WATER POLLUTION CONTROL RESEARCH SERIES • 16110 EXW 06/71
              Factors Affecting
            Pollution Referenda
0.S. DEPARTMENT OF THE INTERIOR • FEDERAL_WAT,ER POLLUTION CONTROL ADMINISTRATION
 ENVIRONMENTAL PROTECTION AGENCY • WATER QUALITY OFFICE

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the results
and progress in the control and abatement of pollution in our Nation's
waters.  They provide a central source  of information on the research,
development, and demonstration activities in the Water Quality Office,
Environmental Protection Agency, through inhouse research and
grants and contracts with Federal, State,  and local agencies, research
institutions and industrial organizations.
                            /
Inquiries pertaining to Water Pollution Control Research Reports
should be directed to the Head,  Project Reports System,  Office of Re-
search and Development, Water Quality Office,  Environmental Protec-
tion Agency, Room 1108, Washington, D. C. 20242.

-------
        Factors Affecting Pollution Referenda
                       Abt Associates  Inc.
                         55 Wheeler Street
                Cambridge, Massachusetts  02138
                             for the


              ENVIRONMENTAL PROTECTION AGENCY
                       Project # 16110  EXV!
                       Contract # 14-12-902
                            June 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $2.50
                          Stock Number 5501-0110

-------
             EPA Review Notice
This report has been reviewed by the Office of
Water Programs, EPA, and approved for pub-
lication.  Approval does not signify that contents
necessarily reflect the views and policies of the
Environmental Protection Agency.
                  11

-------
                             ABSTRACT

Many local governments wanting to finance construction of water pollution
control facilities through the sale of municipal bonds are required to ob-
tain authorization from the electorate in the form of a bond referendum.
Since defeat of such a bond issue can result in significant delay in imple-
mentation of the Federal Water Pollution Control Act, it is important
that EPA understand the factors which affect the outcome of these elec-
tions.

A  sample of 40 bond  referenda for water pollution control projects (20
which passed  and  20  which failed) was  drawn from a comprehensive list
of recent elections.  Six were studied via in-depth field  studies,  and the
remainder were surveyed by telephone.  For  each case  study, at least
the folio-wing municipal employees were interviewed: clerk controller,
planner and engineer.  A representative of the media was also questioned.
Data were collected on variables,  which were analysed using correlation
analysis, contingency tables, and linear regression analysis.

Eleven of these variables  studied are able to explain 62% of the variance
in the results of the election: proportion of the population with family in-
come less than $3, 000; median  income; association of the project with a
pollution problem; rate of population growth; tax rate; dollar amount of
the proposed issue; repayment of the bonds by increased taxes; repayment
of the bonds by user  charges; pressure groups against the issue,  number
of criticisms  raised; partisanship  of groups.

The study contains conclusions  and recommendations about the process
by which the project  is planned, the role of the bond market,  and the
referendum as an approval mechanism for public expenditures.

This report was submitted in fulfillment of Contract Number 14-12-902
under the sponsorship of the Office of Water Programs, Environmental
Protection Agency.
                              111

-------
                    TABLE OF CONTENTS


     SUMMARY
I    PROBLEM DEFINITION
II    THE MUNICIPAL BOND ELECTION PROCESS
III   DATA COLLECTION AND ANALYSIS METHODOLOGY
IV   FACTORS AFFECTING THE OUTCOME OF THE
     REFERENDA                                        49
V    CONCLUSIONS AND RECOMMENDATIONS             79

     APPENDIX A - STATE CONSTITUTIONAL AND
     STATUTORY PROVISIONS CONCERNING BOND
     REFERENDA                                       101
     APPENDIX B - A LOGICAL PROCESS  MODEL
     OF THE MUNICIPAL BOND  ELECTION PROCESS     117
     APPENDIX C - MORE TECHNICAL ASPECTS OF
     THE METHODOLOGY                               121
     APPENDIX D - FILE  OF STATISTICAL DATA:
     CODE BOOK, DATA FILE AND INTERVIEW GUIDES   145
     APPENDIX E - IN-DEPTH CASE STUDIES            205

-------
                          SUMMARY

Nature and Scope of the Problem

The cost of water pollution control facing the nation is immense.   Recent
estimates of the cost of implementing the Federal Water Pollution Con-
trol Act during  the period 1969 to  1973 have been as high as $31 billion.
If present trends continue, the bulk of the burden for financing these
investments  will fall to local governments.

Most local governments finance construction of water pollution control
facilities through the issuance of municipal bonds.   However, because
of the statutory and constitutional  limitations on municipal indebted-
ness, many local governments wanting to sell bonds are required to
obtain authorization from the voters in the form of a bond referendum.
Defeat of a pollution control bond in a local community can result in a
significant delay in the implementation of the Federal Water Pollution
Control Act.  Hence,  it is important that the EPA understand what fac-
tors affect the outcome of municipal bond elections for such projects
so that it can assist local governments where possible and improve
the effectiveness of their own program.

No comprehensive  published data exist which relate the  total number
or amount of water pollution control bonds issued after  referenda to
the total water pollution control bonds issued.   However,  an estimate
made from data provided by the Investment Bankers Association of
America and corroborated by expert opinion  suggests that about 40%
of all bonds issued are subjected to a direct vote by the  electorate,
and about 15% of these fail.  Examination of the dollar amount of the
bonds subjected to  referenda   shows that the more costly projects are
more likely to be voted on, and also more likely to fail.  Since the
larger projects are located in more heavily populated areas, which are
also most seriously affected by pollution problems, these failures have
signigicant implications for EPA's program.

-------
Data Collection and Analysis Methodology

In order to develop policy-oriented recommendations for use by the
EPA and local municipalities, Abt Associates reviewed the entire
referendum process, identifying the actions taken by government
officials and others involved in the planning of pollution control which
might affect the outcome of the vote.  This process became the major
guide for identifying the study's data requirements.   One hundred
fifteen of these factors were selected for study and thus became the
basis for the design of  the interview schedules.

A sample of 40 bond referenda for water pollution control projects
was drawn from a  listing of the municipal bond referenda published in
the Daily Bond Buyer and compiled by the Investment Bankers Assoc-
iation of America.  In order to compare the attributes of the  referenda
that failed with those that passed,  an equal number from  each of  these
two categories was sampled.  For each case study,  at least the follow-
ing people were interviewed: city clerk, city controller or finance  dir-
ector, city planner, city engineer and a representative of the media.

When a referendum had been selected for study and measurements  made
of the variables of interest, then the data were ready to be organized
and analyzed.  A series of  statistical analyses were  performed,  each
one building on the results  of the previous  one.   The first step was to
examine the frequency distribution of each variable.  This provided a
general picture of  results and identified some significant trends  in the
data.  Then, the relationship between each variable and the outcome of
the referendum was examined by means of cross-tabulation.   Calcula-
tion and interpretation of various tests  of significance for the cross-
tabulations  of all of the variables with the outcome of the election identi-
fied 34 independent variables associated with success or failure  in a
non-random pattern.  This suggested that some, or even all of these
variables might explain the  variance  in the results of the bond elections.
The identification of these variables  directed the study to the next  phase
of analysis: linear correlation analysis.

-------
Linear correlation analysis was used to provide a single summary
statistic describing the strength of association between all pairs of
variables.  In addition, it was used to identify indirect relationships
and develop links between them.  The linear correlation analysis and
the cross-tabulations  identified a large number of referenda character-
istics highly correlated with the outcome of the election but not highly
correlated with one another — that  is,  independent of one another.  On
the basis of this information a linear  regression  model was developed.

The linear regression analysis had two major purposes: to clarify the
relationship of each independent variable to the outcome of the referenda,
and to develop an analytical model capable of explaining the outcome of
a referendum in terms of the percentage of votes in favor of the project.
The regression analysis  identified eleven variables which together ex-
plain  62% of the variance in the outcome  of bond elections for water
pollution control.

In addition to the analytical model developed to explain the maximum
amount of variance in election outcome,  a  second model was developed
using only a subset of the variables which could be measured well in
advance  of a bond election.  However, this equation explains only 46%
of the total variance in outcome and is, therefore,  not a practical
predictive model.

-------
Summary of Findings


Eleven variables were identified by the regression analysis which ex-

plained 62% of variance in the outcome of the bond election.  They can
be grouped into four categories: population characteristics, community

characteristics, characteristics of the bond  issue and campaign char-

acteristics.


Population Characteristics
             Proportion of population with family income
             less than $3. OOP.  Bond elections for pollution
             control projects  are more likely to pass in
             communities with a high percentage of low in-
             come families.   This  may mean that poor  people
             tend to vote in-favor of pollution control, but the
             data are not sufficient to warrant this conclusion.

             Median income.   This variable has a weak nega-
             tive correlation with outcome.   It is possible that
             it plays a role in a linear combination of variables
             which together account for  a substantial amount of
             the variance in the election results.
Community Characteristics
             Existance of a pollution problem.  Data indicate
             that this variable is  positively associated with
             approval of the bonds.   This suggests that com-
             munities which were able to capitalize on  current
             widespread concern for environmental quality were
             able to increase  the percentage of votes in favor
             of the issue.

             Population growth.  It was initially hypothesized
             that population growth would lead to pollution prob-
             lems and therefore be associated with the passage
             of the bonds.   On the contrary, population growth
             is associated with the failure of pollution control
             bonds, indirectly through a strong association
             with the number  of groups organized against the
             bond issue.

             Tax rate.  This  variable is negatively correlated
             with bond passage.  It appears that citizens  in
             communities  with high tax rates are  less likely to
             support bond  issues than citizens in communities
             with low tax rates.

-------
Issue Characteristics
             Amount of issue:  The amount of the  issue is
             negatively correlated with passage of the bonds,
             indirectly through an intervening variable: the
             existence of groups against the  issue with which
             the amount of the issue is  positively  associated.
             This suggests that the amount of the  issue may
             be important because it stimulates controversy
             which, in turn,  contributes to defeat.
             Method of repayment.  Two variables measure
             the method of repayment:  repayment by user
             charges and repayment by increased taxes.
             These variables are two of the most  significant
             in the regression equation. Both are highly
             correlated with outcome—user  changes positive-
             ly, and increased taxes negatively.  These  re-
             sults  suggest  that voters generally wish to avoid
             tax increases and therefore prefer the  user  charge
             method of repayment.  The user charge is also
             more equitable, in the sense  that those who  bene-
             fit from the improvement also pay for  it.
Campaign Characteristics
             Existence of opposition groups.  This variable
             emerged in the regression analysis as the most
             important election characteristic;  it accounts for
             41% of the  variance in election outcome.   The
             power of opposition groups lies in  their ability to
             create doubt in the minds of the undecided voters.
             When there is organized opposition,  there is little
             that proponents can do to make an  issue pass.  In the
             absence of organized opposition, it is unlikely to fail.

             jPartisansh.jp of groups  in the campaign.  Most of the
             local  governments in our sample of 40 cases are non-
             partisan governments.  In non-partisan systems, the
             emergence of a partisan group in a bond  election cam-
             paign tends to contribute to the defeat of  the issue at
             hand.

             Number of criticisms raised.  There is a significant
             positive correlation between the  number  of criticisms
             raised and the defect of a bond issue.  These projects
             which stimulate criticism and controversy are much

-------
more likely to be defeated.   Like the emergence
of opposition groups, criticisms tend to create
doubts in the minds  of voters who then  choose
not to spend money on a project of uncertain bene
fit.

-------
Conclusions and Recommendations

This study of the factors influencing the outcome of municipal bond
referenda has identified three major aspects of the  municipal bond pro-
cess which, if changed,  would increase the likelihood that local financ-
ing would be approved for construction of a pollution control facility.
They are the  process by which the facility is planned; the financial en-
vironment, particularly the bond market; and the approval mechanism
required.

The Planning Process.  The  major conclusion to be drawn from this
study is that bond elections for pollution control projects are won or
lost during the  initial planning of the project.  Critical decisions are
made at this time concerning who will benefit from  the improvement,
who will pay the costs,  and how they will be paid.  If everyone poten-
tially interested in the  construction of the project is not contacted at
the outset, or if other mistakes are made at this time, there is little
that the issue's proponents can do during the actual campaign to en-
courage passage of the bonds.

A decision about the  means of financing is an especially important part
of the project plan.   Analysis of the case studies reflects clearly the
voters' aversion to increased taxes and their desire for an equitable
repayment mechanism.  Furthermore, reduction in the amount of the
issue through the use of Federal or State financing seems to  improve
the chances of the bond's passage.

Abt Associates  recommends  that  the Environmental Protection Agency
assist local governments in planning water pollution control projects,
particularly those requiring a bond election.  The objectives of this
assistance would be to:
         •   Increase local officials' awareness of the impli-
             cations of their decisions  regarding the nature of
             the proposed project and the means of financial
             that project for the outcome of the election.

-------
         •   Increase local officials' awareness of the
             importance of consulting all interested
             groups before making these decisions.

         •   To provide local officials with the  information
             and insight necessary to make decisions which
             will be more acceptable to the majority of the
             citizens they serve.


Abt Associates  further recommends that this information be dissemina-

ted through a handbook for local government officials prepared by the

Environmental Protection Agency in cooperation with the state water

pollution control agencies,  and distributed by these state agencies to

local governments.  These  handbooks should be tailored to meet the

requirements of each state  and include the following items.
         •   A checklist of important community groups
             which might be concerned with water pollution
             control and sewer or drainage improvements.
             This checklist should also indicate in detail
             the interests of each group regarding any kind
             of water pollution control project.

         •   Complete and easy-to-use information on grants
             available from the State and Federal governments,
             including a description of the aims of each program,
             the amounts of assistance available,  eligibility re-
             quirements, and  instruction on where and how to
             apply,  and how to obtain qualified technical assis-
             tance where necessary.

         •   Descriptions of alternative  means of financing
             sewer and drainage  improvements,  with emphasis on
             more innovative schemes which are acceptable both
             to  the voter and the  ultimate buyer of the bond.


The Role of the  Municipal Bond Market.   The  second major influence on

the outcome of the bond election is the financial environment  in which

the bonds are formulated. Since  there are a very large number of mun-

icipal bonds and  relatively few buyers, the market is  extremely com-

petitive.  This permits buyers to be very selective about the bonds  they

buy.   Major factors in their choice  are the terms  of the indenture agree

ment between them and the seller, especially the method of repayment.

-------
The terms which are of advantage to the buyer generally operate against
the interests of the  seller: the municipality trying to raise capital for
its  pollution control project.  Hence, it would be to the advantage of
local municipalities to reduce market competition sufficiently to permit
small,  unrated towns  to sell  their bonds.  To accomplish this, Abt

Associates recommends the following:
              The Environmental Protection Agency should
              formulate a model indenture for municipal
              water pollution control bonds which would
              eliminate minor differences between  essen-
              tially like bonds and reduce some of the com-
              petition among  sellers.   This would enable the
              proponents to give more consideration to the
              preferences  of  the electorate when formulating
              the bond issue.

              The Environmental Protection Agency should
              investigate the  feasibility of a model  statute to
              establish State bond banks.  Since the states are
              limited in number,  carry up-to-date  credit infor-
              mation, and  have greater financial resources (and
              thus, flexibility in entering the market),  their
              bonds are more marketable.

              The Environmental Protection Agency should in-
              vestigate the possibility of developing new types
              of financing for water pollution control facilities
              which would  have characteristics attractive both to
              the voters and to the bond market.
The Approval Mechanism.   Many local governments must obtain the

approval of their electorates before they can issue municipal bonds for

water pollution  control projects.   For many municipalities, this require-

ment is imposed by state statuatory or constitutional law.  While it is

often desirable  to hold a referendum on an  important matter of public able

policy,  some local governments have the power to avoid this approval

mechanism and would like to do so.  Since the legal requirements can

be quite complex, they are often unaware of their opportunities and re-

quirements.  In order to expedite  local financing for pollution control

facilities,  Abt Associates recommends the following:

-------
•    Local government officials receive current
     information of all available means of obtaining
     financial support for pollution control.   This
     information should be a part  of the local govern-
     ment handbook recommended above.

•    Further, the Environmental  Protection Agency
     should initiate formulation of a model state water
     pollution control act which would expedite both local
     financing and  regional cooperation for water pollution
     control.  This act would expedite local  effort by sim-
     plifying the legal process  required for approval, and
    by providing opportunities for alternative approval
    mechanisms to the referendum.
                       10

-------
                             CHAPTER 1
                       PROBLEM DEFINITION
Problem Definition

The 1968 report entitled the Cost of Clean Water  estimated the  cost of
carrying out the Federal Water  Pollution Control Act during the  period
1969 to 1973 at $24 to $26 billion,  exclusive of the cost of separating
storm and sanitary sewers, and the cost of coping with agricultural
runoff, mine drainage,  oil pollution and other special problems.   In-
cluded in this total were the following:

             $8. 0  billion for municipal treatment works
             $6. 2  billion for sanitary sewer construction
             $2. 6  to $4. 6 billion for industrial waste treat-
             ment
             A maximum of $1. 8 billion for industrial cool-
             ing treatment
             $5. 3  to $5. 7 billion for operation and maintenance
             costs of municipal and industrial treatment facilities

Since this  estimate was made,  further study of the costs of municipal
                                                              2
treatment has  increased that estimate from $8. 0 to $10 billion.   Assum-
ing historical increases  in construction  costs, a more realistic total would
range  between $28 billion and $31 billion.  It  is clear from these projec-
tions that the cost  of water pollution control is immense.   The next ques-
tion which must be addressed is how this investment is to be financed.

If present trends continue,  the bulk of the burden for  the financing of
needed investments will fall to local governments.  Although the  Federal
government has undertaken many grant programs which provide  funds for
the construction of water pollution facilities,   Federal  contributions
account for only a  small portion of the total investment. One estimate,
made in 1966,  showed that the Federal share of the sewage treatment
                                 11

-------
projects with Federal assistance has been about 20%.  In those same
projects, the State share has been only 2%.   More recent information
is provided by the Federal Water Pollution Control Administration
which pointed out that, for every dollar provided to State and local
governments under its program of grants for treatment facilities,  there
has been a contribution of $5. 20 from State and local governments.

A projected investment of $31 billion over a five year time period will
require  significant financial commitment by local governments.  While
the Federal government has excellent estimates of total  investments
needed,  it has very little information about the problems facing local
government which, in the  last analysis, will determine whether this in-
vestment can or will be made.  One of the problems which has adversely
affected  water pollution control in the past is the failure of local com-
munities to raise their share of the  investment.
There are many constraints on local governments which affect their
decisions about investment in water pollution control facilities.  First,
there is the national financial climate: the interest rate, and in parti-
cular the  competition  of the bond market for the most attractive bonds.
Second, there is the financial situation of the local government itself,
which is especially significant in times of rising costs and stable or
declining  revenues. Coupled with this are local political  constraints on
municipal investments which reflect needs for other  services which
compete with funds that might otherwise be available for water pollution
control.   The level of contribution from the State and Federal  govern-
ments to projects of this  type is also influential.   Finally, there is  the
fact that water pollution cuts across political and geographic boundaries,
and the responsibility for controlling it is  often ambiguous, leaving many
problems unsolved.

Most  local governments finance construction  of water pollution control
facilities  through the issuance of municipal bonds.    However,  because of
                                12

-------
statutory and constitutional limitations on municipal indebtedness, many
local governments which wish to sell bonds  are required to obtain
approval from the electorate before  they may proceed.  This require-
ment dates back to the time  of the railroad boom in which many towns
in the West and Midwest invested.  When many of the railroads failed,
the towns through which they passed lost substantial sums of the tax-
payers' money.   To prevent this from happening again,  most of the
towns and states  which had been affected established  referenda require-
ments for all public investments.  Thus, in addition to the general con-
straints  on municipal finance mentioned above,  some local governments
also face the problem of a referendum.

Indeed,  for those governments which are required  to hold a bond  election
(and for  which a  successful referendum is the only route to the construc-
tion of water pollution control facilities),  it is  important for the Environ-
mental  Protection Agency to know what factors affect the outcome of
that election.  Hence,  the  tasks of this study were  to investigate:  What
factors account for voter response to bond referenda?  What strategy
should be followed to maximize the ability of local  governments to pre-
sent water pollution control  issues successfully to  the public?
                                 13

-------
 Extent of the Problem

 The extent to which bond election law affects the ability of local govern-
 ments to  issue bonds for water pollution control projects  has never been
 systematically studied.  Several different approaches might be taken to
 this problem.  One could compare the dollar amounts of bonds sold which
 were approved by referendum with the amounts of those sold without referen
 dum approval.   This would give the best overall picture of the magnitude of
 the problem.  A comparison of these figures with  information on the num-
 bers  of bonds issued with and without referenda would indicate if referenda
 were more often held with large or  small issues.  A second approach
 would be  to survey State constitutions and statutes for referendum require-
 ments.  This would identify the States with statuatory requirements, which
 are more easily amended.  However, because of the importance of local
 charter law, this would underestimate use of the referendum. A third
 approach  would be  to examine the frequency  of referenda held in recent
 years by  State.  This would point out the States which held a large num-
 ber of referenda, whether or not they were required under the State
 constitution or statutory law.  All three of these approaches have been
 taken in defining the role of the referendum in pollution abatement.  The
 results are discussed below.
Dollar Volume Extent of the  Problem.  No comprehensive published data
exist which relate the total amount of water pollution control bonds  issued
after referenda to the total amount of water pollution control bonds  issued.
However, an estimate of the extent to which improvements of this type
a subjected to referenda can be made from the annual reports of the In-
vestment Bankers Association of America.  Included in their reports are
annual data on the number and total dollar amount of bonds is issued
for water pollution control, and the number and dollar amount of bonds
subjected to referenda.  There is poor correspondence between these
                                14

-------
figures because bonds may be approved in one year and sold in a differ-
ent year when the market is more favorable.  In poor market years,
the amount of bonds voted will be much larger than the amount of bonds
sold.  Futhermore,  since the total bonded debt for water and sewer
projects is increasing, this figure tends to overestimate the use of the
use of the referenda as an approval mechanism.  As  a result,  the per-
centage of the yearly debt incurred for pollution  control that is  approved
by the electorate may appear to be more than 100%.  The data for 1966
through 1969 are summarized in the table on the  following page.

The number of bond issues subjected  to referenda in  a single year  appears
to be a more stable statistic than their dollar amount, and varies between
37% and 48%.   The dollar amount of bonds subjected to referenda is much
higher, with an average for four years of 77%.  This difference suggests
that larger bond issues are more likely than smaller ones to require app-
roval of the electorate.
                               15

-------
Number and Dollar Amount of Bonds Subjected to Referenda
Year
1966
1967
1968
1969
TOTAL
Bonds Issued,
Water and Sewer
Number
1,095
1,209
1, 154
828
4,286
*
Amount
$1,509
1,843
1, 854
1, 175
$6,381
Water and Sewer Bonds
Subjected to Referenda
Number
468
469
439
389
1,765
%
42
38
37
48
41
V
Amount
$1,729
678
1,309
1, 178
$4,894
%**
115
37
70
100
77
 ^Millions of dollars.
##The percent can be greater than 100% because bonds may be voted on in one
  year and issued in a later year when the market is more favorable.

-------
State Constitutions and Statutory Law.  The second approach was to sur-
vey the State constitutions and statutory law for relevant referendum re-
quirements.   These surveys,presented in Appendix A,  are  summarized
here.

                   State Constitutions&Referendum Law
             No provision                                31
             Referendum exempted either                 5
             in general or  with respect to
             sewer bonds.
             Referendum reqi ired only for               12
             general obligation bonds.
             Referendum required for general             2
             obligation bonds and revenue bonds
             Total                                       "50"

The significance of this data lies  in its implications for changing State
level referendum requirements.  Although statutory provisions can be
altered fairly easily, constitutioiBl provisions are much more difficult to
change.  Only 14 states have constitutional requirements for bond refer-
enda for water  pollution control projects.  This suggests that most
states could change their laws rather easily if they so desired.

State statutes were also surveyed and tend to  be more restrictive than the
constitutions.  Data were collected on the jurisdictions permitted to issue
sewer bonds, the existence of referendum requirements, the types of
bond possible,  and the limit on the interest paid, if any.

The tables on the following page  summarize the findings which can be
found in detail in Appendix A.

Together; the constitutional and statuatory data indicate that 44% of the
states are free from all referendum requirements.  Requirements,  if
any, appear in town county or authority charters.  Seventy percent of the
states provide for the issuing of some type of bond in some  type of juris-
diction without referendum.
                                 17

-------
              State Statutes:  Referendum Requirements
               	       by Type of Bond	
                   Type of Bond
# of States
           Allow general obligations only
           Allow revenue bonds only
           Allow both
           Total
     5
    45
    50"
              State Statutes:  Referendum Requirements
                      by Type of Jurisdiction
                Type of Jurisdiction
# of States
           Allowed to forego referendum
           in at least one type of juris-
           diction
           Never allowed to forego
           referendum
           Total
    35
    15
    50
Frequency of Referenda by State.   The third approach was to examine
the use of the referenda for water pollution  control projects by  state for
1966 through June, 1970.   Frequency tables by state were compiled for
each year and for the five  year period from the listing prepared by the
Investment Bankers Association from the Daily Bond  Buyer.   They can
be found in Appendix A.  The survey indicated that 41 states had at least
one referendum during this time period.  Ten states had at least twenty
referenda and accounted for  79% of all referenda reported in the Daily
Bond Buyer.   These ten states are identified below.
                               18

-------
State
Texas
California
Illinois
Oklahoma
Ohio
Michigan
Oregon
South Dakota
Louisiana
Virginia
Total
Number of
Referenda
469
96
86
83
52
43
43
34
27
23
956
Percent of
Referenda in
All States
39.0
7.9
7. 1
6.8
4.3
3. 5
3.5
3.0
2. 2
1. 9
79.2
19

-------
Application of the Findings to Bond Issues Not Subjected
to Referenda

Despite the low proportion of bond referenda that fail, an analysis of
the process can provide useful information about the "readiness1  or
"willingness"8 of local communities to undertake disruptive and expen-
sive public  improvements for water pollution control.  When there is a
bond election,  the decision-making process of the local government  is
exposed to the  interested observer.  Interest groups are formed, issues
are raised and openly debated, and the electorate makes a decision.
The proportion of the total votes cast in favor of the issue can thus be
considered  a quantitative indicator of the willingness of the local  com-
munity to invest  in water pollution control.

In contrast, when there is no referendum,  it is  somewhat more difficult
to identify the issues.  Most of the debate will take place among govern-
ment officials within the confines of their offices and council chambers.
There will be no campaign, as such,  since there is no need to pursuade
the electorate.   Those with strong feelings about the project will voice
their concerns to their elected representatives  and lobby for their votes.
In the last analysis,  however, the fate ofthe project may be determined
as much by the local political situation as by the merits of the proposed
project.

Analysis  of the referendum thus permits the interested observer to see
more clearly what issues  are of concern to the  electorate and what fac-
tors may contribute to the defeat of a water pollution control project at
the polls.  Although the two processes differ, the substantive  issues
remain the  same, and therefore the conclusions drawn from the referen-
dum situation,  cautiously applied, have a broader  application  to non-
referendum situations as well.
                                20

-------
                             CHAPTER II
             THE MUNICIPAL BOND ELECTION PROCESS

Introduction

As  indicated in  Chapter I, a major task of this study was to identify the
factors which account for voter response to bond referenda for pollution
control projects.  Previous studies of this type have looked primarily at
the characteristics of the electorate.  One such study of the I960 Calif-
ornia water bond  issue attempted to  clarify the sociological and political
correlates of a  willingness to support the large public expenditure re-
                                                                   9
quired to maintain an adequate water supply for Los Angeles County.
The availability of election statistics by census tract made it possible
to study systematically the relationships between voting behavior and a
number of demographic characteristics, including  age,  education,  race,
home ownership,  occupation and term of residence.  This analysis was
interpreted in the light of the issues and alignments of the campaign, the
roles played by local governing bodies, water  agencies, the media,  poli-
tical party organizations, special  interest groups and other opinion lead-
ers.

In order to develop policy-oriented recommendations for use by the EPA
or local municipalities, the study  team chose to examine the factors in-
fluencing the outcome of bond referenda more  comprehensively.  The
community as a whole, rather than the electorate,  census tracts  or pre-
cincts became the unit of analysis.  Thus, it was possible to draw con-
clusions from the study about the influence of technical aspects of the
projects,  their  financial  characteristics, and the management of the
campaigns.  All of these are policy-related factors which can be  in-
fluenced by the  EPA,  the state water quality boards and the local juris-
diction.
                                21

-------
A review of the process by which a water pollution control project is

formulated and brought to a vote identified many actions taken by

government officials and others which might affect the outcome of the

vote.  This process, briefly summarized below, thus became the major

guide for identifying the data requirements of the study.
                               Problem
                             Recognition
                             Preliminary
                               Planning
                         Project Development
                         and Financial Strategy
                                   I
                                Election
                                Process
                             Follow-Up and
                             Implementation
                     The Bond Referendum Process



The remainder of this chapter contains a description of each stage of this

process and a list of the study variables associated with it.  A detailed
flow  chart of the  process can be found in Appendix B.
                               22

-------
The Bond Referendum Process

Recognition of the Problem. A proposal to improve or provide a pollution
control system undergoes a series of important considerations and in-
fluences before  it may actually be presented to the voters in the form of
a bond referendum.   Most often, the process begins with a recognition
on the part of the local government,  concerned individuals or  citizens'
groups,  or a state water quality board that a pollution or drainage problem
exists and is creating a public hazard.

There are three major sources of such a pollution problem:  it may arise
from a recent period of rapid growth; from a prolonged period of steady
growth which has  gradually put pressure on existing public or private
sewage disposal systems;  or,  in some cases from system deterioration.
A drainage problem can also result from uncontrolled large scale develop-
ment.  However,  it can also arise from a natural catastrophe, such as
heavy rains and subsequent flooding.  Drainage improvements benefit
primarily those people whose property has been flooded, although these
problems may also have breadier implications for the environment.

Not all pollution control referenda result from identification of a deficiency
in the current system. In some cases,  groups within the community may
feel that the  town's growth  potential  is being stifled because it lacks sewers
in some or all areas.  These groups often consist largely of real estate
owners or developers who stand to gain most from any large scale develop-
ment project.  The owners of large amounts  of land with poor run-off
characteristics  may also bring  pressure on the local government to initiate
drainage improvements.

The variables associated with the problem recognition stage of the referen-
dum process are listed below:
                                 23

-------
      Geographic region
      Investment Bankers Association use of proceeds code
      Rate of population growth (1960-1970)
      Use of the waterfront
      Project aimed specifically at counteracting water
      pollution
      Special topographical conditions contributing to the
      need for the project
      Special topographical problems associated with the
      the construction of the project
      Land development contributing to the need for the project
      Future development contributing to  the anticipated need
      for the project

Preliminary Planning.  When faced with a pollution problem, the first
decision made by the local government is  whether to take direct action
or to transfer responsibility to another authority.  Where a special
district has already been established to deal with a  similar problem,
the supply of water for example, it is often in the interests of local
authorities to expand their function to include  sewage disposal.  Depend-
ing on the financial and legal constraints on the municipality, this approach
may avoid unnecessary strain  on their bonding capacity and open up a lar-
ger number of financial options.

If the municipality decides to take  action itself, officials will begin to think
about the kind of  project that would solve  the problem.  Often a project has
already been suggested, either by a private citizen or voluntary organiza-
tion,  such as  a real estate developer or the Chamber of Commerce,  or by
a public  agency such as a  town planning office.  If indeed a. project has
been outlined, it  may be adopted outright  or become the basis for  evalua-
tion of alternative solutions.

The study variables associated with the preliminary planning stage of the
process  are the following:
                               24

-------
      Governmental unit having jurisdiction
      Existence of independent overlapping governmental units
      Specific uses of the  proceeds
      Description of the project

Project Development and Financial Strategy.  It is seldom the case that
a local government is able to evaluate its pollution problem  or determine
what counter measures  should be  undertaken  on  the basis  of the
information immediately available.  Frequently the problem will be  re-
ferred to the town's engineer or a local consulting firm for further study.
A report will be prepared, analyzing the problem and recommending the
alternative solutions.  The study may also suggest that a regional approach
to the project would be more efficient, in which case the town may coor-
dinate with neighboring communities.

A municipality may then conclude  that it is in no position financially to
undertake the project, that there are higher priority problems making
demands  on their resources, and decide to drop the issue.  This con-
clusion is unlikely, however,  if the  problem is a serious one, of if there
has  been  a significant amount  of pressure put on the government by con-
cerned citizens, other interest groups, or the state water quality board.
On the other hand, if the benefits are clearly evident to the decision-
makers, and the major  obstacles are not technical, the decision-makers
will proceed to  consider the costs  of the project and alternative  financing
schemes.

One  of the most important issues considered when contemplating a major
investment in pollution control is the means by which the project will be
financed.   There.are a number of  options available, including general tax
revenue,  municipal bonds, creation  of a special district and Federal and
State grants.  Projects  of this type are seldom financed out  of the munici-
pality's operating budget alone.  This is due partly to their high cost, but
also to the fact that the  improvement is generally expected to benefit future
                                 25

-------
 as well as present taxpayers.   Typically the project will be financed
 through the sale of municipal bonds, with or without State and Federal
 grant assistance.

 There are three types of municipal bonds which may be used for this
 purpose: the general obligation bond,  the  general mortgage bond and
 the revenue bond.  Until  recently, the most common form of financing
 pollution control facilities was the issuance of general  obligation bonds,
 secured by ad valorem taxes.  Since they are backed by the full faith
 and credit of the issuing authority,  they are considered debt and there-
 fore subject to constitutional,  statuatory and local charter debt limita-
 tions and to referendum requirements.  There are a few exceptions to
 this.    In ten states, debt limitation and referendum requirements have
 been removed from bond  issues for sewage systems,    and in three
 states,  specifically New Mexico,  South Dakota and Washington, general
 obligation bonds repaid from special assessment do not constitute debt.
 Appendix A summarizes State  constitutional and statuatory provisions
 regarding debt for pollution control projects.

 The general mortgage bond pledges the taxing power of the jurisdic-
 tion and the physical facilities being built as collateral.  It does not
 differ from the  general obligation bond in terms of the  referendum law.

 A third kind of bond,  which has been employed successfully to avoid debt
 limitations,and  sometimes referendum requirements,  is the revenue bond.
 The revenue bond pledges the revenue from specified sources (frequently
 user charges for the  facility itself) to repayment.  The issuing authority
 is not obligated to pay the bonds,  but only to take no action which would
 cut off the  source of revenue for the bonds, or restrict the ability of the
bond trustee to collect and disburse the  revenues.  Revenue bonds  do  not.
therefore,  constitute debt for the issuing authority,  except in the state of
Idaho.  No states prohibit revenue bonds,  and five states permit only
                                26

-------
revenue bonds for this type of financing.  In about three quarters of the
states, local governments have actually issued revenue bonds for financ-
ing water pollution control facilities.

Another approach to financing major public investments is the creation
of a special district to provide limited services,  such as sewage collection,
treatment and disposal.   Formation of a special district creates another
political authority which is empowered to raise capital through the sale of
revenue bonds unemcumbered by the debt limitations of the one or more
municipalities involved.  In addition,  the boundaries of the  special dis-
trict may be set to be contiguous with the pollution problem.  This may
                                                                     •
be the case where only part of a town needs service,  or where a regional
approach appears to be the most reasonable one to take.  Regional auth-
orities work most efficiently  in very high and very low density areas.  The
special district has its own governing body which usually operates in a
state of relative autonomy. More than half of the  states permit the forma-
tion of special districts for the  purpose of sewage collection treatment and
       T 12
disposal.

The Federal government has  several programs to assist local municipali-
ties in the construction of sewage treatment facilities.   The Environmen-
tal Protection Agency  authorizes grants for up to  30% of the eligible por-
tion of the project, and up to  40% if the plan conforms with  a comprehensive
metropolitan development plan.  In addition,  the Department of Housing and
Urban Development has several financial assistance programs for this
purpose.  One provides interest free advances to municipalities  for plan-
ning purposes,  including  the payment of architectural and engineering fees.
These advances must be repaid when the construction begins.  In addition,
HUD provides money under their 701 program for comprehensive urban
planning which often includes  sewer facilities. HUD will also make loans
to small towns which a re unable to obtain loans on  reasonable terms from
other sources.  The Farmers Home Administration of the Department of
Agriculture will make grants  for up to 50% of the costs of a sewage col-
lection and treatment facility to towns  of not  more  than 5, 500 in population
                                27

-------
when the projects have been approved by the state water quality agency.
 The Economic Development Administration of the Department of Com-
merce has a similar program for areas designated as areas of substan-
tial and persistent unemployment.   In addition to the Federal funding
available,  there are also a variety of State and regional financial assis-
tance programs.

After estimating the total cost of the project and assessing the availabil-
ity of Federal financing, the municipality will decide  on the amount of
bonds to be issued and the means of repayment.   It may also give some
thought at this  point to the bond market, and the prospect for selling its
bonds when it is ready to do so.  With this  in mind, it will decide upon
the amount of debt to be incurred and the type of bonds to be sold.  De-
pending on the  type of bond contemplated, the constitutional and statua-
tory provisions of the state, and the local charter, a  referendum may
or may not be required.  Thirteen  states always require a referendum;
21 never require a referendum; 12 may require a  referendum, depend-
ing on the circumstances.   When the referendum can be avoided,  most
local governments  will do  so in the interests of not jeopardizing the
project.  However,  it is not unusual for a municipality to hold a referen-
dum,  even though  none  may be required, either because they are not
aware of alternative means of financing which do not require a referen-
dum,  or they wish  to obtain an expression of priorities from the  elec-
torate.

The decision on the type of bonds to be  issued marks the end of the pro-
ject development and financial strategy stage of the municipal bonding
process.  The  study variables associated with this stage are the
following:
                                28

-------
      Dollar amount of issue
      Consistency of population served and population
      paying
      Total dollar cost of the project
      Repayment of bonds by user charged
      Repayment of bonds by increased taxes
      Repayment of bonds by increased taxes until
      facility can be supported by user charge
      Repayment of bonds by special assessment use of
      State grants
      Use of Federal grants
      Current tax rate (dollars per $1, 000  of assessed
      valuation)
      Percentage increase (decrease) in tax rate increase
      from 1966 to 1970
      Proportion of project costs  repaid by increased taxes
      Amount of project  costs repaid by increased taxes
      Proportion of project costs  paid with State funds
      Proportion of project costs  paid with Federal funds
      Total dollar indebtedness at beginning of 1970
      Percent increase (decrease) in indebtedness between
      1966 and 1970
      Portion of operating budget spent for debt service
      Existence of a debt limit
      Source of the debt  limit
      Meaning of bond passage in terms of debt limit
      Current jurisdiction credit rating
      Rating  system (Moody's or Standard and Poor's)
      Federal regulations associated with the project
      State or local regulations associated with the project

The Referendum Process.  If the municipality decides to undertake the
project itself,  and a referendum is required or desired, the next step
is to prepare the wording of the proposition.  This may be  an important
                               29

-------
 step, because the wording of the proposition can have an impact on the
 outcome of the election.   If the wording  is not clear, and the project is
 not described in sufficient detail on the ballot, then the voters  may turn
 it down at the polls.   On the other hand,  if too much detail is given about
 a potentially  controversial project, the voters may find fault with it.

 A decision must also be made on whether the proposal is to be presented
 alone on in combination with other projects.  This requires a preliminary
 assessment of the likely outcome of the  issue.  A popular project may
 help to carry less popular ones  if they are offered in combination.  On
 the other hand, unpopular projects  may help defeat a pollution  control
 project which might succeed if presented alone.

 The municipality must also identify the electorate that will be voting on
 the bond issue.  In some cases this will  be  all of the registered voters
 in the town.   When only part of the  population will be served by the im-
 provement, officials may wish to declare only a portion of the  electorate
 eligible to vote,  as incongruities in the population voting,  paying and
 being served  can contribute to a referendum's defeat.

 The timing of the election appears to have little effect on its outcome.
 However,  if it is scheduled to coincide with an election  of public officials,
the turnout will ordinarily be much greater  than if a special election is
held.   This can have a negative effect on the outcome of the election if
the project is at  all controversial.  A large number of expensive proposals
on the ballot,  each requiring a separate  vote, may force the voters to
 choose what they believe to be the highest priority projects, lest the com-
munity "spend too much money".  This,  too,  could spell defeat for a
 sewer bond.   These options must be weighed against the expense of hold-
ing a special election,  which might in itself jeopardize the outcome.  In
one case studied, voters were angered by the local government when it
chose to hold  a costly special  election when a school bond election was
planned for nearly the  same date.
                                30

-------
The variables associated with the planning of the referendum are:

      Date of the election
      Population served same as different than population
      voting
      Type of election of local government (partisan/non-
      partisan; ward/at-large)
      Wording of the referendum
      Way the issue was offered to the voters  (alone, com-
      bined,  split)
      Other issues  presented in the same proposition (if
      combined)
      Outcome for  different issues on the same ballot (if split)
      Existence of  candidates for public office elected in the
      same election
      Other issues  presented for a vote in the same election
      but not combined
      Median  family income (dollars)
      Proportion of families whose income is between $3, 000
      and $10, 000 per year
      Proportion of families whose income is higher than
      $10, 000 per year
      Proportion of population of age between  21 and 35
      Proportion of population of age over 65
      Proportion of population non-white
      Proportion of population resident in the area in 1955
      Proportion of single family housing units
      Proportion of owner-occupied housing units

The variables associated with the referendum campaign are listed below:

      Existence of  pressure groups for the issue
      Existence of  pressure groups against the issue
      Importance of pressure groups for the issue (infor-
      mants' judgement)
      Importance of pressure groups against the issue (informant's
      judgement)
                               31

-------
      Were the groups partisan in nature?
      Experssed position relative to the project of community
      leaders,  including:

             Local officials
             Doctors
             Politicians and party organizations
             State officials
             Chamber of Commerce
             Other local governing bodies
             Ad hoc groups
             Clergy
             League of Women Voters
             School and University Groups
             PTA
             Others

The Election Campaign.  In the context of a bond  referendum, the campaign
refers to the process whereby actors interested  in the outcome of the
election attempt to influence the  votes of citizens eligible to vote.  In
most  cases,  these actors will be important groups which have been or-
ganized for  some other purpose but which have an interest in the outcome
of the referendum.  The organization which is most often active in bond
election campaigns is the local government itself.  The chief executive
or legislative body,  for example, may have proposed the sewer or drain-
age project and may therefore  be its most important proponent.  Other
types  of governmental units which may become involved are State or
regional agencies  concerned with the problems of drainage of water pol-
lution.

There are also several private groups which  may have a clear interest
in the outcome of the bond  election and be active  in the campaign.  The
local  Chamber of Commerce, for example, is usually concerned with
                              32

-------
any issue which will affect the economic growth of the community and
the tax rate—both of which are important to its business constituents.
Another  group which is often active is the League of Women Voters
which, during the past few years,  has made water pollution control one
of its major concerns.  Moreover, the League's  general function is to
inform the voters so that they can make intelligent decisions on election
day.  In  this regard, it will attempt to provide information on the con-
tent of any bond issue  being presented.  In addition, there are other
private groups which may become involved, especially conservationists,
the clergy,  members  of the health profession, professional engineers,
taxpayers' groups, and the PTA.

Finally,  there maybe ad hoc groups,  established either to support or
oppose the proposed bond  issue.  Many opposition groups are composed
of angry taxpayers or citizens opposed to land development which "threa-
tens the  character of the community".  Proponents of bond issues will
often establish ad hoc organizations which may be "front groups" for
other interested organizations.  For  example, the local government,
which is legally not allowed to campaign for a bond issue it is presenting
to the voters, may nevertheless provide assistance to a "citizens for
better sewers committee".

A campaign will generally not begin until there has been a government
decision to put the issue on the ballot, since this decision will define much
of the subject matter for the campaign.  However, political activity may
begin much earlier, while the local government is still  considering whether
to put the issue  on the ballot.  At this stage the objective is to influence
the way in which the issue will appear on the ballot, or whether it will
appear at all.  A campaign may begin even earlier than this, when some-
one in the community decides that he  would  like the local government to
undertake a sewer or drainage project.  He may immediately take his
ideas to  the public in an effort to create pressures on the city government
to consider the project.
                                33

-------
A campaign may take many forms, but always includes the use of some
promotional materials or activities intended to pursuade the voters.
These may include face-to-face communication, newspaper ads and edi-
torials,  pamphlets, films, television,  and so forth.  A list of variables
associated with the role and  types of media investigated includes:

      Role played by newspapers during the campaign
      Role played by other media during the campaign
      Use of newspaper ads
      Use of radio and TV ads
      Use of telephone surveys
      Transportation offered to voters
      Advertisement mode: literature,  brochures,
      leaflets
      Advertisement mode:  Film
      Advertisement mode:  Posters
      Advertisement mode:  News stories
      Advertisement mode:  Letters to the editor
      Advertisement mode:  Speeches and meetings
      Advertisement mode:  Door-to-door  canvassing
      Availability of map to  voters

The content of the actual information transmitted can also vary immense-
ly.  If,  for example,  there is no opposition to a bond issue,  its proponents
may simply announce that there will be a bond referendum and describe
the project.  On the other hand, if there is significant opposition, there
probably will be a vigorous debate over the merits  of the proposed project,
the means of financing it, and so forth.  Many different criticisms may
be raised.  Those which were included as variables in this  study are:

      Cost and size of the project
      Taxes, fees and assessments
      Source of revenue to repay the bonds
                               34

-------
      Discrepancy in the allocation of benefits and costs
      Timing of the construction
      Location of the improvements
      Need for the project
      Opposition to community growth or change
      Opposition to State or Federal  interference
      Management should be regional rather than local (or vice versa)
      Distrust of local officials
      Not enough information given by administration

The  single most important aspect of  the campaign is the level of cam-
paign activity, the chief determinant of which  seems to be the presence
or absence of opposition.   Where there  is no opposition-,  the proponents
of the bond issue have the option (which they often take) of maintaining
a low level of campaign activity.  If the proponents do any campaigning,
it is normally focused on encouraging the voters who are  likely to vote
in favor of the bond issue to go to the polls.  In cases where there is
opposition, the level  of campaign activity is usually higher,  because
there are more controversial issues to be debated.  In some instances
proponents of a bond  issue generate a high  level of campaign activity
before there is any opposition and therefore any need to campaign at
all.   This often arouses an opposition group and leads them into dam-
aging controversy.  There are   at least  two reasons for this kind of
behavior.  Some proponents interviewed during the in-depth case studies
believed that a high level of campaign activity was necessary for elec-
tion  success,  while others were  simply committed ideologically to the
education of the electorate.

One  other factor which may affect the level of campaign activity is the
position of the press  or other news media covering the campaign.  It is
possible for the media to create  a campaign (or the impression of a cam-
paign) even when the  two  sides of the issue do not want one;  and con-
versely, an uncooperative media can stifle many attempts at campaigning
by groups who favor  or oppose the issue.
                                35

-------
After the campaign, the issue is in the hands of the voters who will
either approve or defeat it.  The variables investigated in this study
which are associated with the actual balloting and its results are:

      Percentage  required to pass the issue
      Outcome (pass or fail)
      Outcome (percent of votes in favor)
      Outcome for the different issues on the ballot (if split)
      Percentage  of the total number of issues  on the ballot
      which received a significantly higher percentage of the
      favorable votes  (5%  or more) than the issue under study.
      Percent of the total  number of issues on the ballot which
      received a lower percentage of the favorable votes  (5% or
      more) than the issue studied
      Issue which received the highest vote
      Issue which received the lowest vote
      Percentage  of the population registered
      Largest turnout for  any purpose in that election

Follow-Up and Implementation.  If the referendum passes,the issuing
authority will probably seek expert advice on the sale of the bonds.   If
the market is favorable,  it may sell them right away; if not, it may decide
to delay.  In a few cases,  as when there  is a change in the local  govern-
ment administration and policy, the bonds may never be sold at all.

