71 11,915
-
MC GUIRE, Michael Edwin, 1938AN APPROXIMATION OF MICHIGAN'S SPATIAL WATER
NEEDS.
Michigan State University, Ph.D., 1970
Geography
U niversity Microfilms, A XEROX C o m p an y , A nn A rbor. M ichigan
AN APPROXIMATION OF MICHIGAN'S
SPATIAL WATER NEEDS
By
Michael Edwin Me Guire
A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Resource Development
1970
ABSTRACT
AN APPROXIMATION OF MICHIGAN'S
SPATIAL WATER NEEDS
By
Michael Edwin Me Guire
Water has been in the past, and will continue to be
in the future, of vital importance to Michigan's continued
growth and development.
Demands on the state's water re
source base are increasing, however, and if all of the var
ious uses of water are to be satisfied forecasts of future
use will have to be made.
The purpose of this study, there
fore, was to examine the factors which have been instrumen
tal in affecting the present spatial pattern of water use in
Michigan, to consider changes in these factors over time,
and to estimate what the future demands for water in Michigan
may be.
Water withdrawals in any area have a certain struc
ture, the principle withdrawals in Michigan being made for
domestic, municipal, and commercial uses, industrial uses,
and for agriculture.
Estimates of future spatial water d e
mands were made by applying unit withdrawal factors to fore
casted magnitudes of the withdrawal u n i t s .
Withdrawals for domestic, municipal, and commercial
purposes are expected to increase at a rate greater than the
corresponding growth of population, a result of improved
socio-economic status.
These estimates were made by m ulti
plying a per capita withdrawal rate times the projected
population.
The results indicate that domestic, municipal,
and commercial withdrawals were approximately 162 billion g al
lons per year in 1940 and had grown to over 250 billion gal
lons by 1960.
13y the year 2000 annual withdrawals are ex
pected to be almost 500 billion gallons.
Withdrawals for manufacturing purposes were estimated
by multiplying a per employee withdrawal rate times the num
ber of manufacturing employees expected in the state.
Manu
facturing withdrawals are believed to have been approximately
275 billion gallons per year in 1940.
With the rapid growth
of manufacturing this increased to about 775 billion gallons
in 1960.
By 2000 withdrawals for manufacturing should be
over 1,000 billion gallons per year.
Domestic, municipal, and commercial withdrawals and
withdrawals for manufacturing are highly concentrated in the
southern third of the state;
15 counties account for over 80
per cent of these withdrawals.
In addition, manufacturing
withdrawals were concentrated in four manufacturing types.
Approximately one-half of one gallon of water is re
quired to produce one kilowatt minute of electricity.
In 1940
this amounted to slightly over 330 billion gallons per year.
With the expected growth of population and economic activity
this may increase to over 6,500 billion gallons per year by
2000.
Most of the withdrawals will be concentrated along the
Great bakes because of the tremendous quantities of water
available.
Rapid changes in Michigan's agriculture has precluded
specific estimates of withdrawals for agricultural purposes.
However,
the expansion of irrigated acreage and technological
developments in livestock production are expected to affect
significant withdrawal increases.
These should be concen
trated in the southeastern and southwestern parts of the
state, the major areas of irrigation at the present time.
ACKNOWLEDGMENTS
The participants in this thesis are legion and it would
certainly be some form of plagiarism not to give full recogni
tion to each.
Unfortunately space does not permit this oppor
tunity and a less ambitious attempt must be made.
Greatest help was received from members of my family
who, through both encouragement and freedom from family re
sponsibilities, provided an environment conducive to study.
Perhaps over the years the debt can be repaid.
Faculty help from the Department of Resource Develop
ment has guided me around numerous pitfalls and renewed my
confidence when this was sorely lacking.
cially
Appreciation espe
is due to Drs. Barlowe and Steinmueller whose doors
were always open.
Fellow graduate students in the department combined to
establish a situation in which academic standards were held
high.
without the competition offered by these men the com
pletion of this thesis would have been in doubt.
Finally, gratitude is expressed to the Michigan Water
Resources Commission whose financial assistance permitted un
divided attention to the thesis for a year.
ii
TABLE OF CONTENTS
LIST OF TABLES
..................................
. . . . .
V
LIST OF F I G U R E S ............................................. V i i
LIST OF A P P E N D I C E S .......................................... Viii
Chapter
I.
BACKGROUND
TO THE S T U D Y ........................
1
Purpose
Conceptual Framework
Objectives
Scope
Introduction to Methodology
Assumptions
II.
FACTORS CONTRIBUTING TO DOMESTIC, MUNICIPAL
AND COMMERCIAL WATER WITHDRAWALS .............
22
Introduction
Composition of Withdrawals
Elements of Domestic Withdrawal Increases
III.
GENERAL POPULATION CHARACTERISTICS AFFECTING
DOMESTIC, MUNICIPAL, AND COMMERCIAL WATER
WITHDRAWALS IN MICHIGAN
..............
35
Introduction
Population Growth
Changes in Population Characteristics
Population Projections
IV.
DETERMINATION OF DOMESTIC, MUNICIPAL,
AND COMMERCIAL WATER WITHDRAWALS ..............
Introduction
Possible Approaches
Data Availability and Deficiencies
Examination of Possible Approaches
iii
51
Chapter
V.
FACTORS CONTRIBUTING TO MANUFACTURING
WATER W I T H D R A W A L S .................................. 77
Introduction
Nature of Water Use in Manufacturing
Manufacturing Activity in Michigan
VI.
DETERMINATION OF MANUFACTURING
WATER W I T H D R A W A L S ................................. 110
Estimates of Manufacturing Employment
Water Withdrawal Estimates
VII.
WATER WITHDRAWALS FOR POWER GENERATION ..........
148
Importance
Nature of Withdrawals
Ouantity Withdrawn
Location of Withdrawals
VIII.
AGRICULTURAL WATER WITHDRAWALS ...................
154
Importance
Land Use in Michigan
Water Use in Agriculture
IX.
SUMMARY t FINDINGS , AND R E C O M M E N D A T I O N S ........... 211
Summary
Findings
Implications and Recommendations
BIBLIOGRAPHY
...............................................
228
A P P E N D I C E S ................................................... 236
iv
LIST OF TABLES
Table
Page
1.
Michigan Surface Water Resources .................
2.
Price Elasticity of Residential Water Demand
3.
Population Growth: US, ENC, M i c h i g a n ............
37
4.
Regional Population Concentration
..............
45
5.
Michigan Urban and Rural Population
............
49
6.
Regional Service from Municipal Systems
7.
Regional Non-Industrial Withdrawal
Coefficients
....................................
76
8.
Employment Trends in Michigan
...................
88
9.
Total State Employment, Projections
to 2020 ...........................................
117
10.
Total State Employment,
1964-1968
. . . . . . .
118
11.
Regional Manufacturing Employment
..............
132
12.
Error R e d u c t i o n ...................................... 141
13.
Water Withdrawal Coefficients
14.
Number of farms; State Total and
Regional Distribution ...........................
. .
. . . .
. . . . . . . . .
2
31
70
145
175
15.
Loss of Farms in M i c h i g a n .......................... 177
16.
Land in Farms: State Total and Regional
Distribution
....................................
178
17.
Average Farm Size in M i c h i g a n ..................... 179
18.
Cropland Harvested in Michigan,
............
180
19.
Irrigated Acreage: State Total and
Regional Distribution
..........................
v
183
1964
Taole
20.
Page
IJumbor of Livestock: State Total and
Regional Distribution
.........................
205
21.
Livestock Specialization in Michigan
.........
206
22.
Summary of Michigan's Water Withdrawals . . . .
214
23.
Domestic, Municipal, and Commercial
W i t h d r a w a l s ...................................... 217
24.
Composition of Manufacturing Withdrawals
by T y p e ........................................... 219
25.
Regional Composition of Manufacturing
W i t h d r a w a l s ...................................... 220
vi
LIST OF FIGURES
Figure
Page
1.
Planning R e g i o n s ................................
14
2.
Normal Potential Evapotranspiration ............
28
3.
Normal D e f i c i t ...................................
29
4.
Price-Deraand for Domestic Water:
Western Cities ..................................
32
Michigan Demographic Rates: Live Births,
Deaths, Natural Increase 1900*1964 ...........
40
6.
Population L o s s ...................................
43
7.
Effective Buying Income Per Household, 1968 . .
47
8.
Relationship Between Population Served
and Total Withdrawals
.........................
66
Total Employment and Per Cent Manufacturing . .
90
5.
9.
10.
Total Employment
................................
11.
Manufacturing Employment
12.
Installed Generating Capacity
13.
Power Plant Location, 1970
14.
Length of Growing S e a s o n ..........................161
15.
Major Irrigated Crops, 1964 .....................
195
16.
Irrigated Acreage, 1964 ..................
196
17.
Major Cattle, Hogs, and Sheep
Producing Counties .............................
209
vii
.......................
................
.
120
128
149
153
LIST OF APPENDICES
Appendix
Table
Page
1.
Sources of Michigan Population Growth ..........
236
2.
County Population Concentration .................
237
3.
Comparison of Population Estimates
............
238
4.
Non-Industrial Withdrawals, United States . . .
239
5.
Independent Variables for Regression
M o d e l .............
239
Per Capita Non-Industrial Withdrawals,
1965 .............................................
241
Per Cent Served by Municipal Systems,
M.D,P.H. Regions ...............................
242
8.
Increase in Withdrawal Rates
244
9.
Manufacturing Composition, Michigan . . . . . .
244
10.
Regional Manufacturing Employment ..............
245
11.
Structure of Regional Manufacturing
E m p l o y m e n t ......................................... 246
12.
Spatial Composition of Regional
Manufacturing Employment ......................
249
Manufacturing Employment as a Proportion
of Total Employment
...........................
256
Regression Coefficients for Lower
Peninsula Regions
.............................
256
15.
Regression Coefficients for Counties
..........
257
16.
Derivation of State SIC Projections ............
259
17.
Regression Coefficients for State SIC
E m p l o y m e n t ......................................
259.
6.
7.
13.
14.
viii
...................
Appendix
Table
18.
Page
Water Recirculation ..............................
260
Appendix
Figure
1.
Michigan Department of Public Health
Planning Regions .................................
ix
243
CHAPTER I
BACKGROUND TO THE STUDY
Purpose
That water is of inestimatable value to an area is
seldom disputed.
Whatever the nature of the area, unless it
is completely unpopulated, water is of major importance.
It
is not necessary to expand the discussion of this point other
than to say that in virtually all aspects of water use the
demand is increasing in almost every area of the nation.
Pop
ulation growth and the expansion of economic activity dictate
that increasing amounts of water will be utilised in the fu
ture.
Water use in Michigan is particularly important.
Mich
igan ranked seventh in population in 1960, and in value added
by manufacturing in 1963 it ranked sixth.
Thus from sheer
size, vast amounts of water are required and more will un
doubtedly be required in the future as population and eco
nomic growth continue.
In Michigan water quality has not been a major prob
lem.
With location on four of the Great Lakes, with a large
amount of inland water, and with abundant groundwater re
sources,
the state has not felt the pressure of the demand
1
2
for water impinging on a limited supply as have other states,
une measure of Michigan's surface water resources is indi
cated in Table 1.
Table 1
Michigan Surface Water Resources*
Source
Area
Great Lakes
24,688,000 Acres
Inland Lakes
841,000 Acres
Inland Streams
36,350 Miles
*Clifford Humphrys, Michigan Lakes and Ponds (East
Lansing: Agricultural Experiment Station, Michigan State Uni
versity, 1965), quoted in Raleigh Barlowe, Implications of
Land and Water Use Developments in Michigan tor Future public
Water Resource Policy (East Lansing: Department off Resource
Development, Michigan State University, 1966), p. 1.
If all of the various uses of water in the state are
to be satisfied in the future expectations of future uses
will have to be made.
Although adequate water resources to
satisfy present uses are available there is no assurance that
they will be adequate in the future.
Water policy made today
will affect the adequacy of water resources tommorow.
These
policy decisions can be made with some degree of certainty if
some idea of future requirements can be obtained.
The developments which influence the amount of water
utilized have not in the past been equally important in all
areas of the state.
Some sections of the state, notably the
northern parts, are probably using less water in 1970 than in
3
19 50.
On the otlier hand, most of the overall increase in
population and economic activity, and as a result increases
in water use, are taking place in 15 or 20 of the southern
counties.
If the supply and demand situation for water is to re
main adequate information must be attained on the structure
and amount of water use throughout the state.
Consequently,
the purpose of this study is to examine the factors which
have been instrumental in influencing the present spatial pat
terns of water use in Michigan, and to consider the changes
which may be expected in the future.
The focus of this ef
fort will be on the non-marginal changes which have taken
place over the years in the elements of water use in order to
indicate the probable developments in the next few decades.
Estimating water use for the state as a whole is a
very complex matter and involves the use of more resources
than were available for this project.
Nevertheless, in order
to give a better indication of what may be expected in the
future patterns of water use, tentative estimates of selected
uses will be made for I960 and 2000.
Although these esti
mates may, in themselves, be important to personnel concerned
with planning for Michigan's future water requirements, it is
believed that the value of these use estimates will not be so
much with the amount of use predicted, but with the identifi
cation of spatial trends.
Water policy established which is
based upon an appreciation of the spatial aspects of water
4
use will be a most; important addition to the planning effort.
Conceptual Framework
A prerequisite for the initiation of any research is
the establishment of parameters within which the research is
to be carried out.
Research which does not contain this
essential element will very likely bewilder the reader with
conflicting terms and with an unclear idea of the purpose and
direction of the work.
For the researcher, too, the early
setting of the framework will better insure that the research
plan is followed and that the purpose and objectives of the
study are completed.
The major concepts which must be established at this
point are the nature of water use and the types of water use
with which this study will be concerned.
The Nature of Water Use
Water may be used but it is never destroyed in the
sense that it is diminished in physical quantity.
The form
of the water may change, causing it to be less available for
use, but the potential amount of water remains unchanged.
This is exemplified by the concept of the hydrologic cycle
whereby the world's potential supply of moisture is constanly moving from the atmosphere to the surface of the earth and
back again.
At some stages in the cycle the moisture is more
available for use than otners, but it is never destroyed.
It is necessary at this point to make a distinction
between water uae and water withdrawal.
Water withdrawal is
a rather straight!orward concept, and it is essentially that
amount of water which is removed from a surface water or
groundwater source.
This is also known as water intake and
says nothing about the use made of water.
Water use is a more difficult term to define.
Water
use is not the same as water consumption yet they are related.
To be consumed water must be used, yet not all water used is
consumed.
To be consumed water must be used in such a way
that it is not readily available for subsequent users, al
though it is physically undiminished.
Consider, for example, a water user who withdraws 100
gallons of water from a water source.
These 100 gallons may
all be used in some industrial process but only 10 gallons may
be consumed.
The remaining 90 gallons are returned to a water
source where they are again available for use.
If a later
user also withdraws 100 gallons and consumes 25, returning 75
gallons, the total water withdrawal will have been 200 gallons
and the total use will have been 200 gallons.
Consumption,
however, will only have been 35 gallons.
There is one further consideration which complicates
the issue of water use and consumption.
Water which is with
drawn, minus that amount which is physically altered in form
(for example, transfer to a vapor state)
so as to make it less
available, may be used and left in such a quality that it is
virtually impossible to make further use of the water.
Thus
6
water may be left unfit for some uses but not be actually
physically altered in form (the original definition of con
sumption) .
Therefore, a further criteria for consumption may
be necessary, one that takes into consideration changes in
water quality that cause it to be unavailable for other uses,
liowever, different types of activities are tolerant of dif
ferent quality water.
Thus water may be consumed, quality
wise, for some types of uses but not for others.
To compli
cate matters even more, water used and left unfit for certain
uses may, with passage downstream or by certain types of water
treatment, again become fit for use (thus being renewed)•
Types of Use
There appear to be two major types of uses, those uses
which internalize water in the production of a tangible good,
and those uses which do not use water internally but use it
as a median for the achievement of objectives.
When water is used for industry, for agriculture, or
for domestic purposes it is used for the production of some
thing, although it is not necessarily consumed, either in
form or in quality, in the process.
Water in the factory may
be used for the production of an automobile, on the farm for
the irrigation of crops, and in the home for the growth of
people.
Other types of uses do not internalize water, nor do
they alter it significantly as a result.
The use of water
for transportation, for recreation, or for aesthetic purposes
7
satisfies these criteria.
Carried to extreme these types of
uses may diminish the quality of the water for any of these
purposes; excessive boating on a water body can decrease the
quality of that particular resource for subsequent boaters.
As a general rule, however, these uses of water do not dimin
ish the water in quantity or quality, and thus impose no costs
on subsequent users.
For the purpose of this paper, the water "use" which
will be considered will be water withdrawals for those uses
which internalize water in its use.
These types of uses will
be restricted to domestic, municipal, and commercial uses,
agricultural use for irrigation and for the production of
livestock and related products, and water use in manufactur
ing and power generation.
The other uses of water, such as for transportation
and for recreation, are recognized as equally important uses
of water.
The rationale for excluding such uses from this re
search is two-fold.
actly the same.
First, the types of problems are not ex
While all water use is related to people,
those uses which use water "externally" are sufficiently dif
ferent from those which use water "internally" to require
special types of research.
The second reason for excluding
these uses from the research is that there are special agen
cies whose main responsibility is directly concerned with
transportation and recreation.
While all uses of water must
necessarily be related if a complete picture is to be
8
achieved,
initial research is probably best done within the
confines of a special research effort.
Objectives
Objective A
The first objective is to identify spatial and tempo
ral trends in the factors which are contributing most to the
changing pattern of water use in Michigan.
The factors which are responsible for broad, spatial
differences in water use patterns are non-marginal.
These
have been identified in a number of different research ef
forts, and are largely the characteristics of the population
and the characteristics of economic activity.
The focus of
this objective will be on the historical development of the
elements of population and economic activity which are most
important in explaining changes in the amount and type of
water use in Michigan over time.
With both population and economic activity the most
critical elements influencing water use will depend to a
great extent on the scale of the study.
If the area selected
for study is extremely small the important elements will be
different than if the study is oriented toward a larger area
or toward a comparison between areas.
Within any one region
the amount of water used will be influenced by more detailed
characteristics, but when the inquiry is the spatial pattern
of use, as between regions, it is likely that detailed
9
information about individual regions will be less important
than more general characteristics for which there are consid
erable variations among regions.
This is true of both population and economic activity.
For example, the elements of population which may be impor
tant when the object of study is a single city may be income,
age structure, housing characteristics, etc.
When the study
is directed toward the difference between cities, however,
the difference in the amount of water used may be determined
so much by sheer differences in numbers of people that the
detailed characteristics of the populations are either ob
scured or become of much less importance.
Similarly, the elements of economic activity important
for determining the amount of water used in a single city may
be a number of detailed characteristics, such as the general
availability of water, or the pricing policy facing each pro
ducer.
When looking at the differences between cities or be
tween regions factors such as these may become less important
than the balance between agricultural or manufacturing acti
vity, or the differences between types of agricultural or
manufacturing activity.
Objective B
The second objective is to make tentative projections
of selected water uses to 1980 and 2000 for Michigan and for
regions thereof.
The Michigan Water Resources Commission has recently
10
completed
(19b8) an inventory of existing water use through
out the state.^ Although the reports have several deficien
cies, they are the most comprehensive treatment of water use
on a state-wide basis that have ever been completed for Mich
igan, and represent a relatively accurate picture of water
use for 1967-1968.
The contribution which this thesis can make to an un
derstanding of water use in Michigan is an estimation of fu
ture water use.
There has been sufficient work done in this
general area to determine that the use of water is going to
increase.
However, there have been no studies which have
attempted to establish a spatial estimate of what this future
use will be.
Even with abundant resources of money and personnel it
is very difficult to make an accurate projection of water use.
And the difficulty increases in direct proportion to the de
gree of detail which is attempted.
rate projections have been made.
For this study four sepa
of these, three are directly
related to the economic activities which are most important in
influencing water use, manufacturing, power generation, and
agriculture.
The third projection which is presented repre
sents a combination of the remaining types of water uses.
For the purpose of this study it will be called domestic use,
*Five separate regional reports (Southeastern Michigan,
Lower Lake Huron, Lower Lake Michigan, Upper Peninsula, and
Northern Lake Michigan and Lake Huron) from February, 1968 to
December, 1968 (Lansingi Michigan Water Resources Commission)•
11
although it includes far more than just household uses.
In
cluded within this category, in addition to household uses,
would be uses by commercial establishments, by institutions,
by governments, and by other potential non-manufacturing
users which might draw water from municipal water systems.
The importance of these projections is not so much
with the actual amount of water estimated to be withdrawn for
use, but with the spatial patterns which develop.
Therefore,
as long as the estimates which are made are reasonable, and
consistency is maintained between regions, the usefulness of
these projections is not impaired.
Objective C
The third objective is to identify spatial patterns of
future water uses which will be of importance to water plan
ning efforts in the state.
The focus of this objective is on a regionalization of
the 1980 and 2000 water use estimates.
This will be concerned
not only with the total amount of water use, but with changes
in the type of uses which can be expected among the several
regions.
It is possible for the state as a whole to exhibit
a relatively continuous increase in water use, but for there
to be quite a variation among the regions.
Therefore, it is
necessary that regional changes be identified if the state
agencies concerned with water planning are to meet their re
sponsibilities.
12
Scope
In the consideration of a research effort the scope of
the study is of great importance.
With a given amount of re
sources which can be expended for a study, the magnitude of
the problem, either in geographic size or in complexity, will
significantly influence the purpose of the research, the ob
jectives, and the methods used.
The geographical area covered by this research is
state-wide.
The examination of such a large area will neces
sitate a more general approach than if the focus was on one
county or even on a group of counties.
Both the general the detailed approaches have merit.
The detailed study of a small area will enable the identifi
cation of important variables, and the magnitude of their in
fluence, to be more accurate, and as a result any quantita
tive estimates of water use at a future date will undoubt
edly be more precise.
The accuracy of such a study, however,
does not indicate the nature of the situation for the larger
area, a region or a state.
There are important spatial con
siderations that can be understood only when the complete
situation can be realized.
In fact, a thorough understand
ing of water use relationships of the small area can often
be seen only within the context of the larger area.
The manner in which data is presented, the size of the
enumeration tract, is important with regards to the unity of
the data.
By reporting data in small tracts it is possible
13
to reorganize it to conform to any number of regional areas.
When the data is presented on a large area basis the flexi
bility of the data becomes less.
And, data for small areas
is often hidden within the larger ones.
For this research the preliminary data and resulting
water use estimates have been organized according to a re
gionalization of counties which was recommended by the Mich
igan Water Resources Commission
(see the map on page 14).
Regions I, II, III, IV, and V represent the groupings of coun
ties which most closely correspond to major watersheds of the
state.
Regions II and III have each been divided into parts
A and B because of the relatively large size of regions II
and III.
In estimating future conditions of water use the time
framework is always a significant consideration.
Some bal
ance must be achieved between the extreme long range pro
jections which are likely to be greatly in error because of
the long span
of time, and
very short range projections which,
although more
precise, are
less useful because of the re
stricted time
period.
For
this study 1980 and 2000 were cho
sen as target
dates.
These were selected completely arbi
trarily, and represent a compromise between short and long
range projections.
Introduction to Methodology
The general research plan for the study is to iden
tify the principle factors responsible for water use, to
ranij
tawaw m
IftM
LVCf
iat-aaa
KMMUWr
■tf
ioicoa* t m x m m
M M OL*
CLM I
H (M T « I M I I l U l m i n ,!
U
aa*Ti»T 1 1***11**
IONM
(HIITfai1
cat
•«***■* — - — - " — T
c«at
mu
Figure 1.
Planning Regions
li>
project tlie magnitude of these factors to 1980 and 2000, and
from these to estimate selected future water uses.
A more
detailed description of the methods used in making these
estimates will be found in later chapters.
However, there is
one point which much be discussed at this tine since it pro
vides a basis for the rest of the study.
This most important
topic involves the nature of forecasts and projections.
Forecasts and Projections
In its most elementary form a forecast is nothing
more than a statement about the condition in which something
is expected to be at some future time.
Such forecasts can be
made on a continuum which ranges from pure guesses to those
which are based on sophisticated models which rely upon per
fect knowledge of future conditions of those factors which
are responsible for the thing which is being examined.
At one end of the continuum, the guess, forecasts are
generally rejected in favor of a method which has more pre
dictive power.
At the other extreme, the detailed knowledge
which is needed if the relevant factors associated with the
object under investigation is seldom present.
Consequently,
forecasts are generally based upon a compromise between the
"guess'* and perfect knowledge.
Most "compromise" forecasts rely to a great extent
upon the observance of the past history of the object of the
forecast and of the factors which are responsible for it.
These compromise forecasts which utilize data on past
lb
conditions arc commonly called projections.
Kuzncts defines
projections as "...statements about the future that claim to
be derived from empirically tested propositions concerning
the past."*
The Mature of Projections
2
Statements about the future which are based upon the
past appear at first to be simple logic; since the future can
never be known by the observation of empirical evidence it
would seem realistic to assume that the events of the future
would resemble those of the immediate past.
Yet there are
two criteria which must be met before such a statement can be
valid.
First, the past and the future must be interrelated
in some logical manner.
And second, the past events must
have some semblance of order about it.
The first criteria, that of relationship between the
past and the future,
is an obvious necessity.
Unless this
relationship is present any attempt to assess the future
would be in vain.
Zt would be futile to expect that the a-
mount of water used in Michigan in 1980 would be in any way
related to the trends in the ratio of male-female births.
However, it would be quite realistic to expect that changes
*Simon huznets, "Concepts and Assumptions in Long-Term
Projections of National Product," Studies in Income and
Wealth, Vol. 16 (Washington; National bureau of economic Researcli, 1954) , p. 9.
2
This section is based upon the points developed in
Kuznets, Ibid.
17
in population levels and economic activity might have an
effect.
The second criteria which must be met before a projec
tion can be made is that there must be some systematic order
ing of past events.
If the relevant factors have occurred
randomly in the past there is no reason to expect that they
will occur in any other way in the future.
And it would be
impossible to project a random past in any manner except ran
domness, a useless effort if the purpose is to determine what
the future may hold.
Therefore, unless past conditions are
arranged in some meaningful pattern it will be useless to
make projections based upon them.
exists in past data can be tested.
The extent to which order
Therefore, it is possible
to proceed with the realization of the probable validity of
the projection.
The question of prediction versus forecast is an im
portant one.
The principle distinction is related to the
existence of a causal relationship between the object of
study and one or more associated variables.
A prediction im
plies that one or more factors cause another factor to assume
a certain condition.
It assumes that specific magnitudes of
A, B, and C will cause 0 to exist in a certain form.
A forecast does not necessarily involve the function
of causality, yet may equally well approximate the future.
physicist and an illiterate may both throw a stone into the
air and state that it will fall to earth.
The physicist, to
A
18
the extent that he applies the concept of gravity,
ing that the stone will fall.
is predict
The illiterate may have no un
derstanding of physics, but from past experience believes that
the stone will behave as previous stones have behaved.
then is making a forecast and not a prediction.
He
In a socio
logical example, there is a strong correlation in Michigan
between a county's total population and its population den
sity.
Thus one may make a statement about future population
levels based upon trends in population density.
This does
not, however, involve a complete explanation since the matter
of the area of the county has not been considered.
The extent to which either prediction or estimation is
satisfactory will depend to a large extent on the purpose of
the projection.
If the object is to develop an understanding
of why and how the causal variables operate one must work
toward prediction.
On the otherhand, if the intent is merely
to determine the magnitude of a certain condition, an expla
nation may not be necessary.
If it was possible each year to
accurately forecast the first snowfall by the date of the
last snowfall in the previous year this would be satisfactory
if the object was just to determine the date of the first
snowfall and not provide an explanation of why.
Such a spurious relationship is seldom, if ever, found.
Most estimations as opposed to predictions are based upon re
lations which are obvious but are not well enough understood
to be fully explained, or for which the lack of data is a
iy
restricting factor.
usefulness.
This, however, does not prevent their
If the object is to estimate how much timber
will be cut in the nation next year, and this can be done by
projecting the quantity cut in previous years, it is not nec
essary to know what the demand for housing, furniture, etc.
will be.
Often the addition of successive factors will increase
the ability to forecast the future with accuracy.
To the ex
tent that data is available that permits this, this should be
the goal.
When data are lacking forecasts must be made with
data at hand.
The present research must, of necessity, be more of a
forecast than of a prediction.
The detailed, historical data
which is required for the more complex projection methods can
not be obtained and, therefore, these methods must be passed
over in favor of methods which can utilize the data which is
available.
The forecasts of water use which are presented
are based, however, on sound relationships which have been
tested in other research which has had access to the type and
quality of input data which is necessary to utilize the more
sophisticated methods.
Assumptions
Forecasts of the future must necessarily involve as
sumptions about the future.
There are certain basic condi
tions which must be stabalized if any projections can be
made.
It must be assumed that water availability will re
main the same, and will play essentially the same role that
it has played in the immediate past.
There have been many
schemes to increase the water in certain areas of the United
States.
The North American Water and Power Alliance and
weather modification are examples.
In Michigan, there has
been discussion of a bake Michigan to Lake Huron canal which
would provide increased water for some areas of the state and
undoubtedly influence the spatial pattern of water use.
In
addition, there are a few potential reservoir sites which
could possibly be utilized to increase the available water.
It is assumed that these developments will not take place to
any appreciable extent prior to 2000.
Further, it is assumed
that there will be no major technological "breakthrough"
which substantially lowers the per capita demand for water.
Equally restrictive, it must be assumed that there will be no
major droughts in 1980 or 2000.
The "normal” available water
must be considered as existing in the target years.
A second assumption is that there will be no major war
or large inflationary or deflationary periods in the economy
in the years 1980 and 2000.
Such wars or economic changes may
very well occur prior to these target years, but it must be
assumed that by 1980 and 2000 their effect will have been
ameliorated so that it is no longer significant.
There are several other assumptions which must be made
with regards to each of the domestic, industrial, and
21
agricultural projections.
spective chapters.
These will be discussed in the re
CHAPTER II
FACTORS CONTRIBUTING TO DOMESTIC,
MUNICIPAL, AND COMMERCIAL
WATER WITHDRAWALS
Introduction
The category of water use labeled "domestic, municipal,
and commercial" is a very important one in that it is one of
the necessary elements of a modern urban system.
This cate
gory of water use includes three principle types of with
drawals:
dences,
(1) domestic - water withdrawals by private resi
(2) municipal - withdrawals for use in schools, hos
pitals, and other urban services,
(3) commercial - laundries,
restaurants, car washes, etc.
The purpose of this chapter will be to examine the
factors behind differences in domestic, municipal, and com
mercial withdrawal rates among different parts of an area.
No attempt will be made to relate these to the Michigan sit
uation in particular because of the inability to handle such
detail on a state-wide scale, but to just discuss in general
the findings of related research.
Composition of Withdrawals
Per capita domestic, municipal, and commercial water
22
23
withdrawal rates combined have increased considerably in the
past and are expected to increase further in future years.
The Mater Resources Council has estimated that, nationally,
the daily per capita withdrawal from municipal systems for
all non-industrial uses will rise from 121 gallons in 1965
to 123 gallons in 1980 and 125 gallons in the year 2000.^
Of the total withdrawals, the amount required for com
mercial uses is expected to remain constant at 28 gallons per
capita per day.
The amount required for public or municipal
purposes is expected to decrease from 20 gallons per capita
per day in 1965 to 18 in 1980 and 16 in the year 2000.
Thus
the total category increase in per capita withdrawals is ex
pected to result from increased domestic use.
For the total
Great bakes region this is expected to increase from 73 gal
lons per capita per day in 1965 to 77 gallons in 1980 and 81
in the year 2000.2
Elements of domestic Withdrawal Increases
The majority of the research which has been done on
residential use points to three factors as being primarily
responsible for the increased per capita use ratest
(1) the
development of modern home appliances which utilize large amounts of water,
(2) increased incomes which allow the
^water Resources Council, The Nation's Water Resources
(Washington: U.S. Government Printing Office, 1968), p. 4-1-2.
2 Ibid.
24
purchase of the aforementioned appliances,
(3) the movement
of larger numbers of people to the suburbs.
Modern Appliances
There have four appliances which have been respon
sible for increasing the per capita withdrawal rates for d o
mestic purposes.
These are clothes washers, dishwashers, ga r
bage disposals, and air conditioners.
