techniques,  each of those conditions
 imposed on the  oil company implies an associated concentration limit at
 the farmer's well.  If the oil  company meets the standard,  the gain to so-
 ciety is equal to area A minus the transactions costs (i. e. , monitoring
 and other enforcement costs) that must be incurred in order to deter-
 mine,  and to obtain compliance with, the optimal standard OS regardless
 of who does the  monitoring.  The cost to the oil company of compliance
 with the standard OS is equal to B.  Conversely, the gain to the oil com-
 pany of noncompliance is B while the cost to society would be A + M, the
 transactions costs.  (M might be zero  if no  effort were made to enforce
 compliance.  However, M might be positive if some transactions (moni-
 toring) costs are incurred but compliance is still not achieved. )  In  this
 case it is  assumed that the permittee oil company will comply  with the
 terms of its permit only if the expected cost of noncompliance  exceeds the
 cost of compliance and will not  comply if the expected cost of noncompli-
 ance is less than the cost of compliance,  although it is  recognized that
 not all permittees  are entirely selfish and thus will not automatically vio-
 late the terms of their permits for economic reasons.  The problem in
 such circumstances is analyzed subsequently in the subsection  Permittee's
 Optimal Violation.  It transpires that no substantive modification of  the
 analysis is required.

 The Expected Cost  of q Standards Violation
   This subsection treats criminal behavior as a rational  reaction to the
 costs and benefits that the criminal  subjectively estimates to be associated
with his decisions,  not as some form of social or psychological deviancy.
It is in the vein,  therefore,  of the modern economic approach to criminal
activity originally developed by Becker (1968 and  1974) and pursued  by
Stigler  (1970) and Ehrlich (1973  and  1975).

   The expected  cost to a  permittee  of violating a standard is equal to

                       E(C) = Pd • Pc  ' E(F)                     (3)
                                 35

-------
where PJ is the probability of detecting violation of the standard.  Pc is
the probability of conviction for the violation, given its detection,  and
E(F) is the expected fine,  given conviction.  A few observations are in
order.  First, since E(C) is multiplicative in its arguments, if Pj,  Pc»
or  E(F) is zero,  then  E(C) will be zero.  A violation will therefore occur.
Second,  the arguments of Equation 3 can be traded off against each other.
Any given expected cost level can be maintained by increasing one argu-
ment and appropriately decreasing another argument.  Third, even  if the
probabilities of detection and conviction are one (unity), the expected cost
of violating the standard could still be less than the expected gain from
such a violation if the expected fine is small enough.

    It is apparent that the EPA's water quality strategists are cognizant
of this problem since  they have stated that the "EPA considers the mag-
nitudes of the fines that it intends applying as a major economic lever to
assure compliance. " (USEPA,  1974). Frequently in the past, it was cheaper
to pay a small fine than to install and operate control equipment.  Now
PL 92-500 (Section 309(c)(l))  provides that "Any person who willfully or
negligently violates .  . .  any permit condition or limitation .  . . shall be
punished by a fine of not less than $2500 nor more than  $25,000 per day
of violation .  . .  . "

    Section 1423,  (b)(2) of the Safe Drinking Water Act provides for simi-
larly stiff penalties.   "Any person who violates any requirement of  an ap-
plicable underground injection control program to  which he is subject . . •
(A) shall be subject to a civil penalty of not more than $5,000 ...  or  (B)
if such violation is willful, such  person may, . .  .  , be fined not more
than $10,000  for  each day of such violation. "

    Fines in the upper ranges of those allowed by PL 92-500 and PL 93-523
have the potential to make the expected cost of a standards violation high
ji the probabilities of detection and conviction are not trivially small.  How-
ever,  as reported, the EPA's administration of the Law has the effect of
decreasing the severity of a fine. According to R. Johnson, Acting EPA
Administrator for Enforcement,  in a speech before the Practicing Law In-
 stitute on February 14, 1975, that appeared in the February 21,  1975,
 issue of Environmental Reporter, the EPA intends to follow generally a.
 four-step approach in enforcing  compliance schedules set forth in the Na-
 tional Pollution Discharge Elimination System  permits.  First, the EPA
 will telephone a discharger found to be in violation of the compliance sched-
 ule written into his permit.  Next, EPA will send  a letter to the discharger*
 and this  will  be followed by an administrative  order if noncompliance per-
 sists.  The final step, according to Johnson, would be court action; only iH.
 serious violations will EPA bypass these  steps and go directly to  court.
 Clearly, if this procedure is followed its  de facto  effect is to reduce any
 expected fine below the de jure level of fines.  (Emphasis added. )
                                   36

-------
    The probability of conviction (given detection) is treated as an exoge-
 nous variable in this report since to explore the reasons why detection of
 an offense sometimes does, and sometimes does not,  lead to a conviction
 would be tributary to the main purpose of this report.  For concreteness,
 and without loss of generality,  the specific exogenous value given as Pc
 is  assumed to be one (unity).

 The Probability of Detection Function
    The probability of being detected violating a groundwater quality stand-
 ard is a function of the amount  of compliance monitoring undertaken—as
 measured, say, by expenditure on compliance monitoring—and the sever-
 ity of the violation.  Thus,  the probability of detection function is

                             Pd=f(Mc,V)                        (4)

 where MC is the expenditure on compliance monitoring and V is the size
 of the violation as measured by the difference between the actual level of
 pollution emitted and that allowed by the  standard.  In addition, it is as-
 sumed that
while
                       SPd
                           >  0 and
   The positive first derivative with respect to Mc for this probability-
of-detection function indicates  that Pd increases monotonically with MC
while the negative second derivative indicates  that it does so at a decreas-
ing rate.   This is based on the presumption of diminishing returns to moni-
toring. Mc might be zero if no effort were made to enforce compliance,
or positive if some monitoring costs are incurred but compliance  is still
not achieved.

   Graphically, Pd as a function of MC would appear as in Figure  12.
The  curve increases monotonically and asymptotically approaches unity.
The  curve is drawn for a given size of violation.  The solid curve  labeled
Pd is predicated on the assumption that the size of the violation is equal
to Vj.  If the size of the violation increased to V2, then the whole  curve
would  shift up to Pd' because the first derivative of Pd with respect to
V is  positive.  Figure 12 also assumes that any given level of compliance
                                 37

-------
  PROBABILITY OF
  DETECTION,
                                                 COMPLIANCE MONITORING
            Figure 12.  The probability of detection function.

monitoring  expenditure which is undertaken is allocated over the monitor-
ing techniques relevant to the pollution situation in such a way as  to max-
imize the probability of detection for that level of expenditure.  The prin-
ciples governing this allocation are discussed in the following subsection.

   The dependence of Pjj on V can be clarified by reference to Figures
13 and  14.  In Figure 13,  if the permittee operated so as to generate a
level of pollutants equal to OU, he would be guilty of a violation of the
standard by an amount of  (OU-OS) = V.  The gain  to the permittee of such
a marginal violation is equal to Dj.

   Figure 14 can be used  to explain that, for a given monitoring expendi-
ture, the probability of detection of a violation of the standard will increase
as the  severity of the violation V increases (i. e.  , dP
-------
  MARGINAL BENEFITS
  FUNCTION
                                      MARGINAL DAMAGES
                                      FUNCTION
                                                   POLLUTION LEVEL
                                                   (mg/l OF Cl)
 Figure 13.  The gain from a standards violation.
Figure 14.  Detection and the size of the violation.
                       39

-------
pollution part of the plume is between Wells A and B.  Wells A and B are
yielding readings  of 250 mg/1,  and thus do not indicate that the permittee
is violating the standard with a concentration of 500 mg/1 over a large
area beyond his property boundary.  If he increases the  severity of his
violation by emitting more pollutants from  P,  then the plume will expand,
the 300 mg/1 isopollution contour will arrive at Well A or B, and a vio-
lation will be detected.  Thus,  the probability of a violation being detected
increases  as the severity of the violation itself increases.

   There seems to be no good reason to suppose that this probability of
detection would increase at either  an increasing or decreasing rate as the
severity of the violation increases.  Consequently, the second derivative,
d2P(j/dV2, is postulated to be zero.   (Nothing substantive in the analysis
which follows would be affected if the second derivative  were either  posi-
tive or negative.  Only the indisputably positive first derivative is essen-
tial to the analysis. )  Since the probability of detection would quite obvi-
ously be zero when V is zero,  it may be  inferred that the probability of
detection will increase in proportion to the size of the violation.  Graph-
ically,  therefore, when P
-------
    PROBABILITY OF
    DETECTION, Pd
     1.00
                                                                SEVERITY OF
                                                                VIOLATION, V
                                                                (mg/l OF Cl)
      Figure 15. The probability of detection and the severity of the violation.

 various  locations and various frequencies.   The feasible set of monitor-
 ing strategies consists of all possible combinations of monitoring tech-
 niques applied at all possible intensities.

    Associated with each monitoring strategy within the feasible set of
 strategies is  a certain probability of detection.  The "output" of differ-
 ent quantities of compliance monitoring undertaken is different proba-
 bilities of detection.   In order to maximize the probability of detection
 from a given monitoring budget,  the efficient monitoring strategy to adopt
 is that associated with an allocation of the budget such that the increase
 in the probability of detection per dollar allocated to  one monitoring tech-
 nique is  equal to the increase in the probability of detection per dollar
 allocated to any other feasible monitoring technique.   (For any given
 probability of detection,  following this rule will lead  to a minimization
 of monitoring expenditures. )

   If such a budget allocation is followed and  there are diminishing re-
 turns to compliance monitoring, then the marginal cost of such monitor-
 ing as a function of the probability of detection will increase.  Graph-
ically,  the marginal cost of monitoring as a function of the probability of
detection will  appear  as in Figure 16.
                                  41

-------
MARGINAL COST, $
                     MARGINAL COST OF
                     COMPLIANCE MONITORING,
                                                           PROBABILITY OF
                                                           DETECTION, Pd
                                                    1.00
          Figure 16.  The marginal cost of compliance monitoring.

Implications of Independent versus Audited
Self-Monitoring
   The probability-of-detection function is independent of who does the
monitoring, i. e. , whether monitoring is done by some independent age
or whether self-monitoring subject to independent audit is the adopted
practice.

   By definition,  independent monitoring is monitoring done by anyone
but the permittee.  Usually, it would be done by the permit issuing agency*
but it could be delegated to some other body  (except the permittee).  °e
monitoring that is subject to audit is monitoring generally done by the
permittee but where the permit-is suing agency reserves the right to con
duct  audit monitoring to ensure the integrity  of the data collected.  Both
the Federal Water  Pollution Control Act and the Safe Drin^np; Water^Act.
 envisage self-monitoring backed up by audit  monitoring.  For example,
 Section 308(a)(4) of PL, 92-500 states:

       (A) The Administrator shall require the owner or operator
      of any point source to (i) establish and maintain such records,
       (ii) make such reports,  (Hi) install, use, and maintain such
      monitoring equipment or methods .  .  .  , (iv) sample such
       effluents (in  accordance with such methods, at such locations,
                                   42

-------
        at such intervals, and in such manner as the Administrator
        shall prescribe) . . .  ; and (B) the Administrator .  . . (i)
        shall have a right of entry to ... any premises in which an
        effluent source is located. . , , and (ii) . .  .  have access to
        records,  .  . .  monitoring equipment . .  .  , and sample any
        effluents . , .

     Analogously,  Section 1445 of the Safe Drinking Water Act states;

        (a) Every person who is  a supplier of water .  . . shall es-
        tablish and maintain such records, make such reports, con-
        duct such monitoring,  and provide such information as the
        Administrator may reasonably require . . . (b)(l) .  . . the
        Administrator ...  is authorized to enter any .  . . property
        ... [in order to audit]  records, files, papers, processes,
        controls,  and facilities .  . .

     The functional relationship between the probability of detection of a
 standards violation and the amount of monitoring  undertaken when the
 party to be monitored does the basic monitoring,  but is subject to audit,
 does not differ substantively  from that of independent monitoring.  This
 can be illustrated by the isopollution contour and monitoring well situa-
 tion of Figure  14.  The readings of wells A and B do not indicate to the
 permitting authority that a violation of the standard is occurring because
 a chloride concentration of 500 mg/1 exists over a large area outside the
 permittee's property boundary.  Clearly, the probability of this violation
 being detected  would increase if additional expenditures on monitoring
 were incurred  to sink more wells, especially between A and B.

    If self-monitoring subject  to audit is substituted in this situation, the
 permit-issuing agency would make as a condition of its permit issuance
 that the permittee himself sink wells at A,  B, and C and report the re-
 sults of samples  taken therefrom to the agency.  The integrity of their
 results would be  preserved by the agency's random audit activities.  How-
 ever,  once again, the standards  violation which is occurring would not be
 detected by samples from A and  B,  and the probability of such detection
 would  increase if the agency imposed on the permittee the obligation to
 sink more monitoring wells along its boundaries.

   If the monitoring burden is placed on the permittee, the permit-issuing
 agency may be tempted  to regard monitoring as a free good since it does
not come out of its budget.  This distinct possibility of a temptation to
impose "over-monitoring" when self-monitoring subject to  audit is the
chosen modus operandi  arises because of the incentive system inherent
in the institutional structure of the situation.  However, the objective of
                                 43

-------
compliance monitoring should not, in general, be to maximize the proba-
bility of detecting a standards violation.  To make this probability arbi-
trarily  close to unity, monitoring wells would have to be sunk arbitrarily
close together along the permittee's property line at great expense.   Thus,
the gain to society in enforcing the standard would, if the objective of max-
imizing the probability of detecting a standards violation were adopted, be
dissipated in monitoring  expenditure.  If wealth maximization is society s
goal, monitoring should be undertaken to  ensure compliance with a stand-
ard only to the extent that the gain to society (which has, in terms of Fig-
ure 9,  a maximum value of $A) is in excess of the monitoring expenditures*
($M) necessary to  enforce compliance.  However, the permit-issuing agency
may perceive its function as maximizing the number of  standards viola-
tions detected since these are immediately visible, whereas the more re-
mote, but appropriate, goal of an overall improvement  in society's wel-
fare would not be so obvious.  While some constraint would be imposed on
the monitoring activities undertaken by the permit-issuing agency if they
had to be financed  from its own budget, no such constraint exists when they
are financed by the permittee.  This argument is not, of course, suffi-
cient to justify independent monitoring to the  exclusion of self-monitoring
subject to audit because there are other reasons why the latter might be
advocated. The most important of these  is that self-monitoring subject
to audit may be the least expensive per unit of monitoring.