If the voters  defeat the bond  issue, the proponents have several different
alternatives.  They  may give up the fight for the  project entirely, accep-
ting the will of the voters, or they may attempt to pass the project any-
way in the same or a different form.   If they decide  not to change their
strategy, they will simply reschedule the election.  On the other hand,
they may use what they have learned from defeat to make  some changes
in the characteristics of the  project,  the bond issue,  or the campaign.
If the defeat has been a serious one, and they are committed to the pro-
ject,  they may choose to establish a special district and avoid another
referendum.   They may even change the  character of the project by
                                36

-------
adding or subtracting from the population that will be served.  They
may also try to develop regional interest in the project so as to have
more flexibility in financing as well as a different electorate.
                                 37

-------
                             CHAPTER III
        DATA COLLECTION AND ANALYSIS METHODOLOGY

Data Collection Methodology

Developing the Questionnaire.  As indicated in Chapter II,  there are a
large number of factors at each stage of the municipal bonding process
which can affect the passage or failure  of a referendum.   One hundred
fifteen of these factors, listed in Chapter II, were isolated as the major
variables to  study.

Having  identified a comprehensive list of variables which might be dir-
ectly or indirectly related to the outcome of the vote,  it was then neces-
sary to identify the most knowledgeable local information  sources and
to formulate questions to collect data about those variables.  It was
obvious that  at least the following  people would have to be questioned:
town clerk, controller or finance director, town planner,  town engineer
and a representative of the media.  In cases in which  there had been
little campaigning and the information was not readily available from
these sources, other members of  the community were also interviewed.

Selecting the Sample.   The sample was drawn from a  listing of all of the
municipal bond referenda published in the  Daily Bond  Buyer and compiled
by the Investment Bankers' Association of America.   This computer list
includes about 75% of  all the referenda held between January 1966 and
June 1970 by cities, counties, townships,  districts and statutory authori-
ties to approve projects for the construction of sanitary sewers,  storm
and drainage projects, and treatment plants.

The IBA list indicated that during  this period 1047 such referenda passed
and 207 failed.  Since a major purpose of the study was  to  compare the
attributes  of the referenda  that failed with those that passed, an equal
number was  sampled from  each of the two categories.  Therefore,  the
                                39

-------
 initial population was stratified according to the pass-fail criterion.
 A systematic random sample was picked within each stratum, avoiding
 any bias inherent in the order of the list. 13  The sample was not further
 stratified on the basis of social economic characteristics,  because the
 purpose of the study was to identify what factors affect the success or
 failure of bond elections and not to focus particularly on the effects of
 socio-economic variables.

 The time and budget available for the study made it possible to study
 40 referenda, that is 20 that passed and 20 that failed.  This represents
 a  sampling rate of approximately 1/50 for those that passed and 1/10
 for those that failed.  This lists of these two sets of 20  referenda follows.

 Statistical Analysis of the Quantitative Data

 This section describes briefly the series of statistical analyses performed
 on the data collected and the kinds of conclusions that can be drawn from
 them.  Some  explanation of the potentials and limitations of the techniques
 is also included.  Those readers who are not interested in the statistical
 methodology may prefer to skip to Chapter IV.  Those who are familiar
 with statistical  techniques will find further comments about them  in
 Appendix C.

 Frequency Distribution.  When a sample has been selected, and a series
 of measurements have been made for some  variable,  then the data are
 ready to be organized and analysed.   There are  many kinds of analyses
 that could be performed and even more generalizations which might be
 made as a result.   However,  not all  of this  information  is very useful.
 Therefore, it is important at the outset of the analysis to organize the
 data so that some  general patterns emerge.  Since the variables being
 investigated are class if icatory, (that  is, they can be measured in terms
 of a limited number of discrete categories or values), the number of
 cases which fell into each class was  examined first.  This  is a frequency
distribution,  which is just an ordered rearrangement of the original data.
                                40

-------
                  Sample of Referenda Which Failed

(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(ID
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
State
Georgia
Minnesota
Texas
California
Alaska
Illinois
Ohio
Oklahoma
Texas
Texas
Louisiana
Montana
Oklahoma
Texas
Colorado
Louisiana
Ohio
Oklahoma
Texas
California
City
Dekalb County
Mankato
Portland
Simi Valley Co. Sant Dist.
Douglas
Ros.elle
Oakwood
Oklahoma City
Round Rock
Brookshire Mun W. D.
Donalds on ville
Great Falls
Cache Water and Sewer
Groves
Hotchkiss Sewer
Morgan City Drain
Avon Sewer
Moore Sanitation
Cuero Water and Sewer
Los Angeles Co. Sewer
Year
1970
1969
1969
1969
1968
1968
1968
1968
1968
1968
1967
1967
1967
1967
1967
1966
1966
1966
1966
1966
Use of
Proceeds*
28
28
28
21
21
28
21
28
21
22
21
21
22
21
21
28
21
21
22
21
* 21   = Water and Sewer
  22   = Sewer
  28   = Storm Sewer or Drainage
                               41

-------
                   Sample of Referenda Which Passed

(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
State
Texas
Texas
Texas
Massachusetts
South Dakota
Texas
California
Michigan
Oregon
Texas
Wyoming
California
Missouri
Texas
Texas
Louisiana
California
North Carolina
Texas
Texas
City
Columbus
Harris Co. Water Dist. #73
Rockport
Athol
Batesland
Harris Co. Water Dist. #10
Palo Alto
Warren
Bend
Kyle
Jackson
Mill Valley
Rolla
Alvin
Royse City
Jefferson Parris Consol.
Drainage District
Petaluma
Brevard Sewer District
Belton Drainage District
Van Alstyne
Year
1970
1970
1969
1969
1969
1969
1968
1968
1968
1968
1968
1967
1967
1967
1967
1967
1966
1966
1966
1966
Use of
Proceeds*
21
22
28
21
22
22
21
21
21
21
21
21
28
22
21
28
21
21
28
22
* 21   = Water and Sewer
  22   = Sewer
  28   = Storm Sewer or Drainage
                               42

-------
For example, the frequency distribution of the amount of the issue indi-
cates that 72. 5% of all issues are quite small, that is less than one mil-
lion dollars.  A comparison of  this information with  the distribution of
another variable, the total cost of the project, shows that only 61% of
the projects cost less than one  million dollars.   This discrepancy sug-
gests the use of other forms of financing--for example, State and  Fed-
eral grants.  However,  both figures corroborate the fact that there are
a great many referenda for rather small projects.

Another application of this descriptive statistic is the frequency distri-
bution of referenda by geographical region.  Sixty-seven percent of
the referenda were held in the  West and Southwest, while  only 18%
were held in the  Midwest,  13%  in the South,  and 2% in the Northeast.
This picture is consistent with  the previous analysis of the statutory and
constitutional provisions concerning the requirements for bond referenda.

Cross-Tabulations.  The relationship between each variable  and the out-
come of the referenda was examined next.  One of the most commonly
used techniques for studying the relationship between variables is  the
cross-tabulation.  The cross-tabulation,  or  contingency table is a tabular
presentation of percentages and frequency distributions of cases accord-
ing to two class if icatory variables.  The principal use of this analysis
was  in displaying the  relationship between each variable and  the outcome
of the referenda.

The  relationship between each  pair of variables was  analysed statistically
by various tests  of significance to determine the probability that the ob-
served distribution could have  occurred by chance.   The strength  of the
relationship was  also tested by means  of another series of statistical
measures.  A detailed discussion of both  the tests of significance  and
the measures of  association is  presented  in Appendix C.
                                 43

-------
 The potential of cross-tabulation tables can be illustrated with an exam-
 ple from this research.  Consider, for example, the contingency table
 below which summarizes the relationship between the outcome of the
 referenda and whether or not the bond  is to be repaid by increased
 taxes.
Outcome
Passed
Failed
Total
Repaid by
Increased Taxes
6
13
19
Not Repaid by
Increased Taxes
14
7
21
Total
20
20
40
 The table reveals that there are clear differences in the outcomes of the
 referenda depending on whether or not the bonds are to be repaid by in-
 creased taxes.  This suggests a strong relationship between the  two
 variables.   Projects not repaid by increased taxes  are more than twice
 as likely to pass as those which are.  The table also shows that this
 phenomenon is not always considered when the strategy for repaying
 the bond is being formulated.  Of  the 40 referenda, close to half pro-
 posed repaying the bonds by increased taxes.

 It is often useful to compare two contingency tables to get a better
 picture  of the  relationship between two variables  that are both related
 to outcome.  The example below shows that projects which are not re-
 paid by  user charges are three times as likely to fail as those which are.
A comparison of the two tables discloses that these two means of re-
 payment have  opposite but nearly  equivalent impacts on the outcome.

 The two tables display clearly the preferences of the  voters.   The re-
 sults of the  other contingency tables were generally less dramatic,  and
the question arose as to whether the observed differences are  statistically
 significant.  This question has been answered by  means of the tests of
                                44

-------
Outcome
Passed
Failed
Total
Repaid by
User's Charges
11
5
16
Not Repaid by
User's Charges
9
15
24
Total
20
20
40
significance cited above and discussed in Appendix C.   These tests deter-
mine whether discrepancies in percentages are due solely to sampling
errors or reflect statistically significant relationships.

Calculation and interpretation of the tests for all the cross-tabulations
between the  114 variables identified and the results of the bond elections
identified 34 independent variables associated with success or failure
in something other than a random pattern.  This suggested that  some,  or
even all, of these variables might explain the variance in outcome of the
bond referenda.   The identification of these variables led to the next
phase of analysis.

Linear Correlation Analysis.  Linear correlation analysis has been  used
in this research to provide a single summary statistic describing the
strength of association between all pairs of variables, whereas  cross-
tabulation was used only to relate the outcome to all the other variables.
Like two-way cross-tabulation tables, correlation coefficients enable
the specification of the relationship between two variables.  In cross-
tabulation the degree of association is obtained by examining the fre-
quencies of the two variables in tabular  form.   In correlation analysis
the strength of association is indicated by a single summary statistic:
the coefficient.   However,  correlation analysis can be applied only to
variables which are dichotomous, or at  least ordinal in scale, as ex-
plained in Appendix C.  Therefore, the correlation coefficients  were
determined only for the 89 initial quantitative variables and the  10 new
variables described in Appendix C.  These 99 variables summarize
most of the relevant information collected, since many variables are
somewhat redundant with the nominal variables.
                                45

-------
 Besides  its use in identifying relationships between variables and des-
 cribing their direction and strength, linear  correlation analysis was
 also used to identify indirect relationships and develop links between
 them.  For example, there is a significant relationship between popu-
 lation growth and the number of groups active in the campaign (+.47)
 and between this later variable and the outcome of  the election (-. 28).
 This type of analysis points  out that the initial characteristics are not
 independent variables, and that even if two variables are not directly
 related,  they can very well be connected through a series of interven-
 ing variables.  Correlation analysis has been applied in this way, where
 relevant in the interpretation of the data presented in the following
 chapter.

 Several sets of correlation coefficients were calculated:  the Pearson's
 product-moment correlations for the data that are  interval in scale,
 Kendall's rank-order correlation coefficients for the ordinal variables,
 and point-biserial and tetrachoric coefficients for dichotomous data.
 While the computational procedures for these types of correlation differ,
 they all have a common and  intuitive representation through the scatter
 diagram.  In this diagram,  assume that the  vertical axis represents the
 outcome  of the referenda in terms of percent of favorable votes,  while
 the horizontal axes represent respectively,  in diagram

       (1)    the percent of present population already resident
             in the community in 1955 (variable 104),
       (2)    the index of coincidence of those who  voted, pay
             and are  served (variable 117),  and
       (3)    the number of criticisms types  raised (variable 119).

 A referendum  is represented by a dot on these diagrams at the inter-
 section of the values recorded for it on the two variables,  and the inter-
 pretation of these diagrams  is straightforward.   In diagram 1, there
is little or no relationship between the two variables.  The dots  are ran-
domly distributed and  the correlation coefficient for these two variables
approaches zero (+0.003).  Scatter diagram  2 reveals a weak positive
                                46

-------
>   *
  4   *
        *    *
        »  *
 *
f »
      1
relationship in which increase  in either one  of the variables is mildly
associated with an increase of the other,.  In  that case, the correlation
coefficient is positive and small in quantity (0. 189).   Finally, the re-
lationship revealed by diagram 3 is strong and negative: the coefficient
is negative  and large (-0.52).

Linear Regression Analysis.   The linear correlation analysis and the
cross-tabulations  identified a large number  of characteristics of the
referenda which were highly correlated with outcome, but not highly
correlated with one another; that is, independent of  one another.  Fac-
tor analysis was also used to identify the combinations  of these inde-
pendent forces  underlying the outcome  of the referenda.  In fact, as
explained at the end of Appendix  C, factor analysis was carried out as
an intermediate step in an attempt to derive clusters of referenda dis-
criminated by the  pass-fail criterion.  However, the additional infor-
mation resulting from these two techniques was  not used to derive con-
clusions  for the reasons explained in Appendix C. On the basis of this
information, the analysis proceeded to the development of a linear re-
gression model.

The linear  regression analysis had two major purposes.  The most im-
portant of these was to  understand better the relationship between each
independent variable retained for this analysis and the outcome of the
referenda.  This understanding could be gained by examining the sign
and size of the coefficients of the equation.   These coefficients reflect
                                  47

-------
the strength as well as the direction of the relationship between a
dependent variable and the independent variable.

The second purpose of the linear regression was to develop a model
capable of predicting the outcome of a referendum, in terms of the
percentage of votes in favor.  This  was achieved by means of a linear
combination of the  values of those known characteristics which together
correlate as highly as possible with the outcome.

A more powerful variation of multiple regression analysis, namely
stepwise multiple regression was used.  It provides a means  of choosing
those independent variables which will predict best with the fewest inde-
pendent variables.  This method constructs a prediction equation using
one independent variable at a time.   The first step is to choose the single
variable which is the best predictor of outcome.  The next variable added
is the one which provides the  best prediction in conjunction with the first
variable.  This process continues until no additional variables can make
a statistically significant contribution to the predicting  of the  variance in
the outcome.

Single and multiple regression used in conjunction with assumptions about
the time sequencing of the independent variables allowed us to analyze
some chains of inter-relationships which can be considered "causal".
However, such causal chains  are really only cascades of inter-relationships
between variables.  The cause and effect relationships  are not provided
by the analysis, but derived either from obvious time precedence or from
assumptions introduced by the analysts about which variables among those
retained have a prior effect on the dependent variable.
                                48

-------
                            CHAPTER IV
              FACTORS AFFECTING THE OUTCOME OF
                          THE REFERENDA
Introduction

As explained in the previous chapter, the regression analysis has  iden-
tified the minimal  set of referendum characteristics that explains  a
satisfactory proportion  of the variance in the outcome of the election.
Sixty-two percent of the variance is accounted for by the following
study variables:

      035      Existence of pressure  groups against the issue
      078      Bonds repaid by increased taxes
      109      Project aimed at counteracting a water pollu-
               tion problem
      119      Number of criticisms raised
      098      Proportion of families with incomes  of less than
               $3,  000 per year
      037      Groups partisan or not
      077      Bonds repaid by user charges
      015      Rate of population growth, 1960-1970
      097      Median family income
      002      Dollar amount of bond  issue
      083      Current tax rate

A list of variables, however, even  in  a regression equation, tells  little
about the dynamics of a bond election  and the role of the independent
variables in determining its outcome.  Therefore,  the direct relation
between each of these variables and the  outcome of the election will be
discussed prior to  the interpretation of the regression analysis.  This
discussion will incorporate examples  from the in-depth field studies and
                                49

-------
telephone surveys.  However,  these examples are only illustrative and
should not mislead the reader  into thinking that the conclusions have
been drawn from only a few observations.

In addition to the variables included in the regression analysis, there
are also several others which  were hypothesized at the outset of the
study to have significant influence on the outcome of the referendum.
Although they did not emerge as significant in the analysis, they will
also be discussed,  since they are generally expected to have an impor-
tant impact.

The independent  variables have been  grouped into four categories:  popu-
lation characteristics; community characteristics; characteristics of
the bond issue; and campaign characteristics.  Population characteris-
tics are basic socio-economic  data which provide a profile of the citizens
in the communities that were studied.  Some  examples are percent of
the population over 65 years of age, percent with incomes under $3, 000
per year and percent non-white.  Community characteristics, on the
other hand,  describe the community as a whole, for example, the tax
rate.  The variables which are characteristics of the bond issue are
those which  describe the issue, eg, the amount and the  manner  in
which it was presented to the voters,  (either as a split issue,  singly,
or in combination with other issues).   The campaign characteristics
describe the political process which precedes the vote  on the bond--the
number of criticisms raised, for example.
                              50

-------
Population Characteristics

Population characteristics are basic socio-economic data such as age,
education, income, race and the like which describe the residents  of
the communities that were studied.  These characteristics are generally
assumed to be important correlates of voting behavior; in addition,  they
can be obtained easily and inexpensively from census data and town plan-
ning studies.   For these reasons, several population characteristics
were included as variables in this study.

Surprisingly, the statistical analyses showed that only two of these var-
iables were directly or indirectly related to the outcome of the bond elec-
tion: percentage of families with incomes less than $3, 000 (variable 98),
and the median income of the population,  (variable  97).  Of these two,
the percentage of families with incomes less than $3, 000 is the more  sig-
nificant,  although  its correlation with outcome is only +.23.  As the
table below suggests, bond elections for pollution control projects are
more likely to be successful in communities with a high percentage of
families with incomes  of less than $3, 000.
Election Outcome
Passed
Failed
Total
Percent of families with
incomes less than $3, 000
0-19%
12
17
29
20+%
8
3
11
Total
20
20
40
This relationship must be interpreted carefully.   It may be that the
data indicate that poor people vote in favor of pollution control bonds--
perhaps because they expect to benefit from the project without having to
pay for it.  This interpretation would be consistent with other political
                                        14
analyses of voting behavior  in referenda.    It would also be consistent
with the assumptions of political leaders interviewed during the in-depth
                                 51

-------
field studies.who felt that they cound "count on the poor wards" to vote
in favor of the bonds.     In spite of the apparent plausibility of this inter-
pretation, this study's data do not provide sufficient evidence to warrant
this conclusion without further analysis.  There are other explanations
which are equally plausible.  For example, it may be that communities
with a higher  percentage of poor persons  have tradiationally been unable
(or unwilling)  to maintain a high level of public  services.  If these were
so, the  need for sewer and drainage improvements  in these communities
might be more critical,  and the urgancy of the need, rather than the per-
centage of poor persons, would contribute to the number of votes in favor
of bonds for such improvements.

All that can safely be said in  interpreting  this relationship is that there
is a positive correlation between communities with a high percentage of
families with  incomes of less than $3, 000 and affirmative votes in bond
elections.  As such, this variable is a predictor (albeit weak) of bond
election outcomes.  These data do not mean that individual families with
incomes of less than $3, 000 vote in favor  of bond  issues.  This kind of
interpretation falls in the category of "ecological  correlations. "
Making  inferences about individual behavior from aggregate data are of
questionable validity.

The other population characteristic which appears in the final regression
equation (variable 97,  median income) has a weak direct correlation with
outcome: -. 05.  If there is any direct relationship between median  in-
come and the outcome of the election, this low correlation may suggest
that the  relationship is not linear.
Election Outcome

Pass
Fail
Total
Median Income
< 4000
8
8
16
4000
2
2
4
5000 6000
2 4
0 3
2 7
7000+
4
7
15
Total

20
20
40
                               52

-------
The lack of a strong direct correlation with election outcome does not,
of course, eliminate a variable for the regression.   It may be that
there is a  relationship between median income and some other  variable,
(or combination of variables) which are,  in turn,  related to the outcome
of the election. Median income apparently plays a role in a linear com-
bination of variables which together account for a substantial amount of
the variance in the dependent variable.

None of the other  population characteristics emerged as significant de-
terminants of the  outcome of the bond  election.  One possible explana-
tion is that the unit of analysis  for this study was  the community, where-
as the relationship between socio-economic characteristics and voting
behavior has generally been observed  for the behavior  of voters as indivi-
duals, or as carefully delineated groups.  While socio-economic  character
istics may have substantial influence on individual voting behavior, it
would be erroneous to assume that they should have the same effect on col-
lective  behavior.   This kind of interpretation runs into the  problem of
"aggregation effects:" the interpretation of data at one  level of aggregation
                                                      17
when it was actually collected and analysed at another.
Another possible explanation for the fact that few socio-economic varia-
bles were  related to election outcome may result from the fact that there
was little variance in several of the socio-economic variables.  The
communities used in this survey were selected by sampling  randomly
from a list of communities whose bonds had  passed and a  list of those
whose had failed.   The resulting sample varied little in terms of socio-
economic characteristics:  many were small towns or suburbs, populated
predominantly by white homeowners.  In order to determine the impact
of homeownership or race on the outcome of the  vote, it would have been
necessary to stratify the sample according to these characteristics or
study the voting behavior of small,  homogeneous groups within the com-
munity.

-------
Community Characteristics

There are three variables which are characteristics of the community
as a whole which appeared in the regression equation: variable 109,
whether or not the proposed project intended to solve a present or
future pollution problem; variable  15, the rate of population growth
from I960 to 1970;  and variable 83, the current tax rate.  Each of
these is discussed below.

Existence of a pollution problem.  Mo.=t of the communities studied re-
proted that their proposed projects were intended to solve a present or
future pollution problem, as illustrated by the following table.  A pol-
lution problem, by definition,  affects everyone in the community be-
cause of the possibility of drinking water contamination and other health
hazards.   Recent intensified concern for environmental quality has drawn
attention  to these hazards, as  well as to a resulting decline in the attrac-
tiveness of the environment.  A town's  ability to capitalize on this  concern
greatly increased the likelihood that its  bonds would be approved.  A few
examples from the case studies should  help to illustrate this point.
Outcome
Passed
Failed
Total
Present or Future
Pollution Problem
No
1
4
5
Present or Future
Pollution Problem
Yes
19
16
35
Total
20
20
40
Although there are only five cases in which there was not reported pol-
pution problems, it is interesting to note that 80% of them failed.  In
contrast,  the association of the project with a pollution problem does not
necessarily command success; it results in failure almost as  often.  This
suggests that the presence  of a pollution problem may be a necessary but
insufficient condition for  election success.
                              54

-------
In Brevard,  North Carolina,  for example, a water pollution problem
•was threatening the tourist business on which a major part of the town's
economy depends.   The seriousness of this problem was the  major issue
in the  campaign for bonds to  finance major sewer improvements, and
the bond issue passed easily.

In contrast,  in those cases where there is no clearly identified pollution
problem, the bond issues had a  more difficult time  passing.  For exam-
ple, in DeKalb County,  Georgia, the voters rejected a drainage bond
issue in part because drainage was simply not a problem for the vast
majority of residents.  At the same time, the same voters approved
by a substantial margin, a new solid waste disposal facility to end a
solid waste problem which was clearly a  pollution problem affecting
the whole county.  This suggests that the voters were in favor of re-
ducing pollution, but were not in favor  of incurring  a debt to  construct
improvements which would assist relatively few people  in the community.

Even where  a pollution problem clearly exists,  however,  the local  resi-
dents may be unwilling to support a bond  issue.  For example,  Avon,
Ohio had been declared a major polluter by the  Ohio Water Pollution
Control Board who  had even  imposed a building ban on the community.
Many of the  residents in the community responded that the OWPCB
"had something against A.von" or were  "in cahoots with  the land develo-
pers who want to build sewers, " and defeated the issue.  Similarly, in
Malibu,  California, the pollution of the beach was a clear and serious
problem, but the residents were unwilling to approve a  bond  issue for
sewer improvements because they  felt  that the  pollution problem was
also the responsibility of non-residents who enjoyed the beach.
Population Growth.  One of the assumptions underlying the model of
the municipal bonding process is that population growth contributes to
a pollution problem,  which in turn is associated with the  approval of
                               55

-------
municipal bonds for pollution control projects.  However,  statistical
analysis suggests that the  relationship between population  growth and
the outcome of the election may be somewhat different.  The correlate
coefficient between variable 15 (population growth) and 109 (existence
of a pollution problem ) is only . 07.  Furthermore, the correlation
coefficient between population growth and the  outcome of the election
is -.16: population growth is associated with election defeat, not success.

This suggests that population growth has other effects  on  the community,
and that some of these are important determinants of the outcome of the
election.  For example, population growth may bring new groups into
the  community with values and ideas which are different from those
held by the older residents.  Many of the communities in this  survey
are small towns or old towns now becoming suburban  communities for
a larger city.   Previous  studies have  shown that differences between
 older resident groups and "newcomers" are one of the major  social
divisions in rapidly growing communities.18  Thus, one possible  inter-
 pretation of these data is  that population growth contributes to political
 instability which increases  the likelihood of disagreement over the pro-
 posal of a bond issue. Some evidence in support of this interpretation
 is provided by the fact that  population growth is highly correlated with
 the number of groups participating in  the bond election campaign, and
 the number of groups is negatively correlated with bond passage.   The
 relationship between population growth and outcome might be  described
 by means of an intervening variable,  as shown  in the causal diagram on
 the following page.

  Tax Rate.  Like population growth, the tax rate (variable 83) is negatively
  correlated with bond passage.  Therefore, it appears that citizens  in
  communities with high tax rates are less  likely to support bond issues than
  citizens in communities  with low tax  rates.  This is presumably because
  Population Growth
     Variable 15
                                  .-.16
Election Outcome
   Variable 21
                            Number of Groups
                               Variable  123

                                  55

-------
they are less willing to undertake major new public improvements which
involve the potential for even higher tax rates in the future.

Some caution should be exercised in interpreting these data, however.
The relationship between tax rate and election outcome is not very strong,
and there are some problems involved in the interpretation of a variable
like the tax rate.  First,  it is not unreasonable  to expect that there may
be communities with high tax rates whose citizens prefer high levels of
public service.   In such communities,  one might reasonably predict that
bond issues for sewer and drainage improvements would have a good
chance of success.  Second, the means  of measuring and collecting data
on tax rates  varies from  community to community,  except in those few
states where such data have been standardized.   In spite of attempts to
keep these data as  uniform as possible, complete consistency was not
possible.  This reduces somewhat the utility of  such a variable as a pre-
dictor of election outcome.  Evidence from the in-depth case studies
suggests that it may be a recent increase in the  tax  rate that has a strong
negative effect on the election outcome, rather than the absolute rate.

 Issue Characteristics

 It is clear from the previous discussion that the population characteris-
 tics and community characteristics alone are poor predictors of bond
 election outcome.   The most important determinants of election outcome
 identified by this study are the characteristics  of the issue itself, and
 the politics  of the  campaign which precedes the bond election.  The issue
 characteristics which were considered as  possible determinants of
 outcome are the amount  of the  issue, Federal or State financial assistance
 to  the project,  the proposed means of repayment, and the presentation  of
 the issue on the ballot.

 Amount of issue.  The amount  of the bond issue, (variable 2) is negatively
 correlated with outcome: as the amount of the issue increases, the likeli-
hood of its being passed by the voters decreases.  The direct correlation
                                57

-------
between the amount of the issue and election outcome,  however,  is rather
small (-.18).   On the other hand, the amount of the issue  is significantly
correlated with the existence of groups against the bond issue (variable
35) which, in turn, is associated with bond failure.  This  suggests that
the amount of issue may be important because it stimulates controversy.
This is  consistent with our analysis of the campaign which demonstrated
that criticisms  regarding the costs of the proposed project accounted for
a significant number of the criticisms  raised during the campaign.  Thus,
the relationship between the amount of issue and outcome  could be. ex-
pressed by the following model, where the existence of groups against the
bond issue is an intervening variable between the amount of the issue and
the outcome of the election.
Amount of Issue
  Variable 02
                                .-.18.
Election Outcome
   (% in favor)
   Variable 21
                         Existence of Groups
                         Against Variable 35
 Federal and State financial assistance.  One of the issue characteristics
 which was expected at the  outset of the study to be significant was the
 effect of Federal and State support of the project on the outcome of the
 bond election.  One might hypothesize, since the amount of the issue had
 a negative correlation with election outcorre , that anything which would
 contribute to a decrease in the amount of the  issue (eg, Federal or State
 funds) would also contribute to the probability of  election success.  The
 statistical analysis  indicates that the relationships between Federal and
 State contributions (variables 87 and 88) are not significant.   In the case
 of State  contributions, for  example, the number of cases  is simply too
 small to be significant.  Although the number of cases of Federal sup-
 port is larger, the correlation is  only slightly positive, +.04.   More-
 over,  Federal contribution is not significantly correlated with any other
 variables which might give some idea of its place in the total election
 process.  All that can be concluded is that any definitive statements
 about the effectiveness of Federal and State funding as an incentive for
 raising local capital for pollution  control must be based on a study of a
 larger number of local governments with this experience.
                                58

-------
Method of repayment.   Two variables measure the method of repayment;
variable 77;  repayment by user charges; and variable 78,  repayment by
increased taxes.  An analysis of the  relationship between these two vari-
ables and election outcome shows that the method of repayment has one
of the clearest and most significant relationships.
If user charges are proposed as the means of repayment, it is more
likely that the issue will pass, as the table below indicates:
Means of Repayment
Election
Outcome
Pass
Fail
Total
User Charge
Not Proposed
6
13
19
User Charge
Proposed
14
7
21
Total
20
20
40
If a tax is proposed to repay the bonds,  then there is a greater likelihood
that the issue will fail.
Means of Repayment
Election
Outcome
Pass
Fail
Total
Tax Not
Proposed
11
5
16
Tax Proposed
9
15
24
Total
20
20
40
There are two ways of interpreting the voters' reaction to these means
of repaying bonds.  The simplest interpretation suggests that voters
generally wish to avoid tax. increases and therefore prefer the user
charge which does not involve a tax increase.  In many of the jurisdic-
tions studied  in this survey, the tax issue was  important, and community
                                59

-------
taxpayers  organizations were major actors in some of the election cam-
paigns, often leading the opposition.   For example,  in Roselle, Illinois,
a taxpayers' organization was revitalized specifically to fight the drain-
age and storm  sewer bond issue and was successful in defeating it. Simi
larly,  in Great Falls, Montana, the Taxpayers'  League was a major
opponent of the bond issue.

A second reason why voters prefer user charges to increased taxes has
to do with the notion of equity: those who benefit from the improvement
should be the same as those who pay.   If the local government is pro-
posing to build sewers for only part of the town,  (the argument  runs)
then only that part of town should pay.   This argument is not valid  in
cases where part of the town which is not going to benefit directly from
this  project is  expected to benefit from future projects that represent
comparable levels of investment.

In Round Rock, Texas,  for example,  the government has proposed a
bond issue to finance improvements and extension of the sanitary sewer
collection system; these charges were to take place only in a wealthy
(and predominantly white) part of the city.  Non-white  residents were
not willing to support an increase in taxes to  repay general obligation
bonds for projects from which they would not benefit.

Unfortunately,  sewer and drainage improvements are particularly  vul-
nerable to this  charge--they often serve only part of the community at
the expense of  the whole.   An extension of sewer lines serves only that
part of the city where these new lines will go. Drainage improvements
serve only the  persons who reside in the immediate area where these
improvements  are constructed.  The data indicate that a more popular
solution to the  problem is to repay the bonds with user charges, thereby
allocating the costs of the project solely to those who benefit directly.
                                60

-------
Campaign Characteristics

A number of different variables can characterize the campaign preceding
a bond election.  The study gathered data on the number and types of
interest groups active in the campaign; the types of activities engaged
in; the types of advertising or other promotional materials used and
their cost; the role of the media in the campaign; and the types of
issues raised during the campaign by both the proponents and opponents
of the bond  issue.

Existence of opposition  groups. Variable 35, the existence of groups
opposed to the bond issue, emerged clearly in the statistical analysis
as the most important election characteristic.  The data show that in
communities where opposition  groups were present, the bond issue
was  much more likely to fail;  in communities where there was no organi-
zed  opposition, the issue was  likely to pass.  The relationship between
this  variable and election outcome  is  illustrated in the following table:
Election
Outcome
Pass
Fail
Total
Existence of Opposition Groups
Non-Existence Existence
18
4
22
2
16
18
Total
20
20
40
It is evident from the data that this is the most significant relationship
between a single independent variable and the out come of the election.
In only two cases in which there were pressure groups organized against
the bond issue were the proponents able to prevail.   The results of the
regression analysis further support this observation; this variable
accounts for 41% of the variance in the  outcome, even when combined
with other variables.
                                61

-------
  The apparent success of opposition groups in ensuring the defeat of
  bonds points to one of the most striking aspects of bond elections.
  A campaign against a bond, even one which is poorly organized, poorly
  financed or poorly reasoned can have a significant impact on the out-
  come of the election.  The reason  for this  seems to be the ease with
  which one can create doubt in the minds of the voters.  As Grain,  Katz
  and Rosenthal point out in their study of water fluoridation referenda,
  one of the most successful arguments against fluoridation was that
  very few of the citizens who voted  against the fluoridation referenda
  actually believed that fluoride was  poison; but they were not completely
  sure that it wasn't either.

  Once an opposition group has begun to  make charges about a proposed
  bond issue, the proponents face some difficult strategic decisions.
  They can either ignore the criticisms, respond to them,  or attempt to
  discredit the opposition on some other grounds.  Any of these approaches
  can have negative consequences for the proponents.

  One of the most common  responses by proponents to criticism by
  groups  opposed to the bond issue is to  increase the level  and intensity
  of their campaign,  on the assumption that they can win if they just try
  harder.  Study data suggest that this may not be the  case. As the
  following table indicates,  no matter what the level of campaign effort,
  the bond issue is likely to fail when there is opposition.


 The reasoning behind  this  phenomenon is complex.  Once  opposition
has been voiced and the proponents  have responded, the level of cam-
paign activity begins to es.calate.  More and more controversial issues
are raised,  sparking the involvement of more special interest groups
and ultimately involving more people in the  campaign.  In  the face of
all this controversy, the voter is forced to make a decision.

The construction of a pollution control facility involves complex  questions
of engineering, municipal finance, tax policy and so forth.  Even the best
                                 62

-------
Presentation of the issue on the ballot.   The bond issue may be presented
on the ballot as a single issue;  it may be combined with other bond issues
on the ballot; or it may be divided into several parts,  each of which must
be voted on separately.  Some local officials have strategic reasons  for
choosing a particular method for presenting the issue on the ballot.  For
example,  if a bond for sewer or drainage improvements is unpopular, it
maybe combined with other bond issues which are more popular in the
hopes that the more popular issues will help it to pass.   If it is a large
bond  issue, it may be split into several parts so that the voters have an
opportunity to consider each part separately and approve the parts they
geel are most important.

At the outset of this study,  it was hypothesized that the way in which the
issue was presented would make a difference in the outcome of the elec-
tions.  Indeed,  the data  show a small correlation between these variables
and the outcome: variable 121,  the  combination of the issue with other
bond  issues on the same ballot, had a weak (+. 14) positive correlation
with passage.  Variable 122, the splitting of the  issue on the ballot, had
a slightly stronger, but  negative, correlation (-.22).  These variables
do not appear in our final regression equation because the number of cases
of each variable is too small.   There are only six instances of a combined
issue, and five  instances of a split  issue.  In order to study systematically
the effects of presentation on the ballot on outcome, a sample  of jurisdic-
tions should be  chosen which would represent each of  these types equally.
                                63

-------
                  Consequences of Campaign Activity
Opposition
Level of Proponent
     Activity*
                                        Number of
                                          Cases
Election Outcome
    Yes
    No
     Low


     Medium


     High



     Low


     Medium


     High
                                            11
     0 Pass
     4 Fail

     1 Pass
     7 Fail

     1 Pass
     7 Fail


    1Q Pass
     1 Fail

     3 Pass
     0 Fail

     5 Pass
     1 Fail
*Some groups endorsed the issue; others contributed funds to financa a
campaign effort; some published leaflets, printed posters, or sponsored
newspaper advertisements; some made speeches or appeared in public
debates;  others went door-to-door soliciting votes; etc.  If a group
merely did one of these things,  its level of activity was rated as  "low".
If a group performed two or three of these activities,  its level  was
rated "medium".   If a  group did more than three,  its  level was rated
"high".
educated and informed voters cannot fully understand all of the important

questions contained in that decision.  Thus,  voters look for advice on how

to vote: they look to a person or an institution they respect—someone

they think does understand the issues.
This advice-seeking process takes place  in partisan elections for public

office as well.  Although the voter may not understand all of the issues

at stake in a presidential election, he does know which of the candidates

-------
seems to stand for the policy views he advocates.  If this is not clear,
he will probably look to the political party which has traditionally es-
poused his views for a cue to the way he should vote.

In bond elections, the simple choice between parties is not available.
One cannot simply vote a straight party ticket; one must  actually decide
on public policy.  As a result,  some unique voting phenomenon occur.
Some voters will cast their votes for either the negative  or positive
option, regardless of the content of the item,  because they are what has
been described as "yeasayers" or "naysayers".    Others who have cer-
tain grievances or hostilities toward the political order,  ie, the "alienated
                                                  21
voter", will  vote against anything that is proposed.

In spite of the scholarly interest in these phenomena,  they account for
                                     22
only a small fraction of the total vote.    Most voters are simply trying
to make a difficult decision on the basis  of inadequate and conflicting
information.  Without a political party to offer a cue,  they look to the
news  media or to influential citizens, (eg, the mayor,  councilman,
selectmen, Chamber of Commerce,  League of Women Voters) for advice
on how to vote.  When the advice of respected community leaders is in-
consistent and the information they give  out is contradictory,  the voter
becomes  even more confused.  Rather than vote  to spend money for a
project for which the benefits are perceived to be uncertain, the doubtful
voter will vote against the bond  issue.

Analysis  of the survey data collected during this study also shows that
organized opposition is particularly  effective when it is led either by an
elected official or someone else who is respected in the community.
Opposition by someone "who ought to know all the facts"  is taken by the
voters to mean that there must be something wrong with  the bond issue.
In those  eight cases where there was opposition and a medium level of
campaign activity, there was one case where  some local officials were
campaigning in favor of the bond and some against.  In the five cases
                                65

-------
where there was  opposition and a high level of campaign activity,  there
were three cases of local officials  campaigning on both sides of the issues.
All of these cases resulted in election defeats.  In many of the in-depth
case  studies where bond issues had been defeated--Avon,  Ohio; DeKalb
County, Georgia; and Round Rock,  Texas; for example--the  respondents
pointed to the  "lack of a united front by elected officials" as  one of the
major contributing factors to the defeat  of the bond election.

The town of Avon,  Ohio is a good example of the power of opposition
groups in defeating a proposed bond issue.   Sewers were first proposed
to the voters  in 1963.  Since that time, there have been six more  elec-
tions,  and each time the proposal has failed.  The reason for failure
cited by most  of the respondents in Avon was that the majority of  the
residents simply did not want sewers because they opposed the land
development which they felt would be a natural consequence of sewer
construction.  Despite substantial opposition,  the proponents continued
to try to convince the voters of the sewer's benefits.  In 1968,  for exam-
ple, the Ohio Water Pollution Control Board placed a building freeze on
the city which would be lifted only when the bond issue was passed and
the sewers constructed.   This  pressure  was resented by the  opponents
who then tried to discredit the proponents.  They claimed that the  OWPCB
had "singled out" Avon rather than concentrating on the more important
polluters of Lake Erie, like Cleveland.   They also argued that the OWPCB
was in collusion with the land developers and other groups who stood to
gain financially from the sewer construction.

Several respondents felt that another important reason for the proposal's
repeated failure was the lack of a "united front" on the part of community
leaders.  In each referendum there was  some elected official, either  the
mayor or some member of the council, who opposed the sewers.  This
legitimized the opposition of many undecided residents whose doubts
about the worth of the project were confirmed.  In summary, the  active
opposition and the decision of community leadership has contributed sub-
stantially to the defeat of seven sewer proposals in Avon.
                                 6:6

-------
Role of community leaders.  In addition to local government officials,
the only other groups which were at least moderately active in the cam-
paigns studied were the State government,  the local Chamber of Com-
merce, the local League  of Women Voters, and ad hoc groups which
emerged as a result of the bond proposal.  Except for the ad hoc groups
all were in favor of the bond issue in their communities  on  the other hand,
ad hoc groups were typically taxpayers organizations which had been formed
with the specific purpose of fighting the bond issue.

 These groups varied in the level of effort they contributed to the cam-
 paign.  The State governments and League of Women Voters rarely con-
 tributed more than a low level of campaign effort.   This was because
 the State  government was concerned that its campaign efforts might be
 interpreted as outside interference in  local affairs and the  League of
 Women Voters  is typically interested only in educating the  voters.  The
 Chamber of Commerce and  the ad hoc^ groups, on the other hand, were
 much more active in the campaign, as the following table indicates.
Group
State Government
Chamber of
Commerce
Ad Hoc Groups
League of Women
Voters

Low
11
7

2
8
Level of Effort
Medium
1
6

6
1

High
-
9

2
~
 The most interesting data on these four organizations concern their success
 of failure in getting the voters to support their side of the issue.  The fol-
 lowing table shows the position taken by each group during the campaigns
 in which they were active,  the results of those campaigns, and the per-
 centage of the time when the group was  successful in supporting or opposing
 a bond issue.
                                 67

-------
The data in this table show that the two groups which are the most active
in bond election campaigns--the State government and the local Chamber
of Commerce--are also the least successful in obtaining support for their
position.  Both groups campaigned overwhelmingly in favor of the bond
elections, and  in a majority of cases they were unsuccessful. On the
other hand,  the two groups which had a lower level of campaign activity
were  more successful—particularly the League of Women Voters.  Part
of the League's success may be due to its research of the issues  involved
in the proposed projects; the League rarely supported issues where there
was opposition.
Group
State
Government
Chamber of
Commerce
Ad Hoc Groups

League of
Women Voters
Activity
In Favor
Against
In Favor
Against
In Favor
Against
In Favor
Against
Pass
4
1
9
1
2
0
6
0
Fail
7
0
12
0
2
6
1
1
% Success
36%
0
42%
0
50%
100%
85%
100%
 The data contained in these last three tables indicate the variety of
 interest groups which can become involved in a campaign,  heighten
 controversy and contribute to the defeat of the bond.  This suggests
 that, to ensure passage,  the proponents of the bond issue should be
 concerned about the ideas and  interests of these groups and attempt
 to gain their support for the project when it is being formulated.

-------
Partisianship of special interest groups.   There were 35 jurisdictions
which reported  that interest groups had been active in the campaign
over the bond issue.   One of the questions asked of respondents in all
of the communities studies was whether the  groups active in the cam-
paign had been partisan,  ie, were political parties, or closely allied
with a political  party.  Of the  25 communities where groups were active
in the campaign, 10 reported that one or more of the  groups were parti-
san, and 15 reported that all of the groups were non-partisan.  Relating
partisanship (variable 37) with the outcome of the election indicates that
the alignment of an interest group with a political part will frequently
contribute to this defeat  on a bond issue,  as evidenced by the following
table.
Partisanship of Groups
Election
Outcome
Pass
Fail
Total
No Groups
11
4
15
Non-Partisan
Groups
8
7
15
Partisan
Groups
1
9
10
Total
20
20
4a
It is clear from this table that the most desirable situation for a success-
ful election is to have no groups at all; the next most desirable situation is
to have non-partisan groups; and the worst situation, in terms of success,
is to have partisan groups active in the campaign.   In fact, where partisan
groups were present only 1  of the 10 elections was successful.

On the surface, it would seem that this phenomenon  contradicts the idea
that political parties are able to mediate  conflicting  interests and thus
contribute to a government's ability to take action.   But this  concept
applies only to cities where the political parties are strong and take
leadership positions.   These conditions do not apply to the majority of
local American governments, and especially not to the non-partisan elec-
toral systems which predominated in our sample.
                                 69

-------
Not only do political parties fail to resolve conflict in this situation,
but the introduction of partisanship into a non-partisan campaign actually
increases the level of conflict.   As Wood points out, in non-partisan
communities there is:

            ". .  . an outright reaction against partisan activity,
            a'r'efusal to recognize that there may be persistent
            cleavages in the electorate,  and an eithical disap-
            proval of permanent group collaboration as  an ap-
            propriate means for setting public disputes. "^

Vote  react negatively to the introduction of party  politics in a bond
election campaign.  They become suspicious of the motives of the parti-
san actors, since they believe that only  "the public interest"  should
motivate voters and actors in elections.   Thus, the work of partisan
actors contributes substantially to election defeat.

 Criticism Raised During the Campaign.   Another variable which charac-
terizes tne campaign,  and  one which is closely related  to the existence
of opposition  groups,  is variable 119 the number  of criticisms  raised
during the campaign.   As the following table illustrates, there  is a
 significant positive correlation between  the number of criticisms raised
and the defeat of bond issues.   Those projects which  stimulate  criticism
and controversy are  much  more likely to be defeated by the voters.
Election Number of Criticisms Raised
Outcome 0-3 4-10 Total
Pass
Fail
Total
17
9
26
33
11
14
20
20
40
 The data permit an analysis of the frequency of different types of criticisms
 raised during the election campaigns and the relation of these types of
 criticisms to the election outcome.   The five criticisms most frequently
 raised during bond election campaign are found  in the following table.
                                70

-------
This table indicates that there are three different ways that the cost of
the project can become  an issue in the campaign: the opposition can say
that the bond will cause an unacceptable tax increase or a sewer fee in-
crease; it can  say simply that it is too expensive; or it can say that the
distribution of costs and benefits is  inequitable.  Although the  last crit-
icisms is raised least frequently, it is most likely to contribute  to fail-
ure.   This may be because it is easiest to prove.  For example, if the
government is proposing a project which will benefit only part of the com-
munity and is attempting to tax the entire community for its  financing, the
inequity of the situation can easily be pointed out to  those persons whowill
be taxed but receive no benefits.
Frequency
Mentioned Pass
Passage of bond will
cause an unacceptable
tax or sewer fee in-
crease
Incorrect level of gov-
ernment is proposing
the project (eg, should
be regional if local)
Cost or size of project
is too large
Inequitable distribution
of costs and benefits
Project is not technically
necessary
26 12

16 5

13 3

10 . 1
10 4

Fail % Fail
14 54%

11 69%

10 77%

9 90%
6 60%

 The criticism that project is not really necessary is difficult to respond
 to.  For example,  in Avon,  Ohio, the proponents answered this criticism
 by pointing to a judgment by the Ohio Water  Pollution Control Board that
 the town was polluting  Lake Erie.  The opponents replied that it was
 doubtful that Avon was polluting Lake Erie as much as Cleveland, and
                                 71

-------
they would be happy to control pollution by some means other than the
elaborate sewer project in the small town of Avon was appropriate solu-
tion to the problem.

The argument concerning the appropriateness of the level of government
taking action is another difficult criticism to handle in a campaign.   It is
clear, for example, that the Federal government has endorsed the idea
of regional solutions to water pollution control.  But when this argument
is raised  during an election campaign--either for or against regional
solution--it tends ultimately to defeat the bond issue.  The criticisms
creates another decision which has to be made  by local government and
therefore another issue around which opposition forces  can organize.
                               72

-------
An Analytical Model of the Factors Influencing the Outcome of Bond
Elections
As indicated in the introduction to this chapter^  the major purpose of the
regression analysis was to develop an analytical model which would explain
the relationship between the important factors in the water pollution bond
election process and the outcome of the referendum.   A series of regres-
sion analyses were performed on different groups  of variables.  The
regression analysis presented here identified the eleven variables dis-
cussed in the preceeding sections which explained a satisfactory amount
of the  variance in outcome.

Byway of review, the basic concept of regression analysis is to produce
a linear combination of independent variables which in combination will
correlate as highly as possible with the dependent  variable.  This equation
can then be  used to "predict" values  of the dependent variable. The
difference between the value of the dependent variable and the value pre-
dicted by the linear  combination of the independent variables is the
"unexplained variance, " also known as the "residual. "  The regression
equation is then written as follows:

                   D - bill  + b2I2 .  . .   + bnln + c + r

where D is the dependent variable, the I's are the  independent variables,
the b's are the regression coefficient (unnormalized), c  is a  constant and
r is  the  residual.

