OKDES
SITE-SPECIFIC SOCIOECONOMIC IMPACTS:
SEVEN CASE STUDIES IN THE
OHIO RIVER BASIN ENERGY STUDY REGION
PHASE II
OHIO RIVER DASIN ENERGY STUDY
-------
SITE-SPECIFIC SOCIO-ECONOMIC IMPACTS:
SEVEN CASE STUDIES IN THE
OHIO RIVER BASIN ENERGY STUDY REGION
By
Steven I. Gordon
Anna S. Graham
The Ohio State University
Columbus, Ohio 1)3210
Prepared for
Ohio River Basin Energy Study (ORBES)
Grant No. EPA R805589
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20^60
-------
Site Specific Socioeconomic Impacts:
Seven Case Studies in the ORBES Region
By
Steven I. Gordon and Anna S. Graham
Department of City and Regional Planning
The Ohio State University
Columbus, Ohio September, 1979
-------
Table of Contents
Section Title Page
1.0 Introduction 1
2.0 Review of Related Literature on Local and
Regional Socioeconomic Impacts of Power Plants . . 3
2.1 Introduction 3
2.2 Socioeconomic Impacts 3
2.3 Classifying Potential Socioeconomic Impacts. ... 5
3.0 Description of Case Study Areas 10
3.1 Jasper County, Illinois 17
3.2 Jefferson County, Indiana. ... 17
i
3.3 Spencer County, Indiana 18
3.4 Trimble County, Kentucky 18
3.5 Adams County, Ohio 19
3.6 Beaver County, Pennsylvania 19
3.7 Mason County, West Virginia 19
4.0 Impact Analysis 20
4.1 Introduction 20
4.2 Impact Analysis for the Period 1965-1975 ..... 20
4.2.1 Jasper County, Illinois 22
4.2.2 Adams County, Ohio 25.
4.2.3 Beaver County, Pennsylvania 31
4.2.4 Overall Impacts, 1965-1975 Impact Analysis .... 3^
4.3 Impact Analysis, 1975-2000 39
4.3.1 Projections of Employment Demand 39
-------
Section
Table of Contents (Cont'd)
Title
4.3.2 Population Impacts 49
4.3.3 Housing Impacts 50
4.3.4 Public Water Systems Impacts 55
4.3.5 Tax Impacts 56
5.0 Classification of Case Study Counties for
Impact Evaluation 60
5.1 Classification in Other Studies 60
5.2 Classification of ORBES Candidate Counties 62
6.0 Summary and Conclusions 76
Appendix A List of Data Sources Used for Site
Specific Analysis 81
Appendix B ORBES Labor Impact Model 84
-------
List of Tables
Table Number Title Page
1 Potential Impact Variables 6
2 Existing and Planned Electric Generating
Units for Case Study Counties 11
3 Demographic Characteristics of Case
Study Counties and ORBES 1970-1975 13
4 1970 Census of Housing Data for Case
Study Counties and ORBES 15
5 Labor Force, Employment and Income Data
for Case Study Counties 16
6 Population and Population Growth for
Jasper County Selected Years 23
7 Employment for Selected Sectors and Annual
Rates of Change - 1970, 1972, 1974, 1976 . . 2k
8 Property Valuations for Jasper County
for Selected Years 26
9 Expenditures and Revenue Data for
Jasper County - 1970-1976 26
10 Population and School Enrollment for
Adams County, Selected Years 27
11 Employment and Unemployment for Adams
County - Selected Years . . 29
12 Expenditures and Revenues for Adams
County - Adams County 30
13 Property Valuations for Adams County
Selected Years 30
14 Population and Population Growth for
Beaver County - Selected Years 32
15 Employment and Employment Growth for
Beaver County - Selected Years 33
16 Property Valuations for Beaver County .... 35
17 Expenditures and Revenues for Beaver County. 35
ill
-------
List of Tables (Cont'd)
Table Number Title Page
18 Partial Correlation Analyses Results ....... 37
19 Regression Results for Actual and Predicted
Employment ...................... 38
20 Information Sources for Power Plant
Construction and Operation Schedules ....... Uo
21 Jasper County Energy Facility Information,
Power Plant Employment and Immigration
1975-2000
22 Jefferson County Energy Facility Information,
Power Plant Employment and Inmigration
1975-2000
23 Spencer County Energy Facility Information,
Power Plant Employment and Inmigration
1975-2000 ............. .
24 Trimble County Energy Facility Information
Power Plant Employment and Inmigration
1975-2000
25 Adams County Energy Facility Information,
Power Plant Employment and Inmigration
1975-2000 ....................... 1*6
26 Beaver County Energy Facility Information
Power Plant Employment and Inmigration
1975-2000 ....................... 1*7
27 Mason County Energy Facility Information,
Power Plant Employment and Inmigration
1975-2000 ....................... 1*8
28 Total Inmigration, Including Families and
Percentage of 1975 Population - Case Study
Counties, 1975-2000 ................. 51
29 Annual Building Rate and Housing Demand
for Case Study Counties - 1975-2000 ....... 54
30 Estimated Jasper County Tax Impacts ....... 58
31 Oak Ridge National Lab Indicator of
Social Impact .................... 51
iv
-------
List of Tables (Cont'd)
Table Number Title Page
32 Variables Used in the Taxonomy of
Candidate Counties A3
33 Descriptive Statistics on Groupings
Derived Using All Variables 66
34 Group Statistics for Selected Variables
Using Alternative Classification Schemes ... 72
35 Group Membership for Case Study Counties
Groups Based on Variables In2 Ifh
36 Description of the Classification of
Candidate Counties and Potential for
Socioeconomic Impacts 75
-------
List of Illustrations
Figure Number Title Page
1 Power Plant Site and Socioeconomic Issues .... ^
2 Impacted Years Associated with Power
Plant Construction in Case Study
Counties 1969-2000 21
3 County Impact Groups Using All Variables 65
k County Impact Groups Using Income Variables ... 68
5 County Impact Groups Using Employment Variables . 69
6 County Impact Groups Using Housing Variables ... 70
7 County Impact Groups Using Population Variables . 71
VI
-------
1.0 Introduction
The general focus of the Ohio River Basin Energy Study (ORBES) is
on the regional impacts of various alternative energy development
futures in the study region. The full array of environmental, economic,
and social impacts are being examined and evaluated in the context of a
number of potential policies.
Given the focus at the regional level, the study must utilize
readily available data at a scale congruent with regional analysis.
Some generalizations must be made in order to give a concise summary of
potential events and their impacts. These generalizations are subject
to some variance, to a greater or lesser degree, depending on the type
of data limitations and the nature of the impacts being evaluated. In
addition, certain impacts at the local level defy generalization to the
regional scale. For example, it makes little sense to generalize about
the impacts of power plant development on regional housing since there
really is no housing market on the scale of the ORBES region. This is
in fact a rather local phenomenon.
Certain generalizations that are being made at the regional level
remain incompletely tested or analyzed at the local scale. We may make
generalizations about the migration impacts of energy development but
in different types of communities the number and importance of these
migrants may differ markedly.
These questions are particularly important for some of the
socioeconomic impacts associated with energy development. The purpose
of this report is to examine seven case study areas in the ORBES region
in order to explicate and examine the nature of some of the socioeconomic
impacts of power plants and their relationship to regional level impacts.
The purpose of this report is to summarize our findings in the
socioeconomic impact analysis of power plants in seven case study
counties in the ORBES region. The first chapter of the report provides
an overview of regional socioeconomic impact analysis. Socioeconomic
issues in siting are illustrated and discussed in the context of
previous site specific and regional studies.
The next chapter begins our analysis of the seven case studies. A
descriptive overview of the cases is given. Comparisons are made to
the ORBES region as a whole. Then, each county is discussed in more
detail in terms of population, housing, economic status, and public
services.
The following chapter describes in detail our site specific impact
analyses. Impacts on population, housing, services, and taxes are
presented. This chapter also notes problems associated with this type
of impact analysis stemming from lack of data or incomplete knowledge,
while at the same time pointing out how indirect evidence can be brought
to bear to obtain an assessment of impacts.
-------
Finally, we discuss our own method for generalizing our results.
A set of classifications of potentially impacted counties is presented
and compared with our case study examples.
-------
2.0 Review of Related Literature on Local and Regional Socioeconomic
Impacts of Power Plants
2.1 Introduction
In recent years, several regional environmental/energy impact
assessments have been undertaken (1). Each has taken a different
approach to solving the question of how to generalize from local impact
experiences to the regional setting. In fact, one of the major linkages
of this report to the remainder of the ORBES work is to answer this
question. Do certain communities have characteristics in common which
make them more or less prone to various social impacts or is each
situation unique? Are there data available at the local level which
allow a more complete analysis of socioeconomic impacts?
This chapter reviews other, similar attempts to answer these
questions. First, we define what is meant by the term "socioeconomic
impact." Then we review past attempts to delineate and quantify such
impacts in various circumstances. Finally, we review the other regional
studies to see to what degree they have been able to answer the questions
posed above.
2.2 Socioeconomic Impacts - Definition
A number of issues are included under the rubric of socioeconomic.
Each impact analysis that has been carried out has chosen to emphasize
and investigate a different set of such impacts.
Figure 1 illustrates the general flow of events following a power
plant siting decision along with the major socioeconomic issues.
Throughout the siting process, each event has the potential for having
both positive and negative impacts on the community. As is shown in
Figure 1, land acquisition may entail a change in the tax-base of the
community. The magnitude of the impact will vary depending on the size
of the parcel, its former use, and the demand for land for other purposes.
If some of the land on or adjacent to the site contains archaelogical
or historical sites these too could become an issue. It is of course
possible to choose sites in a manner which will minimize these problems.
Nevertheless, they are potential socioeconomic issues which have arisen
under a variety of circumstances.
At this time in the siting process, some tax benefits may accrue to
the local community and/or the state. When these changes in assessed
valuation and tax revenues occur depends both on the state tax laws and
on local and state assessment practices.
Following the start of construction, several more sets of issues
come into play. The construction activity will involve the use of heavy
equipment and thus may have impact on local roads (along with the new
-------
Figure 1
Power Plant Siting and Socioeconomic Issues
ISSUES
EVENTS
ISSUES
Siting Decision
is made
Competition for
Land: Agricultural,
Archeological, and
Historical Sites,
Recreation, etc.
Land Acquisition
Tax Benefits
Begin?
{Demand for Labor
Local Job
Benefits
Construction Beginsk.
Immigration
of New
Population
Income
Benefits
Commercial
Service
Impacts
Impacts on
Schools,
Housing,
Water, and
Sewers, etc
Impacts on Land,
Roads, Noise,
Dust, etc.
[plant Operationfr-
Major Tax
Benefits-After
Major Service
Demands? Where?
Changes in
Social Values
Competition for
Water, Air
Resources
-------
traffic), noise, dust, and so on. The construction work force itself
may create both negative and positive effects. Some of the labor force
may be local residents while others will inmigrate or commute from
elsewhere. Thus, there are issues of income benefits, public service
impacts, commercial service impacts (i.e. more local business) and a
potential change in social values caused by the newcomers to the
community. It is here also that the local tax issue is important. The
peak labor demand and thus the peak public service impacts may precede
the maximum tax benefits by several years. There can be short term
problems in the community to pay for new service demands as well as
long term tax benefits.
Finally, the operation of the plant will have similar demands on
services along with an affect of competition for land, water, and air
resources with other land use activities.
Figure 1 illustrates the wide range of issues which can arise
from a power plant siting. A number of demographic and economic data
can serve as indicators of the nature and extent of potential impacts.
Ideally, the variables used would measure all aspects of the social
system including descriptors such as population, housing, social services,
health care, crime, social values, etc. Unfortunately, our investigation
of available data in the first year of ORBES found that many such
indicator data are either unavailable or when available, are collected
for only parts of the region and on an inconsistent basis. For this
reason, we have had to rely on information which is collected on a
consistent basis even though it may be partially incomplete. These
data are from the 1970 Census of Population, census estimates, and
other special census materials such as County Business Patterns.
The types of variables available on this consistent basis is given
as Table 1. In looking at Table 1, several things should be pointed
out. First, it is obvious that there are many ways one might measure
a socioeconomic impact. Second it often remains a matter of judgement
as to whether some of these impacts are positive or negative. Such
judgement can be based only on individual values and perceptions. There
are almost as many perceptions of impacts as there are impact measures.
One less obvious observation is that the data used as indicators are
readily available only in the U.S. Census and related documents. Because
of this, it is very difficult to accurately trace the socioeconomic
status of an area where a power plant is sited and built within a time
frame of 5-8 years (on the average).
2.3 Classifying Potential Socioeconomic Impacts
Looking at the impact variables in Table 1, it becomes evident that
some method of grouping or classifying both impacts and communities
where various impacts might occur must be undertaken in order that
useful generalizations might be drawn and, where called for, appropriate
ameliorative actions can be taken. Several studies have focused on
particular types of impacts and/or methods of generalization. It is
-------
Data Type
Population
Table 1
Potential Impact Variables
Variable
Income
Employment
Housing
Agriculture
Government
Total County Population
Total White Population
Total Negro Population
Population by Age
Total Urban Population
Total Rural Population
Migration Rate
Per Capita Income
Median Family Income
Family Income by Income Class
Number of Families below Poverty Level
Total Labor Force
Occupation of Employed Persons
Employment by Skill
Total Employment
Unemployment Rate
Total Number of Housing Units
Age of Housing Units
Water Source of Housing Units
(public or private)
Housing lacking plumbing
Sewage disposal type
Persons per Room
Vacancy Rate
Land in Farms
Number of Farms
Expenditures and Revenues
6
-------
useful to review these here before summarizing the methods of analysis
we have chosen to use.
Sanderson and O'Hare review a number of socioeconomic impact
methods as applied to energy development (2). They rate each of the
methods they review according to a series of potential problems:
l) The method ignores important relationships
2) Same parameter used to measure two different impact phenomena
3) Dynamics ignored
k) Relationships among variables represented incorrectly
5) Regional parameters used to project subregional changes
6) Parameters estimated in the wrong time period
7) Undersized sample used
8) Parameters calibrated from biased or unrepresentative
sample. (2, p. 1-12 - 1-13)
They also point out the positive aspects of each of the models.