   Although it is not reasonable from society's point of  view to equate the
detection of a violation obtained by monitoring with the benefit of that
monitoring program, others  have erroneously done so.   Ward and Vander-
holm (1973),  p 543, for example,  state that "Figure 2 illustrates the rela-
tionship between effort expended  [annual  cost of monitoring]  and benefits
gained [percentage of spills detected] . . . . "  Moreover, and highly sig-
nificant given the above  strictures,  they  continue, "the information pre-
sented [in Figure  2] answered the water-quality manager's question, ho
much effectiveness am I [sic] getting for my money. "

 Expected Fine Function
    This section previously noted  that PL 92-500 and PL 93-523 contain
statutory scales of fines for  water-quality violations.   These scales of
fines are exogenous variables whose upper bound is  indicated in Figure
 17 by Fmax .  However  the expected fine is not an exogenous variable,
but a function of the severity of the  violation  as is apparent from the re-
marks of Richard Johnson, Acting EPA Administrator for Enforcement.
 Therefore, the following expected fine function is postulated:

                             E(F) = J(V)                           (5)
                                  44

-------
                                                            SEVERITY OF
                                                            VIOLATION, V
                                                            (mg/l OF Cl)
         Figure 17.  The expected fine and the severity of the violation.
 with
                    dCE(F)]
d2[E(F)]
                                             >o .
 The signs of these derivatives are imposed to indicate what might be as-
 sumed to be plausible judicial attitudes; namely, that the size of the fine
 will be increased as the severity of the violation increases and at an in-
 creasing rate.  (Less confidence can be placed in the sign asserted for
 the second derivative, but it is not crucial to the analysis. )

   The expected fine as a function of the severity of the violation is illus-
trated graphically in Figure 17.   The maximum violation that the permit-
tee will perpetrate is SR.  The courts may, or may not (as, in fact, illus-
trated),  impose the maximum fine before this violation is reached.
                                  45

-------
Expected Cost of Violating a Standard, Compliance
Monitoring Expenditures, and Size of Violation
   By substituting Equations 4 and 5 into Equation 3 we find that:

                     E(C) =  f(Mc, V) • Pc -  j(V)                (6)

                         =  f(Mc, V) • j(V)

since the probability of conviction is assumed equal to unity.  The func-
tion of Equation 6 relating the expected cost of violating a standard to the
expenditure on monitoring for a given size violation (V) is  simply a scalar
transform of the function relating the probability of detection to the ex-
penditure on monitoring (Figure 12).  Thus, it will share the properties
of increasing monotonically at a decreasing rate while asymptotically
approaching the expected fine (E(Fj))  for that size violation.  That is,
                                and         < o
Such a function is illustrated by the solid curve in Figure 18.  The broken
curve indicates the situation if the  size of the violation is larger.

   The function relating the expected cost of violating a standard to the
degree,  or severity, of that violation (for a given level of expenditure on
compliance monitoring) will emanate from the origin, be convex from be-
low, and terminate at Fmax (the maximum fine) — as shown  by the solid
curve in Figure  19.  The  shape of this curve is dictated by  the shape of
the function of Equation 5 since the probability of detection  is  assumed to
be proportional to the size of the violation (see Figure 15).   The broken^
curve in Figure  19 indicates the situation if the level of compliance moni-
toring undertaken was increased from, say, M* to M2..

Permittee's  Optima! Violation
   The permittee's total benefits function with respect to the chloride con-
centrations imposed on the farmer's well is discussed in Section III and
illustrated in Figure 7.   That part of the total benefits function which ex-
ceeds  OS chloride concentrations (the standard) is shown in Figure 20.
It increases monotonically between S and R where it  reaches a maximum,
which is the chloride level the permittee will  generate if no attempt is
made to regulate his activities.  Thus,  algebraically, the permittee's
total benefit as a function of the size of his violation  of the  standard is
given by
                                   46

-------
EXPECTED COST
E(C), $

 E(F,)
                            'E(C)' FOR V=V,
                                                        COMPLIANCE MONITORING
                                                        EXPENDITURE, $
  Figure 18.  The expected fine and expenditure on compliance monitoring.
                                                             SEVERITY OF
                                                             VIOLATION, V
                                                             (mg/l OF Cl)
     Figure 19.  The expected cost and the severity of the violation.
                                  47

-------
                B  =  g(V)  ;  -f£ >0  ,  2-5- <0   .              (7)
   Figure 20 also includes the permittee's expected cost as a function of
the size of his  violation for a given scale of fines and level of monitoring
(taken from Figure  19).

   The permittee's  total benefit and expected cost functions can be used
to establish the level of emissions that it is optimal for the permittee
to generate and, therefore,  the optimal severity of the standard violation
for the permittee to undertake.  The assumed objective is  to choose V so
as to maximize net  expected gain (H) where H is defined to be the differ-
ence between total benefits and expected cost.  That is, the permittee is
to maximize H = B  - E(C).  Diagrammatically,  this occurs at Vj_  where
the slopes of the total benefit and expected cost curves are equal.  Ana-
lytically, the maximum occurs when the marginal benefit from increasing
the violation is equal to the marginal expected cost from increasing the
violation.

   Up to this point the  analysis assumes that the sole objective of the per"
mittee is to maximize  the permittee's net expected gain (even though that
involves illegal activities).  On this basis Vj_ in Figure 20 is the permit-
tee's optimal violation in  that it maximizes net expected gain after allow-
ance is made for the expected cost of the standards violation.  However,
the permittee may also derive disutility from engaging in illegal activities
(i. e. , violating a legally promulgated groundwater quality standard),  which
introduces the consumption of nonpecuniary goods into the utility  function.
(This analytical innovation was pioneered by Becker (1957).  In addition,
see Williamson  (1963 and 1964),  and Alchian (1965).)  Essentially, the
permittee has  indifference curves like those illustrated in Figure 21,
which show a diminishing marginal rate of substitution between the net
expected gain  and illegal  activities.  113 is preferred to U2 and 1^2 to
because for a  given net expected gain  equal to IIj a larger violation
must leave the permittee with a lower  level of utility than a smaller vio-
lation (Vj).  Analogously, for a given violation (V^) a higher level of net
 expected gain  dI2)  must generate a higher level of utility for the manager
than a smaller level of wealth (II j).  Completely unscrupulous permittees
would have horizontal indifference curves; completely scrupulous permit-
tees would have vertical indifference curves.

    The permittee's opportunity set under various conditions is illustrated
 in Figure 22.   The  XX curve  illustrates the net expected gain from the per*
mittee1 s activity as a  function of the pollution generated when no  ground-
water quality  standards must be met.  Wealth is maximized at II   where
the level of pollution generated is  R0   The effect of the imposition of
                                   48

-------
TOTAL BENEFIT AND
EXPECTED COST OF
VIOLATION, S
 max
                                                        TOTAL BENEFIT, B
                                                          SEVERITY OF
                                                          VIOLATION, V
                                                          (mg/l OF Cl)
                 Figure 20.  The optimum violation.
        IT,  NET
        EXPECTED
        GAIN
                                                        U,
                                                                SEVERITY OF
                                                                VIOLATION, V
                                                                (mg/l OF Cl)
              ~  V,         V2                    R

               Figure 21.  The preference function.
                                 49

-------
        NET EXPECTED
        GAIN, $
                                                   SEVERITY OF
                                                   VIOLATION, V
                                                   (mg/l OF Cl)
              Vi
                  Figure 22. The revised optimum violation.
standards is to shift the net expected gain curve downwards everywhere
to the right of the standard (S) as illustrated by curve XY with a maximum
between S and R.

   As long as illegal activities do not generate disutility, wealth is maxi-
mized at n2 where the violation undertaken is Vj.   This is where XY is
tangential to the highest horizontal indifference curve (not shown).  As an
aside, where the permittee is completely scrupulous the optimum viola-
tion is zero.   This is where XY crosses the vertical axis.  Here it inter-
sects the highest indifference curve (not shown) at a corner solution.  The
net gain generated is EN.

   If, however, illegal activities do generate disutility,  the violation whic"
maximizes permittee utility is Vj where XY is tangential to the UU indif-
ference curve. Since UU slopes upward from left to right, the optimum
violation (V\) when illegal activities do generate disutility must be less
than the optimum violation  (Vj) when illegal activities do not  cause disutil-
ity.   The substantive implication of recognizing that a permittee may de-
rive disutility from  engaging in illegal activities is minimal;  the only effect
is to reduce the size of the optimal violation.  Since this makes no differ-
ence to  a purely qualitative analysis such as this,  the remainder of  the
                                   50

-------
  analysis assumes for simplicity that a permittee does not derive any dis-
  utility from illegal activities.

     Since the effect of additional compliance monitoring expenditure is to
  increase the slopes of the expected cost function, this would make the
  slopes of the permittee's total benefits and expected cost curves (see
  Figure 20) equal at a lower value of V.  Consequently,  the permittee's
  optimal violation will decrease as expenditure on compliance monitoring
  increases.  If expenditure on compliance monitoring is increased suffi-
  ciently, the permittee's optimal violation will enventually become zero.
  This occurs when the slope of the permittee's expected cost curve ex-
  ceeds that of his total benefits curve for all V —i. e. , when the marginal
  benefit from a violation is less  than the marginal expected cost for all
  violations.   Although it is possible to reduce the permittee's violation to
  zero by increasing compliance monitoring  expenditures, it may not be
  optimal from society's point of  view to do so, depending on the social
  gains and losses associated with monitoring.

  Optimal Compliance Monitoring Expenditure

    The marginal cost of compliance monitoring is a function of the prob-
 ability of detection (the "output" of compliance monitoring).  The optimal
 expenditure on compliance monitoring will be that level of expenditure
 for which the marginal cost is equal to the marginal  benefit derived.
 Since a permittee's optimal violation of a standard will decrease as the
 expenditure on compliance monitoring increases and  vice versa,  this al-
 lows establishment of the functional relationship between expenditure on
 compliance monitoring and the social gain from that expenditure.

    The social gain from monitoring is defined to be the social loss avoided
 by that monitoring.  To illustrate this, consider Figure 23.  If no monitor-
 ing is undertaken, the expected cost to the permittee  is zero and he will
 produce OR pollutants  at the farmer's well.   In terms of Figure 20, the
 expected cost function is contiguous with the horizontal axis  (i. e., zero)
 and thus the net expected gain (H) is maximized at T.  The loss to society
 is equal to the area bounded by the marginal damage and benefit functions
 between S  and R (amount A).  If sufficient monitoring is undertaken to re-
 duce the permittee's violation by one unit (to SRj), the loss avoided is
 given by area (Tj).  Analogously, if enough  additional monitoring is under-
 taken to  reduce the violation by one more unit (to SR2),  the total loss
 avoided is  given by area (Tj + T£) and  the marginal loss avoided by area
 (T2).   Thus the total loss avoided increases  as the violation decreases and
it does so at a diminishing rate towards a maximum equal to A (which oc-
curs at zero violation—see Figure 24. )
                                 51

-------
BENEFITS,
DAMAGES, $
                                 MARGINAL DAMAGES
                                 FUNCTION
        MARGINAL BENEFITS
        FUNCTION
                                                      POLLUTION LEVEL
                                                      (mg/l OF Cl)
        Figure 23.  The benefits from violation reduction.
SOCIAL LOSS
AVOIDED, $
                                                               SEVERITY OF
                                                               VIOLATION, V
                                                               (mg/l OF Ct)
          Figure 24.  Diminishing returns to violation reduction.
                                 52

-------
        The relationship between various levels of probability of a violation
     being detected and the total social loss avoided may be inferred from Fig-
     ure 25.   The upper left quadrant shows the relationship between the prob-
     ability of detection and expenditure on monitoring.   The lower left quad-
     rant shows the expected cost of a violation to the violator as a function of
     the probability of detection—which, from Equation 1, is a proportional
     relationship.  The lower right quadrant shows the violation as a function
     of the expected cost.  This relationship shows that as the expected cost
     of a violation increases due to an increase in the probability of detection
     brought about by additional compliance monitoring, the optimal violation
     decreases but at a diminshing rate  (this relationship may be inferred from
     both Figure 20 and the derivatives of Equations 5 and 7).  Finally, the upper
                                              TOTAL SOCIAL
                                              LOSS AVOIDED
PROBABILITY OF
DETECTION, P,
EXPENDITURE ON
COMPLIANCE
MONITORING
                                                         NOTE:
                                                         ALL AXES ARE POSITIVE
                                 EXPECTED COST, E(C)

         Figure 25.  Compliance monitoring and the total social loss avoided.
                                     53

-------
right quadrant shows the relationship between the violation and the social
loss avoided.  There is  no connection between the upper left and the upper
right quadrants.  This is indicated by the two separate vertical axes — one
measures expenditure on compliance monitoring; the other measures the
social  loss avoided.

   To establish the relationship between the probability of detection  and
the total social loss avoided, assume that the expenditure on monitoring
                    1     ^                    *
is increased from M£ to M£ in order to increase the probability of detec-
tion by one unit from P^ to P,.   This increases the expected cost from
E(C)1 to E(C)2 which decreases the optimal violation from V^  to V2.  As
a result the total social  loss avoided increases from L]_ to L^. If expen-
diture  on monitoring is  increased to M^, probability of detection is in-
creased by one more unit and the total social loss avoided increases from
L«2 to 1,3.  While the total social loss avoided increases as the probability
of detection increases,  it does so at a diminshing rate (because  L,^-Li2 *-s
less than L^-Li).  Thus, the relationships between total social loss avoided
and the probability of detection (i. e. , the total benefit from compliance
monitoring) and the marginal social loss avoided and the probability of
detection (i. e. ,  the marginal benefit from compliance monitoring) will ap-
pear as in Figure 26 and 27, respectively.

    Figure 28  combines both the marginal benefits curve from  expenditure
on compliance monitoring and the marginal cost of compliance monitoring
(illustrated in Figure 16).  The optimal probability of detection is that
probability where marginal  benefit  is equal to the marginal cost, i. e. ,
where the two curves intersect at
   Once the optimal probability of detection to aim for is established the
optimal expenditure on compliance monitoring may be  inferred.  (See the
discussion contained in preceding subsection,  "Monitoring for Compliance,
and Figure  12. )  This identifies the optimal compliance monitoring budget.
The optimal monitoring strategy  (i. e. ,  combination of monitoring tech-
niques) may also be inferred because a unique monitoring strategy is as-
sociated with each level of probability of detection.

   The optimal detection probability is  not, in general, that which would
induce perfect compliance with the standard.   This is because it pays
society to undertake additional compliance monitoring that pushes a per-
mittee closer towards total compliance with the standard only so long as
the incremental gain to  society exceeds  the incremental cost of the addi-
tional monitoring required.  Thus, it should not be surprising to see mar-
ginal violations of quality standards.  To use a rather precise analogy,
the maximum highway speed limit is a standard which is enforced by po-
licing the highways.  The highways could be policed so intensively that
no one would  exceed the speed standard.  However, they are not because
                                  54

-------
  TOTAL SOCIAL GAIN
  FROM MONITORING
  (i.e., LOSS AVOIDED), $
                      TOTAL BENEFIT CURVE
                      FOR COMPLIANCE
                      MONITORING
                                                      PROBABILITY OF
                                                      DETECTION, Pd
     Figure 26.  The total benefits from compliance monitoring.
MARGINAL SOCIAL GAIN
FROM MONITORING  (i.e.,
LOSS AVOIDED), $
                  MARGINAL BENEFIT CURVE FOR
                  COMPLIANCE MONITORING
                                                     PROBABILITY OF
                                                     DETECTION, Pd
Figure 27.  The marginal benefits from compliance monitoring.
                               55

-------
  MARGINAL SOCIAL GAIN
  (i.e., LOSS AVOIDED),
  AND COSTS, $
           MARGINAL BENEFIT OF
           COMPLIANCE MONITORING
                                                MARGINAL COST OF
                                                COMPLIANCE MONITORING
                                                          PROBABILITY OF
                                                          DETECTION, Pd
            Figure 28. The optimal probability of detection.

the marginal social cost of such an increase in policing intensity would
exceed the marginal social benefit obtained from deterring the relatively
few persons who speed.