The  regression equation produced by our analysis  is as follows:

Outcome = -1.71180 (Var 035)  1.64499 (Var 078)  +2.04666 (Var  109)
           -0. 19989 (Var 119) +0.20139 (Var 098)  -0.88602 (Var  037)
           -0.58100 (Var 077) -0.99979 (Var 015)  -0.00003 (Var  097)
          + 0.00001 (Var 002) -0.03124 (Var 083)  -6.58936+ residual
                                73

-------
In other words, to obtain the outcome in terms of the percentage of votes
in favor of the issue (variable 21), one multiplies the values taken by
each independent variable by the coefficient preceeding it; adds together
the results of each of these multiplications; then add that sum to the con-
stant 6. 58936.  The extent to which this  sum differs from the actual value
of variable 21  in any given case is the amount of unexplained variance or
the residual.

The selection of this equation has been guided by several considerations.
First, the variables to be included had to be either quantitative or dichot-
omous.  This constraint is imposed by the technique itself as explained
in Appendix C.  However, as also indicated in this appendix, most of the
variables can be expressed in these terms and are consequently eligible.

A second  and more restrictive requirement  is  that the variables be as
independent as possible from one another (that is, not strongly inter-
correlated) though individually correlated with the outcome.  By choosing
such variables, one attempts to get as close as possible to a "factor
analysis" model where the variables are independent and represent a
whole set of variables with which they are  correlated but which are kept
out of the equation.

The next requirement   is imposed by the size of the sample and the
nature of  the regression technique,  which together necessitates that the
total number of variables  retained should be minimum and in any case
lower than 40, (the size of the sample).  For a given amount of variance
accounted for by the equation, the lower the number of variables  in the
equation,  the  better the result.  This minimization is done automatically
by the computer model that was used.

The amount of variance explained by this model is 62%  and falls within
the range  customarily explained  by regression analysis  of  social science
problems.
                              74

-------
The regression equation developed with these constraints explains 62%
of the variance, which is comparable, or even slightly higher than what
is generally obtained for this type of research.

In order to evaluate the weight of a variable in the total result, one
generally "standardizes" the variables,  that is, reduces the range of
variation of each variable  to the interval 0-1.  This is  done by sub-
tracting from each variable  its mean  and dividing the result by its
variance.  It should be noted that there is no constant in the standardized
regression equation.  The new equation obtained is the  following:

Variable 21 = -0. 34963 (Var 035) -0. 33363 (Var 078) +0. 28022 (Var 109)
Outcome      -0. 21816 (Var 119) +0. 14343 (Var 098) -0. 14451 (Var 037)
              -0.12012 (Var 077) -0.08437 (Var 015) -0.04491 (Var 097)
             + 0.0482  (Var 002) -0.02221 (Var 083)

Under this form, it is easy to see how much each variable contributes
to the result,  (that is, to the value of variable 021), since all of  the
variables now vary between  zero and  one.  In addition,  the sign of each
of the coefficients indicates  the  direction in which each variable
influences the outcome.

 Thus, the variables which tend  to reduce the proportion of votes in
 favor of the referendum are most importantly the political variables
 (especially the existence  of special interest groups against the issue,
 variable 35,  and the associations of increased taxes with the  issue,
 variable 78.  The  variable  that acts the most in favor  of a positive out-
 come is the existence of a pollution problem (variable 109).   These
 three variables together would  explain 55% of the variation of the out-
 come.  All the other variables  add only 7% more in the numerical
 result.  However, they add substantially more to the full understanding
 of the process.
                               75

-------
In both cases the standard error is approximately 1. 7  This means that
the difference between prediction and actual value of the percentage of
votes in favor of the issue is in average 1. 7%.  However, if the predic-
tion error is low in average, it can be very high  on a particular  case.
If the regression equation is used on a very large number of referenda,
the errors made  on each one will cancel out but the average error on
the whole set of predictions  cannot be reduced below 1. 7% no matter
how large the sample.
                              76

-------
A Model for Forecasting Bond Election Outcome

In addition to the analytical model described above,  we were also
interested in developing another analytical model using only the variables
which could be measured well in advance of a  bond election. Unlike
the first model which is highly dependent on the values of the campaign
characteristics,  this second model would enable the user to forecast the
outcome of the election before data on the campaign variables were
available.

The forecasting model relates the following variables  to the outcome,
with the associated coefficients (using standardized variables):

Variable 21 =+0.38574 (Var 109) -0.48607 (Var 078) -0.19497  (Var 002)
Outcome     -0. 31035 (Var 015) +0. 20034 (Var 116) -0. 17686  (Var 097)
             + 0. 12395 (Var 098) -0. 06237 (Var 083) +0. 04372  (Var 102)
              -0. 02243 (Var 077)

This  equation explains only 46% of the total variance in outcome.   The
small amount of variance explained makes this an uncertain tool for
forecasting outcome.  But this fact,  in itself,  further  emphasizes the
importance of the political variables.

The existence of a pollution problem (variable 109) still plays a con-
siderable role  as  does the repayment of the bonds by increased taxes
(variable 78) even more important than in the previous equation.  Such
variables as the amount of the issue (variable 02) and  the rate  of popu-
lation growth (variable 15) now also intervene with more emphasis.
Two variables  which did not add any significant amount to the variance
in the previous equation are now reintegrated.  These  two variables are
the indebtedness  per capital (variable 116) and the percent of persons
over age 65 in  the population (variable 102).  The difference in the pre-
sent case is that  it is not possible to envisage reducing the  number of
variables to the first two or three as the full equation  explains  a smaller
percent of the variance than do the first three variables of the  model in
                                77

-------
in the previous section.  One of these three variables, however,  sum-
marizes the whole campaign process whereas the second equation in-
cludes mainly data of demographic type.

Although it has been possible to explain a significant amount of variance
in the outcome when the campaign variables are included, one must con-
clude that it does not appear to be feasible to develop a significant re-
gression model which excludes them.
                              78

-------
                            CHAPTER V
             CONCLUSIONS AND RECOMMENDATIONS

Introduction

Critical to the  success of any water pollution control program is the
capacity of the  local municipality to raise their share of the financing.
Whenever the financial decision of the  municipality is subject to a
bond election or referendum, there is  a risk that the voters will decide
against the proposed project.  Unfortunately, there are no systematic
data on the prevalence of the referendum for this purpose or the per-
centage of referenda that are defeated. However, expert opinion,
corraborated by analysis of the existing data, suggests that about 40%
of all sewer and other water pollution  control projects are subjected
                                            25
to referenda and that of these, about 15% fail.    While  this amounts to
only about 6% of all of the  local financing raised for this purpose, it
still represents a major factor in the successful completion of water
pollution control projects.

Moreover, analysis of referenda can provide some useful infor-
mation about the "readiness" or "willingness" of local communities
to undertake disruptive and expensive public improvements for water
pollution control.   When there is a referendum, the decision-making
process of the  local government is exposed to the interested observer.
Thus, the bond election permits the observer to see more clearly
what issues are of concern to the electorate and what factors may
contribute to the defeat of  water pollution control projects in general.
Therefore,  implications may be drawn from the bond election situation
which have broader application than the referendum situation alone.
                               79

-------
This study of the factors influencing the outcome of municipal bond
referenda has identified three major aspects of the municipal  bond
process which, if changed, would increase the likelihood that  a pollution
control facility would be constructed: the process by which the facility
is planned; the financial environment, particularly the bond market;
and the approval mechanism required.

The major conclusion of this study is that bond elections for projects
of this type are won or  lost during the initial planning of the project.
Critical decisions are made at this  time concerning who will benefit
from the improvement,who will pay the costs, and how they will be
paid.   If mistakes are made at this  stage, or if there is  poor coordina-
tion with everyone potentially interested in the project,  there  is little
that the issue's proponents can do to rectify the situation.   A decision
about the means of financing is an especially important part of the
project plan.  The results of the case studies reflect clearly the voters'
aversion to increased taxes and their desire for an equitable repayment
mechanism.  As  seen in Chapter IV,  the total amount of the project has
a substantial indirect influence on the outcome of the  bond election.
Anything which local jurisdictions can do to reduce the amount of the
project with outside  funding will improve the chances of the bond's
passage.  Thus,  the first series of conclusions and recommendations
deals with the process of planning the facility.

A second major  influence on the outcome of the bond election is the
financial environment in which the bonds are formulated.  Since there
are a large number of bonds and a relatively small number of buyers,
the buyer of the  bonds exercises considerable influence  on the legal
and financial terms of the  bond, tailoring them to his own advantage.
Further, since the bond market is virtually immune to any influence
from the local political situation, changes  in the relationship  between
the buyer and seller of the bonds will have  to come from some outside
institution.  Hence,  another series  of recommendations  has been
included which are intended to reduce the amount of influence  that the
                               80

-------
bond market has on the financial planning of the bond, and thereby shift
the focus of the local government's attention from the interests of the
buyer of the  bond to the interests of the electorate.

The third aspect of the process has to do with the mechanism for
approving pollution control facilities.  While many municipalities are
required to hold elections to authorize the  sale of municipal bonds,
others have alternative means of approval available to them.  In some
cases it is desirable to hold a bond referendum even when none is
required by law.  However, many local governments have the power to
avoid this mechanism and would prefer to do so,  but are unaware of
the alternatives.  Consequently, the third set of recommendations
focuses on increasing the local government's knowledge of alternative
financing and approval mechanisms available to their jurisdiction when
initiating a water pollution control project.
                                 81

-------
The Planning Process

The Importance of Participatory Planning. An analysis  of the data from
the 40 case studies indicates that a high level of campaign activity is
the most reliable predictor of a bond's election failure.  In only two
cases in which there was any opposition was  the referendum able  to
pass.  Further, where there were a large number of criticisms,  the
referendum passed only three times.  Additional analysis of the level
of campaign effort by local officials shows that a project facing opposi-
tion is nearly always  unsuccessful, regardless of the amount of effort
put into the campaign, while any level of campaign effort without opposi-
tion meets with success.   The major factor determining the outcome
of the referendum is therefore the presence of opposition groups and
the availability of issues •which they can criticize.

Evaluation of the in-depth case study material corraborates this conclu-
sion.   During these studies, the  researchers frequently  asked why
particular bond issues failed.  One of the most common  responses was
"because  of politics. " However; the analysis of our data has shown
that it was not  "politics, " but the absence of  politics which  was the
main cause for defeat  of many of these bond elections.  Politics can be
defined as the art of compromise, a means of resolving  conflict,  or a
mechanism for representing all interests.  When a bond issue failed,
it was  frequently because all interests had not been represented.  The
bond issue offended some group who then publicly opposed it, and the
proponents of the bond issue refused to compromise.

In Round Rock  Texas,  for example, persistent and vocal opposition
arose against a sewer project which the opposition charged slighted
poorer Black and Mexican-American sections of town (which had exper-
ienced very serious problems),  while  benefiting affluent land developers.
The voters made aware of inequities in the project,  defeated it at the
polls.  The mayor, however,  recognizing that the leader of the opposition
                               82

-------
had some valid criticisms and was gaining political strength on the
basis of the sewer issue, restructured the project to respond to his
criticisms.  Despite some hard campaigning by the opposition leader,
the issue passed easily in the next election.  The mayor had been
flexible enough and realistic enough to recognize that the project itself
had serious weaknesses and failed to satisfy the needs of important
interest groups in the community.  Once their needs were  incorporated
in the project, the opposition lost its power and the issue passed.  In
sum, when all of the relevant interest groups are consulted before,
the project is formulated, and their interests are taken into consideration
in the design of the project and the selection of a financing  scheme,
the opportunity for any serious opposition is significantly reduced and
the election campaign itself assumes much less importance.

Since the existence or non-existence of opposition to the bond issue
generally determines the outcome of the election, anything more than
a low level of campaign activity will either contribute to the election's
failure or  turn out to have been unnecessary.  If there is no opposition
to the project, there is no need for a high level of campaign effort.   On
the other hand,  when there is opposition, any level of campaign - low
or high - is unlikely to be successful.  If the proponents decide to keep
the level of campaign activity low, they run the risk that the opponents
will succeed in convincing a large proportion of the electorate to vote
against the bond issue.  On the other hand,  if they decide to respond
to the charges of the opposition, they risk becoming embroiled in a
controversy which can decrease the voters' confidence in the bond issue
and result in defeat.

This  study provides no answer to the question of how to deal with oppo-
sition in a bond election.  This depends on the facts of each individual
case  and on a careful assessment by local leaders of the potential costs
and benefits of each strategy. However,  it is possible to recommend that
proponents of the project perform all of the groundwork necessary
before the campaign begins so that opposition will neve arise.  The
                              83

-------
preparation that needs to be done includes consulting interested indi-
viduals and groups,  choosing on equitable means by financing, and seeking
out Federal and State financial support where available.

Groups to be Consulted.  In each community there are a number of
different groups which may be affected by a proposed water pollution
control project.  Many local officials do not realize the importance of
consulting  these groups,  primarily because they regard decisions
about water pollution control projects as technical in nature  to be  de-
cided by experts.  But it is clear from the evidence of the 40 case
studies that these decisions  also involve fundamental questions about
the values  and objectives of the community, and the allocation of re-
sources in a manner consistent with  these aims.

Even when a local government is aware of the importance of consulting
all of the groups which may be affected by the improvement, they may
not know whom to contact.  Although the types of groups that are likely
to be interested in water pollution control vary from community to
community, there are several which can be found in many communities
and which behave in relatively predictable  •ways.  A partial listing of
these groups, their  interest in water pollution control projects, their
likely position on a proposed project, and the resources they can bring
to bear in a bond election campaign are found in the table on the follow-
ing page.

In addition to this  information on the interests of special groups,  local
officials should bear in mind two additional facts.   First,  very few of
these interests are monolithic.  For example,  even though the Chamber
of Commerce may support the bond issue,  there may be divisions  of
opinion within the  business  community over the nature of the sewer or
drainage improvements.  The extension of sewers into a new residential
development may give one developer a comparative advantage over other
developers and therefore may generate  opposition from the latter.
                               84

-------
                             GROUPS TO BE CONSULTED
    Group
        Interest
                                 Likely posi-
                                 tion on bond
              Resources
Chamber of
Commerce
Taxpayers
organization

Good govern-
ment organiza-
tion (League of
Women Voters)

Local and
regional planning
agencies

Homeowners
Minority groups
Land development,  economic
development, employment,  low
tax rate

Low tax rate, low assessment
Ending water pollution
Informed electorate
Honest government
Ending water pollution
efficiently
Consistency with existing plans

Preservation of character of
   community
Low tax rate

Equitable distribution of costs
   and benefits of project
Improved public  services
For
Against
For
For
Against
Unknown
Money
Prestige
Numbers of
   members

Prestige
Research
Research
Authority to
  withhold funds

Number of
  members
Cohesiveness

Number of
  members
Cohesiveness

-------
Second, there may be groups which do not appear until the issue is
proposed to the voters.  For example, some of the most active groups
are ad hoc groups which form around some aspect of the bond issue.
These groups usually form to rectify an inequity in the distribution of
costs and benefits of the proposed bond issue.  One of the simplest
ways  of preventing this type  of opposition is to exercise care in the
planning process to eliminate inequities from the proposal.

Alternative Means  of Financing.  The method of repayment has one  of
the clearest and most significant relationships with the outcome  of
bond elections.  If user charges are proposed as the means of  repayment,
it is more  likely that the issue will  pass; if a tax is proposed,  then
there is a greater likelihood of failure.   There  are two possible  inter-
pretations  of the voters' preference for  user charges over a tax  increase.
The most direct is that the voters simply balk at a tax increase of any
kind and are more  likely to defeat projects which require them.  The
second interpretation involves the notion of equity:  those 'who benefit
from the project should also  pay for it,  and user payments are inher-
ently more equitable than a tax levy.  Both of these interpretations may
apply to an individual bond referendum.

Given the  voters' explicit disdain for tax increases,  it is  surprising
that the general obligation bond, with its clear consequences for the tax
rate,  is chosen so  often by local governments as the means of  repayment.
The major reason for this is the fact that there is a larger market for
general obligation bonds,  primarily because the risk of default is pre-
sumed to be lower,  since the full faith,  credit and taxing  power of the
local  community is supporting it.

Revenue bonds are riskier and hence  more costly to finance.  In
addition to requiring a higher interest rate,  (commensurate with the
amount of risk being assumed by the buyer),  they carry higher admin-
istrative expenses.  Since the amount of revenue to be collected is based
on the value of services rendered (user  charges) or the value of
                              86

-------
property benefit (assessment),  they must be equitably assessed.  This
can be a costly process for a local government to undertake.  In
addition, because of the variety of ways of determining the fair charge
administratively and judicially, these assessments are particularly
susceptible to lawsuit.

Although the most prevalent means of financing improvements of this
type are general obligation or revenue bonds issued by local govern-
ments,  there are at least two other options which are seldom used.
One  such alternative is the special assessment bond which is  repaid
from an assessment against those who benefit from the facilities
constructed with the proceeds.  The major reason why special assess-
ments are  often avoided is  because the whole problem of assessment
is complex and unfamiliar to local government officials.  Moreover,
when there is a decision to make  a  special assessment,  the means of
assessmentmay be inequitable, and hence a campaign issue around
which opposition can organize.

A second alternative means of financing water pollution control projects
is the creation of a special district. A special district is effectively
another layer of government placed on top of an existing layer.  It
provides a means for tapping the  same revenue  base  without exceeding
debt limitations.  From the viewpoint of  the  voter, the major advantage
of the special district is that it assures equity of those who are served
by the improvement and those who pay.

Federal and State Financial Support.  As  the total amount of the issue
increases, the likelihood of its passing decreases.  Although the  direct
correlation between the size of the issue and the outcome is not large,
it does appear that larger issues  are more likely to arouse controversy
in the campaign, which in turn leads to the bonds' defeat at the polls.
This  evidence suggests that local governments  should seek alternative
means of financing water pollution control projects in order to keep
the total cost of the bond issue  as low as possible.
                              87

-------
According to the National League of Cities' publication entitled Federal
Aids to Local Governments, there are at least twelve different grant
and loan programs providing funds for sewer, drainage and other water
pollution control projects.  These range from the Housing Act of 1954
which provides loans for the financing  of plans for all types of public
works  projects in urban areas,  to the Appalachian Regional Develop-
ment Program, which provides grants for public facilities which contri-
bute to the economic growth of Appalachia, to the Federal Water Pollu-
tion Control Program which provi des financial assistance  of all kinds
to local governments for the planning and construction of various types
of water pollution abatement projects.   The Department of Agriculture,
through the Farmers Home Administration, also provides  loans for
public  facilities in rural areas.   Many states contribute to the planning
and construction Of water pollution control  projects,  although their
programs are generally smaller in scope.

The  Federal government, as well as the National League of Cities
and similar organizations,  has attempted to simplify the process of
applying for and obtaining grants by publishing catalogs of grants
available for different types of projects. But the process  is inherently
complicated, as reflected in the existing catalogs, especially for the
small municipality which is not large enough to have  the resources to
spend on "grantsmanship. " This is especially true for the small water
and sewer districts which are often responsible for administering  sewer
and drainage projects.  Moreover, none of these catalogs  provides in-
formation on state programs available  to local governments.

Recommendations Concerning the Planning Process.   Abt  Associates
recommends that the Environmental Protection Agency assist local
governments in planning water pollution control projects,  particularly
those requiring a bond election.   The objectives of this assistance  would
be to:
                              88

-------
    Increase local officials' awareness of the implications of
    their decisions regarding the nature of the proposed project
    and the means of financing that project for the outcome of
    the election;

    Increase local officials' awareness of the importance  of
    consulting all interested groups before  making these
    decisions;

    To provide local officials with the information and insight
    necessary to make decisions which will be more acceptable
    to the majority of the citizens they serve.


Abt Associates further recommends that this information be disseminated
through a handbook for local government officials prepared by the

Environmental Protection Agency in cooperation with the State water

pollution control agencies, and distributed by these State agencies  to
local governments.  The major reason for involving State agencies in
this process is to  take into consideration the important legal and admin-

istrative conditions concerning the bond election process which vary

from State to State. Therefore, these handbooks  should be tailored to
meet the requirements of each State.


The following  items should be included in the handbook:


    A  check-list of important community groups which might be
    concerned with water pollution control and sewer  or drainage
    improvements. This check-list should also indicate in detail
    the concerns each group would be likely to have regarding
    any kind of water pollution control project.

    Complete and easy-to-use 'information on grants available
    from the State and Federal governments,  including a des-
    cription of the aims of each program,  the amounts of  assis-
    tance available, eligibility requirements, and instruction
    on where  and  how to apply, and  how to obtain qualified
    technical  assistance where necessary.

    Descriptions of alternative means of financing sewer and
    drainage improvements, with emphasis on more innovative
    schemes which are acceptable both to the voter and the
    ultimate buyer of the  bond.
                               89

-------
The Role of the Municipal Bond Market

As indicated previously,  the financial character of the bond, and in
particular its method of repayment, has a significant impact on the
outcome of the bond election: those which are repaid by taxes  are
much less likely to pass than are those repaid by a direct charge to
the users of the improvement.  Local government officials are often
not aware of the implications that this aspect of the bond has for the
outcome of the election.  Often  this choice is delegated to a controller
or treasurer, who may in turn hire outside financial consultants.
However, even when local officials are well informed about the elec-
torate's preferences for revenue financing, they are not always able
to choose this approach without considering other factors.

The most important of these factors is the potential market for the
bond.  Three major forces  affect the  operation of the municipal
bond market.  First, there are a very large number of municipal
bonds issued at one  time  and a relatively small number of buyers.   By
way of contrast, it is interesting to note that the nation's 500 largest
corporations account for  nearly all of the common stock, preferred
stock and corporate bond issues in any given market period, but the
New York Metropolitan area alone  has 1,400 governments whose bonds
compete for buyers.  Because of technical variations in each individual
issue, "guides" are required to lead the issue into the market. These
guides are actually municipal bond counsellors who act as agents of
the bond market by communicating to  the local government the terms
under which the market will buy a bond issue.   Even when employed
by a local government,  the  bond counsellor's major responsibility  is
to the eventual buyer of the bonds.   Although he may be concerned  with
the general question of  whether the electorate  will approve the bond
issue, his responsibility  to his  profession dictates that he prepare a
technically perfect and  highly marketable bond issue.  Unlike the local
politicians,  his professional performance will be judged in the bond
                               90

-------
market and not at the polls.   Thus, local political considerations are
ignored in favor of the market conditions.

The  second major force which influences the market is the predominant
use of underwriting, with no risk assumed by the underwriter.  Thus
no money changes hands until a buyer is found.  Pressure to sell the
bond is therefore  on the issuer,  not on the underwriter, although he
is the one  trying to place the bonds.  As a result, the  underwriter may
adopt a "do it my  way or don't do it" attitude in trying to sell the
issuer's bonds.  Once a buyer is found,  the issuer is free to accept the
buyer's terms or  not sell at all.  Thus,  in the  realities of the  market,
the underwriter tends to act more in the interests of the buyer than
the seller.

"While all bonds  are fair investments, some bonds are excellent invest-
ments.   Thus, the third major influence on the market is the attract-
iveness of a few very good bonds.  The  industrial revenue bond exper-
ience shows that when a very attractive bond hits the market,  weaker
bonds are  not sold.  This does not stem from a lack of investment
money, (since nearly all municipal bonds can compete favorably with
corporate  securities and loans), but  from the pervasive use  of private
placement. This  transaction mechanism is slow because issuers who
need money quickly must compete for "early money" and limited be-
cause only a certain portion of the investment market,  namely the
institutional buyer,  can be reached.  However, due to the impossibility
of marketing the great bulk of municipal securities which are in large
unwieldy "one-shot" issues, or a large  number of small issues,  the
market must continue to be private.

The  most important terms which the bond market influences is the means
by which the bond will be repaid.  This is the major distinguishing char-
acteristic  between  general  obligation and revenue bonds.  Although  the
interest is higher  on revenue bonds,  the investor   generally prefers the
lower risk general obligation bonds which are backed by the  taxing
                               91

-------
power of the local jurisdiction.  On the other hand, as  the data have
indicated,  voters prefer revenue bonds because they do not appear
to result in an increase in taxes. However, revenue bonds have a
number  of subtle hidden difficulties which make them equally unsatis-
factory.

First, when revenue bonds are proposed, there is a tendency for the
local government to ignore consideration of their population's ability
to pay for the services which the  new facilities will provide.  Although
this is done as a matter of course when a tax increase is proposed,
it is not  done when a user charge or a property reassessment will be
required.  Consequently, there is a greater chance of default if it
turns out that the population is unable  to afford the new service.  To
avoid default, the local government will find itself faced with higher
administrative expenses than it anticipates as it attempts to collect the
revenue  for the facility or locate new sources or revenue.

The unexpected additional cost of collecting revenue will fall on the
already much higher administrative expenses of revenue bonds,  resulting
in part from the additional costs of computing the user charge or assess-
ment,  but much more from the increased number of hearings, suits,
and adverse legal action in general.   Since user charges or assessments
are theoretically based on the value of the service or the  increase in
the value of the property, they must by law be equitably assessed. How-
ever, from the  standpoint of the local government,  the basis for the
calculation of the user charge or property assessment is  not the value
added to  the individual household, but an equitable portion of the total
amount of revenue required to service the debt.  In matching valuation
to debt service, therefore, there is an implicit assumption that the
increased benefit to all the individual owners and the  capital expenditures
are equal.   Yet, there may be a need for capital expenditure to benefit
the community as a whole,  though benefit to individuals would not be that
great.  In this situation, the bonds should theoretically default or

-------
standards of valuation would have to be bent.  Generally,  the latter
occurs, and thus the revenue bond breeds lawsuit, as administrative
and judicial standards of value  start to diverge.

These and the other terms  of the bond are set forth in the indenture,
the contract between the  buyer  of the bond and the issuing authority,
which states the rights and obligations -of both parties.  In this highly
competitive market situation which the bond buyers can pick and choose
their investments,  the  terms of the indenture agreement may determine
whether or not a local community's bond will sell.

In order to lessen the amount of influence the bond market has on the
financial character of municipal bonds, Abt Associates recommends the
following:
     1.  The Environmental Protection Agency should formulate
        a model indenture for municipal water pollution control
        bonds.
     2.  The EPA should investigate the feasibility of a model
        statute to establish state bond banks.
     3.  The EPA should investigate the possibility of developing
        new types of municipal financing for water pollution
        control facilities  which would have characteristics
        attractive both  to the voters and to the investors.
The Model Indenture.   Through the indenture,  the bond buyer is present-
ly able to exercise a major influence on the financial terms of a munici-
pality's bond issue. This situation could be  changed by standardizing the
indenture utilized by investment bankers and bond counsellors.  The
present informal indenture practice increases the selling costs to the
municipality,  while providing only ambiguous protection  to the holder
of the bonds.  By contrast, a model indenture would enable local gov-
ernments to predict difficulties in the indenture formulation more
accurately. It would also provide holders with more effective protection,
thereby lowering the amount of interest  that  the bond would have to pay
                                93

-------
to attract a buyer.  Although there would necessarily be a wide  range of
financial options possible under the model indenture,  the increase in
standardization would eliminate minor differences between essentially
like bonds and reduce some of the unnecessary competition among
sellers.   This would enable local government officials to give more
consideration to the preferences of the electorate when formulating a
bond issue.

It is recommended that, through research and conference with legal
and financial experts, the Environmental Protection Agency formulate
a model indenture for municipal water pollution control bonds.  The
Trust Indenture Act may be used as a starting point,  with an eye toward
rewriting it where financing realities diverge.  Once formulated,
promulgation of the  model indenture should proceed in two directions:
toward acceptance by State  legislatures as a part of the municipal
code, and toward  informal  acceptance by the legal and financing commu-
nities.   Initiative and leadership will be required on the part of the
EPA in the acceptance as well as the formulation stage.

State Bonds for Water Pollution Control Facilities.  While the model
indenture addresses itself to some of the  legal technicalities •which
influence the formulation of bond issues,  the basic problem  of the
buyer's market still remains.  Another approach to this problem which
is worthy of further investigation is State bonding of water pollution
control projects.  This approach is attractive because State securities
avoid many of the difficulties inherent in municipal bonds.  Unlike
municipalities  and other local issuing authorities, the  states are limited
in number and carry  continuous, up-to-date  credit information.  As a
result, their bonds are more marketable.   Since the  states have
greater financial resources and more flexibility in the way they spend
their money, they can issue bonds periodically,  or when market condi-
tions are favorable.
                                94

-------
A system for  state bonding of water pollution control projects would
function like a state bank:  local governments would issue their own
securities,  and when they encountered problems in the market, the
State would buy the bonds.  They could buy them either at a depressed
                                   o /
rate or at close to the market rate.    They would then issue their own
bonds to provide the money to buy the local bonds.  Additional bonds
would be issued for actuarial purposes, -with the state taxpayers footing
the bill.   However, in a theoretical model, no new state taxation need
be imposed.  The State would make enough profit on this  to be able to
pay for administrative expenses and still build an actuarial reserve.  In
Vermont, where a State bond bank is operating,  this has  been the ex-
perience to date.

The advantages of such a bond bank would be:

     Market control over local issues would be pared to a
     minimum.
     State bonds   more  independent, less costly and more
     competitive- would  be substituted for local bonds.
     A potential for public sale to small investors,  as  in
     the case of U.S.  Savings Bonds, would exist.
     The local government  should realize a lower cost of
     financing sewer and drainage improvements.

Although  control over local issues by outsiders  would not be eliminated,
control by the market and its representatives would be replaced by State
control.   State control is still public control, and therefore sensitive
to political problems and answerable to the voters.  The  involvement
of the State might also facilitate a more regional approach to water
pollution  control.
                               95

-------
One forseeable disadvantage to this system is that state taxpayers must
assume the burden of speculation by local governments.  However, if
the State bears an interest in the debt of the local municipalities, it
may insure that local government's expertise and education in financing
is thorough and effective.

Thus Abt Associates recommends  that a model statute to establish a
State bond bank be investigated further.  Research should include the
details of such a  statute, its impact on the market,  and identification
of those  states most likely to be aided by such a provision.  Since the
state of Vermont has enacted such a proposal, its history will be impor-
tant in the investigation.  Research should then be refined to determine
whether the scope should be limited to water pollution control financing.

If research is fruitful, then the model statute should be promulgated
among the states.

New Types of Financing.  It is not clear that the  financing of water
pollution control  facilities should be  limited, as it currently is, to either
the revenue bond or the  general obligation  bond.  As discussed above,
both of these bonds have substantial undesirable characteristics.  The
general obligation bonds frequently result in an increase in taxes and
therefore are more  likely to fail in bond elections.  Revenue bonds have
several hidden difficulties related  to the assumptions underlying the
user charge.  The problem, then is to combine the political desirability
of the revenue bond  with the marketability  and low administrative cost
of the general obligation bond.

To accomplish this,  more innovative techniques for financing public
expenditures for  water pollution control  should be explored.  This will
require stimulating  the financial community's creative resources
through a variety of means.  Private conferences with municipal bond
attorneys and investment banking experts should prove most fruitful,
                              96

-------
especially in view of the informality of the municipal bond underwriting
process and the lack of literature and even verbal communication of
information within the financial community on municipal financing.
Small seminars for  leaders in field with the  goal of formulating new
financing  options should also prove effective.

Disseminating Information on Financial Innovations.  It is also recom-
mended that information on innovative financing  schemes currently being
utilized throughout the nation be disseminated to local governments
through the states.  This could be accomplished most efficiently through
a periodical newsletter.  Information included would cover the general
topic of municipal bond financing of water pollution control facilities,
describing new and newly proposed projects  throughout  the nation and
highlighting the most innovative ones.   Such a newsletter could also
double as a means of disseminating engineering  information. It would
be sent to the appropriate agency in each state and would be made
available, in addition, to local government and to the financing community.
                               97

-------
Alternatives to the Referendum

Many local governments must obtain the approval of their electorates
before they can issue municipal bonds for public facilities,  in particular
those for pollution control.  For many municipalities, this requirement
is imposed by state statuatory or constitutional law (see Appendix A);
for others,  it is a requirement of their  local town or city charter.
However, since s'.tate law is always binding over local charter law,
changes  in the legal approval mechanism must ultimately come through
changes  in state law.  As pointed out in Chapter I,  most of the State
mandates for the use of referenda are statutory and not constitutional.
Thus  substantial changes can be made in the limitation on municipal
financing without constitutional amendments or revisions.

Beyond the legal requirements, there are often good reasons for holding
a referendum for a major  public expenditure.  The major reason is that
it provides citizens of the  community with  an opportunity to express
their  opinions directly  on a matter of public policy.  This can be useful
for local government officials wishing to obtain guidance  from the elec-
torate on their priorities for public expenditure.

Although some local governments have the power to avoid this approval
mechanism,  and would like to do so, they are unaware of the alterna-
tives.  Since the statuatory and constitutional requirements regarding
bond elections are quite complex, they can be easily misunderstood.
For example, they may apply to general obligation bonds repaid by
taxes, but not to revenue bonds.   Similarly,  referendum  requirements
rarely apply to special districts created for limited purposes  such as
sewage collection and treatment.  Since both of these financing methods
have significant disadvantages as  well as advantages,  it is important
that municipal decision makers be better informed about  them.
                               98

-------
In order to expedite local financing for pollution control facilities, Abt
Associates recommends  the following:
     1.  Local government officials should receive current
        information on all available means of developing local
        financial support for pollution control facilities.
        This should include discussion of the advantages and
        disadvantages of revenue and special district financing,
        and a methodology by which local officials can evaluate
        the opportunities and risks to their community of these
        approaches. This information should be included in
        the handbook for local officials described earlier in
        this chapter.

     2.  The Environmental Protection Agency should initiate
        formulation of a model State water pollution control
        act which would expedite both local financing and
        regional cooperation for water pollution control.
A Model State Act for Water Pollution Control.  A model act for •water

pollution control would have  several advantages:
     1.  It would expedite Federal/State and Federal/local
        effort through the efficiency of a uniform approach to
        pollution control.

     2.  It would expedite regional interstate and intrastate
        cooperation, within the context of uniform legislation
        on all the jurisdictions involved.

     3.  It would expedite local effort by simplifying the legal
        process, by opening avenues of communication to the
        State and Federal government and by providing oppor-
        tunities  for alternative approval mechanisms to the
        referendum.
To achieve these advantages,  a model state water pollution control act

should do the following:


     1.  Specify the public purpose and goals of the legislation.

     2.  Define the scope of the water pollution problem.
                                99

-------
     3.  Authorize local planning, action and funding.
     4.  Permit State and Federal contributions of resources.
     5.  Describe and permit alternative financing techniques,
        and alternative approval mechanisms.
     6.  Establish whatever State and Federal technical
        assistance  seems to be  desirable to further the
        overall goal of a  planned, cooperative effort to control
        water pollution.

Implementation of the act at the  State level might be facilitated by
choosing a number  of states which are especially responsive to Federal
technical assistance in this area.  Recommendations for new legisla-
tion could be formulated by the state water quality boards,  with the
assistance of the Environmental Protection Agency, and then proposed
to the state legislature. This experience would be studied with a view
toward moving to other states  with a similar approach.  A  snow-ball
effect could result as the  slower states follow the lead of the more
progressive ones in altering their sewer and sewage  treatment law.
                             100

-------
         APPENDIX A

 STATE CONSTITUTIONAL AND
    STATUTORY PROVISIONS
CONCERNING BOND REFERENDA
              101

-------
Alabama
Alaska
Arizona
Arkansas



California


Colorado
            STATE CONSTITUTIONAL PROVISIONS
               CONCERNING BOND REFERENDA
A.C.A.  1958 Constitution,  Art.  12,  s.222.  Specific
exemption of sewer bond issues from referendum require-
ments when the special assessment system  is used.

A. S.  Constitution, Art.  9,  s. 9.  Referendum required
for contracting of debts  for capital improvements.

A. R.S.  Constitution, Art.  9, s.8.  No referendum re-
quired to contract debts within a specified debt limit of
4% of assessed valuation.  A referendum may raise this
limit to 15%.

A.S.A.  1947 Constitution,  Art. 16,  s. 1.  Referendum
required for general obligation bonds, exempted for
revenue bonds. Bond percent set at 6.

A. C. C.  Constitution, Art.  11, s. 18. Referendum required
for all debts which exceed revenues  for that fiscal year.
C.R. S. 1963 Constitution, Art. XI, s.8.
required for the contracting of any debt.
Referendum
Connecticut  No constitutional provision

Delaware    No constitutional provision

Florida      F.S.A. Constitution, Art. DC, s. 6.  Referendum required
             for issuance of bonds.  Florida Supreme Court held
             revenue bonds are certificates, not bonds.

Georgia     C.G.A. Constitution,  s. 2-6001.  Referendum required for
             counties  and municipalities to contract debt.   Bond per cent
             set at 6.

Hawaii      No constitutional provision

Idaho        I.C. Constitution,  Art.  8, s.3.  Referendum required to
             contract debt.  Only revenue bonds are permitted.  Idaho
             Supreme Court has held revenue bonds are debt of issuing
             authority.

Illinois      No constitutional provision
                                103

-------
Indiana

Iowa

Kansas

Kentucky

Louisiana



Maine

Maryland

Massachusetts

Michigan

Minnesota

Mississippi

Missouri
Montana
No constitutional provision

No constitutional provision

No constitutional provision

No constitutional provision

L.R.S. Constitution,  Art.  14,  s. 14.  Counties, munici-
palities and districts  are required to hold referenda
before contracting debt.  Bond per cent set at 6.

No constitutional provision

No constitutional provision

No constitutional provision

No constitutional provision

No constitutional provision

No constitutional provision

V. A. M. S. Constitution, Art. 6, s.26(c) and (d).  Counties
municipalities,  and special districts are required to hold
referenda prior to issuance of general obligation and
revenue bonds.

R. C.M.  1947 Constitution, Art. 13,  s. 6.  In the question
of sewers, the debt limit may be extended by referendum;
otherwise, no constitutional provision.
Nebraska    No constitutional provision

Nevada

New Hampshire

New Jersey  No constitutional provision

New Mexico
No constitutional provision

No constitutional provision
N.M.S.  1953  Constitution, Art. IX, S. 10, 12.  Referendum
required for the contracting of debt.  New Mexico Supreme
Court held revenue bonds  and general obligation bonds
based on special assessments not debt.
New York    No constitutional provision
                               104

-------
North Carolina  G.S.N. C. Constitution, Art. VII, s. 7.  A referendum
                is required for all debts except those which are
                "necessary expenses. " In Reed v. Howerton Engineering
                Co.. 188 N. C. 39,  123. S.E. 479 (1924),  the North
                Carolina Supreme Court held sewers to be  a necessary
                expense.  They have also held  revenue bonds not debt.
North Dakota



Ohio

Oklahoma

Oregon

Perm s y 1 vania

Rhode  Island

South Carolina

South Dakota



Tennessee

Texas

Utah


Vermont

Virginia

Washington
N. D. C.C. Constitution, s. 183.  No referendum
required when not exceeding debt limit.  Referendum.
enables public utility to exceed debt limit.

No constitutional provision

No constitutional provision

No constitutional provision

No constitutional provision

No constitutional provision

No constitutional provision

S.D. C.L. 1967 Constitution,  Art. XIII, s.4.  Refer-
endum required to contract debt.  South Dakota
Supreme  Court held sewer assessment bonds not debt.

No constitutional provision

No constitutional provision

U. C.A.  1953 Constitution, Art. XIV, s. 3.  Referendum
required  to contract debt.

No constitutional provision

No constitutional provision

R.C.W. A. Constitution,  Art.  VIII, s.6.  (Amend. 27).
Referendum  is required to increase the debt limit.
Assessment  sewage bonds are not counted in debt
limit computations.  Smith v.  Seattle,  25Wn. 300,
65P.612 (1901).
                              105

-------
             27
West Virginia   W. V. C. Constitution, Art. X, s. 8.  Referendum
                is required for contracting of debt.  Insofar as this
                provision  requires  a  60% majority rather than a
                simple majority, it is unconstitutional.   Lance v. Board
                of Education, 	W. Va.  	,  170  S. E.
                2nd 783,  (1969).

Wisconsin      No constitutional provision

Wyoming        W.S. 1957 Constitution,  Art.  16, s.4.  Referendum
                required for contracting of debt.
                             106

-------
USE OF REFERENDA BY STATE
State
ALABAMA
ALASKA
ARIZONA
ARKANSAS
CALIFORNIA
COLORADO
CONNECTICUT
DELAWARE
FLORIDA
GEORGIA
HAWAII
IDAHO
ILLINOIS
INDIANA
IOWA
KANSAS
KENTUCKY
LOUISIANA
MAINE
MARYLAND
MASSACHUSETTS
MICHIGAN
MINNESOTA
MISSISSIPPI
MISSOURI
MONTANA
NEBRASKA
NEVADA
NEW HAMPSHIRE
NEW JERSEY
1970 1969
-
2
3
1
12 18
1 2
2 6
-
-
1
-
1
9 15
-
-
-
-
3
1
-
1
2 5
2
2
1
-
5 3
-
1
_ _
1968
2
1
-
1
19
3
3
-
2
2
-
1
29
-
-
-
-
6
1
1
1
15
2
3
6
1
1
-
1
_
1967
1
3
-
1
17
8
4
1
1
-
-
8
21
-
-
3
-
7
-
-
-
8
1
6
13
1
4
-
5
-
1966
-
1

-
30
2
5
-
4
-

1
12
-
-
3
-
11
-
-
-
13
2
1
2
2
1
-

-
Tots
3
7
3
3
96
16
20
1
7
3
-
11
86
-
-
6
-
27
2
1
2
43
7
12
22
4
14
-
7
_
         107

-------
     State
NEW MEXICO
NEW YORK
NORTH CAROLINA
NORTH DAKOTA
OHIO
OKLAHOMA
OREGON
PENNSYLVANIA
RHODE ISLAND
SOUTH CAROLINA
SOUTH DAKOTA
TENNESSEE
TEXAS
UTAH
VERMONT
VIRGINIA
WASHINGTON
WEST  VIRGINIA
WISCONSIN
WYOMING
L970
_
_
-
_
1
-
5
-
3
-
4
70
-


-
_
2

1969
1
_
3
-
10
18
7
1
2
-
3
127
1
3
6
2
_
2

1968
6
-
2
-
19
25
7
1
5
-
9
95
5
5
4
5
_
5

1967
2
-
3
1
9
17
11
1
4
-
11
101
1
3
9
2
1
4

1966
3
1
6
-
13
23
13
1
6
1
7
86
1
3
4
2
2
2

Total
12
1
15
1
52
83
43
4
20
1
34
469
8
14
23
11
3
16
1,212
                              108

-------
STATE STATUTORY PROVISIONS
Statu
Alabama
Alaska
Arizona
Arkansas
California

Colorado

Connecticut
Delaware
Florida
Georgia
Citation
C.A.R. 1958 Tit. 37 s. 308
A.S. s.29. 50.010-050
A.R.S. s. 35-41
A.S. A. 947 s. 10-4101
A.C.C. s.4600

C.R.S. 1963 s. 139-52-1

C.G.S.A. s. 7-245
D.C.A. Art. 9 s.2201
F.S.A. 8.167.01
F.S.A. s.170. 01
C.G.A. s.87-201
C.G.A. s. 87-801
Authorized Political
SubdivisionlEo
counties, municipalities,
special districts
counties, municipalities
special districts
counties, municipalities
special districts
municipalities and special
districts
municipalities
counties and special
districts of counties
municipalities
special districts
municipalities, special
districts
counties
municipalities
municipalities (special
assessment)
counties
municipalities
Referendum
Type of Bond Requirement
general obligation
r e venue
general obligation
revenue
general obligation,
revenue
revenut only
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
no
no
yes
yes
yes
yes
no
yes
yes
yes
yes
yes
no
yes
yes
no
no
no
no
yes
no
yes
no
5
Bond Percent
none specified
none specified
none specified
none specified
none specified
none specified
6% (Const. )
none specified
none specified
6%
6%
6%
6%
6%
6%
6%
6%
none specified
none specified
none specified
none specified
6% (Const. )
6% (Const. 1

-------
   State

Hawaii


Idaho


Illinois
Indiana
Iowa
Kansas
Kentucky
 Louisiana
 Maine
         Citation

H.R.S. s.67-5


See State-Constitutional
provisions '

I.R.S. 1969  s. 11-14-1
                                                     Authorized Political
                                                         Subdivision
                                                   counties
                                                   municipalities
                                                   special districts
   Type of Bond

general obligation
r eve nue
 Referendum
Requirement

       yes
       no
                                                                                      revenue
                                                                                      revenue
Bond Percent:

6% (Const,)
6% (Const. )

none specified
                                                                                                            (1) by order of
                                                                                                               state, no
                                                                                                            (2) no, unless 15%     none specified
                                                                                                               petition for one
                                                                                                            no, unless  15%       none specified
                                                                                                            petition for one
B.I.S. s. 48-2701



I.C. A. H.358,7

K.S, A.."s. 12-617
K.S.A. s. 12.3104
K.R.S. 107.010

K.R.S. 76.010
See State Constitutional
Provisions
M.R.S.A. s. 30-4251
M.R.S.A. s.30
municipalities

counties

counties, special districts
municipalities
municipalities
municipalities
municipalities, special
districts
counties
municipalities, special
districts, parishes
municipalities

general obligation
revenue
general obligation
revenue
revenue
revenue
general obligation
revenue
revenue

revenue
revenue
general obligation
revenue
general obligation
no
no
no
no
no
no
yes
yes
no

yes
yes
yes
yes
yes
none specified
none specified
none specified
none specified
none specified
5%
none specified
none specified
none specified

none specified
6%
6%
6%
6%

-------

State
Ma r yla nd


Massachusetts

Michigan





Minnesota



Mississippi







Missouri


A.
A.
A.
M.

M.



M.

M.

M.

M.





M.



C.
C.
C,
G.

C.



C.

S.

s.

C.





C.



M
M
,M
L

L



L


Citation
. Art. 43 s.409
. Art.43 s.428
. Ajt. 43 s.445
.A. Ch. 83

.A. s.46. 171



.A. s. 123. 241

A. s. 115. 01


A. s. 115. 41

A





A


. 1942 s.3374-116





. 1942 S.3598-01C

Authorized Political
Subdivision
municipalities
municipalities
special districts
municipalities, special
districts
counties



municipalities, special
districts
special districts

municipalities

municipalities





special districts


Type
general
revenue
revenue
general
revenue
general

revenue

general
revenue
general
revenue
general
revenue
general




r evenue
general
revenue

of Bond
obligation


obligation

obligation



obligation

obligation

obligation

obligation





obligation

Referendum
Requirement
yes
no
no
yea
no
no, uiiless %
petition for one
no,uunlcss %
petition for one
yes
no
yes
yes
no
no
yes-for new
plants; no-for
improvements
or for amounts
under $50, 000
"



Bond
5%
6%
5%
none
none
6%

6%

6%
6%
6%
6%
6%
6%
6%




6%
6%
6%

Percent



specified
specified


















See State-Constitutional
provisions
Montana

R.

C.

M

. 1947 s. 11-2216

municipalities, special
districts
general
obligation
municipal.! tie s
yes
yes
none
none
specified
specified

-------

State Citation
Nebraska R.S.N. 1943s. 18-501

Nevada N.R.S. s. 318. 170



N.R.S. s. 710. 400

N.R.S. s. 267. 250
N.R.S. s.271.265
New Hampshire N.H.R.S.A.







New Jersey N.J.S.A. S.40-14A-1

N. J.S.A. s.5a-12

New Mexico N. M. S. 1953 s. 14-40-9

N.M.S. 1953 s. 75-18-1

Authorized Political
Subdivision
munic ipalitie s

special districts



counties

municipalities
municipalities
counties

municipalities

districts

towns

counties, municipalities

special districts

municipalities, special
districts
special districts


Type of Bond
general obligation
revenue
general obligation
general obligation
assessment
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
revenue

general obligation
revenue
Referendum
Requirement
yes
no
yes
no

no
no
no
no
no
no
no
no
no
yes
yes
yes
yes
no
no
no
no
no

yes
yes

Bond Percent
5-1/2%
6%
6
-------
u>
                       State                   Citation

                    New York         McKinnoy's Local Finance
                                      Law s. 34.00
                                      McKinney's Town Law
                                      8.190C

                    North Carolina     C-.S.M.C.  s. 160-424. 1
                    North Dakota


                    Ohio
                    Oklahoma
                    Oregon
N.D.C.C. s.40-34-01


O.R.C.A. s.729


O.R.C.A. s.6117


O.R.C.A. s.6115,6119


O.S.A.1 s. 11-271


O.S.A. s. 19-871

_8.224.010
                    Pennsylvania     P.P.S.A. Tit. 53 8.6201
Authorized Political
Subdivision
municipalities
districts
municipalities
special districts
municipalities, special
districts
municipalities
counties
special districts
municipalities
counties, special districts
municipalities, counties,
special districts

municipalities, special
districts

Type of Bond
general obligation
revenue
general obligation
revenue
general obligation
revenue
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
general obligation
revenue
revenue
general obligation
revenue
general obligation
revenue
Referendum
Requirement
no
no
no
no
no
no
no
no
no
no
ro
no
no
no
no
no
no
yes
yes, unless state
requires the is-
suance of bonds
"
no
no

Bond Percent
none specified
no.ne specified
none specified
none specified
none specified
none specified
5%
none specified
none specified
none specified
none specified
none specified
none specified
none specified
none specified
none specified
none specified
none specified
6%
6%
none specified
none specified

-------
State
Rho'ln Island

South Carolina

South Dakota
Tennessee
Texas
Utah

Vermont
Virginia
G,
G.
C.