Depending on the type
of washing machine the mean rate of water use is between two
and seven gallons per minute.
The normal load requires from
36 to 50 gallons.^ The higher rates are found in the automatic
washers, and the trend toward the use of automatic washers
suggests that increasing amounts of water will be required for
this purpose.
dishwashers.
The same general pattern is found in automatic
An Ohio study found that the amount of water
used for a single washing in an automatic washer was approxi
mately 40 per cent greater than when the washing was done by
hand.
2
This survey was taken in 1956 and it was suggested
that modern washers probably use more water than did the old
er models.
A study made in St. Louis indicated that homes
with air conditioners used between four and one-half and sev
en times as much water in the summer as in the winter.
In
*J.G. C a m s , "Service Lines and Meter Requirements of
Domestic Water Connected Devices," Journal of the American
Water Works Association, Vol. 58, NoT T5 (October, 1966),
pTT25T6-----------------2
Rudolfo Silva, "Land Uses and Water Consumption Re
quirements," Public Works, Vol. 90, Ko. 4 (April, 1959),
p. 123.
25
homes without air conditioning summer use was less than one
and one-half times the winter rate.^
Increased Incomes
Increased incomes allow the populus to enjoy the
fruits of technology, generally resulting in increased per
capita withdrawal rates.
In a study of per capita withdrawal
rates in Illinois, a strong relationship was found between
family income and per capita use rates.
Families with high
incomes used, on the average, five times as much water per
person as did families with low incomes.
In addition to home
appliances, higher family incomes have allowed the construc
tion of homes with more complete sanitary facilities, includ
ing multiple bathrooms.
Even home swimming pools are be
coming commonplace.
A more detailed study of this
Rhode Island.
type
of use was made in
The purpose of the study was to see if the
amount of water used in households was related to social sta
tus.
The significance of a number of variables proporting to
represent social status were tested.
Only three of these,
however, were found to be significantly correlated with the
amount of household water use.
These variables were
^Kenneth S. Watson, "Water Requirements of Dishwashers
and Food Waste Disposers," Journal of the American Water
Works Association, Vol. 55, No. 5 Ulay, IH-J), pp. 5^5-^59.
2
Brent O. Larson and li.E. Hudson, "Residential Water
Use and Family Income," Journal of the American Water Works
Association, Vol. 43, No. 7 (July, 1951), p. 6lo.
2t>
household income, house value
(which is directly related to
income), and household size.^
Other social variables have also been found to be re
lated to residential water use.
In a study in Kankakee,
Il
linois, assessed valuation of the house, family income, edu
cation, and occupation were statistically correlated with
water use per household.
2
It is interesting to note that all
of these variables are interrelated to a certain extent.
Suburban Living
The trend toward suburban living has accelerated the
per capita water withdrawal rates for residential use.
This
is overwhelmingly the result of larger lawns and the demand
for water to irrigate them.
In fact, the increase in the de
mand for lawn sprinkling has changed the design of water sys
tems in surburban areas.
With the pre-World War II condi
tions most water systems were designed with the controlling
factor being the quantity of water needed in case of a bad
fire.
In many suburban areas it has become necessary to make
the potential lawn sprinkling requirements the controlling
factor.
Demands for any fire load are almost always covered
^Irving A. Spaulding, Household Water Use and Social
Status (Kingstons Agricultural Experiment Station, University
of Rhode Island, 1967), pp. 7-25.
2
Dorthy F. Dunn and Thurston E. Larson, wRelationships
of Domestic Water Use to Assessed Valuation, with Selected
Demographic and Socio-Economic Variables," Journal of the
American Water Works Association, Vol. 55, No. 3 (April,
1963), pp. 441-449.
27
under the sprinkling capacity.^
The tremendous growth rate of the suburbs almost in
sures that this trend will continue.
Between 1950 and 1955,
98 per cent of the population increase was in metropolitan
areas, and the suburban growth rate was seven times as great
2
as that of the central cities.
Although these are national
data this trend is also evidenced in Michigan
(see Table 5)•
The amount of water required for watering lawns will
not be uniform throughout a large area because of differences
in climate, largely temperature and precipitation.
Even in
Michigan there is sufficient variety in climatic patterns to
cause a potential difference in sprinkling requirements in
different parts of the state.
Figures 2 and 3 indicate the
situation for the bower Peninsula.
The map of potential
evapotranspiration indicates how much moisture would be evap
orated and transpired if there was sufficient moisture avail
able.
The values for potential evapotranspiration generally
diminish toward the northern part of the state.
The map of
normal moisture deficit shows the difference between the
maximum possible evapotranspiration
and the actual evapotranspiration
(given adequate moisture)
(which may be less than the
maximum possible because of inadequate soil moisture).
The
greater the deficit the more critical is the lack of moisture.
^Angus D. Henderson, " The Lawn Sprinkling L o ad,”
Journal of the American Water Works Association, Vol. 48,
Mo. J ('
A pril, 1955), pp. 361-362.---------------2
p. 121.
Silva,
"Land Uses and Water Consumption Requirements,"
Figure 2.
Normal Potential Evapotranspiration
a Id
aA. Stephen Messenger, "The Water Balance of the Lower
ninsula of Michigan," Papers of the Michigan Academy of
ience, Arts, and Letters, V o l . XLVII (1962), p~! 586.
^Millimeters
AII4U
M N III
9»CI0lA I ct*it
1 A ItI MAC
OCtANA .««»AT*0 'm c COITA i
U»«f.‘
HINT
r iOAIA
.CLIATOH
1 CALHOUN
r
ilMiti r
Figure 3.
Normal Deficit
A. Stephen Messenger, "The Water Balance of the Lower
Peninsula of Michigan," Papers of the Michigan Academy of
Science, Arts, and Letters, Vol. XLVII (1962), p. 387.
^Mi 11 ime 16r s
30
It can be seen that there are definite spatial differences,
at least in the bower Peninsula,
in the soil moisture deficit.
Pricing Policies
Although the withdrawal rates for domestic, municipal,
and commercial uses have risen in the past and are expected
to rise considerably in the future, the rate of increase is
probably lower than it would have been without the operation
of a very basic economic principal.
This principal is the
price elasticity of demand for water and the tendency for
less water to be used when the cost of that water increases
in certain kinds of situations.
a
number of studies have shown empirically that this
process actually works in the area of water utilization.
In
a study of municipal water systems throughout the nation a
direct relationship was found between water use in urban
areas and the price of water.
2
When SO per cent of the water
was metered there was no firm trend in the reduction of use
rates.
In the 50 to 95 per cent metered category there was
a sharp decrease in the per capita use rates
six per cent reduction in per capita u s e ) .
(approximately a
This trend was
^For a most complete discussion of the water balance
concept see C.W. Thornthwaite and J.R. Mather, The Water Balance (Centerton. New Jersey: brexel Institute ot technology,
l 9 in>)
.
2
Ralph Porges, "Factors Influencing Per Capita Water
Consumption," Water and Sewage Works, Vol. 104, No. 5 (May,
1957), pp. 199-264.
31
accentuated in the over 99 per cent metered category.
On the
basis o£ these data the savings in water would be about 25
per cent when going from an unmetered system to 100 per cent
metering.
In an analysis of the water use statistics of 441
cities as published by the American Water Works Association a
close correlation was found between the price of water and
the per capita use rate.
A summary of their findings is pre
sented in Table 2.
Table 2
Price Elasticity of Residential Water Deroanda
Cost of Water
(c/1000 ga l . )
Per Capita Use
(gal. per day)
.70
.60
.50
.40
.30
.20
.10
137
143
154
178
205
246
296
aH. Seidel and E. Baumann, *A Survey of Operating Data
on Water Works in 1955," Journal of the American Water Works
Association, Vol. 59, No. 5 (May, 1967), p. SS!j .
Figure 4 shows the demand-price relationship for water
use for domestic purposes in western cities.
This data again
substantiates this economic relationship.
Despite these apparently strong relationships overall,
it is often difficult to identify them when dealing with any
32
70
50
o
40
>
'
30
*>\
H*44 l;Nli:
*V* W *
44f.fcA£
iwm™™
iiiil
;.;,V i
10,000
and
over
Figure 7.
Effective buying
Per Household, 19GS
Income
a "iyG9 Survey of buying Power," Sales Management,
102, No. 12 (July 10, 1969), pp. Ubb-D9!7T
~~~
Vol.
48
more water per person than do rural people.
a number of reasons.
This is true for
Urban areas usually have higher incomes
which give them greater access to objects which use water.
A
greater proportion of people who live in urban areas have
lawns which are watered periodically.
work in cities
of water.
Generally maintenance
(street washing, etc.) consumes large amounts
In 196 5 it was estimated that per capita domestic,
municipal, and commercial water use in urban areas was over
three times as great as for rural a r e a s .^
The population of Michigan is rapidly becoming urban
ized.
The number of residents who were classified as urban
and rural each decade since 1850 is illustrated in Table 5.
It can be seen that until the early 1900's the number of r u
ral residents have been greater than the urban residents.
Since 1920, however,
larger.
the urban population has been constantly
Since 1840 the urban population as a per cent of the
total has increased steadily except for the 1930 to 1940 p e
riod .
Population Projections
The projection of the population of an area is fraught
with uncertainty.
One is almost sure that any projections
which are made will contain a certain amount of error;
ide
ally this error will be kept to a minimum.
The hazards of making estimates of future populations
^Water Resources Council, The Nation's Water Resources,
P*
4-1-2.
Table 5
Michigan Urban and Rural Population
Year
1960
1950
1940
1930
1920
1910
1900
1890
1880
1870
1860
1850
Population^
Urban
Rural
5,739 .1
4,503.1
3,454.9
3,302.1
2,241.6
1,327.0
952.3
730.3
405.4
238.0
99.7
29.0
2,084 .1
1,868.7
1,801.2
1,540.3
1,426.9
1,483.1
1,468.7
1,363.6
1,231.5
946.1
649.4
368.6
Per Cent
Urban
Rural
73.4
70.7
65.7
68 .2
61.1
47.2
39.3
34.9
24.8
20.1
13.3
7.3
26.6
29.3
34.3
31.8
38.9
52.8
60.7
65.1
75.2
79.9
86.7
92.7
aCompiled from U.S. Bureau of the Census, U.S. Census
of Population: 1850-1960. Characteristics of the Population,
Michigan (Washington: DTs. Government Printing O f f i c e ) .
^Thousands
is best exemplified by examining projections which have been
made by different sources for the same area.
It is safe to
say that all of them strived to provide the forecast with the
greatest amount of accuracy.
However, it is apparent that
there are considerable differences among the several projec
tions which have been made for Michigan.
Some of the more
common of these are listed in Appendix Table 3.
The population projections adopted for use in this
study are based upon the work done by Dr. David Goldberg,
Population Studies Center, University of Michigan
(see foot
note, Appendix Table 3) with modifications made by the State
Resource Planning Division, Office of Economic Expansion,
Michigan Department of Commerce.1 The work of the Resource
Planning Division was to extend these projections to the year
2000
.
The results of this projection for each of the seven
planning regions are presented in Table 4.
The most impor
tant information revealed by this data is the relative pop
ulation changes among regions.
Each of the seven regions ex
cept one is expected to experience an absolute increase in
population between 1960 and 1980, and between 1980 and 2000.
The big difference is in the share of the state total.
Five
of the seven regions are expected to have a declining share
of the state's population.
Only Regions 1 and I1IA are ex
pected to receive a greater proportion of the total state
population.
Population estimates for 1980 and 2000 for the
top 15 counties in 1960 are presented in Appendix Table 2.
Donald E. bailey. Preliminary Population Projections
for Small Areas in Michigan (Lansingi State Resource Planning
bivision. Office of Economic Expansion, Michigan Department
of Commerce, 19b6).
CHAPTER IV
DETERMINATION OF DOMESTIC, MUNICIPAL,
AND COMMERCIAL WATER WITHDRAWALS
Introduction
The amount of water withdrawn for domestic, municipal,
and commercial purposes is a function of the per capita with*
drawal rates for each activity and the level of each of these
activities in the region.
It is possible to develop a rela
tively accurate predictive model of this type of water use by
examining separately the various components of withdrawal and
then applying withdrawal rate factors to forecasted future
levels of each activity.
However, the precise data which
would be required for this approach is staggering and is com
pletely beyond the resources of this thesis.^
For this research it was necessary to use a modified
approach in estimating future levels of withdrawal.
proach involved two processes:
This ap
(1) the development of a water
withdrawal rate per level of activity represented by a combi
nation of domestic, municipal, and commercial uses,
(2) the
^The most detailed model encountered for forecasting
municipal water needs was developed by Hittman Associates,
Inc., "Main I" A System of Computerized Models for Calculating and Evaluating Municipal Water Requirements (Columbia,
Maryland, 19SB). ---- ---------------- 3---------51
application of ttiis withdrawal rate to the expected future
level of activity,
in this case population totals.
This type of an approach necessitates two basic as
sumptions.
First, for the category domestic, municipal, and
commercial use there is a characteristic rate of water with
drawal per level of activity.
For this category of use the
level of activity is measured by the number of people who are
potential water users.
Consequently,
the amount of water
withdrawal is measured on a per capita basis.
Realizing that there will be considerable variation in
water withdrawal rates among individuals and even among dif
ferent parts of the state, the second assumption is that the
rate of per capita water withdrawal multiplied times the to
tal population, will reasonably approximate the total water
withdrawals.
Possible Approaches
The technique which one uses is often a function of
two factors, the resources with which one has to work and the
purpose of the research.
The combination of these two will
often determine the approach which is used.
After a review
of the techniques used in other states in the estimation of
their water needs two possible approaches were considered for
use in this study.
These were modeling and the application
of per capita withdrawal rates to projected levels of popu
lation.
Modeling
The first possible approach which was considered was
the development of a model which would incorporate as many e x
planatory variables as possible in order to predict the mag
nitude of domestic, municipal, and commercial water with
drawals.
V4
changing the values of the inputs to more closely approxi
mate the real world the model can be "tuned" to a greater de
gree of accuracy.
The third advantage follows closely the
second.
This involves the amount of detail which can be pre
sented.
This is entirely up to the researcher, but in the
more complex models a great deal of detail can be achieved.
The major disadvantage involves the amount of data
wnich arc required in using this approach.
Ideally every
force which operates to influence water withdrawals should
be incorporated into the model.
Obviously this is impossible.
The researcher must usually be satisfied with a more general
representation of the real situation.
And in the use of this
approach for estimating future levels of water withdrawals
one is also faced with the difficulty of forecasting the fu
ture levels of those variables which will be used in fore
casting the end product.
Thus the estimate of water withdraw
al, itself a very tenuous undertaking, is based upon variab
les which are themselves estimates and subject to all the
weaknesses of estimates.
Per Capita Withdrawal Rates
The second approach is a less sophisticated one, and
involves the application of domestic, municipal, and commer
cial water withdrawal rates to estimates of population at some
future time.
This approach does not attempt to assess the in
fluence of individual variables on withdrawal rates, but con
siders them to be inherent in the water use coefficients
(or
the per capita withdrawal rates).
It is possible, however,
to adjust for changing conditions by assuming an increase
decrease)
(or
in per capita withdrawal rates to compensate for
conditions which may change over time.
To the extent that
this compensatory change is accurate, future levels of with
drawal can be relatively accurate.
The chief advantage of this approach is in the data
requirements.
The only type of information necessary is that
pertaining to withdrawal rates and estimates of future pop
ulation levels.
Thus, for a study which is being conducted
on a limited budget, or one in which results are needed
quickly, this approach is an appropriate one to use.
The main disadvantage with this approach is that it
docs not identify what factors are instrumental in causing
changes in withdrawal rates.
If the purpose of the study is
more to estimate future levels of withdrawal rather than to
completely explain the system then this is not a major dis
advantage.
However,
to the extent that it reduces the re
liability of the estimate it is a limiting factor.
data Availability and Deficiencies
The data requirements for each of the two approaches
are different, although certain elements are necessary for
each.
The availability of data of the type required may be
the most important factor determining which approach will be
used.
56
Per Capita Withdrawal Kates
basic to both approaches is an accurate determination
of per capita water withdrawal rates.
Since the object of in
quiry is domestic, municipal, and commercial water withdraw
als, per capita rates used must represent only this type of
use and not include any industrial use.
The only data readily available for non-industrial
withdrawals for Michigan in particular is from the Michigan
Department of Public Uealth.
These water withdrawal figures
apply only to water pumped by municipal systems and do not
include information on per capita withdrawals by those who
have their own wells or in some other manner acquire water
from non-municipal sources.
The Department of Public Health is responsible for the
municipal water systems in the state, and periodically col
lects data on the amount of water pumped by these systems.
The most recent data available are for 1965.
At that time
approximately 76 per cent of the population in Michigan re
ceived water from municipal systems.
The Department of Pub
lic Health has separated the total water pumped into two
categories, industrial and non-industrial.
The category
"non-industrial" is a very broad one and encompasses a varie
ty of different uses.
Included within this category would be
the domestic, municipal, and commercial uses which are the
focus of this chapter.
The principle disadvantage with the data from the
57
D e p a r t m e n t of Public Health involves the reliability.
These
Data are reported by individual municipal systems, and the
quality of the reports vary considerably among systems.
In
the larger systems the accuracy of the data are thought to be
quite high.
In the smaller systems, however, the records are
generally less reliable.^ Data on per capita non-industrial
water withdrawals can vary greatly if a water system fails to
accurately differentiate between industrial and non-industrial
uses.
With inadequate staffs and without the more refined
methods of data processing the validity of the data from the
smaller municipal systems is questionable.
Since the data
reported by the Department of Public Health are for a county
it includes both large and small systems.
The records for
individual systems are hidden within the county figures and
can not be extracted.
In addition to the quality of the records there is al
most always some water pumpage which is unaccounted for.
This is generally a result of mechanical malfunctions some
where in the water system or, again, from inadequate record
keeping.
The following clipping from the hansing
(Michigan)
State Journal indicates a specific example of the problem
mentioned above.
^Telephone conversation with William A. Kelly, Michi
gan Department of Public Health, June, 1969.
W h e r e ' s Our Wa t e r ?
iiissing: 20 million gallons of water.
dian Township has lost track of it.
Meri
When the annual report of the public works
department was submitted at the township
board meeting this week, one of the trustees
questioned the large amount of water NlostM
in an area of the township which buys its
supply from bast bansing.
Figures showed 117,35b,QUO gallons were pur
chased from the city and 9b,tt7 5,00U gallons
sold to customers, a difference of 20,480,000
or 17.45 per cent.
a normal water
loss should not exceed 10 mil
lion gallons or about eight to ten per cent,
said Gaylord Smith, public works superinten
dent.
lie said three known water leaks in the area
last year were not large enough to account
for 20 million gallons even when combined
with normal loss from fire department and
road commission use.
MThe loss could possibly be in the differ
ence of when bast bansing read its meters
and when we read ours," Smith said.
MIt is
possible that we will show an excess instead
of a loss the next time the water meters are
read.”
It is possible, he said, that there could be
an undetected underground leak.
"The only thing we can do is to keep on
checking until the difference is accounted
for," Smith said, "right now we just don't
know what happened. "
An evaluation of the Department of Public Health data
reveals a very wide range in the per capita withdrawal rates.
The range of water withdrawal coefficients can be seen in
Appendix Table 6.
The median coefficient was approximately 36,000 gallons
per capita per year, while the mode coefficient was in the
35,000 to 36,000 block.
analysis
of the total counties used in the
(75),* 76 per cent had a range of between 20,000 gal
lons and 50,000 gallons per capita per year.
There was, however, a considerable range overall, from
a low of 7,058 gallons per capita per year to 103,636 gallons
per capita per year.
A certain amount of difference among
counties can be expected because of differences in the fac
tors responsible for water withdrawals, water losses, etc.
however,
it was felt that a range of 7,058 to 103,636 was
probably the result of inaccuracies in the reporting system.
An alternative source of data on water withdrawal rates
for domestic, municipal, and commercial uses is revealed by
the recent report by the Water Resources Council
1, page 23).
(see footnote
In this report data were presented on national
non-industrial withdrawals.
pendix Table 4.
This data is summarized in A p
Water withdrawn from public systems is com
posed of domestic uses
and commercial uses.
ral domestic systems.
(household), public uses
(municipal),
Individual systems are essentially r u
while these data are for the nation
as a whole it is possible to modify them so that they repre
sent more accurately the Michigan situation.
*The remaining eight counties either had data which was
missing (two counties) or the data of more than one county
were grouped together (six counties).
Independent Variables
There are essentially two method of acquiring data on
explanatory variables to incorporate into a predictive model
for water withdrawals.
The first of these is through the use
of an instrument such as a questionnaire or personal inter**
view.
Such an approach could involve simultaneously collect
ing information about the dependent variable
water withdrawal)
(the amount of
and the independent variables
(the explana
tory factors).
The main disadvantage of this approach for an area the
size of Michigan is the cost.
With approximately 8,645,200
people in the state the cost of sampling a large enough seg
ment of the population to be meaningful would be very pro
hibitive and beyond the resources of this study which is in
vestigating industrial and agricultural withdrawals in addi
tion to withdrawals for d o mestic, municipal, and commercial
purposes.
The second method is to utilize existing data.
The
type of information needed is often available from various
state agencies as well as from published census material.
While much less costly to acquire, these types of data suffer
from two weaknesses.
First, the exact type of data needed
are not always the type available.
It is often possible to
"make do" with inappropriate data, but the research generally
suffers.
The second weakness is that the time periods in
which these data are collected do not always coincide with
each other, nor do they always coincide with the data on per
capita water withdrawal rates.
Again, it is possible to
utilize data from two different periods but the results will
not accurately represent the real situation in any year.
And
the greater the period of time between the data collection
dates the greater will be the discrepencies in the results.
Population Estimates
Accurate estimates of future population levels are
necessary for forecasting domestic, municipal, and commer
cial water withdrawals.
This aspect of data requirements has
been covered in Chapter 111.
Examination of Alternative Approaches
In order to achieve the best possible result both the
modeling and per capita withdrawal rate approaches were in
vestigated.
Method Ii
The Model
The method used to estimate annual per capita domestic,
municipal, and commercial water withdrawal rates by relating
them to several explanatory variables was the least squares
regression technique.
The function of regression is to eval
uate changes in a dependent variable
(in this case annual per
capita withdrawal rates) with changes in the independent
(ex
planatory) variables.
The dependent variable, annual per capita domestic,
municipal, and commercial water withdrawal was determined by
62
utilizing the 1965 Department of Public Health data on munic
ipal water systems.
The variable was generated by subtract
ing the industrial withdrawal from the total withdrawal, and
by dividing this figure by the number of people served by mu
nicipal systems
(also available for 1965 from the Department
of Public health).
This information was available for the 75
counties with complete records.
A listing of the independent variables used in the re
gression analysis, the source of the data, and the simple
correlation coefficient of each of the dependent variables is
presented in Appendix Table 5.
Correlation coefficients re
fer to the efficiency of the regression equation in estimat
ing the dependent variable and are determined by the devia
tion of the actual values of the dependent variable from the
predicted value of the dependent variable which is represent
ed by the regression line.
When only one independent variable is considered the
results are a simple correlation and a simple regression.
When more than one variable are used the results are a m ul
tiple correlation and a multiple regression.
When all of the
independent variables are included the multiple correlation
coefficient is .64 which means that approximately 41 per cent
of the variation in per capita water withdrawal was explained
by the independent variables.
There are a number of factors which could account for
this relatively low explanatory power of the model.
It is
felt, however,
that the most significant reason is inadequa
cies in the available data.
discussed.
Many of these have already been
However, there are other data characteristics
which may be important.
The aggregate nature of the data, both dependent and
independent variables, is undoubtedly partially responsible.
When aggregated, as on the county basis, the data represent
merely a mean figure for that variable in that county and do
not reveal information about individual situations.
If water
withdrawal information on individual households could be com
pared with corresponding data on size of family, income, home
value, etc. it is very possible that a greater amount of ex
planation could be achieved than if the mean figures for
counties are used.
One significant factor could not be included in
the analysis.
water.
These are data on the pricing policies of
The extent of metering and the water rates differ
among water systems.
Some systems are not metered and charge
a flat rate regardless of the water used.
Even for the sys
tems which do meter their water, the actual charges vary con
siderably, thus providing for differences in the efficiency
of water use.
The inclusion of this important variable would
have an important influence on the results of the model.
Method lit Per Capita Withdrawal Rates
As a result of the relatively low amount of explana
tion afforded by the model it was decided to try the second
64
technique, that of applying the annual per capita withdrawal
figures for non-industrial use, which were supplied by the
Michigan Department of Public Health, to the projected popu
lation for 1980 and 2000.
It was not possible to apply these data directly to
estimates of future population, however.
Before this could
be done a number of intermediate steps were necessarys
(1) the establislunent of a more reliable per capita with
drawal rate from municipal systems for each county in 1965,
(2) estimating the corresponding withdrawal rates for those
not served from municipal systems,
(3) estimating the per
cent of population served by municipal systems in 1980 and
2000,
(4) accounting for increases in per capita withdrawal
rates to 1980 and 2000, from both municipal sources and
non-municipal sources.
Per Capita Withdrawal Rates. Municipal Systems
Mention was made earlier of the wide range of per cap
ita withdrawal rates reported by the separate municipal water
systems to the Department of Public Health.
A discussion with
officials of that department resulted in the decision not to
use coefficients that were on the extreme ends, high or low,
of the range.
It was decided to eliminate those coefficients
which were more than twice as large as the median coefficient
(36,000 gallons per capita per y e a r ) , and to assign to these
counties the median coefficient.
nine counties.
This adjustment applied to
6L>
In addition to the counties to which the above dis
cussion applies there were two for which the Department of
Public Health did not have information.
For these counties
the median coefficient was used.
There were two groups of counties
(three counties
each) for which the data were reported as an aggregate.
The
coefficient which applied to the group was considered to be
representative of each county in the group.
A listing of the counties, their corresponding coef
ficients, and the derivation of the coefficient are presented
in Appendix Table 6.
Because of the large discrepancies among county with
drawal coefficients a test on the sensitivity of factors oth
er than numbers of people was run.
A correlation was made
using the total amount of water withdrawn as the dependent
variable and the number of people served as the independent
variable.
It was expected that because of the large dif
ferences in the per capita withdrawal rates there would be a
very low correlation.
The opposite was true, however.
The
correlation in this case was only slightly less than 1.00, a
perfect correlation.
A scatter diagram of this relationship
is presented in Figure 8.
This means that the number of
people served accounted for over 99 per cent of the water
withdrawn from municipal systems for non-industrial uses.
This merely indicates that while changes in the per capita
withdrawal rates are important in determining total
2500
-
1500 -
1000
CJ\
t>), p. s-i.
88
Table 8
-
L
Employment Trends in Michigan
Industry
Agriculture
Forestry/Fishing
Mining
Construction
MANUFACTURING
Transportation/
Communication
Trade/Finance
Services/Public
Administration
Not Reported
1880
1900
1940
1950
1960
283.5
1.4
9.2
38.9
82.0
26.1
345.3
1.7
29.3
61.5
130.9
61.9
214.0
2.1
15.8
73.1
700.0
100.3
159.9
2.1
15.5
118.4
978.3
152.7
92.1
1.5
15.3
125.6
1,035.9
155.6
48.3
79.3
113. to
146.8
343.8
352.3
485.7
459.6
573.6
651.2
.4
15.0
23.4
31.7
89.6
a 1880 and 1900 from Simon Kuznets, Population Redistri
bution and Economic Growth, United States TB7b-ii?5Q (Philadel
phia: American Philosophical Society, 1^57), p. 62£.
1940, 1950, and 1960 from U.S. census of population
compiled in Leonard D. Bronder and John M. Koval, M i c h i g a n 1s
Economic Past: Basis for Prosperity (Lansing* State Resource
Planning Division, 6ffi.ce o£ Economic Expansion, Michigan De
partment of Commerce, 1967), pp. 43-107.
^Thousands
Throughout the 1950 and 1960 census years manufacturing em
ployment continued to increase while employment in the ex
tractive industries registered constant declines, thus widen
ing the gap between them.^
This rapid industrialization of Michigan is seen more
clearly when compared to other areas.
Value added by manu
facturing is one measure of manufacturing activity.
For the
It should be noted that the increase in manufactur
ing employment was made possible by the great increase in
the productivity of agriculture, not only in Michigan but in
the nation, which released employment to work in the growing
manufacturing industry.
period of 1900 through 1940, the period of most rapid indus
trial growth in the United States, Michigan had a growth rate
in value added by manufacturing almost three times that of
the nation and exactly double that of the next highest state
in the bast North Central Region
(includes Illinois, Indiana,
Michigan, Ohio, and Wisconsin).1'
There is considerable evidence to suggest that Michi
gan (and the nation) has now entered a new era which will see
a continued increase in manufacturing employment but a de
cline in its relative position.
This era has been dubbed the
"human resources era" and is characterized less by the actual
fabrication of raw materials than with the development of
technology which will further remove employment from expend
ing physical effort in the production process.
This general trend is best seen when presented vis u
ally.
Trends in total employment in manufacturing and manu
facturing employment as a per cent of the total employment
are presented in Figure 9.
The two trends which stand out
roost clearly are, first of all, the increase in the early
1900's of total manufacturing employment and of manufactur
ing employment as a per cent of the total employment.
The
second trend is the recent decline in manufacturing employ
ment relative to other types of employment.
Despite
^Leonard b. Bronder and John M. Koval, Michigan *s Econ
omic P a s t > Basis for Prosperity (Lansing* State Resource Plan
ning Division, Office of Economic Expansion, Michigan Depart
ment of Commerce, p. 7.
1200
1000
100
800
80
600
60
Total
Employmen
vO
o
400
40
Per Cent
30
ianufacturing
200
20
10
1880
1900
Figure 9.
1940
I960
1968
Total Employment and Per Cent Manufacturinga
aData for 1880 through 1960 were taken from Simon Kuznetz, Population Redistribution and Economic Growth. United States 1870-1950 (Philadelphia: American Philosophical Society, 195?). The data for 1968 were £rom unpublished statistics from the
Michigan Employment Security Commission, Lansing, Michigan.
yi
continued increase in manufacturing employment, the per cent
of the total employment in manufacturing has been declining
since approximately 1950.
Uronder and Koval have estimated
that manufacturing's share of the total employment will con
tinue to decline to 29 per cent in 1980.^ This research has
extended this rate of decline to 23 per cent in the year
2000.
This has been accompanied by an increase in absolute
numbers of employees.
The reason why manufacturing employ
ment can increase but have a decreasing share of the total
employment is because total employment is expected to in
crease at a faster rate than employment in manufacturing.
Composition of Manufacturing Employment
Almost equally as important as the amount of employ
ment to the water withdrawals for manufacturing is the compo
sition of the employment.
As mentioned earlier, there are
considerable differences in the water withdrawal rates per
employee among the several two-digit SIC categories.
For
this reason an awareness of the importance of each of the SIC
categories in the state's employment structure is necessary.
The proportion of the state total manufacturing employment in
each of the two-digit SIC categories for 1958, 1963, 1968,
and estimates for 1980 and 2000 are presented in Appendix
Table 8.
^Leonard D. Bronder and John M. Koval, Michigan's
Futuret Its Population and Its Economy (Lansingt ^tate Re
source Planning bivision, Office of Economic Expansion, Mich
igan Department of Commerce, 1967), p. 2.
92
As this table indicates, the bulk of manufacturing
activity in Michigan has been concentrated in a few major
industry types.
In 1968 over 70 per cent of the manufactur
ing employment in Michigan was in four industries, transpor
tation equipment, non-electrical machinery,
products, and primary metal products.
fabricated metal
None of the remaining
17 industry types accounted for more than 100,000 employees,
and none of them accounted for more than five per cent of the
total manufacturing employment in the state.
Combined they
total less than does the transportation industry alone.
Projections are for the transportation equipment in
dustry to further increase its dominance and to have over 36
per cent of the total state manufacturing employment by 2000.
The other manufacturing types which combined with transporta
tion equipment to total 70 per cent of the state employment
in 1968 are expected to either increase only slightly, or to
decline somewhat, to 1980 and 2000 so that by 2000 they will
comprise only 69 per cent of the state total.
This, however,
is not a major change and the industrial structure of the
state can be considered to remain relatively unchanged as far
as the dominance by the four industries is concerned.
Regional Manufacturing Employment
When one examines a map of the manufacturing employment
of Michigan it is immediately apparent that the location of
manufacturing activity is not evenly distributed throughout
the state.