The errors listed above stem from a number of factors. The most
important of these is probably the fact that data on these types of
impacts are very limited. Where data have been collected, they are
generally inconsistent from case to case. Thus, socioeconomic impact
models are constructed based on (frequently untested) indirect measures
of potential impact, theoretical models which are overly simplistic,
and some local data.
The few monitoring studies that have been undertaken, however,
do appear to make many common observations. These include studies by
Mountain West Research, TVA, and Pennsylvania Power and Light
(3 through lU). All of these studies involved construction-worker
surveys during and following the construction period for plants of
varying size and location. Although there is a good deal of variance
in the study results several trends are apparent:
1) The construction projects in the Western U.S. show a much
higher percentage of the work force moving to the construction
area instead of commuting. Large metropolitan areas have
mostly local workers because there is a larger available
work force.
7
-------
2) The percent of the labor force coming from the local area
varies widely. For the TVA projects, the percentage of
non-local workers goer, up as the labor force increases.
(See reference 15 for a more complete discussion) This may
be explained by the variations in the need for skilled labor
as the construction project progresses.
3) Non-local workers tend to rent rooms or live in mobile homes
more often than local workers.
4) Construction workers tend to have median incomes $5000-$7000
above that of the local populace.
Combining these factors we can see that the potential for large
socioeconomic impacts should increase in those areas with fewer local
workers available, less available rental housing, and fewer available
public services. Based on this hypothesis, several regional studies
have sought to characterize power plant candidate sites and their
surrounding communities based on the potential for impacts.
Urban Systems Research and Engineering (16) uses a factor analytic
approach to classify 262 SMSA's (Standard Metropolitan Statistical
Areas) into groups with similar characteristics based on 200 initial
variables. The typology developed included data on ambient environ-
mental quality, urban form and the physical environment, pollution
residuals, and demographic characteristics. Representative cities
from each group are suggested as case study sites which reflect the full
range of possible variations.
Argonne National Laboratory used a classification scheme in a
more specific application to socioeconomic impacts (1, 17). Counties
where energy facilities might be sited were classified based on:
1) The size and age/sex composition of the population
2) The population density of the county and surrounding areas
3) The amount of service employment relative to basic
(or industrial) employment in the county
k) The size and location of nearby regional trade centers
A clustering algorithm was used to assign counties to one of three
groups: counties with low assimilative capacities and thus a high
probability of adverse socioeconomic impact from energy development,
counties with moderate assimilative capacity, and counties with high
assimilative capacities.
A similar approach was taken in Phase I of the ORBES project (18).
Here, six groups of candidate counties were derived, each with varying
-------
capacities to absorb new development. This classification also
differed in that a much larger array of socioeconomic indicators was
used as input.
In all of these attempts to generalize local impacts to the
regional level, no in depth analysis was performed to evaluate the
usefulness or predictive power of the classification scheme. The
monitoring studies reviewed above show a great variance from site to
site in local impacts depending upon a range of local conditions. Yet,
the regional studies choose only one, two, or three variables to
characterize local areas and then to generalize about the potential
for impact. This approach seemed overly simplistic to us. If
generalizations are to be used with any confidence, their basis in
fact must be more closely examined.
For these reasons, we have performed a test of various classification
schemes. We compare the differences associated with utilizing different
criteria as the indicator of potential socioeconomic impact. Further,
we attempt to test the accuracy of these indicators using data from
the ORBES case studies. Our methods and results are reported in the
following chapters. (For further discussion of past classification
efforts, please see Chapter 5).
-------
3.0 Description of Case Study Areas
Several criteria were used in selecting the counties for the
site specific analysis. The most important was that an energy
facility be planned for a site within the county during the 1975
to 2000 period. Additional requirements were: l) that the set of
counties represent a wide range of socioeconomic and environmental
characteristics, 2) that the power generating facilities in this set
of counties represent the various size, fuel and cooling types that
exist in the ORBES region, and 3) that counties undergoing additional
analysis by support researchers be given some priority. The results of
this selection process are listed below:
Jasper County, Illinois
Jefferson County, Indiana
Spencer County, Indiana
Trimble County, Kentucky
Adams County, Ohio
Beaver County, Pennsylvania
Indiana County, Pennsylvania
Mason County, West Virginia
A list of the planned and existing energy facilities for these
counties as reported in the Electric Generating Unit Inventory prepared
for ORBES by S. Jansen is shown as Table 2. We have attempted to
update this inventory by contacting the utilities involved. From GPU
we learned that Seward 7 was postponed indefinitely so Indiana County
was dropped from our analysis. The other updates are included in the
text and in the subsequent analysis.
It is important to understand the socioeconomic background of the
case study areas for it is the existing character that is the prime
determinant of the magnitude and severity of socioeconomic impacts that
may result from energy development. For the purposes of this study
1975 was chosen as the baseline year; however, much of the socioeconomic
data available comes from the 1970 Census. Housing data, one of the
most important variables in socioeconomic impact analysis, is rarely
available outside of the 1970 Census. Utilizing what information is
available several tables have been compiled which demonstrate the
range of characteristics present in the case study counties. A brief
description of each of the counties will then follow.
The demographic characteristics of the counties are shown on
Table 3. Beaver county, the largest county, is the only SMSA county
as a part of the Pittsburgh SMSA. Only Beaver is larger than the ORBES
average population with the remaining counties being one-half or less
this number. Growth rates for the 1970-1975 period vary considerably
although there are no negative rates. Also shown on this table is
the dependency ratio. This ratio is equal to the dependents of the
10
-------
Table 2
County
Jasper
Existing and Planned Electric Generating Units
for Case Study Counties3
Plant Name
Newton
Jefferson Clifty Creek
Spencer
Trimble
Marble Hill
Rockport
Trimble
Adams
J M Stuart
Beaver
Killen
Mansfield
Shippingport
Beaver Valley
nit
1
2
3
1
2
3
4
5
6
1
2
1
2
1
2
3
4
1
2
3
4
1
2
1
2
3
1
1
2
; Size
617
600
600
225
225
225
225
225
225
1130
1130
1300
1300
495
495
675
675
610
610
610
610
600
600
917
917
917
60
885
885
On-Line
Data
1977
1981
1984
1955
1955
1955
1955
1955
1955
1982
1984
1981
1982
1983
1985
p
7
1971
1970
1972
1974
1985
1982
1976
1977
1980
1977
1978
1984
Type Ownership
coal
coal
coal
coal
coal
coal
coal
coal
coal
nuclear
nuclear
coal
coal
coal
coal
coal
coal
coal
coal
coal
coal
coal
coal
coal
coal
coal
nuclear
nuclear
nuclear
CEIP
CEIP
CEIP
INKE
INKE
INKE
INKE
INKE
INKE
PSIN,
PSIN,
INME
INME
LOGE
LOGE
LOGE
LOGE
DAPO,
COSO
DAPO,
COSO
DAPO,
COSO
DAPO,
COSO
DAPO,
DAPO,
CLEI,
OHEC,
CLEI,
OHEC,
TOEC
CLEI,
OHEC,
TOEC
DULC
OHEC,
OHEC,
WVPA
WVPA
CIGE,
CIGE,
CIGE,
DICE,
CIGE
CIGE
PEPC,
DULC
PEPC,
DULC,
PEPC,
DULC,
DULC
DULC,
CLEI, TOEC
11
-------
County
Indiana
Plant Name
Conemaugh
Homer City
Seward
Table 2 (continued)
On-Line
Unit Size Data
Mason
Phil Sporn
1
2
1
2
3
3
4
5
7
1
2
3
4
5
Mountaineer
820
820
660
660
693
35
62
156
800
153
153
153
153
496
1300
1971
1971
1969
1969
1977
1941
1950
1957
1984
1950
1950
1951
1952
1960
1980
Type Ownership
coal PEEC, i4EEC
coal PEEC, MEEC
coal PEEC, NEYE
coal PEEC, NEYE
coal PEEC, NEYE
multi PEEC
multi PEEC
coal PEEC
coal PEEC
coal APPC, OHPC
coal APPC, OHPC
coal APPC, OHPC
coal APPC, OHPC
coal APPC, OHPC
coal APPC
Source:
Steven D. Jansen, Electrical Generating Unit Inventory 1976-1986,
Prepared for ORBES, November, 1978.
Notes: a)
Actual on-line dates for planned plants may differ from this
list. These will be noted in the text.
b) Ownership Codes:
CEIP Central Illinois Public Service Co.
INKE Indiana-Kentucky Electric Corp.
PSIN Public Service Co. of Ind., Inc.
WVPA Wabash Valley Power Assoc.
INME AEP: Indiana and Mich. Electric Co.
LOGE Louisville Gas and Electric Co.
CIGE Cincinnati Gas and Electric Co.
COSO Col's, and Southern Ohio Electric Co,
DAPO Dayton Power and Light Co.
CLEI Cleveland Electric Illuminating Co.
DULC Duquesne Light Col.
OHEC Ohio Edison Co.
PEPC Penna. Power Co.
TOEC Toledo Edison Co.
PEEC GPU: Penna. Electric Co.
MEEC GPU: Metropolitan Edison Co.
NEYE NY State Electric and Gas Corp.
APPC AEP: Appalachian Power Co.
OHPC AEP: Ohio Power Co.
12
-------
5.
Table 3
Demographic Characteristics of Case Study Counties and ORBES
1970-1975
1970
1970 1975 % Change 1970
County
Jasper
Jefferson
Spencer
Trimble
Adams
Beaver
Mason
ORBES Mean
Sources: U.
U.
Population
10,741
27,006
17,134
5,349
18,957
208,418
24,306
54,435
S. Bureau of
S. Bureau of
May 1977.
Population
11,198
27,622
17,631
5,617
22,299
209,328
25,254
55,589
the Census,
the Census,
Population
4.3%
2.3
2.9
5.0
17.6
0.4
3.9
4.7
1970 Census
-*• -" • ** *^V*£X\*. AAVt^Al ^, V
% Rural Ratio
71.9%
40.4
85.0
100.0
100.0
23.4
74.8
66.9
80.1
63.7
73.8
72.1
76.6
59.7
65.8
67.9
of Population.
Current Population
Reports, Series P-25,
13
-------
population (those under 15 years of age plus those 65 or over) divided
by the working-aged population (those between 15 and 6k years of age)
times 100 ^ In general, these dependents demand more of the public
service sector in terms of education, health and welfare services. At
the same time, their input into the public sector in terms of tax
revenues from income, property and sales tax is generally very small.
Therefore, a high dependency ratio can mean a disproportionate demand
for public services compared to a county of equal size and lower or
average dependency ratio. There appears to be a negative relationship
between population size and the dependency ratio for this group of
counties. The smaller counties have a higher proportion of dependents
than the larger counties. This indicates that the smaller counties
could be more heavily burdened by public services demands, such
as schools and health services for the elderly. The range of dependency
ratios across the case study counties is almost as great as that for the
ORBES region.
Housing data from the 1970 Census are shown on Table U. The
condition and attractiveness of the housing stock is indicated by the
availability of plumbing, water and sewer services. The two relatively
urbanized counties (Beaver and Jefferson) fare the best in terms of
these services. The other counties rank below the ORBES average.
The vacancy rate indicates the percentage of housing units that
are vacant at the time of the survey (1969). A high percentage of
vacant housing may be the result of l) a period of population growth
and building followed by out-migration or 2) the lack of public
services available to a portion of the housing stock combined with
substandard housing conditions. In the case of increased housing demand
induced by power plant construction, most of these vacant units could
be occupied.
Income, labor force and employment characteristics for the case
study counties and ORBES are shown on Table 5. Income levels range
from a low of $5563 in Adams County (also Ohio's lowest) to $9^)28 in
Beaver County. The range for the ORBES region is $2*4-07 (in Kentucky)
to $1169^ (in Ohio).
Manufacturing employment is a relatively small proportion of total
employment in Adams and Jasper counties - both rural areas. Jasper
also has the highest unemployment rate. Almost half of the total
employment in Beaver County is in manufacturing.
A description of the socioeconomic characteristics of the counties
on a county by county basis is provided below. A list of the references
used to compile the detailed information for these counties is shown
as Appendix A.
-------
Table 4
1970 Census of Housing Data for Case Study Counties and ORBES
Total % Lacking . % Public I % With Public
Jasper
Jefferson
Spencer
Trimble
Adams
Beaver
Mason
ORBES mean
ousing Units
3901
8553
5479
1798
6846
65942
8193
18450
Some Plumbing
18.1
11.2
16.7
31.8
37.9
3.7
24.9
19.2
Water
38.1
73.1
47.9
58.6
43.9
79.7
53.8
55.5
Vacant
7.3
7.8
7.7
5.7
11.3
2.8
9.5
7.7
Sewers
29.1
55.6
24.0
2.5
30.0
66.8
42.3
42.6
Source: U.S. Bureau of the Census, 1970 Census of Housing.
15
-------
Table 5
Labor Force, Employment and Income Data for Case Study Counties
1970
County
Jasper
Jefferson
Spencer
Trimble
Adams
Beaver
Mason
ORBES mean
Median
Family
Income
$8031
8556
7785
6596
5563
9428
6768
7672
Total
Labor
Force
3,828
10,253
6,498
1,864
6,042
77,734
7,576
20,847
Unemployment
Rate
10.6%
3.4
3.7
2.0
5.4
3.7
6.6
5.9
Total
Employment
3,423
9,902
6,256
1,826
5,714
74,864
7,074
19,726
Emplo;
i:
Manuf
506
3,093
2,131
525
1,039
35,791
1,989
6,109
Employment % of Total
Employment
;. in Manufact,
14.8%
31.2
34.1
28.8
18.2
47.8
28.1
27.7
Source: U.S. Bureau of the Census, 1970 Census of Population.
16
-------
3.1 Jasper County, Illinois
Jasper County is located in southeastern Illinois, off the main
stem of the Ohio River, approximately ^5 miles from Terre Haute,
Indiana, the closest city. It is a small rural county that has a large
elderly population (16.1% of population is 65 or over) and therefore
a high dependency ratio. In spite of the high number of dependents the
per capita income level ($2592) is the second highest among the case
study counties, although it must be noted that the state of Illinois
average per capita income ($3^95) is considerably higher.