S.
T.
Citation
, L.R.I. 1956 s.45-14-l(m)
.L.R.I. '1956 s.46-21-l(d)
L.S.C. 1967 s. 59-618

D.C. L. 1967 s. 9-48-1
C.A. s. 6-1401
Vernon's Ann, Civ. St.
Art. 1112
U.

V.
C.
C.A. 1953 s. 17-6-1

S.A. Tit.24 s. 1751
V. 1950 s. 15. 1-1239
Authorized Political
Subdivision
special districts
municipalities
municipalities
special districts
municipalities, special
districts
municipalities
municipalities
counties, special districts
municipalities
municipalities, counties
counties, municipalities fc
special districts
Type
general
revenue
general
revenue
general
revenue
general
revenue
general
revenue
revenue
r evenue
general
revenue
general
revenue
general
revenue
general
revenue
Referendum
of Bond Requirement
obligation
obligation
obligation
obligation
obligation


obligation
obligation
obligation
obligation
no
no
no
no
yes
no
yes
no
yes
yes
no
yes
yes
yes
yes
yes
yes
yes
no
no
Bond
6%
6%
6%
none
none
6"'o
none
none
none
6%
none
none
none
7%
7%

6%
6%
Percent


specified
specified
specified
spec ified
specified

specified
specified
specified




-------
   State                   Citation



Washington        R.C.W.A. s. 35. 67. 030





                  R.C.W.A. s. 56.04,010
West Virginia      W.V.C. s. 8-18-1





                  W. V.C. s. 16-1Z-14





                  W.V.C. s. 16-13A-1




Wisconsin         W.S.A.  s.67.05







                  W.S.A.  s.59.9



Wyoming          W.S. 1957 s. 15. 1-410





                  W.S. 1957 s.35-162
Authorized Political
Subdivision
municipalities

counties, special districts




municipalities

municipalities, special
districts
counties, special districts
municipalities, special d
districts

counties
municipaliaies

counties, special districts



Type of Bond
general obligation
revenue
general obligation
revenue



general obligation
revenue
revenue

revenue
general obligation

revenue
general obligation
general obligation
revenue
general obligation

revenue
Referendum
Requirement
yes
no
yes
yes, except for ad-
ditional issues for
up to 20% of origi-
nal
yes
yes
no

no
no, unless peti-
tioned by 10%
"
no
yes
yes
no, unless peti-
tioned by 40%
ii

Bond
6%
6%
6%
6%



6%
6%
6%

5%
none

none
none
6To
6%
6%

6%

Percent












specified

specified
specified






-------
           APPENDIX B

A LOGICAL PROCESS MODEL OF THE
MUNICIPAL BOND ELECTION PROCESS
                117

-------
                            PRELIMINARY PLANNING
                    ELECTION PROCESS
118

-------
--
s
propo
TUT the
• tudy

1C
r



dolag&ta to
city engineer

-
undertake
study

-
formulate and
publish rcc-
ommendation*


 PROJECT DEVELOPMENT AND FINANCIAL STRATEGY
decide
.oil bo
to
ndl
/

/
pressure from
• late u
rater
quality board
/ 1
hold public
hearings

/

/
carry
effort
voters
\

1
t
\
\
out
to
dc

\

\
hold
election

1
1
1
1
plan another
election

*» /^.r"",-\
"XT'/
^^ FOLLOW-UP AND IMFLE
rcformulato
project
MENTA'
•ION
119

-------
         APPENDIX C
MORE TECHNICAL ASPECTS OF
     THE METHODOLOGY
              121

-------
The Population

The purpose of this appendix is to provide additional information on
some technical aspects of our study,  namely the statistical population,
the sampling procedure and the potentials  and limitations of the
statistical methods used for processing the data we collected.

As indicated in Chapter 3,  the set of  referenda  adopted as our basic
statistical "population" was the computer listing of all the municipal
bond referenda published in the Daily Bond Buyer and compiled by the
Investment Bankers Association of America.  This computer list
includes about 75% of all the referenda held between January 1966 and
June 1970 by cities,  counties, townships,  districts and statutory
authorities to approve projects for the construction of sanitary sewers,
storm and drainage projects and treatment plants.  These projects
are classified in the  IBA system as codes  21 and 28.  Projects desig-
nated code 22, water and sewer combined, were also  sampled, but
code 20, water, was excluded.

Some special consideration was given at the outset of  the study to the
inclusion of drainage projects in the statistical  population.  Although
they are not often associated with  pollution control, they are intended
to prevent flooding and consequent health hazards  arising from water
stagnation, flooded basements and overflowed septic tanks.  They
therefore play a role in environmental protection.  Furthermore,
classifying projects into categories according to their purpose often
requires difficult decisions of whether a project is predominantly
for drainage or sewer purposes.  After completing our interviews,  we
found that  several of the  projects  classified as  drainage projects
actually consisted also of sewer construction and  were associated with
a pollution problem.  For these reasons, it is useful that they have
been included.
                             123

-------
The Sampling Procedure

Our statistical population includes 1254 referenda, 1047 that passed
and 207 that failed.  Since we were interested in comparing the
attributes of the referenda that passed with those that failed,  an equal
number was sampled from each of these two categories.  The time
and budget available for the study allowed us to document 20 referenda
that passed and 20 that failed.   This represents  a sampling rate of
1/50 for those that passed and 1/10 for those that failed.

For each year, the referenda held are classified by lexicographic
order within each geographic subdivision (city, county, etc.).  This
means that within these geographic subdivisions the referenda are
classified in alphabetical order by state and within each state they are
classified in alphabetical order according to the name of the location
where  they were held.  Since this list has a definite structure,  that
structure had to be taken into account when the sample was selected
in order to avoid any geographical biases.

In order to do so, we selected the  procedure of systematic  sampling.
It consists  of  taking  every pth  sampling unit after a random start.
After ordering our list sheets  by year  starting from 1970 and ending
in 1966,  we considered first the 207 referenda that failed.  As the
sample size of this stratum is  20,  our sampling interval is
p = 207/20  = 10.  Then we drew a number at random from 1 to 10
out of an excerpt of Kendall and Babington Smith's table of random
numbers.  This number turned out to be 9.  The  ninth referendum of
our list was then used as the random start.  We then selected the
nineteenth,  twenty-ninth, thirty-ninth, etc.,  referendum  on our list
until exhaustion which is reached almost exactly  when the 20 referen-
da are drawn.
                              124

-------
Without changing the order of the list, we repeated the same procedure

for the 1047 referenda that passed.  This time the sampling interval

was p = 1047/ZO = 50, approximately,  and the random start was 43.


The properties of systematic sampling can be summarized as follows:


              A systematic sample selected with interval p
              after a random start yields a sample in which
              the population elements have equal probabil-
              ities  of selection.

              If any attribute is observed on the sample, the
              sample mean of this attribute is an unbiased
              estimate of the population mean.

              If the population units were thoroughly shuffled
              and mixed before they were ordered on the list,
              the systematic sample would be equivalent to a
              simple random sample.

              If the nearby elements of the list are similar
              with  respect to the attribute under study,  the
              systematic sample is then more efficient than
              the simple random sample.  That is,  the variance
              is lower  than random selection formulas would
              indicate.  The description given above of our
              population list features  geographical clustering,
              size  clustering and chronological clustering.
              Moreover,  these clusters do not present a
              systematic periodical character as the number
              of referenda in each state or city is not the  same
              every year.   This provides enough evidence
              that the systematic sampling procedure that we
              adopted is more efficient than the simple random
              procedure and is therefore more appropriate.

              Systematic sampling allows a simple replacement
              procedure if some of the cities selected are
              unwilling to participate.  There is no need to
              replace the sample by another  one; it is sufficient
              to consider the initial list as circular and to add
              new units to the list  by adding new intervals p
              to. Hie last unit selected in  the previous  sample.
              However, this procedure should be considered
              as a last resource as it does not yield a strictly
              equal probability sample.  We  did not have to
              replace any units initially selected except for one.
                            125

-------
 The frequency distributions derived from our final set of data show
 regularities with respect to the population as a whole which confirm
 the random character of our sample and its reasonably accurate
 representation of the IBA list.  For instance, the distribution of the
 referenda in our sample according  to the year when they were held is
 similar to that of  the IBA list.  There are more referenda  in our
 sample from the West, Southwest, and Midwest than from the East,
 but this also reflects the regional distribution of referenda in the
 population.  The sample is  also fairly representative with respect to
 the type of governmental unit issuing  the bonds.  The only direction
 in which the sample appears to be biased is in its  over-representation
 of drainage projects, (code 28).  However,  since several of these
projects also include sanitary  sewer construction  or  are associated
with a pollution problem in  other ways,  this does not appear to be a
serious deficiency.
                           126

-------
Statistical Inference

Two types of questions arise in connection with the possibilities of
making inferences from our sample data to the population as a whole.
The first problem is the familiar one of drawing inferences about how
individual voters will behave inanelection on the basis of our data.
Since no survey research was  conducted on individual voter attitudes,
it is not possible to predict how different types of voters will behave.

The second question is related to the size of the sample  relative to
the size of the population  of interest, (all referenda for municipal
bonds for water pollution  control projects).  It is often assumed that
a larger population requires a  larger sample in order to give results
that have the accuracy of  a smaller sample drawn from a smaller
population. The fallacy of this assumption can be explained both
mathematically and by an appeal to intuition.   The complete equation
for specifying  sample size includes the factor ( 1 -  •=RJ), known as
the finite population correction, where N is the size of the population,
and ~ is called the sampling fraction.  When the sampling fraction
is .05 or less, the finite population correction is so close to 1 that
it has minimal effect on the size of the sample.  Therefore it is
generally ignored by statisticians when calculating  sample size for
large populations.

The intuitive argument considers two populations, one of 1,000 persons,
and another of 1,000,000  persons.   If you take a sample  of one  person
from the first population and a sample of 1, 000 from the second
population, the sampling fraction is the  same  in both cases.  Estimates
of a distribution of characteristics  based on the second sample,  how-
ever, would clearly be more reliable than the first.  The principle
behind accepting a sample of 1,000 from 1,000,000 population  is the
same principle that allows the  assertion that  if a  sample of 200 has
been taken, and the sampling fraction less than 5%, then it does not
matter  significantly what  the size of the population  is.
                             127

-------
A homey analogy may help to illustrate this phenomena.  Assume that
you wish to compare a bowl of one kind of soup with a bathtub of
another kind  of soup.  Although there is much more soup in the bath-
tub than in the bowl, it isn't necessary to sample more than a  spoon-
ful to know how the re st will taste,  regardless of whether it comes
from a bowl or a  bathtub,  assuming the soup is homogeneous.
                           128

-------
The Statistical Techniques Used and the Level of Measurement of
the Data
The data collected to characterize each referendum have been arranged
along a variety of scales.  These scales can be classified into types,
each of which is associated with a given level of measurement.  In
this section each characteristic is classified in its  scale category and
the different statistical models appropriate to  each category are dis-
cussed in connection with their potential ability to provide information
on the factors affecting the  outcome of the referenda.

Levels of Measurement of the Data:  Type of Scales.  The customary
categorization of scales represents a progression from low to high
levels of measurement. The lowest level of measurement corresponds
to a classification according to certain characteristics and the assign-
ment of labels to the various  categories.  Although numerical values
may be assigned to these categories,  they are equivalent to the labels.
The type of scale that corresponds to this level of measurement is a
nominal scale.  For instance, variable 001 representing the region
where the referendum was held has been arranged along the following
scale:

                       1 = Northeast
                       2 = South
                       3 = Midwest
                       4 = Southwest
                       5 = West

When it is possible to order categories with respect to the degree to
which they possess a certain characteristic, we can derive an ordinal
scale. Such a scale provides only a rank order and gives no infor-
mation about the magnitude of the intervals separating the ranked
objects.   For example, the role  played by a newspaper  during the
campaign towards the passage of the issue can be arranged along  the
                             129

-------
the following ordinal scale:

                      1 = Important against
                      2 = Low against
                      3 = Low in favor
                      4 = Important in favor

Most of the dichotomous data collected in our survey (fail, pass;
absence, presence) can be considered on ordinal scale.

When it is  possible not only to rank the referenda with respect to the
degree to which they possess a certain characteristics, but also to
indicate the distances  between them, we can derive an interval scale.
This type of scale implies that numerically equal distances on the
scale represent equal  distances in the property being measured. In
other words, it implies the existence of a unit of measurement.  In
this  type of scale, it is meaningful to add and subtract intervals as
well as to take ratios of differences between scale values.  One
cannot,  however,  state that an object with a scale value of 6  is twice
as good as  an object with a scale value of 3 because  the scale has no
natural zero.  Ratio scales have both a constant unit of measurement
and a natural zero and therefore permit all arithmetic operations.
The  typical example of interval scales is that of the  temperature
(fahrenheit or centigrade)  scales.   Besides the year of the issue and
some others,  most of  our  referendum characteristics are  not strictly
measured according to an interval scale; indeed, in  most cases when
a unit of measurement is  available, a natural zero is also  often
available and the corresponding scale is ratio  scale.  For  example,
the population can be arranged according to the following ratio scale,
by adopting 500 persons as a unit:
                            130

-------
                                 1 = 0-499
                                 2 = 500-999
                                10 = 1,000-4,999
                                20 = 5,000-9, 999
                                etc 	

           For the purpose of using some statistical techniques, it may be useful
           to transform an interval or ratio scale into an ordinal scale.  Such
           is the case for cross-tabulations where categories can be regrouped
           and labeled in a different way than according to their original interval
           scale.

           The following table presents a classification of each of the variables
           according to the strongest type of scale they can be associated with,
           given its  nature and level of measurement.  Descriptors of the
           variables (region = character  1; amount of issue = character 2; etc.)
           can be found in  Appendix D in the Code Book.
       Nominal                 Ordinal            Interval         Ratio
1,3,6,7,8,12,13          2,5,9,10,11,20          4,22,27  14,15,21,31,32
16,17,18,19,23           25,33,34,35,36,40       28       76,83,84,86,87,88,89
24, 26, 29, 30, 37,38,39     41, 42, 43, 44,45,46,47             90,01,97,98,99,100,101
63,64,68                  48,49,50,51,52,53,54             102,103,104,105,106
77,78, 79,80,81,82        55, 56,57,58,59,60,61
92,93,96,107,108         62,65,66,67,69,70,71,72
110,111,112,113,114,115  73,74,75,94,95
                          109
           Statistical Techniques Appropriate to Each Type of Characteristic.
           All statistical techniques can be used with interval and ratio charac-
           teristics, from cross-tabulation (with the series of statistical tests
                                       131

-------
 usually conducted in order to examine the strength and significance
 of the relations among variables) to single and partial correlation and
 to multiple,  partial and stepwise regression models.

 For the nominal and ordinal variable - that is,  the non-quantitative
 ones - all the technique s cited above are not applicable due to their
 restrictive assumptions.  The only traditional statistical tools that
 are applicable are cross-tabulations (with the usual  tests) and partial
 correlations using Kendall's tau (that is, a systematic use of one of
 these  tests).

 The reason for  this is  that the more powerful statistical tools use the
 standard  correlation coefficient,  also called 'product moment"
 correlation coefficient, as a measure of the relationship between two
 characteristics.  This coefficient makes use of the scale value of the
 different  characteristics.   When these scale values are arbitrary
 numbers  that merely represent labels,  such a s :

                      1 =  Northeast
                      2 =  South
                      etc   	

 the value  of the correlation coefficient will vary with the label numbers
 and therefore is meaningless.

 However,  there is a particular case for  which the  value of the cor-
 relation coefficient will not depend on the label numbers and that is
 the case where  the two characteristics are scored on scales having
 only two values.  These scales are called dichotomous scales.  When
 one of the characteristics  is scored on a continuous scale and the
 other on a dichotomous scale, the resulting correlation is called a
 "biserial" correlation.  When both characteristics are scored dicho-
tomously,  we have a "fourfold" correlation. In these two  cases,  the
 application of the product-moment correlation formula with any
                             132

-------
arbitrary values assigned to the two categories of the dichotomous
scale - or scales - will provide the same result and the corresponding
coefficient will therefore be meaningful.

The  two coefficients obtained, respectively called the "point-biserial
coefficient" and the "Phi coefficient" are simply particular cases of
the standard product-moment correlation coefficient. For the purpose
of their calculation, any two numbers may be assigned to  the two
categories, but the computation is minimized by using 0 and 1.

The  overwhelming majority of the non-quantiative variables that we
have collected is actually of dichotomous nature, (pass-fail, presence-
absence,  etc  ).   These  can therefore be analyzed by means of the
same sophisticated techniques as the quantitative variables.  The
remaining variables that are not naturally dichotomous can be redefined
so that the possible values are  regrouped into two categories,  or so
that  each type of answer is associated with a dichotomous  variable
representing  the presence or absence of this particular type of answer.

Introduction of New Variables
The first round of statistical analyses shed light on some of the major
variables in the outcome of the referendum.  However,  in order to
understand  certain relationships better, some new variables were
derived and others were cancelled.  These new variables are based on
manipulations of the initial variables.
        VARIABLE 116   This variable is the indebtedness of the
                         city per capita. It consists of variable
                         89, total indebtedness, controlled for
                         population,  variable 14.
        VARIABLE 117   This is an index of coincidence of those
                         who voted with those who will be served
                         by the project and those who will pay for
                         it.  It is a numerical variable  built by
                         using variable  10, coincidence of
                            133

-------
VARIABLE 118
VARIABLE  119
VARIABLE  120
VARIABLE  121
VARIABLE 122
VARIABLE 123
population served and paying,  and
variable 11,  coincidence of population
served and voting.

This variable represents the level of
campaign activity.  It is  the sum of all
the variables from 52 to  62 which
indicate  the use or non-use of particu-
lar types of media for advertising.
As there is no  reason to  distinguish
among the  variables, they have not been
assigned different weights.  Hence,
variable 118  is simply the sum of these
variables.

This variable represents the amount of
criticism raised against  the issue.  It
is the sum of all variables from 63 to
75 inclusive.  As in the construction of
Variable 118, each of these dichotomous
variables representing types of criticism,
has been receded so that the numbers
zero and one represent the absence and
the presence of a given criticism type.
The means of deriving the variable are
the same as for 118 also hold here.

This variable indicates whether or not
the water pollution bond issue  combined
with any other  issues on  the ballot.

This variable indicates whether issues
on the  same ballot as the water pollu-
tion bond issue, but not combined with
it.

This variable indicates whether the
water pollution bond has been split into
two or more  parts to be voted  on  separately.

This variable represents the number of
people, groups and organizations that
took positions for or against the project.
It is defined as the sum of variables 40
to variable 51 inclusive.  As in variables
118 and 119,  this variable seems to be
more  meaningful than the individual
variables that comprise it.
                      134

-------
       VARIABLE 124  This variable combines variables 110 and
                        111 in the same way as variables 10 and
                        11, and represents a scale on which the
                        topographical characteristics of  the
                        project are  recorded.

       VARIABLE 125  This variable indicates the existence of
                        local or Federal regulations in relation
                        to the project. It operates on variables
                        114 and 115, using  the same transforma-
                        tion as  the one for variable  117.
Besides, several variables have been receded so that they present

only two categories (presence or absence of a given characteristic)

These are:


       VARIABLE  12  Partisanship of the government election.

       VARIABLES 38  Role of the newspaper and role of other
         and 39         media.

       VARIABLES 33, Existence and partisanship of groups for
         35,  and 37      and against.


We have eliminated variables such as variable 116 (use of water front)

and variable  96 (rating system: Moody or Standard and Poor)  as

they showed no explanatory power.


After counting the variables added, those removed, those aggregated,

those replaced and those kept,  we arrive at a set of 99 variables

amendable  to  standard correlation analysis and 5 or 6 variables purely

nominal such as the region, the month,  where the referenda were

held, etc. , which  are analyzed through cross-correlation only.
                            135

-------
 Tests of Significance and Measures of Association

 As indicated in Chapter 3,  the relationship between each pair of
 variables was  statistically analyzed by various tests of significance
 to determine the probability that the observed distribution could have
 occurred by chance.  The strength of the relationship was also tested
 by means of another set of tests.   This section includes a short
 description of  each of the tests to which our data were subjected.

 Chi-square. The chi-square statistic allows us to test the independence
 between two variables.  Chi-square only indicates the likelihood of
 obtaining a  given distribution of empirically observed values on the
 basis of chance alone; it does not measure the degree of association.
 It is calculated by first squaring the differences between the observed
 and expected values in each cell,  dividing each by the expected value
 for that cell, and summing over all cells, as follows:
                         V~   o "
                       -  L —r-
                                e

-------
been computed automatically and the results (not significant or sig-
nificant and the level of significance) have been printed.

Yates  Chi-square^  This test computes a chi-square which is corrected
for continuity.  Since the chi-square distribution tables are based on
a large sample,  the  sampling distribution in the test statistic will
approximate the sampling distribution in the chi-square table only
when it is large.  When the expected frequencies are as low as 5, the
observed values are  corrected by either adding or subtracting 5 from
each to bring them closer to the expected values.  The value for
chi-square is then calculated as above.

Fisher's Exact Test. Fisher's exact test is used with 2x2 contin-
gency  tables to yield the exact  (rather than approximate) probability
that the observed distribution differs from a random distribution.
Briefly,  the  test requires the computation of the exact conditional
probability of the observed table,  and of all more extreme "hypothetical"
tables  with the same marginal  totals.

Since the Fisher test is an exact test, it is generally preferable to
the chi-square test corrected for continuity.  The chi-square  test
ordinarily results in somewhat lower probabilities than the Fisher
test, and therefore one  risks rejecting  the null hypothesis when it
should be accepted.  Fisher's exact test is more conservative and
avoids this problem.  However, if the smallest expected value is
greater than 5,  and the  continuity  correction is used when N is small,
then the two  tests will yield nearly the same results.

Phi Coefficient,  C Coefficient and Cramer's V.  The three statistics
described above  are  concerned with whether or not a relationship
exists, and the level of  significance and indicates how much certainty
one can have in that existence.  However, none of them says  anything
about the strength or importance of  the relationship.
                             137

-------
One of the most commonly used measures of the strength of an
association is  the phi coefficient which is defined as:
                                'x2
                       phi (0) =^/ ^p
It essentially adjusts chi-square for the total N (sample  size) since the
value for chi-square is directly proportional to the number of obser-
vations.  In a 2 x 2  table, phi will vary from 0 to 1.

When a phi is calculated for a contingency table which is not 2x2,
it has no upper  limit.  Another measure of association,  introduced by
Cramer, corrects for this condition.  It is defined as follows:
                               ^
                   v =.
                         NMin(r-l, c-
where Min(r-l,c-l) refers to the minimum of the two values: rows
minus 1 or columns minus 1.

Cramer's V will vary between 0 and 1 for all tables, and will be
identical to phi in 2 x 2 tables.

Pearson's  contingency coefficient (C) is another indicator, based on the
chi-square, of the strength of a relationship between two variables.
Its formula is as follows:
                              138

-------
In a 2 x 2 table, the upper limit for C is . 707,  although this value
increases as the number of rows  and columns increase.  For this
reason, C should not be used without a correction when comparing
tables  of different sizes.

Measures of Association Based on Probabilistic Approaches.  Several
other tests of association have been developed which avoid some of
the problems inherent in chi-square. Since they involve  various
probabilistic interpretations and vary from 0 to 1,  they have an intuitive
meaning which is easy to interpret.  For this reason, they are often
preferable to the measures of association based on chi-square.  Included
in this group of statistics are the following:  lambda, tau  B,  tau C,  gamma,
and Sommer's D.

The lambda asymmetric and symmetric tests were suggested by
Goodman and Kruskal.  The asymmetric test is defined as follows:
             Lambda = \,  =
                        b
                               max.f..  - max.f. .
                                     Jk         k
                                  N - max.f.,.
                                             k
where 2_  max.f.,   represents the sum of maximum values of the cell
        X.       J
frequencies in each column and max.f.,. represents the maximum value
of the row of totals.
 It can be interpreted as the relative decrease in the probability of error
 in guessing in which row a unit belongs when its column category is
 known,  over the case when there is no information about the column
 category.  When applied to the bond referendum data, it is the relative
 decrease in the probability of incorrectly guessing the outcome of the
 election when it is known, for example,  how the bond is to be repaid,
 compared to the situation in which this information is not available.   The
 lambda asymmetric test assumes that the rows define the dependent
                               139

-------
 variables.  The lambda symmetric avoids this assumption,  but has the
 same interpretation.  It is defined as follows:
     Lambda
                \~ max. f. , + )  max. f., - max. f.  . - max. f. .
              =  /-  k     k  Z-	jk	k	j_
                           2N ~, max.f.. . - max.f..
                                       k          J
where / max. f   and max.f. . are defined for lambda asymmetric,
            k  jk            k             Y^
max. f:. is the maximum column total, and  )   max.f..   is the sum of
      J                                     «—   3   jk
the maximum values of the cell frequencies in each row.

Kendall's Tau B and Tau C can be used to correlate two ordinal scales.
Tau B is used for tables with an equal number of rows and columns, and
tau C for tables with different numbers of rows and columns.  As long
as both variables can be ranked, these statistics will measure a. degree
of association which varies from -1 to +1.   This measure cannot be
interpreted in terms of the percentage of variation which it explains,
since the notion of variation has no meaning with ordinal scales.  Tau
is essentially a measure  of the degree to which pairs in the same order
exceed the number of pairs in the reverse order.

These two  statistics are calculated by determining  the number of possible
pairs in the table (1/2)N(N-1) and then dividing them into three groups:

        P = all pairs in which the order on one variable is  the same
             as the order on the other	concordant pairs
        Q = all pairs in which the order on one variable is  the opposite
             of the order on the other	discordant pairs
        T - all pairs in which at least one variable shows  a tie
        then
        Tau B =   P - Q
                             140

-------
In words, tau B is the number of pairs in which the order on one variable
is the same as the order on the second, minus the number of pairs for
which the order is different, adjusted for the total number of possible
pairs.

Tau B  can be computed with adjustment for ties by the use of the formula

                         P        -        Q
         Tau B =
                                              - T2
where Tj equals the number of ties in the first variable,  and T2 equals
the number of ties in the second variable.
Tau Chas the formula
                         2m(P - Q)
                         N2 (m - 1)
where m represents either the number of rows or the number of columns
in the  table,  whichever is smaller.

The Gamma  Statistic is another ordinal symmetric measure. Its
formula is

                         Gamma  = P " ®
                                   P + Q

where P and Q have the same meaning as they do for Kendall's Tau.

The Gamma measure has a clear interpretation even in the case of
ties: namely,  the probability of "like" or "same order" minus the
probability  of "unlike"  or "different order, " given two units drawn at
random without ties on either variable.
                              141

-------
The Sommer's D Statistic is similar to Gamma,  but considers ties in
the rank as valid information and can be applied  to bgth symmetric and
asymmetric cases.

In the asymmetric case,  the asymmetry assumption is that the columns
of the table represent the categories of the dependent variable while
the rows represent the categories of the dependent variable.   The
statistic could be interpreted as the probability of "like order" minus
the probability of "unlike order" (given no ties on the column variable),
when two units are drawn at random from the table.
                             142

-------
Additional Analyses Performed on the Data;  Factor and Cluster Analysis

In addition to the analysis discussed in the main body of the report, the
99 variables amenable to standard correlation analysis have also been
processed through factor analysis and cluster analysis in order to
derive possible patterns among the set of referenda.   More specifically
we attempted to classify the referenda into groups such that the referenda
within a group  were more similar to other referenda in their  group than
to referenda belonging to different groups.  If in addition,  the distri-
bution of successes and failures differed significantly from a  group to
another, that is, if most of the referenda which failed appeared in a group
and most of those which passed appear in another  — then, the analysis
of the dissimilarities  between each group could be interpreted in order
to define global patterns of referenda associated with passage or failure.

This  type of analysis has essentially the same goal as regression
analysis.  Indeed,  the regression equation obtained in Chapter 4 is  used
to calculate the mix of major variables associated with a given outcome.
However, the set of possible mixes is very large as each variable can
take on several values and consequently the number of combinations of
these values is very high.  If significant clusters are available, the
number of acceptable  values of a given variable is subdivided, each
subdivision being associated with one of the groups.  This  could simplify
the analysis  of the consequences of combinations of changes that could be
made in some  of the variables.  For example, one could see  the effects
of changes in the level of campaign activity.

Two separate techniques and computer programs have been used to
cluster the referenda  in our sample.   The first is the natural grouping
program which forms  part of the Multivariate Statistical Analyses
designed by Kenneth Jones of the Harvard School of Education.  The
second is a cluster program designed by Fred Bookstein at the Harvard
Center for Behavioral Science.  Both programs  use factor analysis as
                               143

-------
 an intermediate step to orthogonalize the input dimensions.

 The results of the groupings given by these two programs are very
 similar and provide groups which unfortunately are not discriminated by
 the pass-fail criterion.  That is,  the number of cases of passage (or
 failure) in each group is about the same, and the proportion of each is
very close to  50%.

 This result suggested the relevance of more powerful clustering tech-
niques using the full information content of our qualitative  as well as
quantitative variables, in particular those related to the political
process and the campaign activity.  Such techniques  have recently been
programmed for computers and are part of the O.S.I. R. I.  S. package
developed at the University of Michigan  and available through the Harvard
 Computing Center.   However,  the application of these programs to our
data was beyond the time and budget limits of the present study.
                              144

-------
        APPENDIX D

FILE OF STATISTICAL DATA:
  CODE BOOK,   DATA FILE
   AND INTERVIEW GUIDES
           145

-------
CODE BOOK
147

-------
The present codebook is prepared to describe the coded data and to
indicate the IBM card columns in which the data are punched.

The first four fields (columns  1 to 10 included) are the identification
fields.  In the DATA-TEXT System which we intend to use, they are not
given variable numbers.

Missing characteristics are indicated by leaving  the relevant columns
blank for the corresponding referendum.

Missing data are differentiated from the statement "not applicable" by
adopting uniformly the code 0 for the latter.

As more than one card is  required to record all the data on each refer-
endum, the card number is part of the identification field and is punched
in column  10.
                               149

-------
LIST NO.  1
Location
Columbus, Tex.
Harris Co. W. Dist. 73, Tex.
Rockport, Tex;.
Athol, Mass.
Batesland, S.D.
Harris Co. W. Dist. 90, Tex.
Palo Alto, Calif.
Warren, Mich.
Bend, Oreg.
Kyle, Tex.
Jackson, Wyo.
Mill Valley, Calif.
Rolla, Mo.
Alvin, Tex.
Royse City, Tex.
Jefferson for CDD, La.
Petaluma, Calif.
Brevard, Sewer, N.C.
Belton, Drainage, Tex.
VanAlstyne, Tex.
DeKalb Co. , Ga.
Mankato, Minn.
Portland, Tex.
Simi Val Co. Sant D. Calif.
Douglas, Ark.
Roslelle, 111.
Oakwood, Oh.
Oklahoma City, Okla.
Round Rock, Tex.
Brookshire Mun W.D. , Tex.
Donaldsonville, La.
Alpha ID
Code
C0LMB
HAR73
R0CPT
ATH0L
BATES
HAR90
PALAL
WAREN
BEND0
KYLET
JACKS
MILLV
R0LLA
ALVIN
R0YSE
JFRSN
PETAL
BRVRD
BELTN
VANAL
DKALB
MNKT0
PRTLD
SIMIV
DGLAS
R0SEL
0AKWD
0KLAC
RNDRK
BR00K
D0NLD
Number ID
Code
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
State
Code
46
46
46
23
44
46
05
24
39
46
54
05
27
46
46
20
05
35
46
46
12
25
46
05
02
24
37
38
46
46
20
  151

-------
Location
Great Falls, Mont.
Cache W/S, Okla.
Groves, Tex.
Hotchkiss SD, Colo.
Morgan City Drain, La.
Avon Sewer, Oh.
Moore Sant. , Okla.
Cuero W/S, Tex.
Los Angeles C. Sew, Calif.
Alpha ID
Code
GT FAL
CACHE
GR0VS
HTCHK
MORGN
AV0N0
M00RE
CUER0
LACTY
Number ID
Code
32
33
34
35
36
37
38
39
40
State
Code
28
38
46
07
20
37
38
46
05
152

-------
CARD  1

Identification

   Col. 1-5         Alphanumeric:  Location of referenda (see
                    attached list no. 1)
   Col. 6-7         Numerical identification of the referenda
                    (see attached list no.  1)
   Col. 8-9         Identification of state  where the referenda
                    were held (see attached list no. 1)
   Col. 10          CARD NUMBER: 1

Data from ISA List                                           VARIABLE
                                                              NUMBERS
   Col. 11          Region (see attached map)                     1
                    1 =NE
                    2 = S
                    3 = MW
                    4 = SW
                    5 = W
   Col. 12          Amount of Issue                              2
                    1 = Less than $99,000
                    2 = $100,000 - 499,000
                    3 = $500,000 - 999,999
                    4 = $1 million - 4,999, 999
                    5 = $5 millions and above
   Col. 13-14       	
                    01 = JAN
                                etc. ...
                             153

-------
   Col.  15          Year
                    6 = 1966
                    7 = 1967
                    8 = 1968
                    9 - 1969
                    0 = 1970
   Col.  16          Outcome
                    1 = Pass
                    2 = Fail
   Col.  17-18       IBA Use of Proceeds Code
                    21 = S ewers
                    22 = Water and. Sewers
                    28 = Drainage  Sewers
Data from City Clerk
Data Sheet
   Col.  19          Largest Governmental Unit
                    1 = City
                    2 = County
                    3 = District
                    4 = Township
                    5 = Statutory Authority
                    6 = Sewer  District
                    7 = Others
   Col.  20          Type of Government
                    1 = Strong Mayor-Council
                    2 = Weak Mayor-Council
                    3 = Council-Manager
                    4 = Commission
                    5 = Other
   Col.  21          Existence of Independent Governmental Units
                    1 = Yes
                    2 = No
                             154

-------
Col. 22          Is population to be served the same as          10
                 population who pays?
                 1 = Yes
                 2 = No
Col. 23          Is it the same as the population who voted?     11
                 1 = Yes
                 2 = No
Code 24          Type of Election of the Local Government      12
                 1 = partisan
                 2 = non-partisan
Col. 25          Type of Election of the Local Government      13
                 1 = Ward
                 2 = At-large
                 3 = Ward and At-large
Col. 26           Population                                  14
                 1 =       0 -    499
                 2 =     500 -    999
                 3 =  1,000 -  4,999
                 4 =  5,000 -  9,999
                 5 = 10,000 - 24,999
                 6 = 25,000 - 49,999
                 7 = 50,000 - 99,999
                 8 = 100, 000-and over
Col. 27          Rate of Population Growth (1970/1960)          15
                 1 = Decline
                 2 = Stationary
                 3 = 0-49%
                 4 = 50-99%
                 5 = 100-499%
                 6 = 500% and more
                          155

-------
Col. 28          Use of the Waterfront                         16
                 0 = Not applicable
                 1 = Commercial
                 2 = Residential
                 3 = Industrial
                 4 = Recreational
                 5 = Future Development
                 6 = Combination of the Above
Col. 29          Use of the Proceeds (Detailed)                 17
                 1 = New
                 2 = Extension or Addition
                 3 = Remodeling or Alteration
                 4 = New,  Extension and Remodeling
                 5 = New and Extension
                 6 = New and Remodeling
                 7 = Extension and Remodeling
                 8 = Acquisition
                 9 = Acquisition and Any of the above
Col. 30-31       Description of the Project                     18
                 01 = Treatment
                 02 = Collection, Sanitary
                 03 = Disposal, Sanitary
                 04 = Drainage
                 05 = 01 and 02
                 06 = 01 and 03
                 07 = 01 and 04
                 08 = 02 and 03
                 09 = 02 and 06
                 10 = 03 and 04
                 11=01 and 02 and 03
                 12 = 01 and 03 and 04
                 13 = 02 and 03 and 04
                 14 = 01 and 02 and 03 and 04
                 15 = 01 and 02 and 04
                         156

-------
Col. 32-33       Analysis of the Wording of the Referendum     19
                 01 = Positive Wording
                 02 = Existence of description of the improve-
                     ment more detailed than in Col. 30-31.
                 03 = Indication of the location of the improve-
                     ment.
                 04 = Explanation of the means  of financing.
                 05 = 01 and 02
                 06 = 01 and 03
                 07 = 01 and 04
                 08 = 02 and 03
                 09 = 02 and 04
                 10 = 03 and 04
                 11 = 01 and 02 and 03
                 12 = 01 and" 02 and 04
                 13 = 02 and 03 and 04
                 14 = 01 and 02 and 03 and 04
                 15 = None 01,  02,  03, 04
Col. 34          Way the Issue was Offered to the Voters       20
                 1 = Single Issue (Yes to Question 4)
                 2 = Combined Issue (Yes to Question 5a)
                 3 = Split Issue (Yes to Question 5b)
Col. 35          Outcome of the Vote                          21
                 0 = 0 -  9%
                 1 = 10 - 19%
                 2 = 20 - 29%
                 3 = 30 - 39%
                 4 = 40 - 49%
                 5 = 50 - 59%
                 6 = 60 - 69%
                 7 = 70 - 79%
                 8 = 80 - 89%
                 9 = 90 - 100%
                          157

-------
Col. 36          Percentage Required to Pass                   22
                 1 = 50%
                 2 = 67%
Col. 37          .(If combined issue) Other Issues  Presented     23
                 In the Same Package
                 1 = Streets
                 2 = Water
                 3 = Other Sewer-drainage
                 4 = Fire
                 5 = Law Enforcement
                 6 = Hospitals,  School, Other Buildings
                 7 = Parks
                 8 = More than one of the above
Col. 38          (If split issue) Outcome for  the Different        24
                 Issues
                 1 = All passed
                 2 = All failed
                 3 = Part  passed and government intends to
                     use the proceeds
                 4 = Part  passed and government does not
                     intend to use the proceeds
                 5 = Part  passed and government uncertain
                     to use the proceeds
Col. 39          Existence of Candidates for  Public  Office       25
                 Elected in the  Same Election
                 1 = No
                 2 = Yes,  Local Election (primary,  general,
                     special).
                 3 = Yes,  State Election (primary, general,
                     special).
                 4 = Yes,  National Election (primary,
                     general, special).
                 5 = Yes,  others.
                         158

-------
Col. 40          Other Issues Presented to Vote the Same      26
                 Day (but not combined with the project at hand)
                 1  = Streets
                 2  = Water
                 3  = Other Sewer/Drainage
                 4  = Fire
                 5  = Law Enforcement
                 6  = Hospital, School Other Building
                 7  = Parks
                 8  = More than One of the Above
                 9  = None of the above
Col. 41          Of the total number of issues on the ballot,     27
                 what percentage received a significantly
                 higher percentage of favorables votes (5% or
                 more) than the issue at hand did?
                 0=0-   9%
                 1  = 10 - 19%
                 2  = 20 - 29%
                 3  = 30 - 39%
                 4  = 40 - 49%
                 5  = 50 - 59%
                 6  = 60 - 69%
                 7  = 70 - 79%
                 8  = 80 - 89%
                 9  _ 90 - 100%
Col. 42          Same Question as Col.  41,  but replace         28
                 "higher" by "lower"
                 Same code:
                 0=0-   9%
                 1  = 10 - 19%
                 2  = 20 - 29%
                 3  = 30 - 39%
                 4  = 40 - 49%
                 5  = 50 - 59%
                 6  =60 - 69%
                          159

-------
Col. 42          Same Question as Col. 421, but replace
(cont. )           "higher" by 'lower"
                 7 = 70 - 79%
                 8 = 80 - 89%
                 9 = 90 - 100%
Col. 43          What issue received the highest vote for?      29
                 1 = Streets
                 2 = Water
                 3 = Other Sewer /Drainage
                 4 = Fire
                 5 = Law Enforcement
                 6 = Hospitals, Schools, Other Buildings
                 7 = Parks
                 8 = More than One of the Above
                 9 = The Issue at Hand
Col. 44          What issue received the lowest  vote for?       30
                 Same code as in Column 43.
Col. 45          Percentage of Population Registered           31
                 Usual code for percentages as above:
                 [0 - 9] with 0 =  0-9%
                            1 = 10 -  19%
                            etc
Col. 46          Largest Turnout for any Purpose in that        32
                 Election
                 Usual code for percentages as above:
                 [0 - 9] with 0=0- 9%
                            1 = 10 -  19%
                            etc
                        160

-------
Data from the Newspaper
Reporter Data Sheet
   Col.  47           Existence of Pressure Groups for the Issue    33
                    1 = Yes
                    2 = No
   Col.  48           Importance of These Groups in the Outcome   34
                    1 = Yes
                    2 = No
   Col.  49           Existence of Pressure Groups Against the Issue,,-
                    1 = Yes
                    2 = No
   Col.  50           Importance of These Groups for the Outcome  36
                    1 = Yes
                    2 = No
   Col.  51           Were  These- Groups Partisan-jja Nature        37
                    1 = No
                    2 = Yes,  Republican
                    3 = Yes,  Democrat
                    4 = Yes,  Others
                    5 = Yes,  Republican and Democrat
   Col.  52           Role Played by Newspapers during the        38
                    Campaign
                    1 = Important For
                    2 = Important Against
                    3 = Important For and Against
                    4 = Low For
                    5 = Low Against
                    6 = Low For and Against
                    7 = None
                             161

-------
Col. 53          Role  Played by Other Media during the        39
                 Campaign
                 1 = Important For
                 2 = Important Against
                 3 = Important For and Against
                 4 = Low For
                 5 = Low Against
                 6 = Low For and Against
                 7 = None
Col. 54-55       Expressed Position Relative to the Project:    40
                 Local Officials
                 00 = None
                 01 = Pro-Money Contribution
                 02 = Pro-Door to  Door
                 03 = Pro-Speeches
                 04 = Pro-Public Endorsement
                 05 = Pro-Leaflet, Newspaper  Ad
                 06 = Pro-Other
                 07 = Pro-Combinations of 2 or 3 of the above
                 08 = Pro-Combinations of more than 3 of the above
                 09 = Against-Money Contribution
                 10 = Against-Door to Door
                 11 = Against-Speeches
                 12 = Against-Public Endorsements
                 13 = Against-Leaflet, Newspaper Ad
                 14 = Against-Other
                 15 = Against-Combination of 2 or 3 of the above
                 16 = Against-Combinations of more  than 3 of the
                     above
                 17 = Pro and Against - Money Contribution
                 18 = Pro and Against - Door to Door
                 19 = Pro and Against - Speeches
                 20 = Pro and Against - Public Endorsement
                 21 = Pro and Against - Leaflet, Newspaper Ad
                          162

-------
   Col.  54-55
   (cont.)
   Col.  56-57



   Col.  58-59



   Col.  60-61

   Col.  62-63


   Col.  64-65


   Col.  66-67

   Col.  68-69

   Col.  70-71


   Col.  72-73


   Col.  74-75

   Col.  75-76
Expressed Position Relative to the Project:
Local Officials
22 = Pro. and Against - Other

23 = Pro and Against - Combination of 2 or  3
     of the above
24 = Pro and Against - Combination of more
     than 3 of the above

Expressed Position Relative to the Project:
                             Doctors

[Same code as for Col.  54-55]
41
                              Politicans and  42
                              Party Organ-
                              izations
                              State Officials  43

                              Chamber of     44
                              Commerce
                              Other Local
45
                              Governing Bodies

                              Ad hoc Groups  46

                              Clergy         47

                              League of      48
                              Women Voters

                              School and Uni- 49
                              versity Groups

                              PTA           50

                              Others         51
CARD 2


Identification


   Col.  1-5
Alphanumeric:  Location of the referenda
(as in card 1)
                             163

-------
   Col. 6-7         Numerical Identification of the referenda
                    (as in card 1)
   Col. 8-9         State where the referenda were held (as  in
                    card 1)
   Col. 10          Card Number:  2
Data from Newspaper
Reporter Data Sheet
(continued)
   Col. 11          Use of Newspaper Ads                        52
                    1  = Yes
                    2  = No
   Col. 12          Use of Radio and TV Ads                      53
                    1  = Yes
                    2  = No
   Col. 13          Use of Telephone Surveys                     54
                    1  = Yes
                    2  = No
   Col. 14          Transportation Offered to Voters              55
                    1  = Yes
                    2  = No
   Col. 15          Advertisement Mode: Literature, Brochures,  56
                    Leaflets
                    1  = Yes
                    2  = No
   Col. 16          Advertisement Mode: Film                    57
                    1  = Yes
                    2  = No
   Col. 17          Advertisement Mode: Posters                 58
                    1  = Yes
                    2  = No
                              164

-------
Col. 18          Advertisement Mode: News Stories            59
                 1 = Yes
                 2 = No
Col, 19          Advertisement Mode: Letters to Editor        60
                 1 = Yes
                 2 = No
Col. 20          Advertisement Mode: Speeches and Meetings   61
                 1 = Yes
                 2 = No
Col. 21          Advertisement Mode: Door to Door            62
                 Canvassing
                 1 = Yes
                 2 = No
Col. 22          Criticism: Cost and Size of Project            63
                 1 = Too Large
                 2 = Too Small
                 3 = Both
                 4 = Not Raised
Col. 23          Criticism: Taxes,  Fees, and Assessment     64
                 1 = Increase  in Taxes
                 2 = Increase  in Sewer Fees
                 3 = Type of Assessment
                 4=1 and 2
                 5 = 1 and 3
                 6=2 and 3
                 7 = 1 and 2 and 3
                 8 = Not Raised
Col. 24          Criticism: Source  of Revenue                65
                 1 = Raised
                 2 = Not Raised
                          165

-------
Col. 25          Criticism:  Allocation of Costs Benefits        66
                 1 = Raised
                 2 = Not Raised
Col. 26          Criticism:  Timing of Construction             67
                 1 = Raised
                 2 = Not Raised
Col. 27          Criticism:  Location of Improvements          68
                 1 = Treatment Plant
                 2 = Sewer and Drainage
                 3 = Both
                 4 = Not Raised
Col. 28          Criticism:  Not Necessary                     69
                 1 = Raised
                 2 = Not Raised
Col. 29          Criticism:  Opposed to Growth                 70
                 1 = Raised
                 2 = Not Raised
Col. 30          Criticism:  Opposed to State or Federal        71
                 Interference
                 1 = Raised
                 2 = Not Raised
Col. 31          Criticism:  Local versus Regional Manage-     72
                 ment
                 1 = Favor of Local Authority
                 2 = Favor of Regional Government
                 3 = Both
                 4 = Not Raised
Col. 32          Criticism:  Distrust of Local Officials          73
                 1 = Raised
                 2 = Not Raised
                            166