Large portions of the state account for very
yj
little of the total manufacturing employment, while almost
all of the manufacturing activity is concentrated in a rela
tively few counties.
This distribution is most obvious when presented in a
regional framework.
The manufacturing employment in each of
the seven planning regions
total in each region)
(plus the per cent of the state
for 1958, 1963, 1967, and estimates for
1980 and 2000 are presented in Appendix Table 10.
Region I
The dominant position of Region I in the total state
manufacturing picture is easily seen in Appendix Table 10.
During the 10 year period between 1958 and 1967 Region I has
averaged over 58 per cent of the state total manufacturing
employment.
Its rate of increase in total numbers of manu
facturing employees has been comparable to most of the other
regions.
Between 1958 and 1963 Region I had a growth rate of
1.5 per cent per year in actual numbers of manufacturing em
ployees.
This was fourth highest among the regions of the
state, a lower increase than three other regions.
However,
between 1963 and 1967 Region I registered a 4.2 per cent per
year growth rate which was largest of all the regions.
Region I is expected to maintain its dominance in
manufacturing employment in 1980 and 2000, although its po
sition will probably be slightly lower.
The region is ex
pected to increase its total manufacturing employment .2 per
cent per year betweeJ 1967 and 2000.
It should have
*>4
approximately 5 7.8 per cent of the state's manufacturing em
ployees in 1980 and 57.6 per cent in 2000, a slight decrease
from the 1958 to 19b7 period.
The composition of manufacturing employment in Region
1 is presented in Appendix Table 11.
It can be seen that the
bulk of the manufacturing employment is concentrated in three
categories.
SIC categories 34, 35, and 37 accounted for 68.7
per cent of the total regional manufacturing employment in
1968.
Although there are some minor changes expected among
categories with respect to per cent of regional total, the
overall situation will probably not change significantly by
2000.
Fabricated metals, machinery, and transportation
equipment are expected to account for 66*4 per cent of the
total manufacturing employment in the year 2000.
In order to facilitate a more complete understanding
of the direction of shifts in manufacturing employment, pro
jections were made for each county and are essentially a d is
aggregation of the regional forecasts.
Because of the de
tailed nature of the county forecasts less confidence is ex
pressed in them than in the regional forecasts*
Nevertheless,
it is felt that they do give a relatively accurate indica
tion of the direction of sub-regional changes in manufactur
ing activity.
Each county's share of Region I's employment
projections are presented in Appendix Table 12*
The pattern of manufacturing employment within Region
I has changed much more rapidly than has Region I as a whole
95
relative to the other regions, and considerable changes are
forecasted for the future,
Wayne County has dominated the
region for as far back as any type of data are available, and
is expected to continue to be the most important manufactur
ing county in the region at least to the year 2000.
its
However,
position relative to the other counties in the region
is declining.
In 1958 Wayne County had 69.3 per cent of the total man
ufacturing employment in Region I.
Despite continued in
creases in manufacturing employment this had fallen to 64.8
per cent in 1963, and by 1967 Wayne County represented only
60.8 per cent of Region I*s manufacturing employment.
Dur
ing the 1958-1963 period Wayne County increased its manufac
turing employment only 1.0 per cent per year; this was the
lowest of any of the counties in Region I.
The forecasted manufacturing employment for Wayne
County in 1980 is 377,867.
This represents a decline in the
number of manufacturing employees of ,25 per cent per year.
A continued decrease in manufacturing employment is fore
casted to 2000 for Wayne County.
This is expected to be a
decrease of less than ,1 per cent per year, or approximately
5,500 for the 20 year period.
Three counties have profited most by the declining im
portance of Wayne County.
Oakland County has realized the
greatest absolute increase in manufacturing employment be
tween 1958 and 1967.
Employment increased by 43,630 for the
96
period, an increase of 7.2 per cent per year.
The per cent
of the state total manufacturing employment represented by
Oakland County increased from 9.75 per cent in 1958 to 14.54
per cent in 1967.
The rapid rate of increase in Oakland
County can be appreciated by comparing it with Macomb County.
In 1958 Macomb County represented a larger share of the reg
ional employment than did Oakland County by approximately 2
percentage points.
This gap increased slightly by 1963.
Between 1963 and 1967 Oakland County's manufacturing employ
ment increased roughly 8.9 per cent per year while Macomb
County was second only to Wayne County in the number of manu
facturing employees.
Oakland County is expected to increase
its manufacturing employment through 2000 at a rate second
only to Washtenaw County.
Macomb County has also profited considerably from Wayne
County's declining relative position.
Macomb County's manu
facturing employment increased by 26,915 during the 1958 to
1967 period.
This was the third largest increase in absolute
numbers in the region as was the 4,1 per cent per year rate.
Macomb County increased its relative position in the region
from approximately 12 per cent in 1958 to 13.7 per cent in
1967.
The forecasted increases for Macomb County will result
in that county having -5.18 per cent of the regional manu
facturing employment in 1980, and 16.58 per cent in 2000.
Although having a relatively smaller number of employ
ees in manufacturing in 1958, Washtenaw County has increased
97
at the highest rate of any of the counties in Region I (an
average of 7.7 per cent per year between 1958 and 1967).
If
Washtenaw County grows in manufacturing employment at the
rate forecasted it will have almost tripled its 1958 percent
age share of the state total by 2000.
The remaining counties in Region I, with the exception
of St. Clair, have all indicated a constant increase for the
1958 to 1967 period? all counties reported more manufactur
ing employment in 1967 than in 1958.
Some of these counties
have had a growth rate rivaling the larger counties
(Sanilac
4.6, Livingston 4.2) but because of the small initial size
have not made large contributions to the total manufactur
ing employment in the region.
Only one county
(Monroe) is
expected to have a declining manufacturing employment to
2000, and this is so small
(.1 per cent)
that it is insig
nificant.
Region IIA
In terms of total manufacturing employment Region IIA
has increased from 1958 to 1967, and a continued increase is
projected at least until 2000.
Region IIA had an annual rate
of growth of 2.0 per cent between 1958 and 1963, and a 1.8
per cent per year increase between 1963 and 1967.
As with
all of the other regions this rate of increase is expected
to be considerably less to 1980 and 2000.
The expected rate
of growth per year between 1967 and 1980 is expected to be
.3 per cent and .2 per cent between 1980 and 2000.
The
98
growth trend of Region IIA can be seen in Appendix Table 10.
Region IIA is expected to continue to rank second among
the planning regions in terms of manufacturing employment, al
though its relative position is expected to decrease slightly.
In 1958 Region IIA had 25.27 per cent of the state manufactur
ing employment.
After a slight rise in 1963 this decreased
in 1967 to a percentage lower than in 1958.
A continued,
although slight, decline is expected for the region to 1980
and 2000, but it should remain well ahead of the next most
important region.
The composition of manufacturing employment in Region
IIA is similar to that in Region I.
The structure of manu
facturing employment in Region IIA is indicated in Appendix
Table 11.
There is not the great reliance on SIC categories
34, 35, and 37 as in Region I.
These categories are still
the most important in Region IIA but they represent only
47.2 per cent of the total regional employment in 1968 as
opposed to the 68.7 per cent in Region I.
Five categories in Region IIA appear to be the probable
beneficiaries of the declining dominance of SIC*s 34, 35, and
37.
Where in Region I SIC's 20, 25, 26, 36, and 38 accounted
for only 9 per cent in 1967 they accounted for nearly 27 per
cent in Region IIA.
This is particularly important for in
dustrial water withdrawals since SIC 26 is the 4th most
demanding of water of the 21 SIC categories and SIC 20 is
8th.
In terms of the total amount of water withdrawn for
99
manufacturing in Region IIA the contribution of SIC 26 helps
to compensate for lower employment totals.
Examining the region in more detail
12) it can be seen that Kent, Kalamazoo,
(Appendix Table
Ingham, Muskegon, and
Berrien Counties have dominated the region's manufacturing
employment since 1958 and are expected to continue to be the
most important in 1980 and 2000.
These counties had 60.6 per
cent of the regional manufacturing employment in 1958.
This
percentage share fell slightly in 1963 but rose to 62.2 per
cent in 1967.
Sub-regional forecasts for Region IIA indicate
that these counties will probably maintain 62 per cent of the
regional manufacturing employment in both 1980 and 2000.
Within this group of counties Kent County has been the
most important and has steadily increased its share of the
regional manufacturing employment.
Ingham County has had a
continual increase in its percentage share of the regional
manufacturing employment since 1963 and in 1967 displaced
Kalamazoo County as the second county in that region.
other growth county in the group has been Berrien.
The
Except
for a slight decline at the 1967 period Berrien has had a
continual increase in its percentage share.
Muskegon and
Kalamazoo Counties have experienced a general decline in
their share of the total although their absolute manufactur
ing employment has increased.
Three other counties
(Calhoun. Jackson, and Ottawa),
while not among the top counties in the region, all had over
100
10,000
manufacturing employees in 1967, and combined they
represented a substantial proportion of the regional employ
ment.
Jackson and Ottawa Counties have registered absolute
increases from 1958 to 1963, and are expected to continue to
do so to 1980 and 2000.
Their percentage share has been in
creasing, although not steadily, during this period.
The
trend in manufacturing employment in Calhoun County, on the
otherhand, has been down.
Calhoun County's per cent of the
regional total has been decreasing constantly from 1958 to
1967 and forecasts indicate this will continue.
Absolute em
ployment is expected to continue to decline as well.
The remaining 12 counties are small in manufacturing
employment and contributed only about 18 per cent to the re
gional manufacturing totals in 1967.
Region IIB
Region IIB has never occupied an important position
in the manufacturing schema in Michigan.
Since 1958 this re
gion has never had more than 2.0 per cent of the total state
manufacturing employment.
Despite its small manufacturing
employment. Region IIB has been constantly increasing its
manufacturing activity, both in numbers and as a per cent of
the state total.
Between 1958 and 1967 Region IIB had the
largest rate of growth in manufacturing employment of any
region.
The region is expected to reach 2.0 per cent of the
total state manufacturing employment by 2000.
Region IIB's
relative position can be seen in Appendix Table 10.
101
The manufacturing structure for 1968 and forecasts to
1980 and 2000 are indicated in Appendix Table 11.
Regional
manufacturing employment is fairly well diversified, having
at least 2.0 per cent of its manufacturing employment in 15
of the categories, with no single category being of
overwhelming importance.
Slightly over 50 per cent of the
region's manufacturing employment in 1968, however, was in a
few categories.
The dominance of the food processing indus
tries in the region is undoubtedly a reflection of the re
gion's position in the major fruit region of the state.
Fab
ricated metals, electrical and non-electrical machinery each
accounted for more than 10 per cent of the regional employ
ment in 1968 and are estimated to continue their importance
to 2000.
users.
Fortunately these industries are not large water
Those industries which use large amounts of water
(over 1,000,000 gallons per employee per year) accounted for
16.3 per cent in 1968 and by 2000 they are expected to de
cline slightly in importance to 15.5 per cent.
The county pattern of manufacturing employment in
Region IIB is presented in Appendix Table 12.
Examination of
this table indicates that there does not appear to be any
major reordering of the ranking of the individual counties
in the region to 2000.
In 1967 Grand Traverse, Manistee,
Mason, and Osceola Counties accounted for almost 50 per cent
of the total manufacturing employment in the state.
By 2000
102
these counties are expected to have increased their combined
share of the regional manufacturing employment by slightly
less than 1.0 per cent.
Most of the growth appears to be
taking place in Grand Traverse and Osceola Counties, while
Mason County has experienced a rising absolute employment but
so slow in comparison to the regional growth that it has lost
in per cent of regional total.
This is expected to continue
to 2000.
With the exception of three counties the remaining
ones have had a rising manufacturing employment since 1958
and are expected to maintain this to 2000.
The rate of in
crease has not been uniform, however, and the relative posi
tions among the counties has changed.
Leelanau, Newaygo,
and Oceana Counties have had a declining manufacturing em
ployment since 1958, and are expected to continue to decline.
Kegion IIIA
Region IIIA has consistantly been the third largest of
the planning regions
(see Appendix Table 10).
It has in
creased its manufacturing employment 30,631 between 1958 and
1967 which was a growth rate of 3.1 per cent per year.
This
was only slightly less than the 3.4 per cent per year for
Region IIB, and was well ahead of any of the other large manu
facturing regions.
Kegion IIIA had approximately 11.38 per
cent of the manufacturing employment in the state in 1967.
This was down somewhat from the 1963 per cent.
Region IIIA
is expected to continue to grow in manufacturing activity to
103
19BU and 2000 although, as with the other regions, at a much
slower rate.
Manufacturing employment is expected to increase
in Kegion IIIA by approximately .8 per cent per year.
This
is only a little under the 1.0 per cent per year for Region
IIB, and represents the highest rate for any region with over
100,000 manufacturing employees.
The region is expected to
increase its position relative to the other regions at a rela
tively uniform rate, increasing from 10.8 per cent in 1958
to 11.4 per cent in 1967, and forecasted increases to 12.3
per cent in 2000.
Structurally Region IIIA is dominated by the manufac
turing of transportation equipment.
The composition of m anu
facturing employment is shown in Appendix Table 11.
With
only 11.4 per cent of the state manufacturing in 1967, Region
IIIA had approximately 20 per cent of the transportation
equipment industry.
When the next largest manufacturing
types are included
(primary metals, fabricated metals, and
non-electrical machinery)
it can be seen that the manufactur
ing structure in Region IIIA is very concentrated.
These
four manufacturing types constituted over 81 per cent of the
manufacturing employment in Region IIIA in 1968.
By 2000 it
is expected that this concentration will have risen to 82.6
per cent.
Of the manufacturing types,
13 have less than 1.0
per cent of the regional total each.
This manufacturing structure is very important to in
dustrial water withdrawals in Region IIIA.
SIC 33, with over
104
lb thousand employees expected in 2000, is a very heavy water
user.
over 4,000,000 gallons of water per employee are with
drawn each year in the primary metals industry.
Each employ
ee in SIC 33 withdraws as much as 13 employees in the trans
portation equipment industry.
This means that in the year
2000 the primary metals industry, with approximately 15,000
employees will be withdrawing about twice as much water as
the transportation equipment industry with approximately
94,000 employees.
With only a few exceptions, there has been a continual
increase in manufacturing employment for each county from
19SB to 1967 and continued increases are expected in each
county to 19B0 and 2000
(Appendix Table 12).
Genesee County
has been, and will probably continue to be, the major county
in the region with about 60 per cent of the manufacturing em
ployment.
This per cent has stayed uniform since 1958 and is
expected to continue steady until 2000.
Saginaw County, the
next largest, has increased its position somewhat and is ex
pected to have about 25 per cent of the manufacturing em
ployment in the region in 2000.
Saginaw and Genesee Counties
are expected to have well over 85 per cent of the regional
manufacturing employment in 2000.
Bay County, the third largest county in Region IIIA,
has increased its total manufacturing over the years and
is expected to increase to 1980 and 2000.
However, the rate
of increase has been less than the region as a whole and,
1U5
therefore, Day County's per cent of the regional total is d e
clining .
Region IIIB
Of all the regions in the Lower Peninsula, Region I1IB
seems to be the weakest in manufacturing activity
Table 10).
(Appendix
In fact, this region appears to be much like the
Upper Peninsula regions.
Manufacturing employment in Region
IIIB has never contributed a great deal to the state total
manufacturing, and it is expected to contribute even less in
the future.
Despite slight increases in absolute employment
to 1967 the region's percentage share of the state total d e
clined .
A very slight increase in manufacturing employment is
forecasted to 1980, but as a result of the general declining
importance of manufacturing throughout the state Region IIIB
was forecasted to have fewer manufacturing positions in 2000
than in 1980.
However, since the general assumption was made
that no region would have less employment in manufacturing
in 1980 or 2000 than it had previously, the 2000 employment
estimate was increased by 487 to make it equal to the 1980
estimate.
The unique position of the chemical industry
(SIC 28)
in Region IIIB is easily seen in Appendix Table 11.
In 1968
the production of chemicals accounted for 58 per cent of the
manufacturing employment in the region, and 31 per cent of the
employment in the chemicals industry in the state.
The
100
remaining 4 2 per cent of the manufacturing employment in the
region is relatively scattered, no manufacturing type having
over 7.u per cent of the t o tal.
The main factor in Region IIIB's slowly declining
position in manufacturing is probably related to the lack of
growth in the chemical industry.
The composition of manu
facturing employment for Region IIIB is indicated by Appendix
Table 11.
As shown, the chemical industry had a smaller pro
portion of the state total manufacturing employment in 1968
than in 1958.
This same relative position is forecasted to
continue into 1980 and 2000.
The absolute employment in
chemicals is expected to increase slightly to 1980 and 2000.
The importance of the chemical industry in Region IIIB
assumes even greater significance when it is realized that
this industry is the second largest water user of all of the
21 manufacturing categories.
Therefore, primarily because of
the chemical industry large amounts of water were withdrawn
for manufacturing purposes in 1968 in Region IIIB and will
continue to be withdrawn in 2000.
County manufacturing employment in Region IIIB is pre
sented in Appendix Table 12.
This table indicates that M id
land County is the primary manufacturing county in the region.
Since 1958 Midland County has had up to 66 per cent of the
manufacturing employment in this 15 county region.
general increases in manufacturing employment
forecasted 1980 to 2000 period)
Despite
(except for the
the rate has been low enough
so that Midland County has been a declining proportion of the
regional total.
Continued declines are forecasted*
Midland County is the only county in the region with
over 5,000 manufacturing employees;
most 15,000.
3,000.
in 1967 Midland had al
The next largest county, Alpena, had less than
Together these counties have approximately 72 per
cent of the manufacturing employment in the region.
However,
because of declining proportion of regional total in both
Midland and Alpena Counties they are estimated to have only
67 per cent in 2000.
The only other counties of any consequence in manufac
turing employment are Cheboygan, Clare, and Otsego Counties.
Each of these, however, had less than 5.0 per cent of the
regional total in 1967.
Regions IV, V
The regions of the Upper Peninsula have never been im
portant in manufacturing.
Combined, the 15 counties in these
regions accounted for only 1.11 per cent of the total state
manufacturing employment in 1967
(see Appendix Table 10).
Not only does the Upper Peninsula not have a large
manufacturing employment but it has declined considerably in
recent years, falling from 1.6 per cent of the state total in
1958 to the 1.11 per cent in 1967.
The absolute decline has
been greater in Region V than in Region IV, although both
regions had less manufacturing employment in 1967 than in
1958.
108
Projected employment figures to 1980 and 2000 indicate
an even smaller amount of manufacturing activity.
However,
since the decision was made that the 1980 and 2000 employment
would be no less than the 1967 level, both regions were sta
bilized at the 1967 employment totals for 1980 and 2000.
This has resulted in a forecasted decline in the per cent of
the state total for the Upper Peninsula, reaching 1.02 per
cent in 2000.
The manufacturing structure of the Upper Peninsula is
indicated in Appendix Table 11.
The most important type of
manufacturing is related to the timber resources of the area.
In 1968 approximately 42 per cent of the manufacturing employ
ment in the regions was in lumber and wood products, and
paper and allied products.
The next most important type of
manufacturing activity was in the non-electrical machinery
industry.
This category employed 16 per cent of the m an u
facturing employees in 1968.
The importance of SIC 26
(paper and allied products)
to the employment structure of the Upper Peninsula means that
this area withdraws far more water than its small manufactur
ing employment would indicate.
SIC 26 is the fourth heaviest
user of water with a withdrawal of over 3,500,000 gallons per
employee per year.
The other heavy water users are not large
employers in the Upper Peninsula.
The sub-regional manufacturing employment in the Upper
Peninsula is listed in Appendix Table 12.
In 1967 three
100
counties accounted for about 44 per cent of the total.
This
relationship is expected to remain at least through the year
2000
.
CHAPTER VI
DLTLRMINAT ION OF MANUFACTURING
WATER WITHDRAWALS
The technique used in estimating manufacturing v/ater
withdrawals in Michigan for 1980 and 2000 involves two major
steps.
The first of these is the identification of manufac
turing employment for each county, broken down into the 21
two-digit Standard Industrial Classification (SIC) manufac
turing categories
(see page 80)•
The second step is the ap
plication of water withdrawal coefficients, or the amount of
water withdrawn annually by each employee, to these employ
ment figures.
Estimates of Manufacturing Employment
Information Needed
There arc three main types of data which are required
to compute the various types of manufacturing employment in
each county for 1980 and 2000.
First of all, estimates of
total manufacturing employment in each county are needed.
Second, data are required on total state employment for each
SIC category for 1980 and 2000. And third, information is
needed on the proportion of each county's manufacturing
110
Ill
employment in each SIC category at some index period in the
past, and estimates for 1980 and 2000.
Virtually none of
these types of information were available in a form which
could be used in this study and, consequently,
they had to be
estimated.
Data Availability
Data on manufacturing employment are available from
five potential sourcesx
of Manufacturers,
(1) Census of Population,
(3) Annual Survey of Manufacturing,
(4) County Business Patterns,
ity Commission.
(2) Census
(5) Michigan Employment Secur
Each of these data sources has its own ad
vantages and disadvantages and, therefore, each was considered
for use in this study.
Census of Population
The Census of population is taken every 10 years by
the U.S. Bureau of the Census.
Information on manufactur
ing employment in each county has been recorded in each cen
sus period and thus represents a relatively long period of
observation.
The Census of Population, however, has a num
ber of disadvantages which preclude its use in the examina
tion of the location of manufacturing activity.
The first disadvantage of these data is a result of
the method of collection.
The employment figures reported
in the Census of Population are based upon the place of res
idence rather than the place of work.
For this study the
112
major concern is the number of people expected to work in a
county in 1980 and 2000 and not how many people who lived in
the county worked in manufacturing regardless of where this
was.
This difference may not appear to be significant.
How
ever , for counties with relatively low manufacturing employ
ment opportunities, but which are within commuting distance
of a large industrial county, the difference may be consid
erable.
In addition to this major deficiency there is a sec
ond one which makes it difficult to work with the Census of
Population data.
The census of Population reports SIC em
ployment as grouped data.
That is, several individual
two-digit SIC*s are grouped together with no indication of
how they can be proportionally separated.
Since the water
withdrawal coefficients are based upon individual two-digit
categories there would be no rational basis for assigning
withdrawal rates to groups of SIC categories.
Census of Manufacturers
The Census of Manufacturers, also taken by the U.S.
Bureau of the Census, does not have the defects of the C en
sus of Population.
In the Census of Manufacturers employment
is reported according to the place where the individual works
rather than his place of residence.
And the data is pre
sented at the individual two-digit SIC level rather than as
groups of two-digit SIC categories.
However, the Census of
Manufacturers does have two disadvantages which makes this
113
data source less than ideal.
l'irst of all, there is the problem of time series.
The Census of Manufacturers is available in usable form for
only two census periods,
1958 and 1963.
The census of 1958
represented a significant change in the manner of classifying
manufacturing activity over previous censuses.
Industries
which were placed in one category in previous censuses were
often found in another in 1958.
The 1958 and 1963 censuses
were taken using approximately the same categories and are
considered to be comparable data.
For censuses taken before
1958 the difference is so great that any comparison between
the two periods would incorporate so much inconsistency as
to make the results invalid.
The second disadvantage of the Census of Manufacturers
is the incompleteness of data due to disclosure reasons.
When reporting employment for a county would result in dis
closing information about individual producers this informa
tion is withheld.
This does not apply to total county em
ployment figures but only when these are broken down into the
two-digit SIC categories or lower.
For many counties well
over half of the employment by SIC categories is withheld
for reasons of disclosure.
Annual Survey of Manufacturing
The Bureau of the Census conducts annual surveys of
manufacturing activity for intervening years between the Cen
sus of Manufacturers which are taken every five years.
114
Information is reported on manufacturing employment for se
lected types of manufacturing in 12 or 15 of the largest
counties in the state.
However, not all counties in the
state are reported and even in those which are there are con
siderable gaps in the data.
The principle strength of the Annual Survey of Manu
facturing are data on state employment in each of the
two-digit SIC categories.
This information is relatively
complete and when used with the 1958 and 1963 Census of M an
ufacturers provides a reasonably long time series on state
manufacturing employment by SIC category.
County Business Patterns
The County Business Patterns, also a publication of
the U.S. Bureau of the Census, reports data only for individ
uals who are covered by social security.
Publication has
been every three years, except for two years between 1965 and
1967.
The main disadvantage with this source is that it is
possible that the distribution of covered employment might
not be proportional to the actual distribution of manufactur
ing employment, thus causing some areas to be over repre
sented while others could be under represented.
In addition to the question of validity, the County
Business Patterns suffer from the same defect as the Census
of Manufacturers, that of disclosure.
The amount of data
withheld is approximately the same for each source.
11!j
Michigan Employment Security Commission
The Michigan Employment Security Commission collects
and disseminates data on employment which is covered by un
employment compensation.
Every firm which employes one per
son in addition to the owner is required to report to the
Employment Security Commission.
Thus, while their coverage
is not complete it may be regarded as complete without any
substantial error.
Information is readily available from the Employment
Security Commission on total state manufacturing employment,
on employment in each of the two-digit manufacturing SIC cat
egories, and total manufacturing employment for some counties.
These are generally the larger counties or counties in which
there has been substantial and persistent unemployment.
These data are usually reported as grouped data, however,
with the employment of two or more counties grouped together.
However, no information was reported for 48 of the 83 coun
ties in the state.
While these counties are generally the
ones which contribute the least to the total manufacturing
employment in the state it is a major disadvantage when the
purpose is to derive a complete state-wide picture of future
manufacturing activity and related water withdrawals as this
leaves a very large portion of the state unaccounted for.
Data on individual counties by SIC categories is gen
erally withheld, as in the Census of Manufacturers and the
County Business Patterns, for disclosure reasons.
However,
116
for the purpose of this study the researcher was able to o b
tain data on county employment by SIC categories for 196U.
This data was relinquished with the agreement that it was not
to be reported in such a manner that would violate the ini
tial reasons for withholding it.1
Methodology
The estimation of manufacturing employment for each
county stratified by SIC categories involves five major steps:
(1) estimates of total state employment for 1980 and 2000,
(2) estimates of state manufacturing employment for 1980 and
2000,
(3) sub-state estimates of manufacturing employment
(regional and c o u n t y ) , (4) estimates of state manufacturing
employment by SIC categories,
(5) estimates of county m an u
facturing employment by SIC categories.
Estimates of Total State Employment
Before making estimates of state manufacturing employ
ment it was necessary to make estimates of the total state
employment expected for 1980 and 2000.
Since employment in
The 1968 Michigan Employment Security Commission data
on county employment by SIC categories was for one period
early in 1968.
These data were used for both county totals
and as an indication of manufacturing structure.
It was felt
that this data may or may not reflect unusual conditions at
the time it was gathered.
For this reason, when substate
(regional and county) total employment figures were needed
the 1967 monthly average for the state as a whole was used by
disaggregating this figure proportionally based upon the 1968
data.
When referring to manufacturing structure (the m anu
facturing employment in each SIC category) the data are re
ported as 1968 since this information is reported in propor
tions and not in actual figures which would violate the con
fidence of the data.
117
manufacturing is the prime mover of the Michigan economy, and
because estimates of manufacturing employment will be exam
ined as a changing proportion of total employment, it was
mandatory that accurate estimates of total state employment
be made.
Projections of total employment were made for Michi
gan in 1966 to be used in another study.
These projections
for 1970, I960 and 2020 are presented in Table 9.
Table 9
Total State Employment, Projections to 2020a
— ■' - "
Year
1
■
~
Employment
2020
6,876,0
1980
3,769.0
1970
3,222.0
1960
2,727.0
Battelle Memorial Institute, Grand River basin (Mich
igan) Comprehensive Water Resources Study Appendix O; Econ
omic Base Study (Columbus, Ohio, 1 & 66), p. 1-15.
^Thousands
Data on total employment in Michigan for the last five
years are presented in Table 10.
These data are from the
Michigan Employment Security Commission.
the employment for 1968
estimate for 1970
It can be seen that
(3,243,600) exceeds the battelle
(3,222,999).
booking at the years immediately preceeding 1968,
118
Table 10
Total State Employment, 1964-1968*
Year
Employment
1968
3,243.6
1967
3,208.8
1966
3,183.6
1965
3,097.8
1964
2,936.0
aUnpublished statistics from the Michigan Employment
Security Commission, Lansing, Michigan.
T ho usands.
there was an increase of 25,200 employees between 1966 and
1967, and an increase of 34,800 between 1967 and 1968.
Using
an annual rate of increase of 30,000 the 1970 employment
would be 3,303,600.
This would be 2.5 per cent greater than
the estimate made by Battelle for the same period.
The Michigan Office of Economic Expansion has esti
mated that the 1980 employment in Michigan will be approxi
mately 3,954,890.1 For 1980 battelle estimates employment to
be approximately 3,769,000.
This is 4.9 per cent lower than
the estimate of the Office of Economic Expansion.
Since the Battelle employment estimate for 1970 was
approximately 2.5 per cent too low, and because the Battelle
^Bronder and K o v a l , M i c h i g a n ’s Future;
and Its Economy, pp. 18-29.
Its Population
estimate for 1980 is also lower than the 1980 estimate of the
Office of Economic Expansion, it is assumed that the Office
of Economic Expansion's estimate for 1980 represents more
realistically the actual situation.
No estimates for total employment for 2000 could be
found.
However, Battelle has estimated that employment in
Michigan will increase by 1.5 per cent per year between 1980
and 2020.* Applying this rate of increase to the 1980 employ
ment estimate from the Office of Economic Expansion which was
3,954,890 the result is 5,326,683 in 2000.
For the purpose of this study, then, the Office of
Economic Expansion's figure of 3,954,890 for 1980 and
5,236,683 for 2000
(an increase in the 1980 figure of 1.5
per cent per year for 20 years) were accepted as employment
levels for the target years.
Figure 10 is a graphical pre
sentation of these data, and indicates that the estimates
are realistic when compared to previous trends.
Estimates of State Manufacturing Employment
There is a paucity of information on expected future
levels of manufacturing employment in Michigan.
Apparently
the only estimate of future manufacturing employment that has
been made for Michigan as a whole was made by the Office of
Economic Expansion.
However, these projections extend only
to 1980 and provide no data on levels of manufacturing
^Battelle Memorial Institute, Grand River Basin,
p. 1-15.
6 ,000,000
5,000,000
,
4 000,000
3,000,000
1940
1950
1960
1970
1980
1990
2000
Figure 10 . Total Employment
&Data for 1940 through 1960 taken from U.S. Bureau of
the Census, U.S. Census of Population; 1960. Vol. I, Char
acteristics of the Population. Part 24, Michigan (Washington:
U.S. Government Printing Office, 1963). Projections were
taken from battelle Memorial Institute, Grand River Basin
(Michigan) Comprehensive Water Resources fitudy Appendix Ot
Economic Base Study (Columbus, Ohio, 1965).
employment in the year 2000.
The initial plan was to use the estimate of total e m
ployment as a base and to estimate the proportion of the
total employment made up of manufacturing employment.
Sev
eral techniques were tried in an attempt to derive the 1980
and 2000 proportions.
These are listed below and are fol
lowed by a short discussion of the results.
The following techniques were tried as a means of
estimating the proportion of total employment made up by m a n
ufacturing employment:
(1) the application of the most recent
proportion to the estimates of total employment,
(2) a
straight line projection of the changes over the years in the
proportion manufacturing employment is of total employment,
(3) a straight line projection but constraining the propor
tion to .29 in 1980,
(4) extension of the per cent change in
the total employment/manufacturing employment relationship.
Technique 1
The most recent information which is available on the
proportion of manufacturing employment to total employment
for the state as a whole is from the Michigan Employment Se
curity Commission and is for 1968.
One approach in estimat
ing manufacturing employment in the state for 1980 and 2000
is to assume that the proportion of manufacturing employment
to total employment will remain constant, and to use the
1968 proportion, applying this to the estimates of total e m
ployment for 1980 and 2000.
122
Examination of the data for the 1950 to 1968 period
indicates that this relationship has not remained constant
over the years but has changed considerably.
presented in Appendix
Table 13.
This data is
The use of a constant pro
portion would not reflect these basic changes in employment
structure and, therefore, was considered unsatisfactory.
Technique 2
This technique used to estimate the 1980 and 2000 m an
ufacturing employment was to determine the proportion of the
labor force engaged in manufacturing each year and to project
this proportion over time.