Monitoring Priorities
    Figure 28 may be used to illustrate the issues  involved in establishing
monitoring priorities.  Greater marginal benefits from compliance moni-
toring (for given marginal costs) will increase the optimal probability ol
detection.  This depends on the social loss avoided through monitoring
which, in turn, depends on the  difference between the marginal damages
and the marginal benefits associated with a certain level of pollution.
The larger the difference between the marginal damages and marginal
benefits from  pollution, the higher the optimal detection probability and
consequent expenditure on compliance monitoring should be.  This  infer-
ence is consistent with the notion that the larger the threat posed by a
pollutant the stricter the standard imposed should be,  and the more assid-
uously the situation should be monitored in order to  assure standard com-
pliance.  Thus, if the marginal cost of monitoring is the  same for dif-
ferent pollution situations,  then those situations can be ranked according
to  the threat that they pose for society.
                                   56

-------
     Conversely, the optimal probability of detection will increase with
  lower marginal costs of monitoring (for given marginal benefits).  Thus,
  different pollution situations that pose an equal threat to society can be
  prioritized by ranking them according to the marginal  cost of monitoring
  them.  The lower the marginal cost, the higher the probability of detection
  which should be aimed for and the larger the fraction of the total expendi-
  ture on monitoring which  should be allocated to those cases.  Considera-
  tions such as these argue for a greater allocation of monitoring expendi-
  tures to point sources as  opposed to nonpoint sources (such as sewer lines,
  agricultural runoff, etc. ) since the former have lower  marginal costs of
  monitoring than the latter.

  Corruption

    Although the objective analysis of dishonesty may be distasteful, this
 does not make dishonest behavior any less prevalent.  Regrettably, gov-
 ernment regulations are sometimes  broken either willingly or on payment
 of a bribe (perhaps disguised as a wage premium).  Even more regret-
 ably,  it is not unknown for government employees (cloaked as "inspec-
 tors" or "monitors") to collude in the violation of regulations.

    To this point the analysis has assumed that  when it is beneficial for a
 permittee to violate a groundwater quality standard he will do so with im-
 punity.  This  assumption implies that the permittee's employees obey the
 permittee's directions even when they are illegal, but it is highly likely
 that some of the employees will balk at taking actions which they know to
 be illegal.  One of the permittee's options if he is faced with a recalci-
 trant employee might be to fire him and replace him with a more trac-
 table employee. It is  to prevent such blatant victimization of honest em-
 ployees that the Federal Water Pollution Control Act states, in Section
 507(a), that "No person shall fire, or in any other way discriminate
 against . . .  any employee (who) . .  . caused to be filed .  .  .  any pro-
 ceeding under this Act. . . "  If,  after an investigation and hearing,  a
 fired employee is vindicated,  the permittee  may be enjoined to reinstate
 the employee with  retroactive pay and compensation for costs and ex-
 penses in  filing the complaint.  (PL 92-500, Section 507(b).) As an aside,
 such a weak protection clause would not seem to provide much incentive
 for employees to assist in  enforcement of the law.

   The permittee's  second alternative is to move a recalcitrant employee
 to other activities and  replace him with a more tractable employee.  Pre-
 sumably, if the employee were not demoted he would find it difficult, and
would have little incentive, to prove "discrimination. "
                                  57

-------
   Up to now it has been implicitly assumed that the permittee's  employ-
ees would willingly follow the permittee's directions even though that
meant breaking the law, or that he was able to replace recalcitrant em-
ployees with tractable substitutes, without incurring any additional costs.
In a competitive job market this might prove to be close to the truth,
especially when adherence to the standard itself would,  if anything, tend
to make jobs scarcer by reducing the optimal scale of the permittee's
activity.  The possibility of quality standards adversely influencing em
ployment opportunities is  recognized in Section 507(e) of PL 92-500,  \
states that "The Administrator shall conduct continuing evaluations of
potential loss or shifts of  employment which may result from the issuance
of any effluent limitation .  . . including .  . . investigating threatened
plant closures or  reductions in employment ..." However, it is  possi-
ble that none of the permittee's employees would be willing to violate the
law without additional compensation.  Thus,  the permittee's third option
is to pay employees a wage premium (i. e. , a bribe) in order to induce
them to implement the standard violation he  has decided upon. It is plaus
ible to  assume that this bribe would be an increasing function of  the size
of the violation desired by the permittee.  That is,

                                    SB,
                      Bj  = ' •""
where BI is the cost of the bribe to the permittee.  Addition of this cost
to the permittee's other expected costs, as given by Equation 6,  results
in
                  E(C) = f(M,V) •  j(V) + h(V)  .                 (9)
 Instead of the original expected cost curve shown in Figures 19 and 20,
 there would be a new curve everywhere above the original curve and di-
 verging from it as V increases  (i. e. ,  with a greater slope at any V than
 the original curve),  with the implied optimal violation less than Vj.
 Apart from this minor modification, none of the previous analysis is al-
 tered by making allowance for the permittee to bribe employees to vio-
 late groundwater quality standards.  In fact, even this minor modification
 does not arise if the required_bribe is  not a function of the size of the vio-
 lation but is a fixed  amount (Bj) independent of the size of the violation.
 If this is the case, then the cost curves in Figures 19 and 20 shift up by ^
 BI without changing their slope and, therefore,  without changing the opti-
 mum violation (Vj).  The only effect in this situation is on the distributi°n
 of the permittee's gain.  The permittee would now  retain only B - BI
 while "Bj would be transferred to his employees.
                                   58

-------
      While it is not common for government employees to collude in the vio-
   lation of regulations they  are supposed to enforce,  it is not unknown.  Thus,
   for completeness1  sake the analysis should consider what would happen
   if the government employee responsible for enforcing compliance is
   corruptible.

     It seems plausible that the bribe demanded by the government "moni-
  tor" B2 would be an increasing function of the violation proposed by the
  permittee.  Thus,


                       B2 = k(V)   ;

  In general, it must be assumed that employees would continue to  require
  bribing  to violate standards notwithstanding collusion by the government
  monitor.  Thus, the permittee's new expected cost function is given by

                         E(C) - h{V) +k(V)  .                       (11)

  In this new expected cost function  the first multiplicative part of the ex-
  pected cost function, f(M,V)  • j(V) , in, for example,  Equation 9 has
  dropped out.  That is because the probability of detection is assumed to
  be zero  given that the government  monitor has been  suborned.  Instead
  of the original expected  cost curve shown  in Figures 19 and 20 there would
  be an entirely new linear curve with a slope equal to [(BBj/dV) + (BB2/dV)]
 Naturally the slope of this new curve would be equal  to the  slope of the per-
 mittee's expected gain curve  (which is unchanged) at a higher optimum
 value for V because only then will collusion by his employees and the
 government monitor have increased his maximum net expected gain.

    If the  bribes paid to employees and the government_monitor_are not a
 function of the size of the violation but fixed amounts B-^  and B2  that are
 independent of the size of the violation, a_wors_t^case occurs; the new ex-
 Pected cost curve is a horizontal line at (Bi + B£) •  Consequently, the
 Permittee maximizes his net expected gain  by violating the  standard by
 SR •  In short, he operates as though no standard existed.  Referring
 back to Figure 9, the cost to society would equal A +  M where  M  is
 equal to the alleged "monitoring" expenditure,  or an amount that exceeds
 the loss to society which  occurs if no standard  and monitoring are imposed.
Apart from imposing on society the additional cost of  M for monitoring,
corruption of the monitors merely imposes a transfer_of the permittee's
gain. The permittee would now retain only B - (B! +  B2). with BI being
transferred to his employees and B2 to the  government monitor.
                                 59

-------
Implementation Problems of Monitoring
   Some practical difficulties attend the implementation of the optimal
compliance monitoring strategy.  While the optimal probability of detec-
tion to aim for is that level at which the marginal benefits from compli-
ance monitoring equal the marginal costs of compliance monitoring  (see
Figure 28), it is inconceivable  that this level (and implicitly optimal level
of expenditure on compliance monitoring) will remain unchanged over
time.  It would only remain unchanged over time if all the underlying ele-
ments of the marginal benefit and cost curves remain unchanged. For
example, the marginal cost curve for compliance monitoring will change
as technological innovations in  the monitoring industry occur and as the
relative prices of various alternative monitoring techniques change. Sir*1"
ilarly, the marginal benefit curve will change as the marginal social gam
from monitoring (i. e. , the loss avoided) changes.  And this will change
as the polluter's marginal benefits  function from emitting pollutants and
the pollutee's marginal damage function from the  impact of pollutants
change  (which changes the optimal level of pollution).  These marginal
benefits from pollution emission and marginal damages caused by pollu-
tion are themselves dependent  on the cost of the factors of production
and the prices of the products that the polluter and his victim are con-
fronted with.  These factor costs and product prices will inevitably change
with the passage of time.   This will change the optimal level of pollution,
the optimal probability of detection, and the optimal level of compliance
monitoring expenditure.   It is apparent,  then, that the determination of
the optimal level of pollution and, from that, the optimal probability of
detection is not a once and for  all exercise.  Quite the contrary, in prin-
ciple these change continuously through time and,  consequently, any op-
timal strategy would require continuous applications of time and resource
to its discovery and implementation.  Indeed, it is quite conceivable that
the transaction costs  involved in continually  determining and implement-
ing the optimal monitoring strategy would exceed the social gain from its
implementation (in terms of the net social loss avoided).   If that is the
case, then it follows that it is not in society's interest to follow that strat-
egy.  Moreover, the hypothetical case employed in this analysis to illus-
trate the economic principles involved in the determination of an optimal
monitoring strategy was chosen for its simplicity.  The example involved
one identifiable polluter emitting one pollutant which damages one other
party.  If the practical implementation of an optimal monitoring strategy
on a continuing basis is likely  to prove so difficult in such a simple situa-
tion, one  can only conjecture how much more complicated and costly the
implementation of such a strategy would prove to be in situations where
multiple polluters, multiple  pollutants, and multiple pollutees exist, °r
for nonpoint sources  of pollution.  However, this  does not necessarily
mean,  purely on the grounds of economic efficiency,  that nothing should
be done to preserve the quality of the Nation's groundwaters.
                                   60

-------
 THE  PRAGMATIC ALTERNATIVE-SECOND-
 BEST SOLUTIONS
 Standard Setting
    Since implementation of the optimal monitoring strategy on a continu-
 ous basis through time may impose more costs than benefits on society
 (in terms of social loss avoided), this section examines  second-best
 solutions  as a pragmatic alternative to this dilemma.  (The term  "second-
 best" is used in the theory of welfare economics in reference to non-
 pa reto-optimal situations (i. e. , when one transactor can be made better
 off without any other transactor being made worse off).   The term as used
 here is only loosely related to the rigorous way in which it is used in for-
 mal theories of second-best choices. )   Essentially, second-best solutions
 capitalize on the fact that rational action can be predicated on limited in-
 formation concerning the polluter's total benefit function from emitting
 pollutants and the pollutee's total damage function from the impact of pol-
 lutants.

   The limited information that is required in order to implement  some
form of second-best solutions is illustrated in Figure 29.  The two dashed
        ' TOTAL BENEFITS AND
        DAMAGES, $
              X
              r
     o'f
 x
T

                 W'
                           COMPANY'S
                         TOTAL BENEFITS
                         FUNCTION
                         \/
                          A
                                         /FARMER'S TOTAL
                                        / DAMAGES FUNCTION
                                     ,!/
               ~~ \
                      Cl CONCENTRATION (mg/l)
             T      W                 R
                Figure 29. Second-best alternatives.
                                61

-------
curves in this figure are the farmer's total damages function and the oil
company's total benefits function first presented in Figures 5 and 7,
respectively.

   These two curves are shown dashed in Figure 29 as a reminder that
their precise, or complete,  configuration is assumed not to  be known.
Information on these damages and benefits is restricted as follows:

      1.   The damages sustained by the farmer when the oil  com-
          pany's activities are unconstrained.  These are the dam-
          ages associated with the level of chloride concentration
          at the farmer's well equal to  OR ; namely, RR' .   The
          total benefits to the oil company are equal to RX (a
          maximum).

      2.   The chloride concentration at the farmer's well at  which
          the damages sustained by the farmer would be zero.
          This is the threshold-level of chloride concentration
          in Figure 29.  (The threshold may, of course,  be zero
          chlorides.  In general, however, it will be some posi-
          tive level. )

      3.   The cost to the oil company of reducing the chloride
          concentration at the farmer's well to the threshold level
          OT .  This is the reduction in the present value of  the
          oil company that is implied when it adopts the least-
          cost way of reducing the chloride concentration at  the
          farmer's well to  OT  (an amount XT'  in Figure 29).
          In principle,  XT7  may be infinite if no economically
          feasible techniques exist for  lowering the chloride  con-
          centration level  to  OT .  In general, however,  it will
          not be.

      4.   The damages sustained by the farmer,  and the cost to
          the oil company, associated with the reduction of the
          chloride concentration level at the farmer's well to
          that attainable when the "best practicable control tech-
          nology [BPT] currently available, "  is adopted (USEPA,
          1974).  These damages and costs are WW and XW",
          respectively, while the chloride concentration level
          associated with  BPT  is OW.  (The chloride concentration
          level associated with  BPT  may be less than the thresh-
          old level, i. e. ,  OW may be less than OT.  When the
          threshold is less than BPT,  XT7 —i. e. ,  the reduction in ,
          the present value of the oil company implied when  it adopts
                                  62

-------
            the least-cost way of reducing the chloride concentra-
            tion at the farmer's well to the threshold—would be
            expected to be very large.  In other words, the total
            benefits function falls off precipitously to the left of
            W" because of the extremely high cost of lowering
            pollution levels below BPT.  In principle,  the  O'w"
            part of the total benefits curve could become a negative
            value.)

     If complete information were available on the total damage and bene-
  fits curves, the optimal level of pollution to adopt would be that level at
  which the divergence between the two curves was a maximum.   This
  would correspond to the level  OS  (not shown, but between  OW  and OR ).
  However, assuming that this complete information is costly to acquire
  compared to the cost of acquiring the smaller amount of information dis-
  cussed in items  1 through 4, above, the question is not whether to im-
  plement  OS  (the optimal, or "first-best, " policy when information is
  complete) but which of the second-best policies   0 ,  OT , OW ,  or  OR
  to implement given the limited information that  is available or may be
  acquired relatively inexpensively.

    The principle governing the decision is  the same; namely, select that
  second-best alternative with the largest net benefit to society.  Diagram-
 rnatically, this amounts to selecting the second-best policy from the sub-
 set,  0 ,  OT , OW ,  and OR at which the difference between the oil com-
 pany's benefits and the damages to the farmer is the largest.   For pur-
 poses of illustration, in Figure 29 this is seen to be the  BPT policy, OW ,
 since W'W" exceeds  00', TT7 , and XR7 .  Of course, the  0,  OT , and
 OW policies also involve monitoring and other enforcement costs which
 the OR policy does not.  Thus,  OW  is,  in fact,  the desired second-best
 policy only if W'W" exceeds  XR' by at least as much as  the monitoring
 and enforcement costs required to implement that policy.