-------
   Col.  33          Criticism:  Alternative is Better              74
                    1  = Raised
                    2  = Not Raised
   Col.  34          Criticism:  Administration Did Not Give       75
                    Enough Information
                    1  = Raised
                    2  = Not Raised
Data from City
Comptroller Sheet
   Col.  35          Total Cost of the Project (Dollars)             76
                    1  = 0 -99,999
                    2  = 100,000 - 499,999
                    3  = 500,000 - 999,999
                    4  = 1,000,000 - 4, 999, 999
                    5  = 5,000, 000 and over
   Col.  36          Bonds Repaid by User Charges                77
                    1  = Yes
                    2  =.No
   Col.  37          Bonds Repaid by Increased Taxes              78
                    1  = Yes
                    2  = No
   Col.  38          Bonds Repaid by Increased Taxes Until         79
                    Facility Can Be Supported by Users Charges
                    1  = Yes
                    2  = No
   Col.  39          Bonds Repaid by Special Assessment           80
                    1  = Yes
                    2  = No
   Col.  40          Bonds Repaid by State Grants                  81
                    1  = Yes
                    2  = No
                               167

-------
Col. 41          Bonds Repaid by Federal Grants               82
                 1 = Yes
                 2 = No
Col. 42          Current Tax Rate ($ per  $1, OOP of Assessed    83
                 Valuation)
                 1 = 1-24
                 2 = 25-49
                 3 = 50-74
                 4 = 75-99
                 5 = 100-124
                 6 = 125-149
                 7 = 150-174
                 8 = 175-199
                 9 = 200 and more
Col. 43          Tax Rate Increase from 1966  to 1970          84
                 0 =  0-9%
                 1 - 10 - 19%
                 2 = 20 - 29%
                 3 = 30 - 39%
                 4 = 40 . 49%
                 5 = 50 - 59%
                 6 = 60 - 69%
                 7 = 70 - 79%
                 8 = 80 - 89%
                 9 = 90 - 100%
Col. 44          Proportion of Project Costs Repaid by         85
                Increased Taxes
                Usual code for percentages as above
                [0  - 9]
                          16 8

-------
Col. 45          Amount of Project Costs Repaid by            86
                 Increased Taxes
                 1=0- 499,000
                 2 = 499,000 to 999,000
                 3 = 1,000,000 to 9,900,000
                 4 = 10,000,000 to 24,900,000
                 5 = 25, 000, 000 to 49, 900, 000
                 6 = 50,000,000 to 99,900,000
                 7 = 100,000,000 to 499,900,000
                 8 = 500,000,000 to 999,900,000
                 9 = 1,000,000,000 and over
Col. 46          Proportion of Project Costs Paid with         87
                 State Funds
                 Usual Code for Percentages [0 to 9].
Col. 47          Proportionof Project Costs Paid with         88
                 Federal Funds
                 Usual Code for Percentages [0 to 9].
Col. 48          Total Indebtedness at  Beginning of 1970        89
                 (Dollars)
                 1 = 0 to 49,000
                 2 = 50,000 to  99,000
                 3 = 100,000 to 900,000
                 4 = 1,000,000 to 2,400,000
                 5 = 2, 500,000 to 4,900,000
                 6 = 5,000,000 to 9,900,000
                 7 = 10,000,000 to 49,900,000
                 8 = 50,000,000 to 99,900,000
                 9 = 100,000,000 and over
Col. 49          % Increase of Decrease of Indebtedness        90
                 Between 1966-1970
                 1 = High Decline (20% or more)
                 2 = Decline
                 3 = Stationary (less than 5%)
                 4 = Increase
                 5 = High Increase (20% or more)
                          169

-------
Col. 50-51       Portion of Operating Budget Repaying          91
                 the Bonds
                 01 = 0 to 2.4
                 02 = 2. 5 to 4.9
                 03 = 5 to 7.4
                 04 = 7. 5 to 9. 9
                 05 = 10  to 12.4
                 06 = 12.5 to 14.9
                 07 = 15  to 17.4
                 08 = 17.5 to 19.9
                 09 = 20  to 29.9
                 10 = 30  to 39.9
                 11 = 40  to 39.9
                 12 = 50  and above
Col. 52          Is there a debt limit?                          92
                 1 = No
                 2 = Yes, but water or sewers  excluded
                 3 = Yes and water or sewers included
Col. 53          Who set the debt limit?                        93
                 1 = State Constitution
                 2 = State Law
                 3 = Local Charter
                 4 = Local Law
Col. 54          What would passage of bond mean in terms      94
                  of approaching debt limit?
                 1 = Nothing
                 2 = Close
                 3 = Exceed
                           170

-------
   Col.  55-56       Current Credit Rating                         95
                    01 = A
                    02 = AA
                    03 = AAA
                    04 = A AAA
                    05 = B
                    06 = BA
                    07 = BAA
   Col.  57          Rating System                                96
                    1  = Moody's
                    2  = Standard and Poor
                    3  = Both
                    4  = Others
                    5  = None
                    0  = Not Applicable
Data from Census
City Planner Data
Sheet
   Col.  58          Median Family Income (Dollars)              97
                    1 = 0 to 2,999
                    2 = 3,000 to 3,999
                    3 = 4,000 to 4, 999
                    4 = 5,000 to 5, 999
                    5 = 6,000 to 6, 999
                    6 = 7,000 to 7, 999
                    7 = 8, 000 and above
   Col.  59          Proportion  of Families Whose Income Is      98
                    Less Than $3, OOP/Year
                    Usual Code  for Proportions [0 to 9].
                             171

-------
Col. 60          Proportion of Families "Whose Income Is       99
                 Between $3, OOP and $10, OOP/Year
                 Usual Code for Proportions  [0 to 9].
Col. 61          Proportion of Families Whose Income Is      100
                 Higher than $10, OOP/Year
                 Usual Code for Proportions  [0 to 9].
Col. 62          Proportion of Population of Age Between      101
                 21 and 35
                 Usual Code for Proportions  [0 to 9].
Col. 63          Proportion of Population of Age Over 65       102
                 Usual Code for Proportions [0 to 9].
Col. 64          Proportion of Population Non-White           103
                 Usual Code for Proportions [0 to 9].
Col. 65          Proportion of Population Who Was             104
                 Already Resident in the Area in 1955
                 Usual Code for Proportions [P to 9].
Col. 66          Proportion of Single Family Housing Units    1P5
                 Usual Code for Proportions [0 to 9].
Col. 67          Proportion of Owner Occupied Housing        106
                 Units
                 Usual Code for Proportions [0 to  9].
CANCELLED     Most Serious Growth Problem Facing the Area
                 1  = Housing
                 2  = Transportation
                 3  = Schools
                 4  = Hospitals
                 5  = Water
                 6  = Sewers
                 7  = Lack of Economic Base
                 8  = Others
                 9  = None
                           172

-------
   Col.  68-70       BLANK
   Col.  71-72       State where referendum was held             107
                    (same code as col.  8-9)
Data from City
Engineer Data
Sheet
   Col.  73          Availability of Map to Voters                 108
                    1 = Yes
                    2 = No
   Col.  74          Project Aimed at Counteracting Water        109
                    Pollution Problem
                    1 = No
                    2 = Present
                    3 = Future
                    4 = Present and Future
   Col.  75          Special Topographical Conditions             110
                    Contributing to the Need of the  Project
                    1 = Yes
                    2 = No
   Col.  76          Special Topographical Problems              111
                    Associated with the Project
                    1 = Yes
                    2 = No
   Col.  77          Land Development Created Project Need?     112
                    1 = No
                    2 = Yes, Residential
                    3 = Yes, Industrial
                    4 = Yes, Commercial
                    5 = Yes, 2 and 3
                    6 = Yes, 2 and 4
                    7 = Yes, 3 and 4
                    8 = Yes, 2 and 3 and 4
                              173

-------
Col. 78          Will Project Serve Future Development       113
                 Area?
                 1 = No
                 2 = Yes, Residential
                 3 = Yes, Industrial
                 4 = Yes, Commercial
                 5 = Yes, 2 and 3
                 6 = Yes, 2 and 4
                 7 = Yes, 3 and 4
                 8 = Yes, 2 and 3 and 4
Col. 79          Federal Regulations Associated with Project  114

                 1 = No
                 2 = Yes, Technical
                 3 = Yes, Financial
                 4 = Yes, Both
                 5 = Yes,  Others
Col. 80          Local Regulations  Associated with Project     115
                 1 = No
                 2 = Yes,  Technical
                 3 = Yes,  Financial
                 4 = Yes,  Both
                 5 = Yes, Others
                         174

-------
DATA FILE
  175

-------
CDLMB
HAR73
ROCPT
ATHD'L
BATES
HAR90
PALAL
WAREN
BENDD
KYLET
JACKS
MILLV
RC3LLA
ALV IN
ROYSE
JFRSN
PETAL
BRVRD
BELTN
V A N A L
DKALB
M N K T 0
PRTLD
S I M I V
DGLAS
ROSEL
OAKWD
OKLAC
RMDRK
BROOK
DDNLD
GTFAL
CACHE
G R n V S
HTCHK
MORGN
AVOND
M O O P F
CUERC
LACTY

Locality
1
2
3
4
5
6
7
8
0
1
2
3
4
c
7
8
c
20
21
22
23
24
25
26
2 7
28
29
30
3 1
32
33
34
35
36
37
38
l 0

bhwnber
ve 1 4
-4-e 4
? 3
44 7
(-6 4
05: 1 S
24 1 3
•614
54 1 =
05 1 e
27 i:
461
2012
05 1 5
3512
4- € 1
461<:
121;
25 1
161'
05 If
02 1
24 1 ;
37 1
38 1^
461
461 '
201
28 1 f
381
461'
07 1
20 1
371
381
461
05

If
en H
2040
3030
3069
3039
1029
3089
4118
5118
309 8
2118
3088
3047
1097
3047
1067
40 16
2036
1036
2 1 26
4050
5059
2079
5079
1048
2018
3118
5058
2 1 28
2118
2057
5107
2017
i057
2057
3116
3116
2066
2116
5036

-t N to. i *
000 c
000 c
•'-"t •,•
1
12111
126 12
~ 21 4 5
1224 =
2235
12113
2111
121 7C
12111
12141
12113
12611
12213
12111
1 2£ 35
12113
12113
126 1 2
1 22 1 1
22624
22P 13
22 f 13
22160
22 1 1 2
22? 75
221^'
22^13
2^111
p P p -3 c:
22114
22111
22212
2 ? 1 ]_ ~-
2? 1 1 2
2 281 1
22112
22131
22160

i ifl sO* f*~ OO
j o o-| o o
5 O O O O

11223
1221
11223
1215
11221
1 2 3 1
1 1227
1223
11225
11223
121 3
1 2 P 5
11215
21225
11223
I 1 2 ' 7
11225
1221-
22224
11223
11133
11235
11223
11037
1112V
1 1 3
11223
11223
1 1 1 2 '4-
11117
1222"
l 1 22 S
11222
L 1 1 2 5
1 1 22 i
12244
2100

0202
f 406
6 T 05
1C 1 0 r
01 11
34111
0405
3C £ 05
0 4 0 5
sc e 1 1
6403
3C 4 04
3702
3C 106
64 f 5
4 f 4 05
SO 7 05
3C 4 1 5
S0602
0401
30404
36 ^01
>09l 1
3f. l 1 4
i 0 4 04
10108
36709
6405
30411
36402
1.041 4
30702
30302
20705
36404
0111
50702
:4 C 50
6111

OOOOOOOOO 0
i r i '

1
01
0.1
1 f
>l
01
l'l
01
di
3 1
as
3 5
01
30
01
3 1
0.1
37
05
30
0 1
3 1
1 5
1 1
L '2

37
0.1
15
01
37
01
37

34
'"*' ~7
31
) 5

. 9
-6
d
L 8
6
8
6
7
38
7
38
7
19
.6
I 8
8
28
28
14
p
12
5
1
4
i
14
5
1 7
4-
1 3
2 3
33
3
14
4-
3 1
5

. •
a"° N
O O O


itt
1311
1301
'301
1 C 01
1311
200^
1 30 1
100
1 3 0 1
ICO]
2301
1 001
I 31
1C 1 ]
f 3 l" 1
20 1
1311
102
1211
1 C 02
L 30
ICO]
2301
1C01
r ) o i
1 00
L 30 1
1 02
l' 31
101 1
1 32 1
102
1 30 J
1 00
1301
1 00
1321
0 2 1 f
2301

2C 0
900
9C 0
930
900
9 30
^ 900
8
900
930
900
9 3 3
900
2 3 3
900
933
900
6 3 0
800
9 ) 3
5 15
9 3 0
L 900
93 3
c n p
93" 3
300
S 3 3
81 9
33 3
L 900
2 3 3
L 203
33 3
2900
333
900
2 3 3
3069
330

00
00
00
00
00
9 9
0 0
00
00
92
0 0
00
00
06
91
00
6
00
99
00
00
00
00
S '4
1 9
00
77
2 3
3 .7
9 S
00.
99
OC
13
0 0

4 £
3 2
7 4
30
5 t
4
4 2
3 2

4 1
3 5
3 1
5 p
3 1
3 2
3 3
*2
4. 4-
34
3 3

H
!1
4- 2
3 ?
-! 4
33

3
2 6
1-2
3 1


1
1
1
2
1
1
1
1
2
3
1
1
1
1
2,
2
1
1
1
1
>
p
1
1
1
1
1
I
2,
1
1
1
1
1

c
c
c
1
c
i
1
1
0
0
c
1
1
1
1
0
"
1
2
1
1
£_
0
1
1
1
1
1
1
0
1
1
1
1
2

?. c
20 C
2 0 C
222
2 0 (
201
2 0 '
1 1 1
2 0 '
221
2 0 (
201
2 0 <
20
2 0
03 -
2 0
1 1 1
2 Of
2 3
1 1 ;
1 2
1 1
1 1 :
1 1
2 3
1 L 1
1 1
23"
1 1 (
1 . :
232
1 .
'<. ?s
1 I 1
1 1 2
1 L
1 1 2
1 1

1
OOOOOOOOOOOOOO<~
II
II
4 7
)T 7
1 7
~27
327
07
34 7
77
347
1 7
3 1 7
1 7
377
) 57
1 1 7
l"7
L 16
1 5
31 71
) t 7
267
I 5 7
1 67
> t 1
1 ~li
257
3 47
17
47
T "7
347
> 5 5
267
1 7
3 '47
3 7
57
210
I 4-7

04
04
07
04
04
38
04
yf
04
07
01
3 4
04
38
07
38
OS
08
0£
01
Oc
33
23
37
04
04
00
37
08
34
08
20
Oti
37
0 0
23
2 3
38
70
04

•^ 00 
-------
HAR7."
ROCPT
ATHOL
BATES
HAR9C
PALAL
WAREK
BENDO
KYLET
JACKS
MILLV
ROLLA
ALVIN
ROYSE
JFRSN
PETAL
BRVRO
BELTN
VANAL
DKALB
M M K T 0
PRTLD
SI 'A IV
DGLAS
ROSEL
OAKWD
OKLAC
RNDRK
BROOK
DOMLD
GTFAL
CACHE
GR 0 V S
HTCHK
MORGN
AVONO
MOORE
C U E R O
LA'CTY
Locality
01
02
03
04
05
0£
07
oe
09
1C
1 1
12
13
14
15
16
17
is.
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Number
4C
4£
2;
44
4C
05
24
39
4<:
54
oe
27
4f
54
20
05
35
46
46
12
2d
46
oe
02
24
37
3f
46
46
20
2f
3R
46
07
20
37
3£
46
05
OJ
3
53
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2.
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
2
1
p

1
2
1
1
1
1
1
1
1
2
1
1
2
p
1
1
2
1
1
p
2
p
1
1
1
P
1
, (M r
m i
V
'OiS
$1
in>
2
2
2
2
2
2
2
2
2
2
p

2
1
2
i
2
2
1
2
2
2
1
?
?
1
1
2
1
p
2
P
2
2
2
2

2
1
p
p
1
p
1
1
1
p
p
1
1
2;
1
1
2
2
1
2
2
1

1
2
1
1
1
1
2
2
2
2
1
2
2
?
1
1
1
1
1
p

2
2
2
1
1
1
1
2
2
1
2
2
1
2
1
1
1

2
1
1
C
?
2
i
i
i
.J
i
i
2
0
1
i
2
1
1
1
1
1
2
2
2
2
2
2
2
2
1
2
?

2
2
2
0
2
0
2
2
1
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
2
2
1
2
1
1
2
2
1
2
2
2
1
p

p
2
1
1
1
0
p
2
1
2
1
1
2
9
1
1
2
3
1
p
1
p
1
p
1
1
2
1
1
1
2
3
1
1

1
1
1
1
0
1
1
1
1
1
2
1
1
1
2
3
2
p
1
2
1
1
1
1
1
2
2
2
2
2
2
1
1
2
2
2
1
2
0
2
1
2
2
2
1
1
2
2
0
1
2
2
2
1
1
1
1
1
2
1
2
2
1
1
p
2
1
2
1
1
0
p
1
1
2
2
1
2
p
1
2
2
0
p
1
p
1
1
1
1
p
2
2
2
2
2
2
1
2
p
2
2
2
2
p
p
p
2
1
2
p
1
2
2
3
2
?
P
p
p
2
4
4
1
4
4
4
4
4
4
4
4
4.
4
1
4
4
1
4
1
1
2
4
1
4
4
4
4
4
1
4
1
1
1
I
4
R
S
1
1
S
f.
R
4
1
1
R
1
B
1
R
1
R
1
1
1
1
e
2
4
1
R
4
1
5
1
1
1
4
i
i
4
1
8
jooooooooooo













e.
1
2 i
c
£ t.
C. i
1
2 ^
2 <
2]
2
1 ?
2
1 1
? i
i
2 1
2 ]
2
p
2
2 1
1 <
p
2

2.4
24
14
2 4
24
24
p 4
2 -^
2 4
24
p p
24
? i.
2 l
24
24
2 -^
24
> p 2
; 22
: 24
p K(-
L 21
p 3
224
P i
224
3 ^ 4-
21 4
£ 2 )-
L 1 4
224
2 2 i
f 21
»o ^ o 3
o o o o
1

1
222
122
2 2 ]
2 2 f
122
2 2 2
222
222
21
22?
p 22
22 1
1 ?2
2 2
1 1 ?
22?
1 22
12?
£ 2 ^
2 2 c
21 E
2 2 c
2 2

12
2 2
22
1 1 L
2 2
2 2'
1 1
: 4 2
4 ^
p j
4
42
> p p
2 2
2 4 i
I1 ?

i?
242
4
) t, '
'j_ -,
4?
tt-P"
' 42
't
1 S
24^
'L ^
24
4 1
24?
',
' P ',

I
1
24;
. 'i 1
2 4
O"» O ft M
xo r- r- r-
o o o o
11
j 2
p
1
> p
2
2 1
p
p

2
2
p
12
p
2
L?
t 1
2 1
2
p
> P
> p
21
> P
12
P
2 2
1
-1 P
•O 't1
? '^
3 1
31
41
1 1
4 2
4 1
51
31
21
-2 1
32
i p
21
1 1
42
.4 1
221
on
3 p 2
4 2
152
> p p
p ^
' •)• 1
222
> p
Ibl
'1 >

24 1
p
1 ri 2

2 b -1
2
12
p
2 2
1 1
22
2 2
1 2
2 ?
2
22
12
? 2
12
p 2
1 1
p 2
p
2: 1
1 2
12
p p
12
i 2
2 2
1 ?
1 2
I 2
12
1 2
12
1 2
1 2
2 2
p
22
2 2
1
i ;
2 'c.
1 2
2
1 2
1 P
p ^
1 2
21
p ;
2 *
1 =
1
2 <
2
1
^
Ifl *£ r- O3 CT* O
o o b b c
1 I
II
> o c
j
3 O

1
2
2
1
1
2
^
2
1
¥
>?
2
2
2
Jp
O
2 1
p
) p
3 p
3 I
2 1
3 P
21
> P

> P
21
3 3
21
J 3

** p
00 0
O C

1C
1C
1 1
29
1C
3 1
1 C
3
1C
1C
1 3
1 3
1 1
1 0
2 )
91
p p
:3
4 0
1 0
1 0
1 0
1 3
1 2
3 L
4 J
03
63
C 4
01
9 1
17
C £
26
00
91
o
S 6
05
04
C 1
09
04
63
9 7
4
1 2
0 '+
68
91
3?.
D ^
i-
4
96

So <

700

2
-i
0
4
2
3
0
2
1
0
0
0
0
o
0
0
0
0
3
0
4
0
C
0
c
4
0



1
1
1
4
e
C
3
>
+
7
e
i
0
(^
0
2
3
1
1
o
D

2
4
3
5
4
5
4
4
4
4
p
4
0
4
4
iL
/r
4!
L
4
4
C
i

072
122
2
01 2
01 2
12 'i
01 1
01 1
1 1 C
061
b; 5
092
102
1C 2
393
392
092
3
1 C 2
oqi
2
0222!
1
073
0 1 2
1 1 3
0=2
1
£
0 1 2
073
001
00 1
D i-l (\) '
O O CM
1 I
4!
1005
: j DCS
2 1 o"n 1
: : 021
1003
i:00f:
1C 02 3
1003
1 0 1
: j
; o i i
21005
= 'f. 07 1
102 1
:2 005
21005
1 0 6 5
102 L
= To 1 1
1003
i ooo|
j c 6"i i
21005
I 1 00 5
120 71
?2 021
1 1 0 0 5
?goe i
21000
r2 1 Ot 1
11005
31024
3 C 0 Q 0
•* o^ os cr- 1
3 O o O <
\ \ \

.1 6 C 1 1
3 6 0 l)o
flic i i
91 0
C S 4 1 1
C 7 1 2 3
17121
1 1
20
C 63 1 1
26130
2 6 i 2 r
26111
2 6 1 ? C
35021
1 1
1 o 2J2b
17121
16120
1 7|22 1
17110
1 6 1 2 C
11 1
1 1
44161
17120
3 1
C 7 1 2j 0
26120
0/510
2 0
45311
-coo-o gg
3 o o •* -"-*
™ II
III If
9 T
9
5 B 6
' 9 7

575
(b 38
537
36
1
4
736
35
=,
o Jo

3

sss ./
1


461222
46 1 2 22
4 2
4 £ 2 ^ 2 ?
05 14 1 1
?42 22
39 1.2
46 ? 11
4 112
27 4
6 1 '? 2 1
62222
514??
352 222
614.22
4~6 2 2 ? 2
22111
25? 1 ] 1
461311
05 14 1 p
02
?4 1122
3714 1 1
38 1422
1-61411
4 & 1 2 2 ?
20 ?422
28 14 ?;
38 2222
461221
5 2 1 4 1 1
3721 21
381422
34 1 i- 22
35 1 1
O O 0 i-t 1^

•

, - „
2 1 1
722
1 2 1
2 f C 0
44
P5P1
.111
614?
_ 	 rrl ..
3?2 4
D o t[. .
1444
88 31
26 1 1
] 3 1 ^
Mm** 1040 CARDS



-------
INTERVIEW GUIDES
      179

-------
        (This will replace the City Assessor Data Sheet.)

                   ABT  ASSOCIATES  INC.
          55 WHEELER STREET. CAMBRIDGE. MASSACHUSETTS O2138
                  TELEPHONE   •  AREA 6 1 7 - * 9 3 - 7 1 O O
                 CITY PLANNER DATA SHEET

CENSUS DATA
       The following information is to be collected for the primary
governmental unit under study.  If it is available with more geo-
graphic detail,  then collect it as well.  Most governments  will have
I960 census data, and some will have more recent data..  Gather
I960 data for all cities; gather more recent data when available and
specify the date and source of the more  recent data.

1.     Population
           1950 census
            I960 census
            1970 census  (if available)

            Other recent census (specify)

2.     Median Family Income

3.     Income Distribution:

       Low:         Less  than $3,000/year _
       Medium:      $3, 000  to $10, 000	
       High:         $10,000 to $20,000	
       Very High:    Over  $20,000
                            180

-------
4.     Age




       Median Age
       Percent Over 65
       Age Distribution (if available):





                                           Percent
                    0 - 4




                    5 - 9




                   10 - 14




                   15 - 19




                   20 - 24




                   25 - 29




                   30 - 34




                   35 - 39




                   40 - 44




                   45 - 49




                   50 - 54




                   55 - 59




                   60 - 64




                   65 - 69




                   70 - 74




                   75 and over






5.     Percent non-white	






6.     Percent foreign - born
                               181

-------
 7.      Type of Housing

        Number of single-family units	

        Number of multi-family units	
        Number of owner-occupied units

        Number of rental units
 8.      Percentage of unemployment	


 9.      Occupation classification:


                       Percent white collar


                       Percent blue-collar
10.      What is the most serious growth  problem facing the primary
        governmental unit (if any)?
                       Housing

                     __ Transportation

                       Schools

                       Hospitals

                     _ Water

                     _ Sewers

                       None

                       Others (Specify)
11.      What is the tax rate for the primary governmental unit (in
        dollars per $1,000 assessed  valuation)?
                                 182

-------
11.     What is the  tax rate for neighboring governmental units?
       (in dollars per $1,000 assessed valuation)?
       Name of Governmental Unit           Tax rate
12.    Have there been any recent increases in taxes in neighbor
       ing governmental units ?  If yes,  specify.
13.    Get copy of 701 plan (or master plan).
                               183

-------
                    ABT  ASSOCIATES  INC.

           55 WHEELER STREET, CAMBRIDGE, MASSACHUSETTS O2138

                   TELEPHONE   •  AREA 6 1 7 - -4 9 2 - 7 1 O O
              DA T
 1.      What is  the exact nature of the improvements to be financed
        by the bonds?
2.     Where were the improvements  to be located?
       a.   (Get map if possible. )

       b.   Was a map showing the location of the improvements
            available to the voters before the referendum?

                       Yes
                       No
3.     Was this project designed to counteract a present or
       future water pollution problem?

                      Present
                      Future
                      No
4.     Are there any topographical conditions which contribute to
       the need for the project? Specify.
                               184

-------
5.     Are there any topographical conditions which affect the design
       of the project?  ^pecify.  (Obtain a topographical map of the
       area if available. )
6.     Was the need for the project created by recent land development?

       	Yes



       	No


       If ye s:

       a.  What sort of development?

           	Residential


           	Industrial


                         Commercial
       b.  Location of the development.
7.      Is the project designed to serve an area -where future develop-
       ment will be accomodated?

       	Yes

                    No
       If yes:

       a.   What sort of development?

           	Residential	Commercial

           	Industrial


                               185

-------
       Are there any federal regulations or requirements associated
       with any characteristic of the project:  Specify.
                       Yes


                       No


                       Technical


                       Financial


                       Others
9.      Are there any state or local regulations or requirements in
       connection with any characteristic of the project?
                       Yes
                       No


                       Technical


                       Financial


                       Other
       Specify.
                               186

-------
                    ABT  ASSOCIATES  INC.

           55 WHEELER STREET. CAMBRIDGE. MASSACHUSETTS Oai38

                   TELEPHONE  •   AREA 617-492-71OO
               CITY CONTROLLER -- DATA SHEET


1.    What is  the total cost of the project proposed to the voters in this
      bond election?  $	


2.    How -will money be raised to repay the bonds?

      	 user charges

      _     increased taxes

      	increased taxes until facility can be supported by
            user charges

      	Special assessment

        	 State grants

        	Federal grants


3.    What is  the tax rate in your town currently?  $	

      per/$l,000 of assessed valuation.

      [if rate is different for property, schools,  etc. , list all amounts. ]
4.    What has the tax rate been for the past four years?

      1969 $             /$1000 assessed valuation


      1968


      1967


      1966

      [Note year of bond election  with *. ]


                                18?

-------
 10.    Total cost of the project
11.    Portion of the cost -which is paid for by federal grants
12.   Portion of the cost -which is paid for by state grants
                              188

-------
 5.     What portion of the cost will be repaid with increased taxes?
      _$	% of the total project cost

 6.     What is  the most recent total assessed valuation of the primary
       governmental unit?
       $                              Date
 7.     How much of the total project would be paid for with State or
       Federal funds ?
       State $	

       Federal $
 8.     Can you give me a general idea of the total amount of State and
       Federal funding in the town,  exclusive of compensation to indi-
       viduals,   such as social security or welfare  payments?
       $	

 9.     $	of outside funding/assessed value/person, from
       questions  6 and 8 (controller) and 1 (City Planner).

10.     What was  the total indebtedness of  (primary governmental unit)
       as  of the beginning of the 1970  fiscal year?
       1970  $ 	

11.     How does  this compare with  previous  years?
       1969 --  $	
       1968 --  $	
       1967 --  $	._	
       1966 --  $              	
                                   189

-------
12.    How much of the total operating budget was allocated for repay-
       ing bonds and interest on them at the time of the bond election?
       $	

13.    Is there a debt limit imposed on (the primary governmental unit)?
       Are water and sewer projects regulated by this limit?

14.    If yes to 13,  by whom was the limit set?	
15.    If yes to 13,  what would passage of this bond mean in terms of ap-
       proaching this limit?
16.    What is (the primary governmental unit's) current credit rating?

       Whose system?
17.     Is there a prospectus available describing these bonds or any
       others recently issued by the (primary governmental unit)?

       [if so,  request it. ]
                               190

-------
                     ABT  ASSOCIATES  INC.
            55 WHEELER STREET. CAMBRIDGE, MASSACHUSETTS O213S
                    TELEPHONE   •  AREA 617-492-71OO
                   CITY CLERK -- DATA SHEET

1.    What is the largest governmental unit, the population of which
      voted on the referendum?
      	city
      	 county
      	 district
      	township
      	 statutory authority
      	sewer district
      	    other (specify)
NOTE:  In all the following questions,  we shall refer to this unit as
the primary governmental unit^

2.    If the answer to question 1 is "city" or "town":
      What is the type of this primary governmental unit?
      	 strong mayor-council
      	weak mayor-council
      	council-manager
             commission
       	 other  (specify)

3.    Are there independent governmental units within the boundaries
      of the primary unit?
             yes
             no
                                 191

-------
4.     Is the population which will be served by the sewers project the
       same as the population which pays for  the bonds?
                   Yes
                   No
       Is it the same as the population which voted on the referendum?
       	Yes
       	No

5.     What is the type of election of  the government of the primary
       governmental unit?
                   partisan
              	  non-partisan
                   ward
       	        at-large
       	        ward and at-large

6.     Population of  the primary governmental unit
       1950 Census	
       I960 Census
       1970 Census  (if available)
       1970 other source 'specify)

7.      The use of the -waterfront
       	not applicable
       	commercial
       	  residential
                     industrial
                                192

-------
REFERENDUM CHARACTERISTICS
1.
2.
Date of the referendum

Month  	             Day               Year


Use of the proceeds:

2. 1     Type of construction

       	new

       	    extension or addition

       	          remodeling or alteration
       2.2     Description
                          sewer

                          water and sewer

                          storm sewer, drainage sewer, drainage
                          district


                          treatment plant

                          outfall sewer

                          intercepting  sewer

                          •waste stabilization ponds

                          other (specify)
3.     Wording of the referendum: [Attach.]


       Was the meaning clear from the wording?

       	Yes

                   No
                                 193

-------
       Other  characteristics of the  referendum  (stated in negative

       terms,  etc. )
4.      Was the  sewer (drainage) project offered to the voters as a

       single issue?                   Yes                   No

              [If NO, skip to Question 5.
              If YES, what was the outcome of the vote?]
       Total votes cast

       Total for	

       Total against

       Total blank
              [If the answer to Questio.n 4 was NO,  or
              if it were YES,  but the wording of the
              referendum indicated that it was not a
              single issue, ask Question 5.]
5. (a)  Sewer (drainage)projects like the one which was voted on in

       (name of the primary governmental unit) are some tine s part of

       a larger package of capital improvements (like hospitals, roads,

       parks,  drainage improvements, etc.)  which are presented to

       the voters as a single bond issue.  Was this the case in  (the name

       of the primary governmental unit)?

       	Yes

                   No
                                194

-------
       If YES, what was the wording and outcome of the vote:
       Wording:
       Outcome:
              Total Votes  cast
              Total for	
              Total against	
              Total blank
              [if answer to 5(a) was YES, skip part (b)
              and go on to question 6. ]

5. (b)   Sewer (drainage) projects  like the one which was voted  on in
       (name of the primary governmental unit) are sometimes divided
       into several  separate bond issues so that the voters may vote on
       each of the parts of the project (e. g. , treatment plant,  intercep-
       tor) separately.  Was  this the case in (name of the primary govern-
       mental unit)? If YES, what was the wording and outcome of the
       vote for each of these  issues:

       Wording of Issue 1
       Wording of Issue 2
                              195

-------
       Wording of Issue 3
       Wording of Issue 4
       Outcome:





                         Issue 1    Issue 2    Issue 3    Issue 4




       Total votes cast	




       Total for	




       Total gagainst    		




       Total blank       		










6.      Were there candidates for public office elected in this election?




       	Yes




       	No




                 	 Local election (primary, general, special)




       	State election (primary, general, special)




       	National election (primary,  general,  special)




                     Others
                               196

-------
7-      Were there any other bond issues (not related to sewers or
       drainage) on the ballot at the same time as this sewer (drain-
       age) bond issue?

       If YES, what was the wording and outcome of the vote for each
       of these bond issues:

       Wording for Issue 1
       Wording for Issue 2
       Wording for Issue 3

       Wording for issue 4

       Wording for Issue 4
       Outcome:
                         Issue 1    Issue 2    Issue 3    Issue 4
       Total votes cast
       Total for	    	
       Total against
       Total blank
8.     Number of voters registered and eligible to vote on the ref-
       erendum?	,	
                               197

-------
 9.     Largest number of votes cast for any purpose in that election.
10.     Number of voters registed in the different parties on the
        day of the election.
                  	Republican
                        Democrat
        	Other (specify)

11.     Same question for January 1970
                   	  Republican
        	 Democrat
                        Other (specify)

12.     Same question for November; 1968.
        	 Republican
        	Democrat
        	Other (specify)
13.      Same question for January, 1965.
        	Republican
        	Democrat
        	Other (specify)
                                19

-------
                      ABT  ASSOCIATES  INC.
             55 WHEEL-ER STREET. CAMBRIDGE, MASSACHUSETTS O2138
                     TELEPHONE  •   AREA 617-492-71OO


          STUDY OF WATER POLLUTION  BOND REFERENDA

              NEWSPAPER REPORTER  - DATA SHEET

Campaign Characteristics


       1.   Did pressure groups for the passage of the bond issue exist?

           	Yes

           	No

       2.   Were they of major importance in the passage or the rejec-
           tion of the bond?

           	Yes

                      No
       3.  Did pressure groups against the passage of the bond issue
           exist?

           	Yes

                      No
       4.  Were they of major importance in the passage or rejection
           of the bond?

           	Yes

                      No
       B.   Were these groups partisan in nature?
       6.   What was the role played by newspapers,  radio, and tele-
           vision during the campaign?
       7.   List the expressed position relative to the project of

           GROUP              PRO       CON        PARTY AFFILIATION

        a. Local officials       	       	        	
        b. Doctors              	       		
        c. Politicians &:
           party organi-
           zations
        d. State officials
                                   ~~ 199

-------
     GROUP         PRO       CON       ROLE       PARTY AFFILIATION
e.   Chamber of
     Commerce

f.    other local
     governing
     bodies

g.   water distri-
     bution net-
     works &
     agencies

h.   clergy

i.    League  of
     Women
     Voters

j.    school &:
     university
     groip s

k.   PTA

1.    other
NOTE:  THE CODE FOR ROLE MEASUREMENT FOR THE ABOVE QUESTION
        IS.:

  PRO     1.  Money contribution         4.  Public Endorsement

   &      2.  Door to door               5.  Leaflet,  newspaper ad
  CON    3. Speeches                    6.  Other

    8.  Were newspaper advertisements used?

         	 Yes

                   No
    9.  Were radio and television advertisements used?

        	Yes

        	No

   10.  Were telephone surveys used?

        	Yes

       	No


                                  2 CO

-------
11.    Was transportation offered to the polls for the voters?

      	Yes

      	No

12.    Other methods of advertisement used.
13.   How much money did each group spend on newspaper ad-
      vertisements ?
14.  How much money did each groups spent on television or
      radio advertisements?
15.  How much money was spent in other ways?
16.  If possible, try to obtain a copy of the advertising used
      during the campaign.
17.   What were the issues offered as  criticism or opposition to
      the bond?
18.   Were the voters aware of the real,  issues?

      	Yes

                 No
                              201

-------
               ABT  ASSOCIATES  INC.

      SS WHEEUER STREET, CAMBRIDGE. MASSACHUSETTS 03133

              TELEPHONE   •  A R E A 6 1 7 - A 9 2 - 7 1 O O
           DATA SHEET FOR CITY ENGINEER
1.  Was this project designed to counteract a present or a
    future pollution problem?

    	 Present

    	Future

               No
2.  Is there any other growth problem:

    	Housing

    	 Transportation

    	Water

    	School

    	Hospitals

    	None

    	Others

3.  Is the project designed to serve a zone where

    	Industrial

    	Commercial

    	Residential

    growth will be accommodated?

4.  Are there any federal regulations or requirements associated
    with any characteristic of the project:   Specify.

    	Yes       	No

    	 Technical

    	Financial

               Others
                           202

-------
 5.   Are there any state or local regulations or requirements
     in connection with any characteristic of the project?

     	 Yes        	No

     	 Technical

     	Financial

                Other
     Specify.

 6.   Location of improvements  (on map) (for in-depth studies,
     essentially).

 7.   Total cost of the project.

 8.   Portion of the  cost which is paid for by increased taxes.

     Specify type of taxes.

 9.   Portion of the  cost which is paid for by special assessment,

10.   Portion of the  cost which is paid for by federal grants.

11.   Portion of the  cost which is paid for by state grants.

12.   Copy of 701  plan of the city.
                              203

-------
      APPENDIX E
IN-DEPTH CASE STUDIES
          205

-------
ATHOL, MASSACHUSETTS
           207

-------
                               ATHOL
On March 3, 1969, the Town Meeting of Athol, Massachusetts approved
unanimously,  the following article:
      Article 30, VOTED, unanimously:  That $350, 000. 00
      (three hundred fifty thousand dollars) is appropriated
      for the construction of sewerage systems and sewage
      treatment and disposal facilities; that to raise this
      appropriation $50, 000. 00 (fifty thousand dollars) is
      appropriated from the stabilization fund and the trea-
      surer with the approval of the  selectmen is authorized
      to borrow $300, 000. 00 (three hundred thousand dollars)
      under Chapter 44 of the General Laws as amended;  that
      the board of public works are authorized to contract for
      Federal or State aid which shall be spent for the project
      and that such aid may be spent in addition to the amount
      appropriated herein provided that no amount shall be
      borrowed under this vote in excess of the amount nec-
      essary, with other available funds, to pay the total pro-
      ject costs, all as then estimated by the  selectmen,  and
      that the board of public works  are  authorized to acquire
      necessary land,  rights of way and  easement for this
      project.


This was not technically a referendum.   Nevertheless, because of the

similarities between referenda and town meetings, it is included in our
     i  30
sample.
Because of the prevalence of the town meeting form of government in

the New England states (and the infrequency of referenda in these states),

we were pleased to have an opportunity to analyze this governmental

mechanism and its impact on community decision-making on the issue

of sewers.


Population


The primary governmental unit involved  in this vote is Athol, Massachu-

setts, a town of approximately 12, 000  persons located about 75 miles
                               209

-------
 west and slightly north of Boston.  One striking feature of the town is its
 slow rate of population growth.   In 1950,  the population was 11, 554; in
 I960, it was  11, 637.  Although 1970 data are not yet available,  respon-
 dents expect  the 1970 population  to be relatively unchanged from I960.

 The population stability is associated with some other interesting feature
 of the community.  For example, as the Athol Master Plan pointed out:

       Athol is an unusually self-contained community.  The
       great majority of its resident workers have jobs within
       Athol (90% of its residents who work in manufacturing
       are employed within A.thol), and the majority of the
       labor force which staffs Athol jobs  come from Athol
       (about 75% in manufacturing).  The bulk of the shopping
       done by Athol residents is  done within Athol,  and few
       from outside of Athol shop there.  The regional junior
       and senior high schools are in Athol, with less  than 10%
       of their enrollment from outside of the town.  Lack of
       nearby urban centers,  poor access north-south, and
       only fair access east-west accounts for this condition,
       common in other areas many years ago, but vanishing
       in this  age of super metropolitan areas and influences.

 The lack of population growth correlates with the fact that certain attri-
 butes of the town's population have remained virtually unchanged over
 the past 20 years.  For example, the town has never  had  very many non-
 whites; only six appear in the I960 census.  Also, most residents have
 lived  in Athol most of their lives.  According to the I960  census,  92%
of the  residents in I960 had also been residents (or  residents of nearby
communities) in 1955.  Given this stability, the only significant change
in population characteristics  is that the residents have gotten older.  In
fact, the age distribution in Athol is quite dissimilar from most of the
other communities in our  sample--particularly those which are exper-
iencing rapid  growth.   Only 15% of the entire population of Athol was
between the ages of 20 and 34, and another 15% of the  population was
between 60 and 75.  Furthermore, Athol's median age was 35--nearly
10 years older than most cities in our sample.    (Data showing the
distribution of Athol residents by age appears in  Table 1).  According
                               210

-------
to the town's Master Plan,  the decade of 1950-1960 saw an increase in
the  number of children and an increase in the number of persons  of re-
tirement age, but a decrease in the number of persons of working ageT
                                                                    32
Table 1. Distribution of
Age
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85 and over
Athol Residents by Age
Number of Residents
1028
898
908
688
459
478
522
610
729
672
640
548
535
495
423
278
142
108
Source: I960 Census
 Athol is not a wealthy town.  The median family income in I960 was
 $5,400.  As the data in Table 2 show, a large portion of the population
 is concentrated in the $3, 000 to $7, 000 bracket.  Also, it is  significant
 to note that only 15% of the population have family incomes of less than
 $3, 000, while only 12% have incomes  over $10, 000.

 In spite of this low family income,  there is relatively little reported
 unemployment in  Athol.  Unlike neighboring New England towns which
 have relied on the declining industries of textiles and shoes,  Athol has
 a stable manufacturing  base.  Its two major employers,  UTD Corporation
 and L. S. Starrett Company, both tool manufacturers,  have provided the
 town with a stable source of employment. At the same time, the peculiar
                                211

-------
nature of the employment in these companies also contributes to the low
average family income.  One important characteristic of these industries
-,and hence of the town in general) is that they do not provide employment
opportunities for women.  Hence,  few women in Athol work,  and few
even look for employment.  Since looking for work is  a prerequisite for
being included in data on unemployment ,  the unemployment data for
Athol underestimate the number of unemployed women.  This  results in
a lower median family income in Athol than in other manufacturing towns
where women are more  likely to be employed.
Table 2. Family Income Distribution for

Family
0
$ 1, 000
$ 2, 000
$ 3, 000
$ 4, 000
$ 5, 000
$ 6, 000
$ 7, 000
$ 8, 000
$ 9, 000
$10, 000
$15, 000
$25, 000
Source:
Residents of Athol
Income
- $ 999
- $ 1, 999
$ 2,999
$ 3, 999
$ 4,999
$ 5, 999
- $ 6,999
- $ 7,999
- $ 8,999
- $ 9, 999
$14, 999
$24,999
and above
Number of Families
62
123
229
319
466
425
350
250
250
103
164
51
8
I960 Census
A bare majority of all housing units  in Athol are single-family,  owner-
occupied.  Only 55. 3% of all units are single-family,  and only 60% of all
units are owner-occupied.  This is not unusual for factory communities
in Massachusetts,  where two-, three-, and four-family houses  (•with the
owner residing in one of the units) is common.  In Athol,  21. 7% of all
housing units are two-family;  13. 0% are three- and four-family; and only
9. 9% are five-family or more.
                               212

-------
In spite of the low median family income, the city's financial position is
not particularly bad.  As mentioned earlier, Athol has a large, stable
industrial base.  Of the  $15 million in assessed valuation in 1969,  UTD
Corporation and L. S.  Starrett Company were the leading taxpayers, each
assessed at over $1 million.  In addition, Massachusetts Electric Com-
pany was assessed at approximately ,$700, 000, and New England Tele-
phone and Telegraph was assessed at approximately $450, 000.  Still,
low incomes  discourage the  town's  voters from taxing themselves
unnecessarily for public expenditures.  A relevant example cited by one
respondent is the fact that the town has  never seen fit to hire an assistant
for the Superintendent of Public  Works.  Consequently, when the Super-
intendent is in his office, his workers in the field are unsupervised,
thereby reducing the quality of that department's work.

                                                          33
The Finance  Committee Report to the Town Meeting of 1966   further
illustrates the financial concerns of the voters.   The Finance Committee
showed the Town Meeting that Athol's tax rate had increased  significantly
since 1953, and that in 1966 it was the highest tax rate for any Massachu-
setts town with a population exceeding 10, 000.   The Committee argued
that not only was this a heavy burden  on local taxpayers, but  that it also
discouraged new business from locating in Athol.  Finally,  the Committee
suggested that this high tax rate resulted from poor management and
excessive spending.  According  to the Committee,  this was most visible
in the increase in the cost of debt service for the town during the previous
year.  The solution proposed by the Committee  was a reduction in
spending, and  one of the specific items  mentioned by the Committee was
their recommendation that the new treatment plan proposed by the
selectmen not  be financed by taxation.

Government

Athol is  governed by a Representative Town Meeting,  which has full legis-
lative authority for  the town.  The Town Meeting does not consist of all
eligible voters in the community, but rather of  180 elected  representa-
                                213

-------
tives.  The town is divided into six precincts, and 30 representatives are
elected to represent each precinct.  A non-partisan election is held on
the first Monday in February of each year, and each precinct elects ten
representatives to serve three-year terms.  Thus,  one-third of the
Town Meeting must face re-election every year.   The annual Town Meet-
ing which considers appropriations and other business for the upcoming
year is held on the fourth Monday after the election.  Other special meet-
ings of the Town Meeting may be called when necessary during the year.

The executive power of this form of government resides in the office of
selectmen.  There are three  in Athol,  and they are  elected at the same
time as the  representatives to the town meeting,  in  non-partisan, at-
large elections.  These selectmen have responsibility for  the Police
Department, the Fire  Department, and the general administration of
Town business.   In actual practice,  the selectmen spend their time sign-
ing warrants, issuing  permits and licenses,  meeting citizens who have
problems with the town, and appointing people to office.  The selectmen
have full power to make appointments without consent of the Town Meet-
ing.  This appointment power  is not so great as  it might seem,  since
many of the key appointments  are actually made by special boards or
commissions.  For  example,  the following elected boards have full res-
ponsibility for the town business which is  within their authority; Board
of Public Works, School Committee, Planning Board,  Board of Health,
Library  Trustees,  and the Housing Authority.  As one selectman put it:
"The people just don't understand how  limited our authority is.   People
come to  me with problems concerning  streets and roads,  and alii can do
is go to the  Department of Public Works and  ask them to do something.
I have  no authority over that department.   It  belongs to the Public Works
Commissioners. "

The selectmen also have reponsibility for drawing up the  "warrant",
which  is essentially a  list of proposed  appropriations and policies to be
                              214

-------
passed upon by the Town Meeting.   The warrant is divided into articles,
each consisting of some appropriation of town funds or some action to
be taken by the town.

At some unspecified time before the Town Meeting, the selectmen are
supposed to present the warrant containing their proposals to the Town
Meeting's Finance  Committee,  composed of 15 members selected by the
Town Meeting who  serve for three-year terms.  The Finance Committee
then holds hearings on these proposals and publishes a leaflet containing
its recommendations.  In essence,  the Finance Committee acts as an
advisory agency, recommending its findings on all budgets or articles
placed before the voters at the  Town Meeting.  In actual practice, how-
ever its power is much greater because, as most respondents pointed
out,  . . . .the Town  Meeting almost always does what the Finance Committ-
ee recommends.   (In 1969, for  example, there were 75 articles: 66 of
them passed unanimously; one passed after a tally; and eight were defeated. )
The warrant and the  recommendations of the Finance Committee are
both made public before the meeting; the selectmen are required by the
By-Laws  to post the  warrant in three public places at least seven days
before the meeting.

History of the Sewer Issue

The history of the  sewer issue  in Athol dates back to the 1940's, when
the Commonwealth of Massachusetts and local conservationists  began to
push to clean-up of the Millers  River which flows through the town.   In
1947, the Federal  Housing and  Home Finance Agency (HHFA) financed a
study of the sewer  problem.  Athol received a $10, 850 loan toward the
design of  a treatment plant.