A participation rate could then
be determined for 1980 and 2000 which would reflect the d y
namic nature of the variable.
The data used for this pro
jection were those listed in Appendix Table 13.
This approach is essentially a time series projection
where observations are taken of the participation rates each
year from 1956 through 1968, and
1980 and 2000.
this “trend" projected on to
Such a model will assume that the conditions
operating in the 1956-1968 period will continue to operate in
1980 and 2000 in approximately the same magnitude, and that
no new factors will enter which will disrupt this relation
ship.
The model for participation rate then becomes Y*f(t)
where Y is the participation rate and t represents time.
There will almost always be a number of factors which
combine to influence the magnitude of the participation rate.
These factors can be combined into an equation which normally
has a greater amount of predictive power than a projection
based solely on time.
To be used, however, the future values
of these "other factors" must be known, and this usually in
volves additional projections.
The result is an end figure
(in this case a participation rate) which is a result of a
projection based upon other factors which are in turn pro
jected.
Projections of a time series implicitly includes the
effect of these other factors even if the individual effect
of each can not be separated from the combined effect.
Thus,
while weaker than a model that incorporates discrete varia
bles into a mathematical formula, the time series model does
incorporate the indirect effect of these other variables.
The actual placing of the trend line can be accom
plished by visually locating the line so that it appears to
"fit" the data.
However, a more precise method is available
with the use of the least squares regression technique.
This
method assures that with any given set of data the trend line
will be placed so that the errors between it and the actual
values which it purports to estimate will be smaller than if
the line was in any other location.
This second method of
fitting the trend line was the one employed.
When the initial regression was run with the propor
tion of the employment in manufacturing as the dependent v a r
iable and with time as the independent variable the correla
tion coefficient was only .48.
This regression was run once
124
more,
this time with two independent variables,
time and the
variable "unemployment as a per cent of the total labor
force." With the addition of the second variable the correla
tion coefficient became
.87.
Since unemployment was to be included in the equation
it was necessary to estimate the rate of unemployment for
1980 and 2000.
Because there are so many unknowns involved
in determining the rate of unemployment, the mean unemploy
ment for the 1956 to 1968 period was considered to be the un
employment rate for both 1980 and 2000.
Utilizing this model produced a trend line which
reached zero in 1990 and negative proportions after that.
This is obviously unrealistic, and it was necessary to in
corporate into the model criteria which would produce a more
plausable estimate.
Technique 3
This technique was a modification of the previous one,
differing in two important ways.
First, in addition to year
and unemployment, the variable "unemployment squared" was
added to the model.
This had the effect of lessening the de
gree of slope of the regression line over time as it ap
proached zero.
The second modification in the model was to
constrain the 1980 estimate to a proportion that had been
previously estimated by the uffice of Economic Expansion;
the
Office of Economic Expansion has estimated that manufacturing
12 5
employment in 1980 will be 29 per cent of the total employ
ment in the state.^ The effect of this constraint was to
force the regression line through the .29 level for 1980
while continuing to utilize the 1956 through 1968 data.
Using this model to estimate the 1957-1968 proportions
produced results which were very close to the actual propor
tions for those years.
However, using this model for 2000
produced a figure of .33 for manufacturing as a proportion of
total employment.
This is not the expected result when the
trend of the previous data and the estimate of .29 for 1980
are considered.
Instead of increasing from .29 to .33 the
trend should have been towards a further decrease.
It also
produced a "jump** in the trend line that could not be ac
counted for.
Projections which exhibit an abrupt change from past
trends are normally the result of very strong forces.
Since
these forces were not evident in Michigan it was assumed that
these abnormalities were a result of inadequacies in the model.
The critical value of the model appeared to be the use
of the mean rate of unemployment.
To test the sensitivity of
this variable the estimated proportion for 2000 was computed
using the highest rate of unemployment in the 1956-1968 peri
od
(13.6) and the lowest rate of unemployment ever experi
enced during this same period
(3.5).
There was some change
^Bronder and Koval, Michigan's Future:
and Its Economy, pp. 18-29.
Its Population
126
in the proportion for 2000 but this was relatively small.
Using the high unemployment rate the value of 30.5 was the
indicated proportion, or 2.46 per cent lower than when the
mean unemployment rate was used.
When the low unemployment
rate was used the value was 36.V, or 3.88 per cent higher
than when the mean unemployment rate was used.
Technique 4
The per cent manufacturing employment was of total em
ployment for the 11 year period between 19 57 and 1968 is in
dicated in Appendix Table 13.
The Michigan Office of Econ
omic Expansion has estimated that employment in manufactur
ing will be 29 per cent of total employment in 1980.
This
29 per cent appears to be a reasonable figure when considered
in the light of the 1957 to 1968 trend, and it was accepted
as the rate to be used for this research.
Applying the 29
per cent rate to the estimated total employment in 1960, the
result is 1,146,918.
The estimated fugure for manufacturing employment as a
per cent of total employment for 2000 using Technique 4 was
developed by extending the rate of decrease past the 29 per
cent for 1980.
Referring again to Appendix Table 13, it can
be seen that there has been a sizable change in the per cent
from year to year.
In order to start from a base which repre
sents a more stable condition rather than yearly fluctuations
the mean per cent for the last three years was used as a base
rather than the 31.6 per cent for 1968.
Taking the mean of
127
the last three years the result is 33.95.
Accepting this 33.95 rate for the 1966 to 1968 period
as a more accurate rate than the actual 1968 rate, there has
been a decrease of approximately 1.25 per cent per year at a
compound rate of decrease to the 29 per cent estimated for
1980.
At this 1.25 rate of decrease the rate would be 22.62
in 2000.
Applying this 22.62 per cent to the estimated total em
ployment for 2000
(5,326,683)
the result is 1,205,288.
This
figure is accepted as the total state manufacturing employ
ment for 2000 for the purpose of this research.
These two figures, then, 1,146,918 for 1980 and
1,205,288 for 2000 are chosen as the estimated manufacturing
employment levels for this study.
Figure 11
indicates graph
ically the trend in manufacturing employment from 1940 through
1968 with the estimates for 1980 and 2000.
These estimates
represent a reasonable continuation of the trend in manufac
turing employment since 1950, and reflect the increasing ab
solute importance of manufacturing employment, but the de
clining relative importance.
Sub-State Estimates of Manufacturing Employment
Sub-state estimates of manufacturing employment have
been made for two levels of generalization.
Employment esti
mates were first made for the seven planning regions.
were then disaggregated to the county level.
These
128
1,200,200
1 , 100,000
1 ,000,000
900,000
800,000
700,000
1940
1950
1960
Figure 11.
1970
1980
1990
2000
Manufacturing Employment*
aData for 1940, 1950, and 1960 were taken from U.S.
bureau of the Census, U.S. Census of Populationt I960* Vol.
I, Characteristics of the Population. Part 24, Michigan
(Washington: U.S. Government Printing Office, 1963). Data
for 1963 were taken from U.S. Bureau of the Census, Census of
Manufacturers: 1963. Vol. Ill, Area Statistics. Part 23,
Michigan (Washington: U.S. Government Printing Office, 1966).
Data for 1967 were from unpublished statistics, Michigan Biployment Security Commission, Lansing, Michigan. Projections
for 1980 and 2000 were made by the author.
12 9
kuyional Estimates
After the estimates of state manufacturing employment
were made it was necessary to break this total down into the
seven planning regions chosen for this study.
Two different
approaches were utilized in this effort, one applying to the
two regions of the Upper Peninsula and one applying to the
remaining five regions of the bower Peninsula.
Upper Peninsula
The contribution of the Upper Penin
sula to the total state manufacturing employment has never
been great.
In the period between I960 and 1967 the entire
Upper Peninsula had never accounted for more than 1.56 per
cent of the total manufacturing employment in the state
(see
appendix Table 10).
The future of the Upper Peninsula as a contributor to
the total state manufacturing employment is also expected to
remain slight.
The Office of Economic Expansion has esti
mated that the Upper Peninsula will not only fail to in
crease in manufacturing employment to 1980, but that it will
actually lose manufacturing employment and in 1980 will ac
count for only 1.07 per cent of the total manufacturing em
ployment in the state.
For this reason the assumption is made that the posi
tion of the two regions in the Upper Peninsula will not change
from the 1967 situation.
In 1967 Region IV had approximately
7,193 employees in manufacturing, while Region V had 5,026.
It is assumed that these figures will remain constant for
130
1980 and 2000.
This assumption is based upon four reasons.
First, the population of the Upper Peninsula, which has been
decreasing in recent years, is projected to increase to 1980
and 2000.
Second, it is believed that the efforts of the
various organizations charged with the economic development
of the Upper Peninsula will begin to stem the decrease in
manufacturing employment.
Third, the reductions in the Mack
inac Bridge tolls will contribute to making the Upper Penin
sula a more integral part of the total Michigan economy.
Fourth, the increased attraction of the Upper Peninsula as a
recreation area will contribute to maintaining its present
manufacturing position.
Lower Peninsula.
Projections of the manufacturing em
ployment of the Lower Peninsula were made by doing a time
series projection of the actual employment in each of the
regions to 1980 and 2000, determining what per cent of the
total was in each of the regions for the two years, and then
applying these per cents to the already projected state manu
facturing employment figures exclusive of the Upper Penin
sula.
These estimates were derived by using a time series
regression based upon data from the 1958 and 1963 Census of
Manufacturers and the unpublished 1967 statistics from the
Michigan Employment Security Commission.
It is recognized
that these data have three disadvantages:
(1) the time span
131
is relatively short,
(2) there are only three observations,
(3) the data were taken from more than one source.
The first two disadvantages reflect the general lack
of available data.
It is believed that the inadequacies of
the sources used are less, however, than the alternate
sources of data which were available for use.
The third d is
advantage is minimized by the fact that the coverage is v ir
tually synonymous.
The 1967 Employment Security Commission
data covers all employment in firms which employ one person
besides the owner.
The Census of Manufacturers' data covers
all full or part-time employees.
The regression equations for regional employment for
1980 and 2000 are of the form -1,003,640 + (523 x 1980) “
31,900, and -1,003,640 + (523 x 2000) * 42,360.
The number
-1,003,640 is a constant and 523 is the regression coeffi
cient.
Regression coefficients and constants for the regions
are shown in Appendix Table 14.
This same form is used in
estimating county employment and SIC employment.
The results of the regression are shown in Table 11.
The column "adjusted" refers to the proportional adjustments
of the estimated figures so that when totaled they do not ex
ceed the total state manufacturing employment exclusive of
the 12,219 employees estimated for the Upper Peninsula.
County Estimates
Although the primary objective is to estimate water
withdrawals for 1980 and 2000 in each of the seven planning
132
Table
11
Regional Manufacturing Employment
Estimated
R
1980
I
.94
865,094
1,198,014
662,551
694,109
1IA
.97
376,116
519,676
288,100
301,002
Ilti
.99
29,507
42,727
22,581
24,695
IIIA
.99
178,985
256,205
136,618
148,414
IIIB
.91
31,900
42,360
24,849
24,849
Region
2000
Adjusted
1980
2000
regions estimates of manufacturing withdrawals were made for
each county.
This was necessitated because of the overall
importance of manufacturing to the economy of Michigan.
Estimates of manufacturing employment by county is a
much more tenuous procedure than the estimation of regional
manufacturing employment.
When working with regions abrupt
changes in the employment level in one county or another will
normally cancel out
(increases in one county being balanced
by decreases in another) thus providing a more uniform trend.
At the county level, however, abrupt changes from one year to
the next are directly reflected in the trend and often bias
the extension of the time series.
For this reason, in esti
mating county employment it was necessary to make adjustments
to counteract this tendency.
The final estimate of county employment, the one which
will be used for the final computation of manufacturing water
133
withdrawals,
is designed to reflect the nine year trend in
manufacturing employment from 1958 to 1967.
This trend was
determined by doing a time series regression for each county.
It was discovered that 20 counties out of the 83 in the state
had a trend line which indicated a decreasing level of manu
facturing employment.
The remaining 63 counties had a posi
tively sloping regression line which indicated an increasing
level of manufacturing employment.
Separate techniques were
used for each group.
For the 20 counties which registered a decreasing
level of manufacturing employment it was assumed that each
one had already reached its lowest level of employment in
1958, 1963, or 1967 whichever one was the lowest.
Employment
figures were initially assumed to remain at this level to
1980 and 2000 subject to later proportional adjustment so
that the total employment of all counties in a region would
equal the regional employment.
The rationale for this assumption stems from the be
lief that those firms which are more prone to lower employ
ment levels will register the first losses, and that sub
sequent losses will be fewer because the status of the re
maining industries as a group will be improved and less prone
to losses than before the initial losses occured.
If em
ployment changes in subsequent years are allowed to reflect
the earlier losses, employment would reach zero before 2000
in nine counties.
134
For the 63 counties which indicated an increasing
level of manufacturing employment the final employment level
represents a median figure between a high and low estimate.
It is felt that a final figure which represents the median
position between two probable extremes will more accurately
reflect the situation in 1980 and 2000.
The first estimate was made by assuming that the 1967
per cent of regional total remains unchanged to 1980 and
2000.
These per cents, when applied to the regional esti
mates for 1980 and 2000, will represent a minimum level of
employment for the 63 counties with increasing levels of
manufacturing employment.
The application of a constant proportion,
is often used as a means of forecasting.
in itself,
By assuming that
future conditions will be related to a base condition by the
same proportion one is almost assured of being reasonably
close to the future conditions, provided the base condition
has been correctly determined, unless there have been dras
tic changes in the interim.
However, by using a constant
ratio one fails to take into consideration possible changes
over time.
The high estimates for the 63 counties which have ex
perienced an increasing level of manufacturing employment is
derived by the time series regression.
Each county's 1980
and 2000 manufacturing employment is allowed to reach the
maximum possible based upon the given data for 1958, 1963,
135
ami 1967.
The regression coefficients and constants are pre
sented in Appendix Table 15.
The final estimate for each county was then made by
choosing the figure which is half way between the high esti
mate and the low estimate.
When these final employment estimates for counties were
grouped according to regions and totaled they did not always
equal the 1980 and 2000 previously estimated regional totals.
These counties v/ere then “forced" to fit the regional total.
This was done by inflating or deflating them proportionally.
For example,
if a county accounted for five per cent of its
regional total, five per cent of the difference between the
projected regional total and the sum of the county estimates
would be added to or subtracted from that county.
When this
is done for each county in the region the county totals, when
summed, will equal the projected regional totals.
Estimates of state SIC Employment
The data available for use in the estimation of SIC
category employment for the state was more pleantiful than in
the estimation of regional and county employment.
There were
data for 11 years available for the period between 1958 and
1967.
These data were from the 1958 and 1963 Census of Manu
facturers, and from the 1959, 1960, 1961, 1962, 1964, 1965,
and 1966 Annual Survey of Manufacturing which are normally
taken between the regular censuses.
These data are collected
by the same agency, for the same purpose, and with essentially
130
the same criteria.
The only major difference being that one
is a census and the other a survey.
It is felt, however,
that in spite of this disadvantage these are still the best
data available.
The basic method used in the estimation of employment
by SIC category was the extension of a time series by regres
sion.
However, because of missing data in four categories
it was necessary to use two approaches.
Employment data for four categories
(19, 21, 31, and
39) were not reported in the Annual Survey of Manufacturing
or were missing from either the 1958 or 1963 Census of Manu
facturers.
These categories combined accounted for a very
small proportion of the total state employment
(1.6 per cent
in 1968) and the lack of data is not believed to be vital to
the projections.
Estimates for 1980 and 2000 for these cate
gories were made by assuming that the per cent of the total
state manufacturing employment for each of the four
ies in 1968 would not change before 2000.
categor
The 1968 per cent
of state total was then applied to the estimated total manu
facturing employment for Michigan in 1980 and 2000.
The manufacturing employment in 1980 and 2000 for each
of the remaining 17 categories was projected by a time series
regression.
These categories were then divided into two
groups based upon the strength of the trend of the data.
R^ of .50 was arbitrarily chosen as the breaking point be
tween those categories exhibiting a "weak trend" and those
An
137
exhibiting a "strong trend." Six of the categories had a R
of less than .50 (SIC 25, 26, 27, 20, 29, 32).
These cate
gories accounted for 12.9 per cent of the state total in 1968.
For these categories the assumption was again made
that the 1960 per cent of state total for each category would
not change before 2000.
These per cents were then applied to
the remaining total state manufacturing employment
(that re
maining after subtracting out the employment in the first
four categories).
For the categories with an R
2
more than .50 the number
of employees for 1980 and 2000 were estimated by means of the
time series regression,
when summed, these category projec
tions totaled more than the remaining employment 85.5 per cent
of the state total)•
These categories were then "forced" to
fit the remaining total employment by adjusting them propor
tionally as in the estimation of county employment.
The d e r
ivations of the SIC estimates and the data for the regression
equations are listed in Appendix Tables 16 and 17 respectively.
Estimates of County SIC Employment
Detailed and complete employment figures for each coun
ty by SIC category are not readily available.
For this r e
port the 1968 Michigan Employment Security Commission data
were utilized.
In addition to county totals, data were
available at the two-digit SIC category level for all 83
counties.
This provided an 83 by 21 matrix for the state
providing detailed information on the structure of Michigan
138
manufacturing for a very recent period.
The 1968 data on manufacturing employment serves as a
point of reference for determining future cell employment for
1980 and 2000.1 To compute these future data the 1968 cell
figures are adjusted so they conform with changes expected in
total county manufacturing employment and employment in the
two-digit manufacturing SIC categories for 1980 and 2000.
The first step in this process is to determine the
difference between the total county employment in 1968 and
that expected in 1980.
if the 1968 employment is less than
that expected in 1980 the 1968 total must be increased.
This
is done proportionally based upon the percentage of the total
county employment in each of the SIC categories.
For example, imagine a county which had manufacturing
employment in only four of the two-digit SIC categories.
SIC
20
25
30
35
Emp
25
75
50
50
In this case the total manufacturing employment for the coun
ty would be 200.
Each of the categories would also contain a
certain per cent of the total.
SIC
20
25
30
35
Emp
12.5%
37.5%
25.0%
25.0%
Imagine that the manufacturing employment in the coun
ty now increased by 100.
These 100 employees must be added to
cell is defined as one SIC category in one county.
Each county has 21 cells and each category 83.
This provides
an 83 by 21 matrix of 1,743 cells.
139
the county totals.
This is done proportionally, based upon
the per cent each of the cell employment figures was of the
total county employment.
In the example, the following would
be added to the initial cell employment figures.
SIC
20
25
30
35
Lmp
13
37
25
25
This would then make the total employment in these cells for
19b8 as follows.
SIC
20
25
30
35
Lmp
38
112
75
75
When this process is carried through for each of the
83 counties the sum of all the cells in each county will
equal the estimated county totals for 1980.
Step two is to adjust the 1968 employment in the 21
SIC categories for each county to conform to the 1980 esti
mates of total county manufacturing employment.
The same pro
cedure is followed as in adjusting the county totals.
If the
employment for the state in SIC 20 is less in 1968 than the
expected employment in 1980 the employment must be adjusted
to conform to the 1980 totals.
ally.
This is also done proportion
For example, if 15 per cent of the 1968 employment in
SIC 20 was found in Wayne County, then 15 per cent of the in
crease in SIC 20 must take place in Wayne County.
After the adjustment to correct the SIC totals is
made the county totals may be off.
Therefore, it is again
necessary to correct for the county totals, which will
140
probably throw the SIC totals off again.
This process of em
ployment adjustment may be completed as many times as neces
sary, each time reducing the amount of error between totals.
For this study these adjustments were made five times.
The
amount of error reduction with eacli adjustment is shown in
Table 12.
The greatest reauction in error occurred with the
first run, and with subsequent runs the amount of error re
duction became progressively smaller.
Since the important figures are those of county employ
ment rather than of SIC employment, the last adjustment is
made so that the sum of the 21 SIC categories for each county
will equal the projected 1980 employment.
Again, there will
probably be some error remaining in the SIC totals, but this
error can be tolerated at that point where it can not be
tolerated in the county totals.
The same procedure would then be followed for the year
2000, again using the 1968 Employment Security Commission
data as a starting point.
The result of this adjustment process would be a sec
ond and third 83 by 21 matrix of manufacturing employment
(the 21 SIC categories for each of the 83 counties).
Water Use Estimate
After the estimates of manufacturing employment were
made the next step was to develop estimates of annual per
capita water withdrawal per employee in each of the two-digit
SIC categories and to apply these to the employment estimates.
Table 12
Error Reduction
County Totals
Iteration
Cumulative
Per Cent Error
SIC Totals
Highest Individual
Per Cent Error
Cumulative
Per Cent Error
Highest Individual
Per Cent Error
1980
Base
29.67
128.55
35.27
6.71
1.37
.53
.43
2000
Base
5th run
1422.00
98.80
221.00
37.30
7.32
1.13
.21
.09
14 1
5th run
758.00
142
Source of Data
The most complete data on water use by manufacturing
establishments has been collected by the U.S. Bureau of the
Census in connection with the 1963 Census of Manufacturers.
In the census water withdrawal information was asked of all
firms.
Those which reported using more than 20 million gal
lons of water per year were asked to complete a more detailed
questionnaire on their water use characteristics.
These
findings were presented in a special report.*
Developing the Coefficient
The coefficients representing water withdrawal per em
ployee were developed by dividing the number of employees
into the amount of water withdrawn.
The resulting coeffi
cients represent the annual water withdrawal per employee.
To accomplish this, however, necessitated two further con
siderations.
Determining Employment Figures
It was not possible to utilize the employment figures
reported in Water Use in Manufacturing because these repre
sent employment only in those establishments which were large
water users.
If the water withdrawal coefficients were based
upon this employment and then ascribed to total employment it
would give undue weight to the large water users and result
*U.S. Bureau of the Census, Census of Manufacturers!
1 963. Subject Statisticss Water Use in Manufacturing.
Wash
ington t U.S. Government Printing Office, 1966.
143
in an inflated total withdrawal figure.
Since the withdrawals reported in Water Use in Manu
facturing represented virtually all of the water use in manu
facturing in 1963 it was possible to divide the total employ
ment for each SIC category reported in the 1963 Census of
Manufacturers into the total withdrawals for each SIC cate
gory as reported in Water Use in Manufacturing.
It was then
possible to apply this withdrawal coefficient to total em
ployment without fear of over representing the larger water
using industries.
Levels of Generalization
While water withdrawal data were given expressly for
Michigan for only 11 of the SIC categories these account for
the bulk of the water withdrawals in the state.
However*
the remaining types of manufacturing account for some with
drawals and must be taken into consideration if a complete
picture of manufacturing demands is to be had.
The withdrawal coefficients for seven of the remain
ing SIC categories were developed by utilizing national data.
This, essentially* results in ascribing water withdrawal
characteristics for the nation as a whole to the Michigan
situation in particular.
While this certainly results in
some misrepresentation it is felt that this is less than if
these categories were ignored.
Water withdrawals in three
categories were considered so insignificant that they were
ignored even at the national scale.
Thus it was impossible
144
to develop withdrawal coefficients for these categories.
The
resulting coefficients are presented in Table 13.
Compensatory Effect of Labor Productivity
and Water Technology
It has been recognized in a number of studies that
water use in manufacturing is more directly related to levels
of output than to numbers of employees.
However, the diffi
culty of working with physical output, because of the great
variety in types of output, has generally precluded the use
of this measure as a basis for estimating water use.
Data on value added by manufacturing are consistent
among types of products and provide a proxy for levels of
output.
fects.
However, value added suffers from two important de
The level of value added is difficult to compare from
one period to another because of inflation.
And because of
the reluctance of firms to divulge information on levels of
income it is impossible to acquire information at the county
level for value added by manufacturing.
To the extent that level of employment reflects levels
of output employment may be used as a proxy for output.
This
relationship is complicated, however, by the increasing phys
ical output per man hour as the manufacturing industry ad
vances technologically.
Increased automation has meant that
the same amount of physical output can be produced by fewer
employees.
For example, if there were 100 employees in a
base year and productivity increases three per cent per year.
145
Table 13
Water Withdrawal Coefficients8
SIC
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Water
.
Withdrawals
1963
Employment
_
16
3
148
151
3
101
177
38
163
1
19
332
6
6
3
86
29
13
Withdrawal
Coefficient
_
52,472
77,330
863,246
-
563,135
376,548
28,078
33,595
2,816
414,959
3,454
17,838
81,807
86,619
134,256
32,231
281,870
305,452
390,760
304,924 c
36,794 d
171,445 d
-
268,141 d
7,967 d
3,597,122 c
-
5,268,641
13,494,318
392,809
289,519
1,065,141
4,058,332
69,268
44,690
93,078
305,105
94,941
33,268
c
C
d
C
c
c
c
c
c
c
d
d
U.S. Bureau of the Census, Census of Manufacturers*
1963.
Subject Statistics* Water Use in Manufacturing
(Washingtons U.S. Government Printing Office, 1966).
Billion gallons
c Based on Michigan data
d Based on national data
146
five years lienee, still with 100 employees, the production
will be equal to 115 employees.
nomics,
The Office of Dusiness Lco-
U.S. Department of Commerce has estimated that labor
productivity will continue to increase at three per cent per
y ear.^
Counteracting this tendency is the process of water
conservation through reuse.
This effort toward water e co
nomy was discussed in Chapter V.
Appendix Table 18 presents
data on water recirculation for 18 of the 21 two-digit SIC
categories for 1959 and 1964.
It can be seen that there is
not a clearly demonstrated trend in water conservation in all
of the SIC categories.
Of the total categories presented,
11
of them demonstrated an increase in the reuse ratio between
1959 and 1964 while seven exhibited a decrease.
Nor is there
an established correlation between the magnitude of the coef
ficient and the change in the rate of reuse.
It was expected
that those industries which had a higher withdrawal rate per
employee would have the greatest incentive to reuse water.
This was not evident, however.
Some of the categories with
the largest withdrawal coefficients had a decrease in the
rate of reuse while categories which had a low rate of w i t h
drawal per employee often had a significant increase in the
rate of reuse.
The appears to be a concensus that there will be a
^Regional Economics Division, Office of Business Econ
omics, Preliminary Report on Economic Projections for Sel
ected Geographic Areas, 192& to 2o2t), VolT I (Washington:
Water Resources Council, 1968), p. 11-12.
147
general increase in the tendency for water reuse in the fu
ture .
The Water Resources Council has estimated that manu
facturing water recirculation will increase approximately 2.3
per cent per year between 1965 and 2000.^
For the purpose of this research these two forces are
considered to balance out, and the water use coefficients
will continue to be based upon numbers of employees*
Since
the increase in labor productivity is greater than the in
crease in water recirculation this will probably result in an
over estimation of the water needs in 1980 and 2000.
The
difficulties involved in making estimates of labor produc
tivity specifically for Michigan and in estimating recircu
lation for each SIC give further credence to accepting the
cancellation of these forces.
Application to Employment Figures
After the water withdrawal coefficients for each SIC
category were developed it was only necessary to multiply
them times the number of employees estimated for that cate
gory.
This was done for each SIC category in each county.
When summed, these water withdrawal figures total the esti
mated water withdrawal for manufacturing for each county.
These counties can then be organised into the seven planning
regions.
The resulting estimates of water withdrawals for
each region and for each type of manufacturing are presented
in Chapter I X •
^Water Resources Council, The Nation's water Resources,
p. 4-2-4.
-------------------------------
CHAPTER VII
WATER WITHDRAWALS FOR POWER GENERATION1
Importance
Tremendous quantities of electricity are required to
power the modern industrial economy of Michigan and to supply
energy for domestic consumption, and from all indications
this will continue to be a major concern to the state.
The
growth of installed generating capacity in the state for the
last 25 years is indicated in Figure 12.
It can be seen that
there have been significant increases at each five year peri
od.
ture.
These are expected to continue into the foreseeable fu
In July, 1969 Michigan had approximately 10,000,000 Kw
of installed capacity;
by 1990 this is expected to increase
to 39,300,000 Kw.
The effect of this rapidly increasing demand for elec
trical power will be felt in a number of ways, not the least
of which is in the demand for water.
At the current level of
technology approximately one-half of one gallon of water is
1Unle88 cited otherwise the material for this chapter
was obtained from conversations and correspondence with D.E.
Syler, Senior Engineer, Consumer Power Company, Jackson, Mich
igan, and from Wayne L. Wingert, “Present and Future Power
Development for the State of Michigan," Paper presented at
the Governor's Conference on Thermal Pollution, Traverse
City, Michigan, July 18, 1969.
148
149
MILLION
KW
MILLION
KW
12
10
8
1945
1950
Figure 12.
1955
1960
1965
1970
Installed Generating Capacity9
aMichigan Electric Power Capacity and Reliability A d
visory Committee, Michigan Electric Power Capacity and Reliabillty 1966-1970 (Lansing^
, pZ YTZ
150
rcquireu for the production of each kilowatt minute of elec
tricity.^ At the present time this comes to about 1,656 bil
lion gallons in 1970 and by 1990 will approach 6,507 billion
gallons.
Nature of Withdrawals
The primary purpose of water used in the generation of
electrical power is for cooling purposes.
Very little of the
water which is withdrawn for power generation
(cooling)
is
actually consumed; reliable estimates place this at less than
one tenth of one per cent.
The water which is returned to
the water source is unchanged in composition,
fication being in the addition of heat.
the only m o d i
The temperature of
the water is increased by 10° to 20° in the cooling process.
At the present time very little of the cooling water
is recycled
(cooled) and used again; most is returned as
warmer water to the source.
Conservationists are becoming
concerned about the ecological damages which may result from
the audition of warm water to the naturally cool water body,
and it is possible that if these fears are real restrictions
may be placed on water returns and more water will be cooled
and used again.
Should this happen the increased water d e
manded for power generation may not be as great as predicted.
^This is an average figure and is a combination of
nuclear and fossil fueled plants as well as plants of d i f
ferent ages.
11*1
quantity Withdrawn
As stated earlier, approximately one-half of one gal
lon of water is required for the production of one kilowatt
minute of electricity at the present time.
Going on the as
sumption that the average 24 hour production of electricity
from a plant is b3 per cent of the installed capacity, it is
relatively easy to estimate the amount of withdrawals.
For
example, in 1990 approximately 39,300 Kw of capacity will be
installed in Michigan.
Therefore,
(39,300,000) - 12,379,500 x (60) x
(1/2) x (0.63) x
(24) x
(365) « 6,506,665
million gallons per year.
it should be noted that the water requirements esti
mated by such a method, while the best possible for such a
large scale study, produce only approximate results.
Mr.
Syler cites four reasons for this:
1.
The value of one-half of one gallon per kilowatt
minute is only a "rule of thumb" figure based on
an average of our present generating facilities.
Future technologies may reduce this ratio.
2.
With an increase in the number of hydroelectric
pumped storage installations, the capacity fac
tor of steam plants requiring cooling water would
increase.
3.
An increasing number of future generating plants
may be internal combustion and hydroelectric
pumped storage plants requiring no cooling water.
Other plants may utilize a closed cycle cooling
system.
4.
The forecasts of future energy requirements are
based upon an extrapolation from present day
growth rates and could change considerably d e
pending upon the future economic growth of the
state.
152
Location of Withdrawals
because of the large amounts of water required, very
few of the rivers in Michigan are capable of supporting large
scale electrical production.
For that reason the majority of
the generating capacity has been installed on the shores of
the Great Lakes.
Approximately 90 per cent of the installed
capacity utilized water from the Great Lakes in 1969 and e x
cept for a plant in Midland County all proposed power capacity
to at least 197 4 will utilize Great Lakes water.
The location of major power plants in the state in
1970 is shown in Figure 13.
Most of the power production
takes place, as expected, in the southeastern part of the
state, coinciding most closely with Region 1.
At the present
time most of the installed capacity is fossil fueled, but in
creasingly the new capacity is turning to nuclear fuel.
The location of the 39,300,000 Kw of installed capac
ity for 1990 is not known.
It is the opinion of this writer
that this capacity will be installed along the shores of Lake
Michigan and Lake Huron in Regions IIB and IIIB.
for this belief are two.
The reasons
First, water of the Great Lakes are
cooler in the northern parts and the problem of thermal pol
lution will probably not be as great.
But more important is
the fact that a significant proportion of this capacity may
be nuclear fueled and public pressure may force the location
of these plants to more sparsely settled areas of the state
which certainly means outside of the southeastern part of
the state.
(*««.■r
i
c
o
A
•
•
I
t
L.T.u...
