Citation
The evolution of industrial land use with the Knoxville metropolitan region

Material Information

Title:
The evolution of industrial land use with the Knoxville metropolitan region
Creator:
Honea, Robert B
Copyright Date:
1975
Language:
English

Subjects

Subjects / Keywords:
Economic models ( jstor )
Employment ( jstor )
Geodetic position ( jstor )
Industrial plant locations ( jstor )
Kurtosis ( jstor )
Land use ( jstor )
Land use change ( jstor )
Modeling ( jstor )
Statistical discrepancies ( jstor )
Statistical mode ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
ADB3750 ( ltuf )
00585118 ( ALEPH )
14184163 ( OCLC )

Downloads

This item has the following downloads:


Full Text












THE EVOLUTION OF INDUSTRIAL LAND USE WITHIN
THE KNOXVILLE METROPOLITAN REGION: AN ANALYSIS
USING AERIAL PHOTOGRAPHY AND HISTORICAL DATA
FOR THE PURPOSE OF LAND USE MODELING







By




ROBERT B. HONEA


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY





UNIVERSITY OF FLORIDA
1975



































DEDICATED TO:

Dr. (Jim) James C. Wilkinson










AC KNOWLEDGEMENTS


Of the many individuals who have contributed to this study,

foremost is Dr. Clark I. Cross, Chairman of my Supervisory Conurttee.

I am appreciative for his professional and personal assistance. His

periodic encouragement has been most opportune. Dr. Gary Shannon has

contributed both to this study and at other stages of my graduate

career. I am grateful for his help. Dr. James R. Anderson has been a

source of encouragement before and during my graduate tenure. To him

I extend my appreciation. To Dr. David L. Niddrie 1 extend my appre-

ciation for many engaging discussions. His intellectual vigor is a

source of professional inspiration. The lectures of Dr. Clayton

Curti. provided many ideas which arc :c uded in tniS study. To him

my sincere appreciation is extended. Gratitude is also extended to

Dr. Sh3nnon McCune, Chairman of 'he De'partment of Geography and other

menibers of the faculty for their assistance.



who contributed to the research effort: Mr. C. T. N. (Ted) Paludan at

Marshall Space Flight Center; Dr. C. W. (Pete) Craven, Mr. Richard C.

DurCee, anid Dr. Michael M. McCarthy at Oak' Ridge National La:boratory;

rand > i~mcT.'ily ,.'ho patiently occupied tlhe remainder -f ::y huro while i.

hibern ated in my office.










TABLE OF CONTENTS


Page


ACKNOWLEGEMENTS . . .

LIS" OF TABLES . . .

LIST OF ILLUSTRATIONS. . .

ABSTRACT . . .

CHAPTER

I. INTRODUCTION . . .


- i-i


. . . . vii.

. . . ix


Problem Background ..
Statement of Problem .
Problem Operationalization
Model Operation . .
Motivation for Study .
Organization of Study. .


II. THEORETICAL AND EMPIRICAL ST'IDIES RELEVANT 10 'THE
DEVELOPMENT OF INDUSTRIAL SITE-SELECTION ALGORITHM .

Selected Theoretical Works . .

'er IsolerLe Stat . . ..
lUber den Standort der Iridustr -cn .. .....
The Location of Economiic Rehavior. ..
Imperfect Competition Ir.dI the Divlnonl., neb,) pr ..
'ihe Theories of David M. Smith aiu
Melvin L. Greenhut . . .....

Empirical Studies . .

The Studies of Allen Prod and Richard Londale .
Historical Studies of Industrial Location Factors.

Land Use Modeling Studies. . . . .
4A icrtative List of Variables. ..

III. DESCRIPTION UF THE STUDY AREA AND ANALYSIS PROCEDURES. .

The Study Region . . . .....

Industry within the Region . . ..

Food products industries (SIC 20) ..
Tobacco products industries (SIC 21) .
Textile mill product-; industries (SIC 22).


19
22

5 32
35


39
S 39

. '1


. . . .
S. . . .