    Pragmatically, in  setting standards the EPA is confronted with the de-
 cision whether to allow the oil company's activities to continue at the un-
 constrained level (OR) or whether to set  some new standard  such as BPT,
 threshold,  or zero pollution.  To decide which is the appropriate second-
 best standard, the EPA would have to  estimate the reduction in damages
 to the farmer for each policy and subtract from this the additional cost
 to the oil company plus the monitoring and other  enforcement costs for
 each policy.  The policy for which this difference was largest is the ap-
propriate second-best policy (as long as  the difference is positive).  By
way of illustration, in Figure 29 the  OW  policy reduces the  damages to
the farmer by R'W' while increasing the oil company's costs by XW" .
The net benefit to society from changing to this standard is,  therefore,
                                 63

-------
(R'W') - (XW" 4- Mw) where MW is the monitoring and other enforcement
costs associated with the BPT standard.   The net benefit to society from
changing from OR to OT would be (R;T) - (XT' + Mt) where Mt is the moni-
toring and other  enforcement costs associated with the threshold stand-
ard; and the net benefit to society from changing from OR to 0 would be
(RX0) - (XO'  + MQ) where MQ is the monitoring  and  other enforcement
costs associated with the zero pollution standard.  In principle,  MW,
and MQ may differ.


Monitoring
   The previous  subsection argues that,  in a second-best situation, the
appropriate policy is to  select that standard for which the difference be-
tween the reduced damages to the farmer and the additional cost to the
oil company plus the monitoring and other enforcement costs was the
largest. However,  this says nothing about what the optimal expenditure
on monitoring should be.  The reason for this omission  is that,  in a
second-best  situation, there is not sufficient information to infer the op-
timal expenditure on compliance monitoring.  This  can be explained by
referring back to Figure 28 and, in particular,  to the marginal benefit
curve for compliance monitoring.  That curve in turn was derived from
Figure 26 which shows the total social gain from monitoring (in terms ot
loss  avoided) as a function of the probability of  detection.  But the gain
was defined in terms of the area between the marginal benefits and datn-
                                                               •     *? *% \
ages functions associated with various levels of pollution (see Figure L^i1
However, these latter two curves depend on complete information con-
 cerning the total benefit and damage curves associated with pollution
 and, by definition, in second-best situations such complete information
 is not available.  By this chain of reasoning,  in second-best situations
 the marginal benefits curve from compliance monitoring is not discern-
 ible,  and thus the optimal probability of detection to aim for (which also
 is implicitly optimal for compliance monitoring) cannot be identified.

    There seems to  be no answer to this particular difficulty except to
 make certain that if a second-best standard is  set, the monitoring and
 other transactions costs imposed to enforce compliance with the stand-
 ard are no greater than the difference between  the  reduction in damages
 to the farmer and the increase in costs  to the oil company.  As long as
 the expenditure is kept below that difference  society gains; if it equals
 that difference society is no better or worse  off and the  second-best pol-
 icy would only have distributional effects; if it  exceeds  that difference
 society loses.

    The foregoing discussion of implementation problems and second-best
 solutions serves to  show how difficult it proves to be for the government
                                   64

-------
to set and enforce economically efficient groundwater quality standards
by direct regulation.  Because of these difficulties, Section V explores
an alternative method of obtaining an economically efficient level of
groundwater quality which relies on a respecification of the private prop-
erty right to that resource.
                                65

-------
                             SECTION V
         WASTE RELOCATION RIGHTS:  AN ALTERNATIVE SYSTEM
      OF GROUNDWATER MONITORING AND POLLUTION CONTROL
   In general, it is analytically incorrect to argue that pollution should be
eliminated.   As with most other things, there is an optimal amount of
pollution.  The burden placed upon regulatory authorities is to design in-
stitutional arrangements that will achieve that optimum.  The analysis in
Section IV of the set of institutional arrangements required to achieve
this  end through direct government regulation by issuance of permits
shows  that considerable,  perhaps insuperable, difficulties lie in the way
of this option.

   This section discusses  in some detail a second scheme for regulating
groundwater quality which revolves around what is described as a "Waste
Relocation Right. "  The associated monitoring requirements for this reg-
ulatory scheme are also addressed.

   It is recognized that the concept of waste relocation rights is contro-
versial and  would  require considerable modification of existing institu-
tional  and legal mandates in order to be implemented.  As a consequence
it must be stressed that a regulatory scheme based on waste relocation
rights  does  not necessarily reflect  EPA policy.

ADMINISTRATIVE VERSUS NONADMINISTRATIVE
POLLUTION  MONITORING AND CONTROL

   Most attempts to monitor and control groundwater pollution in the
United States have  involved administrative regulations.  Discharge  per-
mits,  for example, set allowable upper limits on the amount of pollutants
that  may be disposed of into the ground at a point source.  This limit is
set by statute or regulation by the cognizant administrative agency.  Most
of the  proposals for new methods to monitor or control pollution,  includ-
ing effluent disposal charges,  also involve mainly administrative tech-
niques.  Although a pollution regulatory system based on effluent "fees"
is thought by some to be a market-oriented type of pollution control, it
too is more an administrative than  a market system.  Although the  pollu-
ter is  allowed to vary his disposal techniques  according to his comparison
of the  cost of the fees paid for injection and the cost of alternative means
for waste disposal, the fee is determined administratively; it is not a
                                 66

-------
  price that fluctuates according to the market processes of demand and
  supply.  Moreover,  most monitoring under the administrative regulatory
  system is performed by public agencies.  Under a nonadministrative sys-
  tem, monitoring would be undertaken more often by private parties through
  the ordinary market processes of transactions among strangers.

    The U. S. system of basically administrative control of monitoring and
  pollution is too new to yield much knowledge  of its overall impact on so-
  ciety.  Among the benefits of administrative  regulation have been moni-
  toring of severe pollution situations, preventing deterioration of some
  important aquifers by pollution, and drastically limiting disposal of such
  dangerous pollutants as chromates.  In spite  of these benefits, however,
  it is difficult to judge whether such  regulation has on net improved soci-
  ety's welfare.   Administrative regulations  are imperfect instruments for
  controlling  and monitoring pollution. The responsible agencies may not
 possess all the information required to ascertain existing or optimal pol-
 lution levels.  They may devote too  many or too few resources to moni-
 toring. Once rules and regulations  on monitoring and pollution control
 are implemented, they are often costly and cumbersome to change even
 if circumstances alter rapidly.  The requisite monitoring and compliance
 data to assure  that no more than the economically optimal amount of pol-
 lution occurs might become available eventually.  But agencies might lack
 incentives to employ these data and to pursue social  optimality single-
 mindedly to the exclusion of internal bureaucratic goals.   Moreover, an
 agency devoting itself too zealously to the reduction of pollution might cur-
 tail certain  pollution-causing economic  activities that would be of greater
 benefit to society than the value of removing the pollution associated with
 them.  Unfortunately, there is little  evidence to suggest that administra-
 tive regulation  of monitoring and pollution has resolved these problems
 in a fashion that would,  overall,  benefit society. Thus, given today's
 meager knowledge, it is literally the case that past experience with ad-
 ministrative regulation of groundwater pollution and monitoring raises
 more questions than it answers.

   A  recognition that  the existing system of pollution and monitoring con-
 trol is imperfect evidently has not troubled many officials or citizens,
 probably because the  existing system of regulation, with all its warts, is
 considered to be superior to  "no regulation, " the only alternative that has
 been conceptualized in the public's mind.  But the criticism against the
 existing administrative system could be taken  seriously if there were a
feasible nonadministrative system of regulation that could prevent uneco-
nomic aquifer degradation while containing fewer of the pitfalls of admin-
istrative regulation.
                                  67

-------
   One nonadministrative system of groundwater pollution monitoring and
control that could be considered as a legitimate alternative to improve
the overall efficiency with which the Nation's aquifers are used has as
its basic element a waste relocation right.* The term "waste relocation
right" is used here in recognition that wastes,  or residuals,  are the in-
evitable  result of man's activities and that, equally inevitably,  they must
be relocated somewhere.  The problem is,  of course, to get them opti-
mally relocated.  The waste relocation right is a grant to  each  land owner
which entitles him to protection from pollution above a specified amount
in the groundwater underlying his land.   This right would  be  exchanged
among persons through ordinary market processes and would be enforced
by the existing U. S.  J.egal system.  It would not require the creation of
new,  untried, and potentially expensive governmental agencies.  This
would be an essentially private system of pollution monitoring  and control,
with limited government supervision, based on the self-regulating forces
of existing and tested market processes  and legal institutions.

   Waste relocation  rights would be analogous to mineral leases and tied
to the ownership of land.  Pollution of groundwater underlying  another
party's land would be prohibited without first purchasing that party's
waste relocation  right.  Such rights would be valuable, freely transfer-
able,  and command prices in the marketplace.  They would be exchanged
privately in much the same fashion as titles to land are now  bought and
sold.  Monitoring for pollution in excess of the allowable  levels would
mainly be through private contracting.   Enforcement of property owners
 rights would be through injunctions or monetary damages awarded  by
 courts.  Thus,  the waste  relocation rights  scheme would  rely  upon and
harmonize with the basic processes by which rights to other resources
in this country  are  exchanged and protected.

    The remainder of this  section describes the waste relocation rights
 system  and its  ability to cope with the peculiar and sometimes  severe
 problems  presented by  groundwater pollution.  Detailed are the nature
 of the rights, the advantages of court enforcement by a rule of strict lia-
 bility, problems in setting the initial limits on the amount of waste that
 may be  relocated, the processes and costs of exchanging rights, the pro-
 tection  given by the system to  nonowners of land and environmental groups,
 and some potential problems in the operation of this property  system in-
 cluding situations where it might not be applicable.
 *For discussion of a surface water waste relocation right scheme,  see
  Dales  (1968), Chapter 6.  Dales calls his contract a  "pollution right. "
  However, the term "pollution right" has been countersloganized in
  the phrase "there can be no right to pollute" which,  though wrong,  has
  emotional appeal to those who traffic in slogans rather than analysis.
                                   68

-------
 WASTE RELOCATION RIGHTS*

    Assume that there are 11 pollutants in the range of 1,. . . , n_ and that
 there are m_holders of waste relocation  rights in the range of 1,... ,m.
 Py would then designate the maximum allowable level of the i^^1 pollu-
 tant underlying the land of the j^1 individual.   For example, if the 4th
 pollutant were chlorides, P^J would represent the maximum concentra-
 tion of chloride pollution (e, g. , 300 milligrams per liter of water) that
 rightsholder j  would legally be  required to sustain under his land.   P£J
 could be the maximum  allowable biochemical oxygen demand for organic
 material for water under j's land; Py* could refer to the maximum parts
 per million for dissolved Folids in the water under j^s land; and so forth.

    Under a waste relocation rights scheme, the owner of each legally
 defined parcel of land or his assignee in  any of the United States  would
 be granted by either Federal or State statute a package of rights  and
 duties regarding the use and enjoyment of the waters underlying his land.
 This package would contain paragraphs to the effect  that:

       1. He  is  granted the  exclusive right to engage  in economic
         or other activities on his land which originally degrade,
         change, or pollute the water underlying that land to any
         extent, provided that the level of pollution which origi-
         nates from his land does not exceed PJJ in the ground-
         water at any point beyond the boundaries of his land.

      2. He is granted  the  right to be free of pollutants in the
         water underlying the boundaries of his land, which
         may originate from activities on the land of any other
         owner or his assignee in the United States,  and which
         degrade, change,  or pollute the quality of the ground-
         water underlying the  boundaries  of his land  beyond the
         levels PJJ referred to in item 1.

      3.  For violation of the duty defined  in item 2,  an owner
         of land or of waste relocation rights shall,  upon proof
         by any other owner of a violation, be required to adopt
         the least-cost method of restoring the quality of ground-
         water to the level  Py specified in item 1 within a

*The basic analysis contained  in pages 69 to 81 of this report is analo-
 gous to and  drawn from the property system for electromagnetic radia-
 tion found in De Vany et al. (1969).  The important contributions of Pro-
 fessor Charles J.  Meyers to the conception and the  legal drafting of
 these property definitions  are gratefully  acknowledged.
                                 69

-------
   period of M_months, where M_is determined according
   to accepted practice for models of groundwater flows
   for particular aquifers, given the maximum levels of
   pollution P^.  In addition to the foregoing relief by
   injunction, the complainant may recover such pecu-
   niary damages that he may be able to prove in a court
   of law as having been caused by said violation.

4.  All owners of land or waste relocation  rights must reg-
   ister their rights to groundwater use,  including any
   subsequent exchange or recombination  of those rights,
   in a  central registry maintained by the cognizant gov-
   ernmental agency for public access by  any owner of
   land.

5.  A party suspecting a violation of his waste relocation
   rights may install monitoring devices  on his own
   property, and he may install  at his  expense such de-
   vices on the property of other owners  with their ex-
   press consent.

6.  There shall be no restraint upon the transfer or ex-
   change between two or more owners of land of  all or
   any  part of the waste relocation rights  which are spe-
   cified in items 1 and 2, either as to:

      a. Changes in the boundaries of land beyond which
         pollution in the underlying groundwater may
         not exceed the limit P;;, or
                               J
      b. Changes in the intensity of the pollution of
         groundwater to a quality different from the
         PJJ defined in item 2 within the boundaries
         of the owners of land or waste relocation
         rights who are parties  to the exchange
   provided, that this  exchange  does not result in levels
   of pollution in excess of the original Pj- specified in
   item 1 for the groundwater underlying the land of any
   owner of waste relocation rights who has not been com-
   pensated for this higher pollution or whose permission
   has  not been obtained.
                             70

-------
  COURT ENFORCEMENT AND  STRICT LIABILITY

     The waste relocation rights regulatory system would be created by
  statute but enforced by  adversary proceedings in courts of law.  This en-
  forcement procedure appears to have several advantages.  First,  there
  is already a long-standing body of case law on groundwater pollution, so
  the courts are familiar  with the often technologically complex questions
  raised by groundwater pollution and monitoring problems.  (See, for
  example, Corker (1971),  Knodell (1967),  and Meyers and  Tarlock (1971).)
  The courts could probably enforce groundwater rights more successfully
  than would  relatively untried administrative  agencies.  Second, the ten-
  dency of the courts  to settle fresh cases on the basis of precedent would
  give owners of waste relocation rights some confidence over the nature,
  extent,  and value of their resources.  Third, the adversary nature of
  court proceedings would focus on the competing harms  involved in pollu-
 tion cases—i. e., whether the polluter  is to impose the  cost of lower water
 quality on the pollutee or the pollutee is allowed to shut down some  or all
 of the polluter's economic activities.  Coase (I960) has argued that courts
 often take these economic factors into  account in reaching decisions.  If
 the  process by which courts  weigh competing harms usually leads to se-
 lecting in favor of the least economic cost, then efficient  resource use
 would be promoted.

   In general, there are two broad alternative rules of  liability for pollu-
 tion that the courts might impose:  fault (or tort) versus strict liability.
 Under tort  law,  liability usually depends not  just on the actual violation
 but upon such factors as the  reasonableness of the violation or whether
 the defendent intended to harm the  plaintiff.  (See Section  II. )   Alterna-
 tively, under the rule of strict liability, a dependent's guilt is determined
 by only the fact of his violation of a duty owed to someone else, and not
 by the degree of the  violation, his state of mind, or other extenuating
 circumstances.