JmmttTtt****0™
i o w w T •w u t i i m n j
L
?L- A .
A //«•••! l a m reie«»C»rctil„
La
t»
\
1
'
^ j
I'c M M o io ” o » ew » « , AlCWM
■
I
-
1
Jip•
(
J .
1
J i
•
.1
/
*mii»e /
1
f
rA
_ >•«/
>1
' l I C M H ,l* » * it.L * l|H O L * * * "!
'•
I
W
!
MMLU
!**_
•u. <
“m 1
"“ J*"I
1 <««t J
■
,_ _
L* -i
A L L C IM p , I t M f
_
• (M il* I
pi_
I
I
,
I
_ _ i_ —
^
_
W■ • * I
:
T c“ "
---‘riifii.>
-i<~m-«
im» l
_L
[ ____ _ f ;
;*•>
••:
I
1
*
, C * T M ^ P « « M|M Ju VIMIT**,
it*«
Figure 13.
I
|M
- -U |M
,
_______
LwrlHfl NrHW
I ••!
f
^
“S*o m m
Power Plant Location, 1970*
*Michigan Electric Power Capacity and Reliability Ad*
visory Committee, Michigan Electric Power Capacity and Re
liability 1966-1970 (Lanaing, 13^6^.
.
CHAPTER VIII
AGRICULTURAL WATER WITHDRAWALS
Importance
agricultural water withdrawals can be grouped into two
main types, irrigation withdrawals and withdrawals required
for the production of livestock.
For the nation as a whole
in 1965 it has been estimated that approximately 171,877 m i l
lion gallons per day are withdrawn for municipal and domestic,
industrial, and agricultural purposes.
Of this total, agri
cultural withdrawals accounted for 111,259 million gallons
per day, about 85 per cent of the total.^ These figures are
heavily weighted by the western states where irrigation is
required for most types of agricultural production.
For the state of Michigan an almost reverse situation
is presented.
In 1955 it was estimated that municipal and
domestic, industrial, and agricultural uses of water withdrew
approximately 11,110 million gallons of water per day for the
Great Lakes region.
Of this total figure, however, only 154
million gallons per day, or about one per cent, were with
drawn for agricultural purposes, distributed almost equally
V a t e r Resources council. The Nation*a Water Resources,
p. 4-1.
154
155
between irrigation and livestock production.1
The much smaller per cent of the total withdrawal
which is accounted for by agriculture in the Great Lakes area
is a result of the more humid environment in which the area
is located.
in Michigan, average annual precipitation ranges
from 26 to 44 inches^ and is normally adequate for most of the
crops commonly grown in the state.
The agricultural irriga
tion which does occur is classified as supplementary irriga
tion and is used primarily to augment the natural precipita
tion.
as
a result, therefore, while the amount of irrigation
water applied will vary considerably from year to year, the
possibility of any agricultural production does not depend
upon continuous irrigation.
Despite the fact that agricultural production in Mich
igan does not rely completely on irrigation as do the more
arid areas of the nation, the potential for significant ag
ricultural withdrawals is present.
In 1968 there were approx
imately 12,670,007 acres of cropland and pasture-range in the
3
state.
While only a very small per cent of this total is
being irrigated at the present time, the irrigated acreage
iIbid., pp. 4-1-1 - 4-4-6.
2
Lynn C. Myers and David W. Van Meer, Michigan Sta
tistical Abstract, 1966 (Last Lansing: Bureau of Business and
Economic Research, Graduate School of Business Administration,
Michigan State University, 1966), pp. 74-83.
^Soil Conservation Service, U.S. Department of Agricul
ture , An Inventory of Michigan Soil and Water Conservation
N e e d s (Last Lansing: Agricultural Experiment Station, Michigan
State University, 1968).
156
has been increasing at a very rapid rate as more and more
farmers realize the potential which supplementary irrigation
holds for increased production.
In addition to the water used in agriculture for irri
gation,
large quantities are needed for the maintenance of
livestock.
While the numbers of animals may be decreasing in
some cases the unit withdrawal rate is not.
As a result, the
production of livestock will continue to place a significant
demand on the state's water resources.
The purpose of this chapter, therefore, is to summa
rize some of the important state-wide and regional trends
which may have an important influence on the amount of fu
ture water withdrawals for agricultural purposes,
because
of the very rapid changes which are taking place in agricul
ture in Michigan any projections of future water withdrawals
would serve more to indicate possible directional change and
not to identify probable arocunts of withdrawals.
For this
reason no attempt will be made to estimate future withdrawals
but merely to identify trends in those factors which may co n
tribute to the magnitude of further withdrawals.
band Use in Michigan
any discussion of possible future water requirements
for an area trie size of the state of Michigan would be in
complete without attention being given to major changes in
land use patterns.
This is true for two reasons.
First of
all, the major types of water use which have been discussed
157
to tiiis point are directly affected by changes in land use
throughout the state.
Larger and more concentrated popula
tions do not come about without significant alterations in
land use pat t e r n s .
Land which was once prime agricultural
land or wooded is taken over by the higher urban and suburban
land uses.
And as this occurs the level of manufacturing
activity normally rises causing a significant increase in
water withdrawals for manufacturing purposes.
For these rea
sons, land use changes can be viewed as indicative of changes
in both the amount and composition of water withdrawals.
A second reason why it is necessary to examine land
use changes is because they are directly related to water
withdrawals for agricultural purposes.
Changes in the ex
tent and character of agricultural water withdrawals are re
flected in changing land use patterns.
As the amount of land
in farms decreases, as the agricultural land under irrigation
increases, and as the extent of livestock holdings decline,
the amount of water used in agriculture will change, both in
amount and in kind.
Consequently, a discussion of agricul
tural water withdrawals without a discussion of the conse
quent lana use changes would be inadequate.
This section will proceed by first examining the a g
ricultural resource base which permits the type of agricul
tural production characteristic of Michigan.
Attention will
next be turned to the land resource history of Michigan,
looking at the different periods in Michigan's development
1‘
->8
and how these have acted to alter the land use patterns,
and last,
the current lana use pattern will be studied, giv
ing major emphasis to how this pattern is significant to the
spatial distribution of agricultural water withdrawals in the
state as exhibited in the seven planning regions.
Agricultural Resource base
The agricultural production which Michigan has real
ized since the middle 1800's has been possible because of the
bounteous agricultural resource base which the state enjoys.
As a generalization one can say that Michigan's climate,
soil,
and topographical characteristics permit a relatively high
rate of production of a number of different types of agricul
tural commodities.
This is not to say that Michigan is without equal as
an agricultural state; in 1964 there were 16 states which
ranked higher than Michigan in value of farm products sold,
one measure of agricultural importance.
Nor is it meant to
suggest that the state is uniform with respect to agricultural
potential;
in 1964 Keweenaw County, the most northerly county
in the state, had less than one million dollars in farm pro
duct sales, while Huron County registered over 35.6 million
159
dollars of farm product sales.^
The varied agricultural productivity exhibited is d i
rectly attributable to the agricultural resource base.
In
Michigan this has been composed primarily of three factors.
The spatial occurrence of these factors will be briefly intro
duced preparatory to a discussion of overall land use changes,
and changes in agricultural land use in particular, which are
responsible for agricultural water use.
Climate 2
Michigan's climate is classified as humid continental,
both long and short summer.
A humid continental climate is
one which has an average temperature of the warmest month
above 50°F and an average temperature of the coldest month
below 26.6°F.
The division between long and short summer is
based upon the average temperature of the warmest month,
above or below 71.6°F?
The dividing line between humid con
tinental long and short summer climates in Michigan falls
through the southern tier of counties, with 95 per cent or
^U.S. bureau of the Census, 1964 Census of Agriculture
reported in K.T. Wright and D.A. Caul, Michigan*s A griculture:
Its Income, Major Products, Locations, and changes (East Lan
sing : Cooperative Extension Service, Michigan &tate University,
1967), p. 34.
2
For a complete description of the varied character of
Michigan's climate see Thomas E. Neidringhaus, "A Climatography of Michigan" (unpublished Ph.D. dissertation. Department
of Geography, Michigan State University, 1964).
^Glenn Trewartha, An Introduction to Climate
McGraw-Hill Book Co., 1954), p^ 383.
(New York:
160
of the state being of the short summer type.
As far as the agricultural potential of Michigan's cli
mate is concerned,
the two most important elements are tem
perature and precipitation.
other factors are important, such
as the amount of sunshine, wind, and others, but their sig
nificance for agricultural production are far less than the
temperature and precipitation factors.
Extremes of temperatures are not as important to Mich
igan agriculture as the length of time between the last kill
ing frost in the spring and the first killing frost in the
fall.
This period, commonly known as the growing season, is
critical because it is in this period that plant growth must
take place.
The average length of the growing season for the
state of Michigan is shown in Figure 14.
it can be seen that
tne longer growing seasons occur in the southern half of the
bower Peninsula.
Growing seasons are generally shorter to
the north and in the higher elevations.
Somewhat longer
growing seasons than expected occur along the shores of bake
Michigan where the prevailing winds carry moisture-laden
winds from the water to the land.
in general, then, one can say that the southern half
of the bower Peninsula has temperature conditions which are
best suited for agriculture and that, with exceptions re
sulting from proximity to large water bodies and elevation,
the agricultural potential of Michigan temperature regimes
decreases in the northern half of the bower Peninsula and in
Figure 14.
Length of Growing Seasona
aElton B. Hill and Russell G. Mawby, Types of Farming
in Michigan (East Lansing: Agricultural Experiment Station,
Michigan State University, 1954), p. 12.
the higher elevations of the western part of the Upper Peni
1
insula.
Precipitation totals in Michigan vary from 25 to 36
inches per year and thus arc normally adequate for the type
of crops which are commonly grown in the state.
The distri
bution of the total precipitation is relatively uniform
throughout the year, with a slight concentration during the
spring and summer months.
While total precipitation amounts are adequate for the
common Michigan crops, droughts
(a period of at least two
weeks during which less than 0.25 inches of rain falls in 24
2
hours) are a frequent occurrence during the growing season.
Droughts commonly occur for periods of one to three weeks but
may be hidden within the monthly totals which may be normal
for the year.
for the farmer, however, a period of drought,
even coming within a month in which precipitation totals are
adequate, may be critical.
In summary, while droughts are a distinct handicap to
Michigan rarely are they of such severity that they preclude
any agricultural production at all; the effect is normally a
reduction in the level of production.
Thus, as a generalisa
tion, one can say that precipitation inadequacies pose no
^There is more than a month's variation in the last
freezing temperature in the spring between the southern coun
ties and the Upper Peninsula.
2
L.ll. Kidder and R.z. Wheaton, Supplemental Irrigation
in Michigan (Last Lansing: Cooperative Lxtension Service,
Michigan State University, 1958), p. 4.
major problems for Michigan farmers throughout the state, al
though some areas of the state have a greater tendency for
moisture deficiency than others
28
and
28
(see Figures 2 and 3 and pages
.
Soils*
It is very difficult to say anything meaningful about
the soil resources of Michigan without a complete description
of the soil formation process; and this is outside the scope
of interest of this work.
Nevertheless, a rudimentary under
standing of the basic differences among the major categories
of soils found in Michigan is necessary for an appreciation
of the agricultural pattern exhibited in the state.
Soils are a result of the simultaneous operation of a
number of factors: parent material, vegetation, climate,
ography, and time.
top
The several combinations in which these
elements can occur cause there to be many different kinds of
soils in Michigan, each with a somewhat different agricultural
potential.
The concern in this brief section will be only
with the broadest classification of soils in order to explain
one of the physical factors responsible for the concentration
of agricultural production
(and agricultural water use)
in
*Most of the information for this section on soils has
been taken from class notes in soil science courses (soil
classification and Mapping and Soil Origin and Classification)
at Michigan State University.
Interested readers may refer to
L.P. Whiteside, I.F. Schneider, and R.L. Cook, Soils of Michigan (bast Lansing: Agricultural Experiment Station and Cooperative Extension Service, Michigan State University, 1868).
164
the southern third of the state.
As the need arises in a
future section on irrigation selected aspects of soils may be
discussed in greater detail.
At the great soil group level there are two categories
of soils found in the state, podzols and gray-brown podzolic.^
The division line between these two categories of soils cuts
across the southern part of the Lower Peninsula,
largely c o
inciding with the traditional Bay City-Muskegon line.
The
podzols are found in the northern part of the Lower and the
Upper Peninsula, while the gray-brown podzolic soils are
found south of the Bay City-Muskegon line.
2
It should be recognized that both the podzol and
gray-brown podzolic soils are classified as zonal
(zonal soils
are those which occur on well drained uplands and reflect the
dominant influence of climate and vegetation).
Within these
broad regions there will be areas which, because of local
conditions, do not conform to either podzol or gray-brown
podzolic criteria.
As a rule, the soils found in the podzol area of the
northern two-thirds of the state are much less fertile than
^The newly developed soil classification (the Seventh
Approximation) places the podzol in the spodosol class and
the gray-brown podzolic in the alfisol class.
Since this new
classification is just starting to be used most readers will
be more familiar with the older termonology.
2
The Bay City-Muskegon line is normally considered to
separate the industrial southern part of the state from the
northern part of the state which has a preponderance of e x
tensive economic activities.
165
those found in the southern third.
For the planning regions
used in this study the podzol regions coincide most closely
with Regions llti, IIIB, XV, and V.
There are a number of rea
sons which account for the lower level of fertility.
nature of the natural vegetation is one.
The
In the northern
two-thirds of the state the natural vegetation was predomi
nately a coniferous or mixed coniferous and deciduous forest.
As the precipitation filtered through the litter from the
fallen leaves and, more importantly, the needles from the
coniferous trees, it produced a leaching material which tended
to remove much of the nutrients from the soil.
The resulting
soil was commonly acidic and was usually of a lower fertility
than the soils in the southern part of the state which d e
veloped under a leaf type of vegetation.
The relatively
sandier textured parent material found in many parts of the
northern two-thirds of the state, and the generally wetter
conditions
(a result of lower temperatures which retards
evapotranspiration)
serves to accentuate the leaching action.
These conditions are lacking or are not nearly as p ro
nounced in the southern third of the state.
Here coniferous
vegetation was not normally found, soils were less sandy,
and the leaching less intense.
southern third of the state
Thus the fertility of the
(largely Regions I, IIA, and IIIA)
has remained higher and today supports the majority of the
agricultural production in the state; in 1964 Regions I, IIA,
and IIIA accounted for slightly over 82 per cent of the total
farm product sales in the state
(with marked concentrations
in the "thumb area" and in the southeastern fruit area)
Topographic Features
2
When considered along with climate and soil conditions
topographic features do not have a major incluence on agricul
tural possibilities.
Most of the southern third of the state
has a relief which is flat to gently rolling, and poses few
problems for agricultural production.
Within this area there
may be areas which are unusually wet and in need of drainage,
and there may also be areas where local relief restricts a g
ricultural activity;
this is particularly true along the Lake
Michigan shore and in prominent moraine areas.
part, however,
For the most
topography is not a major deterrent to agri
cultural production in the southern third of the state.
In the northern two-thirds of the state, already re
stricted by a shorter growing season and generally poorer
soils, there is a greater proportion of the total land area
which lias restricted agricultural potential because of topo
graphic conditions.
These conditions are generally hilly or
rolling relief and serve to prohibit the cultivation of most
row crops.
This is particularly serious in parts of the
*K.T. Wright and D.A. Caul, Michigan's Agricultures
Its Income, Major Products, locations, and changes.
2
Information for this brief section on topography has
been condensed from notes (see footnote on page 163), per
sonal experience, and miscellaneous readings.
Interested
readers may refer to J.O. Veatch, Agricultural Land Classifi
cation and Land Types of Michigan (East Lansing: Agricultural
Experiment Station, Michigan State University, 1941).
167
northern Lower Peninsula and the western half of the Upper
Peninsula.
To summarize,
then, the southern third of the state,
relatively unincumbered either by length of growing season or
soil conditions, does not exhibit topographic conditions
which greatly restrict agricultural production.
Land Resource History
Upon this physical environment, represented by the
critical elements of climate, soil, and landforros, Michigan
has evolved from a land peopled by only Indians and an occa
sional fur trader to the modern industrial and agricultural
state which it is today.
Throughout this approximately 150
years the state has experienced a number of distinct resource
development periods which represented the then current land
use and economic structure of the state.
This section will
introduce these major periods in Michigan's development and
summarize the important land use changes which have resulted.
Major Development Periods*
In an interesting work, Perloff and Wingo suggested
that regional growth can be attributed to a region's resource
^There are innumerable references to M i c h i g a n 's his
torical development.
Unless cited to the contrary source
material for this section has been condensed from miscellane
ous readings and from two books in particular.
These are
M.M. Quaife and S. Glazer, Michigan; From Primitive Wilder
ness to Industrial Commonwealth (New Vork: Prentice-Hall, In c •,
1968), a n d W . F . Dunbar, Michigan; A History of the Wolverine
State (Grand Rapids, Michigan: William B. EercLnans Publishing
Co., 1965).
168
endowment which changes as the level of technology and the
demand for different products changes.^ To a great extent
Michigan's development can be viewed in a like manner, pro
gressing from a purely extractive economy to one dependent on
manufacturing and tertiary activity.
Focus on Extractive Industries
For the first 300 years or so after the initial pene
tration by explorers of European extraction the dominant type
of economic activity in the area which was to become Michigan
was extractive in nature.
A number of successive stages were
experienced during this period, each having a place in the
progression of land use changes to the pattern seen in 1970.
Fur Trading Per i o d .
The earliest type of economic
activity which was carried on in Michigan was the one which
had the least influence on land use.
For a period of 100
years or more, and extending under the political jurisdiction
of France, Britain, and the United States, pelts of beaver
and other fur bearing animals provided the motivating force
for tne early exploration of the Michigan area.
At one time
or another extending over the entire state, this activity had
the most minimal effect on land use.
^liarvey Perloff and bowden Wingo "Natural Resource En
dowment and Regional Economic Growth," in John Friedmann and
William Alonso, Regional Development and Planning (Cambridge,
Massachusetts: The M .I .T . P r e s s , 1964), pp. 215-239.
109
Agricultural uevclopinont in the Doutl..
With the grad
ual extinction ot the Indian land claims land in Michigan be
came open to agricultural development.
The direction of land
clearing and settlement was from the southeast, spreading west
and north.
Tne initial focus of development centered on the
southeast because the Great bakes were the main source of
entry into the Michigan area and Detroit was the main debar
kation point.
The importance of this focal point was so
great that the price of land was determined to a great e x
tent by distance from Detroit.
besides access to transportation facilities, quality
of agricultural land was the second most important determi
nant of the location of early agricultural development.
The
generally inferior quality of the soil in the northern
two-thirds of the state was recognized at an early date and
agricultural development was restricted to the southern part
of the state.
This pattern of development, concentrated in the south,
can be clearly seen by the dates of the establishment of the
counties in the state.
When Michigan achieved statehood in
1837 there were 28 organized counties.
Of these, 25 were
located in the southern four tiers of counties.
The only
northern counties to be organized were Mackinac and Chippewa
in the Upper Peninsula, later divided and reduced to their
present size.
Once started these agricultural developments in the
170
southern part of the state progressed rapidly.
By the latter
part of the 1 8 0 0 's most of the timber below the Bay CityMuskegon line had been removed and the land was in agricul
tural production.
Development in the N o r t h .
The development of the
northern two-thirds of Michigan followed a dissimilar pattern
from the southern third.
Where in the south the initial
rationale for development had, from the start, been agricul
turally motivated, the development of the north was basically
for timber production.
One of the most romantic periods in
Michigan's history, the approximately 40 years from 1860 to
1900, coincided with the most frenzied lumbering era the
state
(or nation) has ever experienced.
Obviously the lumber era had a tremendous affect on
land use in the state.
merchantable timber.
Millions of acres were cleared of
In some cases entire counties were d e
nuded of white pine and other valuable species.
Starting
first in the Saginaw Valley area, lumbering activity spread
throughout the Lower Peninsula and by the 1800's had moved
across the Straits into the Upper Peninsula.
Coincident with the extraction of the timber resources
of the northern two-thirds of the state was the problem of
the land after the timber had been removed.
Commonly known
as the cutover region, large areas in Michigan
(as well as in
northern Wisconsin and Minnesota) were put under cultivation
by farmers who were able to purchase land from timber
171
companies who had already extracted the most valuable reresource . 1 Thousands of a c r e s , already cleared, appeared to
await the plow.
Apparently unaware of the agricultural d is
advantages of the northern soils, hundreds of farmers were
only too happy to oblige.
As a consequence hundreds of farms
and thousands of acres were placed on the agricultural roles.
It soon became apparent that these northern lands had
a number of disadvantages which were not realized earlier,
and that it was impossible to compete with farming areas in
the southern part of the state with their longer growing sea
sons, better soils, and closer proximity to markets.
abandonment on all but the better land was c o m m o n .
band
Large
acreages reverted to state ownership because of non-payment
of taxes.
Barlowe estimates that by 1941 some 4.5 million
acres had so reverted, while thousands of other acres were
purchased by the federal government.
2
The agricultural pro
duction which has remained in the northern areas has been the
specialty types of production in which local conditions pr o
vide selected areas with advantages in the production of par
ticular commodities, and types of production for which a
short growing season and low quality soils are not a major
handicap.
^By 1900 the timber industry had all but disappeared
from Michigan.
2
Raleigh Barlowe, Use of Land and Water Resources in
Michigan, Report No. 52 of Project 80: Rural Michigan Now
and in 1980 (East Lansing: Agricultural Experiment Station
and Cooperative Extension Service, Michigan State University,
1966), p. 3.
172
A period of mining activity, concentrated in both iron
and copper ore, began in the Upper Peninsula in the middle of
the 1800's and maintained a relatively high rate of production
until 1920.
Since then neither copper nor iron production in
Michigan has been able to compete with the higher quality ores
found in the Southwest and in Minnesota, and mining has d e
clined to a relatively insignificant activity.
The actual
land use changes associated with mining have not been great
except in the immediate vicinity of the mines, and their re
lation to agricultural development and agricultural water
withdrawals is insignificant.
Focus on Secondary and Tertiary Activities
Michigan is no longer a state which depends on ex
tractive industries.
While there is still some lumbering and
mining which takes place, and while Michigan is still a sig
nificant agricultural producer, there are few who would clas
sify the state as anything except industrial.
This most
basic change has had a very significant affect on land use
patterns in the state and its affect on agricultural pro
duction in the state can not be over emphasized.
■
Manufacturing.
The
Michigan is well known and
role
ofthe automobile industry in
need not be recounted here.
Suf
fice it to say that the automobile industry has dominated the
Michigan economy since the
turn of the century, particularly
173
through direct employment in the industry but also in the
expansion of other inuustries related to the automobile in
dustry and in the expansion of service activities supporting
the industry employee.
The role of the automobile industry as a factor in
water withdrawals has been covered in an earlier section.
What needs to be mentioned here are the related land use
changes which have influenced the agricultural withdrawals.
The development of major concern has been the growth of urban
and suburban populations and the consequental loss of agri
cultural land.
The conversion of agricultural land to urban
and suburban use is a common experience and there is no in
dication that the end is in sight.
Continued decreases in
agricultural land will reduce the potential for agricultural
water withdrawals.
Coincident with, and indirectly related to, the growth
of the manufacturing economy in Michigan has been a reduction
of agricultural land use.
This is in addition to the reduc
tion as a result of urban and suburban growth, and is a result
of the tremendous growth in agricultural productiveness which
the United States has been able to achieve.
This subject is
presented here only for the sake of completeness and will be
discussed more fully in a later section.
Amenities♦
The most recent period which Michigan has
entered has been called the amenities era.
The focus of at
tention has turned to areas of the state which can provide
174
recreational resources.
This has not been to the exclusion
of areas of the state which continue to provide places of
residence and employment but an addition to them.
Much of
the land use in the northern two-thirds of Michigan is ori
ented exclusively for recreational purposes.
of both federal and state owned land.
This is true
And it is also true
of much of the privately owned land.1 Again, while a sig
nificant change in land use, the only direct relation which
recreational land use has to agricultural water withdrawals
is when agricultural land is withdrawn from production for
recreation use.
Agricultural Land Use
As a manifestation of the major resource development
periods exhibited in Michigan's history, there have been
major shifts in land use patterns.
The purpose of this sec
tion is to briefly examine changes in agricultural land use
which have occurred in the state preparatory to examining
several aspects of water use in agriculture.
These major
changes in agricultural land use will be considered.
are:
(1) the decreasing number of farms,
total acreage in farms, and
These
(2) the decreasing
(3) the increasing size of the
remaining farms.
1Kobert L. Vertrees, MA Survey of Nonresident Land
owners of Ten or More Acres in Antrim and Kalkaska Counties,
Michigan," (unpublished M.A. thesis. Department of Resource
Development, Michigan State University, 1967)•
175
Decreasing Number of Farms
For the nation as a whole there are fewer farms now
than there were 10 years ago, and this same trend is operat
ing in Michigan as well.
The number of farms which were in
existence during the last several census periods is indi
cated in Table 14.
It can be seen that the number of farms
has declined from each census period to the next.
Table 14
Number of Farms: State Total and
Regional Distribution®
Area
State*5
Region0
I
IIA
I IB
IIIA
IIIB
IV
V
1935
1940
1945
1950
1954
1959
196.5
187 .6
175.3
155.6
138.9
111.8
93.5
17. b
37.9
12.1
15.1
8.9
3.3
4.8
18 .6
37 .5
12.4
15.7
8.5
3.1
4.3
18.1
37 .9
11.6
15.7
9.1
3.5
4.2
18.6
38.6
11.7
15.7
8.7
3.1
3.6
18.3
39.3
11.8
16.2
8.7
2.8
3.1
18.7
40.4
10.6
17.1
8.4
2.3
2.4
18.7
40.8
10.4
17.1
8.5
2.3
2.2
1964
aCompiled from U.S. Bureau of the Census, Census of
Agriculture: 1935-1964. Statistics for States and bounties,
Michigan (Washington: U.S. Government Printing Off i c e ).
i_
Number of farms
cPer cent of state total
The number of farms expected in future years has been
tne subject of previous research efforts.
For example, in
"Project 80" an agricultural economist estimated that by 1980
176
Michigan would have approximately 4t>,000 farms.^ This esti
mate was arrived at by taking a median figure between the
projected 1959-64 decline in absolute numbers of farms and
the projected 1959-64 percentage decline in the number of
f arms.*
It is the opinion of this researcher that the rate of
decline in the future will not be as great as forecasted in
"Project 80." The reason for this is that the very marginal
farms which have a tendency to drop out first will do so b e
fore the less marginal farms.
It is believed that we have
already experienced the fastest rate of decline and that in
future years the rate of decline will be less.
In the fu
ture this will mean that the farms which remain in produc
tion at any one time will have a lesser tendency to drop out
of production than those farms which have already dropped.
The decline in number of farms in Michigan is indi
cated in Table 15.
This table indicates that the above dis
cussion may be true but at least one more census period is
need to give further evidence.
From the 1935-40 period to
^"Project BO" was a research effort by the Agricultural
Experiment and Cooperative Extension Service, College of A g
riculture, Michigan State University, and was designed to con
sider changes in rural Michigan between 1966 and 1980.
To
date, 16 separate reports have been published under the title
"Project 80."
2 Karl T. Wright, Economic Prospects of Far m e r s , Report
No. 47 of Project B0: Rural Michigan Now and in 1980 (East
Lansing: Agricultural Experiment Station and Cooperative Ex
tension Service, Michigan State University, 1966), pp. 15-16.
177
the 19 54-59 period there was an increasing rate of decline.
Tne 1959-04 period represented a decline in the number of
farms but at a slower rate of decline than during the previous
periods.
It is felt that this 19 59-64 period represents the
start of this new era.
Table 15
a
boss of Farms in Michigan
Per Cent
beeline
Per Year
Period
3.27
3.90
2.68
2.24
1.31
.91
1959-64
1954-59
1950-54
1945-50
1940-45
1935-40
U.S. bureau of the C e n s u s , Census of Agricultures
1964.
statistics for State and C o unties, Michigan (Wash
ington : U.S. Government Printing Office, 1967), p. 7.
The regional distribution of farms in Michigan since
1935 is shown in Table 14.
An examination of this spatial
distribution finds most of the farms in the southern part of
the state, and a slight tendency toward an increasing con
centration in this area.
This change has been relatively
slight, however, and the distribution is not expected to
change significantly to 2000.
band in Farms
Coincident with the decline in the number of farms has
178
been a decline in the land in farms.
beginning immediately
after World War II there started a decline in the acres of
farm land which has continued to the present time.
This
trend is indicated in Table 16.
Table 16
band in F a r m s : State Tota£ and
Kegional Distribution
Area
1935
1940
1945
1950
1954
1959
1964
State
lb.5
18 .0
18.4
17.3
16.5
14.8
13.6
Kegionc
I
IIA
I IB
IIIA
IIIB
IV
V
16.9
36.0
14.1
14.3
10.9
3.5
4.3
20 .4
43.5
17 .7
17.7
13.0
4.2
4.9
17.1
35.0
14.0
14.5
11.2
4.2
4.1
16.4
35.3
14.1
14.7
11.2
4.2
4.1
16.5
35.7
13.8
15.0
11.1
4.0
3.9
17.0
37.3
12.5
15.9
10.7
3.5
3.2
17.0
37.7
11.9
16.3
10.6
3.4
3.0
aCompiled from U.S. bureau of the Census, Census of
Agriculture: 1935-1964.
Statistics for States and Counties,
Michigan (WashingtonV U.S. Government Printing Office).
^Millions of acres
cPer cent of state total
The regional distribution of farm land is also pre
sented in Table 16.
Here, also, the dominance of the southern
part of the state is o b v i o u s .
Although there have been some
changes in the regional distribution over the years this has
been relatively slight except for the Upper Peninsula coun
ties and by no means presents a clear trend.
As would be expected, the regions with the greatest
179
proportion of the total land area in the region which is in
farms are greatest in the southern regions.
In 1964 Regions
I , IIA, and IIA had over 71 per cent of the total land area
in farms.
Farm Size
Coupled with the decreasing number of farms and total
acreage in farms has been an increasing specialization and
mechanization which has permitted the farmer to work ever
larger units of land thus realizing economies of large scale
production.
Data on average farm size for the state as a
whole and for the planning regions is presented in Table 17.
In each instance the average size has increased over the
years.
Table 17
Average Farm Size in Michigan
Area
State*3
b
Region
I
IIA
1 114
IIIA
IIIB
IV
V
a
1945
1950
1954
1959
1964
96
105
111
119
132
145
88
94
115
91
132
101
87
95
102
132
99
144
119
97
97
104
138
103
155
140
122
105
112
146
109
166
162
147
116
126
163
121
180
194
178
127
138
177
136
196
210
202
1935
1940
94
89
91
112
89
125
94
79
aCompiled fromi U.S. Bureau of the C e n s u s , Census of
Agriculture: 1935-1964.
Statistics for States and Coun
Michigan (Washington: U.S. Government Printing Office).
Acres
Ironically, farms in the Upper Peninsula
(Regions IV
and V) are larger than those in the Lower Peninsula.
This is
undoubtedly because much of the farm land in the northern
part of the state is used for pasture or some other use which
is relatively non-intensive and which permits farmers to ef
fectively manage larger units than if the operation was more
intensive.
The amount of land in farms and the amount from
which crops were harvested is shown in Table 18.
This data
appears to substantiate the above point.
Table 18
cropland Uarvested in Michigan, 1964a
Cropland Uarvested
Per Cent of Total
Area
Total Land
in Farms
Cropland .
harvested
State
13,598,992
6,738,032
49.5
2,319,544
5,125,481
1,622,469
2,219,122
1,444,895
456,650
410,831
1,360,176
2,644,513
509,709
1,451,219
525,911
122,057
124,447
58.6
51.6
31.4
65.4
36.4
26.7
30.3
Region
I
IIA
IIB
IIIA
IIIB
IV
V
aSoil Conservation Service, U.S. Department of Agri
culture , Inventory of Michigan Soil and Water Conservation
Needs
(East Lansing: Agricultural Experiment Station, Mich
igan State University, 1968).
b
Acres
181
Water Use in Agriculture
Water use in agriculture takes a number of forms.
The
use of water by the farm homestead has already been covered
in a previous chapter.
The remaining withdrawals which occur
in agriculture are primarily those for irrigation and for the
maintenance of livestock.