Page

Clothing and related products (SIC 23). .. 58
Lumber and wood products except furniture
(SIC 24) .. .. . . 58
Furniture and fixtures industries (SIC 25) 64
Paper and pulp products industries (SIC 26) 68
Printing and publishing industries (SIC 27) 68
Chemical products industries (SIC 28). . 6
Petroleum refining and paving and roofing
products (SIC 29). . . . 74
Rubber and plastic products industries (SIC 30). 74
Leather products industries (SIC 31) ..... 74
Stone, clay, and gland; products industries
(SIC 32) . . . . .. 81
Primary metal industries (SIC 33). . . 81
Fabricated metal prodLcts industries (SIC 34). 81
Machinery (except electrical) industries
(SIC 35) .... . . .. ... 87
Electrical machinery, equipment, and supplies
industries (SIC 36). ... ... . 87
Transportation equipment industries (SIC 37) .. 94
Instrument and related products industries
(SIC 38) .. ...... ...... 94
Miscellaneous industries (SIC 39). ... .... 94

Data Collection. . . ... . . 101
Statistical Procedures . . . 106

IV. EMPIRICAL ANALYSIS . . . . . ll

Descriptive Statistic' .. .. I.1

Description of Sample. . .. ... .. .. 113
Description of Variable Measurements .. ...... 112

Results of the Factor Analysis .. .. .. ... 148

V. SUMMARY AND CONCLUSIONS . . . .. 158

Summary. ... . . . . 158
Conclusions. .. . . . . 159

Conclusions Regarding the Form of the Site-Selection
Algorithm . . . ... . .. 159
Conclusions Regarding the Proper Form of Model
Operation . . . . . 160
Conclusions Regarding the Use of Aerial Photography
in Model Design. . . . . 160
Conclusions Regarding Recommendations for Future
Research . .. . . 161







vi


Page

RESEARCH BI LIOGRAPHY. . . . . 163

Periodicals ........ ................ 163
Books . . . . . . . 168
Government Doci..,ients and Agency and Institute Research
Reports. . . . . ..... 172
Unpublished Sources. . . . .... .177

APPENDICES

A. Sample Survey Forms, . ... .. . .. 179
B. Map Overlays ........ .. ............. .182

BIOGRAPHICAL SKETCH . . . ..... .. 184










LIST OF TABLES


Table

1. Factors Most Influential in the Location Decisions
of Florida Industries, 1956-1957 . . .

2. List of Possible Factors Influencing Industry Location
as Utilized in the Carrier and Schriver Survey .

3. Ten Factors Most Frequently Mentioned by Tennessee
Firms as Affecting Location Decisions . .

4. Ten Location Factors with Highest Mean Number Points
Assigned by Tennessee Firms Interviewed. . .


Page


. 24


. 26


. 27


. 28


Location Factors Suggested by Bullington .

Tentative List of Variables . ...

Expansion of ETDD Industries Between 1952-1973

Distribution of Sample by SIC Category .


9. Variable -

i0. Variable -

L1. Variable -

12. Variable -

!3. Variable -

14. Variable -

15. Variable -

16, Variable -
Half Mi.lcs

i7. Variable -

[8. Variable -

[9. Variable -

20. Variable -

21. Variable -


Slope of Land . . . . .

Drainage . . . . .

Clearing-Cover Condition . . .

Distance to Center of Town . . .

Distance to Major Throughfare . . .

Density of Land Use in Immnediate Vicinity .

Rating of Price of Land . . .

Proportion of Urban Area Within Two and a


Distance to Major Highway . . .

Distance to Secondary Road . . .

Distance co Railway . . . .

Distance to Waterway . ......

Distance to Airport . . . .


31

37

53

313

114

115

117

118

119

120

122


123

12b

126

127

128

130







viii


Tab e Page

22. Variable Distance to Interstate. . . . 131

23. Variable Oveiall Quality of Accessibility-Then . 133

24. Variable Overall Quality of Accessibility-Now. ... 134

25. Variable City Water Availability . . ... 135

26. Variable City Sewage Availability. . . ... 135

27. Variable Gas Availability. . . .. . 136

28. Variable -- Did Community Have Zoning Then? ... .136

29. Variable Was Site Zoned for Industry?. . .. 137

30. Variable Was Industry Already in Area?. . 137

31. Variable Overall Rating of r'oltigous Land
Use Compatability .... . . ... 159

.'. Vari.able Condition of NeighL ohood ... . 140

33 Yariable Density of Land Use ... .. .. .. ... 141

34. Variable Nearby Community Service. . . .143

55. Variable Was the Site in an Industrial Park? ... .. 144

6. c.'" ibje Was was o Qu .ality of the i'a-?. . 1J,4

37. Variable Proximity to Knoxville. . . ... 46

33. Variable Amount of Other Industry Located Nearby 147

39. Correlation Matrix ... .... .. . 149

A(j. Eigen Values ..... . .. . .. 152

41, Unrotated Factor Matrix. .. . .. .. 154

42. Varimax Rotated Factor Matrix. . . . .. 155











LIST OF FIGURES


Figure

1.
i

2

3.