   According to the waste  relocation right, liability for  violations would
 be strict (De Vany et al.,  1969).  Referring to item 3 of the right, if
 rightsholder A monitors  for pollutant 4 under his land and finds more
 than the allowed P4A which is traceable to rightsholder B, then^AJs
 remedy is to sue B for the violation of duty which B owed to A. A may
 sue B either for an injunction forcing B to reduce the pollution to  an
 amount within the P4A constraint before I£months have elapsed;  for
 damages which A_can prove he sustained as a result of pollution from B
 in excess of P4A; or for  both an injunction and monetary damages.  All
that matters in court is whether B's actions led to more than P4A pollu-
tion being found in the water underlying either A's land or  the boundaries
of his waste relocation right.   It does not matter whether B's violation in
                                 71

-------
exceeding P^ was the result of intent,  carelessness, ignorance, malev-
olence, or stupidity.   It also does not matter whether B may have exceede
the P4A limit by a small or "reasonable" amount for a  short period of
time.  If B exceeds P4A by any amount no matter what  the reason may be,
he is guilty under a standard liability if A chooses to enforce his right.

   A standard of strict liability is rarely found in the existing law of
groundwater pollution.  (See Section II, particularly quotations from
Corker (1971).)  The  guilt of the polluter is usually determined by one
of several tort law concepts that seek to determine the  nature of fault
for pollution.  The determination of fault raises additional questions be-
yond the simple and sole question raised by strict liability—whether B s
pollution under A's land exceeded P4A-   ^n some situations, tort law de-
termines  guilt according to -whether the polluter intended to harm the po
lutee, or whether the  polluter had been careless or negligent in failing to
exercise a prudent degree of care in his pollution-causing activities.  In
still other cases guilt depends upon whether the pollution caused a nuisanc
to the pollutee.

    An economic objection  to enforcement by a rule of tort liability is tna
transactions costs would be raised  and the incentives for efficient use or
 resources would be weakened because the process of making and enforc-
ing transactions consumes additional resources.  It includes the costs
for both buyers and sellers of searching out, negotiating, and enforcing
mutually beneficial exchange opportunities.  This involves hiring survey-
 ors, lawyers, and other consultants.  Sometimes it is  necessary to go
 to  court when voluntary compliance is not obtained  or when the parties
 are in disagreement over whose right is dominant.   When these costs ar
 relatively high,  the incentives for  rightsholders to  enter into exchanges
 will be limited:  if the cost of making an exchange exceeds its value to
 either party it will not be completed.  Relatively high transactions costs
 will dampen and limit the efficiency of the economic process by which
 rights to resources move from lower valued to higher  valued uses  and
 users.  Thus, in designing new property rights, care  should be taken to
 minimize transactions costs in order to encourage beneficial exchange.

    One factor that will almost always impede exchange is a vague defini-
 tion of rights or liabilities.  The more uncertain rights and duties are,
 the greater the transaction and litigation costs will be.  In the cases of
 tort law,  a legal determination of  the meaning of such ambiguous stand-
 ards as "reasonableness, " "negligence, " "good faith,  " or "intent, " wouW
 be so lengthy and expensive that these concepts could  scarcely be terrne
 standards at all.  If they  could be  given legal meaning, it would only be
 by a lengthy ad hoc procedure where each situation is  adjudicated accor
 ing to its particular facts, individuals,  personalities,  intentions, etc.
                                   72

-------
    (De Vany et al., 1969).  Since there really is no "standard" of tort lia-
    bility,  each case would require a separate adjudication and its result
    would be in doubt until the verdict was in (Corker, 1971).  This would
    increase an owner's uncertainty about the value of his  rights,  tempting
    him to try litigation before he bargains with his "adversary. "  This ac-
   tion avoids  his purchasing a right that a court might  later decide he al-
    ready owned, or adjusting his economic activities to reduce pollution
   when not required by law.  Thus, nebulously defined rights  reduce the
   likelihood of bargaining,  increase transactions and litigation costs, and
   reduce the probability that society's scarce resources  will be  used opti-
   mally.                                                            ^

    ^ ^Whether  legal critics have found an acceptable substitute to tort lia-
   bility,  as Corker alleges  they have not, is open to dispute.   In the case
   of waste  relocation rights, a standard of strict liability would lower trans-
   actions costs by raising "one,  and only one, triable issue of fact: What
   was the [level of pollution] at the time in question? "  (De Vany et al.,
   1969).  Did it exceed the maximum Py level? This would not always be
   an easy question to answer.  Substantial resources would often have to
  be devoted to the problem of monitoring and proving groundwater pollu-
  tion.  But it  is probably an easier question to answer  than the questions
  raised by tort liability rules.


  THE ZERO WASTE RELOCATION  "RIGHT"

     The preceding comparison of fault versus strict liability rules also
  suggests that a regulatory system of waste relocation  rights which sets
  the initial Pjj equal to zero would not promote the efficient use of re-
  sources (De Vany et al. , 1969).  Although some might find it appealing
  to specify waste relocation rights that allow zero increments  to ground-
  water pollution,  these "rights" would have little value.   As a practical
 matter,  pollution increments cannot literally be reduced to zero without
 curtailing most human and  economic activities.  The practical result of
 setting an  initial limit of 0  mg/1 of,  say,  chlorides would be to give pol-
 luters the  right to inject chlorides up to the point where they did not cause
 excessive harm to the groundwater underlying the land of their neighbors.
 But, as has just been argued, a right defined in terms of excessive harm
 entails an enforcement mechanism having relatively high transactions
 costs.   Each case of violation under the zero pollution right,  as  with the
 fault rule of liability,  would have to be adjudicated separately in  order
 to determine what constituted excessive harm in various  circumstances.
 This increased degree of uncertainty would lead parties to litigate rather
 than bargain and would curtail exchanges, all of which would diminish the
 efficiency of resource  use.   Thus, on grounds of economic efficiency, the
waste relocation right  should be defined at some positive  level rather than
at zero.
                                 73

-------
SETTING THE INITIAL WASTE LEVELS
   Each landowner has thus far been hypothetically granted a package of
rights and duties: first, a right of protection from a certain level of pol-
lution caused by others in the groundwater underlying his land; and sec-
ond, an obligation to refrain from causing this  same level of pollution to
be exceeded in groundwater underlying all other parcels of land.  Thus,
all landowners have a structure of symmetrical rights and duties concern-
ing groundwater pollution.  The permissible amount of pollution is deter-
mined by the original limits  of the P^ and by the subsequent process of
exchanging these rights in the marketplace.

   The  initial Pjj limits would be set after a series  of once-and-for-all
hydrologic,  biologic, chemical, and economic  studies established toler-
able and economically usable amounts of PJJ.  These would result in
packages of rights  that would have monetary value.   Two of the Py fea-
tures should be stressed.  First, the initial levels should usually be
slightly in excess of ambient levels of pollution.  PJJ greatly in excess of
ambient levels would permit  rightsholders to substantially increase pol-
lution without obtaining consent from the affected parties.  PJJ signifi-
cantly below ambient pollution levels should also be avoided.  This woul
force some landowners to "clean up" the waters  underlying their property
before the land could be used for any purpose,  possibly imposing signifi-
cant losses on them.

   Second,  the initial PJJ must be usable but need not be perfect.  That
is, they need not be economically optimal or ideally suited to the use tha
would  result on a given parcel  and all surrounding parcels of land with
existing technological and market conditions.  The cost of amassing the
information required to determine optimality at the outset would be enor-
mous.  Such information is not presently available and would probably
cost more than the several rounds of exchanges  required to optimize the
original rights packages.  Moreover, the variables determining the  opti-
mality of rights  packages— such as population concentrations,  technolo-
gies, costs, and market  demands—will undergo  constant change.  This
 requires that rights be restructured accordingly.  Therefore, under the
waste relocation rights regulatory system, the economically optimal
amount of pollution and monitoring would be approached iteratively throug
market exchanges once usable packages of rights were  created and as-
 signed. The market exchange approach avoids the  most serious difficul-
ties associated with the administrative maintenance of the optimal level
 of groundwater quality via permits, the problem of the  changing nature
of that optimum through time  (De Vany et al. , 1969).
                                   74

-------
     In most cases, the process of exchange would constitute the least ex-
  pensive process for allocating waste relocation rights from lower to
  higher  valued uses.  To illustrate this, if brine pollution from the oil
  company example of this study were seriously affecting the crop yields
  of adjacent farmers, the farmers would have incentives to purchase the
  oil company's waste relocation right if the right were more valuable in
  agriculture than in industrial production.  However, if it  is known in ad-
  vance that a particular level of pollution would be devastating to highly
  valuable activities located nearby—such as a municipal water supply—
  then exchange may not be the most efficient process for determining the
  optimal amount of pollution.  In these cases it would be uneconomic to
  set the  initial P^ at a level that would devastate the higher valued use,
  and setting it at a lower,  near zero level could avoid the cost of restruc-
  turing rights  later on.


  EXCHANGING WASTE RELOCATION RIGHTS

     Two  conditions must be met in order for  optimality to  be approached
  through  market processes.  First,  the  cost of making exchanges and re-
  combining packages of waste relocation rights  must be low relative to
  their value.  Second, mutually beneficial exchanges between two or more
  transactors must not impose uncompensated costs in excess of Py pollu-
  tants  on  other rightsholders unless  their consent is  first obtained. Just
  as it is socially and economically undesirable for two transactors to im-
  pose harm on third parties without their consent, it is equally undesirable
  for the law to require that the consent of third parties be obtained before
  each exchange is concluded.  The transactions  costs required to achieve
 optimal resource use would be almost the same in each instance; in the
 first instance third parties must bargain with transactors to prevent the
 harm,  and in the second instance transactors must "buy off" all third
 parties (De Vany et al. ,  1969).  The definition of waste relocation rights
 avoids both extremes by allowing third parties to be subjected to pollu-
 tion levels up to the  P^ threshold.   As long as the PJJ limit is not ex-
 ceeded,  exchanges between two parties are not  severely constrained and
 the extraordinarily high transactions costs and  enforcement costs which
 usually accompany exchanges involving a large number of transactors
 can be avoided.  When the effect of the exchange is to exceed PJJ, how-
 ever, third parties must be brought  in and compensated.

   Waste relocation rights are, by the terms of the property definition,
^e jure covenants expressly tied to the ownership of land.   Therefore,
 exchanges of waste relocation rights would be similar to exchanges of
 land with  the exception that the transactors would be  exchanging rights
to alter (either improve or pollute) the quality of waters beneath the land
without necessarily purchasing the land itself.
                                 75

-------
   A landowner's sale of waste relocation rights to a party who seeks to
alter the quality of groundwater is exactly analogous  to a mineral right.
For example, an oil company may buy or lease subterranean rights to
drill or explore without  buying the parcel of land itself.   All the driller
purchases is the owner's promise with respect to one use of the land.
Exchanges of this  sort occur commonly in spite of the large number of
landowners with which an oil company may have to deal,  especially in
cases of slant drilling.

   Land transactions may serve as a model for exchanges of waste relo-
cation rights.  Two primary types of exchanges will occur:  restructuring
all or portions of waste  relocation rights without exchange of the land it-
self; and combining or subdividing rights as the result of the sale of land.


RESTRUCTURING WASTE RELOCATION RIGHTS*

   Consider  the simple  situation of four landowners depicted in Figure
30 where the initial level of groundwater chloride content for each is set
at 300 mg/1.  Assume that landowners A, B, and C are farmers but
that D begins to drill for oil and seeks to dispose  of his brines via an
unlined disposal pit.   Knowing that the flow gradient in an underlying
aquifer is from east to west and that all pollutants are traceable at low
costs, D purchases B's  rights (but not B's land) to create up to 300
mg/1 of chloride pollution at any point beyond B's land.   When B's rights
are combined with D's initial right to achieve up to 300 mg/1 of chlorides
at any point  beyond the boundaries of D's land, one might conclude that
D has  now obtained the  right to create up to 600 mg/1 of chlorides  in the
groundwater in the vicinity of the B/D boundary, but this is  not necessarily
                                              D
             Figure 30.  Land parcel diagram, chloride pollution
                       example (Pc| = 300 mg/1).
 *The analysis of exchanges of waste relocations rights in this and the
  following subsection is based on the model of exchanges of radiation
  rights developed in De Vany et al. (1969).
                                  76

-------
  so.  Purchasing B's right to create 300 tng/1 of chlorides beyond B's
  land automatically  assigns to D the initial duty of B not to achieve more
  than 300 mg/1 of chlorides in the waters underlying anyone else's land.
  Unless he deals with A, D has not obtained the  right to impose more
  than 300 mg/1 of chlorides west of the A/B boundary.   A's protection is
  maintained in  spite of the exchange between D and B.   Thus,  in order
  for D to maintain his duty toward A, he may have to limit his brine con-
  centration to substantially less than 600 mg/1 of chlorides in the vicinity
  of the  B/D boundary.

    Similarly, the exchange between  B and D does not relieve D from
 maintaining the initial protection of the water underlying C's land from
  either D's or J3's  land.  D's brine disposal activities after the purchase
 of B's  rights must take cognizance of his duty toward _C as well as B's
 toward C,  J3's  toward A, and his own toward A.  If any of these duties
 are violated  even  slightly, D could face a lawsuit.  Again, an analogy to
 land law is apt.  If D purchases  from B the right to build a fence a few
 feet to the west of the B/D boundary for its full  length,  this sale does not
 change in any way the position of the A/B.  B/C. or C/D boundaries.

    Several conclusions follow from  even this  simple example of exchang-
 ing rights apart from the sale of land.   First, in order to make room for
 his brine plume by purchasing B's waste relocation rights, D will want
 to  learn the direction of flow in addition to other characteristics  of the
 aquifer and perhaps to model them.   Second,  D will monitor  groundwater
 before  and after the exchange occurs in order to make certain that he is
 not exceeding his own or B's duties  toward other owners of land.  Third,
 the transaction between B and D  does not require the consent of A or C
 unless  their initial rights of protection against more than 300 mg/1 of
 chlorides from either B or D  are violated by the transaction.  Transactors
 are allowed some latitude before having to deal with third parties such as
 A and C, but the extent of that latitude is determined by the initial P^-.
 At  least to  this  extent some transactions including many participants,
 with their high  transactions costs, are avoided.

    The  preceding analysis also applies to two other subcategories of ex-
 changes which amount to combining portions of waste relocation rights.
 In the first subcategory,  B might agree to sell D the  rights applying to
 the easternmost one-fourth of B's land.  This would permit D to achieve
 a maximum of 300 mg/1 of chlorides  at a new waste relocation rights
 boundary (represented by the dashed  line of Figure 30} lying one-quarter
 of the distance westward from the B/D land boundary.  As with exchanges
 involving sales  of entire waste relocation rights, sales of portions of
waste relocation rights would not allow B or D to violate their original
duties toward A and C.  Indeed, the contract between B  and D would
                                  77

-------
provide for D's right to monitor on B's land up to the dashed-line boundary
so that D could make sure that he adhered to his obligation toward B not
to exceed allowable levels of pollution.  B would continue to monitor on
his side  of the boundary, and both A and C would continue monitoring along
the edges of their lands to make sure their rights also were not violated
by the exchange.  D's entire monitoring and modeling effort would indi-
cate by how much he could increase pollution in the waters to the west and
south of  the dashed-line boundary.   This would influence the  price that D
would be willing to pay B for this portion of B's waste  relocation rights.

   The second subcategory of exchanges between B and D would be retain
the initial solid-line B/D boundary  but to increase the  allowable  amount
of chloride pollution which D could  release into the waters underlying B s
land to,  say,  400  mg/1.  However,  this  would not relieve B and D of their
respective duties  to achieve no more than 300 mg/1 of chlorides  in the
waters under A's  or jC's land unless the permission of A or C were ob-
tained.  Again, some modeling and monitoring would be undertaken by
both B and D  before entering into this contract in order to protect them-
selves from liability toward A or C.