The major employers for Jasper County in 1975 were manufacturing
and wholesale and retail trade. Unemployment rates were extremely high
for that year as well as in 1970. Total employment figures show
declining values between 1970 and
Public water and sewerage services were available only to
approximately one-third of the housing units in 1970. Medical services
are meager in the county. There is no hospital within its boundaries
and only one full-time practicing physician.
Jasper County is the site of the Newton Power Plant, owned by
Central Illinois Public Service Co. According to CIPS, three
coal-fired units are being constructed. Site clearing for the first
unit began in 1972, operations began in late 1977. The other two
units are projected for 1981 and 198^ on-line dates.
3.2 Jefferson County, Indiana
Jefferson County is located on the Ohio River in the southeastern
portion of Indiana directly opposite Trimble County in Kentucky (another
case study county). It is a small semi -urbanized (greater than 55%
urban) county approximately 55 miles south of Cincinnati and 35 miles
north of Louisville.
The housing units in Jefferson county are somewhat better served
by public water and sewer services than the other case study counties.
There are plans to expand the sewer systems in 1980 since the septic
tank system could prove inadequate in the event of considerable
inmigration. The percentage of housing units lacking some plumbing is
also lower in Jefferson than in the other ' counties .
The major manufacturing groups in 197^ were lumber and wood
products and machinery except electrical. Madison, the county seat,
is the site of $k.6% of all the manufacturing establishments in the
county.
Two hospitals in the county have 920 beds to serve the public.
New school facilities which serve southwestern Jefferson county were
constructed in 1976 near Hanover.
17
-------
The Clifty Creek Power Plant is situated in the Ohio River
downstream from Madison in Jefferson County. Built in the 1950fs, the
1300 MW coal-fired plant is still in operation. The Marble Hill nuclear
power plant began construction under a limited work authorization in
1977. The first unit is expected to be on-line in late 1982, the
second in 1984. This 2260 MW pressurized water reactor plant is owned
by Public Service of Indiana and Wabash Valley Power Association.
3.3 Spencer County, Indiana
Spencer County is located on the Ohio River across from Owensboro,
Kentucky, 35 miles east of Evansville, Indiana. A small, rural county,
Spencer has no hospitals or clinic facilities. Owensboro, Kentucky's
health service facilities are used by the county's residents. Public
sewer and water systems are available to less than one-half the housing
units in the county.
The major employers in Spencer County are:
Rockport Sanitary Pottery-
Abbey Press
Barmet Industries
each of which is in tLe category of 100-250 employees. Approximately
one-third of the employed are employed in the manufacturing industries.
The Rockport coal-fired generating plant is just north of
Rockport, Indiana, the county seat. Owned by Indiana and Michigan
Electric Co., the two-unit 2600 MW plant is expected to begin operation
in 1983 (1st unit) and 1984 (2nd unit). Construction began in 1977.
3.4 Trimble County, Kentucky
Trimble County is in northern Kentucky along the Ohio River,
roughly halfway between Cincinnati and Louisville. This is a completely
rural area with no public sewers and many housing units lacking some
plumbing. It has a very small population base - only slightly more
than 5,000 people.
Louisville Gas and Electric has planned to build a four unit coal
fired plant in Trimble county. The first unit is expected to come
on-line in 1984, the second in 1985. Construction began in 1979. No
other energy facilities exist in the county although the Clifty Creek
Plant in Jefferson County, Indiana is just across the river.
18
-------
3.5 Adams County, Ohio
Adams county is also situated on the Ohio River, 50 miles east of
Cincinnati. Although Adams county is not the smallest in population
size of the case study counties it is the poorest (median family income
of $5563) and contains the highest percentage of housing units that
lack some plumbing (37.8$). It is a rural area with a high dependency
ratio. Over one-third of the housing units are classified as substandard.
Less than 20$ of the employed work in the manufacturing industries. In
spite of all of these "negative" characteristics, Adams had a 17.6$
growth rate in population over the 1970-1975 period.
The JM Stuart coal-fired electric generating plant with four units
totaling 2khO MW was completed from 1970 to 197^. The Killen plant,
1200 MW in 2 coal-fired units, began full-scale construction in 1976.
Unit 1 is expected to be on-line in 1982, unit 2 in 1985.
3.6 Beaver County, Pennsylvania
Beaver county is the largest in population size and the only
SMSA county of the case study counties. It is part of the Pittsburgh
SMSA in western Pennsylvania on the Ohio River. Beaver County is an
urbanized area with adequate public services (sewer and water,
hospitals); low vacancy rate, low dependency ratio and high median
family income. Almost 50$ of the employed work in the manufacturing
industries, primarily in primary metals industry.
Two power plants are being constructed in the county. The Bruce
Mansfield coal-fired plant will have three 917 MW units with on-line
dates in 1976, 1977 and 1980 respectively. Four utility companies
jointly own this plant. The Beaver Valley plant is a nuclear plant
with 2 units at 885 MW each scheduled to be on-line in 1977 and 1982.
This is also a jointly owned plant.
3.7 Mason County, West Virginia
Mason County is situated on the Ohio and Kanawha Rivers in western
West Virginia, approximately ^5 miles north of Huntington. It is a
rural county with almost one-quarter of the families below the poverty
level. One hospital with 127 beds is located in the county. New wells
were being planned to double the current output for the public water
supply system. Two major manufacturing firms in the area are Foote
Mineral Company with k$l employees and the Goodyear Tire and Rubber
Company with approximately. 500 employees.
The Mountaineer plant is planned to be on-line in Mason county in
1980. It is a single unit, 1300 MW coal-fired plant owned by
Appalachian Power Co. Construction began in 197^ but was interrupted
shortly thereafter and construction was delayed until 1977.
19
-------
h.O Impact Analysis
if.l Introduction
This chapter presents our impact analysis for the seven case study
counties. The impacts examined include population, employment, services,
and taxes. For each case study area, a somewhat different set of data
were available. This means that the emphasis of our impact analysis
differs somewhat from case to case. It should be pointed out that this
difference in emphasis is solely one associated with information
available and in no way implies judgement as to the importance of each
impact.
The impact analysis is divided into several parts. In the first
part we examine impacts which have occurred in the recent past. This
applies to only three sites. In the second, we utilize several
projection models along with the data in environmental impact statements
to gain insignts into the impacts tnat could or are occurring at each
of the seven sites.
The primary criteria for selection as a case study area was that
an energy facility be planned for a site within the county. Each of
the chosen counties, therefore, has been designated by a utility as a
future site for energy development. Information on electric generating
units in the case study areas was presented on Table 2 in the previous
section. Planned units with on-line dates past 1975 are the basis of
the impact analysis performed for the case study counties.
A diagram has been constructed (see Figure 2) to show the
'impacted' years associated with each of the units with on-line dates
between 1965 and 2000. Assuming that one coal unit takes five years
and one nuclear unit takes seven years, the time line represents the
construction period with the asterisk indicating the peak employment
year (assumed to be the fourth year of construction for both types of
plants). When actual construction dates were available these were used
rather than the assumptions. From Figure 2 we can see that three of
the seven counties were experiencing power plant construction activity
during the period 1970-1975 (based on the above assumptions). Since
the OKBES socioeconomi c data base encompasses this time period we can
examine the trends in population, employment and public sector activity
for these counties. From this information we may be able to identify
some changes that could be associated with the construction activity.
k.2 Impact Analysis for the Period 1965-1975
The three counties experiencing power plant construction activity
during the 1965-1975 period are Jasper, Adams, and Beaver. We have
attempted to collect detailed data covering this period. These data
were requested from county and regional planning agencies, state and
20
-------
Figure 2. Impacted Years Associated with Power Plant Construction
In Case Study Counties, 1965-2000.
1965
1970
1975
1980
1985
1S90
1995
2000
Jasper
Jefferson
Spencer
Trimble
Adams
Beaver
Adams
610 MW lc
610 MW lc
610 MW
617 MW
. lc
610 MW lc
917 MW lc
917 MW
885 MW
•M
lc
600 MW
2260 MW
2600 MW
1200
lc
In
917 MW lc
885 MW
1300 MW 1
c
00 MW lc
2n
2c
95 MW lc
495 MW lc
MW 2c
In
675 MW
•
lc
675 MW lc
1965
1970
1975
1980 1985
YEAR
1990
1995
2000
Note: The megawatt size, number of units, and coal (c) or nuclear (n) type
is indicated above the time lines.
21
-------
county auditor's offices, state and county tax offices, boards of
education and so forth. Additional Census data was compiled from the
County Business Patterns (employment data) and Current Population
Reports series. Data sources are noted in Appendix A. Each of the
counties will be examined separately below. The trends or changes that
have occured over this time period may or may not be related to the
construction activity. For the most part we do not have sufficient
information to make that judgement. However we can indicate the change
and suggest possibilities for its cause.
lj-,2.1 Jasper County, Illinois
The Newton Plant, Unit 1, started operations in Jasper County in
1977. Construction began in late 1972. Typically, peak employment
occurs in the third or fourth year of the construction period or
1975-1976 for Newton 1. Although the data are meager and do not cover
the entire construction period there is some evidence of possible
related impacts in terms of construction employment and property
valuations. Trends in population growth are shown in Table 6. Please
note that the 1975 and 1976 figures are estimates. Although the trend
for declining population was broken for the period 1970 to 1975 there
appears to have been no substantial inmigration.
Employment figures for total employment and three major industrial
groups are shown on Table 7. We must be careful when reviewing
employment figures from County Business Patterns for three reasons:
l) They include only those employees covered by workmen's
compensation hence many employees are simply not counted
and an increase over time could mean new 'covered' employees
and not new employees.
2) They come from 'place of work1 data hence comparisons
with census data cannot be made since the census uses
'place of residence' data.
3) Part of the construction work force may be counted in
the county where the contractor's headquarters are located
rather than the'worksite.
For our purposes it is just as important to note that there was a
substantial number of workers in the construction industry in 1976. It
appears that the workers have not, for the most part, migrated to
Jasper and there has yet to be an effect on employment in other
industries, unless it has been one of reversing a declining trend.
22
-------
Table 6
Population and Population Growth for Jasper County Selected Years
Year Population Annual Rate % Change
of Change
1960 11,346
-605 -5.3
1970 10,741
21 0.6
1973 10,805
197 3.6
1975 11,198
-147 -1.3
1976 11,051
Sources: U.S. Bureau of the Census, 1970 Census of Population,
U.S. Bureau of the Census, Current Population Reports,
Series P-2.5.
23
-------
Table 7
Employment for Selected Sectors and Annual Rates of Change
1970, 1972, 1974, 1976
Year
Total Empl.
Annual Rate Construction
of Change
Annual Rate
of Change
1970
1972
1974
1976
Year
1970
1972
1974
1976
1420
1224
1602
2146
Manufacturing
542
253
607
576
-98
189
272
Annual Rate
of Change
-144
177
-15
28
103
46
(500-999)
Services
157
95
157
162
37
-28
na
Annual Rate
of Change
-31
31
2
Source: U.S. Bureau of the Census, County Business Patterns
1970 to 1976
-------
Information on property valuations is presented in. Table 8. The
large increase from 1975 to 1977 may be due to the valuation of the
plant itself. This would be expected since the first unit was
completed in 1977. This increase in property valuations can effect
the county residents in several different ways;
l) property tax rates can sharply decrease
2) government revenue can sharply increase allowing public
expenditures on capital improvements and so forth
3) a smaller tax rate decrease may be accompanied by
increased expenditures.
Neither tax rates nor expenditures data for 1977 were made available
to us so we cannot speculate as to the effect the increased property
valuations had on county residents. Later, we address this question
through the use of a simple model.
Expenditures and tax revenue data are available for selected
years (see Table 9). The most obvious change shown on this table is
for expenditures on highways. During the construction period
(1972-1977) huge amounts of money (relative to total expenditures) were
expended for highways. In 197^ highway expenditures were 60$ of total
expenditures. Highway improvements and maintenance are one of the most
significant impacts from energy development. Commuter traffic and
heavy truck traffic during construction decrease the expected life of
roads and bridges leading, to the site. New roads or 'widened' roads
may be necessary to support the increased traffic. Sometimes, the
utility company will either defray part of the costs or make the
necessary improvements themselves for the highways adjacent to their
property. In most areas, county governments such as Jasper still bear
most of the burden for highway maintenance. Note also that increased
tax revenue from the power plant would not be expected until 1977 or
later (after property valuations increase). Hence the funds used for
highway expenditures in 197*4- and 1975 do not, for the most part,
originate from tax revenue gained from the utility company. Thus, we
have our first indication of a timing problem with regard to tax
revenues and public expenditures required for an energy facility. This
could create short term community problems.
^.2.2 Adams County, Ohio
Four units of the J.M. Stuart plant were on-line in Adams County
in 1970, 1971, 1972 and 197^ respectively. Based on the assumptions
used in Figure 2 for construction periods, construction activity was
in progress from 1965 to 197^. Population and school enrollment data
for this period is presented in Table 10. The enrollment data for
25
-------
Table 8
Property Valuations for Jasper County for Selected Years
Year Property Valuation
1970
1974
1975
1977
$47,241,000
$58,678,448
$66,030,569
$104,630,487
Source: Illinois, Department of Local Government Affairs
Year
1970
1974
1975
1976
Table 9
Expenditures and Revenue Data for Jasper County
1970-1976
Total Expenditures
$ 405,370
896,092
883,481
583,412
Highway Expenditures Tax Revenues
$ 71,878 $ 305,556
535,743 377,625
335,898 428,475
223,387 550,208
Sources: Jasper County Clerk
Illinois Comptrollers Office: Statewide Summary of
Finance
26
-------
Table 10
Population and School Enrollment for Adams County/Selected Years
Year
1960
1965
1970
1973
1975
1976
Population
19,982
na
18,957
21,507
22,299
22,170
Annual Rate School
of Change Enrollment
-102
850
396
-129
5070
5575
5979
5989
5988
Annual Rate
of Change
101
134
-1
Sources: U.S. Bureau of the Census, 1970 Census of Population.
U.S. Bureau of the Census, Current Population Reports, Series
p. 25.
27
-------
1965-1970 suggest that the entire I960 to 1970 period may not have had
negative population growth and that, in fact, 1965 to 1970 may have
"been a period of population growth. However, population data for 1965
are not available to support this.