4.

5.

6.

7.

8.

9.

'10.

11.

12.

13.




15.
7l.






17.

18.

19.

20.

21.

22.

23.

24.


Food Products (Plant It!:.ti

- Food Products (Employment)

- Textile Products (Plant Loc

- Textile Products (Employmen


Urban Land-Use Location Scheme. .

Smith's Model with Comparisons. .

East Tennessee Development District

Physiography . . .

Towns and Communities . .

Plant Location 1943 .

Employment 1943 . . .

Plant Location 1953 ....

Employment 1953 . ....

Plant Locations 1963 .. .

Employment 1963 . .

Plant Locations 1973. ..

Employment 1973 . . .


Apparel and TRelared 'roduci. (Plant L.oc:ti

- Apparel and Related Products (Employricent).

- Lumber and Wood Products (Plant Locacions)

-Lumber and Wood Products (Employment). .

- Furniture and Fixtures (Plant Locations) .

- Furniture and Fixtures (Employment). .

- Paper and Pulp Products (Plant Locations).


ons)


61

62

63

65

66

67

69


Page

. . 11

. . 17

. 40

. . 42

. . 43

. . 45

. 46

S . . 47

. . 48




. 50

. . 51

. .. 52

on,; . 56

. . 57

nations) . 59

t). . . 60


SIC

SIC


SIC


siC
SIC

SIC

SIC

SIC

SIC

SIC









Figure Page

25. SIC 26 Paper and Pulp Products (Employnent) . .. 70

26. SIC 27 Printing and Publishing (Plant Locations). .. 71

2-. SIC 27 Printing and Publishing (Employment) . .. 72

28. SIC 28 Chemical Products (Plant Locations). . ... 73

29. SIC 28 Chemical Products employmentt) . ... '75

30. SIC 29 Petroleum Refining and Paving and Roofing
Products (Plant Locations) ...... 76

31. SIC 29 Petroleum Refining and Paving and Roofing
Products (Employment). . . ... 77

32. SIC 30 Rubber and Plastic Products (Plant Locations) 78

33. SIC 30 R bber and Plastic Products (Fmploymet) 79

34. SIC 31 Leather Products (Plant Locations) ... 80

35. SIC 31 -Leather Products (Employn.ent) .. .... 81

36. SIC 32 Stone, Clay and Glass Products (Plant
Locations) . . . . 83

37. SIC 32 Stone, Clay and Glass Products (Employment). 84

38. STC 33 Primary cetal T.c'ustcies (Plant I.ocation) 85

39. SIC 33 Primary Metal Industries (Employment).. . 86

40. SIC 34 Fabricated Metal Products (Plant Locations). 88

41. SIC 34 Fabricated Metal Products (Employrment) .. .. 89

47. IC .'S MAchi nery ,Pia.nt Locations) ....... .. 90

43. SIC 35 Machinery (Employ.ent) . . 91

44. SIC 36 Electrical Machinery, Equipment and
Supplies (Plant Locations) ... ..... 92

45. SIC 36 Electrical Machinery, Equipment and
Supplies (Employment) . . . 93

46. SIC 37 Transportation Equipment (Plant Locations) 95







xi


Figure Page

47. SIC 37 Transportation Equipment (Employment). ... 96

48. SIC 38 Instruments and Related Products (Plant
Locations) ... . . . 97

49. SIC 38 Instruments and Related Products (Employment). 98

50. SIC 39 Miscellaneous Industries (Plant Locations) 99

51. SIC 39 Miscellaneous Industries (Employment). ... 100

52. 1958 Stereo Image of Beta-Tok Industrial Site .... .102

53. 1974 Stereo Image of Beta-'ek Industlial Site .103









Abstract of Dissertation Presented to the Graduate Council
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy

THE EVOLUTION OF INDUSTRIAL LAND USE WITHIN
THE KNOXVILLE METROPOLITAN REGION: AN ANALYSIS
USING AERIAL PHOTOGRAPHY AND HISTORICAL DATA
FOR T1HE PURPOSE OF LAND USE MODELING

By

Robert B. Honea

March, 1975

Chairman: Dr. Clark I. Cross
Major Department: Geography

This research effort represents a portion of a much larger

research goal, the identification of the determinants affecting land

use change. This study focuses, however, upon only one aspect of this

problem, the conversion of land to industrial use. The approach

assumes that the determinants of Lhe land use conversion process are

found in the "market place," where land transactions among buyers and

sellers occur. Only one side of the market transaction process is

studied, however, namely that of ite purchaser's desires (in this

case, the industrial developer) in securing an ideal or suitable site.

The problem was to identify the ideal qualities, quantities or attri-

butes desired in an industrial site and to formulate a general algo-

rithmic statement to identify potential industrial sites.

Research procedures involv'-!d ihe developing o7 a .ist of var:iabl es

previously noted in the literature to be related to industrial site

selection and streamlining the list to a set suitable for statistical

testing. A sample of 157 industrLes which have located (or relocated)

in the 16-county Knoxville metropolitan region since 1950 was selected

for analyst s. Ui current a&nd 1'slio-. al a:..i l. ,,hctrn.ra ch data






xiii


were collected on the location characteristics of each industrial

site. These data were then subjected to factor analysis to determine

the interrelations of variables, to minimize the list of variables

needed to describe the industrial site-selection process, and to

determine if the preconceived ideas concerning the factors affecting

the process were valid. Seven factors accounting for 72 percent of

the variance found in the original data were identified.

The 30 variables studied did not group as previously conceived

but nevertheless the factors present a logical arrangement of vari-

ables. Four factors are almost singularly identified by accessibility

characteristics. Reexamination of the accessibility factors indicates

they could be combined, thereby reducing the number of indexes to two

or three. The results suggest that a zoned or designated place for

industry to locate, open space suitable for industrial development,

and accessibility to the site acc the three most important considera-

tions in industrial site-selectio~i decisions. Any algorithm intended

to simulate industrial land use coriversion should be constructed with

these parameters as major components.















CHAPTER I

INTRODUCTION


Problem Background


Most urban and regional planners seek to answer the question,

"What would happen if .?" The answer is not simply determined.

"Even though equipped with the best information system, no .

decision maker, however adept, can anticipate or visualize the many

relationships which must be considered in a comprehensive, regional

plan."1 If, however, we could specify the correct development pro-

cesses and represent them symbolically as a set of equations, the

computer can approximate through simulation, years of regional

growth in a matter of minutes. Before we can accurately simulate

regional growth, the development processes must first be defined and

understood, and by observing the limited success of previous land use

modeling efforts we become acutely aware that we do not sufficiently

understand them.2




1Charles R. Meyers, Jr., "New Tools for Regional Planning," AIA
Journal, October, 1971, p. 1.

2Models should not be judged simply on a success-failure basis
but rather in terms of the models' ability to answer questions about
real world. In the past, some models have been more successful than
others. For a review of previous modeling efforts one should consult:
Ira S. Lowry, Seven Models of Urban Development: A Structural Com-
parison (Santa Monica, California: RAND Corporation, 1967); B. W.
Mar and IV. T. Newell, Assessment of Selected RANN Environmental Modeling
Efforts. A report prepared for the Environmental Systems and Resources









We would expect land use change or growth to be stimulated or

brought about by certain "triggering" events or mechanisms. Many

early land use modeling efforts assumed the "triggering" mechanism to

be brought about by the location of new industry. We now recognize

that other stimulating mechanisms such as improvements in transpor-

tation or expanding demand for personal services also may prompt

regional growth.3 This research problem, however, is predicated upon

the assumption that industry (specifically export industry) is the

major stimulus for urban and regional growth and thus deserves initial

consideration in any modeling effort.

Brown et al. recognized "the critical importance of industrial

location. .yet land-use modelers have devoted surprisingly little

effort to analyzing the determinants of industrial location."4 Only

a few research efforts have made serious attempts to model the basic

determinants of industrial location choices and "for the most part

these attempts have been quite limited and crude."5 The focus of




Division, National Science Foundation (Seattle: University of Washing-
ton, June, 1973); and 0. Stradel and B. G. Hutchinson, Notes for a
Short Course on Practical Applications of Regional Development Models
(Waterloo, Ontario: University of Waterloo, The Transport Group,
Department of Civil Engineering, November, 1971).