 SALES OF LAND WITH WASTE RELOCATION RIGHTS

   The waste relocation rights regulatory system allows  rights to be ex-
changed  separately from the sale of land.   But assume now that the land
is sold in its  entirety with the original waste relocation rights intact.
Assume also  that, owing to  the long-lived and slow-moving nature of
groundwater pollution, a violation is not discovered until substantial time
has passed after the pollution-causing activity occurred.  If  A sells his
entire land to Z and C finds a violation of more than 300 mg/1 of chlorides
some time after this  transaction, two questions arise.  First, if liability
is assigned, whether it should go to A or to Z, and whether C would sue
A or Z or both.  Second, whether liability should be assigned at all and
•whether C may sue anyone.

    The critical issue for the first question is  that the liability should be
assigned fully to just one of the two parties (De Vany et al. ,  1969). Divi-
 sion  of liability between the two is  bound to be arbitrary in some degree
and would promote controversy.  This raises transactions and litigation
costs.  A more efficient solution would  be to assign full liability to just
one party according to either of two legal rules.  The  first rule,  "last
in time  is first in right, " would place the full liability on A which he
 could either retain or shift  to Z at a cost to himself.  The second rule,
 "first in time is first in right, " would place the full liability on Z which
he would take into account as he monitored for existing pollution levels
 and decided on his offer price to A.  As before, Z  could  contract the
burden of liability back to A (at a price) or keep it.
                                   78

-------
     Each of these legal rules has  similar implications for groundwater
  monitoring.  Assume that A originally sold his land with waste reloca-
  tion rights to E who eventually sold out to Z.   Assume also that the Pj-
  had not changed  between sales.   Under the "last in time is first in right"
  rule, A would retain liability  until A either died or was dissolved (de-
  pending upon whether A is a person or a firm).  Liability would then pass
  to E even though Z currently owned title to the land and waste relocation
  rights.  Since C has incurred  a violation of his waste relocation rights
  he would have recourse against either A or JE, who would regularly moni-
  tor the groundwater  under what was once A's  land to make sure that C's
  protection was maintained (or to  determine for himself the extent of the
  liability to  C).  Under the "first in time is first in right" rule,  Z would
  be liable for the  violation of C's waste relocation rights.  In this case,
  Z would monitor before purchasing the land from E to determine his
 potential liability to C and other rightsholders and adjust his offer price
  accordingly.

    Thus, monitoring would be  undertaken under either  legal rule.  Under
 the "first in time" rule, it would  be done at Z's expense.   Under the
  "last in time" rule,  it would be performed at  A's or E's expense.  More-
 over,  there does not appear to be any difference  between the quantity of
 monitoring undertaken under the two rules as  long as the level of PJJ
 protection  afforded to C and others remains unchanged.

    The answer to the second question—whether C may sue anyone —in-
 volves  considerations of equity more than economics.   If C were given
 no protection at law  from pollution by others,  the pollution might still
 be reduced or removed if this were worth more to C than the payment
 necessary to get Z to eliminate the pollution source or to take other cor-
 rective action. If transactions costs were  low, the optimal allocation of
 resources would be achieved regardless of whether C were given express
 protection. * Further, monitoring would probably be unaffected  by a fail-
 ure to give C legal protection.   Rather than A  or  Z monitoring to avoid
 violating C's rights,  C would incur monitoring costs to discover the ex-
 tent of the pollution harm and the party with whom he could bargain to  re-
 duce or remove it.

   However, if it  is desirable on grounds of equity to give protection to
 C, there is reason to choose the "first in time" liability rule.  This is
Because of the longevity of groundwater pollution.   Since C may incur
pollution damage long after causative action by A or his assignees, as-
signing liability to Z  on the basis of the most recent ownership assures

*See Coase (I960) and Section HI,  page 24,  "Achieving the Efficient
 Solution. "
                                 79

-------
that _C will always have legal recourse for damages.  To free Z of lia-
bility against C could mean that C would have no recourse if A  and other
intervening owners of that land prior to its acquisition by Z were either
deceased or difficult to locate.  Also,  assigning liability to the most re-
cent owner  avoids the cost of detailed  title searches  to discover which
previous owner bears liability.  Since the "first in time" rule has no off-
setting increase in monitoring costs, it appears to be preferable on eco-
nomic grounds.

   Different possibilities arise if A sells only the southern half of his
land to Z.  C might incur pollution from both A and Z.   A and Z might
divide A's original waste relocation  rights and duties in a fashion to as-
sure that C's protection right was not violated.   But  if the exchange oc-
curred and  C's protection right were violated, then according to the
"first in time"  rule, C would sue Z first.  If modifying or shutting  down
Z's pollution-causing activities were insufficient to achieve C's original
level of pollution protection, then C  could sue A too.   As before, it
would be undesirable to allow C initially to sue  both A and Z together,
and thus reduce the total pollution in the waters beneath C's land by ap-
portioning the reduction between  A and Z by  some method (such as  by
the fraction that the size of  A's and  Z's  land bears to the total  area).
While various methods of apportionment are possible, they would un-
doubtedly raise transactions and  litigation costs.

    Now assume that A sold only the  southern half of his  land in four por-
tions to  Zj_, Z2, Z3, and Z4.   Any subsequent violation of C's  right to
protection against more than 300 mg/1 of chlorides would give  C the
 right to  sue any one of the four, or all of them together, as having vio-
lated A's original duty of not raising the chloride content above 300 mg/1
in the waters underlying C's land, which duty both A and the_Zs assumed
after the exchange.   If a suit against one or  more of  the Zs were insuf-
ficient to reduce the pollution under C's land to under 300 mg/1 or  less
of chlorides, C would ultimately  have recourse against  A as well under
the "first in time" rule.  However many Zs  there may eventually be,
 C would have the right to  join one, two, or all of them in a lawsuit to
protect his rights.  If necessary, C could additionally sue A or whoever
might have subsequently purchased the northern half of  A's land, but this
 could only be done according to the priority  sequence established by the
 "first in time" rule.

    By the same token, if C  subdivided his land  and one or all of the Cs
 experienced a pollution violation, the Cs would have the same  legal right
against chloride pollution in excess  of 300 mg/1 as C himself initially
had against A,  Z,  or any of the Zs.   It is conceivable in this case that
 all the Cs would join together and sue all the Zs (  if  necessary including
                                  80

-------
  A) in a single action rather than each of the Cs suing each of the Zs sep-
  arately.  This outcome is likely since transaction costs would probably
  be lowered (De Vany et al. , 1969).

     If numerous  exchanges and subdivisions of land occur with the passage
  of time, the procedures for enforcement will  grow increasingly complex.
  But this should  not be taken as an  argument against creating a waste re-
  location rights system of groundwater quality regulation.  The increasing
  complexity which may  result constitutes evidence that the rights to allo-
  cate land,  groundwater, and allied resources  to steadily higher valued
  users and uses is being accomplished through ordinary market processes.
  In spite of the apparent complexity, however,  protection is  granted in
  each case by two relatively simple legal rules.  First,  sales or subdivi-
  sions of land are allowed provided that they do not create pollution in the
 waters underlying any other owner's land which exceeds the pollutee's
 original right of protection (before sale or  subdivision) unless he con-
 sents.   Second,  whenever  a violation occurs, the pollutee has legal re-
 course against the most recent acquirer of land; that is, priority for pro-
 tection from liability is given to the owners who are  "earliest in time. "


 WASTE  RELOCATION  RIGHTS AND NONLANDOWNERS
    Most of the foregoing discussion of the creation of a waste relocation
 rights regulatory system—the  rights,  their enforcement, and their ex-
 change—has been limited to owners of land  or their assignees,  since
 most polluters and pollutees would be landowners or  lessees of land.
 These owners or lessees would include nonprofit associations as well as
 the United States Government,  all of which would be given every measure
 of protection under the waste relocation rights regulatory system that
 for-profit owners of land received.

    The waste relocation rights  system also  extends protection to individ-
 uals and groups that  do  not own land.   Since waste relocation rights are
 tied to land as covenants which may be sold separately from the land,
 landowners could transfer those rights to individuals  or groups  who seek
 not to pollute the qualities of groundwater but to preserve and protect
 them.  Thus, environmental groups having a keen interest in improved
 water qualities might buy up important waste relocation  rights and hold
 them "idle" for any use  except pollution.  This would be analogous to the
 Sierra Club and similar organizations now buying choice redwood stands,
 "using" these resources for such activities as recreation and visual en-
joyment  rather than for  lumber products.
                                 81

-------
AN APPRAISAL OF  SOME  POTENTIAL PROBLEMS
   In order for the system of waste relocation rights  outlined in this sec-
tion to achieve the optimal level of pollution,  it must cope with several
problems at a cost which is relatively low compared to achieving the op-
timal level of pollution control by other means  such as discharge permits
or effluent charges.

   In general, the problems caused by the characteristics of the ground-
water resource present  no greater challenge to the planner of the waste
relocation rights model  than to the planner of the discharge permit model
or the effluent charges model, providing that each planner has the same
goal in mind; namely, achieving the economically efficient use of resourc
and thus the optimal level of groundwater monitoring and pollution.  This
subsection analyzes the  viability of the waste relocation rights scheme in
light of (1) the  large number of transactors who would buy  or sell rights;
(2) the difficulty of monitoring groundwater pollution  and of tracing it to
the  correct source; (3) the difficulty of predicting the movement of ground-
water pollutants;  (4) the longevity of some  of these pollutants; and (5) the
potential for some pollutants to combine synergistically.  Each of these
problems presents obstacles (costs) to the effectiveness of the waste re-
location rights scheme.   But the crucial question is whether these prob-
lems are more easily managed through waste relocation rights versus
discharge permits or effluent charges.

The Large Number of Transactors

    Defining rights to relocate wastes in terms of the overlying ownership
of land raises  the question of whether myriad owners of land or waste
relocation rights  would  greatly increase the costs  of exchange.  If the
 cost of exchange  often exceeded its value,  then rights would not usually
 be allocated to their highest valued use and the economic benefits of the
waste relocation  rights  system would be lost.

     There are two reasons, however, that the large number of transactor
 should not hobble the exchange process.   First, when two or more par-
 ties exchanged waste relocation rights,  they would be required by law to
 negotiate with (compensate) only those third parties  who would be ad-
 versely affected by the  exchange; that is,  those who  would incur pollution
 in excess of their Py protection rights.  This constraint would limit the
 number of parties that  must legally be brought into any single transac-
 tion and reduce the potential number of expensive multisided exchanges.

     Second, the number of owners of land or waste relocation rights would
 be greater than the number that would use land to pollute groundwater.
 (Analogously, oil companies seeking to drill under land that they do not
                                   82

-------
  own must purchase mineral leases only from the relatively few owners
  whose property would overlay the path of drilling. ) The number of trans-
  actors would also be reduced since a municipal corporation (a city gov-
  ernment) would be considered an "owner of land" as either a polluter of
  groundwater or a pollutee.  Thus, defining groundwater waste relocation
  rights in terms of the overlying ownership of land should not increase the
  number  of participants beyond those involved under a discharge permit
  regulatory system.   The same consideration applies to the effluent charges
  regulatory system.   The agency charged with setting effluent fees  would
  have to survey and analyze a large number  of economic  activities in order
  to ascertain the optimal degree of pollution and the fee associated  with it.
  Similar  startup costs are involved with the waste relocation rights regu-
  latory system in setting the initial Py. But once the initial PJJ have
  been determined,  the market system  is relatively self-regulating at little
  public expense provided exchange costs are kept low.  This is not  so for
  either the discharge permit or the effluent charges systems.

    The self-regulating character of the waste relocation rights  regulatory
  system should be  emphasized.  Although the initial PJJ would  be set gov-
 ernmentally, they would be modified toward optimal levels through volun-
 tary exchanges.  This process would be similar to the manner by which
 parcels of land are exchanged, subdivided, and recombined into more val-
 uable configurations.  The price of waste relocation rights would fluctuate
 like prices of other resource rights, responding to forces of demand and
 supply.   The prices would be established through competition among vari-
 ous groups in the marketplace.  For example,  environmental, recreation,
 and water-quality  organizations would have incentives to bid up the price
 of waste  relocation rights to use aquifers that are valuable for their pur-
 poses.  They would "use" these rights  by holding them idle—in effect,
 preventing pollution by bidding them away from pollution-causing uses
 and users and reserving them for conservationist goals.  Other groups
 such as industrial  waste disposers and municipal sewage systems would
 bid for waste relocation rights in order to inject pollutants into ground-
 water up  to the allowable PJJ limits.  Whichever individual or group
 had the greater demand would offer the owner of the waste relocation
 rights the highest price,  suggesting that this would be the highest-valued
 social use of these particular rights.   The advantage of the waste reloca-
 tion rights  regulatory system is that the quantity of pollution and ground-
 water quality monitoring would be  determined through impersonal market
 forces rather than  through the favoritism or lobbying that might accom-
 pany an administrative regulatory  system.

   The large number of transactors in the waste relocation rights scheme
offers two additional advantages.  First, there would be so many owners
of rights that  the possibility for monopoly gains from their sale would be
                                 83

-------
rare.  The main economic benefit accruing to owners of large parcels of
land or waste relocation rights would be greater flexibility in planning
their pollution-causing activities because of the greater possible distance
from the point source to the boundary at which pollution could not exceed
the legal P£J maximum.  The  second advantage in having a large number
of transactors is that a detailed administrative  scheme or auction to dis-
tribute the rights to individuals would be unnecessary.  This is because
the creation of these rights would be tied to the existing ownership of
land.  The only requirement for getting the system into operation would
be a tract index of land ownership and the hydrological-economic infor-
mation necessary for setting initially usable Pj.- limits on pollution.


Traceability

   Under the hypothetical waste relocation rights regulatory system, it
is essential that a pollutee be  able,  at relatively low cost,  to trace the
source of the pollution plume to the offending landowner or his assignee.
But tracing groundwater pollution to its precise source  can often be ex-
pensive.  Monitoring  along the edge of the pollutee1 s property may be
insufficient to establish a valid damage claim in court,  especially if the
point source is several miles  away.  Models of groundwater pollution
movements may not be sufficiently accurate to determine the exact source.
Visual inspection of man's activities —e. g.,  an oil company's brine dis-
posal pit being found "up plume" from a farmer's contaminated well—may
give some clues as to possible sources.  Such clues,  however, might not
support a court injunction to move the disposal pit, to otherwise dispose
of the  brine,  or to compensate the owner of the farm  land.

   Monitoring is intrinsic to the problem of tracing groundwater pollution.
But the difficulty of monitoring is not unique to a regulatory system based
on waste relocation rights.  The success of any regulatory  system for
pollution and monitoring—including one based on either discharge permits
or effluent charges—ultimately rests on the costs of detecting, tracing,
and predicting groundwater pollution. Thus,  monitoring technologies,
costs,  and strategies are critical to any regulatory system which may be
adopted.  The workability and transactions costs associated with any
chosen regulatory  system,  including waste relocation rights, will be only
as sound as the quality and cost of the supporting monitoring and modeling
techniques.