The following two sections will
deal with these two types of withdrawals.
Irrigation
Michigan is not a major irrigation state when compared
to other states in the nation.
However, significant water
withdrawals are made for irrigation purposes and it is prob
able that this will continue.
For this reason an understand
ing of the current irrigation situation in Michigan is nec
essary if a complete picture of water withdrawals in Michigan
is to be obtained.
Importance of Irrigation in Michigan
Agricultural areas which receive, on the average, more
than 20 inches of precipitation per year are considered to
have sufficient moisture for growing agricultural crops w ith
out irrigation.
Naturally the greatest amount of irrigation
in the United States is found in the arid western states
which have less than 20 inches of precipitation.
However,
when one examines the expansion of irrigation in recent years
a major portion has been in the humid areas of the East where
precipitation is normally adequate but where an occasional
drought necessitates supplemental irrigation for maximum
yields.
Extensive irrigation projects are a relatively new
type of agricultural development in Michigan.
According to
the U.S. Bureau of the Census there were only 5,567 acres of
cropland which were irrigated in the state in 1935,^ the ear
liest year of record.
It is probable, however, that irriga
tion in the state extends back further than 1930.
The Michi
gan Water Resources Commission reports that in 1930 there
were at least 214 irrigation systems operating in the state.
2
The irrigated acreage in the state since 1935 is shown
in Table 19.
This table shows that the adoption of the prac
tice of supplemental irrigation is largely a post World War II
phenomena.
There were approximately 3,000 acres irrigated in
the state in 1945, a decrease of almost 50 per cent from the
1935 level.
After that the growth was particularly marked.
The 1950 average was almost 14,000 acres, an increase of over
300 per cent in five years.
Since 1950 the irrigated acreage
has almost doubled every census period except for the 1959 to
1964 period in which the increase was much less.
There have been some very real reasons for the rapid
*U.S. Bureau of the Census, Census of Agriculture:
1 964. Statistics for State and Counties, Michigan (Washington: U.S. Government Printing Office, 1967), p. 7.
2
Michigan water Resources Commission, Water Use for
Irrigation; A Survey of Water Use in Agriculture and Municipal
Irrigation (Lansing, 1^59), pi TT
183
increase in the irrigated acreage.
Listed below are some of
the more important factors responsible for this t r e n d .^
Table 19
Irrigated Acreage: State To£al
and Regional Distribution
Area
1935
1940
1945
1950
1954
1959
1964
State
5.6
3.0
2.9
14.0
23.5
40.2
49.0
19.9
53.2
6.0
20.0
.8
.0
.3
13.1
78.7
3.8
3.7
.0
.0
.6
18.8
69.3
4.4
6.8
.1
.6
.0
9.3
69.1
13.5
2.6
3.0
.7
1.9
11.6
65.1
10.3
8.2
2.2
1.2
1.4
11.3
67.5
11.7
5.7
2.1
.9
.9
11.1
70.0
10.5
5.4
1.4
.7
1.4
Region
I
IIA
IIB
IIIA
I IIB
IV
V
c
Compiled from U.S. Bureau of the C e n s u s , Census of
Agriculture: 1935-1964. Statistics for State and C o u n t i e s ,
Michigan (Washington: U.S. Government Printing Office).
^Thousands of acres
cPer cent of state total
1.
Probably the most basic reason is the desire of
farmers to eliminate weather risks in the production of farm
products.
This has been felt most strongly in the production
of crops with a high market value potential and for which a
steady supply of moisture is necessary for optimum production.
Roy Huffman, Irrigation Development and Public Water
Policy (New York: Ronald Press, 1953), and Max M. *Pharp and
C.w. Crickman, "Supplemental Irrigation in Humid Regions," in
Water: 1955 Yearbook of Agriculture (Washington: U.S. Government Printing Office, l4!>5) .
184
2.
A second major contributor has been the reduced
costs associated with the improved irrigation equipment.
This is related primarily to the release o£ war-time develop
ed aluminum for other than military uses.
proven to be light in weight, durable
cial handling care), and lower priced.
Aluminum pipe has
(thus requiring no spe
The development of
quick-coupling devices has made sprinkling irrigation much
more portable than before, allowing a minumum amount of total
equipment to be used on a larger total a r e a .
And improved
pumps and motors have increased the efficiency of the irriga
tion system.
3.
The spread of information on the moisture require
ments of crops and the moisture characteristics of soil has
the farm operator aware of the potential which irrigation
o f fers.
4.
High product prices and resulting farm income im
mediately after world War II released capital for investment
purposes.
5.
The increase in rural electrification permitted
the expansion of irrigation systems and made them much more
portable thus greatly increasing their effectiveness.
Nature of Irrigation in Humid Areas
Irrigation in a humid area such as Michigan normally
follows a completely different pattern than irrigation in
arid lands.
Although most of the following statements are
addressed to irrigation in humid areas in common they apply
equally well
to M i c h i g a n
in p a r t i c u l a r .
Need for Irrigation
The principle feature of the agricultural climate in
humid areas which requires that some type of special atten
tion be given to matters of precipitation and moisture is the
frequency of short, but severe, periods of drought which o c
cur during the growing season.
These drought periods are very
prevelent in Michigan; during one ten year period in Michigan
there were seven periods of one to two weeks in length in
which there was no rainfall registered.
Few, however, extend
beyond the third w e e k .^
If we consider a drought to be at least a two week pe
riod in which less than 0.25 inches of rain falls during any
24 hour period then the occurrence of droughty periods are of
longer duration.
Kidder and Wheaton found that, for a number
of locations in the central part of the Lower Peninsula, pe
riods of drought ranging from 2 to 15 weeks were not uncommon.
Records over a 20 year period indicate that a dry spell of at
least a month can be expected every two years.
2
Droughts of from one to three weeks may or may not be
fatal to agricultural crops, but will normally have a delete
rious affect upon the quantity and quality of yields.
Four
^Orson W. Israelsen, Irrigation Principles and Prac
tices (New York: John Wiley and Sons, Inc7, 1950), p. 348.
2
Kidder and Wheaton, Supplemental Irrigation in Michi
g a n , p. 4.
X86
basic measures have been suggested to counteract drought in
humid areas:
1) increase the storage capacity of the soil
for moisture,
2) reduce the water needs of the crops by using
plants which require less water or by using farming practices
which increase the efficiency of water use, 3) adjust the
planting period so that the critical periods of growth will
occur before the normal drought season, 4) make up the water
deficiency by supplemental irrigation.* The latter measure
has been the one to receive the greatest emphasis in Michigan
as in the other humid areas of the nation.
booking at the basic problem of uneven distribution of
natural precipitation in historical perspective, there have
been two basic approaches to the problem, one emphasizing the
removal of excess water during periods of excessive precipi
tation and the other the gradual at first and then rapid
adoption of irrigation measures.
In chronological order the
agricultural drainage programs predated the irrigation pro
grams by several years.
In the 1960 Census of Agricultural
Drainage Michigan ranked third among the states in terms of
2
land area in drainage projects.
The second response to uneven distribution of natural
precipitation has been through irrigation.
In the humid
^Huffman, Irrigation Development and Public Water
Policy, p. 239.
2
Raleigh Darlowe, Implications of Land and Water De
velopments in Michigan for^public Water Resource Policy (Lan
sing: Michigan Water Resources Commission, 1966), p. 2T.
187
parts of the nation this has taken the form of supplemental
irrigation.
Supplemental irrigation is based upon the fact
that even though average annual precipitation
36 inches in Michigan)
(between 25 and
is generally adequate for the crops
which can be grown under the normal temperature conditions
the available natural moisture varies considerably, both from
year to year and even within any one year.
Therefore, natural
precipitation can not guarantee an adequate amount of moisture
throughout the entire growing cycle of the crops.
If this
lack of moisture occurs during a critical growth period yields
may be drastically reduced.
For this reason it has become
increasingly profitable for farmers in the humid part of the
nation to add irrigation capacity to the capital improvements
of the farming enterprise.
In Michigan and other humid areas this has taken the
form of supplemental irrigation.
Supplemental irrigation
means that natural precipitation supplies the bulk of the
moisture requirements for crops and any deficit is made up
by the application of water through irrigation.
This is a
considerably different situation than that evidenced in the
more arid parts of the nation where almost all types of cul
tivation require irrigation.
In arid areas it is often nat
ural precipitation which is supplemental and not irrigation.
Types of Irrigation
Three general methods of irrigation have been devel
oped.
These are surface, sprinkler, and subirrigation.
In
1118
surface irrigation the water is conveyed to the plants by
means of furrows through which the water flows and seeps into
the ground where it is used by the plant.
Sprinkler irriga
tion applies water to the crops by means of pipes and sprin
klers which allow the water to fall upon the plants and soil
and to then seep into the soil.
In subirrigation the water
table is kept sufficiently high so that the root system of
the crops is kept constantly moist.
In Michigan the majority
of the irrigation is of the sprinkler type.
Advantages of Sprinkler Irrigation.
The use of sprin
kler irrigation has a number of advantages over other types
of irrigation.
1.
These are summarized from a couple of sources.^
There is complete control of water at all times,
permitting the irrigation of land which precludes the use of
surface irrigation techniques because of possible soil ero
sion.
2.
It is possible to apply water uniformally on all
types of soils.
Different textured soils have different
rates of permeability and with a sprinkler system this can
easily be incorporated into the irrigation scheme.
3.
Marginal control of water applications is possible.
If the exact water needs of the crops can be assessed with
^Tyler H. Quackenbush and Dell G. Shockley, "The Use
of Sprinklers for Irrigation," in Water: 1955 Yearbook of A g
riculture (Washington: U.S. Government Printing Office, 195$),
p7 267, Tind Huffman, Irrigation Development and Public Water
Policy, pp. 237-238.
189
some accuracy these needs can be met exactly.
4.
With the use of sprinkling irrigation expensive
land preparation is not required.
Leveling or grading may
be costly or virtually impossible because of the shallowness
of the soil.
These problems are eliminated with irrigation
done by sprinkling.
5.
More land can be cropped because land which would
be occupied by ditches in surface irrigation can be put into
production when sprinkling is used.
6.
Sprinkling irrigation permits the use of water
sources of smaller quantities than does surface irrigation.
7.
The application of fertilizers can be strictly
controlled when applied through a sprinkler irrigation sys
tem.
8.
Labor costs are substantially reduced with sprin
kler irrigation.
There are no ditches or ridges to be main
tained, relatively inexperienced labor may be used, and irri
gation can be worked into the normal scheduling of farm work
so that it requires special attention only once or twice a
day.
9.
Potential frost damage is reduced by using sprin
kler irrigation systems.
10.
Less water is required than in surface irrigation.
This permits either a larger amount of land to be irrigated
or less cost to be incurred on a fixed quantity of land.
190
Disadvantages of Sprinkler Irrigation.
Despite these
distinct advantages there are a number of very real disadvan
tages to the use of sprinkler irrigation.
Some of the more
critical of these are listed b e l o w.^
1.
Excessive winds may so disrupt the sprinkler pat
terns that certain areas receive too much water while others
receive too little.
2.
Water, to be used in a sprinkler system, must be
free of sand and other material which may clog the jets.
3.
A constant supply of water is necessary if the
equipment is to be used to the maximum.
4.
The initial investment in pipes, pumps, sprinkler
heads, etc. is high and sufficient investment capital is nec
essary before a system can be installed.
5.
Power requirements are relatively high and can be
quite costly.
6.
Very fine textured soils do not permit rapid in
filtration and the fields may remain muddy for some time after
the application.
When the rate of application is reduced to
prevent surface collection the loss by wind drift and evapo
ration is often excessive.
Despite these disadvantages sprinkler irrigation r e
mains the dominant type of irrigation in Michigan; virtually
Quackenbush and Shockley, "The Use of Sprinklers for
Irrigation,"p. 267, and Huffman, Irrigation Development and
Public Water Policy, pp. 237-238.
191
100 per cent of the irrigation systems in the state are of
this type.
Benefits and Costs of Supplemental Irrigation
The irrigation of agricultural crops is a rather ex
pensive practice.
Equipment purchases, installations, and
operation are costly.
Whether or not there will be an ex
pansion of irrigated acreage in Ilichigan will depend primar
ily upon the question of whether or not the increased pro
duction as a result of irrigation will raise farm incomes
sufficiently to cover the cost of irrigation.
In this con
sideration there are concerns of both advantages and costs
of irrigation.
Benefits.
There are a number of benefits to the farm
ing operation which have been claimed for irrigation, not all
of them directly related to increasing yields through the
addition of moisture.
Benefits commonly cited are the follow
ing .^
1.
Safeguards against droughts that may permanently
damage or destroy agricultural crops.
2.
Increased yields made possible by the addition of
moisture during critical growing periods.
3.
The production of higher quality products which
^Tharp and Crickman, "Supplemental Irrigation in Humid
Regions," pp. 256-258, and Huffman, Irrigation Development
and Public Water P o l i c y , p. 246.
192
can command a greater price in the market.
4.
Earlier maturation which permits access to the
early markets.
5.
The maintenance of quality pasture conditions d ur
ing dry periods.
6.
An aid in the application of commercial fertili-
zers, insecticides, and herbicides.
7.
Protection against frost damage.
8.
Farmstead fire protection.
Costs.
On the deficit side of the ledger there are
costs which must be incurred in establishing and maintaining
an irrigation system.
ber of factors.
These will vary directly with a num
Some of the more important are listed be
low. 1
1.
The amount of land which is being irrigated.
2.
The basic outlay for equipment.
3.
The costs of maintenance and operation.
4.
The costs involved in transferring the water from
the source to the fields
(normally a function of such things
as nearness to the source, amount of lift required, and costs
of power.
5.
The type of system used.
^Tharp and Crickman,
Ibid.
193
Spatial Distribution of Irrigation in Michigan
More important for this study than the total amount of
irrigation is the spatial distribution.
This will be d is
cussed under the regional framework set up for this study and
under the more germane regions based upon the actual location
of irrigation.
Regional Distribution
Irrigated acreage in Michigan is not distributed
equally throughout the sta t e , but is concentrated in a rela
tively few areas.
Tabulated according to planning regions as
used in this study, the distribution of irrigated acreage is
presented in Table 19
(page 183)-
It can be seen that Region
IIA has been able to maintain between 65 and 70 per cent of
the total state irrigated acreage for the last 25 years.
The
other two dominant regions are Region 1 and Region IIB, each
with approximately 10 per cent of the state total.
The re
maining regions had less than nine per cent of the state
total in 1964.
There does not appear to be any clearly
exhibited pattern of change in any of the regions;
in fact,
the change has been quite erratic.
Distribution by Irrigation Districts
Tabulation by planning regions does not accurately
present the complete picture of irrigated agriculture in Mich
igan.
Certain types of crops tend to be irrigated much more
than others.
Since these crops tend to be grown in certain
194
parts of the state more than in others there is a distinct
concentration of irrigation in selected areas.
Irrigated C r o p s .
According to a 1958 study conducted
by the Michigan Water Resources Commission four types of
crops account for the bulk of the irrigated acreage in Michi
gan.1 These are: truck crops, small fruits and tree fruits,
potatoes, and nurseries.
These crops tend to be those which
can command a relatively high price in the market and for
which an adequate supply of moisture is necessary for quality
production.
Certain areas of the state tend to concentrate on the
production of these crops more than others.
The major pro
ducing counties in the state in 1964 for each of these four
types of crops are shown in Figure 15.
Irrigation Reg i o n s .
A map of the major irrigated
areas in the state is presented in Figure 16.
This shows a
very close correlation between the extent of irrigation and
the acreage in the four categories of irrigated c r o p s .
main areas stand out on this map.
Two
The southwestern counties
represent the major concentration of irrigated acreage, cor
responding most closely with Region IIA of the planning re
gions.
Irrigation in the southwest is based primarily on the
production of fruits
(both small fruits and tree fruits) and
^Michigan Water Resources Commission, Water Use for
Irrigation: A Survey of Water Use in Agriculture and Munici
pal Irrigation.
Figure 15.
Major Irrigated Crops, 1964a
aK.T. Wright and D.A. Caul, M i c h i g a n ’s Agriculture;
Its Income, Major Products, Locations, and changes (East
Lansing: Cooperative Extension Service, Michigan State Uni
versity, 1967) .
/M U M K l
MICOfT* (■•*•<*■*■*
Figure 16.
Irrigated Acreage, 1964a
aU .S . Bureau of the C e n s u s , Census of Agriculture;
1 9 6 4 . Statistics for State and C o u n t i e s , Michigan (Washing*
ton: U.S. Government Printing Office, 1967), pp. 258-265.
197
truck crops.
A major area of irrigated potatoes is centered
on Montcalm County.
In the southeast the concentration of
irrigation is not nearly as great; only one county is in the
top ten counties based upon irrigated acreage.
Truck crops,
nursery products, and, to a lesser extent, potatoes are the
basis for this center of irrigation.
Prospects for Future Irrigation in Michigan
The question of how much water will be needed in Mich
igan in future years to satisfy the demands of agricultural
irrigation is the major focus of this chapter.
There are so
many factors which are involved that it is difficult to iden
tify a directional trend and certainly impossible to indicate
an amount of water which may be withdrawn.
The purpose of
this section is to identify some of the more significant fac
tors which combined will play an important role in determin
ing future withdrawals.
As a general rule there are five main factors which
combine to determine the amount of irrigation water which will
be applied to agricultural crops in any given year.
These
are: 1) the nature of the natural precipitation and tempera
ture in the area, 2) the nature of the soil
(such elements
as slope, texture, waterholding capacity, etc.), 3) the nature
and water requirements of the particular crops that are being
irrigated, 4) management decisions as to the need for irriga
tion, the cost and availability of water, marginal returns,
etc., and 5) the amount of land being irrigated.
198
The nature of the natural precipitation and tempera
ture is given and is beyond the manipulation of man.
That
there will be periods of drought in the future is certain;
the question which will determine the amount of water needed
for irrigation will be the willingness of farmers to adopt
the appropriate irrigation practices.
Droughts will occur on all types of soils; however,
not all types of soils will receive irrigation.
As suggested
in an earlier chapter it is less the level of precipitation
and temperature than the water in the soil which determines
when irrigation is needed.
Depending on a number of factors
some soils, under the same precipitation and temperature con
ditions, will require irrigation sooner and more often than
1
2
others.
According to Kidder, soil texture
is one of the most
important criteria; virtually 100 per cent of the irrigated
acreage in Michigan occurs on soil with a sandy, loamy sand,
or sandy loam texture.
It is Dr. Kidder's feeling that these
three textured soils will continue to be the most heavily ir
rigated .3
It is doubtful if there will be much change in the
basic water requirements of irrigated c r o p s .
The amount of
*Dr. Kidder is a professor in the Department of Agri
cultural Engineering, Michigan State University, and is the
department's expert in the field of irrigation.
2
Soil texture refers to the relative size of the soil
particles.
^Conversation with Dr. Kidder, July, 1969.
water necessary for optimum growth is largely determined by
the physiological conditions of the plants*
However, this
amount of water is greater for some crops than others.
For
example, the average moisture use rate in inches per day for
celery is 0.20 while for potatoes it is 0.30.* While the
basic amount of water needed for the different crops will
probably not change to a great extent, changes in the compo
sition of the crops grown could have a significant effect.
Management decisions may have a very significant im
pact on the total water withdrawn.
Ideally the application
of irrigation water would be governed by the moisture c on
tent of the soil.
When the moisture content dropped below
a critical level a measured amount of water would be applied
to bring the moisture content back to tne optimum.
ideal is seldom achieved in Michigan.
This
Few irrigators are
properly equipped or so inclined to measure changes in soil
moisture content and to apply irrigation water in direct
response.
In actual practice water is often applied on
schedule and in set amounts regardless of the soil moisture
content.2
Probably there is a no more critical element in de
termining the amount of water used in agricultural irrigation
Tentative Irrigation Guide for the Design of Sprin
kler Irrigation Systems in Michigan" (College of Agriculture,
Michigan State University and Soil Conservation Service, U.S.
Department of Agriculture, 1954), (Mimeographed).
2
Dr. Kidder suggests that this will probably average
about eight inches per acre per year.
than the amount of land under irrigation.
If one knew the
amount of land under irrigation at the present time, or pro
jections for some future date, he would be well on the way to
estimating the amount of irrigation water needed.
However,
even the estimation of future irrigated acreage is fraught
with difficulties.
Aside from the problem of the rapid change
in agriculture as a whole and the problems which this presents
for making projections, the projection of irrigated acreage
faces some problems unique to irrigation.
One of the major problems which one confronts in pro
jecting the future of irrigation in Michigan, or in any humid
area, is the incidence of drought.
The adoption of irrigation
practices may be accelerated considerably as a result of cli
mate conditions.
Irrigation systems are often purchased and
installed based upon inadequate moisture conditions during a
given year or immediately preceeding years.
Once the initial
investment is made in equipment these systems may be operated
during periods when the natural moisture conditions, while
inadequate, are not critical and would not, in themselves,
cause the installation of an irrigation system.
Thus there
is a tendency to irrigate more acreage in subsequent years
than would have been irrigated if there had not been earlier
periods of drought.
Therefore, in any given year, the number
of irrigation systems in operation and the amount of acreage
under irrigation may be related as much to natural conditions
outside the realm of human manipulation as to purposely
201
planned management practices.
A second problem which must be faced when assessing
the possible future amount of irrigated acreage which may be
expected in the future is the fact that with the use of port
able sprinkler types of irrigation systems much larger areas
of land may be potentially irrigable than is presently under
irrigation.
Depending on the particular year there may be
several times as much land under irrigation than at other
times.
This will probably become more important as the irri
gation systems become more portable.
There are a number of factors which may combine to re
strict the future expansion of irrigation in Michigan.
Two
of the more important of these are related to the marginal
returns which can be expected from the addition of a supple
mental irrigation capacity, and the relative productivity of
water used in competing p u rposes.
Whether or not any farmer
will expand his irrigation potential will depend on his
assessment of the marginal returns and the cost involved in
producing them.
Should product prices be at a high enough
level to more than compensate for the cost incurred it is
very probable that the acreage under irrigation would in
crease.
If opposite conditions prevail substantial expansion
of irrigation is unlikely.
water used in irrigation is a very consumptive use in
that water is used only once and there is almost no possi
bility of reuse.
On 'the otherhand, water is commonly used
two or more times in the manufacturing process.
This is es-
pecialiy important because water used for irrigation is very
uneconomical when compared to industry.
Even for high value
crops the marginal returns per gallon of water used in agri
culture can not begin to compete with the marginal return of
that same gallon used in manufacturing.
It has been esti
mated that the ratio of value added by manufacturing to value
added by agriculture was about 140 to 1 for a given quantity
of water u s e d .^
Because of this, conflicts arising from competition
between industry and agriculture for limited amounts of water
will be examined carefully to insure that the benefits to be
gained from allocating limited quantities of water to various
uses are maximized.
It is becoming increasingly clear that
water resources must be plentiful to allow them to be used
profitably in agriculture.
In addition, it may be necessary to restrict the
amount of water withdrawn from a water source for irrigation
so that the demands for water for domestic purposes, pollu
tion abatement, etc. will remain adequate.
This point has
not yet been reached in Michigan, but as the competition for
water becomes more intense this possibility can not be over
looked .
Despite these several hindrances to the future
^Sheppard D. Powell, "Relative Economic Returns from
Industrial and Agricultural Water Uses," Journal of the American Water Works Association, Vol. 48, No.“ 5 (August, 1*)56) ,
p. 991.----------------------
203
expansion of irrigation in Michigan,
it is undoubtedly true
that the state has not yet reached the end of the growth pe
riod and that there will continue to be an expansion of irri
gation; the data in Table 19
(page 183) indicates an increas
ing, although at a slower rate, of irrigated acreage in the
state.
No estimates were found of future irrigated acreage
nor does this writer wish to make any; the myriad of physical,
economical, and political influences would put any projection
in only the realm of the possible.
As irrigated acreage does continue to expand it will
probably continue to expand in the same general areas where
it is now found in the greatest concentration, namely in the
southwest, west central, and southeast parts of the state.
It is in these parts of the state where the crops capable of
producing the greatest return from irrigation have found the
most favorable environment.
Livestock
In addition to water used to produce irrigated crops,
the other major agricultural withdrawal is for the production
and maintenance of livestock.
This is a most important ag
ricultural use of water and must be recognized in any water
demand s t u d y .
The total amount of water which is used for livestock
purposes is directly related to the number of animal units
and the water requirements per unit.
Number of Animal Units
In terms of total number of livestock there is no
clearcut increase or decrease in numbers of livestock in the
state.
The numbers of three major types of stock on hand in
the state during several census periods are shown in Table 20
Sheep appear to be the only type of stock which exhibit any
degree of consistency in trend and even here the trend is dis
continuous in terms of direction of c h a n g e .
Of the other two
hogs show the greatest consistency in direction of change,
while cattle do not appear to exhibit any degree of consis
tency.
In terms of other criteria than numbers of livestock
the state does present a very strong directional change and
this is toward a growing specialization of production with
fewer farms producing larger quantities of one or a few types
of stock.
This is pointed out quite clearly in Table 21.
Looking at the data on number of farms in the state
with any of the three types of stock, in each case there has
been a decline in the number of producing farms, indicating
a concentration of production in the hands of a few producers
One might counter this point by noting that the num
ber of total farms has been decreasing and thus one would e x
pect that the number of farms producing any given type of
stock would also be decreasing.
This is true but as shown in
Table 21 there has been a constant movement toward a smaller
and smaller per cent of the total farms producing each type
205
T a b l e 20
Number of Livestock: State T£tal
and Regional Distribution
Area
1940
1945
1950
1954
1959
1964
State
Cattle
Sheep
Hogs
1,541
857
58b
1,955
663
681
1,696
388
690
1,839
465
789
1,610
459
968
1,725
323
716
Region Ic
Cattle
Sheep
Hogs
19.1
27 .0
23.0
18.8
28.9
22.3
18.0
28.9
23.6
18.0
29.2
19.6
19.5
37.7
16.7
19.2
32.9
15.2
Region IIA
Cattle
Sheep
Hogs
35.0
44.4
46.2
34.9
45.3
49.4
37 .0
48.5
51.5
37.1
45.7
56.3
38.5
59.7
62.2
37 .9
46.3
65.2
Region IIBC
Cattle
Sheep
Hogs
12.1
4.2
6.5
12.1
3.6
6.1
11.8
4.0
7.3
11.5
6.5
7.0
10.2
7.3
5.0
10.0
5.0
3.7
Region IIIAC
Cattle
Sheep
Hogs
16. 5
11.0
15.8
16.4
10.5
14.9
15.7
9.0
10.4
15.6
8.2
10.2
15.4
10.2
10.1
15.6
7.1
10.6
Region IIIBC
Cattle
Sheep
Hogs
10.2
12.6
6.6
11.1
10.7
5.9
10.8
8.7
5.9
11.2
9.4
6.2
10.7
11.6
5.5
12.0
8.0
5.0
Region IVC
Cattle
Sheep
Hogs
3.5
.3
1.1
3.4
.3
.9
3.5
.3
.9
3.5
.4
.5
3.2
.3
.3
3.0
.1
.2
Region V c
Cattle
Sheep
Hogs
3.7
.5
.7
3.4
.6
.7
3.2
.5
.5
3.2
.7
.3
2.5
.9
.2
2.3
.6
.2
aCompiled from U.S. Bureau of the Census, Census of A g
riculture: 1940-1964. Statistics for State and C o u n t i e s , M ich
igan (Washington: uTs. Government Printing Off i c e ) .
Thousands of animals
Per Cent of State Total
Table 21
Livestock Specialization in Michigan3
Characteristic
1940
1945
1950
1954
1959
1964
114,154
73
15
96,741
70
19
67,435
60
24
54,150
58
32
9,180
6
42
8,505
6
55
7,310
7
63
5,282
6
61
56,023
36
12
43,256
31
18
34,257
31
28
18,266
20
39
Cattle
Farms with Cattle
Per Cent of Total
Ave. No. Per Farm
148,663
79
10
136,030
78
14
206
Sheep
Farms with Sheep
Per Cent of Total
Ave. No. Per Farm
Farms with Hogs
Per Cent of Total
Ave. No. Per Farm
25,043
13
34
16,689
10
40
88,768
47
7
Hogs
70,303
40
10
aCompiled from U.S. Bureau of the Census, Census of Agriculture: 1940-1964. Sta
tistics for State and Counties, Michigan (Washington: U.S. Government Printing Office).
207
of stock.
This shows that not only is production becoming
more concentrated because of a fewer number of farms but that
it is also becoming increasingly concentrated in the smaller
number of farms which remain at each census.
One of the most outward manifestations of this grow
ing specialization is the average number of animals per farm.
This data is also presented in Table 21, and it is very clear
that the number of animals per farm is increasing and will
probably continue to increase into the future.
It is this
reason which causes the total number of livestock not to show
any clear-cut decline over the years.
Spatial Distribution of Animal Units
The distribution of livestock production in Michigan
is much less concentrated than is the distribution of irri
gated acreage, although it still conforms to the overall ag
ricultural picture of the state with the majority of produc
tion concentrated in the southern third.*
The regional distribution of the three main types of
livestock for the past several census periods is shown in
Table 20.
It can be seen that there has been very little
change in the spatial distribution of livestock production in
the state over the years.
Based upon this it is probable
that there will be very little change in the distribution of
^In 1964 over 81 per cent of the total sales from
cattle, sheep, and hogs was in Regions I, IIA, and IIA which
are the regions which comprise the southern third of the state.
208
production in the next 30 years.
A map of the sales of cattle, sheep, and hog products
in Michigan in 1964 is presented in Figure 16.
This further
shows the concentration in the southern third of the state
and also areas of concentration within that area.
Water Use in Livestock Production
Water is used in the production of livestock in a num
ber of ways.
The most obvious way in which water is used is
for the body maintenance of the animal.
In this use different
types of animals require different amounts of water.
Listed
below are the water requirements for a number of livestock
1
types.
Dairy cow
3.3 gallons per pound of milk
Beef cow
10 gallons per head per day
Hogs
10 gallons per head per day
Sheep
1 gallon per head per day
Chickens
5 gallons per day per 100 birds
The above figures are only averages and will vary depending on
the size of the animal, climate, and other factors.
Partly as a result of the general trend toward larger
and more mechanized farming operations and partly as a means
of maintaining higher sanitary standards and a better quality
product, there has been a gradual adoption of new agricultural
1 Interview with Dr. Harlan Ritchey, Department of Animan Husbandry, Michigan State University, July, 1969.
Figure 17.
Major Cattle, Hog. and Sheep
Producing Counties
aK.T. Wright and D.A. Caul, Michigan*s Agriculture:
Its Income, Major Products, Locations, and Changes (East
Lansing: Cooperative Extension Service, Michigan State Uni
versity, 1967), p. 43.
210
equipment and processes.
Many of these have resulted in a
large increase in the amount of water used per livestock
unit.
Listed below are some of the major livestock related
uses of water which might be found in Michigan today.^ By
2000 there will undoubtedly be a number of additions to the
list.
Washing animals
hosing floors and ramps
Cleaning milking equipment
Refrigeration
Climate control
Liquid manure handling
Automatic waterers
Egg washing
Dressing and processing
C.K. Kline, "Water Systems Analysis to Meet Changing
Conditions" (Agricultural Engineering Department and Coopera
tive Extension Service, Michigan State University, no d a t e ) .
CHAPTER IX
SUMMARY, FINDINGS, AND RECOMMENDATIONS
Summary
Water availability will not be the most critical ele
ment in Michigan's future development.
Although no attempt
has been made to examine the water supply/demand relation
ships of the state, it is believed that, at least in the
foreseeable future, Michigan's bounteous water resources will
be sufficient to satisfy demands made upon them.
true for water quantity only;
This is
the availability of water oJ a
particular quality is another matter and was not an issue in
this study.
While it will probably not be necessary to import
water into the state the demand for water will continue to
expand and will necessitate positive action on the part of
the resource planners of the state if this demand is to be
met in a satisfactory manner.
Increased demand will most
certainly necessitate the construction of new water distri
bution facilities.
In some cases it may require that special
legal, political, and economic regulations be devised to in
sure that future water resource action in the state proceeds
in an orderly and efficient manner.
211
212
The purpose of this study is to identify what the fu
ture demands for water in Michigan may be.
In this accord
emphasis was placed on three aspects of expected demand.
First was the estimation of overall demand in the
state.
There have been a number of studies which have identi
fied the important factors in Michigan's water demand and
have rightfully concluded that these demands will increase in
the future.