3Homer Hoyt was among the first urban economists to study the
growth and development of urban land. His primary axiom was that "no
city could grow by taking in its own wash." Homer Hoyt, The Structure
and Growth of Residential Neighborhoods in American Cities (Washington,
D. C.: U. S. Government Printing Office, 1939).

4james II. Brown and others, Empirical Models of Urban Land Use:
Suggestions on Research Objectives and Organizations, Exploratory
Report 6 (New York: Columbia University Press, 1972), p. 82.

Ibid.









this study is on the conversion of land from prior uses to industrial

use.

Statement of Problem


An initial requisite of any model purporting to replicate and

predict industrial land use development is the identification of those

land parcels having the greatest probability to change to industrial

land use. It is necessary, therefore, to construct an algorithm which

calculates the suitability of some land parcel for a specific indus-

trial land use. The primary objective of this research effort is to

develop a site-selection algorithm which identifies land parcels

having the highest probability of attracting industrial land use.

More specifically this research attempts to identify and measure the

importance of the intrinsic site characteristics which appear to

control the conversion of land to industrial use within the 16-county,

Knoxville metropolitan region and to formulate a simulation algorithm

to identify potential industrial sites.


Problem Operationalization


Intrinsic site characteristics refer to the qualities possessed

by some land parcel which identify its suitability for a specific land

use. In the case of industrial land use, examples of some of the

attributes which should be examined are: proximity to rail service;

access to nearby highways; site preparation costs; availability of

city services; or whether the site is within a developed industrial

park. The importance of these attributes may be measured and subse-

quently aggregated to form a combined measure or index.









The following is a list of indexes which are often identified as

affecting industrial land use choices:

Site Preparation Costs--The cost of converting rural or open
land to manufacturing use. Matters which might be taken
into account are foundation conditions, drainage conditions,
slope of the land, and type of vegetation cover. This
factor does not include the purchase price of the land.

Market Price of Land--The per-acre cost of land to the
manufacturer who wishes to build a new plant or expand his
existing plant. This would include property taxes but not
necessarily taxes on improvements to the property.

Proximity to Suitable Labor Force--The number of suitable
workers within easy access of the site.

Transportation Accessibility--The ease with which people or
materials can be moved from the plant site to roads, rail-
roads, airports, or waterways. In this factor, the concern
is with the linkage between the plant and major transporta-
tion facilities within the immediate area.

Utilities--The kinds and quality of utilities available at
the potential plant site. Some utilities which might be
considered are water, sewer, and gas.

Compatibility with Existing Land Uses--The compatibility of
general manufacturing activity with other existing land uses
adjacent to the potential site.

Neighborhood or Community Attractiveness and Amenities--The
condition and density of dwelling units and business estab-
lishments in the immediate area of the plant site, and
proximity to hospitals, schools, parks, and churches.

Industrial Park Space--The availability of suitable buildings
or land in an industrial park. The appearance of the park
and the quality of services produced are some of the con-
siderations to be examined.

This list was compiled from both a review of literature and

preliminary study. Each factor is a composite of several variables

collectively identified as an index. Particular combinations of

variables are not unique but simply represent logical groupings based

upon the literature search. It is convenient for the purpose of this









study to utilize this list as a beginning point to be refined after

further study.

Each index posited is considered to reflect an information unit

to be evaluted in the selection of industrial sites at the intra-urban

scale. By combining these indices it is possible to develop an

aggregate measure of the suitability of a site for industrial use.

Before doing so, however, a weight should be attached to each index to

reflect its relative importance in the location decision. A general

mathematical form for the site selection algorithm follows:

N
LUS. W. I.
i=l i, ,


where LUS = Attractiveness score for land use category 2
(in this case k = industrial land use),

N = Number of Indices,
th th
W. = Index weight for the i index and the th land
use category, and

.th th 6
I. = i index for the Z land use category.

-7
Model Operation


In modeling industrial land use development attractiveness scores

are utilized in the following manner: Projected industry growth




Oak Ridge National Laboratory, Regional Environmental System
Analysis, A Research Proposal Submitted to the National Science
Foundation, Research Applied to National Needs (RANN) Feb. 1, 1973,
p. 9.