In spite of this incentive, the issue lay dormant for the next twenty years.
The  HHFA frequently inquired about the  status of sewers in Athol (and
about the date for repayment of the $10, 850 loan),  but the town did nothing.
                               215

-------
Local residents tried to stimulate interest.  Conservationists kept the
issue in the minds  of the voters,  and the Athol Daily News editorialized
in favor of sewer improvements several times. Still there was no interest.

Whatever sentiment there was in favor of the treatment plant was  effec-
tively neutralized by opposition to these  improvements.  The most popu-
lar argument against the improvements was  that there was no reason for
Athol to clean up its part of Miller's River if the upstream polluters of
the river failed to clean up their part.   Many local residents felt that the
effluent from paper mills upstream was  responsible for the pollution.
They argued that if the water was dirty when it came into Athol, it would
be dirty when it left,  no matter what Athol did  to it.  Closely related to
this  argument was the feeling that the demands by the Commonwealth of
Massachusetts for Athol to construct a treatment plant were arbitrary
and discriminatory as long as other polluters of the river were not also
compelled to construct treatment plants.  As a result,  no action was
taken on the  sewer  issue during the 1950's and  1960's.
In 1965, the selectmen appointed a special Sewage Treatment Plant
Committee,  headed by a  selectman and former State representative.
The committee proposed that the town construct a treatment plant at a
cost of $1,422, 000.  It further proposed that the Federal government
pay about 30% of the total cost, and that the remainder (approximately
$1, 000, 000) be financed by the town.   The Finance Committee, in its
recommendations to the representatives at the Town Meeting, rejected
the report of the Sewage  Treatment Plant Committee on the grounds
that the plant should not be financed from taxation.  (This was part of a
larger report made to the Town  Meeting by the Finance Committee
which concentrated,  in large part,  on the rising tax rate in the town).

The selectmen, who were still in favor of proceeding with  the treatment
plant, went to the Board  of Public Works (composed of three elected
                               216

-------
commissioners) and requested a trebling of sewage user charges  (from
$5. 00 to $15. 00).   The selectmen intended to use the revenue from the
increased user charges to finance the revenue bonds for the  construction
of the treatment plant.  After 6 months delay, the Board of Public Works
agreed to the increase.

In 1969,  the selectmen proposed again the authorization of construction
of the plant.  The wording of the proposal was as follows:

      ARTICLE 30.  To see if the Town will vote to  raise and
      appropriate a sum of money for the  construction of a
      Sewage Treatment Plant, Pumping Station, Intercepting
      Sewers, related facilities and appurtenances,  authorize
      the Board of Public Works, and/or the Board of Selectmen,
      to acquire land, rights of way, or easements,  to provide
      whether the required funds shall be  raised by taxation,
      transferred from available funds, by borrowing,  or by a
      combination of these  methods,  and apply any Federal or
      State aid and grants of funds to reduce the cost to the
      Town, pass any vote  or take any action relative thereto.

This time, with the construction financed  by revenue bonds instead of
general obligation bonds, the Town Meeting approved the proposal.
 Conclusions

 Two questions about the sewer  issue in Athol are significant.   Why did
 Athol wait until 1966--nearly 20 years after the first sewer study--to
 consider the sewer bond issue? Why did Athol pass the sewer bond issue
 in 1969 after defeating it in 1966?

 In.answer to the first  question,  it would appear that the arguments against
 building the treatment plant were sufficient to convince the residents of
 Athol that they ought not to take immediate action on the plant--even
 though they were receiving pressure from  State and Federal agencies.
 The following arguments  seem  to have been most important:
                               217

-------
            There was no reason for Athol to act alone while
            other polluters  of the Millers River were allowed
            to continue their pollution.
            Action against Athol was arbitrary and unreasonable
            so long as the Commonwealth did not act against
            other, polluters  of the Millers River.
            The Metropolitan District Commission was going to
            use water from the Millers  River  in the Quibban
            Reservoir, and when it did so, it would construct,
            and pay for;  a treatment plant.
            It would be more efficient to build a single treatment
            plant to serve both  Orange and A.thol--and share the
            cost between the two towns.
            The cost  of the  treatment plant was too high.

It is interesting to note that none of these arguments suggest that the
treatment plant was not  needed, nor  that the plant would not be good for
the  town of Athol.   Indeed,  the  town  seemed to be agreed,  at least pub-
licly, on the need to abate pollution of the Miller's River.   But the voters
disagreed about when the project should be undertaken.  Furthermore,
there were many residents  who felt that if the main issue were pollution,
Athol should not act until it was clear that they had chosen the most
efficient and effective means of ending pollution.   Another factor which
may have contributed to delay, (although it was not so publicly expressed)
was the desire of many  residents to  minimize town expenditures.  If the
need for a treatment plant at this place  and at this time were not ab-
solutely clear, then perhaps the town ought not to make the  expenditure
until the need was clear.  In 1966, the issue was apparently clear enough
to the selectmen to justify a proposal in the warrant; but it was apparently
not clear enough to the Town Meeting to justify passage.

Passage of the bond issue in 1969 is  attributable primarily to a change in
financing from general obligation bonds to revenue bonds.   There were
other financial changes as well. For example, the proposal in 1966 in-
cluded only  30% financing by the Federal government.   By 1969, the
                              218

-------
selectmen estimated that 75% of the cost would be paid for by the Com-
monwealth or the Federal government.  Also, interest rates, which were
very high in  1966, had lowered.  In addition, by 1969 State pressure  on
Athol to build the plant had increased to the point where it had an impact
on the residents.  Although there is no way to measure successfully the
pressure  that was being exerted,  it seems  clear from conversations  with
respondents  in Athol that the townspeople felt in 1969 that the demands by
the Commonwealth could no longer be resisted.

These conclusions were offered by most  respondents in Athol; four addi-
tional features of the case seem to be important.  First, this was a case
where State interference in local affairs  was apparently successful.  Al-
though the residents resented State orders  to clean up the Millers River,
and felt discriminated against by the State agency, they did comply when
the State pressure was strong enough.  This was not always the case  in
other communities that  were studied, for example Avon,  Ohio.  However,
there was no opposition  to sewers  in Athol--just opposition to the timing
of the sewers.  In Avon,  by contrast, the argument was over growth  and
development, with  sewers representing development, and there were
strong partisans on both sides.   This suggests that if there  is no opposi-
tion to sewers per se,  and merely differences of opinion over the timing
of the action, State pressure may provide the necessary incentive for
immediate action.
Second,  Athol is a clear case of a sewer issue being defeated because of
its means of financing.  Clearly, citizens are not oblivious to these im-
portant financial matters, and a change in financing can mean success
for the  sewer issue.   Notice also that the change was from general obli-
gation bonds to revenue bonds.   This is consistent with our finding that
the voters prefer bonds repaid by user charges to those repaid by in-
creased taxes.  The history of the vote  indicates that the citizens of
Athol were especially concerned about taxes:  The Athol Town Report for
1966 devoted considerable of attention to the tax issue and the Finance
Committee presented  a special report on taxes at the beginning  of the
Town Meeting.

                              219

-------
The third,  and most interesting feature of this case is the fact that the
issue was passed by a Town Meeting and not a referendum.  The number
of representatives voting on this issue was only 180,  out of a population
of nearly 12, 000.  However, the important difference between a referen-
dum and a  Town Meeting, does not depend entirely on the number of
persons  voting.  More important is the way in which  the vote is taken.
In the referendum,  voters consider the question alone,  by secret ballot,
in a voting booth.  In the  Town Meeting,  voters confront one another
over the issue, discuss and debate the issue, and vote in full view  of one
another.  These conditions,  combined with the fact that fewer voters are
present at town meetings, have a definite effect on the outcome  of the
vote.  Due to psychological pressures, the voters  are more apt to  be
influenced by each other.  Thus, it is doubtful that action would have
been defeated so decisively in 1966 had a referendum been held.  On the
other hand, it is equally questionable whether the final vote in 1969
would have been unanimous, or even passed at all  if it had gone to  a
r ef e r endum.

One reason for the unanimity of the vote  (according to many of our res-
pondents) was the fact that the  Town  Meeting nearly always goes along
with recommendations of the Finance Committee.   The  record seems
to indicate that the  Town  Meeting agrees with the recommendations of
the Finance Committee when either of two conditions  prevail: there is
no serious opposition to the recommendations and  the Finance Committee
is unified in its recommendations.  A closely related phenomenon is
visible in many of the other cases under  studying:  the existence of  a
"united front" of elected community officials makes it much easier to
gain community acceptance of a public improvement.
                                220

-------
AVON ,  OHIO
   221

-------
                                AVON
This bond issue was submitted to the voters of the Town of Avon, Ohio,

on November 8,  1966.  The wording of the proposition was:


      "Shall bonds be issued by Avon City, Lorain County, Ohio,
      for the purpose of paying part of the cost of constructing
      sanitary sewerage collection facilities including the acqui-
      sition of the necessary real estate and interests in real
      estate in the sum of ($800, 000) eight-hundred thousand
      dollars  and a levy of taxes be made outside of the ten mill
      limitation estimated by the county auditor to average
      2. 51939  mills for each one dollar of valuation which amounts
      to  twenty-five and 1939/1000 cents for each one hundred
      dollars  of valuation for a maximum period of thirty (30)
      years to pay the principal and interest of such bonds? "


The project to be constructed with the proceeds of these bonds (plus the

proceeds of special  assessments and mortgage revenue bonds) was a new
sanitary sewer system.  The vote on the referendum was 751  (44%) in
favor, and 967 (56%) against.  The referendum was defeated.


Population


The primary governmental unit involved in this referendum was Avon,
Ohio.  Avon is a city of 7, 000 people, located west of Cleveland, and
four miles south of Lake Erie.  It has traditionally been an excellent
farming  area,  and its truck farms once supplied produce for the residents
of Cleveland.   Farming is no longer profitable,  however,  and only a few

of the truck farms still exist.  The proposed extension of Interstate 90

through the northern part of the  city in an east-west direction and the

proposed extension of State Route 76 through the center of the city in a

north-south direction would seem to make it a prime  candidate for sub-

urban residential development.  .Such development has not occurred.


This is not to say that there has been no growth at all.  In 1950,  the
                              223

-------
population was only 2, 773.  By I960,  it had grown to 6, 000.  And in 1970,
according to preliminary Census data, it was 7, 137.  The physical loca-
tion of this growth can be located easily on the map on the next page.
It occurred primarily in two areas: first,  in the area surrounding the
junction of Detroit Road and Colorado Road, and second,  in the area east
of Jaycox Road and south of Detroit Road.  The rest of the city remains
primarily farmland,  except for the area north of,  and immediately adja-
cent to, Chester Road, which is marshy and unsuitable  for farming.

The rapid population  growth during the 1950's brought few measurable
changes in the socioeconomic characteristics of the population.   For
example,  there  were  no non-whites living in Avon in 1950; there were
none in I960;  and there apparently  were none in 1970.  The same can be
said for Mexican-Americans and Puerto Ricans, although one resident
said that "...  a few of. them [i. e. , Puerto Ricans] may have  been able
to infiltrate during the past couple  of years. "  (Puerto Ricans from
Lorain  are now being used as laborers on the truck farms in Avon. )  One
consequence of the growth in population during the 1950's was a decrease
in the median age of the population.  The influx of young families during
the 1950's has also been reflected in an increase in the  school population
in Avon during the  1960's, requiring the construction of an additional
elementary school  and additions to other schools.  The median age of the
population has remained relatively stable since I960, when it was 25 years

These data do not fully explain the  character of population growth in Avon
during the era 1950-1970.  According to local residents,  the in-migra-
tion to the city occurred from two direction, with two different types of
migrants.  From Lorain came Eastern-European working-class families
who wanted to escape the city and buy a small home in a relatively rural
area.  From Cleveland came younger professional and technical people,
with no  particular ethnic connection, who wanted to live in a  rural or
suburban setting.   The in-migrants from Lorain settled along Jaycox and
Nagel Roads and in the Northgate development on Colorado Road.

Local residents  disagree about the current ethnic composition of the
                             224

-------
CO
ts)
Ul

-------
Avon population.   There is general agreement that the first two ethnic
groups in the city were the English (who were Protestants and became
Republicans) and the Germans (who were Catholics  and became  Demo-
crats).  Many residents contend that there are no other significant ethnic
groups in the community.  Others argue that the newer migrants from
Lorain were Czechs,  Hungarians and Poles  and that these groups now
make up a substantial part of the population.  They  also point out that
this has great significance for the politics of Avon,  since these  Eastern
European migrants were associated with the Democratic machine of
Lorain.  According to the I960 Census,  there were 1286 persons in Avon
who reported themselves as being foreign-born.   Of these,  556,  or 43%,
were from Czechoslovakia,  Hungary  or Poland.  Only 334 were from
Germany  and England.  Eastern Europeans comprise at least 13% of the
total population.

Residents of Avon often claim with pride that their city has  the highest
level of income in Lorain County and one of the highest in Ohio.   Although
there is no evidence to  confirm or deny this  assertion,  it is clear that
the median income in Avon ($7, 215 in I960) would lend  some substance
to these claims.  Moreover,  in I960,  only 6% of the families in Avon had
incomes of $3, 000 or less, and 29% made $9, 000 or more.  The median
family income for urban areas of Ohio in I960 was $6, 442,  and for
Lorain County in I960 it was $6, 241.

This should not suggest, however, that Avon is a white-collar suburb.
In terms of occupation classifications, 51% of all employed  males are in
craft or operative occupations.  Laborers comprise another 8%,  bring-
ing the blue-collar labor force to 60% of the  total.   Professional and
technical  and sales account for only 20% of the population.   Managerial
occupations account for 14%,  but many of those in this category are
farm operators.

Housing in Avon consists primarily of single-family,  owner-occupied
units.  Of the 1534 units occupied in I960, 1337, or 87%, were owner-
occupied;  the remainder were rental  units.   Of the 1661 units which
                              226

-------
were classified as single-family or multi-family,  1530 (95%) were
single-family.   The median value of an owner-occupied home in Avon
in I960 was $18, 800, as compared with $14, 000 for Ohio urban areas.
In I960,  62% of all Avon residents had been living in their current hous-
ing unit for six years or less.

Economy

The economic base of Avon is currently rather weak.  The truck farms
which were once the mainstay of the economy have ceased to be produc-
tive, and the community has been slow to develop  an industrial base.
                                                        35
According  to a city bond prospectus dated August  15, 1969  , there were
only four industries in Avon in 1969, employing a  total of 220 persons.
It is clear  that many residents must travel out  of the city--to Lorain,
Elyria, Avon Lake or Cleveland--for employment.  The city's real
estate  value is composed of 78. 5% residential,  11. 5% commercial, 8. 0%
agricultural, and 2. 0% industrial property.  A  combination of commer-
cial, industrial and public utility values accounts for 35% of the total
assessed valuation of property in Avon  (which was $19, 401, 530  in 1968-
69). Per capita assessed value is $2,436, compared with $3, 359  for the
203 cities in Ohio.

Avon's record  in attracting industry is particularly poor in comparison
with its neighbor to  the north, Avon Lake.  Avon Lake was part of Avon
until 1917, when there was a  dispute over the creation of a municipal
water system.   Those in the northern part of Avon at that time wanted
the system, and they became a separate local government in order to
advance  this project.  Avon Lake has continued to  lead Avon in the pro-
vision  of certain public services, having established a municipal sewer
system during  the 1950's.

Some persons argue that Avon Lake's advancements in the provision of
water and sewer systems explains why it, rather than Avon, has been
able to attract  industry.  This argument,  however, is subject to much
debate.-  Whatever the reason, the facts are unmistakable.  Even though
                              227

-------
Avon consists of more than 4, 000 acres, only 2, 149 acres are developed.
The area north  of Chester Road has been a potential industrial park for
many years, but still stands undeveloped.  This  stands in contrast with
Avon Lake, where a solid industrial base has enabled that city to main-
tain a high level of public services while levying a smaller property tax
on its residents than does Avon.  (See Table 1 , below. )
            Table 1 .  Comparative Economic Development of
                         Avon and Avon Lake
Number of People
Annual Real Estate Tax Revenue
Number of Full-time Firemen
Number of Full-time Policemen
Students per Teacher
Real Estate Tax Millage
    Avon
   8, 200
$917,603
        0
        3
      27
   47. 25
 Avon Lake
     12,800
$4, 132,475
         13
         13
         22
     44.77
The financial position of the government of Avon is closely related to its
economic base.  As the data in Table 1 show, Avon is in a substantially
inferior position to its neighbor, Avon Lake, in terms of quantity of
services and tax rate.  On the other  hand,  in comparison with Avon Lake
and its other neighbors,  Avon's tax rate is about the same.   (See  Table 2. )
In addition,  the  city felt enough need for additional revenues  to adopt a
one percent city income tax in 1968.

Avon is governed by a weak mayor-council form of government.   The
city council is composed of seven members, four elected by wards and
three  elected at-large. The mayor and all members of the council
serve for two-year terms and are elected on a non-partisan basis.  The
city council has  substantial authority to overrule the mayor on important
policy matters,  and all of the mayor's appointments must be confirmed
by a two-thirds  vote of the council.   On the other hand, there are some
parts  of the charter which are to the  advantage of the mayor.  Of  particu-
lar importance is the fact that the mayor can veto any act of the council
                               228

-------
as long as there is one dissenting vote on the council.
Table 2. Tax Rates
City
Avon
Avon Lake
Sheffield
Sheffield Lake
North Ridgeville
Elyria
of Avon and Neighboring
1965 Tax Rate
41. 69
40. 15
49. 60
50. 39
40. 07
40. 91
C ommunitie s
1969 Tax Rate
51. 75
46. 67
53. 97
67. 25
50. 68
50. 58
Avon has recently decided to participate in the French Creek Council of
Governments,  a regional federation of cities along the French Creek
which was established to provide for regional planning and to facilitate
the development of cooperative arrangements for the provision of
services.  The Council of Governments (COG) has been strongly suppor
ted by the mayor of Avon since its inception.  One of the cooperative
projects proposed by the COG is a French Creek sewer system.  Avon's
possible participation in this proposed regional system was a major
issue in the recent sewer bond elections.
                               229

-------
History of the Sewer Issue

The history of the sewer issue in Avon is important for an understanding
of recent bond elections.  In 1955, a developer built a housing subdivi-
sion known as Northgate north of Detroit Road on Colorado Road.  The
developer also provided sewers for these new homes and built an
88, 000-gallon treatment plant to treat the sewerage. Some observers
claim that from the  beginning this plant was plagued by faulty workman-
ship in  construction.  They also claim that the system must process a
much greater volume of liquid than it was designed to handle, because
of recent population in-migration. -Records of the treatment indicate
that it is flooded almost every year and inoperable for brief periods of
time.  They also show that the plant often handles over 100, 000 gallons
a day,  although  its capacity is  only 88, 000 gallons.   The Ohio Depart-
ment of Health evaluated this system in January, 1970,  and rated its
performance 13  on a scale of 1 to 100.   Any score below 40 was  consi-
dered a "complete failure of the  system to meet its intended purpose.  "

The Northgate plant has never  been adequate for the needs of the resi-
dents of Northgate,  and, in fact, it became a major  source of pollution
soon after  it  was built.   The failure of the Northgate plant was used
thereafter  by §tate and Federal agencies as  a reason for sewering the
entire city.  Complaints from the Ohio Water Pollution Control Board
about the Northgate  plant led the city, in the late 1950's, to a study of
the problem.   A loan of $12, 600  from the Federal Government in I960
paid the necessary engineering fees.

Armed with a study  and plans for sewer construction, the mayor
attempted to  get the issue on the ballot so that bonds could be issued.
The city council, failing to see the need for sewers  and unwilling to
propose a raise  in taxes to finance sewer construction,  took no
action on the  matter until 1963.   In a special election on May 7, 1963,
a proposal to increase the millage by 1. 8523 mills to raise $500, 000 for
sanitary sewers was defeated by a vote of 320 to 663.  The plan for the
                              230

-------
proposed sewer system is shown in Figure 1 .  The plan emphasizes the
use of gravity flow--at least for the northern half of the  system.  The
only place where gravity  flow was not to be utilized also became one of
the most controversial aspects  of the system—a  12-inch trunk line
across farmland in the southeast corner of the city.  Although the bond
issue was for only $500,  000, the total cost of the project was much
more.  Although the records are incomplete,  it is likely that the remain-
der of the cost was to be  raised by assessment or by mortgage revenue
bonds.

On November 3,  1964, the  sewer issue was placed on the ballot again.
The clerk's records indicate that the city proposed an increase  in taxes
of 2. 0 mils  for five years for "sewer improvements. "  This was a
straight levy; there were no specific sewer plans.  The issue failed,
with 911 voting in favor and 999 opposed.

In 1965, the city placed on  the ballot a proposed sewer bond issue of
$300, 000, to be paid  for by an increase of 1. 0387 mills in the tax levy.
The 1963 plan, shown in Figure 2 , was used again as the basic  plan for
the system, but the amount of construction to be  paid for by this bond
issue was smaller than in 1963. This  was also an election year for city
officials.  One of the candidates for mayor ran on a  platform of opposi-
tion to sewers,  charging  that the plans called for the public to pay for
capital improvements that would benefit special interests who wanted to
develop certain areas of the city.  The referendum was defeated on
November 2,  1965, by a vote of 794 in favor  to 1036 against.  In the
same election, the anti-sewer candidate was elected mayor.

In 1966, the city council,  led by a political rival  of the new mayor,  made
arrangements with the engineer of the  1963 sewer plan.  The newly
elected mayor,  who was not in full control of the  city government,  was
excluded from these arrangements.  The  sewer plan selected by the
council was essentially the  same as the 1963 plan.  Although he officially
recognized the need for sewer  construction,  the mayor  repeated his
                              231

-------
    .,    .

   CHSSTp.	r


  /^733o7vfeiT.     I

  i^^^d^k, ^o.rn
  >)	" I"; ;; O
      -           "
                    .
   /     .     ycif^
Existing  '%K /f^2^==
Package'    ^^
ge     /
-------
charges about benefits to special interests.  The total cost for the
general obligation bonds part of the 1966 proposal was $800, 000, and the
increase in the tax levy was to be 2. 51939 mills.  The issue  was defea-
ted by a vote of 751 in favor to 967 opposed on November 8, 1966.

In 1967,  the mayor began to investigate the possibility of a regional
sewer system.  He was convinced that the regional idea would save the
city money and encourage the investment of Federal and State funds in
the project.  The council did not agree.  Although there was no plan for
sewers in  1967, the council decided to try a new tactic  for presenting the
issue to the voters.  On the ballot was a question which asked the voters
to approve an increase in taxes for  unspecified "purposes other than debt
service.  "  Part of the money was to have been used for sewers.  The
election was held on May 2, 1967,  and was defeated by  a vote  of 644 in
favor to 720 against.

The next important event in the history of sewer elections in Avon took
place on December 10, 1968, when the Ohio Water Pollution Control
Board imposed a building ban on the city.  The ban would be lifted,
according to the OWPCB, only when Avon built a sewer system.  Impo-
sition of the building ban was the latest in a long series of disputes
between Avon and the OWPCB, beginning with the  complaints by that
Board about the defective Northgate sewerage treatment plant many
years previous.

In 1969,  a new sewer plan was introduced by the city council.   This time
the council had a local engineer draw up a plan  for a sewer system.
This plan is found in Figure 2 .  This  sewer plan was the most ambitious
ever proposed in Avon.  Although it did not contain the  controversial line
south of Schwartz Road,  it  did contain two equally controversial items:  a
line across the north part of town,  above Chester  Road, and a line down
sparsely-populated Stoney Ridge Road.  The first  of these additions was
clearly designed to attract  industry, since there is little population that
far north in the city.   Because of its possible influence on industrial
                               233

-------
to
                                                                                                                  .jj  '    MS    17-C-rry
                                                               Figure  2.   Shafer Plan

-------
development of the city, the chamber of commerce played a larger role
in this campaign than it had in any other.

One possible savings to be  derived from the 1969 plan lay in its complete
reliance on gravity flow,  which  eliminated the.cost of constructing lift
stations  (although there was a possibility that it would require deeper
pipes, which might add cost).  The total cost of the project was set at
$4, 089, 645.  To raise  this money, the council specified the following
sources  of money: special assessments--$l, 590, 000; Federal grants--
$399, 645; general obligation bonds--$1, 200, 000; and mortgage revenue
bonds paid for by sewer rental charges--$900, 000.  Thus, the voters
were  to consider a $1,  200, 000 bond issue to be financed by a  5. 2 mill
increase in taxes.  The voters went to the polls on June 24, and rejected
the issue,  662 in favor to 992  against. As in other years, the man who
was now their mayor led the opposition--this time in the form of a
Citizens' Committee on Sewers.

In 1970,  the mayor finally got his sewer plan on the ballot.  The most
important aspect of this new plan was participation in the French Creek
system.  (See Figure 3 . )   Unlike earlier  plans, the French Creek Plan
does not attempt to sewer  sparsely-populated areas; only populated
areas of the city would be  served.  This plan could make use  of gravity
flow,  thereby necessitating the construction of a lift station on Jaycox
Road.   The other notable  feature of the plan was its reliance  on the pro-
posed French Creek Interceptor. Indeed, the plan proposes the
struction of only a collection system for Avon; treatment would presum-
ably be  provided by the COG.  These  changes  help to  explain the lower
cost of the 1970 proposal:  $1, 738, 350.  Of that total,  $1, 159,  820 was to
be paid by  special assessments,  and $578, 530 was to be paid by general
obligation bonds.  However, the  amount of the bond issue proposed to the
voters was $900, 000.   The reason for this is a ten-mill limitation (in the
Ohio Constitution) on taxes levied without  a vote restricting the amount
of special assessments which can be levied.  The maximum amount of
special assessment bonds that Avon could issue  was $846, 000.  Thus,
                               235

-------
ro
UJ

/ J

i sF'jVAc.r T -•• -r
PL* N T ( A'.T'L, K KW"rk J


c
o
a
-- i
I
J
c>
a:
1

I
£!•
0
a:
                        PROPOSED  FOR IMMEDIATE CONSTRUCTION BY AVON CITIZENS SEWER COMMITTEE
                              E CONSTRUCTION  BY FRENCH C3EEK COUNCIL OF GOVERNMENTS
                                         Figure 3,  French Creek Plan

-------
the remaining $313, 820 to be financed by special assessment must be
included in the voted bond issue.  The total for general obligation bonds
plus the  remainder of the  special assessment bonds is $892, 350, roun-
ded- off to $900, 000.   In order to pay for these bonds, a tax increase
of 3. 9 mills was proposed.  In spite of the mayor's support, the plan
was defeated in an election on May 5, 1970 by a vote of 531 in favor to
703 against.

The subject of this case study was the bond election which took place in
1966.  However, an understanding of this election depends upon an under-
understanding of all seven of the elections which have taken  place during
the last eight  years.   Therefore,  for purposes of analysis, the discur-
sion will not be limited to the 1966 election alone.

The campaigns which  preceded the elections varied widely in their con-
tent and intensity.  In 1967, for example,  the issue was hidden in a gen-
eral purpose tax levy.  In 1969, by contrast,  there was an extensive
campaign.  Although the voters of Avon had already considered five pre-
vious sewer elections, and presumably were  aware of the essential
issues,  the  proponents of the bond issue produced a leaflet which
explained the  issue in extensive detail.  Leaflets, newspaper stories,
and advertisements were used in favor and against the issue.  There
were public debates and guest speakers; the chamber of commerce
offered a free dinner to people who  were interested in hearing a speech
about the need for sewers (300 people attended); and proponents  put up
pro-referendum posters-

The participants in the campaigns varied somewhat over the decade, but
during the last five years, they narrowed down to two main groups.  The
proponents had been led by the  chamber of commerce and the opponents
had been led by the mayor, until the last election when the mayor was
able to propose his own sewer plan.   One characteristic of all the cam-
paigns,  including those in the early 1960's, was the division of the city's
elected  leadership on the issue.  In 1963 and  1964, the mayor was in
                               237

-------
favor and the council opposed; in 1965,  1966,  1967 and 1969,  the council
was in favor and the mayor was opposed; in  1970,  the mayor became a
proponent of sewers only to find opponents to sewers on the council.

The issues in the campaigns changed somewhat from the beginning  of the
decade.  In 1963, for example, neither side  put on much of a campaign.
By 1965, it was clear that those in favor of sewers would have to sell the
issue to the voters,  and they began to develop arguments in favor of their
project.  They focused  on the failures of the Northgate treatment plant
and the ensuing complaints from the OWPCB and on the benefits of  indus-
trial development and concommitant growth in the  city's tax base which
would presumably be derived from sewers.  By 1969, the proponents had
added the issue of cost, arguing that  sewers were  inevitable and that the
cost was constantly increasing.  (Their slogan in 1969 was, "Sewers will
cost more tomorrow. ")  The fact that the OWPCB  placed a building ban
on Avon in December,  1968 was another argument used to support the
bond issue.

In 1963 and  1964, the opposition was just as  inactive  as the pro-sewer
forces.  By 1965, the mayor  was  raising the issue which would be the
basis of his anti-sewer  campaign and his political  career: private inter-
ests  would be the major beneficiaries  of the  public sewer project.   In
1967  he added the issue of the advantages of a regional, as opposed to
strictly local,   sewer system.

In 1970,  with the mayor in favor  of sewers,  the opposition was less
vocal than in the past, aid less focused on the important issues.  Oppo-
nents concentrated  on issues  of cost,  reasons for  not cooperating with
the regional system,  and popular  resentment of State and Federal pres-
sure in favor of sewers.

Since city politics in Avon are formally nonpartisan, the sewer bond
elections were supposed to be free from partisan politics.  For the most
part,  this was  true, but active participants in these campaigns tell a
                              238

-------
different story.   They say it was no accident that the proponents  of
sewers used the  chamber of commerce as the organizational base for
pro-sewer activities and that the  chamber of commerce was primarily
a Republican organization.  One of the most active proponents of sewers
was the Republican nominee for state representative from the Avon area
for the 1970 election,  and his brother, a member of the city council.  On
the other side, the mayor,  who was the chief opponent of  sewers, was
the Democratic committeeman for Avon.  His brother was Lorain County
auditor and a member of the Lorian City Democratic machine. The
rivalry between these  two political factions was more than a simple par-
tisan dispute.  On many occasions during the past five years,  it became
so bitter that it could more accurately be described  in the terms  used by
some of the respondents as a "feud"  or "vendetta. "   The  mayor  accused
this powerful Republican family of trying to exploit public works  (like
sewers) and public office for private gain (by the  family-owned business).
The mayor's opponents denied these  charges and  added that the mayor
was the representative of a small group of wealthy landowners who
wanted to thwart the development of Avon because they didn't want
increases in their taxes.  The form of the dispute took many forms.
             In 1966, when this dominant Republican clan tried to
             obtain the necessary zoning changes for a shopping
             center  on their property,  they had trouble with the
             Avon Planning Commission (and particularly two
             mayoral appointees on the Commission).  The rival
             clan then called for the resignation of the mayor's
             appointees,  charging them with a "conflict of interest. "
             In 1968, two months after the formation of the French
             Creek Council of Governments (in which the mayor had
             played  a leading role), the Avon City Council (led by a
             member of the  opposition) voted to abandon the COG's
             sewer plan in favor of the city's own sewer system.
             On still another occasion,  the mayor charged his Repub-
             lican rival on the  city council with a "conflict of inter-
             est, " alleging that he  was using  his office of city coun-
             cilman to advance  the interests of the family-owned
             business;  the city council promptly censured the mayor
             for his  allegation.
                               239

-------
            When the sewer plan of 1969 was conceived, the
            mayor was not included in the planning process,
            even though the engineer  who presented this scheme
            had been appointed to the position of city engineer
            by the mayor himself.  It was held by some that the
            city engineer was at this  time employed by the local
            company owned by the mayor's political rivals.

Interpretation of Outcome

There were seven major reasons cited by our respondents  for the failure
of the sewer bond referenda in Avon.   The single most important reason,
and the  reason which was cited by every respondent, is that the  majority
of the residents of Avon simply did not want sewers.  This is not because
they favored septic tanks, but rather because they were against  the land
development that they sawas the  natural consequence of sewer construc-
tion.  There was a great deal of discussion in Avon  of the "selfish inter-
ests" who would profit from the construction of sewers and of the  "out-
side developers" who would come into Avon and change the character of
the town if there were sewers.  Many  of the residents are retired far-
mers or people who have  recently escaped the city and who like  the semi-
rural atmosphere of Avon.  They feared development would change all of
this,  and preferred that the town remain  unchanged.

A  second reason cited by  some of the respondents was the unwillingness
of Avon residents to tax themselves for public expenditures in general.
Although there were serious  differences of opinion on this point,  some
respondents argued that it was difficult to pass any referendum in Avon,
with the possible exception of those for school bonds.

A third  reason cited by many respondents, both proponents and opponents
alike, was  that the residents of Avon  resented the pressure being placed
on the community by the Ohio Water Pollution Control Board (OWPCB).
Many residents felt that they were being singled out by the  OWPCB even
though they were not serious polluters of Lake Erie (particularly in com-
parison with Cleveland),  and that the  reason for the OWPCB1 s action was
                               240

-------
that the OWPCB was being paid by developers to do this.

Several respondents argued that the most important reason for failure
was that there was "no united front" put forward by the major  elected
officials of the community.  That is, in each referendum there was some
elected official--either the mayor (from 1965 to  1969) or some other
member of the council--who would opoose sewers.  This legitimized the
opposition of many undecided residents who believed that if elected offi-
cials  were undecided,  then there must be  some doubt about the worth of
the project.  (It is interesting to note that both the mayor and his fore-
most  political opponent cited this reason in almost identical language.)

A fifth reason is the charge that the public was misinformed on the issue
of sewers.  Persons supporting  sewers often  suggest that if the people
only knew the real issues  or only understood the finances or engineering
of sewers, some of their  "irrational" opposition to sewers would disap-
pear.  Closely associated with this argument  is a suggestion that better
publicity might have eliminated the ignorance of the  people and helped the
campaign to succeed.   There is  some difference of opinion on this  point,
however, since some  others argue that the most effective campaign was
the 1967 campaign when the issue  of sewers was most obscured.

A sixth reason mentioned  by several respondents is the partisan politics
associated with the issue.  Some blame land developers,  particularly
the Republican power  structure, for the bad image of self-interest crea-
ted by their support of the referenda.  Others blame the mayor for
making political gains out of opposition to the pro-sewer forces.   Regard-
less of who is to blame, the issue was a highly partisan one, at least  in
the sense of there being two clearly-defined groups with positions on the
issue  who opposed one another for several years.

Finally,  one  respondent suggested that one reason for failure of the last
couple of referenda is the  fact that people in Avon were tired of the sewer
issue.  They had  been debating it for  almost a decade and had defeated
                               241

-------
it every time it had been offered.   Offering it again only antagonized
people who had made up their minds long ago.

Our next task is to evaluate the reasons offered by our respondents for
the failure of the bond elections.  The fact that so many of our respon-
dents—from both sides of the is sue--mentioned several  of these reasons
for the failure of the  referenda lends some weight to their explanatory
value. Although there is no  survey data to substantiate the  claim that
the residents of Avon are opposed to  development of their community,
authoritative observers of political behavior in the city seem to agree
that this  is the  case.

Also, there is little  doubt that some local residents  resent the OWPCB
pressure in favor of sewers.  Indeed, the actions of the  OWPCB could
be described as dysfunctional in this  particular city.  The building ban
imposed  by the OWPCB hurt no one in Avon except the developers and
realtors  who were the major supporters of the  sewer issue.

However, there is evidence with which to evaluate some  of the other
arguments. The results of several elections,  recorded  by precinct,
provide some useful insights.  Table 3  contains the votes, by precinct,
for the 1965, 1966, and 1967 sewer bond elections;  Table 4  contains the
vote  for 1967 charter amendments; and Table 5 contains  the vote for
mayor in 1969.  Some respondents suggested that the sewer issue had
become so politicized, and so associated with the feud between the mayor
and his political rivals, that a citizen's attitude on the feud  may have
determined his position on sewers.  Table 6 shows the vote, by precinct,
on three  separate  issues during the period 1967-1969.  These data seem
to  indicate a correlation between a vote for the  mayor  and opposition to
the sewer bond election*  Moreover,  these two  phenomena are also
related to a vote against the  1967 charter amendments.  This correlation
does not mean that a citizen's attitude toward the mayor  determined his
vote  on sewers (or vice-versa), but it does suggest the existence of two
voting blocs in  the city which voted against one  another on several
issues including sewers.
                              242

-------
ts)
^
uo
Table 3. Summary of Recent Sewer Issue Elections
Precinct
11965 Vote
1965 Vote as % of 1965 Vote
Number of Yes Votes in 1965
% of 1965 Vote Voting Yes
21966 Vote
1966 Vote as % of 1965 Vote
Number of Yes Votes in 1966
% of 1966 Votes Voting Yes
31967 Vote
1967 Vote as % of 1965 Vote
Number of Yes Votes in 1967
% of 1967 Vote Voting Yes
1A
195
100%
54
28%
169
87%
51
31%
145
74%
48
30%
IB
175
100%
76
43%
162
93%
66
43%
126
72%
46
37%
2A
160
100%
77
48%
130
81%
48
37%
145
89%
64
45%
2B
240
100%
115
48%
222
92%
123
56%
128
53%
77
60%
3A
219
100%
110
50%
214
98%
83
39%
186
85%
84
45%
3B
222
100%
82
37%
215
97%
79
37%
199
90%
78
39%
3C
132
100%
35
27%
128
97%
37
29%
72
54%
27
38%
4A
259
100%
135
52%
232
89%
125
54%
192
74%
106
55%

4B
220
100%
107
49%
214
97%
107
50%
180
82%
112
67%
            iNovember election-election of mayor.  The mayor opposed small sewer issue.

            2November election- -election of governor.  The mayor  "supported" big sewer  issue.  Size of issue
             caused decline of yes votes in Precincts 2A and 3A to the  level of the hardcore opposition precincts:
             1A, IB,  3B,  3C.
                y primary election—no candidates in Avon.  The mayor opposed small sewer issue.  % of yes
             votes from Northgate was high, .but the turnout was surprisingly small.  Experiment in neighborhood
             personal contact program in precinct 4B was very successful.

-------
Table 4. May 2, 1967 Vote to Change Charter
Precinct
1A
IB
2A
2B
3A
3B
3C
4A
4B

Yes
47
46
65
77
84
78
27
108
112
644
No
98
80
79
41
102
121
45
86
68
720

Ward
1A
IB
2A
2B
3A
3B
3C
4A
4B

Table 5.
Republican
33%
35%
56%
63%
43%
38%
36%
60%
62%

1969 Vote for
Candidate
68
72
141
117
115
90
65
165
157
990
Mayor
Democratic
67%
65%
44%
37%
57%
62%
64%
40%
38%


Candidate
136
133
113
70
130
144
115 ?
109
95
1045
244

-------
Table 6.
Precinct
1A
IB
2A
2B
3A
3B
3C
4A
4B
Voting Outcome, by
1969 Mayoralty
Candidate
Democrat
Democrat
Republican
Republican
Democrat
Democrat
Democrat
Republican
Republican
Precinct, For Selected
1967 Charter
Ame ndme nt s
No
No
No
Yes
No
No
No
Yes
Yes
Elections
1966 Sewer
Bond Election
No
No
No
Yes
No
No
No
Yes
Yes
In terms of voting on sewer bonds, Precints 1A,  IB  3B,  and 3C seemed
solidly opposed.  Precincts 2B, 4A,  and 4B were in favor.  Precincts
2A and 3A were swing precincts,  and the variable affecting their vote on
sewers was the cost of the project:  the higher the cost, the more  they
are opposed.  (A precinct map of  Avon is found in Figure 4. )

Another of the  reasons suggested  by  our respondents has to  do with the
strategy  of the campaigns which accompanied these bond elections.
Some of the respondents, for example, felt that a high turnout helped
those who oppose sewers; others disagreed.  Similarly, some said that
an informed electorate helped those who opposed  sewers; others held the
opposite  opinion.  While the  data do not provide clear answers to these
questions,  they do yield some interesting information.  For example, as
one can see in  Table  7,   the two obscure sewer issues--the 1964
straight levy and the  1967 general purpose levy--came closest to being
passed.   This  is not  a clear test of the obscurity issue, however.
                               245

-------
  ~	.	 	.    .	
Figure 4.  Precinct A

-------
Table 7. Outcome of Sewer Bond Elections,
Avon, Ohio, 1963-1970
Date
1963
1964
1965
1966
1967
1969
1970
May
Nov.
Nov.
Nov.
May
June
May
Votes Cast
983
1910
1830
1718
1364
1654
1234
Yes
320
911
794
751
644
662
531
31%
48%
43%
44%
47%
40%
42%
No
663
999
1036
967
720
992
703
69%
52%
57%
56%
53%
60%
58%
Other variables were associated with obscurity in 1964 and 1967.  In
1967, for example, sewers were tied to a general tax levy and therefore
tied to other popular  projects; in 1964, the straight levy was small,  and
voters may simply have reacted favorably to the  low cost.

The evidence  on the effect of turnout is even less clear.  The nearest the
sewer bond election came to  passing was  in 1964, and that was also the
election with the lightest turnout. The next closest was in 1967, and the
turnout in that election was also comparatively low.
     /
Conclusions
 Four conclusions can be drawn from the Avon case study.

 First, sewer bond elections in Avon were tied directly to the issue of
 development.  The debate over sewers was a debate over development,
 and development--at least for the present--has lost.   A majority of
 Avon citizens appear  to be opposed to development.  Study respondents
 mentioned this attitude many times, in many different ways, but the
 basic idea was the same:  "Avon is a nice town the way it is, " "There is
 a fear of big outside developers coming in and changing the town, "
 "Growth means buildings, people, and problems. "
                                247

-------
Second, in spite of this substantial opposition,  some proponents of
sewers keep trying.  For example,  in 1968, the OWPCB placed a buil-
ding freeze  on the city; in  1970, the mayor finally changed sides and
came out in favor of sewers after securing acceptance by the council of
the French Creek Plan.  Neither of these actions changed the outcome
of the election.  The OWPCB's action,  in fact,  was actually dysfunction-
al  because it hurt only the developers who were the main supporters of
sewers.

Third, as many of our respondents pointed out,  there was division
among local government officials.  Opposition to sewers by elected
officials, particularly by a man as well respected as their mayor,
created substantial doubts  in the minds of the voters.  Indeed, it would
appear that the mayor was so successful that when he shifted from
opposition to support of the sewer issue, he failed to bring enough of
his followers with him to assure success.

Finally, Avon's most  recent sewer bond election involved the issue of
regional government.  The Federal government is committed to region-
al solutions to water pollution problems, and it provides a financial
incentive to those local governments which will adopt a coordinated
regional waste treatment program.  The availability of more outside
funds was used by the  mayor as an argument in favor of his  regional
plans.  But many voters apparently placed a higher value on local auto-
nomy than the Federal funds,  because the issue was nonetheless
defeated.
                             248

-------
DEKALB COUNTY,  GEORGIA
           249

-------
                    DEKALB COUNTY

The referendum in question was for one of fourteen bond issues submit-
ted to the voters of DeKalb County, Georgia, on May 26,  1970.  The
•wording of the drainage referendum was:
            "Shall drainage improvement bonds in the
            amount of $2, 000, 000 be issued by DeKalb
            County? "
A public notice described the bonds as follows:
            "Drainage improvement bonds, in the princi-
            pal amount of $2, 000, 000 for the purpose of
            installing storm sewers and drainage facili-
            ties, expanding and improving existing drain-
            age  and related facilities, acquiring the  neces-
            sary land and rights-of-way, for the payment
            of damages arising therefrom and expenses
            incident thereto:	"
The vote on the drainage improvements bond issue was 12, 564 (42%) in
favor, and 17,060 (58%) against.

Population of DeKalb County

The primary governmental unit involved in this referendum is DeKalb
County.  DeKalb is one of the five counties that make up the Atlanta
metropolitan region.   The metropolitan area has  a population of approxi
mately 1, 300, 000 people,  and DeKalb County has a population currently
estimated at 400, 000.  Part of the City of Atlanta is in DeKalb County,
as well as eight other incorporated areas.   The names and populations
of these cities  are found in Table 1.
                              251

-------
Table 1. Independent
Located Within
Municipality
City of Atlanta
(Portion in DeKalb)
Avondale Estates
Chamblee
Clarkston
Decatur
Doraville
Lithonia
Pine Lake
Stone Mountain
TOTAL
Census
1960
41, 332
1, 646
6, 635
1, 524
22, 026
4,437
1, 667
738
1,976
81, 981
Governmental Units
DeKalb County
Estimated
1968
45, 781
1, 649
9,167
2,496
22, 827
7, 713
1,792
752
2, 320
94,497
Projected
1969 1970
48, 161
1, 734
9,643
2, 625
24, 014
8, 114
1, 885
791
2,440
99,407
51, 146
1, 841
10,240
2, 787
25, 502
8, 617
2, 001
840
2,591
105, 565
DeKalb is one of the fastest growing areas in the Southeastern United States.
According to data furnished by the Economic Development Department of the
DeKalb Chamber of Commerce, during the year of 1969, 84 new and expanded
industries located in DeKalb County,  representing 1, 944, 137 square feet of
new construction at an estimated dollar value•.of $20, 191, 342.  During the
past 20 years,  DeKalb County has been the focus of much of the residential
growth of the Atlanta metropolitan area.  As the data in Table 2 indicate,
DeKalb County has more than quadrupled its population of the Atlanta SMSA.
During the period from I960 to 1969, 38% of the Atlanta metropolitan areas
population increase occurred in DeKalb County.

The median age of DeKalb County in I960 was 25.  The age and sex distri-
bution of the County's population remained relatively stable during the
period I960 to  1969, in spite of the rapid growth of the County.37 Non-
white population as a percentage of total DeKalb population increased
from  9%  in I960 to 12% in  1969, but it is not expected to increase any higher
                              252

-------
Table 2. DeKalb County Population Percentage
of Atlanta SMSA and State of Georgia 1940 - 1969
PoDulation

State of Georgia


DeKalb 's Percentage
of SMSA
DeKalb 's Percentage
of State of Georgia
1940
3, 123, 723
558, 842
86, 942
15. 6%
2.8%
1950
3,444, 578
726,989
136, 395
1 8. 8%
3.9%
1960
3,942,936
1, 017, 008
256, 782
25.2%
6.5%
1969
4, 568, 000
1, 302, 000
364, 700
28. 0%
7.9%
SOURCE: U.S. Bureau of the Census and Atlanta Region Metropolitan
Planning Commission.
                               38
than 12% during the next decade.   The non-white population is concen-
trated in Atlanta in DeKalb,  Decatur, and in the rural areas and small
textile mill towns of the eastern part of the country..

The population of DeKalb County has the highest income per household
($12, 146) and per capita ($3, 543) of any county in Southeastern United
States with a population in excess of 50, 000.   Percentage 46.4 of all
                                                          39
households in DeKalb have an income of $10, 000 or greater.   It also
has the highest level of education of any county in the Southeast.40

This high level of income does not mean that there  is no poverty in De-
Kalb County.  Indeed,  as Table 3 shows, a majority of all non-white
family incomes are below the poverty level according to the I960 census
Non-whites also occupy a substantial number of the deteriorating and
dilapidated housing units  in the county.  According  to the I960 census,
65% of the non-whites  in DeKalb live in housing which was classified as
       A 41
un s ound.
                              253

-------
Table 3. Population of DeKalb County

Family Income
Under $3, 000
$3 - $6, 000
$6 - $9, 000
$9, 000-above
by Family
White
4,592
13, 178
17, 921
16, 845
Income and
Non-White
1,798
1, 130
153
43
Race
% White
8. 7
25
24
32

% Non -White
57
36
4.9
1. 3
Housing in DeKalb  County is primarily single-family and owner-occupied.
Of the 76, 615 occupied units in I960,  68,374 were single-family, and
only 8, 241 were multi-family.  However, the number of multi-family
units grew during the 1960's, so that  in 1969, of 127, 341 units,  99, 173
were single-family, and 28, 1 68 were multi-family.  Moreover,  when
one considers only new housing construction,  multi-family units repre-
sented only 8% of the total construction in I960, but increased to 50% of
                       42
all construction in  1969.