Employment data (see Table 11) show a slightly different pattern
of change than the population data. The period of negative growth for
employment is 1972-1975; for population it is 1960-1965 and 1975-1976.
Again, we must remember that employment data is by 'place of work1 and
that a declining population could be accompanied by a steady or
increasing number of commuters or vice versa.
We used the ORBES labor impact model (see Appendix B for a brief
explanation of the model) to estimate labor demand for construction of
the JM Stuart Plant. The peak employment occured in 1968 with demand
of approximately 1690 workers. The largest annual rate of change for
employment growth (see Table 10) was 1970-1972; for population,
1970-1973. From a conversation with Mrs. Roush of the Adams County
Regional Planning Commission and from more detailed data in the
County Business Patterns we discovered that a large factory manufac-
turing refrigeration units was opened in the county around 1972. This
plant employs approximately 500 people. The combined indirect effects
from the energy facility and the factory are difficult to ascertain
from Table 11 since total employment decreased by 5^ workers between
1972 and 1975. Even so, there is little chance that we could separate
the factory indirect employment impacts from those resulting from the
construction and operation of the JM Stuart Plant.
According to Mrs. Roush, the growth in Adams County during the
1970-1975 period was caused by a combination of fovr concurrent
situations
1) the Copeland factory mentioned above
2) urban to rural migration
3) increased building of recreational homes
h) JM Stuart and Killen Power plants
Fiscal data for the county is available for the years 1970, 1972
and 1975 (See Table 12). The large increase in expenditures from
1972 to 1975 was made possible, in part, through the issuance of
$1.U million general obligation notes. One of these notes was issued
and redeemed in both 197^ and 1975. They were expended for capital
improvements. We do not have any information on the nature of these
capital improvements so we cannot speculate as to the basis or cause
of the need for these improvements. Although it is not noted in
Table 12, highway expenditures jumped to $965,81*9 in 1973 indicating
a k3% increase over 1972. The highway improvements could have been
required as a result of either the energy facility construction traffic
or the factory traffic.
28
-------
Year
Table 11
Employment and Unemployment for Adams County
Selected Years
Total Empl.
Annual Rate
of Change
Construction
Annual Rate
of Change
1965
1970
1972
1975
1976
Year
1965
1970
1972
1975
1976
Year
1970
1972
1975
1438
1555
2519
1865
2220
Manufacturing
591
404
1064
745
864
Unemployment
Rate
9.9
12.4
21.5
23
482
-218
355
Annual Rate
of Change
-37
330
-106
119
Annual Rate
of Change
1.25
23
122
150
46
50
Services
106
145
176
176
176
20
14
-35
4
Annual Rate
of Change
26
15
0
0
0
Sources: U.S. Bureau of the Census, County Business Patterns
Ohio Bureau of Employment Security
29
-------
Table 12
Expenditures and Revenues for Adams County
Selected Years
Year
1970
1972
1975
Total Expenditures Highway Expenditures Revenues
$ 1,555,834
1,992,085
5,906,415
$ 700,204
677,028
908,293
$ 1,486,378
2,075,057
3,555,053
Source: Auditors of State of Ohio, Financial Reports; Ohio Counties
Table 13
Property Valuations for Adams County
Selected Years
Year
1965
1970
1973
1975
1976
Property Valuations
$ 20,098,530
28,320,380
30,850,370
62,359,400
63,132,610
Source: Ohio Department of Tax Equalization
30
-------
Property valuations had a significant increase over the 1973-1975
period (See Table 13). This was alsc the time period during which the
final units of the JM Stuart Plant were to go on-line. The valuation
of the new factory might also be included in this increase although it
appears that the opening of the plant may have taken place in 1971-1972.
The 1965-1975 period was one of growth for Adams County. Several
phenomena may have been responsible for these increases in the small
rural county. We cannot say with certainty which phenomenon was the
overriding factor of change. Associated with any population change
from any cause are the effects on school enrollment, housing, demand
for public services and tax revenue. However, the size of the work
force needed to build the JM Stuart Plant and the relatively long
construction period indicate that some inmigration (even though it may
be temporary) would be expected. Certainly, the doubling of the
property valuations between 1973 and 1975 must be at least, in part,
associated with the JM Stuart Plant.
k.2.3 Beaver County, Pennsylvania
Two power plants were under construction between 1970 and 1975
in Beaver County. The Bruce Mansfield plant, coal-fired units 1 and 2,
(under our scheduling assumptions) would be under construction starting
in 1971. The construction of the first unit of the Beaver Valley
nuclear plant would have begun work in 1970. The total impacted years
for these three units would be 1970 to 1977, however, a third unit of
the Mansfield plant and the second unit of the Beaver Valley plant
would, under our assumptions, have begun construction in 1975.
Population data for the I960 to 1976 "time period is shown on
Table lU. Although the peak year for population was 1973 both the
absolute and percentage increases for 1970-1973 were small. Counties
of this size would be .expected to have little or no inmigration
resulting from power plant construction activity.
Employment data are shown on Table 15. Except for the changes in
construction employment there appears to be no other trends that can be
associated with the construction activity. The increase in manufac-
turing, 1972-197^, is very short-lived and amounts to less than 6%
increase over that period. The peak years for direct labor impacts,
according to our model, would be 1973 to 1975. The peak labor demand
is projected to be approximately 3,800 workers. According to the
County Business Patterns, the increase in construction employment from
1972 to 197^ is only 833 employees with a total of 2,337 employees.
Many of the workers at the plant must be either not covered by
workmen's compensation, not categorized under 'construction' or have
headquarters in another county. A better indicator of the direct and
indirect employment effects of energy development is needed, such as
direct surveys of workers.
31
-------
Table 14
Population and Population Growth for Beaver County
Selected Years
Year Population Change of Percentage
Population of Change
1960 206,948
147 .7%
1970 208,418
457 .7
1973 209,788
-230 -.2
1975 209,328
-435 -.2
1976 208,893
Source: U.S. Bureau of the Census, 1970 Census of Population
U.S. Bureau of the Census, Current Population Reports
Series p-25.
32
-------
Year
Year
Table 15
Employment and Employment Growth for Beaver County
Selected Years
Total Empl.
Annual Rate
of Change
Construction Annual Rate
of Change
1970
1972
1974
1976
62,514
59,081
63,935
64,050
-1716
2427
57
1,803
1,504
2,337
2,747
-149
406
205
Manufacturing
Annual Rate
of Change
Services
Annual Rate
of Change
1970
1972
1974
1976
40,531
-1963
36,604
1015
38,634
-409
37,816
6,790
6,980
7,656
8,298
95
338
321
Source: U.S. Bureau of the Census, County Business Patterns
1970-1976.
33
-------
The data on property valuations reveal no changes that can be
identified with the energy facilities or any other growth inducer.
(See Table 16).
The data on county expenditures and revenues (see Table 17)
show a change over the 1970 to 197^ period that may or may not be
associated with the energy development. Since there was little or
no inmigration during this period the increased expenditures and
revenues must be due to the demands placed by the energy facilities
themselves (highway maintenance or improvements, other infrastructure
additions or improvements near the plant) or to some major capital
improvements that were funded at this time.
As we had expected there were few traces of the impacts related
to energy development in Beaver County. Labor demands are easily
absorbed by a large population base. The lack of inmigrants also
diminishes the associated indirect employment effect.
k.2.b Overall Impacts, 1965-1975 Impact Analysis
Overall, the historical data has reinforced some of our basic
ideas concerning socioeconomic impacts of construction of electric
generating units. We hypothesized that the larger the population size,
the more urban and the wealthier the county the smaller the impact.
Beaver County, a relatively large county which is part of the
Pittsburgh SMSA, showed little or no evidence of employment, population,
public expenditures or property valuation impacts. Conversely, Adams
County, a small, rural and poor county, experienced large population
increases, increased public expenditures and revenues, employment
changes and property valuation increases. Unfortunately, we can not
attribute these changes to power plant construction alone. Other
concurrent development was underway during this period that could have
had a major impact in these same areas of concern. In this type of
'ex post' analysis the problem arises to identify the causes which
precipitated the documented change such as in population or property
valuation noted above. When linkages are known or straightforward
it is a relatively simple matter. Sources of changes in property
valuation can be traced by going through the county assessor's books
parcel by parcel. It is a rather tedious and time-consuming task but
the documents are in the public domain and accessible. Changes in
population or employment are not as easily traced because we need to
know more than the obvious. For example,we need to know why the
people moved to Adams County or what stimulated industrial growth.
Thet
a) opening of new businesses, factories, military, health
or penal institutions
b) increased demand for rural recreational homes or
c) development of energy or transportation facitities
-------
Table 16
Property Valuations for Beaver County
Year Property Valuations
1970 $ 303,580,750
1972 318,839,440
1975 330,056,590
Source: Pennsylvania Abstracts
Table 17
Expenditures and Revenues for Beaver County
Year Expenditures Revenues
1970 $ 9,035,000 $ 4,920,000
1974 17,363,000 6,059,000
1975 16,386,000 6,163,000
Source: Pennsylvania Abstracts
35
-------
can all have similar effects on population and employment growth.
The closing of such facilities can have an opposite effect. When
several of these events occur concurrently it is very difficult to
separate out the effects of one from the other. It is likely that
there could be synergistic effects as well. With our limited knowledge
of the 'growth inducing' or 'growth reducing' situations in the case
study counties, we must limit our conclusions concerning the observed
indirect and direct effects from power plant construction.
Realizing that employment in manufacturing may be an indicator of
concurrent industrial development, we decided to perform some
statistical tests to see if power plant construction employment correlated
with total employment while holding manufacturing employment constant.
We chose to use partial correlation analysis and multiple regression
analysis as these tests. Using the model-derived construction and
operation employment with total manufacturing employment as taken from
the U.S. Bureau of the Census, County Business Patterns, the partial
correlation analysis results were as in Table 18.There were 26
observations, one for each impacted year between 1965 and 1975 for
Jasper, Adams and Beaver counties. Since the first order partial is
greater than the zero order partial (simple correlation coefficient)
and the significance level of the first order partial is so high, we
can interpret these results as indicating that model-derived
construction and operation employment explains 29 percent of the
variance in total employment.
Again, this provides indirect1evidence that our impact modeling
assumptions are correct since the correlation between model total
employment and actual total employment goes up significantly when
manufacturing employment is controlled for. It ic also evident that
total employment is much more strongly related to manufacturing
employment than to our model values. This is expected since
manufacturing employment generally makes up a much more significant
proportion of the total labor force. However, our analysis still shows
that power plant construction (as per our model) has a significant
albeit lesser impact on total employment changes.
We also tested this relationship using regression analysis.
Here, the actual total employment from County Business Patterns is the
dependent variable and services employment, manufacturing employment,
and either the predicted model total employment or construction
employment were the independent variables. The results are displayed
in Table 19. Although service and manufacturing employment changes
explain most of the ^b^nge in total employment, model total and
construction employment for power plants contributes significantly
to the variance explained. In fact, model employment accounts for
approximately 30$ of the variation in total employment. This confirms
that our model works fairly well in predicting the changes in local
employment caused by power plant construction.
3b
-------
Table 18
Partial Correlation Analysis Results
Correlation Coefficients
TOTEMP MODTOT MANU
Total Employment (TOTEMP 1.00 .5336 .9992
Model-Derived Employment (MODTOT) 1.00 .5131
Manufacturing Employment (MANU) 1.00
First Order Partial
Controlling for Manufacturing Employement (MANU):
MODTOT Significance
Total Employment (TOTEMP) .5888 .001
37
-------
Table 19
Regression Results for Actual and Predicted Employment
Dependent Variable - Total Employment, County, Business Patterns
N=26
R2=.99
Equation 1
Independent Variables
Service employment
Model total employment
Manufacturing employment
Constant
B Standard Error
2.78 0.25
0.32 0.09
1.06 0.05
484.73
F*
120.34
11.35
506.7
B
Equation 2 (same N and R )
Independent Variables
Service employment 2.83
Model construction employment 0.31
Manufacturing employment 1.05
Constant 519.72
Standard Error F*
0.25 124.9
0.09 10.2
0.05 498.5
*A11 values of F are significant at the 0.05 level
38
-------
Overall then, this portion of our impact analysis confirms many
of the impact assumptions we make at the regional level. In addition,
it illustrates the nature and extent of county level impacts vis a vis
county population size. Finally, it shows the variation associated
with actual impacts caused by the uniqueness of the local situation.
k.3 Impact Analysis, 1973-2000
In this section we estimate employment demands for the planned
energy facilities in the seven case study counties, 1975-2000. The
basis for these employment estimates are either the projected manpower
requirements listed in Environmental Impact Statements for these
facilities or the ORBES Labor Impact Model (discussed below). These
employment demands are translated into population impacts. The
population impacts axe the basis for estimating housing, water systems,
public expenditure, and county revenue impacts. Each of these impacts
is discussed for each case study county.
4.3.1 Projections of Employment Demand
Construction and operation employment demand is estimated for
each case study county for the period 1975-2000. The on-line dates and
schedules for the power plants in these counties were obtained from the
utility companies involved, or specific Environmental Reports and
Environmental Impact Statements whenever possible. Otherwise, on-line
dates were taken from the Generating Unit Inventory prepared by S. Jansen
for ORBES and schedules were derived using the ORBES Labor Impact
Model. The sources used for developing employment demands other than
the ORBES Labor Impact Model and the ORBES report mentioned above are
shown in Table 20.
The ORBES Labor Impact Model was developed to estimate annual labor
demands for the construction and operation of planned and scenario-
specific generating units in the ORBES region. The annual construction
labor demands are derived using manpower per megawatt ratios that vary
for the following conditions:
coal unit with scrubbers, single unit
coal unit with scrubbers, part of multiple unit
coal unit without scrubbers, single unit
coal unit without scrubbers, part of multiple unit
nuclear unit, single or part of multiple
Our ratios are based on actual and expected manpower data available
in EIS's, ER's and from the utility companies themselves. These ratios
are higher for units with scrubbers than without, lower for parts of
multiple units than single units, and higher for nuclear units than any
other unit. Scrubbers require additional buildings, equipment and waste
disposal areas. Multiple units can share some facilities and infra-
structure and therefore can reduce the overall cost per unit. Nuclear
units require many highly skilled workers and longer construction periods.
39
-------
Table 20
Information Sources for Power Plant
Construction and Operation Schedules
County
Jefferson, Ind.