Land use models are constantly evolvi;ig and thus one can only
speak of the framework of the model in a hypothetical manner. This
rity or may not become the structure of the nodel and represents only
a tentative view as to the operating manner of the iodel.









provided via a socioeconomic model is distributed initially to various

subregions through a subregional allocation algorithm. Industrial

expansion within the subregion is distributed to available land par-

cels on the basis of attractiveness scores calculated in the above

manner. The availability of various land parcels is determined from

knowledge of the probability that specific land parcels will be

offered for consumption.

The land use model as presently conceptualized is perhaps best

characterized as a stochastic (Monte Carlo) model. Projected growth

in land use will be awarded to various land parcels on the basis of

the land use scores (LUSk). Basically, however, the model is heuris-

tic in that as the land use conversion processes are better under-

stood, analytic-deterministic procedures will be substituted for

simulation-stochastic procedures.


Motivation for Study


It has been noted that "urban spatial organization is the outcome

of a process which allocates activities to sites. In our society,




8The assumption that all land is potentially available is not a
valid premise in that not all property owners are willing to sell
property. Cadastral data and ownership characteristics necessary to
develop a site-availability algorithm are not readily available and,
therefore, very few land use models have included this consideration.
The sociopolitical modeling team at Oak Ridge National Laboratory is
currently studying the methods whereby local opinion and knowledge of
owner characteristics may be utilized to identify land parcels which
may be available for various land use activities. See: Osbin L.
Ervin and Charles R. Meyers, Jr., The Utilization of Local Opinion
on Land Use Simulation Modeling: A Delphi Approach (Oak Ridge,
Tennessee: Oak Ridge National Laboratory, ORNL-NSF Environmental
Program, 1973). Olaf Helmar, The Delphi Method for Systemizing
Judgments about the Future (Los Angeles: University of California,
April, 1966).








the process is mainly one of transactions between owners of real

estate and those who wish to rent or purchase space for homes and

businesses." The market process of "transactions between willing

buyers and willing sellers determines the spatial organization of

urban activities ." and thus should dictate the methodological

structure of land use models.10

The market place however is not perfect. Individual land specu-

lation, family or corporate property gifts, over-building by contrac-

tors, land use planners over-estimating demand, governments exercising

property rights, the tendency for land uses to remain intact and the

perpetuation of mistakes rather than corrections, all spoil the simple

modeling of market transactions. However, after investigating numerous

approaches toward modeling land use development, the market place

appears to be the most viable way to approach the simulation of urban

activities.11 The site-selection algorithm seeks only to replicate

the buyer's considerations in selecting an adequate industrial site.

Seller considerations and market perturbations are not considered at

this time.

The results of this research effort are intended to integrate

with a "holistic" environmental model to be developed later. The

motivation which stimulated this research problem, however, was the




9John P. Crecine, Computer Simulation in Urban Research (Santa
Monica, California: The RAND Corporation, 1967), p. 2.
10
Ira S. Lowry, Seven Models of Urban Development: A Structural
Comparison (Santa Monica, California: The RAND Corporation, 1967),
p. 5.
Ibid.










desire to understand the land use development process. After some

preliminary analysis and review of previous research it became appar-

ent that understanding and modeling land use change contained problems

to occupy several dissertations. The decision was made, therefore, to

focus upon one aspect of the land use evolutionary process, the develop-

ment of industrial land use, in the hope that by expanding the under-

standing of this process, rigor could be added to the eventual

development of a comprehensive land use model.


Organization of Study


The organization of this study is: the present chapter attempts

to introduce the reader to the research problem and to rationalize the

need for the study. Chapter II discusses the literature from which a

tentative list of site-selection variables was developed. In Chapter

III, the test region is described and the research methodologies and

procedures explained. Chapter IV presents the results of an empirical

study conducted to identify and determine the relative importance of

various location factors in the 16-county Knoxville region. Finally,

Chapter V summarizes the conclusions derived from the empirical study

and presents recommendations for the construction of a site-selection

algorithm and recommendations for further research.