 Predictability

   The question of the predictability of groundwater pollution is a sepa-
rate matter from its traceability to a source.   The level of pollution will
vary over time with  rainfall and  other factors.   Although pollution may
                                  84

-------
 vary around some central tendency, it will not always be inexpensively
 predictable.  Yet the property definition sets a flat nonstochastic limit of
 PJJ, not allowing for seasonal or other variations.  Any instance of ex-
 cessive pollution,  some amount greater than Pji, would be a violation of
 duty owed by one rightsholder to another.

    This structure of the right is neither unrealistic nor inefficient for at
 least three reasons (De Vany et al. ,  1969).  First,  the constraint is a
 maximum,  not a specific requirement.  Second,  pollutees would prob-
 ably not incur the costs of court action if they expected the increased
 pollution in excess of the legal PJJ maximum to be minor or short-lived;
 small or infrequent violations, even if they endure for some time,  are
 unlikely to raise a protest.  Third, and most important,  to write a sto-
 chastic definition of pollution into the property right—for example,  that
 the maximum Pj,- could not be exceeded by more  than 5 percent during
 10 percent of the time— would  raise the cost of enforcing the right.   How-
 ever, nothing in the structure of the right prevents the polluter and the
 pollutee from voluntarily making a contract providing for a stochastic
 limit  to pollution.

    In  sum, it would appear that the problems of the predictability and
 traceability of groundwater pollution would become manageable,  i. e. ,
 sufficiently inexpensive, in the waste relocation rights model only if the
 engineering and hydrological sciences  provide  reasonably accurate  and
 reliable groundwater pollution quality models.  However, this is neces-
 sary for monitoring and pollution to be reduced to economic levels under
 any regulatory system.

 The Large Number of Pollutants

    The number of potential groundwater pollutants is large and the types
 vary tremendously— from chlorides to  chromates  to coliform bacteria.
 Each may harm aquifers differently depending upon the nature of the
 aquifer and the use to which it is put (e. g. ,  a rice crop is more sensi-
 tive to salt pollution than is cotton or soybeans). Determining initial pol-
 lution levels when the number of pollutants is so great and the conse-
 quences so varied may be a difficult problem, but it is not unique to the
 regulation of groundwater pollution through a waste relocation rights
 scheme.  Any system of pollution regulation— and the monitoring proce-
dures  implied by it— must set some permissible levels for each pollutant.
 These decisions are made differently under each of the three regulatory
systems.

   Under the discharge permit regulatory system, maximal levels of in-
jections are set for each type of pollutant for each polluter,  or permittee.
This decision is similar to the maximal limits of P  that would initially
                                  85

-------
be set under the waste relocation rights  regulatory scheme.  Hence, the
large variety of pollutants which could make the waste relocation rights
system complex and unwieldy also could cripple the discharge permit
regulatory system.  The same type of decision—setting some level of
allowable pollution for a variety of substances and organisms—would be
necessary under each regulatory system.  However, the waste relocation
rights regulatory  scheme is superior in this regard to the discharge per-
mit scheme.  The initial limits for pollution rights must be usable and
valuable for exchange but do not need to be optimal; rights  can be ex-
changed and recombined to maintain a continuous  optimum  over time.
Since the discharge permit system does not provide for exchanges of per-
mits or voluntary and gradual adjustment of pollution levels, greater care
must be taken under this regulatory system to set the initial maxima fair-
ly close to optimality.  Second, -the optimum under any system will tend
to change over time as populations,  demands,  technologies,  and climates
change.

   A  similar relationship exists between waste relocation rights and the
effluent charges regulatory system.  The agency  charged with controlling
effluents would, in all probability, have to tax not just effluent per se  but
different effluents differently.  Again, this requires a decision similar
to setting the initial P^ of  the waste relocation rights scheme in addition
to setting the effluent charge expected to result in the desired quantity of
each effluent.  As with the discharge permit regulatory system,  the efflu-
ent charges model requires an administrative agency to determine changes
in the fee as populations and other circumstances change.   This  discrete
administrative decisionmaking process also compares unfavorably to  the
gradual decentralized decisionmaking process associated with the waste
 relocation rights  scheme.

   Although the initial Pji under a system of waste relocation rights
 should be determined only  after careful hydrological-economic  studies,
one might tentatively conclude that to  be economically usable they usu-
 ally should be slightly in excess of existing pollution levels for  aquifers
that are relatively unpolluted.  PJJ slightly in excess of existing levels
would give landowners something to exchange and would reduce monitor-
 ing costs.  The PJJ could be set at less  than existing levels so  as to be
 on the  safe side,  although  this is  not recommended since it would impose
 large capital losses  on many landowners by forcing them to clean up
 underlying groundwater.  It could also raise monitoring costs drastically.

 Longevity of Pollution

    Liability for groundwater pollution is complicated under any regulatory
 system by  the historical distribution of pollution. Pollutants often move
                                   86

-------
 slowly through aquifers and may not cause harm until years later when
 the firm or individual who injected the pollutants is dissolved or dead.

    Under the discharge permit regulatory system, protection can be pro-
 vided for pollutees only by setting maxima for injections and monitoring
 inputs accordingly.   Where the future consequences of pollution are ex-
 pected to be especially severe, the discharge permit system would re-
 spond by lowering allowable inputs further and monitoring even more care-
 fully.  This response is open to several criticisms.  First,  the discharge
 permit system is cumbersome administratively.  In general, it would not
 produce the correct value of monitoring information regarding the benefits
 and costs of decreased (or, for that matter,  increased) pollution.  There-
 fore,  it would generate the optimal degree of pollution only by coincidence.
 Second, it is possible that more pollution would result than the administra-
 tive agency intended.  This additional damage could be the fault of poor
 groundwater models, careless monitoring,  or conditions beyond the agen-
 cy's control (e. g. , unpredictable rainfall).   Third, the discharge permit
 regulatory system provides no mechanism to compensate pollutees for
 any damage incurred years later,  especially for damage which is greater
 than anticipated administratively.  As long as the polluter adheres to  the
 maximum injections allowed by his permit and monitored by the agency,
 the future pollutee has no recourse against either the polluter or the
 agency.   Fourth, the discharge permit system gives the polluter no  incen-
 tive to reduce pollution below the maximum level specified in the dis-
 charge permit.  Discharges below this level  are "free" to the polluter
 but they may be  costly to society at large, especially if the particular
 pollutants are long-lived.

   Most of these criticisms can also be  made against the  effluent charges
 regulatory system.  The polluter would  adjust his  effluent according to
 the specified tax, but the total amount of pollution  and the effluent charge
 associated  with it would both be set adminstratively.  The only means by
 which the future consequences of long-lived pollution could be taken into
 account would be for adminstrators to raise the taxes for these substances.
 But fees would be difficult to determine  administratively for a variety of
 pollutants  and difficult to change once set.  Under this  regulatory system,
 any cognizance for future damage must be taken administratively since
 the polluter would not be liable for future damage as long  as he paid the
 full tax associated with his particular emissions.  Pollutees would not
 be compensated for greater-than-anticipated damages resulting from an
 improperly selected  level of pollution or from imperfect groundwater
modeling and predictions.

   Under the waste relocation rights regulatory scheme,  some protection
is provided for future pollutees by linking liability to particular parcels
                                  87

-------
of land.  This is accomplished by two items of the waste relocation rights
detailed in this  section.  In item 1 of the right,  a polluter may not exceed
the maximum P^ unless these levels are changed through voluntary trans-
actions among rightsholders which do not harm third parties without their
consent or compensation.  Hence, liability for pollution in excess of the
level voluntarily agreed to is tied to the particular legally defined land
area.   This liability rule becomes a covenant that goes with the land as
part of its title  whenever it is sold.  The second element of protection to
future pollutees is  provided by item 3 of the waste relocation  right.  This
paragraph gives pollutees injunctive relief and whatever damages they
can prove in a court of law as having been caused by the pollution.  Since
courts  of law recognize the principle of discounting future income streams
to present value in assessing damages in personal injury  or death cases,
they could allow for  compounding past damage values to present values
years later in accounting for the damage proved to have been caused by
past pollution to groundwater.  The skill of courts in awarding accurate^
damages  from pollution would be closely related to the accuracy of moni-
toring and modeling techniques.

    The critical distinction among the three regulatory systems is that the
waste relocation rights regulatory  system provides some protection to
current rightsholders from historical pollution actions since the symmet-
rical right to pollute and the right to protection from pollution alike are
tied permanently to  the ownership or assignment of land.

    A prudent buyer of land could reckon with this potential  liability in two
ways.  First,  before purchasing the land, he might ask to monitor ground-
water to  determine existing levels.  In situations where the value of the
land or the potential liability is large,  he  might even hire consultants to
perform  some modeling to predict future levels.  This activity is exactly
analogous to the termite or structural inspection of a home paid for by
the buyer.  Second,  the buyer could engage in a title search in the cog-
nizant government agency's register for waste relocation rights and lia-
bilities in much the same way that an escrow or title insurance company
 is paid by prospective owners to guarantee the title of a piece of land or
 a house against debts or other encumbrances.  If the waste relocation
 rights title search revealed liabilities which were especially costly for
 the purpose to  which the prospective buyer expected to put the  land,  then
 he would lower or withdraw his offer price for it.


Transacting the Different Pollutants
    Although the foregoing analysis has, for simplicity's sake,  focused
 on transactions involving only one pollutant at a time, the actual number
 of pollutants is large.  While this might appear to greatly increase
                                   88

-------
 exchange costs of a waste relocation rights regulatory system, two forces
 would work to reduce these costs.  First,  it is likely that only a few pollu-
 tants—perhaps a dozen or less—would be important in any particular geo-
 graphical area,  For  example, a given level of boron and brines might
 ruin orange groves in Orange  County,  California, but would have little
 effect on industrial users around  nearby Long Beach.  This  effect would
 limit the number of pollutants  that pollutees would  seek protection from.
 The second factor for reducing transactions costs would be exchanges
 covering more than one pollutant  at once.  Nothing in the waste reloca-
 tion rights regulatory scheme  requires transactors to contract for just
 one pollutant at a time.  The monitoring, modeling, and legal  services  re-
 quired to consummate one exchange of waste relocation rights  for a single
 pollutant could be applied to others,  "spreading" the  lump-sum exchange
 costs over a variety of pollution possibilities.  The costs of monitoring
 actual pollution and of enforcing contracts could be treated similarly.


 SynergisHc Pollution
    This analysis has treated each pollutant as independent in the damage
 it causes; that is,  as if the damage caused  by pollutant P4 could be sepa-
 rated from the damage caused  by  pollutant  Pj20'   Nature, however, is
 not  so simple.  In some cases  groundwater pollutants  combine  into fresh
 chemical compounds which tend to neutralize the  harmful effects that
 might be caused by each separately.  In such cases pollutees would suf-
 fer less harm  than in the absence of one or more of the interacting pol-
 lutants.   Thus, if the limit on one pollutant were  technically violated,
 the pollutee usually would not take legal action if  this "violation" were
 neutralized by the presence of  some offsetting pollutant in the ground-
 water beneath  his land. On the other hand, two or more pollutants might
 combine to cause the pollutee more damage than either would cause sepa-
 rately.

   The pollutee is automatically protected to some extent by  the waste
 relocation right if each of the two  (or more) pollutants is generated by
polluter A and  if A exceeds the maximum legal limit for either  or both
pollutants'.  But the pollutee is  not protected by the  rights system, as it
has been outlined,  if A injects two chemicals  which combine  chemically
to produce damage to the pollutee, provided each  injection is within A's
waste relocation rights of P4A  and P12OA*  Protection of this kind would
require expanding the Pj- list of pollutants to include synergistic com-
pounds for A such as P4A and P12OA'   This modification is probably
feasible  although it would make the property system more complex.
                                 89

-------
   Consider next the situation where the TP^. found under the pollutee's
land is traceable to polluter A but the PI20 is traceable to polluter B.
The p'roperty  system gives the pollutee recourse whenever ?4A or
separately exceed the legal maximum, but this  recourse is limited to
either polluter in isolation. Allowing the pollutee to sue A and B jointly
reintroduces the problem of apportioning liability between the two defen-
dants, but this creates severe legal problems and reduces the possibility
of achieving economically efficient recourse.  In this respect,  the syner-
gistic combination of pollutants is similar to the problem of intermodu-
lation among different radio emitters operating on different frequencies,
again paralleling the analysis  of De Vany et  al. , (1969).

    Consider also the situation where the pollutee incurs damage even
though A injects P^A and B injects Pj20B in amounts that fall within
the legal maxima.   As before, the property  system gives the pollutee no
protection against this form of synergistic pollution.  This  could be change
by  extending the "first in time" liability rule to require in these cases that
the most recent polluter either modify his operations or shut down in orde
to restore pollution under the pollutee's land to legal levels.  However,
this modification is not  recommended for two reasons.  First, it would
greatly complicate the waste relocation rights system by requiring initial
levels for not only the P^- pollutants but for  combinations of them from
different polluters.  This might require modification of some of the rules
for making exchanges, and surely would complicate even the registry re-
quirements of the rights system.   The whole system would become less
flexible and transactions  costs would increase, especially if there were
some apportionment of liability among different polluters.  Second, it
might be difficult for courts to determine what  the critical cutoff  levels
of  synergistic pollution for each polluter would  be.  Therefore, it appears
undesirable a priori to give pollutees protection from synergistic pollu-
tion beyond that given where individual polluters violate their P^j con-
straints and the harms from synergistic pollution can be  traced to these
violations.

    The pollutee would still have options for  reducing synergistic pollution
even if the system of waste relocation rights did not provide formal pr°~
tection.   For example,  he could  pay A or B to  reduce existing synergis-
tic pollutants or to generate nonsynergistic  pollutants.  If the value of
preventing this damage were less than the cost of making and enforcing
contracts  voluntarily, the pollutee's land would probably go unused or be
allocated to other economic activities.  This result is not undesirable if
modifying the waste relocation rights  system to take into account syner-
gistic relationships between two  polluters were to raise transactions and
 enforcement  costs to the point of reducing overall economic efficiency.
                                   90

-------
 SITUATIONS INAPPROPRIATE FOR WASTE
 RELOCATION RIGHTS
    The waste relocation rights regulatory system can cope with most but
 not all pollution problems.  It can even handle some complex pollution
 situations at an increased cost, but several pollution and monitoring prob-
 lems appear to be beyond  its scope of application.

    Pollution involving large numbers of polluters and pollutees may not
 lend itself inexpensively to the waste relocation rights paradigm.   Where
 thousands of septic tanks pollute the water used by hundreds of farmers
 "down plume, " it would probably be uneconomic for  each farmer to sue
 each septic tank owner separately  even if pollution were traceable at low
 cost to its exact source.   The damage caused by each polluter would prob-
 ably be small (and hence his liability would be small) relative to the costs
 to the pollutee of proving damages. Moreover,  the transactions costs
 involved  in all the pollutees joining together to sue all the polluters might
 be great  relative to the expected pecuniary gains unless  each group were
 represented by a municipal corporation. The problem would be more
 tractable if both groups were represented by the same municipal corpora-
 tion, so that liability is settled through local bargaining  and political pro-
 cesses.

   In the  case where polluters or pollutees are unincorporated and the
 transactions costs of bargaining or taking legal action are  relatively high,
 the economic solution appears to be to assign liability by regulations
 rather than to rely on market exchanges to determine the optimal  amount
 of pollution and monitoring.  If economic efficiency is the goal of public
 policy, then the liability in such cases should be assigned by statute to
 the party who can avoid the damage at least cost.