Spencer, Ind.
Adams, Oh.
Mason, W. Va.
Jefferson, Ind.
Jasper, 111.
Spencer, Ind.
Trimble, Ky.
Beaver, Pa.
Report
Marble Hill EIS
Rockport ER
Killen EIS
Utility
Appalachian Power
Public Serivce
of Indiana
Central Illinois
Power Service Co.
Indiana and Michigan
Electric Company
Louisville Gas &
Electric
Pennsylvania Power
Information Obtained
Construction and
operation employment
Construction and
operation employment
Construction and
operation employment
Construction and on-line
dates for Mountaineer
Construction and expected
on-line dates for
Marble Hill
Construction and expected
on-line date for Newton
Construction and expected
on-line dates for
Rockport
Construction and expected
on-line dates for
Trimble
Construction and expected
on-line dates for
Mansfield
-------
The length of the construction period was also derived from the
same information used to develop the ratios. Coal units less than
1000 megawatts were assumed to take 5 years to build, coal units
1000 mw or greater, 6 years and nuclear units, 7 years.
Operation employment demands were also estimated on a per megawatt
basis. Again, the same sources were used. The ratios varied for coal
units with scrubbers, coal units without scrubbers and nuclear units.
A more complete description of the methods and data sources used
to develop the labor demands portion of the OKBES Labor Impact Model
is provided in an ORBES memo from Steve Gordon dated June 19, 1979 and
in the report by Gordon and Graham (26).
Tables 21 to 27 show the employment demands for each of the case
study counties, 1975-2000. The tables exhibit construction and
operation employment and total estimated inmigration by year from
1975 to 2000. The figures reflect annual changes in worker demand
through the construction period (only some of whom become inmigrants)
and the final labor force associated with power plant operation (all
of whom are assumed to inmigrate). These counties exhibit a wide range
of construction schedule types. The Rockport plant in Spencer County
has a very high peak in construction employment (2,988 workers) with
a relatively short duration (6 years for 2 units with total of 2,600 MW).
The magnitude of this demand may strain the labor supply for the peak
years. This short-lived high demand for workers can strain all the
local private and public services. A long construction period with low
extended peaks would have quite a different effect from one like that
at Rockport. For instance, the Killen plant in Adams County (2 units,
total of 1200 MW, 8 year construction period) has a projected construction
employment peak of only ^00 workers lasting for four years. The
prospects of employment lasting for this longer time period is likely
to induce a higher percentage of the workers to migrate and to bring
their families.
Another variation in schedules is shown by the Newton Plant in
Jasper County. Here the construction of three units is lagged over a
12 year period. This 'long-term' demand for workers, again, gives
more incentive for construction workers to migrate and to bring their
families than a short, high-peaked construction schedule such as
Rockport's.
An unusual schedule is shown for Mason County where construction
work began in 197^, proceeded for less than six months and was forced
to shut down for two years because of a work stoppage. There are many
factors that could delay the construction of a power plant, including:
materials supply problems, legal constraints, labor supply problems,
work stoppage (strikes), accidents and so forth. These are the
"unpredictables" associated with power plant development. Any special
problems associated with delayed or interrupted construction such as
ill
-------
Table 21
Jasper County Energy Facility Information,
Power Plant Employment and Inmigration 1975-2000
Plant Type On-Line Date MW
Newton 1 coal 1977 617
Newton 2 coal 1981 600
Newton 3 coal 1984 600
Year Total Construction Operation Inmigration
Employment Employment Total
1975 616 0 68
1976 121 0 12
1977 274 74 101
1978 730 74 174
1979 705 74 145
1980 345 74 109
1981 730 146 219
1982 657 146 212
1983 71 146 153
1984 218 218
1985 218 218
1986 218 218
1987 218 218
1988 218 218
1989 218 218
1990 218 218
1991 218 218
1992 218 218
1993 218 218
1994 218 218
1995 218 218
1996 218 218
1997 218 218
1998 218 218
1999 218 218
2000 218 218
Note: Schedule derived from ORBES Labor Impact Model
-------
Table 22
Jefferson County Energy Facility Information,
Power Plant Employment and Inmigration 1975-2000
Plant Type On-Line Date MW
Marble Hill 1 & 2 nuclear 1984 2260
Year Total Construction Operation Inmigration
Employment Employment Total
1976 7 $ I
1977 180 0 18
1978 923 0 92
1979 1,820 0 182
1980 2,154 0 215
1981 1,864 0 186
1982 1,023 0 102
1983 244 0 24
1984 155 155
1985 155 155
1986 155 155
1987 155 155
1988 155 155
1989 155 155
1990 155 155
1991 155 155
1992 155 155
1993 155 155
1994 155 155
1995 155 155
1996 155 155
1997 155 155
1998 155 155
1999 155 155
2000 155 155
Note: Schedule takea from Marble Hill EIS,
-------
Table 23
Spencer County Energy Facility Information,
Power Plant Employment and Inmigration 1975-2000
Plant Type On-Line Date MW
Rockport 1 & 2 coal 1983 2600
Year Total Construction Operation Immigration
Employment Employment Total
1977 466 47
1978 756 76
1979 2,225 223
1980 2,988 299
1981 1,819 182
1982 150 165 180
1983 335 335
1984 335 335
1985 335 335
1986 335 335
1987 335 335
1988 335 335
1989 335 335
1990 335 335
1991 335 335
1992 335 335
1993 335 335
1994 335 335
1995 335 335
1996 335 335
1997 335 335
1998 335 335
1999 335 335
2000 335 335
Note: Schedule taken from Rockport Environmental Report
-------
Table 24
Trimble County Energy Facility Information
Power Plant Employment and Inmigration 1975-2000
Plant
Trimble 1
Trimble 2
Trimble 3
Trimble 4
Type
coal
coal
coal
coal
On-Line Date
1984
1985
1992
1999
MW
495
495
675
675
Year
Total Construction
Employment
Operation
Employment
Immigration
Total
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
47
318
993
1,372
720
70
9
9
64
370
485
886
96
9
9
64
370
985
886
96
9
69
138
138
138
138
138
138
138
232
232
232
232
232
232
232
326
326
5
32
99
137
72
76
138
138
144
175
237
227
148
232
232
238
269
331
321
242
326
326
Note: Schedule derived from ORBES Labor Impact Model
-------
Table 25
Adams County Energy Facility Information,
Power Plant Employment and Inmigration 1975-2000
Type
Coal
On-Line Date MW
1985 1200
Year
Total Construction
Employment
Operation
Employment
Immigration
Total
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
130
300
400
400
400
400
250
250
9
9
9
9
9
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
13
30
40
40
40
40
25
25
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
Note: Schedule taken from Killen EIS
-------
Table 26
Beaver County Energy Facility Information
Power Plant Employment and Inmigration 1975-2000
Plant Type On-Line Date MW
Mansfield 1 Coal 1976 917
Mansfield 2 coal 1977 917
Mansfield 3 coal 1980 917
Beaver Valley 1 nuclear 1978 885
Beaver Valley 2 nuclear 1984 885
Year Total Construction Operation Inmigration
Employment Employment Total
1975 2,365 0 118
1976 1,143 192 249
1977 1,513 384 460
1978 1,710 463 549
1979 1,143 463 520
1980 1,256 655 718
1981 943 655 702
1982 511 655 681
1983 92 655 660
1984 734 734
1985 734 734
1986 734 734
1987 734 734
1988 734 734
1989 734 734
1990 734 734
1991 734 734
1992 734 734
1993 734 734
1994 734 734
1995 734 734
1996 734 734
1997 734 734
1998 734 734
1999 734 734
2000 734 734
Note: Schedules derived from ORBES labor impact model
-------
Table 27
Mason County Energy Facility Information,
Power Plant Employment and Immigration 1975-2000
Plant Type On-Line Date MW
Mountaineer Coal 1981 1300
Year Total Construction Operation Immigration
Employment Employment Total
1974 104 9 10
1975 0 0 9
1976 9 9 9
1977 594 0 59
1978 1,583 0 158
1979 1,424 0 142
1980 154 0 15
1981 156 156
1982 156 156
1983 156 156
1984 156 156
1985 156 156
1986 156 156
1987 156 156
1988 156 156
1989 156 156
1990 156 156
1991 156 156
1992 156 156
1993 156 156
1994 156 156
1995 156 156
1996 156 156
1997 156 156
1998 156 156
1999 156 156
2000 156 156
Note: Schedule derived from ORBES Labor Impact Model and
Information gained from APPC as to the work stoppage
1974-1977
-------
the effect of increased labor and supply costs, cannot be covered here
since these problems would be specific to the particular situation.
The specifics for the Mountaineer plant in Mason County, other than the
years of construction, were not available to us so we can only address
the direct effects of a break in the schedule.
Just how these various schedules are translated into population
changes and housing, public service, public expenditures and county
revenue impacts is discussed in the remainder of this chapter.
4.3.2 Population Impacts
Once the direct labor requirements have been established for the
construction and operation of.an energy facility, we must then proceed
to determine how many of these workers will migrate and how many will
commute to that county. The number of inmigrants is an important
variable for estimating housing and public service impacts discussed
in the following sections of this chapter. We will review the
assumptions used in estimating the number of migrants and the derived
figures for each case study county.
The number of inmigrants is computed as part of the ORBES Labor
Impact Model. The percentage of workers that migrate is assumed to
vary directly with the distance between the case study county and the
nearest SMSA county. The percentages of migrating workers related to
distance were derived from the TVA and other monitoring studies cited
in Chapter 2.0. A test of these assumptions in year one of ORBES
showed them to be good relative indicators of migration impacts
consistent with the available data. Distances were measured from
centroid to centroid. SMSA counties were assumed to attract only 5$ of
the construction workforce as inmigrants, counties within 50 miles of
an SMSA, 10$, and counties greater than 50 miles away, 30$. All
operation and maintenance workers are assumed to migrate to the county.
Refer to Tables 21 to 2? for lists of the number of inmigrant workers
for each case study county, 1975 to 2000.
Beaver County, the only SMSA county in our case study group, shows
only 5$ of the construction workforce migrating (see Table 26) and
100$ of the operating workers. We must note that the inmigration
totals on these tables is for workers only. Since some of these workers,
the operation workers particularly, will bring their families with them,
the actual number of inmigrants would be approximately from 1.8 to 2.2
times as great.
The rest of the case study group, Jasper, Jefferson, Spencer,
Trimble, Adams, and Mason Counties, are all within 50 miles of an SMSA.
The ten percent migration figure was used for those counties. Using
the conservative figure of 1.8 persons per family, we estimated the
population impact of the power plant workers on each county. These are
-------
listed on Table 28. Also listed is the percentage of the county's 1975
population that the immigrants represent. In only Trimble County,
Kentucky, does this percentage go above 3.5%, and this occurs past 1980
when the base population would be greater. The peak percentage of
inmigrants to 1975 population reaches 10.6$ for Trimble County, however,
the peak is reached in gradual increments from 1979 to 1996. In
Beaver County the percentage never goes above .6%. This information
seems to indicate that the number of inmigrants for any county will
cause no great influx of people relative to the base population.
Numbers of inmigrants alone are not a sufficient indication of
the related impacts in housing, public services and county revenues
and expenditures. Each of these impact areas are covered in the
following sections.
^.3.3 Housing Impacts
One of the most significant impacts resulting from power plant
construction is the demand placed on the local housing stock by
inmigrants. This demand can create a strain on the existing market
and thereby increase housing costs. Housing impacts are estimated
for each of the case study counties based on the number of construction
and operation employees expected to migrate and an estimate of the
rate of building new housing.
Very little housing data exists outside the decennial census.
Using the I960 and 1970 Census of Housing data we derived an annual
new building rate for each case study county. By comparing this annual
building rate with the additional demand created by migrating power
plant workers we can get some idea concerning a potentially strained
housing market. These figures are listed on Table 29.
We assume that an additional demand comprising more than 50
percent of the historical growth rate will create a significant impact
on the housing stock. Five out of the seven case study counties
appear to have at least one year with potential housing shortage
problems. In Jasper County the additional housing demand exceeds
50 percent of the building rate for four years - 1975, 1977, 1978, and
1981. For Spencer County this situation exists in 1979 and 1980.
In Trimble County 50 percent of the building rate is exceeded
six times, 1980-1982, 1988-1989, 1996. In Adams County, where the
construction schedule had very low peaks over a long duration, the rate
is exceeded only in 1985, the on-line date of the Killen plant. The
operation workers are assumed to move to the host county. For Killen,
tie largest number of inmigrants is associated with the start of the
operations phase of the plant. In Mason County, housing shortages are
possible in 1978. The consequences of such demands can be both positive
and negative. They may be negative in the sense that construction
workers with higher salaries than most local residents can bid up the
prices for housing. They might also choose to live in temporary
50
-------
Table 28
Total Inmigration, Including Families and Percentage of 19~75 Population
Case Study Counties, 1975-2000
VJl
Jasper County
Inmigrants As a % of
with Families 1975 Population
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
122
22
182
313
261
196
394
382
275
392
392
392
392
392
392
392
392
392
392
392
392
392
392
392
392
392
1.1
0.2
1.6
2.8
2.3
1.8
3.5
3.4
2.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
Jefferson County
Inmigrants As a % of
with Families 1975 Population
Spencer County
Inmigrants As a % of
with Families 1975 Population
0
2
32
166
328
381
335
184
43
279
279
279
279
279
279
279
279
279
279
279
279
279
279
279
279
279
0
0.1
0.6
1.2
1.4
1.2
0.7
0.2
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0
0
85
137
401
538
328
324
603
603
603
603
603
603
603
603
603
603
603
603
603
603
603
603
603
603
0
0
0.5
0.8
2.3
3.1
1.9
1.8
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
3.4
-------
Table 28 (Cont'd)
Trimble County
Adams County
Inmigrants
with Families
As a % of
1975 Population
Inmigrants
with Families
As a % of
1975 Population
Ul
ro
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
9
58
178
247
130
137
248
248
259
315
427
409
266
418
418
428
484
596
578
436
587
587
0.2
1.0
3.2
4.4
2.3
2.4
4.4
4.4
4.6
5.6
7.6
7.3
4.7
7.4
7.4
7.6
8.6
10.6
10.3
7.8
10.5
10.5
23
54
72
72
72
72
45
45
270
270
270
270
270
270
270
270
270
270
270
270
270
270
270
270
0.