Until very recently, DeKalb County was primarily a bedroom suburb; a
substantial number of persons in the county commuted to other parts of
the metropolitan area (primarily Atlanta) to work. In  1961,  the total
labor force of the county was 90% higher than total employment  in the
county.  There has been a conscious effort on the part of DeKalb to di-
versify its economy,  and this effort has resulted in a significant increase
in manufacturing in the county.  This increase has come primarily in
high-wage industry groups,  thus  contributing to the  preservation of the
high level of personal income in the county.  Nevertheless,  DeKalb  still
continues to  export a significant percentage of its labor force outside the
     ^  43
county.

Although there was no detailed  information on the characteristics of the
labor force in DeKalb  available for this study, authoritative estimates
suggest that it has  a higher percentage of white-collar workers  than  other
areas of the  metropolitan  region and that unemployment is  low.
                              254

-------
Government

DeKalb County is governed by a board of county commissioners.   The
board is composed of five members; four of the members represent
districts, and one, the chairman,  is elected at large.  The elections
are partisan, and the current board is  composed of four Democrats and
one Republican.  Commissioners have  four-year terms, staggered
so that  some part of the commission must face election every two years,
(A list of the commissioners appears below in Table 4).
Table 4. List of the Commissioners
District
Chairman (At-large)
District One
District Two
District Three
District Four
Name
Clark Harrison
Arthur C. Guhl
T. M. Callaway, Jr.
J. C. Haynes
William M. Evans
Party
Democrat
Republican
Democrat
Democrat
Democrat
Term
Expires
1972
1972
1970
1970
1972
Geography

Geographically, development of DeKalb County began in the cities of
Atlanta and Decatur at the center of the western border of the county
and moved north and east along the routes of the highways and railroads.
There have been few topographical barriers to this development although
there are some areas of the county where poor soil and excess water
have created problems.  The area south of and immediately adjacent to
the South River is particularly susceptible to flooding.  Eastern DeKalb
also has flooding problems  particularly along the Yellow River, and some
creeks  in the North such as the Nancy Creek overflow periodically.  Ex-
cessive construction at the  headwaters of  streams contributes to the
drainage problems.  (A map of the county  with major streams and flood
plains indicated, appears in Figure  1).
                              255

-------
i   I
r\
•-»....

  .-/ ? \
               \
                 \
                     '• A
                         \

                      '/\
                   -,
                                                 \
                                                      j.s- /

                                        ^?
            CJUUTt
                                      ^    »    v "'*..
                                          $    \., 1;>
                                      '?   ^   , ^ "i /

                                      •?  ^;!   A ;/

                                      •v?   ;<•">•
                                             -o
                     Figure 1. Streams and Flood Plains




                      256

-------
 During the last two decades of rapid growth in the county, developers
 were allowed to build on the flood plains of many of these streams.  The
 result has been constant flooding of basements and other damage to pro-
 perty located in these areas.  Future building  on the floodplain has been
 limited by zoning changes recently enacted.

 County Finances

 The county's financial position is the subject of much debate.  Services
 provided by  the county out of tax revenue have decreased since sanitation
 and water were put on a user charge system in 1967.  There have been
 no recent major capital improvements.  The gross bonded debt of the
 county has been steadily decreasing over the past five years—from $34
 million in  1965 to $28 million in 1969.  However, in  1969, the county
 had the largest tax increase in its history.  (Tax rates for the past five
 years are  shown in Table 5).

 The most important features of the county's financial position  are directly
 related to DeKalb's recent political history. In 1968, the incumbent
 chairman of the board of commissioners  was  defeated in an election
 where the main issue was the fiscal responsibility of his Administration.
 His opponent charged that the administration had gone deeply into debt
 by borrowing on the county's water and sewer funds,  which were not sub-
 ject to audit. A significant part of the population never accepted these
 charges, and when the new chairman raised taxes by $4. 97 in  the summer
 of 1969, on the grounds  that he had to use that  money to pay for the pre-
 vious administration's debt, a large part of the public rejected this ex-
 planation.  In short,  1970 was not a good time  for a bond referendum.

 County Politics

 The controversial election of 1968, and the voters' quick dissatisfaction
with the victor in that election, is symptomatic of the current  political
 situation in DeKalb County.  It was not always  this way.   In the early
                              257,

-------
                                                  Table 5.
                                  Tax Rates for DeKalb County,  1965 -  1969
Actual Tax Rates per $1000
County









Schools




State



General
Fire
Sanitation
Bond Retirement &: Interest
Hospital
Accumulated Debt
Total County


Maintenance
Bond Retirement & Interest
Total Schools



Total Tax Rate
i— 	 , 	
1969


8.3.
1.80

2.00
1.80
2.46
16.37


NA
NA
NA

NA

NA

1968


7.90
1. 50

2.00


11.40


18.75
4.00
22.75

.25

34.40

1967


8.20
1.50

2.00


11.70


18.00
4.00
22.00

.25

33. 95

1966


10.20
1. 50
2.20
2.00


15. 90


16.00
4.00
20.00

.25

36. 15

1965


10.20
1. 50
2.20
2.00


15.90


16.00
4.00
20.00

.25

36. 15

IS)
(SI
oo

-------
1950's, when the head of one of DeKalb's County's oldest and wealthiest
families was ending his  long tenure as the single county commissioner,
he had more control over DeKalb than the "power elite" had over Floyd
                 43
Hunter's Atlanta.   But  the history of the county ever  since that time
has been one of increasing political instability.  Each of the last three
chairmen of the board of commissioners has tried to succeed himself,
and each has been defeated.   The present chairman is in serious political
trouble after losing the  referendum, and most observers agree that he
would probably be defeated if he ran for re-election.

This political instability stems from two factors:  (1) the rapid growth
over the past two decades,  and (2) growth of the Republican party.  The
county's population growth during the past two decades has had a signi-
ficant impact on the political scene.   Even though the immigrants to the
county have the same  socio-economic characteristics as  the original
residents in terms of  education level,  income, occupation classification
and race, they are still  "new people" to DeKalb.   When they did not find
places for themselves within the existing party structure in the 1950's
and 60's, many of them  formed factions which challenged the incumbent
leadership of the Democratic party.   Since 1964 there has also been a
growing Republican party, and many of the "new  people"  (along with dis-
affected conservative  Demoncrats) have allied themselves with this party.
The Republicans are optimistic about gaining control of the board of com-
missioners  in the 1970 election.

The general instability of DeKalb politics makes  it difficult to describe
voting patterns in the  county,  but political observers make the following
generalizations.  The county is conservative, but its convervatism is of
the Nixon variety.  In. the 1968 presidential election, Nixon won 50% of
the vote; Wallace came  in third.  The Wallace supporters are located
primarily in the southern and eastern parts of the county.  Republicans
are located  in the southwest and northwest of the county.   The liberal
vote  comes  from the whites  living around Emory University and from
the Blacks in Atlanta, and Decatur.
                               259

-------
The Referendum of 1970

It was against this political background that the current referendum was
held.   Plans for the referendum began nearly a year before  it actually
was held.  The first step was the appointment of a Citizens Advisory
Committee on Capital Improvements.  From the very beginning, parti-
san politics was evident in the bond campaign.   The chairman of the
Citizens Committee,  a Republican, resigned from the post shortly after
being appointed.

After public hearings in the winter and spring of 1970,  special sub-
committees of the Advisory Committee turned  in their list of proposed
capital improvements which totalled $104 million.  This was substan-
tially more than the $60 million county bonding capacity.  In April,  the
full Committee reduced its  requests to a package of $35 million of high-
priority projects.  The new administration was not satisfied with this
package, however, and increased the package by $9 million — including
money for  a new jail,  a new police headquarters, and a recreation facili-
ty in Glenwood Park.  The items presented  to the voters are found in
Table  6.

It was apparent from the outset of the campaign that the referendum was
in trouble.  In May, the Republican Party announced its opposition to all
but two of the fourteen proposed items on the ballot.  Republicans charged
that the administration was not living up  to the Bond Accountability Law
which provides that any statement of advertisement about the fete of bond
money is binding on the people who dispose  of the bond money. (This was
a bill which was supported by Republicans and  pushed through the State
legislature after a previous administration had failed to carry out a large
number of  capital improvements which it had proposed during a successful
1961 referendum campaign. )  This violation of bond accountability,  accor-
ding to the GOP,  was most evident in the category of streets and roads,
where the administration had allegedly established a $3 million slush fund
for industrial paving  which the Republicans  charged would be used for the
benefit of contractor  friends.
                               '260

-------
Table 6. Items
Item
Juvenile Detention and Care
Facilities
Garbage Disposal Facilities
Drainage Improvements
School Access Roads and
Sidewalks
Streets, Roads and Traffic
Improvements
Parks and Recreation
Facilities
Fire Department
County Jail
Courthouse Parking Annex
Hospital (South DeKalb)
Health Care Centers
Policy Department
Libraries
Vehicle Storage and Main-
tenance Facilities
in the Bond Referendum
Amount
$ 2, 170,000
6, 080,000
2,000,000
1, 850,000
18, 650, 000
3, 510,000
1, 000, 000
2, 500, 000
1, 185,000
1, 950, 000
1, 140, 000
1, 500, 000
1, 165, 000
300, 000

Result
Pass
Pass
Fail
Fail
Fail
Fail
Pass
Pass
Fail
Pass
Fail
Pass
Fail
Fail
261

-------
Republicans also charged that the administration was unnecessarily
costing the tax payers $30, 000 by holding a special bond election in
May,  rather than waiting until the general election in November.  They
also alleged that the administration had scheduled the election purpose-
fully to encourage a low turnout so that the administration supporters
could  "push through the referendum".  The opposition further charged
the commissioner with a  "credibility gap" because of his promises that
the bonds would not require a tax increase, even though the  county had
a $4. 97 tax increase in the summer of 1969.

On May 14, 1970,  the mayor of Doraville announced  his opposition to
several parts  of the referendum on the  grounds that Doraville was not
getting its fair share of the improvements to be financed by the bonds.
He charged, for example, that the tax digest of Doraville is 10% of the
gross county digest,  and yet,  not one dollar in the  bond issues had been
earmarked for street improvements in Doraville.  Also in May, a mem-
ber of the DeKalb  County Bond Commission from 1961 to 1966 who was
also a member of  the Republican Party Executive Committee of DeKalb
County, mimeographed and distributed a pamphlet  on the 1961 "bond
scandal" and charged that the 1970 referendum would be a repeat of that
scandal because "the same Establishment crowd is in power and control
of the county Government".

Support for the bond issue was being  organized by the information direc-
tor for the board of commissioner s.   He directed an $80,000 campaign
which began six weeks before the election.  It  included a four-minute
film (which cost $6, 000) which was shown to civic and fraternal organi-
zations; speeches  to these organizations;  brochures which were distri-
buted  throughout the county; a newspaper insert which appeared in
county papers and then was mailed to 40, 000 residents; posters made
by women's clubs  and placed in county businesses; a nine-part series
of articles in local newspapers; and two series of programs on Atlanta
television.  The director and his staff had  support from the local  Chamber
ber of Commerce   the League of Women Voters, other women's clubs,
                              262

-------
and several local businesses.   The Atlanta Constitution and Journal
were both in favor of the referendum, and many of the county news-
papers printed the nine-part series on the  issues written by the infor-
mation  director.  The support, however,  was not enough.  Eight of
the fourteen  items were defeated by the voters.

Interpretation of the Outcome

It is difficult to distinguish why the drainage bond failed from the ques-
tion of why the entire  referendum failed.  To the extent that it is possi-
ble to focus on the drainage part of the total referendum package, how-
ever, informed observers and participants in the campaign offer four
major explanations for its failure.

First, drainage is not a big problem for many people in DeKalb County.
Some persons do live  in flood plain areas  where floods could cause sub-
stantial damage to their homes and some  live in areas where there is
currently improper drainage,  causing flooding of basements and other
inconveniences.  But the number of persons who are seriously harmed
by drainage problems is  small in comparison to the total  population of
400, 000.

Second, in comparison with some of the other issues on the ballot,
drainage was not a salient issue to the average voter.  This is related
to the first reason suggested above—that  drainage was not a serious
problem.

Third, the drainage improvements planned by the commissioners were
not specified in the publicity on the referendum.  Therefore, even those
citizens who were seriously affected by drainage problems were not
sure that the county planned to  do something about their problem with
the money from the bonds.
                               263

-------
Fourth, one newspaper reporter suggested that there was a racial issue
underlying the drainage issue.  He argued that drainage improvements
would open up for development (and particularly for multi-family dwel-
ling units) certain areas of the county which had previously not been
developed.  Some residents allegedly fear that such developments might
facilitate  the in-migration of Blacks.

On  the question  of why the referendum as a whole was unsuccessful,
the  following arguments were advanced.  First,  all participants and
observers of the referendum (both proponents and opponents alike)
agreed that May 26 was a bad time for a referendum.  This was a
special election being held at a cost of $30, 000 when it might have been
held in November at no extra cost.  Further, there was  another special
election for school bonds  planned for two weeks after this referendum,
also at a cost of $30, 000.  Finally,  taxes had been raised  substantially
in 1969 and the taxpayers were not anxious to have another increase.

Some opponents and some supporters  claimed that there was little or
no publicity about the referendum and that the voters didn't understand
the  issues or the significance of  the issues.   Opponents of the referen-
dum charged that the administration had consciously limited the publicity
in order to limit the turnout.

Some county officials claimed that the low turnout contributed to the de-
feat of the referendum.  (They also claimed that they had vigorously
publicized the referendum in an effort to increase turnout).


Mearly all of the participants and observers  agreed that there  was a
credibility gap between the administration and the electorate.  Those
                              264

-------
who were favorably disposed to the administration argued that this was
an endemic feature  of all administrations in DeKalb in recent years,
or that it was a product of a previous administration.   Those unfavor-
ably disposed to the chairman of the board argued that he had created
his own credibility gap by raising  taxes and by not providing the voters
with accurate information on the planned uses and means of financing
of the proposed bonds.

The facts of the campaign provide some support for many of these sug-
gested reasons.  Some, however,  like the alleged racial issues  under-
lying and drainage part of the referendum, are  rather difficult to sub-
stantiate.  Others,  like the question of the amount of publicity,  are
difficult to evaluate.  (There was more publicity in DeKalb than  there
was in Avon,  Ohio,  but then, DeKalb is much larger. ) Still others,
like the suggestion that a low turnout contributed to defeat, are rather
dubious in light of what appears to be the  trend  in other jurisdictions.
Finally,  the fact that drainage was only part of  a total referendum
package  complicates the analysis.

On the basis of the analysis of other referenda,  it seems that the most
important reason for the defeat of the referendum package  was the
credibility gap of the county administration.  Given this,  there was little
the administration could have done to improve the referendum's chances
of passage.  Political history shows that all recent administrations in
DeKalb have suffered from this problem.   Because each administration
is unlikely to  have a stable political constituency, opposition  forces
have every reason to attack the incumbent.   Constant  reiteration of
opposition charges creates doubt in the minds of the electorate.   Thus,
even when events  are not under the control of the administration,  the
administration nevertheless gets  the blame from the electorate.  (For
example, the  administration at the time of our study had no control
                               2S5

-------
over the timing of the  school election, and yet this was cited as one
of the aspects of the "bad timing" of the referendum).   This is not to
say that the administration did not make political mistakes.  Indeed,
the  tax  increase in 1969, one year before the largest proposed bond
issue in the county's history,  was a serious error.   But the impor-
tant factor was that the administration had lost what little legitimacy
it had when it assumed office.  In this situation, it is unlikely that an
increase in publicity or turnout would improve the chances of passage
of the referendum.

Given the fact that the whole package was headed for defeat, and given
the  fact that drainage was not a serious problem for most voters, it was
unlikely that the drainage issue could  pass even if there were changes
made in its presentation on the ballot.  One  of the changes now being
considered by the  commissioners is the specification of the improve-
ments to be planned.   But the two issues where this was done in May--
streets  and roads  and  parks recreation--both failed.

Examining the six issues which passed and the reasons for their pass-
age may be useful.  Three Juvenile Detention and Care Facilities, the
County  Facilities, the County Jail,  and the Police Department — received
support, probably from the current concern for  law and order.  The
Garbage Disposal  Facilities received support from current concern with
pollution.  The South DeKalb Hospital received support from residents
of South DeKalb, who voted heavily against other issues on the referen-
dum.  Finally, the Fire Department is a popular county service which
always  receives support from the voters.  Drainage improvements  seem
to lack this sort of appeal to the voters, and  it is unlikely that changes
in packaging will change this.

Conclusions

At least two conclusions can be drawn from  the DeKalb County exper-
ience.   First, it is extremely difficult to pass a referendum when there
                               266

-------
is a credibility gap between the government and the voters; and second,
it is extremely difficult to convince the voters that they ought to support
with public funds the construction of what is essentially for private
benefit.

To say that the DeKalb County administration had a "credibility gap"
actually  raises more questions than it answers.  What is a credibility
gap?  How do we measure it?  What causes it? In answer to the  last
question, there appear to be two  phenomena which are frequently asso-
ciated with credibility gaps: a loss of legitimacy and authority by impor-
tant government officials and a breakdown of traditional political  struc-
tures.  These two phenomena are no doubt related,  and both seem to  be
related to rapid population  growth.  A better conceptualization of the
"credibility gap" phenomenon is required before  the other questions can
be answered.

One important characteristic of drainage improvements is that they are
private goods. This is  particularly interesting in light of the issues
which passed  in DeKalb. All of them were essentially public goods except
for the South DeKalb Hospital,  and the hospital had substantial support
from residents who would be served by it. The important question is: how
does a government convince the public to support bonds for private goods?
Does  it specify the benefits? Analysis  of other bond elections indicates
that this often creates more problems than it  solves.  Does it pay for the
bonds  by a special assessment or by user charges,  rather than through
general  revenues?  Assessment and user charges are complicated issues,
and an administration with  a credibility gap may  have difficulty explaining
the equity of its choice of assessment.  Does it campaign more vigorous-
ly?  There are techniques which the county administration did not use and
which might have helped.  For  example, they might have surveyed the
county to find but where their supporters and opponents were located, and
then they might have made  every  effort within the law to get their suppor-
ters to vote.  This might have put the administration in the position of
registering  Black voters and encouraging them to vote,  which is still a
                                267

-------
delicate issue in Georgia.  Some comparative studies of drainage refer
enda which succeeded might shed some light on these questions.
                               268

-------
OKLAHOMA CITY, OKLAHOMA
            269

-------
                      OKLAHOMA CITY


The referendum in question was one of thirteen bond issues offered to
the voters  of Oklahoma City,  Oklahoma on May 28, 1968.  The wording
of the storm sewer referendum was as follows:
            Shall the City of Oklahoma City,  State of
            Oklahoma,  incur an indebtedness by issuing its
            negotiable coupon bonds in the sum of five mil-
            lion, two hundred twenty-six thousand six hun-
            dred twenty dollars, ($5, 226, 620) to provide
            funds for the purpose of acquiring land,  rights-
            of-way, and equipment for and constructing and
            purchasing improvements, extensions and re-
            pairs to the Storm Sewer System, to be  owned
            exclusively by said city,  and levy and collect an
            annual  tax, in addition to all other taxes, upon
            all the  taxable property in  said city sufficient to
            pay the interest on  said bonds as it falls due, and
            also to constitute a sinking fund for the payment of
            the  principal thereof when due,  said bonds to bear
            interest at not to exceed the rate of five per centum
            5% per annum,  payable semi-annually, and to be-
            come due serially within twenty-five years from
            their date ?
The vote in this storm sewer bond issue was 14, 299 (49%) in favor and

14, 626  (51%) against passage.


Background


The primary governmental unit with which we are concerned is the

City of  Oklahoma  City.  This sprawling city encompasses approximately
75% of Oklahoma County and spills over into three neighboring counties.

The city proper currently has a population of 361, 270 people, but sur-

rounding suburbs  swell the area population of metropolitan Oklahoma

City to well over a half million.   Table 1 contrasts the population
figures  and growth rates for Oklahoma City proper, metropolitan Okla-

homa City,  Oklahoma County, and the State of Oklahoma for I960 and

1970.
                               271

-------
Table 1. Population and Growth Rate for
Oklahoma City and the State of Oklahoma
Area
Oklahoma City
(Proper)
Oklahoma City
(Metropolitan)
Oklahoma County
State of Oklahoma
Population
I960
324, 253
511, 253
439, 907
2, 328, 284
Population
1970
361, 270
629, 305
517, 000
2,500,312
Growth Rate
11.1%
22. 9%
11. 8%
7.4%
This table indicates that Oklahoma, particularly the metropolitan area,
has experienced more rapid growth than the rest of the State of Oklahoma.
This data, however,  is somewhat misleading because  it does not take
into consideration the annexations to metropolitan Oklahoma City in the
past decade and the corresponding discrepancy between  Oklahoma  City
in I960 and  Oklahoma City in 1970.   For the past thirty  years,  suburban
settlements such as Hichols Hills, the Village,  Del City, and Warr Acres
have been attracting residents from the central  city.   Yet,  during  the
1960-1970 period,  Oklahoma City has brought some of these areas plus
some other  suburban areas under its governmental jurisdiction.  Thus,
of the increase in population of 37, 000,  the residents of the annexed
areas comprise most of this growth.

The total land area of Oklahoma City is over six hundred square miles,
making it one of the two largest cities in the United States. This great
land area is the end result of a decade of annexation battles in the  courts.
A little over ten years ago,  a drive to increase  the tax base of the city
was initiated with the annexation of a small community northeast of the
downtown area and on the outskirts of the city.  After  that annexation,
the city grew tremendously in all directions until its city limits came
into contact with the boundaries of several small cities twenty to thirty
miles from  the center of Oklahoma City.  Fearing that Oklahoma City
                               272

-------
would soon surround them, these small cities of Norman, Moore,  Mid-
west City, El Reno, and Edmond embarked on annexation ventures of
their own. Because of annexations to these cities, Oklahoma City is
now almost entirely surrounded by a chain of cities which are large
enough to prevent their annexation.

Oklahoma City was aptly named, for it has always been a city,  never a
town.  On the day it was founded in April,  1889,  during a land rush,  it
had a population of 30, 000 people.  It has always been the largest city
in the state.   Its oil and natural gas resources attracted money, thus
assuring  the early development of the city into a metropolitan area.
Because of its central location and size, it was  chosen as the State
capital.   Although Oklahoma is a rural state by nature, the problems
which Oklahoma City faces are very familiar to the largest Eastern
cities: a deteriorating downtown core,  flight to the suburbs, and a
mounting  list of expenses.  Urban renewal has been one of the hottest
political issues in Oklahoma City for a decade.   The  present mayor is
deeply pledged to restore the downtown area by qualifying the city for
large amounts of Federal funds.  During his administration, a large
downtown playground-mall has been built and a huge convention center
is currently  under construction in hopes of drawing large amounts  of
outside capital from visitors to the city.  Future plans call for  major
changes in almost every downtown building,  and  the total cost will be
in the neighborhood  of 300 million dollars.

Population Characteristics of  Oklahoma  City

Today the downtown area of Oklahoma  City is inhabited largely  by Black
and Mexican-American families.  It is almost entirely contained in the
present Ward Seven (see  Figure  1  ), although it does  spill over into
the southern part of Ward Two.  The northern part of Ward Two has a
predominantly white population, whose numbers  more than offset the
Blacks in the south.   The eastern two-thirds is  sparsely  populated,
having been allocated for future industrial use.   The  six wards  remaining
                                273

-------
7^
 -V
      fyKMfjas


      }y
      MW
      L
         rr
         i£SJl
         IO6"" t ST-
      bean TON
                  >-,
             ~~ "..L_	



              f ^V»>0
             ^S
            l2£L
            PO^P
                                 \ -

                                  5_
                              -'-''  )>-
                         'wA
                         ^
                            •S i
sc
                           J9jte
                               .\
                               ^
_L
      ^
                                        ^  FT
                                        ^T"-JL
                                       B&J.
          vrx
        >^
                                        v/
           T^
                                               o.
                                                 /)->
                  V
\^ -i-

i
                   SQ
                                     —,.o=-5»—-,



                                     f
                                                     —r—•"»
                                                     jEu
                          asi-i
                                                       -H9
                     or
m
                         WWT-
                         f/ms .-
                           5uW

            icmrcrwxai      -l;2k "\LJ   -Ji_
    Asorto BY TMI cm c^vtxa.
        OCTOStS II. IfeLt

_   8 WARD PROPOSAL  _

           LEGEND
!
   EXISTING WARD LIKES

   PROPOSED V/ARD LIMES

   INCORPORATED AREAS

      Figure 1
                         274

-------
-J
(J\

-------
have small numbers of Blacks.  Wards One and Eight tend to be the
areas of affluence, where several country clubs and many large es-
tates are located.  Southern Wards Three, Four,  and Five are popu-
lated predominantly by blue-collar factory workers.  There are several
meat-packing plants in this area which provide employment.   Ward Six
contains many private businesses and some blue-collar workers.

A voter breakdown by parties and sex is given in  Table 3  (see follow-
ing page).  It can be seen that the predominantly Black Ward Two has
the highest Democratic registration; whereas, the more affluent Wards
One and Eight possess the highest percentage of registered Republi-
cans.

The numbers of non-whites in I960 was 42,282 (13%)  and it is believed
that this number as well as the  percentage has increased during the
past decade.  No figures for 1970 were available  regarding this hy-
pothesis; however, the city administration has estimated that 20% of
the city is non-white.

 The median  family income  of $5, 740 was fairly typical of the cities in
the South Central area of the country in I960.  Although the median
family incomes of Dallas and Tulsa are higher, the amounts are not
 significant.   Table 2 gives  family income distribution in Oklahoma
 City for I960.
              Table 2 .  Income Distribution/Oklahoma City
                 Less than $3, 000/year
                 $3,000 to $10,000
                 $10, 0,00 to $20, 000
                 Over $20, 000
17.2%
68. 0%
13.3%
 1.6%
                               276

-------
                                                 Table 3.
                                        Oklahoma City, August, 1968
Ward
1
2
3
4
5
6
7
8
Total

Demo
15,577
16,040
17,753
17, 101
18,501
14,771
16,287
15,615
!3_L-!!5

% D
66.150%
74.030 '
77.724
86.356
81.892
79. 133
89.899
73.214


Rep
7783
5424
4890
2541
3808'
3718
1673
5579
35,416

%R
33.052%
25.033
21.409
12.831
16.856
19.919
9.234
26. 158


Ind
168
180
155
104
205
143
153
118
1,226

Amer
20
23
43
57
78
34
4
16
275

Male
11, 127
10,034
11, 153
9,348
11, 112
7,869
8, 140
9,262


%M
47.252%
46.310
48.829
47.205
49. 186
42.157
44.930
43.426


Fern
12,416
11,623
11,680
10,436
11,468
10, .788
9,957
12,057


%F
52.726%
53.644
51. 136
52.699
50.761
57.795
54.959
56.531


Total
23,548
21,667
22,841
' 19, 803
22,592
18,666
18, 117
21,328
168, 562

                                     B. Oklahoma City, August, 1970
115,477
16,307
17,445
15,536
19,050
13,496
15,529
14,730

127,570
62.497%
72.799
73.888
83.670
78.317
76.438
90.490
69.764


9015
5764
5788
2676
4717
3879
1477
6142

39, 548
36.403%
25.732
24.515
14.411
19.392
21.969
8.606
29.089


184
243
205
134
204
147
147
147

1,491
88
86
172
222
273
134
8
95

1,078
11,654
10,357
11,439
8,679
11,938
7,409
7, 607
9,139


47.06 %
46.236
48.449
46.741
49.079
41.963
44.327
43.284
13, 106
12,039
12, 163
9,877
12, 377
10,239
9,539
11,969
i
1
52.923%
53.745
51'. 516
53.193
50.883
57.991
55.585
56.687


24,764
22,400
23,610
18,568
24,324
17,656
17, 161
21, 114
	
169,597
  1
  2
  3
  4
  5
  6
  7
  8

Total

-------
Government

Oklahoma City has a council-manager form of government.  Under the
city charter, the city council has the responsibility for appointing the
city manager and can remove him from office by a two-thirds  vote.  He
is responsible for all city appointments and acts  as a liaison between
the city council and each department of the city.   The council can con-
duct no business with a city department without the consent of  the mana-
ger.  The mayor is the president of the council,  and is responsible for
the endorsement of all official bonds,  contracts,  etc, approved by the
council.  He may not vote at a council meeting unless there  is  a tie.

Proposed Bond Package of May 28,  1968

The history of the storm sewer bond issue begins in late  1966  when the
mayor and the city council appointed a 100-man Citizens' Committee to
decide on the substance  of a package of bonds to be submitted to the
voters in early 1968.  The formation of this Committee was  well-publicized,
but public interest  in the Committee's hearings was lacking.  The Commit-
tee's  specific  instructions were to itemize all projects which it felt should
be undertaken in the city, obtain preliminary cost estimates, and decide
on the substance of a bond issue package costing  about $100, 000, 000.

Preliminary hearings had established that this $100, 000,  000 figure would
cover about one-tenth of the costs for all  needed  projects, so project
priorities were established and used to fill the bond package.  The final
result was the listing given in Table 4.

The Campaign

In a time of voter apathy, the only projects which sparked any interest
at all were the sanitary  sewer and convention center.  The fate of the
other eleven propositions was closely tied to the  outcome of propositions
two and nine.  The  drainage  or storm sewer  bond, the subject of this  re-
port,  was a minor  issue in comparison with these two more  costly issues.
                              278

-------
          Table 4.  Summary Bond Package of May 28,  1968
                 Propositions                      Cost
         Streets and Bridges                    $ 12, 143, 100
         Sanitary Sewers                         34, 945, 665
         Storm Sewers                            5, 226, 620
         Water Works System                    13, 189,460
         Solid Waste  Disposal                       687, 300
         Municipal Maintenance Facility             700, 000
         Fire Stations                             2,478, 670
         Traffic Control                           3,241,153
         Convention Facility                      18, 000, 000
         Airport                                 13,777,360
         Parks and Recreation                     7,454, 560
         Fairgrounds                              1,400,000
         Public  Transportation                     2, 104, 500
                                     TOTAL   $115,348,688
It was the self-appointed job of the Oklahoma City Chamber of Com-
merce to campaign for the acceptance of these issues.  Because of
Oklahoma City's previous tendency to pass large civic bond issues, the
Chamber officials decided on a low-keyed approach, feeling certain
that the bonds would pass.  Little worry was given to a small, but fairly
vocal opposition which sprang up in the  Black wards,  Two and Seven.

Led by a member of the 100-man citizens' committee, a dissident
group publicly declared its  opposition to the sanitary  sewer issue.  The
opposition wanted a fairly dilapidated sewage treatment plant which was
going to be improved by the bond issue to be moved because of its proxi-
mity to several schools and many homes in the area.   If this were not
feasible  either technically or financially,  they demanded no less than
a complete renovation of the facility—not just the improvements which
the bond slated for the plant.   Although  this group did not really organize
                               279

-------
an opposition campaign, the Daily Oklahoman covered the story thor-
oughly and thereby publicized their views.

The other dissident note was heard in conjunction with the proposed
convention center,  proposition number nine.   Preliminary work by the
Citizens' Committee had established the site of the proposed convention
center.  Located near the downtown area, it was hoped that the center
would be able to qualify for urban renewal funds and reinforce the  econo
mic stability of the deteriorating downtown area.  This proposition was
opposed by a small group of ultra-conservatives, led by the wealthy
land-owner who held the title of the land on which the center was to be
built.  The group never mounted an effective opposition, although they
did put out a  small,  sparsely distributed pamphlet.

By the time of the election, the small number of  voters who came  to the
polls approved  only the fire station proposal.  (The results of this  elec-
tion are given in  Table 5).  Observers credit this issue's success  with
voter identification of  a social need for the project.

Conclusions

The small but well-publicized opposition to the sanitary sewer bond and
the disunity among the members of the citizens' referendum committee
apparently raised doubts about the bond package in the voters'  minds.
These  doubts were not offset by information from proponents of the
package concerning the social utility of the proposed issues.   Oppon-
ents to the package argued that there was very little in the nature of  the
project which would have  satisfied pressing social needs.  There was
no clear need for the convention center,  and improvements on the  sani-
tary  sewer.system were more desperately needed elsewhere.

Opponents and proponents of the bond issue argue that low voter turnout
helped bring  about defeat of the bond package.  The factor of low turnout
                               280

-------
oo
                                                                     Table 5.

                                                                TABUIATIOH 0? R2TUHXS

                                                     OHAKOMft. GUY SPECIAL ELECTION, KAY 28,  1863


         .;-   1           2          3          4          '5          6          7          8           5           10         11          12         13

          -^   y?s   j£3   y-s .  :-:_o   y?s   NO   yz_s    NO  y?;s   NO   vps   NO   y-.'s   NO   y£s    KO  -.YES   jro   yi3   jw   yis  2i2   YES   rj   r-g

         1522 27S4   2235 32SS  2171 3115   2210 3063   2143  3.133  2295 2884  2C09 3237  2184 3106   24t!i 2753  2453 2SC5  2C42 2015  2SC5.27/.4  2-I7C 27£2

         ICiC 2051   1319 23<2  1550 2135   1886 2101   J.MSS  2137  1720 2036  1SOO 2186  1700 2070   ISiS 1003  1S55 1915  175J i£-CS  1S77 1J75  ISii 1C;;
                      -
                 2-iCO 12-13  2233  1S22  2270 1775  2271 1779  2333 16S-1  21-13 1917  22-18  17S3   2446 1571'  20£-1 if.25  23C5  1713   2'.35 1551  2-133 15^-C

      l.:D2 1C1-1.  21-:o 11G3  1037  1C=0  1254 1054  1927 1033  1S66 1021  1865 1153  1929  1071   2021  971  2004  967  1275  1022   2-25  934  £OC£ 1C11

      :;C4 12C-i  25S4 1711  2530  1567  2529 1510  2511 1532  2512 1475  2355 1740  2520  15S6   2707 13C3  2550 1-132  2530  1520   2321 1377  2735 1J.-IJ

      1013 1577  1303 1517  '1172  1339  1193 1343  1172 1353  1230 12S7  115S 1302  1174  1305   1281 1251  1271 125S  1243  1237   1273 1255  1232 13C3

      1C5D  502  1581   542  1235   507  1233  491 J2.25  503 .1235  477  1213  502  1226  477   1173  430  1231  453  122C  433   1223  470  121G  4£-

      !::::- SSOi  1324 23SS  133S  i.;73  1731 2537  1670 2S66  1767 2545  1S09 2715  163S  2825   1384 2420  1737 2401  1792  2SC9   1£G4 2401  1805 2311
                                 14525
                                        14871
             . n sy r. f,
             — ^ £, *4£.
                         15542
                                 14299
13925
       14530
15170      13376
1.4659
15852
15546      152C3
1 f. c •>'
 w s WX
                                                                                                                                                 15721
                                                           14256
    13430      14332      14007      12722
                          12551
                           13453      12531
   CCl'.-.l   2-1321      31501       25S25      237S6      28835     28600       28353      2SS66      23531       235S7      2C5S1      2£612
!Vor key to proposition # and title' of
 see Table 4, Col.  1,  "Proposition. "
                                         proposition,
            TOl'AI. VCTZJS - 33,452  G-CIA-fOH?. CO'JKTY
                               22  CTK23 CC^.TIES

-------
helped bring about defeat of the bond package.  The factor of low turnout
is not an independent cause however; the disinterest of the uninformed
voter in the package as a whole contributed significantly to this low turn-
out.

This  report is mainly concerned with events surrounding the storm
sewer bond of Oklahoma City,  but it is very difficult to distinguish the
causes  for its failure from those  pertaining more directly to other
issues in the package.  However, the drainage issue which failed can be
contrasted to the construction  of three fire stations, the only issue which
passed.  Fire stations are perceived by voters as important to their own
lives.  Drainage,  on the other hand, is not a very interesting or salient
issue to the average voter.  The presence of storm sewers cannot be
seen  as affecting his life as directly as fire stations,  particularly when,
as those concerned with the bond  proposal point out, the storm sewer
deficiencies were minor in comparison to other bond proposals.

Proposed Bond Package of July 16,  1968

After this election, the mayor, City Council,  and the Chamber of Com-
merce declared that the defeated  bonds were vital to the future of the
city and began efforts to hold another bond election on these  same bonds.
As there were no restrictions  to keep them from doing this,  they chose
a date two months later, July 16,  1968, for the second election.

The Campaign

Opposition remained essentially the same.  As one might expect,  one
new argument being heard was that  the administration was acting "high-
handed" toward the voters in resubmitting the proposals after they had
been defeated once.
                              282

-------
A second criticism of the July,  1968 bond package concerned the pro-
posed fairgrounds bond issue for $1,400, 300.  It was pointed out by the
two prime opposition groups that the fairground was governed by a pri-
vate concern.  "Why, "  they argued,  "should the general public pay for
improvements to some  private group's land? "

The  Oklahoma City Chamber  of Commerce countered the opposition to
the administration's resubmittal of the proposal by appealing to the
voters' sense of urgency,  saying that these proposed projects were des-
perately needed.   The voters failed to let their voice be heard in the
election two months previously, but luckily they were going to get another
chance.  The Oklahoma City Chamber of Commerce said that the project
would not have been brought up for a second time and placed before the
voters of Oklahoma  City if the projects were not desperately needed.
They clained further that a true test of the will of the people had not been
provided in the previous election,  so that the results of the upcoming
election were of paramount importance.

The  Chamber of Commerce chose not to answer  the charges that improve-
ment to the fairground was governed by a private concern.  They tried
instead to persuade  voters to vote for the more widely-publicized sani-
tary sewer system and  the convention center.  Officials privately ex-
pressed the hope that the tide of voters approving these issues would
carry over into other areas and provide a winning margin for all bond
issues.

The  Oklahoma City Chamber  of Commerce substantially changed its cam-
paign tactics for the second election.  A new advertising agency was hired
by the  Chamber of Commerce to put on a high  pressure campaign.  This
agency used every means of advertisement available:  radio, television,
handouts,  flyers,  door-to-door  campaigning,  telephone campaigning,
and public meetings, "the works",  according to the vice president of  the
Chamber of Commerce.
                               283

-------
The Chamber of Commerce also distributed fact sheets to each ward,
listing what specific projects would be undertaken in that particular
ward and how much money would be spent.   Chamber leaders indicate
that this action was necessitated by overwhelming the Black vote
against the previous sanitary sewer bond.  The fact sheets were widely
distributed in the two Black wards.

Election Results

The results of the second election, held on July 16, 1968 are given in
Table 6.  The voter turnout was  increased by about 30% over May, and
all but the fairgrounds bond issue passed.  The electorate voted 3-1 in
favor of most of the bonds, with  the exception of the fairgrounds issue.
It is  interesting to observe this phenomenon in the Black Seventh Ward:
although all bonds were defeated both  times  in this ward, the additional
voters in the July election voted  yes by a 2-1 margin.

The most obvious difference between the May and  July elections was
the existence of a campaign for the  second package. Both opponents
and proponents of the package agreed  that the Chamber of Commerce's
advertising campaign, including  the distribution of ward fact sheets,
changed the outcome of the election.  The campaign clearly specified
the problems and equated them with social needs.   This  decreased the
misgivings that some voters  had about the administration's ability to
plan for the future needs of the Oklahoma City area.  Proponents  stated
that they led an "information" campaign, while opponents stated that the
Chamber's pamphlets were distortions, but both agreed that the above
result was accomplished.

A second difference between the  May and July, 1968 elections was the
feeling which emerged and was voiced by members of the opposition
that voters felt that nothing could stop the city administration from
holding such reelections until the package passed.  Even the Chamber
of Commerce conceded this point and  argued that a quick approval would
save money used to hold more  special elections.
                             284

-------
                 Table 6.
            TA3U1ATIOK OF R2TOHX3
OKLAHOMA CITY SPECIAL SIB CTION, JULY 16, 1368
ST.6 33. ffi'.SSWSg

Fnc?.*' i 2
i".r.S YI3 MO "F.3 "C
1 4751 IS CO SOSS 2685
2 o'iCS 1'
^^\j£-^r^
• '^: LC-?>
_PA3g

10
YJS j!0
4'25 2595
2838 2023
244G 2756
1503 2160
2323 3112
1718 1275
855 1470
3433 1929
19346
17331
36877
c-n-y •JLX^-'
/x'
?AT;?^.:^JD

il
YS3 >'0
3391 2396
264-J 21S7
2276 2015
1333 2231
2C59 3353
1673 1318
825 14S8
3279 2053
18046
18475
35524
^~

"n«'ve~
**^h".- ,.,"»
12
VT<3 —p.
^.--^ ..y
^C/.r, ?7sn
2773 2rsO
2342 2373
1/-50 2?,35
20SC 3369
1-703 "">•>•)
-•' - •J ^wO6
859 1473
3446 1S33
18782
17931
30773


-------
There was unanimous agreement that the sewer and convention center
bonds carried the other bonds—hence,  the passage of the storm sewer
bond issue.
Special Election - June 9,  1970

After the July adoption of all but one of the bonds,  the administration of
Oklahoma City looked  forward to selling the bonds to investors; however,
rising interest rates made the sale of all the bonds at the 5-1/2% rate
set by the City Charter impossible.   The City  Council and the mayor
proposed a special charter election  to be held  early in 1970.   Its pur-
pose would be to propose a raise in  the municipal bond interest rate
from 5-1/2% to 8%.  However, since all bonds from the July 16, 1968
election were not sold, the Charter  also provided that the remaining
bonds must be  placed before the electorate again.  The resulting pack-
age was scheduled for balloting on June 9, 1970, and is summarized in
Table  7.
         Table 7.  Summary of Special Election,  June 9,  1970
                Propositions
       Streets and Bridges
       Sanitary Sewers
       Storm Sewers
       Water Works System
       Solid Waste Disposal
       Municipal Maintenance Facility
       Fire Stations
       Traffic Control
       Airports
       Parks and Recreation
       Municipal Libraries
       Public Transportation
    Cost
$ 8, 168, 000
 32,475, 665
  1, 666, 620
 13, 189,460
    282,300
    200,000
  1, 738, 670
  2, 936, 153
  8, 972, 360
  4,909,560
  1,100,000
  1,624,500
                              286

-------
Campaign
The third campaign for the bond package was again initiated by the
Oklahoma City Chamber of Commerce with the help of an outside
agency.  Although very similar to the second 1968 campaign in its
style,  the 1970 campaign attracted far less interest.  Local officials
and even previous political opponents felt that the voters would be
very reluctant to overturn a previously approved but partially  completed
bond issue.   Indeed,  the Chamber of Commerce underplayed the twelve
bonds  relative to the  charter amendment, preferring to spend  most
of their time on the more subtle charter amendment proposal,  realiz-
ing that it was the key to the package.  No organized opposition emerged
during this campaign, but the conservative opposition apparently main-
tained its posture of  opposition to all urban renewal projects.

As before, the Chamber of Commerce sponsored television and radio
programs promoting  the election.  It conducted door-to-door cam-
paigns, telephone surveys, regional ward fact sheet distribution,  ex-
tensive brochure mailings, as well as  relying on standard techniques
such as posters, billboards, newspaper advertisements and articles, and
free rides to the polls.  Even this extensive effort,  however, did not bring
out the crowds of voters that the second campaign had in 1968.   The re-
sults are given in Tables 8 and 9.
Table 8. Tabulation of Returns
Oklahoma City Special Election
Charter Amendment, June 9, 1970
Ward
1
2
3
4
5
6
7
8
Total
Yes
2,425
1, 701
1,356
776
1,297
1,250
542
2,052
11,399
No
741
562
661
725
816
532
211
669
4,917
                              287

-------
     Table 9.
TABULATION OF RETURNS

wv?n.
l
2
3
4
5
to 6
CO
oo 7
8
TOTAL
YES
NO
GRAXD
TOTAL
ST. 4 BP.
,f 1
YES NO
1836 579
1331 458
1048 538
583 613
961 654
1020 446
432 140
1661 564

8872
3992
12864
SAN. SEWER
2
YES NO
2783 908
1764 643
1688 663
869 862
1524 1092
1117 599
656 223
2123 786

12524
5S81
18 5 OS
ST. SEWER
3
YES NO
2634 846
1668 606
1609 808
789 800
1478 1035
1058 540
613 210
2040 703

11889
5548
17437
OKLAHOMA CITY
WATER SOLID WASTE
4
YES NO
2625 850
1654 613
1580 823
781 794
1447 1052
1050 544
604 211
2007 693

11748
5580
17328
5
YES NO
2650 833
1669 604
1611 798
786 795
1461 1049
1051 537
608 207
2041 708

11877
5531.
17408
SPECIAL ELECTIOI-;, J'JI.'E 9
MAIUT.FAC. FIRE
6
YES JJQ
2554 901
1620 639
1547 850
752 816
1425 1074
1012 565
589 213
1972 739

11471
5797
17268
7
YES NO
2671 794
1666 596
1638 770
803 769
1491 1023
1049 544
601 200
2035 695

11954
5391
17345
, 1970
TRAFFIC
8
YES NO
2637 837
1644 617
1599 803
775 796
1475 1026
1043 544
602 201
2014 704

11789
5528
17317
9
JES NO
2500 951
1577 668
1503 894
731 832
1394 1100
1005 569
588 207
1942 755

11240
5976
17216
PARKS
10
YES
2541
1610
1541
736.
1434
1003
586
1946

11397
NO
911
637
850
817
1062
562
206
746

5791
17188
LIBRARY
11
YES
2502
1580
1406
718
1414
986
594
1904

11194
NO
951
661
890
840
1082
576
205
785

5991
17185
FUBL.1?.V:SP,
12
YES NO
2431 S6S
1577 670
1460 927
737 825
1363 1130
1007 567
600 203
1920 770

11145
6057
17202

-------
Conclusions

Two conclusions can be drawn from the Oklahoma City experience.
First, it is difficult to  convince the voters that they should support
something  for which a need has not been demonstrated, or that they
should support, with public funds, what is essentially a private ven-
ture.  (Witness the rejection of the fairgrounds proposal in both  the
May and July 1968 elections).

Second, there is in the Oklahoma City experience of July 1968 one
of those rare cases where a local government was able to reverse a
defeat at the  polls by an increased level of, campaign activity, even
while facing  an opposition.   In our sample of 40 cases, there were
only two instances of a successful bond election where there was an
opposition; and only one case of a successful election when there
was a high level of campaign activity  in the face of opposition. In
this particular case, it would appear  that the officials of Oklahoma
City accurately analyzed  the failure as being caused not by a  signi-
ficant division  of opinion  in the community, but rather by the gen-
eral lack of any opinion at all.  Thus,  all that was needed was a
campaign by the Chamber of Commerce pointing  out the benefits  to
be derived from the bond issue,  and it passed.
                              289

-------
PETALUMA, CALIFORNIA
            291

-------
                        PETALUMA
This bond issue was submitted to the voters of the Town of Petaluma,

California on January 18, 1966.  The wording of this sanitary sewer
bond issue was:
            "Shall the City of Petaluma incur a bonded
            indebtedness in the principal amount of
            $2, 350, 000 for the  acquisition,  construction,
            and completion of the following municipal im-
            provement,  to wit:  Sewer system improve-
            ments including additions to present  sewage
            treatment plant and appurtenant facilities
            additional to the sewage collection system,
            construction of new pump stations,  collection
            mains, trunk sewers,  outfall lines,  intercep-
            tor  sewers,  force mains, pipes, pumps,
            machinery,  manholes; lands,  easements, and
            rights of ways; and other works, property,  or
            structures necessary or convenient for sewer
            system improvements for the City of Petaluma?"
The purpose of this bond issue was to extend and remodel the present

sewer system  of Petaluma.  A two-thirds majority of those voting was

needed for passage of the proposal, according to the town charter, and

the final vote was 2, 793 (84%) in favor and 520 (16%) against passage.