What these studies have not done is to make a
quantitative estimate of future withdrawals.
Separate studies
have been made for particular areas of the state, and for
particular types of water uses, but none have made estimates
of future water demand for the state as a whole.
It is be
lieved that this study will be a contribution to this end.
The second focus of the study was on the types of
water use.
Estimates of total withdrawal are useful and pro
vide the planner with valuable information.
However, much
more useful are estimates which are broken down into differ
ent types of uses.
of purposes.
Water withdrawals are made for a number
For this research four uses were considered.
These were domestic, municipal, and commercial withdrawals,
withdrawals for power generation, and manufacturing and ag
ricultural withdrawals.
Specifically excluded from the study
were water uses for non-withdrawal purposes such as for rec
reation and transportation.
The third emphasis was placed on identifying the spa
tial occurrence of expected withdrawals.
Increased amounts of
213
water will not only be demanded in the state in the future,
but will be demanded in specific areas of the state.
Many
times this spatial aspect of withdrawals is equally important,
or more so, than the total amount of withdrawals.
Citing
only the total amount of withdrawal and ignoring where these
withdrawals occur is analogous to saying that the river has
an average depth of one foot, ignoring major holes.
Conse-
quently, this study has attempted to place the demand esti
mates within a spatial framework.
In some instances it was impossible to accomplish all
of these goals.
withdrawals.
This was particularly true of agricultural
The causal factors of agricultural withdrawals
are identifiable but their rapidity of change places any
estimates of withdrawal in the realm of the possible which
certainly would not allow a great deal of confidence to be
placed in them.
For this reason it was necessary to be con
tent with a discussion of directional changes in those fac
tors responsible for agricultural withdrawals and to forego
any specific estimates of withdrawals.
It was also impossible to place withdrawal estimates
for power generation in a regional framework.
With contin
ued technological developments the mobility of power has
increased tremendously.
Thus it is now possible to transmit
electrical powe: • over relatively long distances, greatly re
ducing the pull to any one specific location.
Therefore,
estimates of water requirements for power generation were
214
made for the state as a whole.
Suggestions are made as to
the general area of occurrence, but no attempt has been made
to ascribe parts of the total withdrawal to specific regions.
Findings
The demands on Michigan's water resources have been
large in the past and will continue to increase at least to
the year 2000.
The trend of water withdrawals in the state
for selected uses is presented in Table 22.
It is obvious
that the problem of adequately meeting the state's demand for
water will be increasingly important in the future.
Table 22
Summary of Michigan's Water Withdrawalsa
Type of Use
Domestic
Manufacturing
d
Power
r
1940
1950
1960
1980
2000
162
198
256
352
498
275
345
775
1,005
1,049
331
538
1,151
4,700
6,507
aBillions of gallons per year.
u
Adjusted for increased per capita consumption and for
differences in municipal and non-municipal withdrawals.
In
cludes domestic, municipal, and commercial withdrawals.
cBased on a per employee withdrawal rate.
^Computed using a withdrawal coefficient of one-half
of one gallon per kilowatt minute of electricity produced.
eFor 1990.
The following is a discussion of some of the findings
215
in the major areas of withdrawal selected for examination in
this study.
Domestic, Municipal, and
Commercial Withdrawals
Water withdrawal for domestic, municipal, and commer
cial purposes has been a major contributor to water withdraw
als in the past, and from all indications will continue to be
an important use of water in the future.
As technology devel
ops, and as higher incomes give access to, larger numbers of
water using devices, and as the trend toward urban and subur
ban living continues,
crease.
However,
the per capita withdrawal rate will in
it appears that the level of population will
be the most important in affecting levels of future water
withdrawals in the state and among the several regions.
Estimates of water withdrawal were made by determining
a per capita withdrawal coefficient and then multiplying this
coefficient times the number of people expected in the state.
The withdrawal coefficients were developed on a regional
scale and involved a consolidation of county withdrawal rates
for municipal systems.
These withdrawal rates were based
upon unpublished data for 1965 from the Michigan Department
of Public Health with adjustments made for municipal and
non-municipal systems, and for increases in the withdrawal
rates over time.
ized.
Existing population projections were util
These were projections which had been developed for
state planning purposes.
216
The resulting estimates of regional withdrawals for
domestic, municipal, and commercial purposes is presented in
Table 23.
It can be seen that increased withdrawals are ex**
pected to 11)80 and 2000 for the state as a whole and for each
of the seven planning regions.
Region II will probably con
tinue to account for the bulk of the withdrawals, approaching
58 per cent of the state total, but the dominance of Region I
is expected to decline slightly.
Each of the regions is ex
pected to have an ever increasing proportion of its popula
tion served by municipal systems.
In the predominately urban
ized Region I this will approach 100 per cent by 2000, and
will also be relatively high in the other regions as well.
This will place an increasing burden on municipalities to pro
vide adequate service.
Manufacturing Withdrawals
As one of the nation's top industrial states, Michigan
has had tremendous demands placed upon its water resources by
manufacturing activity.
In 1963 Michigan ranked in the top
ten states in terms of manufacturing employment.
Between
1963 and 2000 manufacturing employment is expected to in
crease by over 240,000 employees.
This is over 25 per cent
of the 1963 totals and will be accompanied by significant in
creases in water demand.
In 1963 Michigan accounted for some
what over 900 billion gallons of water intake for manufactur
ing purposes; by 1980 this is expected to increase to 1,004.9
billion gallons and to 1,048.9 billion gallons by 2000.
217
Table 23
Domestic, Municipal, and <
C ommercial Withdrawalsa
Region
With
drawal
Per Cent
of Total
Per Cent
Municipal
Per Cent
Non-Municipal
1
1965
1980
2000
159,229.3
204,469.0
286,434.7
58 .6
58.0
57.5
95.2
96.8
99.8
4.8
3.2
.2
ZIA
1965
1980
2000
59,085.0
78,163.3
111,490.2
21.7
22.2
22.4
79.3
85.3
91.2
20.7
14.7
8.8
I ID
1965
1980
2000
6,504.8
7,786.4
10,527.1
2.4
2.2
2.1
66.2
71.7
79.9
33.8
28.3
20.1
IIIA
1965
1980
2000
31,787.3
43,467.9
64,322.3
11.7
12.3
12.9
79.4
84.9
90.2
20.6
15.1
9.8
II IB
1965
1980
2000
5,574.7
7,353.8
10,891.3
2.1
2.1
2.2
73.3
80.4
87.3
26.7
19.6
12.7
IV
1965
1980
2000
3,397.4
3,723.8
4,552.2
1.2
1.1
.9
83.9
87.6
93.1
16.1
12.4
6.9
V
1965
1980
2000
6,248.5
7,283.7
9,715.2
2.3
2.1
2.0
82.3
84.9
89.5
17.7
15.1
10.5
-
88.3
91.5
95.7
11.7
8.5
4.3
State
1965
1980
2000
271,827.0
352,247.9
497,933.0
aMillions of gallons
-
*
Water withdrawals for manufacturing purposes were es
timated by multiplying a per employee withdrawal rate times
the number of manufacturing employees expected in the state
in 1980 and 2000.
Because of the great diversity in manu
facturing types separate withdrawal estimates were made for
18 separate manufacturing types.
This resulted in a con
siderably more accurate estimate than if all manufacturing
employees were grouped together.
The basic data from which
these withdrawal coefficients were developed were taken from
the 1963 Census of Manufacturers report. Water Use in Manu
facturing.
Recognition was made of the effect of water con
servation techniques
(recirculation) and industrial with
drawals, but it was assumed that the compensatory effect of
increased labor productivity would balance this force and the
1963 coefficients were considered to remain essentially un
changed to 1980 and 2000.
Available estimates of manufacturing employment were
considered to be inadequate for this study and it was neces
sary to develop original estimates.
These were made by
disaggregating previously made estimates of total state em
ployment , and were made for each of the 83 counties in the
state.
In addition, the county estimates were further strati
fied into 21 manufacturing types based upon unpublished Mich
igan Employment Security Commission data for 1968.
The results of the water
withdrawal estimates for
1968 and for 1980 and 2000 are shown in Tables 24 and 25.
Table 24
Composition of Manufacturing Withdrawals by Type
1968
a
1980
2000
SIC
With
drawals
Per Cent
of Total
With
drawals
Per Cent
of Total
With
drawals
Per Cent
of Total
20
21
22
24
25
26
28
29
30
31
32
33
34
35
36
37
38
39
15,732.5
7.1
594.0
3,008.3
176.8
100,647.5
246,651.5
31,698.2
4,176.8
1,221.7
14,951.4
394,494.2
9,272.6
7,890.7
4,357.3
121,499.6
1,078.6
329.9
1.6
.0
.1
.3
.0
10.5
25.8
3.3
.4
.1
1.6
41.2
1.0
.8
.5
12.7
.1
.0
15,831.7
7.1
586.6
3,474.0
177.1
100,989.1
247,494.5
31,725.2
10,800.6
1,224.9
21,021.7
425,629.8
8,201.4
7,312.0
4,030.0
124,439.9
1,666.8
331.3
1.6
.0
.1
.3
.0
10.0
24.6
3.2
1.1
.1
2.1
42.4
.8
.7
.4
12.4
.2
.0
16,600.4
7.4
756.1
3,170.0
186.3
106,154.7
260,097.1
33,384.9
12,064.8
1,287.8
22,098.4
436,989.0
8,586.5
7,222.3
4,229.7
133,099.3
1,819.0
348.2
1.6
.0
.1
.3
.0
10.1
24.8
3.2
1.2
.1
2.1
41.7
.8
.7
.4
12.7
.2
.0
aMillions of gallons
Table 25
Regional Composition of Manufacturing Withdrawalsa
1967
1980
2000
Withdrawals
Per Cent
of Total
Withdrawals
Per Cent
of Total
Withdrawals
Per Cent
of Total
1
498.2
52.0
517.5
51.5
536.1
51.2
IIA
245.0
25.6
261.5
26.0
274.5
26.2
IIB
16.5
1.7
18.2
1.8
19.9
1.9
IIIA
100.0
10.4
110.8
11.0
121.5
11.6
IIIB
85.6
8.9
84.4
8.4
82.8
7.9
IV
6.6
.7
6.7
.7
7.1
.7
V
5.8
.6
5.9
.6
6.2
.6
Region
aBillions of gallons
221
Two important points stand out.
First of all, water with
drawals are highly concentrated in a relatively few industry
types.
In 1968 it is believed that four manufacturing types
accounted for the bulk of the withdrawals.
Listed in order
of importance, primary metals, chemicals, transportation
equipment, and the pulp and paper industry were responsible
for almost 90 per cent of the withdrawals.
Three of these
industry types were significant water users because of the
nature of the use made of the water; primary metals, chemicals,
and the pulp and paper industries all had a withdrawal coef
ficient of over 1,000,000 gallons per employee per year.
The
manufacturing type with the greatest per employee withdrawal
rate was the petroleum industry.
But because employment in
this category was so low total withdrawals were significantly
less than the pulp and paper industry, the next highest cate
gory.
The transportation industry was a major water user
even though its per employee withdrawal coefficient was not
extremely large.
This was a result of the magnitude of the
automotive industry in Michigan
(40 per cent of the total
manufacturing employment in 1968).
At the other end of the spectrum were those manufac
turing types which accounted for a relatively low amount of
the total withdrawals.
Eleven categories account for less
than one per cent of the total manufacturing withdrawal.
In
addition to these there were three categories whose water d e
mands per employee were so insignificant that they were not
»
222
included in the census study.
These were the ordinance, ap
parel, and printing and publishing industries, and in 1968
accounted for only about five per cent of the state's manu
facturing employment.
The distribution of withdrawals among industry types
is not expected to change to any degree to 2000
(less than
one per cent difference in the relative share among the 18
industries).
The regional distribution of withdrawal followed a
pattern similar to that of total manufacturing.
As expected,
the southern part of the state dominates the manufacturing
withdrawals.
of the total
state total).
Region 1 and IIA accounted for over 77 per cent
(Region I itself accounts for over half of the
On the otherhand, three of the northern regions
together accounted for only three per cent of the total state
withdrawals for manufacturing.
This regional pattern of withdrawals is not expected
to change significantly to 1980 and 2000.
It will take par
ticularly strong economic forces to alter the present pattern
of manufacturing activity in the state and affect a relocation
of water withdrawals.
Withdrawals for Power Generation
As population and manufacturing activity in Michigan
continue to expand, and as modern living and industry continue
to use greater numbers of technological devices, the amount of
electrical energy required in the state will increase
223
significantly.
As tremendous quantities of water are required
to cool the generating equipment this will mean a very large
demand on the state's water resource base.
Estimates of future withdrawals for electrical power
production were made by multiplying a withdrawal coefficient
of one half of one gallon of water per kilowatt minute times
the estimated production of electrical energy in the future.
Both the withdrawal coefficient and the estimate of future
power requirements were obtained from interviews and corres
pondence with engineers in the electrical power industry and
are the most reliable estimates available.
Based upon the withdrawal coefficient of one half of
one gallon of water per kilowatt minute, and an estimated
production of 39 million kilowatts of installed capacity in
1990, the estimated withdrawals will be approximately
6,506,665 million gallons per year in 1990.
This figure is
based upon 1970 technology and future advances may alter both
the withdrawal coefficient and the estimates of required
power.
No attempt was made to place this withdrawal estimate
within a regional framework.
However, it is almost certain
that most, if not all, of the new installed capacity will
utilize water from the Great Lakes.
It is the opinion of
this writer that this will occur primarily in the northern
part of tne state.
224
Agricultural Withdrawals
When compared to the other types of withdrawals, water
used for agricultural production is much less important in
Michigan.
Nevertheless, significant amounts of water are
used in agriculture in Michigan and may possibly increase in
importance in the future.
The significance of total with-
drawals is compounded by its spatial localization.
For areas
of the state where large quantities of water are demanded it
matters not whether agricultural withdrawals for the state as
a whole are relatively unimportant; for any one area they may
be very important.
Agricultural activity is a southern Michigan phenomena.
Climate and soil conditions
(and to some extent topographic
features) combine to form an agricultural resource base which
is very limited in the northern part of the state.
Conse
quently most of the agricultural production occurs in the
southern half of the Lower Peninsula;
in 1964 over 82 per
cent of the value of farm products produced in Michigan came
from Regions 1, IIA, and IIIA, the southern three regions of
the state.
Water withdrawals for agricultural production in Mich
igan occur for two basic reasons, for livestock maintenance
and for supplemental irrigation.
Irrigation in Michigan is
similar to irrigation in other humid areas of the nation;
natural precipitation supplies most of the needed moisture
and irrigation is used to make up any deficiencies which may
225
occur during periods of drought.
In livestock production
water is used in body maintenance and for other types of a c
tivities such as maintaining sanitary conditions, as an aid
in feeding, and other non-essential but desirable uses.
The discussion of agricultural withdrawals differs
from the previously mentioned withdrawals in that no specific
estimates of withdrawals were made.
The agricultural pattern
in Michigan is changing so rapidly that any estimate of fu
ture withdrawals would be little better than a guess.
For
this reason it was decided that greater service would be pro
vided by a summarization of trends in those activities and
conditions which affect agricultural withdrawals.
The location of irrigated acreage in Michigan is con
centrated in the southwest and the southeast parts of the
state, with almost 70 per cent of it being found in the fruit
belt of the counties in the southwest which are riparian to
Lake Michigan;
IIA.
this area is largely coincident with Region
In addition to fruits this is also an area of truck
crop production which is a heavily irrigated type of produc
tion.
The concentration in the southeast is largely a result
of truck crops and of nursery products.
There is no one concentration of livestock production
in the state outside of the concentration in the southern
half of the Lower Peninsula.
Although no specific estimates of water withdrawals
have been made for agriculture in Michigan a number of trends
226
have been identified which have implications for future with
drawals.
1.
The major ones have been listed below.
Decreasing number of farms and land in farms
throughout the state.
2.
An increasing concentration of agricultural activ
ity in the southern third of the s t ate, especially in Regions
IIA and IIIA.
3.
A significant increase in the size of the farming
unit and larger livestock herds.
4.
Increase in irrigated acreage and an increasing
concentration in Region IIA.
5.
Increased tendency to make greater use of the
irrigation equipment once installed, resulting in an increased
application rate per acre.
6.
An expanded use of water for non-essentials in the
production of livestock.
These trends, and others, make it clear that agricul
ture in Michigan will continue to demand increasing quantities
of water in the future, and that these demands will probably
be relatively concentrated in a few areas.
Implications and Recommendations
The implications of this study are quite clear.
Mich
igan will continue to require increased amounts of water to
support the population and economic activity of the state.
With the exception of certain areas of the state the water
resource base should be adequate to satisfy the demand, but
the supply/demand situation will be much more critical in
some areas of the state than others.
It is the recommendation of this author that an ade
quately financed,
three-pronged study be initiated by the
Michigan Water Resources Commission to explore in depth some
of the conditions presented in this study.
This should be
focused toward agricultural, manufacturing, and domestic,
municipal, and commercial withdrawals, each type being the
focus of a separate study.
turing establishments
Typical farming units, manufac
(separated into the major SIC cate
gories) and municipal water systems
vice areas)
(of various size ser
should be identified and their cooperation sought
for a long range study of water withdrawals.
It is felt that
a ten year period should be the minimum length of time for
such a study.
If every attempt is made to insure that the
units chosen are representative of that type of activity in
the state it would be possible to make reliable generaliza
tions to the rest of the activity.
Strict consistency should be maintained among the
several areas of the state and over time.
This is something
very much lacking in the regional water use reports
and makes a comparison among regions of th
if not impossible.
(page 10)
state difficult,
BIBLIOGRAPHY
BIBLIOGRAPHY
Books and Reports
Bailey, Donald.
Preliminary Population Projections for Small
Areas in Michigan. Lansingi State Resource Planning
Division, Office of Economic Expansion, Michigan De
partment of Commerce, 1966.
Barlowe, Raleigh.
Implications of Land and Water Use Develop
ments in Michigan for Future Public Water kesource Pol
icy. Lansing: Department of Resource Development, Mich*
igan State University, 1966.
Barlowe, Raleigh.
Use of Land and Water Resources in Michi
g a n . Report No. 52 of Project 80: Rural Michigan Now
and in 1980.
East Lansing: Agricultural Experiment
Station and Cooperative Extension Service, Michigan
State University, 1966.
Battelle Memorial Institute.
Grand River Basin (Michigan)
Comprehensive Water Resources Study Appendix O: Econ
omic Base Study. Columbus, Ohio, 1966.
Battelle Memorial Institute.
The Michigan Manpower Study: An
Analysis of the Characteristics of Michigan's Labor
Force in the Nexi~i5 Y e a r s . Columbus, Ohio, 1966.
Bronder, Leonard D., and Koval, John M.
M i c h i g a n 1s Economic
Past: Basis for Prosperity. Lansing: &£ate Resource
Planning Division, Office of Economic Expansion, Mich
igan Department of Commerce, 1967.
Bronder, Leonard D . , and Koval, John M. Michigan*s F u t u r e :
Its Population and Its Eco n o m y . Lansing: State hesource Planning Division, Office of Economic Expan
sion, Michigan Department of Commerce, 1967.
Dunbar, Willis F.
Michigan: A History of the Wolverine State,
Grand Rapids, Michigan: William B. Eerdmans Publishing
Co., 1965.
228
229
Flack, J.E., and Martinez-F, Fortunato.
Urban Water Use
Study. Denver: American Society o F C i v i l Engineers
Viater Resources Engineering Conference, May 16-20,
1966.
Haber, William.
The Michigan Economy: Its Potentials and Its
Problems. Kalamazoo, Michigan: W.E. Upjohn Institute,
1959 .
Haber, William, Spivey, Allen, and Warshaw, Martin (eds.).
Michigan in the 1970's: An Economic Forecast. Ann
Arbor: ¥ure a u of Business Research, Graduate School of
Business Administration, The University of Michigan,
1965.
Hill, Elton B . , and Mawby, Russell G.
Types of Farming in
Michigan. East Lansing: Agricultural Experiment Sta
tion, Michigan State University, 1954.
Hittman Associates, Inc.
"Main I" A System of Computerized
Models for Calculating
and Evaluating Municipal water
C
kequirementss. Columbia, Maryland, 196^.
Huffman, Roy. Irrigation Development
lopment and Public Water Policy.
Mew YorJTi konald Press , 1953
.
Humphrys, Clifford. Michigan Lakes and P o nds. East Lansing:
Agricultural Experiment Station, Michigan State Uni
versity, 1965.
Indra, R.R.
Michigan Population Handbook. 1965. Lansing:
Michigan Department of Public health, 1765.
Israelsen, Orson W.
Irrigation Principles and Practices.
New Y o r k : John Wiley and Sons, Inc., 1950.
Kidder, E.H., and Wheaton, R.Z. Supplemental Irrigation in
Mic h i g a n . East Lansing: Cooperative Extension ser
vice, Michigan State University, 1958.
Kline, C.K. Water Systems Analysis to Meet Changing Condidit i o n s . feast Lansing: Cooperative Extension Service.
Michigan State University, no date.
Kuznets, Simon. "Concepts and Assumptions in Long-Term Pro
jections of National Product," in Studies in Income
and Wea l t h . Vol. 16.
Washington: National Bureau of
Economic Research, 1954.
Kuznetz, Simon. Population Redistribution and Economic Growth.
United States 1870-1950. Philadelphia: American Philo
sophical Society, 1957•
2 JO
Michigan Department of Agriculture.
Michigan Agricultural
Statistics, 1968. Lansing, 1965.
Michigan Department of Public Health.
Michigan State Plan,
1967-68, For Hospital and Medical Facilities Construction. Lansing, 1565.
Michigan Department of Public Health.
Report on Water Re
quirements for Municipal U s e . Lansing, 1966.
Michigan Electric Power Capacity and Reliability Committee.
Michigan Electric Power Capacity and Reliability
1566-1970. Lansing, 1966.
Michigan Water Resources Commission.
The Water Resources of
the Lower Lake Huron Drainage Basin: An overview o T
Region Water U s e s . Lansing, 1568.
Michigan Water Resources Commission.
The Water Resources of
the Lower Lake Michigan Drainage B a s i n » An Overview of
Region Water U s e s . Lansing, 1968.
Michigan Water Resources Commission.
The Water Resources of
the Northern Lake Michigan and Lake Huron Drainage
Area Lower Peninsulas An Overview of Region Water Uses,
Lansing! 1965.
Michigan Water Resources Commission.
The Water Resources of
Southeastern
istern Michigan: An Overview of kegion Water
Lansing! 1968.
Uses.
flichigan Water Resources Commission.
The Water Resources of
the Upper Peninsula Drainage A r ea: An Overview of Re
gion Water U s e s . Lansing, 1568.
Michigan Water Resources Commission.
Water Use for Irriga
tion: A Survey of Water Use in Agricultural and Munic
ipal Irrigation. Lansing, 1959.
Myers, Lynn C . , and Van Meer, David W. Michigan Statistical
Abstract, 1 9 6 4 . East Lansing: Bureau of business and
Economic Research, Graduate School of Business Admin
istration, Michigan State University, 1966.
Perloff, Harvey, and wingo, Lowdon. "Natural Resource Endow
ment and Regional Economic Growth," in Friedmann, John
and Alonso, William.
Regional Development and Plan
ning.
Cambridge, Massachusetts: The M.I.T. Press,
1964.
231
guaife, M . M . , and Glazcr, S.
Michigan: From Primitive Wilderncss to Industrial Commonwealth. New Vork: l»renticc-hall, Inc., 1968.
Regional Economics Division, Office of Business Economics,
U.S. Department of Commerce.
Preliminary Report on
Economic Projectio n s for Selected Geographic Areas,
19^9 to 2 0 2 0 , Vol. FI Washington: Water Resources
Council, 1908.
Soil Conservation Service, U.S. Department of Agriculture.
Inventory of Michigan Soil and Water Conservation
Ne ed s . East Lansing: Agricultural kxperiment Station,
Michigan State University, 1968.
Spaulding, Irving A.
Household Water Use and Social Status.
Kingston: Agricultural kxperimenk Station, University
of Rhode Island, 1967.
State Resource Planning Division, Office of Economic Expan
sion, Michigan Department of Commerce.
Population and
Labor Force Projections for Michigan. Lansing, 19&6.
Thornthwaite, C.W., and Mather, J.R. The Water Balance.
Cen
ter ton, New Jersey: brexel Institute of Technology,
1955.
Trewartha, Glen T.
An Introduction to Cli m a t e .
Me Graw-Hill Book C o . , 1954.
New York:
U.S. Bureau of the Budget.
Standard Industrial Classlfication Manual.
Washington: U.S. Government Printing
CTH ce, T . W 7,
Veatch, J •O. Agricultural Land Classification and Land Types
of Michigan. East Lansing: Agricultural Experiment
Station, Michigan State University, 1941.
Water Resources Council.
The Nation*s Water Resources.
ington : U.S. Government Printing bffice, 1968.
Wash
Whiteside, E.P., Schneider, I.F., and Cook, R.L.
Soils of
Michigan. East Lansing: Agricultural Experiment £ Na
tion and Cooperative Extension Service, Michigan State
University, 1968.
Wright, Karl T.
Economic Prospects of Farmers. Report No.
47 of Project 80: Rural Michigan tJow and in 1980.
East Lansing: Agricultural Experiment Station and
Cooperative Extension Service, Michigan State Univer
sity, 1966.
232
Wright, Karl T . , and Caul, D.A.
Michigan A griculturet Its
Income, Major Products, Locations, and tihangest
East
Lansing; Cooperative Extension Service, Michigan State
University, 1967.
Newspapers and Periodicals
Cams,
J.G. "Service Lines and Meter Requirements of Domes
tic Water Connected Devices," Journal of the American
Water Works Association, Vol. 58, No. 10 (October.
!96S7" pp. "T756-1253.
Cleary, Edward J. "Some New Facts for Forecasting Industrial
Water Needs," Public Works, Vol. 92, No. 11 (November,
1961), pp. 78,' 80, 02.
Derby, Ray L. "Water Use in Industry," Journal of the Irriqaof the Ameri-
Dunn, Dorthy F., and Larson, Thurston E.
"Relationships of
Domestic Water Use to Assessed Valuation with Sel
ected Demographic and Socio-Economic Variables," Jour
nal of the American Water Works Association, Vol. 55,
tfo. 4 (April, 1^63), pp. 441-449.
Henderson, Angus D.
"The Lawn Sprinkling Load," Journal of
the American Water Works Association, Vol. 48, No. 4
(April, l4^6), pp. 361-364.
Iloak, Richard D.
"Water Resources and the Steel Industry,"
Iron and Steel Engineer, Vol. 41, No. 7 (May, 1964),
pp. 0/-9U.
Larson, Brent O . , and Hudson, H.E. "Residential Water Use
and Family Income," Journal of the American Water
Works Association. Vol. 43, Mo. i (July, 1951}, pp. 610.
Messenger, A.S.
"The Water Balance of the Lower Peninsula of
Michigan," Papers of the Michigan Academy of Science,
Arts, and Letters, Vol. X l VIY (1962), pp. 38^-392.
Patterson, W. "Demand Rates for Water Service,” Journal of
The American Water Works Association. Vol. 5^, No. 10
(October, 1961}, pp. 1261-1269.
233
Porges, Kalph. "Factors Influencing Per Capita Water Consump
tion," water and Sewage Works, Vol. 104, No. 5 (Mav.
1957), pp.' 193-204.— 3-------
Powell, Sheppard L). "Relative Economic Returns from Indus
trial and Agricultural Water Uses," Journal of the
American Water Works Association, Vol. 48, No. 8 Ta u gust, 1956) , pp. 991-992.
Ouackenbush, Tyler H . , and Shockley, Dell G. "The Use of
Sprinklers for Irrigation," in Water; 1955 Yearbook of
Agriculture. Washington; U.S. Government Printina Of-
1ice ,"1955.
Ruble, Earl H. "Industrial Water Requirements," Journal of
the American Water Works Association, Vol. 57, No. 7
(July, 1965), pp. 831-833.
Seidel, Ii., and Baumann, E. "A Survey of Operating Data on
Water Works in 1955," Journal of the American Water
Works Association, Vol. 59, No. 5 (Hay, '1967), pp. 553.
Silva, Rudolfo. "band Uses and and Water Consumption Require
ments," Public W o r k s , Vol. 90, No. 4 ( April, 1959),
pp.
120-124.
"1969 Survey of Buying Power," Sales Management, Vol. 102,
No. 2 (July 10, 1969), pp^ E>8b-D95.
Tharp, Max M . , and Crickman, C.w. "Supplemental Irrigation in
Humid Regions," in Water; 1955 Yearbook of Agriculture.
Washingtons U.S. Government Printing Office, 1955.
Thomas, Morgan D. "A Regional Model for Projecting Industrial
Water Consumption," Papers of the Michigan Academy of
Science, Arts, and betters. Vol. xLlll (1957), pp.
251- 25’b:----
--------------
Uliman, Edward b. "Amenities as a Factor in Regional Growth,”
The Geographical Review, Vol. XblV, No. 1 (January,
m y y , " pP . i i b -t t ?.----
Watson, Kenneth S. "Water Requirements of Dishwashers and
Food Waste Disposers," Journal of the American Water
Works Association, Vol. ^5, No. 5 (May, 1963), pp.
555-559.
"Where *s Our Water?"
State Journal
February 6, 1969.
(bansing, Michigan),
Unpublished Material
College of Agriculture, Michigan State University, and Soil
Conservation Service, U.S. Department of Agriculture.
"Tentative Irrigation Guide for the Design of Sprin
kler Irrigation Systems in Michigan." August, 1954.
(Mimeographed)
tianke, Steve H.
"The Effects of Metering Urban Water." Paper
presented to the Seventh Annual Institute for City
Engineers, University of Colorado, Boulder, Colorado,
February 6-8, 1968.
Kollar, K.L. , and Brewer, Robert. "Water Requirements for
Manufacturing." Unpublished material from Water Indus
tries and Engineering Services Division, Business and
Defense Services Administration, U.S. Department of
Commerce, Washington, D.C., May, 1968.
Michigan Department of Public Education.
Unpublished statis
tics on public school and college enrollment.
Lansing,
1968.
Michigan Department of Public Health.
Unpublished statistics
on water use by municipal water systems.
Lansing,
1968.
Michigan Employment Security Commission.
Unpublished statis
tics on county employment by two-digit SIC categories,
1968.
Lansing, 1968.
Niedringhaus, Thomas. "A Climatology of Michigan."
Unpub
lished Ph.D. dissertation, Department of Geography,
Michigan State University, 1964.
Vertrees, Robert L.
"A Survey of Nonresident Landowners of
Ten or More Acres in Antrim and Kalkaska Counties,
Michigan." Unpublished M.A. thesis, Department of Re
source Development, Michigan State University, 1967.
Wingert, Wayne L.
"Present and Future Power Development for
the State of Michigan."
Paper presented at the Gov
ernor 's Conference on Thermal Pollution, Traverse City,
July 18, 1969.
235
Personal Interviews and Correspondence
Kelly, William A.
Michigan Department of Public Health,
Lansing, Michigan.
July, 1968.
Kidder, E.H.
Department of Agricultural Engineering, Michi
gan State University, East Lansing, Michigan.
July,
1969.
Ritchey, Harlan.
Department of Animal Husbandry, Michigan
State University, East Lansing, Michigan.
July, 1969,
Syler, D.E.
Senior Engineer, Consumers Power Company, Jack
son, Michigan.
August, 1969.
U.S. Census Reports
U.S. Bureau of the Census.
Annual Survey of Manufacturers!
1959, 1960, 1961. 1967, 1964. 1^65, 1966. Washington:
U .s . Government Printing Office.
U.S. Bureau of the Census.
Census of Agriculturet 1964.
Statistics for State and Counties, kichigan.
Washingtoni U.S. Government Printing Office, 1967.
U.S. Bureau of the Census.
U.S. Census of Housing: 1960.
Vol. I, States and Small Areas, tart 5, Michigan.
Washingtons U.S. Government Printing Office, 1963.
U.S. Bureau of the Census.
Census of Manufacturers! 1 9 6 3 .
Vol. Ill, Area Statistics.
Part 23, Michigan.
Washington! U.S. Government Printing Office, 1966.
U.S. Bureau of the Census.
Census of Manufacturers! 196 3 .
Subject Statistics: Water use in Manufacturing.
Washington: U.S. Government Printing Office, 1966.