0.
0.
0.
0.
0.
0.
0.
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
-------
Table 28 (Cont'd)
Beaver County
Mason County
LO
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
Inmigrants
with Families
212
448
828
988
936
1292
1264
1226
1188
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
1321
As a % of
1975 Population
0.1
0.2
4
5
0,
0
0.4
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
As a % of
with Families
106
284
256
27
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
281
As a % of
1975 Population
0.4
1.1
1.0
0.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
-------
Table 29
Annual Building Rate and Housing Demand for Case Study Counties 1975-2000
Additional Demand Created
by In-migrating Workers
Annual Building
County Rate
Jasper
61
Jefferson
214
Spencer
114
Trimble
45
Adams
94
Beaver
909
Mason
162
Year Number of Units
1975
1977
1978
1981
1976
1977
1978
1979
1980
1977
1978
1979
1980
1983
1979
1980
1981
1982
1985
1987
1988
1989
1994
1995
1996
1977
1978
1979
1985
1975
1976
1977
1978
1980
1984
1974
1977
1978
68
33
46
72
1
17
74
90
33
47
29
147
76
36
5
27
67
38
1
6
31
62
1
3
62
13
17
10
110
118
131
211
89
169
16
10
49
99
Source: U.S. Bureau of the Census, 1970 Census of Housing,
-------
housing such as trailers which may be perceived as a negative impact
by some local residents. On the other hand, the new demand.'; could
mean increased local employment in the construction and service
industries. The additional payroll of the power plant construction
will bring some positive economic benefits. The extent to which these
negative and/or positive effects will occur depends on local perceptions
and individual decisions which are impossible to predict.
U.3.^ Public Water Systems Impacts
Increased housing demands generate associated infrastructure
demands such as sewer and water services. Based on information obtained
from the U.S. Bureau of the Census and state agencies dealing with
these services, we have attempted to evaluate the potential demand/
supply situations which could exist in the case study counties with
the development of energy facilities as planned.
Many rural counties in the OKBES region have little or no public
sewer systems. Public water systems are more abundant but they often
serve only 50$> of the housing units. (See Table U). Existing water
and sewer systems can have consumption or use levels which are very
near the design capacity of the plant. Inmigration could severely
strain the systems in this case forcing a decline in the quality and
quantity of service available to all customers, or causing the public
outlay of funds to expand or improve their system. In some counties,
an increase in the number of septic tank systems could severely effect
the local water quality.
Information concerning local water systems is often available from
the state department of Natural Resources or the state EPA. The data
usually includes population served, treatment plant capacity, pumping
station capacity, daily consumption, and others. Interpretation of
this information can be difficult. Data on local water systems were
collected for the case study counties using the references listed in
Appendix A. Local consumption and capacity figures were summed to the
county level. We then compared consumption with capacity. At the
county level, there appears to be a great excess of water capacity. We
further examined other ORBES Ohio counties for this excess and found
this excess to be the rule rather than the exception. One of the
reasons for this may be that the consumption figures used are 'average
daily use' rather than 'peak daily use'. We must also note that summing
across local systems to the county level overlooks the potential strain
on an individual system. For example, a few local systems within the
case study counties are relatively close to capacity. The New Haven-
Hartford-Mason service area in Mason County was reported as having a
daily excess of 20,000 gallons per day. Using the 'rule of thumb'
estimate of 100 gallons of water required per person per day, this
water system could handle only 200 additional residents. The Cresville
Heights water system in Beaver County serves approximately 10,500 users
with .85 mgd capacity. These figures indicate that, at capacity, only
55
-------
8l gallons per day is available per person (well below the 'rule of
thumb'). An influx of new users would further reduce the amount of
water available per person for all users in this local service area.
A great deal of variation in water consumption per capita was
evident for the case study counties and the Ohio ORBES counties. This
variation could be due to differences in the availability of sewer
services, the number of seasonal or temporary users (state parks, •„•
colleges) or other reasons. Overall, it appears that problem areas of
water supply shortages associated with demand created by inmigrating
power plant workers will be localized and difficult to predict, both in
location and severity of impact.
4.3.5 Tax Impacts
One of the major local benefits of an energy facility can be its
impact on property tax receipts. Most plants receive a very high
assessed valuation and as such, comprise a large proportion of the tax
base. Thus, we had originally hoped to analyze tax benefits and public
service costs in some detail as a part of our case study analysis.
Unfortunately, little tax information was made available to us either
by the utility companies or by the local or state government agencies
responsible for taxes. Thus, we have had to utilize the limited infor-
mation available to estimate the local tax impacts of power plants.
Information that was reviewed includes state tax laws, environ-
mental impact statements, and related socioeconomic impact studies.
Appendix A shows the EIS's from which information was obtained. The
property tax laws in each of the six states is different, although they
are parallel. Differences come in what is considered taxable as real
property, the revenues from which local and county governemnts are
supported, and what is taxed in other categories, the revenues from
which the state draws part of its income. We are interested in this
case in only the first quantity. Accordingly, we searched for informa-
tion on assessed values of power plant sites over time.
From the limited information available, we found that the
assessed values of a coal fired power plant ranged from $39,500,000 to
$94,809,000. On a per megawatt of nameplate capacity basis, data from
four plants yielded assessed values per MW of $40,385, $49,899,
$64,233, and $70,876. The average of these is approximately $56,000/MW.
We also found that the plant is only taxed at this valuation the
year following the completion of construction. All states have a
provision for assessment during construction. In jfest Virginia, this
assessment is based on the cost of construction materials, while in the
other states it is a function of the value of the land and partially
completed buildings. In many areas, the local assessor will not make
an assessment on the ouildings until the third or fourth year of
construction. Whether or not this will occur will dep_end on the
56
-------
individual assessor and is impossible to predict. Our first set of
calculations was done assuming that an assessment was made each year
of construction. From the scanty data we had available, the percentage
of the assessment charged by year of construction is assumed to be
0.2f0 after the first year, 6.^3$ after the second, 7.0$ after the third,
50/o after the fourth, 100% after the fifth (19.20). For tax purposes,
depreciation is assumed to be allowed at 3fo/year following the completion
of the plant. This is in congruence with state tax laws.
Expenditure estimates are made using the average per capita cost
for all local services derived from the Census of Governments (21).
Figures were derived for each of the case study counties. They were
found to be quite variant when looked at over time. Thus, the state
average per capita cost were assumed to be a more representative number.
Obviously, calculations performed using these assumptions will not
give exact results. However, we believe that the numbers used are.
representative of the average trends in tax revenues and expenditures
over time and will thus give good "ball park" estimates. The most
complete, local information available is for Jasper County, Illinois.
Using the ORBES Labor Impact Model, we derived the number of inmigrants
to the county over time. Additional expenditures were calculated
using an average family size of 1.8 persons per family and the average
state per capita annual services cost of $527. Revenues were calculated
based on the 1975 tax rate of ^.12$, multiplied by the current plant
assessed value calculated according to the assumptions listed above.
Results of these calculations are shown in Table 30. Here, one can
see that there are several potential impacts of the power plant location.
First, one can see that in the short run, namely the first two years
of construction, revenues will be less than the cost of additional
service demands. Second, we can see that' in the longer run, the plant
will bring a tremendous net benefit in terms of revenues. These monies
could be applied to improving services or to reducing the overall tax
rates. This second impact is calculated in the last column of the
table. Probably, a combination of the two effects will occur. We must
also note that these calculations are for Jasper County as a whole.
Settlement of additional construction or operation personnel in other
localities would mean that those localities would have to absorb the
additional costs without any tax benefits.
Overall, the tax impacts of a power plant can be summarized as
follows:
1) Wet, long term tax benefits to the taxing district and county
in which the plant is located.
2) Tax benefits result in the ability of the community to either
increase service levels or decrease tax rates.
3) There is a short term deficit in the amount of taxes collected
versus service costs. This would be exacerbated if the local
assessor delays in making an assessment as has happened
57
-------
Table 30
Estimated Jasper County Tax Impacts
oo
Year
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
Total Plant
Assessed Value
(lOOO's $)
69
149
2418
17276
34619
33659
34852
48402
64333
64615
77195
92183
89418
86735
Declines to
end of period
Revenues
@ 4.12%
lOOO's $
2.84
6.13
99.62
711.77
1426.30
1386.75
1435.90
1994.16
2650.52
2662.14
3180.43
3797.94
3684.02
3573.48
Declines to
end of period
Increased
Expenditures
@ 1.8 %/Person
Family
18.84
64.05
11.30
95.13
138.45
136.57
102.67
102.66
206.27
199.68
144.11
205.33
205.33
205.33
Stable to
end of period
Total New
County Assessed
Value
58678
58758
61027
75885
93228
92268
93461
107011
122942
123224
135704
150792
148022
145344
New
Tax
Rate
4.15
4.22
3.98
3.31
2.74
2.77
2.69
2.35
2.13
2.12
1.89
1.74
1.77
1.80
-------
in the past.
k) There is a possibility of a geographic disparity in tax
revenues versus service costs related to the degree to which
inmigrants settle outside the taxing districts benefiting
from plant tax revenues.
We might conclude, then, that overall, the local tax benefits of a
power plant are significant and positive, but that some policies might
be required to offset disparities in time or geographic distribution of
revenues.
59
-------
5.0 Classification of Case Study Counties for Impact Evaluation
As was mentioned above, each of the case study counties was
selected to represent a range of socioeconomic and environmental
characteristics representative of all candidate counties in the region.
In order to test the generalizability of our results, we derived a
taxonomy of all candidate counties in the region.
This chapter reviews the taxonomies used by other researchers for a
similar purpose. It then goes on to discuss a few of the problems
associated with generalization in socioeconomic impact analysis. Finally,
the classification we derived is presented in detail.
5.1 Classification in Other Studies
Several of the studies cited in the introduction to this report
have sought to use classification as a way to generalize from specific
county information to overall regional impacts. The basic underlying
assumption made in doing this is that counties with similar demographic,
economic, and social attributes will have similar propensities to be
impacted - in both a positive and a negative way. For example, counties
with large populations are generally assumed to easily absorb any
population influx which might occur during power plant construction. In
contrast, counties with few people are assumed to be more heavily
impacted.
With this type of assumption in mind, researchers at Argonne
National labs acted three primary factors influencing socioeconomic
Impacts:
The timing and relative magnitudes of the employment
requirements associated with the energy facility;
The economic characteristics of the local impact area; and
The sociodemographic characteristics of the impact county
(17, p.8-8).
Accordingly, they classified all of their candidate counties in relation
to four variables:
1) Population density at the time of impact;
2) Population density of the county and surrounding areas;
3) Distance in miles to the nearest regional trade center;
k) Relationship between basic and secondary employment.
These criteria were used to evaluate the impact potential for all
counties that were potential sites for coal development. A multi-
variate Euclidean Distance algorithm was used to put counties into one
of three groups. The first group includes counties with "a high
probability of adverse socioeconomic impact from energy development."
(17, p. 8-16). The second group has moderate assimilative capacities
and therefore less chance of adverse impacts. The third group can
accomodate large increases in coal development without major impacts.
60
-------
A similar classification was performed in a related study by
Oak Ridge National Laboratory (22). They classified counties into impact
groups based on the potential annual growth rate for population as
shown in Table 31.
Table 31
Oak Ridge National Lab Indicator of Social Impact
Annual Growth Rate Probability of Social Impacts
Less than 5% Low
% to 15% Moderate
More than 15% High
Source: Ref. 22, p. 9-18.
This is analagous to the population density variables of ML. Oak Ridge
researchers also derived an index of service impacts. They used factor
analysis to derive a linear weight associated with each of six variables:
population, percent of the population in urban places, median family
income, whether the county in a Standard Metropolitan Statistical Area
(SMSA), population density, and retail-wholesale-service trade. Using
the weights they calculated an index of service base capacity:
6 Y Y
I, = £ CW. ( ij i) + k] 100
J 1=1 . Sd.
where I. = the index value for county j
J
W.^ = weight of the ith variable, i = 1, 6
X.. = the level of the ith variable in the jth county
Id
X. = the mean value for the ith variable
0
Sd. = the standard deviation
X = a constant to make the index value equal to 100
k = when all variables take the mean value
(22, p. 9-35,36)
The two studies cited above both indicate that a major problem
associated with socioeconomic impact analyses to date is the scanty
information with respect to the causes and effects of such impacts.
As was pointed out in the introduction to this report, few studies
have been completed which trace actual socioeconomic impacts through
time. Thus, predictions of impact are based on generalizations from
these studies and many assumptions made by various researchers.
61
-------
The implications of these scanty data are that the level of
generalizations made, especially in regional studies, must be very
high. Thus, researchers at the two National Labs only derived three
groups of luipact counties.
It is clear however from the few monitoring studies that have
been made that the variation in socioeconomic impacts associated with
power plants is very great. Variations in many individual plant and
community characteristics may mean that two communities which are
almost equal in population size, population density, and income may
have quite different actual impacts.
The question which remains is whether other available data would
increase the probability of correctly identifying potential impacts.
Is it possible to refine the generalizations that have been made in
previous studies in order to make more explicit statements about
socioeconomic impacts? The remainder of this chapter discusses a test
of such a data set.
5.2 Classification of ORBES Candidate Counties
In an attempt to place our own case study counties into groups
about which some socioeconomic impact generalizations might be made,
we derived a taxonomy of our candidate counties in a manner parallel
to those cited above.
First, we assembled a data set which would represent a large array
of potential socioeconomic impact indicators. These included variables
on population, housing, income, employment, and natural resources.
Table 32 lists the variables used in the final analysis. Here, one can
see that the variables represent a much broader and more specific set
of indicators of the socioeconomic impacts of energy facilities than
did those used in previous studies.
The data were separately input to two analyses in order to classify
the candidate counties into groups based on the potential for socio-
economic impact. In the first analysis, the variables were first
grouped using factor analysis. This was necessary in order to reduce
the redundancy in the data set and because the next algorithm used
assumed that the data used were statistically independent. The factor
scores from the factor analysis represent such an independent data set.*
The second algorithm used was the H-GROUP program (23). It is basically
a Euclidean Distance-based clustering technique. The final result was
a grouping of ORBES candidate counties into four groups. The groups
will be described below.