Population Characteristics of Petaluma


The primary governmental unit involved in this bond issue is the city

of Petaluma, California.  It is a moderate sized  city located approxi-

mately thirty miles north of San Francisco.  In the past,  the area was

heavily dependent on agriculture for its economy.  Now Petaluma is  the

center of the dairy industry of the Northern California-San Francisco

Bay area and supplies eighty percent of all the milk and milk products

for the nine-county San Francisco Bay area.   Although  the number of

people involved in the dairy industry has decreased about 10% over the

past ten  years, this industry still produces over  $35 million per year

in wages.
                               293

-------
Since 1950, the city has been deluged with white- and blue-collar work-
ers from the San Francisco area.  In 1950 the population of Petaluma
was  10, 135,  and by I960 it had increased to 14, 035.  During the decade
between I960 and 1970,  the population of Petaluma increased by 75%
to its present figure of 24, 570.

This population rise has been associated with the rise in employment
opportunities in the Bay area.   Although Petaluma is thirty miles from
San Francisco,  many of its residents travel this distance daily to their
jobs.  It is believed that most of the  Petaluma residents who follow this
pattern are recent  arrivals to the area who have immigrated within the
past ten years.

This mobility of the population  stems directly from the excellent high-
ways provided for the area.   The Redwood Highway, (U.S.  101), the main
artery for northbound traffic from San Francisco,  passes through the
eastern section of Petaluma.   The highway divides the city of Petaluma
into  two parts: the  older section of the city, including the downtown area
to the west and a new and rapidly growing housing development to the
east. (See the map of Petaluma, Figure 1. ) Most houses in this eastern
part of the city have been built within the past ten years.  Residents new
to Petaluma  since I960 comprise most of the development's population.

The  housing  situation in Petaluma is summarized in Table 1.

In I960,  92% of all houses were single family.   This percentage has in-
creased slightly during the past ten years.  There are almost no multi-
family dwelling units in the city.  In  I960,  83% of all houses were owner-
occupied.  However,  the tendency of new commuting residents to rent
rather than own homes  lowered this figure to 71% in 1970.

Table 2 gives the percentages of rental houses in Petaluma within monthly
rental categories.
                               294

-------
Table 1. Housing in Petaluma (1969)
House Classification
New Home (less than 5 years old)
5 to 20 years old - sound
2 to 20 years old - deteriorating
5 to 20 years old - dilapidated
21 to 40 years old - sound
21 to 40 years old - deteriorating
21 to 40 years old - dilapidated
Over 40 years old - sound
Over 40 years old - deteriorating
Over 40 years old - dilapidated
Unclassified
Source: State of California Department
Number
1689
2524
67
29
1111
113
22
1562
271
51
9
of Finance

Table 2. Breakdown of Rental Charges for Rental
Houses in Petaluma
Rent Category ($ per month) % Rental
Under $60
$6l-$75
$76-$100
$101-$125
$126-$150
$151-$175
$176-$200
Over $200

Houses /Category
10%
21%
24%
25%
13%
44%
2%
1%
The median rental is less than $100 per month, which is substantially
less than rentals around the San Francisco area.  This is viewed by
many in  the city as a prime reason for Petaluma's rapid growth in the
past decade.

Forty-five percent of the population is under 20 years of age, whereas
9% is over 65.  This latter figure represents a decrease from the 12%
over 65 in I960.   There also was a slight increase in the numbers  in
the  210 to 34-year-old age group: from 18% in I960 to 20% in 1969.   The
                              295

-------
 RESIDENTIAL

      LOW DENSITY

  is MEDIUM DENSITY

INDUSTRIAL
[• '-'-f''j LIMITED

     GENERAL
Figure 1.  Petaluma Area General Plan
 Source:   LIVINGSTON   AND  BIAVNEY
            CITY AND  REGIONAL  P L A N N E » S
            SAN  FRANCISCO,   CALIFORNIA
— FOUR LANE THOROUGHFARE
— TWO LANE THOROUGHFARE
•- RAPID TRANSIT
— RAILROAD

-------
median age has dropped from 27 to 26 in nine years.  These statistics
reflect the influx of the young families during the 1960's,  a trend which
has created the need for more schools and additions to existing schools
in the city.

Table 3 gives occupation breakdowns and the number of Petaluma resi-
dents in each occupation classification.
Table 3. Occupational Breakdown for Petaluma
(For Principal Wage Earner Only)
Occupation
Professional, technical, and
kindred workers
Managers, officials, and pro-
prietors including farmers
and farm managers
Clerical and kindred workers
Sales workers
Craftsmen, foremen and kin-
dred workers
Operatives and kindred workers
Service workers including pri-
vate household
Laborers, including farm
Retired - unemployed
Not in labor force
Unclassified
TOT
Numb e r
695
694
657
497
1393
630
563
400
928
88
173
^AL 6718

%
10.4
10.4
9.7
7.4
20.8
9.3
8.4
5.9
13. 8
1.3
2.5
100.0

Blue-collar jobs clearly predominate in Petaluma,  but white-collar jobs
are increasing,  Just twenty years ago, the economy of the region was
almost exclusively agricultural.  Now, many more occupations have
been opened,  increasing the heterogeneity of the city's population.   Still,
there are very few non-whites in the area.   The I960 census showed 109
                              297

-------
out of a population of over 14, 000.  The number has increased in the
past ten years, but the percentage has remained constant.

Table 4 gives family income distribution for I960.  The median family
income has increased in the past ten years,  and its increase is greater
than the national increase.  No figures for 1970 are available,  but it is
Table 4. Family Income Distribution - I960
Less than $3, 000/year
$3, 000 to $10, 000/year
$10, 000 and over
19%
62%
19%
Median: $5, 957
thought that the median family income now is around $7, 500, and that
the percentage making less than $3, 000 per year has decreased to
approximately 10%.

Government of Petaluma

Petaluma has  a council-manager government.   The mayor,  elected at
large every four years, heads the city-council and is a  regular voting
member of it.  He has no unique powers which distinguish him from
other members of the council.

The present mayor was elected in 1964 and reelected in 1968, both times
by large margins.  A nonpartisan  mayor, she is the first woman to be
elected to public office from Petaluma.  Her platform during her elec-
tion campaigns had two major points: enlargement of the Petaluma
water and sewer system to handle the increased sewage brought about
by the entry of new residents in the city since I960, and the encourage-
ment  of desirable industry to improve Petaluma1 s economic base.  Al-
though attracting new industry for the purpose of gaining tax dollars and
                               298

-------
new jobs was important,  the mayor believed it could pose hazards
greater than those caused by not having those tax dollars.  Undesirable
development included,  among others, oil companies and certain chemi-
cal companies.   These two planks of the mayor's platform were mutually
reinforcing:  new sewer construction would also be provided for the new
industry which would provide new jobs and more people who would need
new sewers.  These two  planks were  skillfully intertwined in 1964, and
she had no effective opposition.

Petaluma's Sewer System Prior to 1966

The rapid  growth of the area east of U.S. 101 had caused a burden on
the town's  sewer system.  During winter months, (the rainy months for
that section of California) the problem became particularly acute.  A
rising water table caused seepage into the city's leaky pipes and doubled
or tripled  the normal flow volumes.   This resulted in the overflow of
raw sewage from the treatment plant  which  went directly in the Petaluma
River, allowing the unprocessed affluent from the treatment plant to con-
taminate the Petaluma  River was the  only alternative to having it spill
out from manholes right  on the city's main street.

No new treatment plant had been constructed when the houses east of
U.S.  101 were built, and only sewer lines had been provided.  These
were tied into the system of the western part  of Petaluma, and it was
only a matter of time before the1 system's capacity was  exceeded.  This
occurred in 1963.

In a report issued that  same year, the engineers of Petaluma's'General
Comprehensive Plan" noted that sewer lines were inadequate  for the
present population.  Future residential,  commercial and  industrial
growth was limited by the capacity of the  sewer system.   Accordingly,
they mapped preliminary plans for new sewer facilities designed to
handle the  eastern residential development as well as the  sewage anti-
cipated from two areas set aside for industrial development.
                              299

-------
The first of these areas lay between the railroad tracks and the Peta-
luma River.  It consisted of commercial establishments and some
housing, mostly dilapidated and its sewers were presently untrustworthy.
Leaks were quite common,  and ruptures occurred frequently.  The plan
called for a rezoning of this area to industrial use,  a program similar
to urban renewal to refurbish the sector, the repair of existing sewers,
and construction of new sewer lines to handle the proposed new indus-
tries.   The existing treatment plant was to be  remodeled and increased
in capacity.

A  second, more controversial industrial zone  was proposed in the eastern
part of  Petaluma,  south of the new houses,  east of the  intersection of
California Highway 116 and U.S. 101.   The construction of sewer and
water lines was proposed to meet this area's future  needs.  This aroused
some controversy  among members of the 1963 City Council who unani-
mously  felt that sewage lines  for an undeveloped zone were a luxury,  and
therefore not as pressing as the more urgent situation  in the first indus-
trial zone and the new housing development.

The Proposed Bond Issue

After the new mayor took office in  1964, she hired an engineering firm
and a financial advisor for preliminary planning of a sewer bond proposal.
They issued a report before the end of 1965 recommending that the bond
issue include enlargement of the central treatment plant to three times  its
present capacity, construction of a smaller treatment plant for the eastern
sector of town, and city-wide reconditioning of sewer lines.  The contro-
versial  sewer plans for the second industrial zone were dropped, accord-
ing to city officials, in the interests of economy.  They wished to make
passage as easy as possible without having an  unpopular and costly issue
drag the entire bond issue  down to  defeat.

The total project cost was  $2, 350, 000,  to be financed through the  sale of
general obligation  bonds,  which would be repaid with both tax  revenue and
a slight increase in the monthly sewer charges.
                                300

-------
There were two major reasons for recommending general obligation
bonds over straight revenue bonds.   First, property taxes in Petaluma
were very low - only about $3. 00 per $1, 000 of assessed valuation.
City officials  felt that a small increase in the property taxes would not
be offensive to the  voters.  Furthermore, an old general obligation bond
would be retired in 1966, normally resulting in a slight decline in the
tax rate.  Since the sewer bond proposed in the January  election would
follow immediately the retirement of the old bond, a substantial revenue
could be realized from a  very small  tax increase of $. 15 per $1, 000.
Additional income  required to repay  the bonds would be obtained from
a $. 50 increase in  the monthly user  charges.  Finally, it was felt that
general bond  indebtedness would make the acquisition of federal funds
easier that  would  revenue bond indebtedness.

The Campaign

The mayor and members of City Council were restrained by city charter
from taking an active part in the campaign for the sewer, even though
it was general knowledge throughout  the town that they unanimously
supported it.   The  fight for adoption  of the program was  led  instead  by
three independent groups, each of whom had  very similar arguments.

The Sewer Bonds for Progress Committee,  a citizens' group, published
three pamphlets outlining all the arguments which appeared during the
campaign. Briefly, the arguments for passage were:

           Sewer  improvements are needed to handle  increased
           population.
           Sewer  improvements are needed to enlarge treatment
           plant so that  system does not pollute the Petaluma
           River and meets state health standards.
           Sewer  improvements are needed to stop overflowing
           of sewage into streets.
           Sewer  improvements are needed to attract new in-
           dustry  and citizens.
                               301

-------
An interesting feature of the Petaluma Charter allows advertising in
the form of printed ballot arguments to be distributed within polling
places.  The Sewer Bonds for  Progress Committee presented a printed
argument which was approved  by the city clerk for distribution in pol-
ling .places.

A second  group,  the League of Women Voters, coordinated its activities
with the citizens' committee and both groups went door to door, passed
out the three pamphlets, provided rides to the polls,  took out newspaper
advertisements and placed posters in prominent places aroung the city.

A third, less  active,  group was the Sonoma  County Taxpayers Assoc-
iation.  This is a county-wide  organization with branches in each  of the
twelve cities in the county.   Although small  in number, the members
of the  association generally are among the wealthy and influential resi-
dents of the county.  Real estate and farming  interests were known to have
a strong voice in determining the stand of the association.  This group
supported the proposition because of the benefit to land values in Peta-
luma.   However, it qualified support with the stand that taxes  should
not be  raised, and that users'  fees and installation charges should be
utilized instead to repay the bonds.

No group  in opposition to the bond issue emerged.  There was some
opposition,  but none of it was organized.   No particular group seems
to have been more opposed than another,  with the possible exception of
the old people who feared a rise  in taxes  and sewer charges,  but this is
only a   hypothesis advanced by the mayor.

California state law requires a 2- to-1 majority for the adoption of any
bond issue. When the results  were in, it was apparent that this had
been achieved: 84. 3% voted for acceptance.   The  election figures  are
given in Table 5.
                              302

-------
           Table 5.   Election Results of January 18,  1966
         Number of Registered Voters
         Number of Voters Voting
         Total For
         Total Against
         Total Blank
7588
3315  (43.7%)
2793  (84.3%)
 520  (p5. 7%)
   2
On the precinct level, the percentage of voters in favor of the sewer
bond varied 9% around the figure of 84. 3% for the entire city.  No appar-
ent pattern exists, which leads us to conclude that no particular  section
of the city favored or opposed passage more than any other sector  of the
city.

Interpretation  of the  Outcome

Three reasons for passage of the bond issue were offered by the infor-
mants interviewed for this  study.  First,  and most important, was the
professional quality of the campaign.   Persons subscribing to this  view
pointed to the fact that several outside, (and therefore, presumably un-
biased) companies had assessed the Petaluma sewer situation and had
offered their recommendations that the bond be passed.   Other outside
experts at the  sanction of the  citizens' sewer committee and the  League
of Women Voters, presented informative  programs which attempted to
answer all questions posed by the electorate.  Even persons not  associ-
ated with the groups  promoting passage mentioned the fact that the infor-
mation flow to the voters was sufficient to answer most questions about
the bond issue.  Others  attributed the  success  of the Petaluma sewer
bond issue to widespread recognition  of the need for the projects.  Sewage
overflow  in the downtown area was particularly noticeable.  The pollution
of the  Petaluma River was also cited,  but most people interviewed believed
that this was minor in comparison to  the flooding of the downtown area.
                                303

-------
Further,  voters felt that the $2. 3 million bond represented only the
absolute necessities.  No frills were included in this package,  such as
the sewers for the second industrial zone in the eastern part of the town.
The recognition of the need  for the  sewer project is reflected,  respondents
declared, by the high turnout of voters  (43. 7%) for the election.

The final reason cited for passage was  the unity of the city administration
and business leaders, who were represented in the citizens' committee.
Mention was made that one newspaper advertisement listed all important
people within Petaluma as on record in supporting the  sewer issue.   Voters,
our informants felt, became confident of the actions of the city adminis-
tration  when the administration itself possessed strong self-confidence.

Conclusions
Two factors  seem to be most important in the success of the Petaluma
bond election.  First, city officials in Petaluma were united.  There
were no dissident members  of the administration  or the city council
who could legitimize opposition to the election.  Second,  the administration
wisely eliminated the one  possible source of opposition from the pro-
posals for  sewer improvements--the  controversial sewer line to the  in-
dustrial zone in the eastern  part of the city.  Since there was no oppos.i-
tion, the administration did  not need to campaign, and the bond  issue
passed  easily.
                               304

-------
ROUND ROCK,  TEXAS
         305.

-------
                        ROUND ROCK


The $281, 000 sewer bond proposition which was placed before the elec-

torate of Round Rock, Texas on December 5,  1968, appeared as follows:


                         PROPOSITION NO. 2

            "Shall the City Council of the  City of Round
            Rock, Texas, be authorized to issue the bonds
            of said City in the Amount of $281, 000, maturing
            serially over a period of years not to exceed forty
            (40) years from their date,  bearing interest at
            such rate or rates as shall  be determined within
            the discretion of said City Council,  for the pur-
            pose of  constructing extensions  and improvements
            to the City's sewer system  to be issued in accor-
            dance with and secured in the manner provided  by
            Article  IIII,  et seq V.A. T.  C. S. , as amended,
            each bond to be  conditioned that the holder thereof
            shall never  have the right to demand payment of
            said obligation out of funds  raised or to be raised
            by taxation,  and secured by a lien on and pledge
            of the net revenues from the operation of the City's
            combined waterworks and sewer system? "


The vote on this proposition, involving  both the construction of new

sewer lines and pumping stations and the replacement and recondition-

ing of old lines, was 146 (43%) in favor  and  196 (57%) against passage.


Population  Characteristics of Round Rock


Round Rock, the primary governmental unit for this election,  is a city

of 2, 687 people located  in the middle of Williamson County,  as shown

on the map  following.    Williamson County lies in the Blacklands and

Grand Prairie areas of  central Texas.


Following its  establishment in 1850, Round  Rock was predominantly

white.  Slavery did not exist in the  area at any time and residents were

mainly small  farmers and ranchers.  After the Civil War,  however,
                              307

-------
                                                          \
KOUMD  ROCK COHERE HENSIVE  CITY    PLAN
AUSTINECONOMICREGION
                    308

-------
Blacks from the deep South immigrated in small numbers to provide a
labor force for the increasing agricultural production in the area.  They
were followed by Mexicans after the turn of the century.  By 1930,  the
relative proportions  of these three ethnic groups were established,  with
each comprising close to one third of the total population.  This ratio
has  changed little to date.

The population of Round  Rock from 1920 to 1970 is presented in the
table below.
Population of Round Rock, Texas
19
-------
ponsible for a large part of the city's growth, particularly in the past.
The major employer in Round Rock is the Round Rock Lime Company
which currently employs 100 persons and,  according to the personnel
director, anticipates an increase in employment in the future.
Table 2. Tax Rates for Selected Texas Cities
City
Austin
Bartlett
Georgetown
Hutto
Round Rock
Strawn
Taylor
Total Assessed Assessment
Valuation Rate (%)
$808, 812, 140
1, 303, 200
10, 771, 880
5, 855
1, 752, 520
612, 388
17, 921, 514
67. 5
20
60
50
21. 7
33.3
50
Total Tax Adjusted
Rate ($/$!, 000) Tax Rate
12.90
10.00
6.50
10. 00
15. 00
14. 50
10. 00
8. 71
2. 00
3.90
5. 00
3. 25
4. 83
5.00
Geographic Features of Round Rock

The Balcones Fault,  running north-south through Williamson County,
divides Round Rock into two distinct land areas and contributes directly
and indirectly to the cost of developing the area.  New construction is
limited by the difficulty of finding adequate support for foundations.
Sanitation problems are greatest in this area, due to seeps and springs
which result in a low absorption rate and substantial clay deposits.
Consequently, septic tanks in this area are unsuitable.  Principal  drain-
age features in the County include Brushy Creek, which  passes through
northern Round  Rock, Berry Creek,  and the  San Gabriel River.  Brushy
Creek and Lake Creek,  which meander through the city from west to
east,  contribute  significantly to the cost and  difficulty of development
by limiting accessibility from one area to another.   Together with the
Balcones Fault,  they have contributed to expensive scattered patterns
of development.  In addition to these constraints, the western side of
Round Rock is limited to industrial development because of the dust
from limestone  mines in the area.
                               310

-------
Housing Conditions in Round Rock

Even though housing densities  are generally low in Round Rock, three
residential areas of the city are considered overcrowded.  These areas
are the Flats  (marked Section  I on the map in Figure 2), portions of
Old Town (Section II) and newly developed areas along Lake Creek in the
southern half  of the city (Section III).

The  Flats is in the older  downtown area, bounded to the north by Brushy
Creek.  This  sector of the city is of importance to our  study because
opposition to the sewer bond issue emerged most strongly here.  Its
population is a mixture of minorities, predominantly Mexican-Americans,
but with some whites and Blacks.  There is a noted absence of people in
the 19 to 35 year age group and a preponderance of middle aged and senior
citizens.  Incomes are low and families are large.

The  Old Town population  distribution is similar to that  of the Flats.   The
problem of overcrowding in this area contributes to frequent failures in
the sewer system which are already caused by  seepage and underground
springs.

The  area  near Lake Creek and the railroad tracks is inhabited by Mexi-
can-Americans,  whites and Blacks.  They are  generally migrant workers
whose family incomes fall well below poverty level.  Families are often
very large and may consist of  three or four generations, with as many
as a dozen children in each generation.
                                          I
There is a Federal low-income housing development program in the ex-
treme southern part of the city known as Egger's Acres.  It consists
largely of overcrowded multi-family residences.   The small sewer line
leading to the  area is generally considered inadequate for the population.
                              311

-------
The city's white population can be found almost exclusively in two dif-
ferent developments.  One, in the far western part of the city, is be-
coming  less desirable to residents because of its proximity to a lime
plant  to the  south.  This area is being slowly vacated by the whites as
they move in increasing numbers to the  second development, a large
housing development north of U.S.  Highway 79 in the extreme north
part of the city.

The following housing  characteristics were revealed by the I960 Census,

      (1)    Ninety-two percent of the dwelling units were occupied
            and 8% were vacant.  Of the occupied dwellings,  65%
            were owner-occupied and 35% were renter-occupied.
      (2)    Non-whites are most likely to be renters, and whites
            are overwhelmingly single-family home owners.
      (3)    The average number of  rooms in an owner-occupied
            dwelling unit was 4. 9, and the average number of
            rooms  in a renter-occupied  dwelling was 4. 2.
      (4)    The average value of owner-occupied dwellings was
            $7, 000.
      (5)    Seventy percent of the dwelling units were classified as
            standard,  14% as substandard, and 16% as dilapidated.
      (6)    Of the above,  17% of the standard units were deteriora-
            ting,  and 66% of the substandard units were deteriora-
            ting.
      (7)    Non-whites, who comprise over one half the population,
            occupy 11% of the housing supply.
      (8)    Of the houses considered dilapidated, 59% were occupied
            by non-whites.

Government
Round Rock has a strong mayor-council government.  The current
mayor was elected last year in a rather close election after serving for
five years on the City Council.  The government structure also includes
a position known as city administrator.  This officer has no formal power
under the city charter, and his duties incorporate those traditionally be-
                              312

-------
longing to the city secretary, city manager, department of public works,
and city clerk.  He stores city records, handles city correspondence,
oversees city sewer, water, and  garbage  service, and in general, is the
man most, up-to-date on all aspects of the city government.

This system is efficient for the city, but does have some undesirable
consequences.  The city administrator often finds himself explaining and
justifying a particular action of the  mayor and city council,  an action for
which he is not responsible and over which he has no power.  The present
mayor tends to divert citizens' questions from himself to the administrator,
and as a result, the electorate feels it is  insulated from the government.

Previous Attempts to Build Sewers

The  sewer and water proposal was first placed before the voters on
November  4,  1967,   In that election, four questions  detailed below were
placed before the voters.

            Question 1
            The issuance of $50, 000 in revenue bonds for
            extensions  and improvements  to the waterworks
            system.
            Question 2
            The issuance of $180, 000 in revenue bonds for
            extensions  and improvements  to the sanitary
            sewer system.
            Question 3
            The issuance of $25, 000 in general obligation
            bonds for street improvements.
            Question 4
            The issuance of $25, 000 in general obligation
            bonds for construction of a new public library.
                              313

-------
During this  election,  the public was poorly informed on the critical
need for sewer and water projects.   Little information from city hall
was forthcoming; in fact, neither a map nor a description of the pro-
posed improvements  could be found at the city offices.  Consequently,
the public's desire for information was  satisfied by the efforts and
publications of the bond issue's opponents.

Apparently, most of the funds were to be spent in the large housing
development north of the town,  an area  inhabited mostly by whites.
Although the development originally had its own sewer collection sys-
tem, it was not tied to the main system of Round Rock.  Subsequent
additions to the settlement were not serviced by any sewer lines at
all.  While water lines  to all houses existed, they were considered in-
adequate to  handle above-average flow volumes  experienced during
most of the  summer months.

Groups for and against passage of the sewer and water proposal
coalesced quickly.   The buildings of the housing development (the
bank president and two  members of the city council) formed the nucleus
of the group promoting  passage.  The opposition was a citizens'  group
made up of lower-income residents of the central and  southern sections
of town.  A  third faction, led by another councilman (who later became
mayor) opposed only the last two questions in the package  on the grounds
that they would increase taxes.  Both the proponents and the opposition
citizens' group regarded this move as strictly political in  nature.  They
charged that he was planning his campaign for mayor in two years and
did not want to be known as responsible for any tax increase.   Further-
more,  much of his financial background would come from  the wealthy
builders,  and  much of his popular support would come from whites,
who would benefit from the  sewer and water proposals.

Citizens' group voiced its opinion that other sections of the town and
other racial groups were being  slighted in the allocation of improvements.
They noted that an overwhelming proportion of the money would benefit
about 25% of the town's population, and  the town's  most affluent white
population at that.

                               314

-------
The builders countered saying that the situation was most critical in
the northern section of town.   One builder stated  that this project was

only the beginning, and that the Mexican-American and Black areas
would be served by future bond issues.


The  general concensus seemed to be that the builders and real estate

interests, who had considerable influence on Round Rock's government

were planning to pass bond issues one at a time,  knowing that they could
not hope to  pass a large bond issue which would cover all the needed

city  improvements.   But the voters felt that they  were not being given
all the facts, and became suspicious of the builders' motives.  By the

time of the  election,  it was clear to most observers that the bonds  would
not pass.  The results are  given in  Table 3.
Table 3. Election Results of November 4, 1967

Question 1
Question 2
Question 3
Question 4
(water):
(sewer):
(streets):
(library):
For
Against
For
Against
For
Against
For
Against
106
176
107
174
108
174
78
203
Why did the election fail?   The observers of the first election in 1967
have cited the following reasons for the failure of the water and sewer

bonds:
      (1)    The city council was split on the question of
            passage of the package as a whole.  While the
            mayor and council supported the passage of
            the water  and sewer bonds,  the ambitious
            councilman and his group did not support the
            passage of the streets and library questions.
            The voters were primarily concerned with /
            the split among the council members on the
            bond package.   Consequently,  the  council's
                              315

-------
      unity on the water and sewer issues failed to
      ease the voters' misgivings regarding the
      total package.  Moreover,  the voters' mis-
      givings about the city administration were exa-
      cerbated by the overt political maneuvers of the
      ambitious  council member.  The leader of the
      citizens' group felt that this man's desire to
      become  mayor alienated many voters who thought
      that personal politics were out of place in a bond
      election.

(2)    The critical need for new sewage facilities was
      not publicized by groups supporting passage  of
      the water and sewer bonds.  Consequently, claims
      that the  funds for improvements would not be dis-
      tributed equally over the town or among the racial
      groups had a  strong influence on the voters'  decision-
      making.

(3)    Both proponents and opponents  agreed that the pub-
      lic was confused about the  financing of the bond
      package.  The presence of both  revenue bonds and
      general  obligation bonds seemed to confuse the voters
      because they  did not understand the differences
      between the two.  Knowing that one would raise pro-
      perty taxes directly and that the other would not,
      but forgetting which was which,  the voters tended to
      vote against all of the issues.

(4)    All groups agree that the library issue aroused
      opposition from voters who thought it was an unneces-
      sary frill for a town only a few miles from the lar-
      gest library in  Texas.  Its presence on the ballot gave
      a negative air to the entire package.

(5)    The racial issue, while difficult to evaluate, clearly
      played a deciding role in the  rejection of the package.
      The residential lines in the community are clearly
      drawn, with few whites living in the Flats and the
      southern part of the city.   Likewise, few non-whites
      live anywhere else.  The history of the city in fighting
      Federal integration guidelines coupled with the city
      council's unwillingness to authorize the  spending of
      funds outside of areas occupied by whites is cited by
      most observers as a dominant reason for failure.
                         316

-------
The Sewer Bond Election of 1968

The question lay dormant for one year.  During this time the ambitious
councilman appears to have been promised the builders' aid in seeking
the office of mayor, since the incumbent was planning to retire.   He
was pledged to support the entire package  of bonds, and he in turn ob-
tained the cooperation of the builders in replacing the general obliga-
tion bonds with revenue bonds.  It seemed to be possible to finance the
street improvements from revenue bonds  (even though the streets gen-
erate no revenue) by combining revenues from the  city water and sewer
system and paying off the streets bonds from this fund.  The unpopular
library bond was dropped from the ballot which was then presented to
the voters on December 5,  1968, as follows:

            Question 1
            The issuance of $25,000 in revenue bonds for
            extensions and improvements  in the waterworks
            system.
            Question 2
            The issuance of $281,000 in revenue bonds for
            extensions and improvements  in the sanitary
            sewer  system.
            Question 3
            The issuance of $50, 000 in revenue bonds for
            street  improvements.

This time only the citizens'group was leading any opposition.  The builders,
the mayor,  and all but one member of the  city council were in favor.
Many arguments against the project were  the same as in 1967.   The forces
in favor argued that the  improvements  were desperately needed,  and the
citizens' group continued to oppose what they considered to be an unfair
apportionment of the funds.
                              317

-------
New arguments against the bond issues were more specific in nature
than arguments presented the previous year. Specific examples of
pressing needs which would not be alleviated by the passage of the
bond issues were cited.  An elementary school, located on the clay
bed of the western half of Round Rock and attended mostly by non-
whites,  would not benefit from the improvements.  For years the school
grounds had been flooded  by sewage overflow resulting from poor seep-
age.  It was also noted that sewer lines in the Flats were located above
water lines and that their frequent rupture resulted in contamination of
water being delivered  to the houses  in this area.  No  remedial action
was outlined in the  proposed bond issue for either of these health
hazards.

The citizens'group  was indignant that the group of men who had built
the northern housing unit were the same ones decrying the poor sewage
and water systems  in that area.  More specifically, charges of "wheeler
dealing" and profiteering at the expense of the city of Round Rock were
leveled  at the builders' group. The peoples' group further objected to
the arrangement between the mayoral candidate and the builders; they
were well aware that the candidate had the builders' interest at heart
since they were  supporting his campaign.

New arguments in favor  of the bond  package refuted the arguments of
discrimination in the allocation of funds that had been used so effective-
ly the previous year.  They noted that a large part of the sewer bond
money (25%) was being used to add a new pipeline to Egger's Acres.
Further, they cited growth figures over the  past five years and stated
that future growth would be impaired by failure to adopt the  new sewer,
water and street proposals.  New housing would have to be constructed
to ensure progress, they said, because only 2% of the housing  stock
was vacant.  This housing,  they said could not be served by the present
inadquate system.   Finally, they said that they had removed most ob-
jections by promising  no increase in taxes.
                              318

-------
 The citizens' organization's reaction to this was to point out that Egger's
Acres was also a development of the real estate interests of the town.
 The same charge of profiteering at the expense of the city was renewed
against the banker who, they said,  had purposely built inadequate sewer
and water lines to save himself  money.  On the second point, they chal-
lenged the goal of "progress".   They accused the administration and
builders of blindly leading the city down a path with only the nebulous
goal of "progress" as the end result.  They demanded further clarifica-
tion as to the nature of "progress".   Finally, this group stated that while
the tax rates would not rise,  the increased  water and sewer taxes would
be regressive in nature,  affecting the poorer non-whites more drastically
than the white population.

Vehicles for communicating the  arguments  for and against the bond issues
included public debates,  posters, telephone surveys, door-to-door  canvas
 sing   (by the citizen's group),  newspaper advertisements and editorials.

When the vote was in,  it  was clear  that the  citizens,1  campaign had been
the more effective. The results are given below in Table 4.
            Table 4.   Election Results of December 5,  1968
      Question 1 (water):           For                    145
                                   Against                200
      Question 2 (sewer):           For                    146
                                   Against                196
      Question 3 (streets):          For                    175
                                   Against                166
When the subject of the 1968 election was posed to the people who were
interviewed, two new reasons  were suggested for the failure of the water
and sewer bonds.

First, while the mayor, builders and majority of the city council includ-
ing the mayoral candidate had  taken a unified stand in support of the bond
                              319

-------
package, their unity now appeared before the voters'  eyes as a demon-
stration of the collusion which had been hinted at in the 1967 election
campaign.   Effective arguments were made by the opposition that the
builder and the real estate developers in the city government were res-
ponsible for developing the areas  which were cited as desperately in
need of improved water and sewer facilities.  Intense indignation was
to be found in Egger's Acres, where the residents felt that the govern-
ment funds utilized to construct the housing project had been used in-
effectively, resulting in the present undesirable situation.  They blamed
the builder for this and followed their spokesman's  suggestion that the
builder himself be required to remedy this situation to which he had
contributed.

Second, the point that the increased water and sewer  rates would be
regressive in nature was  conceeded by both proponents and  opponents
to have had some impact among the Mexican-Americans and Blacks of
the town, which caused them to vote against passage of the water and
sewer  bonds.   This point may be valid,  but seems to be a relatively
minor  issue in the broad picture of the Round Rock water and  sewer
defeats.

Mayorality Election of 1969

The Round Rock water and sewer  problem entered an important phase
in its  resolution during the mayorality election  of 1969.  The city coun-
cil member, backed by the builders,declared his candidacy  soon after
the second  bond defeat, and was shortly followed by the leader of the
citizens' group.  This put the forces favoring and opposing the water
and sewer bonds in direct competition.

After a fairly lengthy campaign,  three issues stood out as deciding fac-
tors in the  election:  (1) the passage of the water and  sewer bonds; (2)
the nature and desirability of "progress"; and (3) the fitness of the
citizens' group spokesman to be mayor.  This candidate had carried
                              320

-------
over a strong position from the 1968 sewer campaign, but  his opponent
introduced doubts about his ability to run the city since he had been fired
from the school system on the grounds of incompetence.  Observers
agree that,  had the balloting between the two men rested on the issues
of the water and sewer bonds and "progress", the candidate in question
would have won.  However,  the question of his capabilities played a
decisive role  in the  decision-making of the voters, and the city council-
man won in a  close vote.

Sewer Bond Election of 1970

One of the first acts  of the new administration was a plan for rezoning
the city which would have  some effect on future  sewer and water bond
plans.  In line with his plan for "progress",  the new mayor moved to
change the designation of the relatively unpopulated land north and west
of Egger's Acres and near the highway from "residential" to  "industrial
and commercial" in  order to lure more industry to Round  Rock.   Two
families who  lived in the area and the residents of  Egger's Acres ob-
jected on the  grounds that the proposed industries would be too near
their homes.  A new citizens' committee was formed which was  able
to persuade the council to change its plans.  Spokesmen for this  com-
mittee said that they were assured by the leadership of the town  that no
industry would be able to locate east of Interstate 35. The classification
was subsequently changed to the more desirable commercial  zone.

The effect of  this change of zoning had a direct impact on the future of
the water and sewer question in the city.   By placating the  citizens who
had previously followed his  opponent, the new mayor enlarged his own
political base and had greatly enhanced its position in the community.
The fact that  the loser took no part in the zoning debate, when coupled
with his  recent defeat,  was  seen to indicate his demise as a political
force in the town.
                               321

-------
Soon after the zoning debate, the mayor and the city council, all of
whom had supported him in the mayorality election,  drew up plans
for a new water and sewer bond election to be held on May 23,  1970.
The amounts of the bonds were substantially increased, reflecting, an
administration official stated, their confidence in their ability to secure
passage for the entire package.   The package that was presented to the
voters is given below:

            Question 1
            The issuance of $315, 000 in revenue bonds for
            extensions and improvements to the waterworks
            system.
            Question 2
            The issuance of $635, 000 in revenue bonds for
            extensions and improvements to the sanitary
            sewer system.

In a further attempt to respond to former criticisms of inequity in the
allocation of improvements to various sectors of the town, the  new
council added to the bond package of 1968 funds proposed to take care
of the improvements to the elementary school where sewage overflowed,
repair to the central city's faulty water and sewer lines, and expansion
of service to Egger's Acres and other areas south of the railroad tracks.
A detailed map of the improvements attached to a circular describing
the nature of the  improvements was delivered to every house in the com-
munity.

The defeated mayoral candidate again opposed the water and sewer bonds,
but found his support greatly decreased from previous years.  His  stand
was that the mayor and city council were in collusion,  but his most per-
suasive arguments  against passage were no longer relevant, and he was
unable to develop any significant new reasons to vote against the bond.
                               322

-------
There were no circulars or posters of opposition to the bond.  Speeches,
primarily by the unsuccessful candidate were the only means of con-
veying the message of opposition.  The results are given in Table 5
below.
Table 5.
Question 1

Question 2

Results of Election of May 23, 1970
(water):

(sewer):

For
Against
For
Against
267
111
269
109
 Conclusions

 The Round Rock sewer issue is an excellent example of a case where
 local officials presented a bond issue which was inequitable in its dis-
 tribution of benefits and costs,  and where the voters who were being
 discriminated against recognized this fact and defeated this proposal
 at the polls.  In effect, what the officials  had proposed in the 1967
 election was that the entire town should support a project which would
 benefit  only the affluent white population of the town, while important
 sewer needs of the Black and Mexican-American populations were ig-
 nored.  On the other hand, once these inequities were eliminated, the
 opposition to the bond issue was significantly reduced and the bond
 issue passed easily in 1970.
                                323

-------
                       ACKNOWLEDGEMENTS

This study has been accomplished through the efforts of a large group of
people.  The project staff is especially indebted to the local officials and
other community leaders in the forty cities which were studied who took
time to answer our endless questions and direct us to additional  valuable
sources of information.   Without their cooperation, the  study could not
have been accomplished.

The Investment Bankers Association of America,  and in particular, its
research director Mr. Robert King, provided us with the data with which
to identify a sample and evaluate the extent to which the referendum poses
an obstacle to municipalities  in financing of water pollution control.

The research was carried out under the direction of Ovadia Salama,
and supervised by Robert H.  Rea.  Ralph Jones and Holly Kinley are
largely responsible for the interpretation of the results  and the prepara-
tion of this report.  Others who made important contributions  to the
research were John Durkey,  Ron Patton and Kenneth Marr.  The support
of the  Environmental  Protection Agency and the assistance of  Robert
Brown and Jack Daunt are acknowledged with thanks.
                               325

-------
                             REFERENCES
          1 U. S. Department of the Interior,  Federal Water  Pollution
Control Administration,  The Cost of Clean Water , January,  1969, p. 1.

          2U. S. Department of the Interior,  Federal Water  Pollution
Control Administration,  The Economics of Clean Water,  Summary
Report,  March 1970,  p.  5.

          -^For a list of all the  Federal programs providing  grants for
construction of water pollution control facilities, see National League
of Cities, Federal Aids to Local Governments.

           U.S. Congress, Joint Economic Committee, State and Local
Public Facility Needs and Financing,  Vol.  1,  Dec.,  1966, p. 178.
                Economics of Clean Water,  Summary Volume, March
1970, pp. 30-31.

          "Controller General of the United States,  Examination Into
the Effectiveness of the Construction Grant Program for Abating,
Controlling and Preventing Water Pollution, Federal Water Pollution
Control Administration,  Department of the Interior,  Nov.  1969,  p.  2.

          7There are, of course, other means of financing.   If the
project is small,  for  example, the municipality may wish to finance it
through short-term public borrowings, revolving construction funds  or
sinking funds related  to previous borrowings,  or direct appropriations
(i. e. ,  the "pay-as-you-go" approach).   The last form of financing is
particularly popular for  small additions or renovations of existing
systems.   See U.S. Congress, Joint Economic Committee, op. cit. ,
p. 148. Also,  private land developers and homebuilders often undertake
capital outlays for sewage collection systems.  Indeed, according to
data collected by the  Joint Economic Committee  of the Congress, 25%
of all sanitary sewer  construction has been undertaken by private home
builders.   Loc. cit. , p.  147.

          ^U. S.  Department of the Interior,  Federal Water  Pollution
Control Administration,  The Economics of Clean Water, Volume I,
Detailed Analysis, March 1970.

          9Dwaine Marvick, The I960 California Water Bond Issue and
the Los Angeles  County Electorate:  A Statistical and Historical  Analysis,
University of California at T os Angeles,  1967.   (unpublished)
                                 327

-------
          l°Donald J. Robinson.,. "Financing Water and Pollution Control
Facilities, "  Investment Dealers  Digest, Section II, May 27, 1969, p. 35.
            Ibid.
          13 The theory and procedure for systematic sampling is
described in Appendix C.

          !4James Q.  Wilson and Edward C.  Banfield,  "Pub.lic-.Regard
ingness as a Value Premise in Voting Behavior, " American Political
Science Review,  58 (December,  1964),  pp.  876-888.
                  of these same leaders found out that their assumptions
were not always correct.  For example, in Oklahoma City, the residents
of two predominantly poor, non-white wards opposed the. May 1968
bond issue because it did not  provide  for either the complete  renovation
or the removal of a noxious treatment plant which was  located in proximity
to several schools and many homes in those wards.  This opposition
caught the proponents of the bond issue  unaware,  since these wards had
traditionally provided substantial support for bond issues.

           l6This  problem was first raised by W.S. Robinson in an
article entitled:  "Ecological  Correlations and  the Behavior of Individuals, "
American  Sociological Review,  15 (June,  1950), 351-357.  There are
means of minimizing the problems involved in  the use of aggregate data.
See, e. g. , W. Phillips Shively,  "Ecological1 Inference:  The Use of
Aggregate Data to Study Individuals, " American Political Science Review,
63 (1969),  pp.  1183-1196.   A.n example  of a recent study which uses
aggregate  correlations to infer  individual-level relationships, is Wilson
and Banfield's "Public Regardingness as a Value Premise in Voting
Behavior, " op. cit.  One maj or difference between the analysis of
Wilson and Banfield and the analysis  of this study is  that Wilson and
Banfield studied the voting behavior of carefully selected homogeneous
neighborhoods, each containing individuals with a particular  income level,
ethnic group, etc.

           l?See Gerald T.  Slatin, "Ecological Analysis of Delinquency:
Aggregation  Effects,"  American Sociological  Review, 34 (December, 1969),
pp. 894-907.
           I Q
           1 "Robert Wood, Suburbia   (Boston:   Houghton Mifflin Company,
1958), pp. 178-179.
                    L.  Grain, Elihu Katz and Donald B.  Rosenthal,
The  Politics of Community Conflict, (Indianapolis:  The Bobbs-Merrill
Company,  Inc. , 1969).

-------
                    Couch and  Kenneth Keniston, "Yeasayers and.Nay>
sayers:  Agreeing Response Set as a  Personality Variable, " Journal of
Ab normal and Social Psychology,  (March, 1950), pp. 151-175.

           21 Robert L. Grain,  Elkhu  Katz and Donald B.  Rosenthal,  The
Politics of Community Conflict, (Indianapolis: The Bobbs-Merrill  Company,
Inc. ,  1969).                    '

           22 John E. Mueller,  "Voting on the Propositions:  Ballot
Patterns and Historical Trends in California, "   LXII American Political
Science  Review (December, 1969), pp.  1197-1212.

           23See, for  example Eugene C.  Lee, The Politics of Non-
partisanship: A Study of California City Elections (Berkeley: University
of California Press, I960) and Robert C. Wood,  Suburbia:  Its People
and Their  Politics (Boston:  Houghton Mifflin Company, 1958).
                    op.  cit. ,  p. 155.
           25
             See Chapter I for an explanation of these data.  U. S.  Depart-
ment of the Interior, Federal Water Pollution Control Administration,
March  1970.
           7 h
             This is feasible  from a market point of view since high rates
are based on political default and difficulties the trustee then encounters
in securing bond payment revenues.  But the state would quickly and forci-
bly collect revenues through  local taxation to forestall default,  and so the
high rates would not be necessary.

           27
             A number of the jurisdictions which require referenda for
whatever purpose specify that the  approval majority exceed 50%, either to
60% or  to 2/3.  In Lance v. Board of Education,            W. Va.
 170 S. E.  2nd 783 (1969), the highest West Virginia court held such a pro-
 vision in the West Virginia Constitution to be repugnant to the Fourteenth
 Amendment of the United States Constitution in that it deprived those voting
 pro of votes  equal in weight to those voting con. The case will be heard by
 the United States  Supreme Court,  no citation as yet.
           28
             Special districts are actually broken down into two quite distinct
 political entities.  One is the "subdivision" of a municipality, designed to
 provide area sewer service within a town which does not require service
 throughout the town.   Politically,  it is usually a servant of the town govern-
 ment.  The other is the "authority, " the large inter-city arrangement of a
 multiplicity of sewer  systems, designed either in congested areas to serve
 efficiently inter-community populations or in rural areas to function on a
 county-wide  or inter-county basis.  This type of district has its own govern-
ing body which usually operates in a state of relative autonomy.
             Both types are legally considered municipal corporations, how-
 ever; and, in terms of referendum law,  no distinction is drawn between the
two types. Where the district was a simple adjunct of the county government,
it was listed as a county.  Where it was a simple adjunct of the municipal
government,  it was listed as a district,  since most small special districts fit
into this category.

                                  329

-------
            Bond percent also refers to two terms--rate of return (or
face rate) and cost of bond.  For the most part, statuatory bond per-
cents are face.percents.  This is the maximum percent the issuer is
allowed to pay out to the bondholder as interest.  Some states also limit
the cost of the bond.  Cost of bond includes interest,  trustee charges
and issuing expenses.   The face percent will be the major part, but not
all of the  cost percent except where the issuer is allowed to issue addi-
tional bonds to cover expenses.  In such a case, the percents will be id-
entical.  There are frequently provisions which limit the amount of the
issue to the cost  of the project.  Designed to prevent arbitrage, these
provisions usually  have the undesirable side effect of lowering the face
rate where the percent ceiling is a cost rate.  None  of these constraints
are very significant.

          30
            The  Investment Bankers Association of America, in pre-
paring our list of referenda, included votes of town meetings.  The
apparent reason for inclusion of town meetins was the similarity between
these meetings and referenda (e. g. ,  both are  commonly considered to be
examples of "direct legislation").  Such similarities are most evident in
the caseof simple town meetings where every citizen may vote,  as opposed
to r ep r e sentative town  meetings where only selected representatives may
vote.  Both types certainly encourage participation by local residents in
community decision-making.

            Economic  Development Associates, Inc. , "A Summary Report
on the Athol Comprehensive Plan, " (mimeo),  p. 1.

            Ibid. ,  p.  3.
          33
            See "Reports of the Town  Officers of the Town of Athol,
Massachusetts for  the Year Edning December 31,  1966," pp. 44-49.
          34
            All details on the form of  government in Athol are from the
"By-Laws of the Town of Athol, Massachusetts," March 1, 1951,  as amended
to June 23,  1969.
          35
            All data used in this paragraph are derived from this bond pros-
pectus.
          0/1
            This is the basic chronology reported by the mayor at that time.
The  records in the  city clerk's  office in Avon are incomplete, therefore it is
not possible to substantiate fully the  events which took place in the early
1960's. All materials in the following  section on the bond elections are drawn
from data available from the city clerk's files and from recollections of in-
formed observers.

          37
            DeKalb County Planning Department,  Preliminary T^nd Use
Plan, DeKalb County,  May 1970,  p.  29.
                               330

-------
          38Ibid. , pp. 29-32.


          39
            Furnished by Economic Development Department, DeKalb

Chamber of Commerce.


          40
            Preliminary Land Use Plan, op.  cit. ,  p.  33.


          41
            League of Women Voters of DeKalb  County, "Housing Survey

of DeKalb County, Georgia, "  February 1970.


          42
            Preliminary Land Use Plan, op. cit. , p.  32.


          43
            Ibid. , pp. 34-35.


          44
            Floyd Hunter,  Community Power Structure.  (Chapel Hill:

University of North Carolina Press, 1953. )
                              331

-------
1

5
Accession Number
/% Subject Field fa Group
o6c
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
A~W?'1fss'ociates Inc.
Cambridge, Massachusetts  02138
      Title
       Factors  Affecting Pollution Referenda
  10
      Authors)
                                 16
                                    Project Designation
                                      21
                                          Note
                                          This report was submitted in fulfillment of
                                        Contract #14-12-902 under the sponsorship
                                        of the Office of Water Programs, EPA.
  22
      Citation
  23
      Descriptors (Starred First)
       *Referenda,*Bonds,*Election,5Tax rate,*User Charges,*Bond Market,^Public
 Expenditures, ^Increased Taxes > ^Government Finance
  25
      Identifiers (Starred First)
  27
Abstract
Man;
        y local governments wanting  to finance construction of water pollution control
facilities through the sale of municipal bonds are required to obtain  authorization
from the electorate in the form of  a bond referendum.  Since defeat of such a
bond issue can result in significant delay in implementation of the Federal Water
Pollution Control Act, it is important that EPA understand the factors which affect
the  outcome of these elections.
The  study contains conclusions and  recommendations about the process by which the
project is planned, the role 6£ the bond market, and the referendum as an approval
mechanism for public expenditures.
 Abstractor
                                Institution
   WR:I02 
-------