   On the one hand, if it would be cheaper to  regulate the design of septic
 tanks to reduce damage from leakage, then the polluters  should legally  be
 given this responsibility and the costs should be  borne by them.  Placing
 the liability on the owners  of tanks  would save extra transactions  costs
 since pollutees would otherwise have to contract with tank owners  to re-
 duce damage.  In this situation groundwater monitoring would amount to
 checking  randomly on construction  standards  and leakages, probably by
 a public agency in rather the same  way that other building code provisions
 are enforced.  On the other hand, if fewer resources were consumed by
having pollutees either treat the water polluted by septic  tanks or adjust
their activities in some other manner, then economic efficiency would
command that they be given responsibility for adjustment.  Again,  fail-
ure to assign liability in this way would lead polluters to  waste resources
in contracting with pollutees to assume  liability.  In each of these  cases,
                                 91

-------
economic efficiency requires that liability be in ad e unambiguous and be
initially assigned to the least-cost avoider of the damage.

   The waste relocation rights  regulatory system could not be expected
to allocate resources  efficiently when pollution is not traceable at low
cost to its point source.  This matter can scarcely be overemphasized.
Monitoring and modeling for information  as to where pollution originates
is essential for efficient resource allocation under this  regulatory sys-
tem.  Corker's adage that the "effectiveness of law is limited by hydro-
geology"  applies equally to the  effectiveness of market processes in
allocating waste relocation rights  (Corker, 1971).

   The waste relocation rights  system would also be inappropriate for
pollution  problems associated with "common pools."  Consider the prob-
lems associated with  seawater  intrusion in the diagram depicted in Fig-
ure 31.  Pumping from A's well lowers the water table to the point where
A and B experience seawater intrusion, with B's salinity in excess of
the hypothetical legal maximum of 300 mg/1 of chlorides.  Whether B
has  recourse against  A depends upon the  interpretation of the phrases
"economic activities" and "originally degrade" in item  1 of the original
waste relocation rights definition.  A would be liable if the courts inter-
preted these phrases  to include a failure  by A to take into account the
effect of his pumping  on  seawater intrusion into the aquifer beneath B's
land.  This would be a logical interpretation of the right if B would not
have experienced pollution in the absence of A's pumping, which is con-
sidered to be an "economic activity. " Alternatively, this problem could
be handled under the body of law governing pumping from common pools
rather than the waste relocation rights regulatory system.
                                            B           \ OCEAN
                   Figure 31.  Seawater intrusion example.
    A final problem that may not be handled inexpensively by the waste
 relocation rights regulatory system is pollution caused by nonpoint sources
 (such as  sewer lines,  road salts, etc. ).  Assume that a road coincides
 with the A/B boundary in Figure 31 and that it is salted during winter to
 the extent that more than the legal limit of 300 mg/1 of chlorides is found
 in  the groundwater under B's land.  If the road is owned and salted by A
 or B there is no problem in placing liability provided the illegal salinity
 can be traced to the salting operation.  If the roads are municipally owned
                                  92

-------
  and salted by a public agency, then liability is  still clear, provided that
  it is politically feasible for public agencies to assume it.

     Since the pollution rights regulatory system gives protection to ground-
  water underlying publicly owned land,  it would  be logical to assign lia-
  bility to such public land in the same way  that symmetrical rights and
  duties are assigned to privately owned land.  If this  assignment of liabil-
  ity is not practical,  then landowners could be given protection against
  publicly caused road  salt pollution only by municipal regulations  or State-
  imposed regulations on municipalities that limited the quantity of salts
  in areas especially prone to saline pollution.


 CONCLUSIONS

    This section has compared and contrasted alternative pollution regula-
 tion systems and highlighted the relative attractiveness of the waste relo-
 cation rights system.  There is an important relationship between this
 discussion and the analysis of monitoring in the remainder of the report.

    Both the nature and extent of monitoring are  dependent upon the pollu-
 tion  regulatory system adopted.  The nature of monitoring and  the value
 of resources devoted to it will be different when regulation is by discharge
 permits rather than by waste relocation rights.  Both systems  involve
 monitoring for information  and for compliance.  For example,  monitor-
 ing by public agencies near sources of pollution would probably be more
 common under the discharge permit system; monitoring by private par-
 ties near the consequences  of pollution would be more common  under the
 waste relocation rights system.

   A virtue of the waste relocation rights regulatory system is  that moni-
 toring would be done by private parties  and would  be largely "self-enforc-
 ing" through the courts.  A  large bureaucracy, a large budget,  and large
 groups of "monitoring marshals" could be avoided while the Nation's
 groundwater quality would still be guarded  in most situations under this
 regulatory system.  Instead,  private parties would take steps largely in
 their own interest to protect their individual waste relocation rights which
 would be steps in the national interest as well.   Thus,  the  "privatization"
 of most monitoring would not  only reduce costs to  the general taxpayer
 and avoid the difficulties of centralized governmental  administration but
be more likely than any alternative to provide an economically efficient
degree of protection to the Nation's groundwater  supplies.
                                 93

-------
                            REFERENCES
Alchian, A., Some Economics of Property.  RAND Corp. paper P-2316,
   Santa Monica, California,  May 26, 1961.

Alchian, A. , and R. Kessel,  "Competition,  Monopoly, and the Pursuit
   of Money, " Aspects of Labor Economics. National Bureau of Eco-
   nomic Research, Princeton University Press, 1962.

Alchian, A. , "The Basis of Some Recent Advances in the Theory of
   Management  of the Firm, " Journal of Industrial Economics, Novem-
   ber 1965.

American Law Reports,  Annotated, Second Series (38 ALR 2d),  "Lia-
   bility for Pollution of Substream Water. "

Arrow, K. ,  "An Extension of the  Basic  Theorems of Classical Welfare
   Economics, " Proceedings  of the Second Berkeley Symposium on
   Mathematical Statistics and Probability.  Berkeley,  California, 1951-

Arrow, K. ,  "Economic Welfare and the Allocation of Resources for
   Invention," The Rate and Direction of Inventive Activity;  Economic
   and Social Factors, National Bureau  of Economic Research,  Prince-
   ton University Press,  1962.

Arrow, K. ,  and G.  Debreu,  "Existence of an Equilibrium for a Com-
   petitive Economy, " Ecqnornetrica. July 1954.

Barrett, L. E. , and T. E.  Waddell, Cost of Air Pollution Damage;  A
   Status Report. U.S. Environmental Protection Agency,  February
   1973.

Becker, G. , Economics of Discrimination.  University of Chicago Press,
   1957.

Becker, G. , "Crime and Punishment: An Economic Approach, " Jour-
   nal of Political Economy.  March/April 1968.

Becker, G. , Essays in the Economics of Crime and Punishment.  Na-
   tional Bureau of Economic Research,  1974.   (Contains Becker (1968),
   Stigler (1970), and Ehrlich (1973).)

Blower, B. , Water Quality Planning. Interstate Commission on the
   Potomac River Basin,  May 1973.
                                 94

-------
  Buchanan, J. , and C. Stubblebine, "Externality, " Economica.  Novem-
    ber 1962.

  Cheung,  S. ,  "Transactions Costs, Risk Aversion, and the Choice of
    Contractual Arrangements, " Journal of Law and Economics.  April
    1969.                                                 ~~~

  Coase,  R. H. ,  "The Problem of Social Cost, " Journal of Law and
    Economics. October  I960.

  Corker,  C.  (ed), Ground-water Law.  Management and Administration.
    National Water Commission, NWL-L-72-06, Final Report, October
    1971.

  Dales, J. H. , Pollution.  Property, and Prices. University of Toronto
    Press, 1968.

 Debreu,  G. ,  Theory of Value. Wiley, 1959.

 Demsetz, H.  , "The Exchange and  Enforcement of Property Rights, "
    Journal of Law and Economics.  October 1964.

 Demsetz, H.  , "Some Aspects of Property Rights, " Journal of Law
    and Economics.  October 1966.

 Demsetz, H.  , "Toward a Theory of Property Rights, " American
    Economic  Review. Papers and Proceedings,  May 1967.

 Demsetz, H. , "Information and Efficiency: Another Viewpoint, "
    Journal of Law and  Economics. April 1969.

 De Vany,  A. S. ,  R. D.  Eckert,  C. J. Meyers,  D. J. O'Hara, and R. C.
    Scott,  "A Property System for Market Allocation of the Electro-
    magnetic Spectrum: A Legal-Economic-Engineering Study, "
    Stanford Law Review.  June 1969.

 Dorfman,  R., and N.  Dorfman, Economics of the  Environment.  Norton,
    1972.

 Ehrlich,  I. ,  "Participation in Illegitimate Activities: A Theoretical
   and Empirical Investigation, (1 Journal of Political Economy, May/
    June 1973.

 Ehrlich, I. ,  "The Deterrent Effect of Capital Punishment, " Ameri-
   can Economic Review.  June 1975.

Furubotn,  E. , and S,  Pejovich, "Property Rights and Economic
   Theory: A Survey of Recent Literature, " Journal of Economic
   Literature, December 1972.

Knodell, A. ,   "Liability for Pollution of Surface and Underground
   Waters. " Rocky Mountain Mineral Law Review. 1967.

Maler, K. G. , Environmental Economics.  A Theoretical Enquiry.
   Resources  for the Future,  1974.
                                95

-------
Meyers, C. , and A. Tarlock,  Water Resource Management, The
   Foundation Press,  1971.
Mishan,  E. , "The Postwar Literature on Externalities:  An Inter-
   pretative Essay. " Journal of Economic Literature,  March 1971.
Pinder,  G. F. , "A Galerkin-Finite Element Simulation of Ground-
   water Contamination on Long Island,  New  York, " Water  Resources
   Research.  Vol 9, No.  6,  December 1973.
Quirk,  J. , and R. Saposnik, Introduction to General Equilibrium
   ^Theory  and Welfare Economics,  McGraw-Hill, 1968.
Ridker,  R. ,  Economic Costs of Air Pollution, Praeger,  1967.
Stigler,  G. ,  "The Optimum  Enforcement of Laws, " Journal of Poli-
   tical Economy. March/April 1970.
Stigler,  G. , "The Economics of Information, " Journal of Political
   Economy,  June 1961.
Turvey, R. , "On Divergence Between Private Cost and Social Cost, "
   Economica. August 1963.
U. S. Council on  Environmental Quality, "Environmental Quality,
   1971. "  Second Annual Report.  1971.
U.S. Council on  Environmental Quality, Fourth Annual Report. 1973.
U. S. Environmental Protection Agency, Application for NPDES Permit
   to Discharge Treated  Waste Water to U.S. Waters. Region IV Ap-
   plication No.  AL 074 OYM  2 000570.
U. S. Environmental Protection Agency, Rehabilitation of Brine-
   Polluted Aquifers, Environmental Protection Technology Series,
   EPA-R2-72-014,  December 1972.
U. S. Environmental Protection Agency, Water Quality Strategy Pa-
   per,  Second  Edition, March 1974.
U.S. Public Law 92-500,  Federal Water Pollution Control  Act.
   S. 2770, 1972.
U.S. Public Law 93-523, Safe Drinking Water Act. 93rd  Congress,
   S. 433,  1974.
 Ward,  R. W.,  and D.  Vanderholm, "Cost-Effectiveness Methodolo-
   gies for Data Acquisition on Water Quality Management, " Water
   Resources Research, Vol 9,  June 1973,
 Williamson, O. , "Management Direction  and Business Behavior, "
   American Economic  Review.  December 1963.
 Williamson, O. , The Economics of Disciplinary Behavior;  Mana-
   gerial Objectives in a Theory of the  Firm. Prentice-Hall,  1964.
                                  96

-------
                               APPENDIX
                        CONVERSION FACTORS
   English to International System (SI) conversion factors for units used
in this report follow:
Unit
gallons per
minute
acres
inches
feet
gallons per day
per foot
feet per mile
feet per year
gallons per year
Abbreviation
gpm
not used
in
ft
gpd/ft
ft/mi
ft/yr
gpy
Multiply by
3.785412
0.404 686
2.540000
0.304 800
3. 785 412 x gal
0. 304 800 x ft
0.304 800 x ft
1. 609 340 x mi
0.304 800
3.785412
To find
liters0 per
minute
hectares"
centimeters
meters
liters per day
per meter
meters per
kilometer
meters per
year
liters per year
Abbreviation
Ipm
ha
cm
m
Ipd/m
mAm
m/yr
Ipy
Notes:
°One liter = 0.001 cubic meter
One hectare = 10,000 square meters
                                                  
-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
 REPORT NO.
 EPA-600-4
                            2.
                                                          3. RECIPIENT'S ACCESSION-NO.
TITLE AND SUBTITLE
 MONITORING GROUNDWATER QUALITY:  ECONOMIC FRAMEWORK
 AND PRINCIPLES
                                                          5. REPORT DATE
                                                           September  1976
                                                          6. PERFORMING ORGANIZATION CODE
 AUTHOR(S)
  Robert L. Crouch, Ross  D.  Eckert, Donald D. Rugg
                                                         8. PERFORMING ORGANIZATION REF

                                                           GE75TMP-71
 PERFORMING ORGANIZATION NAME AND ADDRESS
  General Electric—TEMPO
  816 State Street
  Santa Barbara, California 93101
                                                         10. PROGRAM ELEMENT NO.
                                                           1HA326
                                                         11. CONTRACT/GRANT NO.

                                                           68-01-0759
2. SPONSORING AGENCY NAME AND ADDRESS
  Environmental Monitoring  and Support Laboratory
  Office of Research and Development
  U.S. Environmental Protection Agency
  Las Vegas, Nevada 89114            	
                                                          13. TYPE OF REPORT AND PERIOD COVERED
                                                         14. SPONSORING AGENCY CODE

                                                            EPA-ORD
5. SUPPLEMENTARY NOTES
6. ABSTRACT
  Discusses the economic  considerations in selecting  an  optimal groundwater quality
  monitoring system.   Section I argues that poor specification of the property rights
  in groundwater is a  major cause of excessive pollution.   Section II examines
  groundwater adjudication and legislation and notes  that  government intervention
  through the authority of PL 92-500 will take the  form  of government-established
  and -enforced groundwater quality standards.  Section  III discusses the overall
  costs and benefits to society involved in groundwater  quality monitoring.  Section
  IV discusses monitoring needs for establishment of  quality standards and their
  enforcement, and develops a cost-benefit framework  for the analysis of groundwater
  quality monitoring.   Section V examines an alternative regulatory approach based on
  "waste relocation rights" for property owners.  These  rights would protect property
  owners' groundwater  from pollution by others through specifying allowable pollutant
  levels.  They would  be  transferable in the marketplace (like mineral rights) and
  enforcement of them  would be carried out in the courts.
17.
                               KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                             b.IDENTIFIERS/OPEN ENDED TERMS
                                                                          COSATI Field/Group
  Groundwater
  Economic analysis
  Management methods
  Water rights
  Underground water
  Water resources
                                               Economic framework
                                                and principles
                                               Monitoring groundwater
                                               Water pollution
                                               Water quality
    08H
    13B
    14A
18. DISTRIBUTION STATEMEN1

  RELEASE TO  PUBLIC
                                             19. SECURITY CLASS (ThisReport)
                                               UNCLASSIFIED
21. NO. OF PAGES
     104
                                             20. SECURITY CLASS (Thispage)
                                               UNCLASSIFIED
                                                                        22. PRICE
EPA Form 2220-1 (9-73)

-------