*For a complete review of factor analysis and classification, the
reader is referred to Abler et al. (2k) and Rummel (25).
62
-------
Table 32
Variables Used in the Taxonomy of Candidate Counties
Variable Type
Population
Income
Housing
Employment
Variable
Total 1970 Population
Net Migration 1970-76
Dependency Ratio
Source
1970 Census
1970 Census and
Census Estimates
Derived from Census
Comments
Population
0-14+ 65 and
Over Divided
by Population
15-64
Natural Resources
Total Urban Population
Population Density 1975
Median Family Income
% Families Below Poverty
Level
% Persons on Public
Assistance - Aid to
Dependent Children
% Persons on Public
Assistance - Old Age
Median Effective Buying
Income
% of Housing Units Built
Before 1939
% of Housing Units Built
1960-70
% of Housing Units with
Public Water
% of Housing Units with
Public Sewer
% of Housing Units Vacant
Year Round
Total Housing Units
% Housing Units Lacking
Some Plumbing
% Housing Units with 71.51
Persons per Room
Total Employment 1970
% Workers Employed in
Agriculture
% Workers Employed in
Services
% Workers Employed in
Mining
% Workers Employed in
Manufacturing
% Workers Employed as
Craftsmen
% Land in Forest
1970 Census
1975 Census Population
Estimates
1970 Census
1970 Census
City and County
Data Book, 1972
City and County
Data Book, 1972
Sales Management, 1975
1970 Census
1970 Census
Measure of
Housing Age
Measure of
Service
Measure of
Service
Measure of
Vacancy
Measure of
Housing
Quality
Measure of
Crowding
1970 Census
1970 Census
1970 Census
1970 Census
1970 Census
1970 Census
63
-------
A second method of grouping was tested in order to get a measure
of the efficiency of the first method. Here, the original variables
were input to a discriminant analysis program. The groups derived
using the first procedure were input as well. The discriminant
analysis program derived three linear discriminant functions (very
analagous to the factor analysis functions) and tested the ability of
the functions to correctly classify the candidate counties. Only seven
of the ll^ counties were found incorrectly classified. After changing
these seven to the correct group, the analysis was repeated. The
resulting new discriminant functions placed 96$ of the counties into
the correct group.
Putting the technical aspects of these analyses aside, this means
that based on either approach to classification, the vast majority of
the counties were placed into groups which represent their differences
with respect to the socioeconomic variables. Theoretically then, these
groups whould be distinct enough to discern differences in potential
socioeconomic impacts.
Figure 3 shows a map of the ORBES region with the four county
impact groups. Table 33 shows the statistical breakdown of the groups
with respect to each of the original variables. What becomes immediately
apparent is that the values of the variables for some of the members of
each group overlap with values for the other groups. The mean values
are generally different but the range around the mean indicates this
overlap. The implication of this is that the classification does not
yield a distinct set of counties where socioeconomic impacts will be
the same.
An example will serve to illustrate the difficulty one encounters.
If one uses the mean housing vacancy rate as an indicator of housing
impacts, group 1 counties would have the highest probability of impact
because they have the lowest vacancy rate at 6.5$. Groups 3 and 4 would
be about equal at 7.2$ and group 2 would be the "best" at 8.1$. Yet,
if we look at the range or minimum and maximum we see that one county
in group 1 has a rate of 13.0$ while one in group 2 has a rate of 2.8$
Even though the classification algorithms work efficiently in statistical
terms, averaging characteristics over an array of variables yields a
grouping which is difficult to interpret for impact analysis. This same
problem must have arisen in previous studies but was not reported.
Thus, the generalizations made here and in those studies are subject to
so many exceptions as to make them meaningless.
It becomes apparent that the degree of generalization must be
relaxed. One way of doing this is to perform separate taxonomies using
the separate groups of variables - housing, population, income, and
employment. The resulting classifications will still yield some
exceptions to the generalizations that are made. However, more discrete
groups should form making the interpretation for impact analysis easier.
-------
Figure 3
COUNTY IMPACT GROUPS USING
ALL VARIABLES
ON
vn
GROUP 4
GROUP 3
GROUP 2
GROUP 1
PREPARED FOR OHIO RIVER BASM ENERGY STUDY
BY CAGIS/UKC, MARCH. 1980
-------
Table 33
Descriptive Statistics on Groupings Derived Using All Variables
Variable
% Older Houses
% New Houses
% Houses Served
by Public Sewers
% Houses Vacant
% Lacking Some Plumbing
% Families Below Poverty
% Net Migration '70-'76
Dependency Ratio
cr
^ Total Urban Population (1000's)
Total Population (1000's)
Median Family Income
Total Employment (1000's)
% Manufacturing Workers
% Agricultural Workers
% Mining Employees
Group 1
N=22
Mean Range
62.9
15.8
46.9
6.5
12.3
11.2
1.6
71.2
) 8.0
31.1
8463.0
11.4
24.0
15.3
2.3
26.2
10.6
58.5
10.2
26.6
11.4
23.5
15.0
120.8
156.3
3418.0
57.3
28.3
24.6
11.9
Group
N=48
Mean
60.0
16.0
38.0
8.2
18.2
15.6
3.5
68.2
18.7
49.1
7667.6
16.4
32.0
5.3
4.4
2
Range
45.4
23.1
64.9
16.0
47.0
30.5
35.8
40.9
159.6
206.7
5746.0
73.6
50.2
16.5
24.2
Group 3
N=42
Mean
58.2
19.2
47.8
7.2
15.2
12.3
2.2
68.3
16.2
34.0
8099.3
12.3
32.4
8.2
1.9
Range
58.8
32.8
83.9
8.7
33.0
31.2
25.1
23.9
77.0
102.6
4890.0
36.3
27.9
26.4
11.3
Group
N=2
Mean
40.0
24.5
85.2
4.8
3.4
8.6
-7.0
63.7
773.0
809.5
10153.0
311.2
32.3
0.5
0.1
4
Range
10.8
4.6
15.5
0.5
0.6
0.6
3.3
1.7
230.1
229.0
667.0
85.1
0.4
0.1
0.1
Sources: 1970 Census of Population and Housing and
1976 Population Estimates of Bureau of the Census
-------
Figures k to 7 show the results of the separate classifications
using the separate sets of variables. Here, one can see that the
groupings change significantly depending on the set of variables being
analyzed. Table 3^- summarizes the differences in the groups for each
variable set . One should, note that group k is alwe.ys only two rather
unique counties - Jefferson County, Ky. (Louisville); Hamilton County,
Ohio (Cincinnati).
A complete discussion of the groupings is given in another report
(26). For now, we will point out where our case study counties fall.
This is shown in Table 35. As one can see, the case studies come from
a range of groups. The group membership changes for some of the
counties as one goes from one variable set to another. Table 36
summarizes what are the characteristics of the groups in terms of the
probability of impacts.
If we compare Tables 35 and 36, we can derive the set of expected
impacts associated with power plant development in the case study areas.
These can then be compared to our findings based on actual data. We do
not expect that these results will be absolutely clear cut. However, we
would expect that the expected impacts would put us in the right "ball
park". If it does not, we must restrict our regional generalizations.
If we are correct, we can make our generalizations with more confidence.
Such a comparison will be drawn in the final chapter of this report.
67
-------
Figure k
COUNTY IMPACT GROUPS USING
INCOME VARIABLES
H GROUP 4
Bi GROUP 3
E3 GROUP 2
E3 CROUP 1
mCPAREO FOR OKU RIVER BASK ENERGY STUD*
BY CAGtS/UKC. MARCH, 1980
-------
Figure 5
COUNTY IMPACT GROUPS USING
EMPLOYMENT VARIABLES
VD
GROUP 4
GROUP 3
GROUP 2
GROUP 1
PREPARED FOR OHIO RIVER BASIN ENERGY STUDY
BY CAGtS/UCC. MARCH, 1980
-------
Figure 6
COUNTY IMPACT GROUPS USING
HOUSING VARIABLES
GROUP 4
GROUP 3
GROUP 2
GROUP I
PREPARED FOR OHIO RIVER BASM ENERGY STUD*
BY CAGIS/UCC, MARCH. 1980
-------
Figure 7
COUNTY IMPACT GROUPS USING
POPULATION VARIABLES
GROUP 4
GROUP 3
GROUP 2
GROUP t
PREPARED FOR OHIO RIVER BASH ENERGY STUDY
BY CAGrS/UKX, MARCH. 1980
-------
Table 3k
Group Statistics for Selected Variables Using Alternative Classification Schemes
Classification Based on Housing
Classification Based on Income
Variable
% Old Housing
% New Housing
% Housing Vacant
% Net Migration '70-'76
Total Urban Population (1000's
-0 Median Family Income
ro
Total Employment (1000's)
% Manfac. Employees
% Agricultural Employees
Group 1
N=28
Mean
64.8
14.7
6.2
1.6
;) 8.0
8450.
11.4
30.5
11.3
Group 2
N=46
Mean
61.6
15.4
8.7
3.4
15.4
7420.
14.7
30.4
8.6
Group 3
N=38
Mean
54.2
21.2
7.0
2.5
21.5
8328.
14.7
30.9
5.9
Group 4
N=2
Mean
40.0
24.4
4.8
-7.0
773.0
10153.
311.2
32.3
0.5
Group 1
N=21
Mean
62.6
15.3
6.1
1.4
15.3
8481.
13.0
25.7
13.5
Group 2
N=44
Mean
61.5
15.6
8.2
2.8
15.1
7473.
14.8
31.2
5.7
Group 3
N=42
Mean
58.4
19.2
7.9
3.4
12.5
8023.
10.3
31.9
9.2
Group 4
N=7
Mean
44.9
22.3
3.9
-1.9
255.4
10112.0
117.4
34.5
2.3
-------
Table 3^ (Cont'd)
Variable
% Old Housing
% New Housing
% Housing Vacant
(jo % Net Migration '70-'76
Total Urban Population
(1000's)
Median Family Income
Total Employment (1000*s)
% Manufacturing Employment
% Agricultural Employees
Classification
Group 1
N=38
Mean
62.5
16.4
6.8
1.0
9.5
1359.
11.4
28.5
12.6
Group 2
N=47
Mean
58.9
16.6
8.0
3.2
21.6
7667.
17.7
31.8
4.4
Based on Population
Group 3
N=27
Mean
58.0
19.3
7.5
4.1
13.9
8015.
10.7
31.5
9.2
Group 4
N=2
Mean
40.0
24.5
4.8
-7.0
773.0
10152.
311.2
32.3
0.5
Classification Based on Employment
Group 1
N=29
Mean
62.7
16.2
7.7
3.2
5.4
7498.
7.0
23.9
17.8
Group 2
N=46
Mean
60.1
16.4
7.3
2.0
24.3
8039.
19.2
33.5
3.4
Group 3
N=37
Mean
57.6
18.9
7.5
3.1
13.0
8301.
12.7
32.2
7.1
Group 4
N=2
Mean
40.0
24.5
4.8
-7.0
773.0
10152.
311.2
32.3
0.5
-------
Table 35
Group Membership for Case Study Counties
Groups Based on Variables In2
County Housing Population Employment Income
Jasper, 111. 11 11
Jefferson, Ind. 33 32
Spencer, Ind. 21 13
Trimble, Ky. 3 3 1 3
Adams, Ohio 22 12
Beaver, Pa. 22 21
Mason, W. Va. 2 2 2 2
-------
Table 36
Description of the Classification of Candidate Counties
and Potential for Socioeconomic Impacts
Variable Type
Population
Housing
Income
Employment
Group Group Descriptions
1 Smallest populations, density, most
rural
2 Largest populations, density, most
urban, lowest dependency ratio
3 Medium size, density, dependency
ratio
1 Fewest units, lowest vacancy, many
with public sewer, water
least crowded units
2 Largest # units, largest vacancy,
most crowding fewer with public
sewer, water
3 Medium # units, vacancy, crowding,
most with public sewer, water
1 Fewest below poverty, largest median
income, largest buying income,
fewest old age on assistance, largest
ADC*
2 Highest families below poverty, lowest
median income, lowest buying income,
medium # persons on public assistance
3 Median income between year 2 & 3,
families below poverty, ADC, buying
income, highest old age public
assistance
1 Most people in agriculture, lowest
workforce, lowest in manufacturing,
services, craftsman
Fewest in agriculture, most manu-
facturing, mining, total employees
medium in services
Medium in agriculture, total
employees, craftsmen, manufacturing,
highest in services, lowest in mining
Potential for
Impact
High
Low
Medium
Medium to
High
Low
Medium to
High
Low
High
Medium
Highest-
Induced
migration
but lowest
employment
benefits
Lowest,
induced
migration
Highest
employment
benefits
Medium
75
-------
6.0 Summary and Conclusions
Our case study analysis set out to delineate those site specific
socioeconomic impacts that can occur due to power plant location and to
test several of the assumptions and models being used in the regional
analysis. Data limitations have prevented us from doing a complete
evaluation on all the case study areas. However, we have been able to
accomplish our major goals. Our results from the case study analysis
can be summarized as follows:
- A wide range of socioeconomic issues can arise from any of the
three phases in power plant siting: land acquistion, construction
or operation.
- From the few monitoring socioeconomic impact studies that have
been undertaken several trends have become apparent that lead to
the hypothesis that the potential for large socioeconomic impacts
should increase in those areas with fewer local workers avail-
able, less available rental housing and fewer available public
services.
- In attempting to generalize local impacts to the regional level
some classification of impacts and/or counties is necessary.
- The case study counties are indicative of the wide range of
demographic, housing, income and employment characteristics
that exist in the ORBES region.
- From historical data, 1965-1975? in energy development and
socioeconomic conditions in Adams County, Beaver County and
Jasper County we found that
-- the most substantial evidence of population, employment and
county fiscal impacts occurred in Adams County - a small,
rural, low income county.
— little or no evidence of energy-related socioeconomic
impacts was found in Beaver County, an SMSA county that is
highly urbanized.
— concurrent development of manufacturing industries,
institutions, highways, or any other sectors can obscure the
effects from power plant construction making it impossible to
separate out cause and effect.
- Partial correlation and regression analysis was used to confirm
that our model-derived construction employment figures are
reasonable estimates of the observed changes in employment.