U.S. Bureau of the Census.
U.S. Census of Population! 1960.
Vol. I, Characteristics of tne Population,
part 24,
Michigan.
Washingtoni U.S. Government Printing Of
fice, 1963.
APPENDICES
APPENDIX TABLE 1
Sources of Michigan Population Growth*
Period
1960-1963b
Population
Increase
Natural
Increase
Per Cent
Migration
Increase
Per Cent
Annual Rates
Births
Deaths
176.7
-76.7
2.4
.9
1950-1959
1,421,500
89.5
10.5
2.7
.9
1940-1949
1,140,895
65.1
34.9
2.2
1.0
1930-1939
420,805
91.9
8.1
1.8
1.1
1920-1929
1,157,690
39.8
60.2
2.3
1.2
1910-1919
855,615
41.0
59.0
2.4
1.3
1900-1909
401,706
44.4
55.6
2.0
1.3
*R. Raja Indra, Michigan Population Handbook 1965 (Lansing: Michigan Department
of Public Health, 1965), p. 8.
Estimated
236
267,000
APPENDIX TABLE 2
County Population Concentration
1920
1940
c
Per
Cent
Popu
lation
„ c
Per
Cent
Wayne
Oakland
Macomb
Genesee
Kent
Ingham
Saginaw
Washtenaw
Kalamazoo
Muskegon
Berrien
Calhoun
Jackson
St. Clair
Bay
1,177.6
90.1
38.1
125.7
183.0
81.6
100.3
49.5
71.2
62.4
62.7
72.9
72.5
58.0
59.5
32.1
2.5
1.0
3.4
5.0
2.2
2.7
1.3
1.9
1.7
1.7
2.0
2.0
1.6
1.9
2,015.6
254.1
107.6
227.9
246.3
130.6
130.5
80.8
100.1
94.5
89.1
94.2
93.1
76.2
75.0
TOTAL
2,315.1
63.1
3,815.7
1980
^ c
Per
Cent
2000
o
Popu
PerC
lation Cent
lation
Cent
Popu
lation
38.3
4.8
2.1
4.3
4.7
2.5
2.5
1.5
1.9
1.8
1.7
1.8
1.8
1.5
1.4
2,666.3
690.3
405.8
374.3
363.2
211.3
190.8
172.4
169.7
150.0
149.9
138.9
132.0
107.2
107.0
34.0
8.8
5.2
4.8
4.6
2.7
2.4
2.2
2.2
1.9
1.9
1.8
1.7
1.4
1.4
2,700.5
1,136.4
954.3
554.0
459.5
295.7
245.2
300.4
222.1
171.5
191.5
147.3
141.1
117.1
117.3
27.4
11.5
9.7
5.6
4.7
3.0
2.5
3.0
2.3
1.7
1.9
1.5
1.5
1.2
1.2
3,113.1
1,678.6
1,663.5
790.8
592.9
400.2
318.0
553.7
295.4
210.0
250.4
174.4
175.7
138.2
140.3
23.7
12.8
12.7
6.0
4.5
3.1
2.4
4.2
2.3
1.6
1.9
1.3
1.3
1.1
1.1
72.6
6,029.0
77.1
7,757.1
78.6
10,495.3
80.0
^J.S. Bureau of the Census, U.S. Census of Population: 1960. Vol. I, Characteris
tics of the Population. Part 24, Michigan (Washington: U.S. Government Printing Office,
1963) and Donald E. Bailey, Preliminary Population Projections for Small Areas in Michi
gan (Lansing: State Resource Planning Division, Office of Economic Expansion, Michigan
Department of Conmerce, 1966).
h
Thousands
c
Per cent of state total
237
County
Popu
lation
1960
_ c
r
.
O Per
Popu
238
APPENDIX TABLE 3
Comparison of Population Estimates**9
Source
a
Bureau of the Census
National Planning
Association
Stanford Research
Institute
Dr. David Goldberg
Michigan in the 1970's
Michigan Energy Study
1970
1975
1980
8,658
9,271
9,999
11,615
9,384
8,645
8,891
10,268
9,192
9,494
11,275
9,868
10,204
10,700
These estimates were compiled in State Resource Planning Division, office of Economic Expansion, Michigan Depart
ment of Commerce, Population and Labor Force Projections for
Michigan (Lansing, 1966).
^Thousands
Estimated
U.S. Bureau of the Census, Current Population Reports.
Population Estimates, Series P-25, No. 301 (Washingtons U.S.
Government Printing Office, February 26, 1965).
U.S. Department of Labor, Bureau of Labor Statistics,
Projections to the Year 1976 and 2000* Economic Growth, Pop
ulation. Labor frorce. Leisure, and Transportation. Outdoor
kecreatlon Resources Review Commission Study Report 23 (Wash
ingtons U.S. Government Printing Office, 1962).
Pietro Balestra and Rao Koteswara, Basic Economic
Projections, United States Population. 1963-1980 ( M e n l o P a r k ,
Californiai Stanford Research institute, 1963.
David Goldberg, Population Projections M i c h i g a n »
1960-1960 (Lansing* State Resource Planning Division, Office
of Economic Expansion, Michigan Department of Commerce, 1966).
William Haber, Allen Spivey, and Martin Warshaw (eds),
Michigan in the 1 9 7 0 *s An Economic Forecast (Ann Arbor *
bureau of Business Research, university of Michigan, 1965).
Michigan Energy Study (Lansing* Office of Economic
Expansion, Michigan Department of Commerce, 1963).
239
Appendix Table 4
Non-Industrial Withdrawals, United States4**
Public Systems
Year
Domestic
Public
Commercial
2000
1980
1965
81
77
73
16
18
20
28
28
28
aPer Capita daily withdrawals
Total
Public
Individual
Systems
125
123
121
71
58
51
(gallons)
Water Resources Council, The Nation's Water Resources
(Washington: U.S. Government Printing Office, 1968) , p"! 4-1-? .
Appendix Table 5
Independent Variables for Regression Model
Variable
Per cent of population which was urban.
1960
Buying income per family, 1968
Number of people per housing unit
Per cent of homes in 1960 which were
over $20,000 in value
Median value of homes in 1960
Per cent of homes heating with steam
or hot water in 1960
Per cent of homes with complete plumbing
facilities in 1960
Per cent of homes with clothes washing
machines in 1960
Per cent of homes with air conditioners
in 1960
Per cent of homes with hot and cold run
ning water inside structure in 1960
Per cent of homes in 1960 which were built
within the last 10 years
Number of acute care hospital beds (1968)
per 1000 population (1960)
Number of nursing home beds (1968) per
1000 population (1960)
Sourcea
K
1
.11
2
3
3
.09
.05
.05
3
3
.12
.23
3
.01
3
.02
3
.05
3
.08
3
.02
4
.20
4
.09
240
Appendix Table 5— Continued
Variable
Number of public school students (1968)
per 1000 population (1960)
Number of public four year college students
(1968) per 1000 population (1960)
Number of public two year college students
(1968) per 1000 population (1960)
Number of non-public college students
(1968) per 1000 population (1960)
Source8
R
5
.02
5
.09
5
.33
5
.07
a Source;
^U.S. Bureau of the Census, U.S. Census of Popula
tion: 1 960 . Vol. I, Characteristics o £ the Population, part
24, Michigan (Washington: U.S. Government Printing Office,
1963).
2
"1969 Survey of Buying Power," Sales Management,
Vol. 102, No. 2 (July 10, 1969), pp. D88-D95.
3U.S. Bureau of the Census, U.S. census of Housing:
1960. Vol. I, States and Small Areas"!
Part 5, Michigan.
(Washington: U.S. Government Printing Office, 1963).
^Michigan Department of Public Health, Michigan State
Plan, 1967-68, For Hospital and Medical Facilities construc
tion (Lansing, 1968).
^Unpublished statistics from the Michigan Department
of Education, Lansing, Michigan.
241
Appendix Table 6
Per Capita Non-Industrial Withdrawals, 1965J
County
Alcona
Allegan
Antrim
Baraga
Bay
Berrien
Calhoun
Charlevoix
Chippewa
Clinton
Delta
Eaton
Genesee
Gogebic
Gratiot
Houghton
Ingham
Iosco
Isabella
Kalamazoo
Kent
Lake
Leelanau
Livingston
Mackinac
Manistee
Mason
Menominee
Missaukee
Montcalm
Muskegon
Oakland
Ogemaw
Osceola
Otsego
Presque Isle
Saginaw
St. Joseph
Schoolcraft
Tuscola
Washtenaw
Wexford
Coefficient
32
34
58
35
48
46
50
45
42
27
37
39
35
35
29
35
38
35
24
26
38
29
19
62
57
41
35
34
27
36
43
38
41
33
35
35
48
35
70
30
45
35
000
982
780
536
281
287
251
541
974
053
000
465
500
500
173
500
139
536
349
444
744
091
342
174
826
709
500
561
941
497
692
108
538
333
500
500
281
035
851
256
523
500
c
c
c '
e
f
c
c
c
c
c
c
c
d
e
c
e
c
c
c
c
c
c
c
c
c
c
e
c
c
c
c
f
c
c
e
e
f
c
c
c
c
e
County
Alger
Alpena
Arenac
Barry
Benzie
Branch
Cass
Cheboygan
Clare
Crawford
Dickinson
Emmet
Gladwin
Grand Traverse
Hillsdale
Huron
Ionia
Iron
Jackson
Kalkaska
Keweenaw
Lapeer
Lenawee
Luce
Macomb
Marquette
Mecosta
Midland
Monroe
Montmorency
Newaygo
Oceana
Ontonagon
Oscoda
Ottawa
Roscommon
St. Clair
Sanilac
Shiawassee
Van Buren
Wayne
Coefficient
35.800
35.974
28.095
17.989
33,333
29,100
32,711
41,630
43,278
49,524
26,624
64,967
35,692
62,871
25,126
37,244
40,754
29,310
26,571
20,526
27,826
41,684
30,125
46,129
38,108
36,692
23,333
48,281
35,151
35,500
28,610
49,353
45,000
35,500
44,349
27,777
35,500
26,300
29,997
42,631
38,108
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
f
c
c
f
c
e
c
c
c
d
c
c
e
c
c
c
f
242
appendix Table 6 --Continued
a Unpublished statistics from the Michigan Department
of Public Health, Lansing, Michigan.
^Annual per capita withdrawal
(gallons).
C o e f f i c i e n t used as reported in Michigan Department
of Public health statistics.
^Missing data, median coefficient used
£
Extreme data, median coefficient used
^Grouped data, group coefficient used
Appendix Table 7
Per Cent Served by Municipal Systems,
M.D.P.H. Regions
Region
1940
1960
1980
2015
1
2
3
4
5
6
7
8
9
68.2
42.1
50.1
36.4
50.8
62.7
63.1
91.8
53.4
72.0
53.5
46.1
37.4
59.0
60 .0
58.1
89.0
52.4
85.2
52.3
56.9
61.8
67.6
71.2
73.1
92.7
78.9
95.0
88.1
83.8
92.1
88.5
89.3
93.8
97.8
94.5
aMichigan Department of Public Health, Report on Water
Requirements for Municipal Use (Lansing, 1966), pp. 3-20.
IIU Ntltl
inn
'•CMMLCflAPT
c O
'Hli IlMfl
Itll
ch*»l*Voi*L
IU( ” roic«r* rc7 ^ fMl
in* 'newM
U
III
M A T W T I ••••»*•
mam,■«<*
Appendix Figure 1. Michigan Department of
Public Health Planning Regions
aMichigan Department of Public H e a l t h , Report on Water
Requirements for Municipal Use (Lansing, 1966).
244
Appendix Table 8
Increase in Withdrawal Rates
Public
System
Individual
System
1965-2000
3.31
40.00
1965-1980
1.65
13.72
Period
Computed from data reported in Water Resources Coun
cil, The Nation's Water Resources (Washington: U.S. Govern
ment Printing Office, 1568), pZ T-l-l.
kper cent increase
Appendix Table 9
Manufacturing Composition, Michigan3*5
SIC
1958
1963
1968
1980
2000
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
.4
7.2
.1
.3
1.4
1.4
2.3
3.4
3.5
4.4
.4
1.8
.6
2.2
8.5
9.9
15.6
3.4
31.6
1.1
2.1
.2
6.0
.0
.3
1.8
1.4
2.1
3.2
3.2
3.8
.3
2.2
.4
2.0
9.3
9.9
15.3
3.7
32.1
1.5
1.3
.4
4.5
.0
.3
1.8
1.0
1.9
2.4
2.6
4.1
.2
2.1
.4
1.7
8.5
11.7
15.5
4.1
34.9
1.0
.9
.4
4.5
.0
.3
2.4
1.1
1.9
2.4
2.6
4.1
.2
2.4
.4
1.7
9.1
10.3
14.3
3.8
35.6
1.5
.9
.4
4.5
.0
.4
2.8
1.0
1.9
2.4
2.6
4.1
.2
2.5
.4
1.7
8.9
10.3
13.4
3.8
36.2
1.6
.9
245
Appendix Table 9— Continued
al'or Tables 9 through 12 the following conditions are
applicable.
The 1958 and 1963 data were taken from U.S. Bu
reau of the Census, Census of Manufacturers: 1958 and 1963
(Washington: U.S. Government Printing O f f i c e ) . The 1^67/68
data were taken from unpublished statistics from the Michigan
Employment Security Commission.
The 1967 figures were based
upon an apportionment of the 1967 monthly average for the
state according to the county employment figures available
for a period in 1968 (see page
).
Projections for 1980
and 2000 were made by the author in accordance with tech
niques outlined in Chapter VI.
b
Per cent of state total.
Appendix Table 10
Regional Manufacturing Employment
Region
1958
1963
1967
1980
2000
I
513,350
58.3
551,905
57.4
644,296
58.3
662,551
57.8
694,109
57.6
IIA
222,414
25.3
245,519
25.6
278,257
25.2
288,100
25.1
301,002
25.0
IIB
14,835
1.7
18,198
1.9
20,079
1.8
22,581
2.0
24,695
2.1
I IIA
95,085
10.8
111,368
11.6
125,716
11.4
136,618
11.9
148,414
12.3
IIIB
20,040
2.3
20,830
2.2
24,043
2.2
24,849
2.2
24,849
2.1
IV
7,352
.8
6,696
.7
7,193
.7
7,193
.6
7,193
.6
V
6,947
.8
6,574
.7
5,026
.5
5,026
.4
5,026
.4
a24,849 absolute.
2.2 per cent of total.
246
Appendix Table 11
Structure of Regional Manufacturing Employment3
SI
1968
1980
I
19
20
21
22
23
24
25
20
27
28
29
30
31
32
33
34
35
36
37
38
39
.6
3.6
n
.2
2.2
.3
.6
1.1
2.4
3.0
.2
2.3
n
1.7
8.7
12.6
16.5
3.1
39.6
.7
.5
.6
3.5
n
.2
3.0
.4
.6
1.1
2.4
3.1
.2
2.6
n
1.7
9.2
11.1
15.5
2.9
40.3
1.2
.6
IIA
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
.0
6.6
.0
.6
1.7
1.2
5.4
5.5
3.5
3.3
.1
2.4
.9
2.0
8.6
11.6
18.2
.0
6.7
.0
.6
2.3
1.5
5.4
5.5
3.6
3.4
.1
2.7
.8
2.0
9.4
10.3
16.5
247
Appendix Table 11— Continued
1968
1980
7.3
17.4
2.1
1.6
6.6
18.0
3.1
1.6
IXB
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
.0
15.3
.0
n
3.3
5.8
3.1
3.0
3.1
3.8
.2
7.1
5.4
2.3
7.3
9.7
11.8
10.0
5.4
1.9
1.3
.0
16.4
.0
n
2.6
5.2
3.2
3.1
3.1
4.0
.2
5.9
5.1
2.3
6.8
10.6
12.6
10.5
5.6
1.4
1.3
I IIA
19
20
21
22
23
24
25
26
27
28
29
30
31
.1
3.2
.0
.2
.3
.4
.7
.4
1.5
1.3
.6
.5
.3
.1
3.3
.0
.2
.3
.4
.7
.4
1.5
1.3
.6
.6
.3
248
Appendix Table 11— Continued
38
39
1968
1980
.8
9.3
9.9
5.6
1.6
61.9
.1
.3
.8
10.3
8.7
5.3
1.3
63.3
.1
.3
IIIB
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
.0
1.2
.0
n
.4
5.6
.4
5.5
2.0
57.0
.3
1.2
.0
4.4
1.6
5.2
6.6
1.6
5.0
n
2.3
.0
1.1
.0
n
.3
4.6
.3
5.2
2.1
58.2
.3
1.0
.0
4.7
1.4
5.4
6.8
1.4
4.3
n
3.0
IV and V
19
20
21
22
23
24
25
26
.4
8.3
.0
.0
2.6
22.6
7.5
17.5
.4
7.7
.0
.0
2.9
24.8
7.2
18.3
249
Appendix Table 11— Continued
SIC
1968
1980
2000
27
28
29
30
31
32
33
34
35
36
37
38
39
4.7
.4
.0
.3
.2
2.4
5.5
.9
17.0
5.0
2.2
.1
2.6
4.3
.4
.0
.3
.1
2.3
5.4
.8
16.1
4.2
2.1
.1
2.7
4.3
.4
.0
.4
.1
2.3
5.3
.8
16.5
4.1
2.2
.1
2.9
Per cent of total.
Letter n signifies less than one-tenth of one per
cent.
Appendix Table 12
Spatial Composition of Regional
Manufacturing Employment
County
Lenawee
Livingston
Macomb
Monroe
1958
1963
1967
1980
2000
10,001
2.0
10,201
1.8
13,141
2.0
13,603
14,666
2.1
2.1
1,715
.3
2,303
.4
2,484
.4
2,795
.4
3,106
.4
61,459
12.0
69,722
12.6
88,366
13.7
95,726
14.4
97,044
13.8
6,151
1.2
6,837
1.2
6,864
1.1
6,875
1.0
6,825
1.0
250
Appendix Table 12— Continued
County
1958
1963
1967
50,042
9.8
66,528
12.1
93,666
14.5
110,184
16.6
130,527
18.5
St. Clair
7,755
1.5
9,741
1.8
9,278
1.4
9,658
1.5
10,052
1.4
Sanilac
2,227
.4
3,083
.6
3,320
.5
3,754
.6
4,212
.6
18,009
3.5
25,827
4.7
35,214
5.5
42,089
6.4
50,228
7.1
355,991
69.3
357,663
64.8
391,963
60.8
377,867
57.0
377,449
53.6
Oakland
Washtenaw
Wayne
1980
2000
IIA
Allegan
3,665
1.6
3,871
1.6
4,681
1.7
4,868
1.7
5,112
1.7
Barry
2,399
1.1
3,000
1.2
3,115
1.1
3,309
1.1
3,498
1.2
20,714
9.3
23,533
9.6
26,291
9.5
27,435
9.5
28,751
9.6
2,844
1.3
3,276
1.3
3,214
1.2
3,247
1.1
3,229
1.1
20,551
9.2
19,407
7.9
20,080
7.2
18,248
6.3
16,922
5.6
2,174
1.0
2,655
1.1
3,071
1.1
3,331
1.2
3,626
1.2
811
.4
914
.4
524
.2
483
.2
447
.1
Eaton
1,978
.9
2,279
.9
2,732
1.0
2,925
1.0
3,161
1.1
Hillsdale
1,921
.9
2,468
1.0
2,959
1.1
3,334
1.2
3,711
1.2
Berrien
Branch
Calhoun
Cass
Clinton
251
Appendix Table 12— Continued
County
1958
1963
1967
1980
2000
Ingham
22,910
10.3
25,188
10.3
33,578
12.1
36,510
12.7
40,147
13.3
3,799
1.7
3,894
1.6
3,310
1.2
2,995
1.0
2,825
1.0
Jackson
14,902
6.7
16,501
6.7
19,521
7.0
20,492
7.1
21,734
7.2
Kalamazoo
25,063
11.3
25,884
10.5
29,151
10.5
28,963
10.1
29,234
9.7
Kent
44,468
20.0
50,090
20.4
58,287
21.0
61,440
21.3
65,136
21.3
Montcalm
3,452
1.6
3,966
1.6
6,282
2.3
7,260
2.5
8,418
2.8
Muskegon
23,504
10.6
25,103
10.2
25,883
9.3
25,351
8.8
24,900
8.3
Ottawa
12,012
5.4
14,512
5.9
15,748
5.7
16,771
5.8
17,788
5.9
St. Joseph
7,375
3.3
8,590
3.5
8,650
3.1
8,792
3.1
8,921
3.0
Shiawassee
4,158
1.9
5,429
2.2
6,321
2.3
7,097
2.5
7,877
2.6
Van Buren
3,714
1.7
4,959
2.0
4,859
1.7
5,249
1.8
5,565
1.8
Ionia
I IB
Antrim
711
4.8
965
5.3
1,028
5.1
1,226
5.4
1,340
5.4
Benzie
432
2.9
483
2.7
535
2.7
581
2.6
610
2.5
Charlevoix
912
6.1
1,490
8.2
1,702
8.5
2,120
9.4
2,467
10.0
252
Appendix Table 12- -Continued
County
Emmet
1958
1963
1967
1980
2000
656
4.4
933
5.1
846
4.2
949
4.2
1,016
4.1
2,082
14.0
2,291
12.6
3,015
15.0
3,379
15.0
3,740
15.1
107
.7
126
.7
224
1.1
228
1.0
314
1.3
38
.3
41
.2
163
.8
181
.8
250
1.0
Leelanau
229
1.5
195
1.1
118
.6
114
.5
111
.4
Manistee
1,871
12.6
2,533
13.9
2,511
12.5
2,853
12.6
3,078
12.4
Mason
1,899
12.8
2,358
13.0
2,169
12.5
2,290
10.1
2,341
9.5
Mecosta
1,003
6.8
1,267
7.0
1,604
8.0
1,886
8.3
2,134
8.6
22
.1
27
.1
38
.2
19
.2
21
.2
1,826
12.3
1,758
9.7
1,814
9.0
1,733
7.7
1,654
6.7
Oceana
487
3.3
434
2.4
263
1.3
254
1.1
242
1.0
Osceola
947
6.4
1,417
7.8
2,052
10.2
2,607
11.5
3,098
12.5
Roscommon
116
.8
139
.8
188
.9
252
1.1
275
1.1
1,497
10.1
1,741
9.6
1,809
9.0
1,909
8.4
2,002
8.1
Grand Traverse
Kalkaska
Lake
Missaukee
Newaygo
Wexford
Appendix Table 12— Continued
County
1958
1963
1967
1980
2000
IIIA
9,315
9.8
9,630
8.6
10,290
8.2
10,257
7.5
10,360
7.0
Genesee
57,536
60.5
66,918
60.1
76,459
60.8
83,102
60.8
90,437
60.9
Gratiot
2,970
3.1
3,657
3.3
3,345
2.7
3,482
2.5
3,551
2.4
Huron
1,428
1.5
1,362
1.2
1,684
1.3
1,673
1.2
1,749
1.2
Lapeer
1,151
1.2
1,957
1.8
1,830
1.5
2,241
1.6
2,541
1.7
Saginaw
21,201
22.3
25,771
23.1
29,793
23.7
33,146
24.3
36,694
24 .7
Tuscola
1,484
1.6
2,073
1.9
2,315
1.8
2,717
2.0
3,082
2.1
Bay
11 IB
Alcona
84
.4
193
.9
319
1.3
399
1.6
487
2.0
Alpena
3,451
17.3
2,908
14.0
2,525
10.5
2,322
9.3
2,032
8.2
Arenac
200
1.0
231
1.1
259
1.1
236
1.0
248
1.0
Cheboygan
325
1.6
463
2.2
1,010
4.2
1,263
5.1
1,404
5.7
Clare
408
2.0
672
3.2
1,073
4.5
1,334
5.4
1,551
6.2
Crawford
294
1.5
268
1.3
799
3.3
617
2.5
416
1.7
58
.3
152
.7
330
1.4
482
1.9
578
2.3
Gladwin
Appendix Table 12— Continued
County
1958
1963
1967
1980
2000
Iosco
482
2.4
506
2.4
577
2.4
565
2.3
551
2.2
Isabella
716
3.6
535
2.6
555
2.3
502
2.0
688
2.8
13,004
64.9
13,843
66.5
14,793
61.5
14,983
60.3
14,538
58.5
74
.4
114
.5
353
1.5
499
2.0
591
2.4
Ogemaw
289
1.4
340
1.6
468
1.9
532
2.1
567
2.3
Oscoda
80
.4
36
.2
92
.4
82
.3
78
.3
Otsego
371
1.9
439
2.1
755
3.1
910
3.7
1,012
4.1
Presque Isle
204
1.0
130
.6
135
.6
123
.5
108
.4
Midland
Montmorency
IV and V
Alger
906
6.3
794
6.0
674
5.5
629
5.1
576
4.7
Baraga
229
1.6
663
5.0
710
5.8
989
8.1
1,259
10.3
Chippewa
775
5.4
580
4.4
247
2.0
231
1.9
211
1.7
Delta
2,174
15.2
2,337
17.6
2,546
20.8
2,680
21.9
2,777
22.7
Dickinson
1,951
13.6
986
7.4
1,231
10.1
1,062
8.7
993
8.1
795
5.6
713
5.4
857
7.0
786
6.4
763
6.2
Gogebic
255
Appendix Table 12— Continued
County
1958
1963
Houghton
1,293
9.0
1,091
8.2
726
5.9
678
5.5
621
5.1
Iron
229
1.6
608
4.6
254
2.1
364
3.0
372
3.0
Keweenaw
115
.8
113
.9
100
.8
94
.8
86
.7
Luce
177
1.2
203
1.5
149
1.2
139
1.1
128
1.0
84
.6
93
.7
56
.5
54
.4
50
.4
Marquette
2,223
15.5
1,761
13.3
1,111
9.1
1,035
8.5
950
7.8
Menominee
2,573
18.0
2,345
17.7
2,830
23.2
2,770
22.7
2,757
22.6
Ontonagon
434
3.0
656
4.9
452
3.7
445
3.6
432
3.5
Schoolcraft
341
2.4
327
2.5
276
2.3
263
2.2
244
2.0
Mackinac
a 263 absolute.
2.2 per cent of total.
1967
1980
2000
Appendix Table 13
Manufacturing Employment as a Proportion
of Total Employroenta
Year
1968
1967
1966
1965
1964
1963
Proportion
.32
.34
.36
.35
.35
.35
Year
Proportion
1962
1961
1960
1959
1958
1957
.34
.33
.35
.35
.34
.37
U n p u b l i s h e d statistics from the Michigan Employment
Security Commission, Lansing, Michigan.
Appendix Table 14
Regression Coefficients for Lower Peninsula Regions
Region
1
IIA
IIB
I IIA
11 IB
Constant
-23,093,986
-13,836,324
- 1,279,273
- 7,465,795
- 1,003,640
Regression
Coefficient
16,646
7,178
661
3,861
523
Correlation
Coefficient
.94
.97
.99
.99
.90
257
Appendix Table 15
ucgrcssion Coefficients for Counties
Constant
Regression
Coefficient
Correlation
Coefficient
50,702
212,142
69,401
12,608
- 107,043
- 156,781
- 199,602
21,878
-1,190,072
81,428
- 192,909
- 173,148
- 144,879
- 142,605
- 104,742
78,069
- 160,410
43,381
-4,038,948
58,217
10,859
- 196,024
86,154
- 223,643
50,986
-2,242,997
19,732
10,962
- 976,213
- 841,459
24,622
-2,928,928
26,110
- 153,324
- 634,131
- 168,118
-5,688,791
141,775
61,709
- 128,647
- 373,345
3,398
- 153,759
- 595,976
26
110
36
7
55
81
107
11
618
43
100
89
74
73
54
41
83
23
2,092
30
6
101
46
115
28
1,157
10
6
506
442
13
1,518
13
79
329
87
2,936
73
32
66
197
2
82
306
.99
.92
.96
.99
.93
.95
.97
.99
.99
.83
.99
.98
.92
.98
.81
.99
.98
.72
.99
.97
.37
.93
.60
.99
.71
.93
.94
.12
.97
.92
.91
.98
.84
.82
.85
.97
.96
.88
.64
.99
.99
.96
.91
.91
County
Alcona
Allegan
Antrim
Arenac
Daraga
Barry
Bay
Benzie
Berrien
Branch
Cass
Charlevoix
Cheboygan
Clare
Crawford
belta
baton
Emmet
Genesee
Gladwin
Gogebic
Grand Traverse
Gratiot
Hillsdale
Huron
Ingham
Iosco
Iron
Jackson
Kalamazoo
Kalkaska
Kent
Lake
Lapeer
Lenawee
Livingston
Macomb
Manistee
Mason
Mecosta
Midland
Misaukee
Monroe
Montcalm
-
258
Appendix Table 15--Continued
Constant
Regression
Coefficient
Correlation
Coefficient
58,812
- 500,309
-9 ,319,868
37,890
6,822
- 237,185
1,732
80,881
808,175
15,280
-1 ,844,833
341,766
278,617
239,480
467,925
181,311
255,397
-3 ,698,068
-7 ,179,912
67,494
30
268
4,784
19
4
122
1
41
419
8
953
179
146
123
241
93
132
1,898
3,846
35
.90
.99
.98
.95
.14
.99
.14
.91
.99
.96
.99
.78
.91
.97
.99
.98
.86
.99
.85
.96
County
Montmorency
Muskegon
Oakland
Oqemaw
Ontonagon
Osceola
Oscoda
Otsego
Ottawa
Roscommon
Saginaw
St. Clair
St. Joseph
Sanilac
Shiawassee
Tuscola
Van Buren
Washtenaw
Wayne
Wexford
—
259
Appendix Table 16
Derivation of State SIC Projections
2
R
19
.83
20
21
22
23
24
25
26
27
28
29
.79
.92
.51
.15
.37
.14
.03
.03
a
Derivation
2
SIC
a
c
a
c
c
c
b
b
b
b
b
30
31
32
33
34
35
36
37
38
39
Deriation
R
.91
c
a
b
c
c
c
c
c
c
a
.44
.84
.74
.68
.52
.76
.60
Missing data, taken at 1968 proportion
b
Weak trend, taken at 1968 proportion
c
Strong trend, time series regression, proportional
deflation
Appendix Table 17
Regression Coefficients for State SIC Employment
SIC
20
22
23
24
30
33
34
35
36
37
38
Constant
2,167,209
235,234
- 2,293,174
415,205
- 1,861,958
- 5,448,579
- 6,571,920
- 7,385,764
- 2,422,061
-24,240,613
- 1,110,291
Regression
Coefficient
Correlation
Coefficient
- 1,706
121
1,176
217
958
2,818
3,395
3,834
1,251
12,503
571
.91
.88
.95
.71
.95
.91
.86
.82
.72
.87
.77
Appendix Table 18
Water Kecirculationa
1964
SIC
16
3
148
151
3
101
177
38
163
16
19
332
6
6
3
86
29
13
Gross
Water
Used
Reuse
Ratio
23
19
311
217
4
243
215
79
336
18
24
475
10
14
3
192
76
22
1.43
6.33
2.10
1.43
1.33
3.19
1.21
2.07
2.06
1.12
1.26
1.43
1.66
2.33
1.00
2.23
2.62
1.69
Total
Water .
Intake
12
3
135
140
3
95
241
18
127
12
64
257
6
11
2
86
23
14
^J.S. Bureau o£ the Census, Census of Manufacturers; 1963. Subject Statistics: Water Use in Manu
facturing (Washington: U.S. Government Printing Office, 1966).
Gross
Water
Used
19
44
182
184
4
244
278
38
218
14
68
375
7
22
4
155
60
20
Reuse
Ratio
1.58
14.66
1.34
1.31
1.33
2.56
1.15
2.11
1.71
1.16
1.06
1.45
1.16
2.00
2.00
1.80
2.60
1.42
Per Cent Change
Reuse Ratio
1959-1964
- 9.49
-56.82
■*■56.71
+ 9.16
.00
+24.60
+ 5.21
- 1.89
+20.46
- 3.44
+18.86
- 1.37
+43.10
+16.50
-50.00
+23.88
+ .76
+19.01
Billion gallons
Based on Michigan data
Based on national data
c
d
d
d
d
c
c
c
d
c
c
c
c
c
c
c
d
d
260
20
21
22
24
25
26
28
29
30
31
32
33
34
35
36
37
38
39
Total
Water .
Intake
1959