- Spencer County had the highest peak construction employment,
2,988 workers; Adams County the lowest at UOO workers.
- In only Trimble County does the number of inmigrants (including
families estimated as 1.8 times the inmigrant workers) go above
of the 1975 population.
76
-------
- An annual building rate was estimated from 1970 Census of
Housing data to compare with the additional demand created
by inmigrants for any given year of the scenario. Using an
additional demand comprising more than 50% of the historical
building rate as a cutoff point, five out of the seven case
study counties show potential housing shortage problems.
Jefferson and Beaver Counties, both semi-urbanized, are not
among these five.
- Housing shortages can create both positive and negative effects.
Construction and service industries may be stimulated. However,
housing costs will almost certainly go up. Temporary housing
may arise and be considered a negative effect by some but not by
others.
- Water supply shortages associated with demand created by
inmigrating power plant workers will be localized if they
occur at a,11 and are difficult to predict.
- Tax impacts of a power plant include net, long term tax benefits
to the taxing district and county in which the plant is located.
- Tax benefits result in the ability of the community to either
increase service levels or decrease tax rates.
- In the short term there can be a deficit tax revenues versus
public service costs.
- There is a possibility of a geographic disparity in tax revenues
varsus service costs related to the degree to which inmigrants
settle outside the taxing district receiving the revenues from
the plant.
- A wide array of socioeconomic variables was used to classify
ORBES candidate counties into h groups. From our analysis of the
classifications and in attempting to make generalizations based
on the classifications we found that it was more meaningful to
use separate classifications based on separate groups of variables
relating to a specific impact area. Our resulting classifications
are thus more refined than those employed in previous regional
studies.
In conclusion, we can state that there are positive and negative
socioeconomic impacts associated with power plant development. We have
identified many but not all of these in our site specific analysis. We
have been able to confirm many of our modeling assumptions as giving
correct results and thus can proceed to use these methods in our
regional analysis with greater confidence.
77
-------
References
(1) Argonne NationalLaboratory, A Preliminary Assessment of
the Health and Environmental Effects of Coal
Utilization in the Midwest/ Argonne/ 111;
Argonne National Laboratory, January 1977.
(2) Sanderson, Debra and Michael O'Hare, Predicting the
Local Impacts of Energy Development; A Critical
Guide to Forecasting Methods and Models, Cambridge,
Mass: Laboratory of Architecture and Planning,
M.I.T., May 1977.
(3) Mountain West Research, Inc., Construction Worker
Profile; Community Reports (7 volumes), prepared
for The Old West Regional Commission, Washington,
D.C., December 1975.
(4) Mountain West Research, Inc., Construction Worker
Profile; Summary Report, prepared for The Old
West Regional Commission, Washington, D.C.,
December 1975.
(5) Mountain West Research, Inc., Construction Worker
Profile; Final Report, prepared for The Old
West Regional Commission, Washington, D.C.,
December 1975.
(6) Pennsylvania Power and Light, A Monitoring Study of
Community Impacts for The~Susquehanna Steam
Electric Station, Allentown, June 1976.
(7) Tennessee Valley Authority, Semiannual Construction
Employment at TVA Plants, Employment by Kraft,
obtained through correspondence with G.R. DeVeny,
TVA, 1978.
(8) Tennessee Valley Authority, Hartsville Nuclear Plants:
Socioeconomic Monitoring and Mitigation Report,
Survey Reports for September 1976, March 1977, and
September 1977, Knoxville, Tennessee, 1977 and 1978.
(9) Tennessee Valley Authority, Beliefonte Nuclear Plant;
Construction Employment Impacts, Survey Reports for
April 1975, May 1976 and August 1977; Knoxville,
Tennessee, 1975, 1976 and 1977.
78
-------
References (Cont'd)
(10) Tennessee Valley Authority, Watts Bar Nuclear Plant;
Construction Employment Impact, Survey Reports for
May 1971, March 1973, July 1974, and May 1976;
Knoxville, Tennessee, 1971, 1973, 1974 and 1976.
(11) Tennessee Valley Authority, Sequoyah Nuclear Plant:
Construction Employment Impact, Survey Reports
for October 1970, July 1972, August 1974 and
February 1977; Knoxville, Tennessee, 1970,
1972, 1974, and 1977.
(12) Tennessee Valley Authority, Browns Terry Nuclear Plant;
Construction Employment Impact, Survey Reports
for September 1969, April 1971 and July 1973;
Knoxville, Tennessee, 1969, 1971 and 1973.
(13) Tennessee Valley Authority, Cumberland Steam Plant;
Construction Employment Impact, Survey Reports
for September 1969, May 1971, and April 1973;
Knoxville, Tennessee, 1969, 1971, and 1973.
(14) Pennsylvania Power and Light Company, A Monitoring
Study of Community Impact: An Update.
Allentown, Pennsylvania: P.P. & L. 1978.
(15) von Rabenau, B., "The Employment and Population Impact
of Energy Facility Construction on Urban Areas
in the ORBES Region", Draft Report, prepared for
The Ohio River Basin Energy Study, September 1979.
(16) Urban Systems Research and Engineering, Classification
of American Cities for Case Study Analysis by
Elizabeth Cole et al. Report for Office of
Research and Development, USEPA, Washington, D.C.:
Urban Systems Research and Engineering, July.1976.
(17) Argonne National Laboratory, An Integrated Assessment of
Increased Coal Use in the Midwest; Impacts
and Constraints. Argonne, 111.: ANL/AA-11
(Draft Report), October 1977.
(18) The Ohio State University and Purdue University,
Preliminary Technology Assessment Report, Vol. II-A,
Part 2 of Ohio River Basin Study, Indiana University,
May 15, 1977.
79
-------
(19) Darling, David, Data on Indiana Power Plant Tax Payments,
Miraeo, 1979.
(20) Ohio Department of Taxation, Report on Kyger Creek
Tax Payments, Mimeo, 1968.
(21) U.S. Bureau of the Census, 1972 Census of Governments,
Washington D.C.: U.S. Government Printing Office,
1974.
(22) Oak Ridge National Laboratory, National Coal Utilization
Assessment, Oak Ridge, Tenn.:Oak Ridge National
Laboratory, October 1978.
(23) Veldman, D.J. (ed.), Fortran Programming for the Behavioral
Sciences, New York: Holt, Rinehart and Winston,
1967.
(24) Abler, R.; Adams, J.S.; and Gould, P., Spatial
Organization, Englewood Cliffs, N.J.: Prentice-
Hall, Inc., 1971.
(25) Rummel, R.J., Applied Factor Analysis, Evantson, 111.:
Northwestern University, 1970.
(26) Gordon, S.I. and Graham, A.S., Regional Socioeconomic
Impacts of Alternative Energy Scenarios for the
Ohio River Basin Energy Study Region, Columbus,
Ohio: Department of City and Regional Planning,
The Ohio State University, forthcoming.
80
-------
Appendix A
List of Data Sources Used for Site Specific Analysis
County
All
Jasper, 111,
Source of Information
U.S. Bureau of the Census, County
Business Patterns 1965-1976
U.S. Bureau of the Census, 1970
Census of Population
U.S. Bureau of the Census, 1970
of Housing
U.S. Bureau of the Census, Current
Population Reports, Series
P-25, 1977
U.S. Bureau of the Census, 1977
Census of Governments
Illinois Tax Department
Jasper County Treasurer's Office
Alan Rudolph, Central Illinois
Public Service Company
Illinois Institute for Natural
Resources
Jefferson, Ind. Gary Stegner, Region 12 Development
Commission
Indiana Department of Public
Instruction
Jefferson County Auditor
Public Service of Indiana
Spencer, Ind. Spencer County Auditor
Indiana 15 Regional Planning
Commission
Indiana Dept. of Public Instruction
Indiana and Michigan Electric Co.
81
Type of Information
Employment by Industry
Population, Income,
Occupations
Housing
Population, Per Capita
Income
County Revenues and
Expenditures
Property Valuations
County Disbursements
and Tax Revenue
Newton Plant
Water Systems
Housing, Property Tax,
Employment, Labor Force
School Enrollment
Assessed Valuations,
County Disbursements
and Revenues
Marble Hill
County Expenditures
and Revenues, Property
Valuations
Housing, Population
School Enrollment
Rockport Plant
-------
County
Spencer, Ind.
Trimble, KY
Adams
Appendix A (continued)
Source of Information Type of Information
Martin T. Pond Memo to Spencer Impact Study Report
County Study Committee
Concerning Rockport Plant,
April 2, 1979
"An Assessment of the Greater
Grandview Community and Its
Survey and Impact Analysis
Opportunities for Improvement"
by Community Development
Staff and Students at Purdue
University
Kentuckiana Regional
Planning and Development
Agency
Trimble County Clerk
Kentucky Auditor of Public
Accounts
Housing, Police, File,
Water System
Property Valuations
County Expenditures
Louisville Gas and Electric Trimble Co. Plant
Mrs. Helen Roush
Adams Co. Reg. Planning
Commission
Jeffrey A. Spencer
Ohio Valley Regional
Development Commission
Ohio Valley School District
Ohio Tax Equalization Dept.
Ohio Bureau of Employment
Services
Financial Reports for
Counties
Auditor's Office
Bob Berker
Killen Power Station
ODNR, Inventory of
Municipal Water Supply
Systems by County, Ohio
Ohio Water Inventory Report
# 24 82
Causes for Population
Growth
Water and Sewer Systems,
Housing Conditions,
Health Services,
Fire and Police Services,
Manufacturers
School Enrollment
Property Valuations
Labor Force, Unemployment
Expenditures and
Resources
Killen Plant
Water System
-------
County
Beaver, Pa.
Mason
Appendix A (continued)
Source of Information
Type of Information
Pennsylvania Bureau of Employment Labor Force,
Security Unemployment
Pennsylvania Abstracts
Green International Inc.,
Comprehensive Water Quality
Management Plan May 1976
for Pa. Dept. of Environmental
Resources
Ohio Edison
Region II Planning and
Development Commission
Mason County Board of Education
Appalachian Power Company
State Tax Commissioner "Study
of Property Valuations As
They Relate to Levies Laid for
the Support of Schools in
West Virginia"
West Virginia Tax Department,
Local Government Relations
Division
Enrollments, Property
Valuations, County
Expenditures &
Revenues
Water Systems
Beaver Valley Plant,
Bruce Mansfield Plant
Major Industries,
Community Facilities,
Sewer and Water,
Police and Fire
School Enrollments
Mountaineer Plant
Property Valuations
County Expenditures
and Revenues
83
-------
APPENDIX B
ORBES Labor Impact Model
The ORBES Labor Impact Model estimates the following
impacts by year for every county hosting the construction of
an electric generating facility:
1. Construction manpower requirements
2. Operations manpower requirements
3. Inmigration from construction workforce
4. Inmigration from construction and operation workforce
5. Employment multiplier effect
6. Payroll generated by total construction and operation
workforce
7. Payroll generated by inmigrant workforce (local payroll)
8. Income multiplier effects
9. State income tax generated from local payroll
Regional impacts are also estimated for each year of the scenario:
10. Total construction manpower requirements
11. Total operation manpower requirements
12. Total construction manpower requirements by skill
Each of these items will be discussed briefly below.
Construction and Operation Manpower Requirements (Items 1 and 2)
-------
A ratio of construction manpower requirements per megawatt
capacity was derived for the following types of energy facilities:
coal, single unit, no scrubbers
coal, part of a multiple unit, no scrubbers
coal, single unit, scrubbers
coal, part of a multiple unit, scrubbers
nuclear
A similar ratio was derived for operation manpower
requirements for coal units with scrubbers, coal units without
scrubbers and nuclear units. These ratios were derived as an
average of actual or reported manpower requirements per megawatts
as published in EIS's, ER's, or other reports/publications.
The construction schedules used in the model are:
coal units less than 1000 MW 5 years
coal units 1000 MW or greater 6 years
nuclear units 7 years
Immigration (Items 3 and 4)
Rates of inmigration of construction workers are permitted
to vary for three different situations: 1)SMSA counties
2) counties within 50 miles of an SMSA, .and 3) counties outside
the 50 mile range. In our basic runs we use 5,10 and 30 percent
as the rates of inmigration for the respective situations listed
above.
For operation workers it is assumed that fully 100%
of the workforce migrates to the host county
85
-------
Employment Multiplier (Item 5)
Employment multipliers by sub-region for manufacturing
and construction were taken from an Argonne National Laboratory
report written by Erik J. Stenehjem and James E. Metzger entitled
A Framework for Projecting Employment and Population Changes
Accompanying Energy Development Phase I (August, 1976). These
multipliers are:
Indiana 1.5
Indiana 1.2
Kentucky/Tenn 1.5
Ohio 1.5
Northeastern 1.4
States (Penn)
West Va/Va 1.1
Maryland
Payroll (Items 6 and 7)
Payroll is input into the model as an average for
all construction and operation workers. Our initial sums include
$18,000 as this value.
86
-------
Income Multiplier (Item 8)
The income multipliers used in the model are approximately
15% less than the employment multipliers. These are:
Illinois 1.28
Indiana 1.05
Kentucky 1.28
Ohio 1.28
Pennsylvania 1.19
West Virginia 1.00
State Income Tax (Item 9)
Average state income tax rates are part of the input
to the model. Our i-itial sums use the following 1975 rates
derived from a U.S. Bureau of the Census report entitled State
Tax Collections in 1976.
Illinois $150 per $10,000 income
Indiana $134
Kentucky $151
Ohio 77
Pennsylvania 142
West Virginia 135
87
-------
Regional Construction and Operation Manpower (Items 10 and 11)
Regional Manpower Requirements are simply the sums
of the county totals by year.
Skill (Item 12)
To break down the labor demand by skill it was necessary,
first, to separate the construction workers at coal-fired plants
from those at nuclear units. The percentage of workers by skill
category for coal and nuclear units was derived as an average
of the figures presented in the literature. These percentages
were applied to the yearly labor demand for coal and nuclear
units and then summed to give a total regional labor demand
by skill.
A detailed explanation of Items 1,2, and 12 was contained
in an ORBES memo from Steve Gordon dated June 9, 1979 and is
also given in another report. (Gordon and Graham, 1979) Please
refer to these writeups for an explicit statement of methods and
references used to derive manpower and skill demands.
------- |