CHAPTER II

THEORETICAL AND EMPIRICAL STUDIES RELEVANT TO
THE DEVELOPMENT OF INDUSTRIAL SITE-SELECTION ALGORITHM


The purpose of this chapter is to relate to this research effort

ideas expressed within selected theoretical and empirical studies and

to discuss the evolution of the indexes listed in the previous chapter.

Discussion of some works may be abbreviated depending upon their

relevance to industrial site-selection processes. Theoretical works

involving classical location theory are discussed first, followed by

a discussion of empirical studies of industrial location factors.

The final sections discuss approaches utilized in other land use

models and the development of a tentative list of variables.


Selected Theoretical Works


The location of industry in an intrametropolitan area is the

result of a complex interaction of variables that can best be under-

stood only through careful examination of historical events. Classi-

cal location theory, however, abstracts from reality by use of the

principle ceteris paribus where one or two variables are permitted to

vary while all other variables are held constant. Von Thtinen's and

Weber's studies represent early contributions to location theory and,

thus, deserve fir.:t consideration in this study. More recent theo-

retical works follow.









Der Isolierte Staat1

Von ThUnen's work dealt with a hypothetical agricultural land use

system whereby transportation costs and economic rent were used to

explain the location of specific types of agricultural production as

they were concentrically displaced around a market center. In a

similar manner, one might expect industrial, commercial, and residen-

tial land uses to arrange concentrically around an urban center (Figure

1). For a specific land use the utility (accessibility) derived from

any location would decline with increasing distance from the CBD as in

(a). This, of course, assumes that the CBD is the most accessible

point within the urban area. The optimal price one would pay for

utility or accessibility is illustrated in (b). Cost incurred in

obtaining utility (Y) is represented by the area (XYZA) but the profit
2
to be derived at utility (Y) is (ABCZ). The profit is similar to the

land rent of von Thinen's agricultural model. Thus, rent for various

land uses can be expressed as a function of distance from the CBD as

in (c). Based upon the comparative bid-rent capabilities of each land

use, one would expect (RR1) to represent the rent function for

residential land use, (QQ1) for wholesaling and industry land use,




Johann Heinrich von ThUnen, Der Isolierte Staat in Beziehung
Auf Landwirtschaft Und Nationalikonomies (Berlin: Schumacher-Zarchin,
1875) in K. W. Kapp and L. L. Kapp, Eds., Readings in Economics (New
York: Barnes and Noble, 1949).

2Michael E. Eliot Hurst, A Geography of Economic Behavior: An
Introduction, Belmont, California: Duxbury Press, 1972, p. 231.






Page
Missing
or
Unavailable









and (PP1) for commercial and service activities. The intersections at

Y and Z would form the boundaries between the various land uses.

Commercial activities are the highest bidders for sites and

usually occupy the most accessible places within the city. But commer-

cial land uses are also found in the industrial and residential zones.

Similarly, residential uses may also be found in other zones. For

example, a four-story apartment building may yield several times more

rent as a one-story commercial operation on the same site and thus may

out-bid competitors for the property.4

Bid-rent functions are a very important real world phenomenon

encountered in explaining industrial site selections. Industries

seeking highly accessible sites upon which to build must compete

against other bidders for those same sites. Consequently cost per

unit acre, distance from CBD and proximity to transportation facili-

ties were included as variables to be examined in this study.


Uber den Standort der Industrien5

Alfred Weber was among the first economists to pose a general

theory of industrial plant location. The optimal location for an

industrial plant was seen to be a formation of three factors:




3Ibid.

Ronald Reed Boyce, The Bases of Economic Geography: An Essay
on the Spatial Characteristics of Man's Economic Activities (Atlanta:
Holt, Rinehart and Winston, Inc., 1974), p. 264.

5Alfred Weber, Uber den Standort der Industrien (Chicago: Univer-
sity of Chicago Press, 1928).









transportation costs, labor costs, and agglomerative forces. Weber

theorized that the optimal location would be found:

1. where total transportation costs per unit of output were at

a minimum or,

2. where transportation diseconomies were offset by savings

through agglomeration factors or access to labor.

Of particular interest to this study are Weber's ideas concerning-

agglomerative factors and proximity to labor. In general, the ideas

posited by Weber are of greater importance at the regional or sub-

regional levels of industrial location than at the site level. Some

of Weber's agglomerative factors, however, relate to industrial site

characteristics and, thus, are considered in this analysis.

Weber's agglomeration factors are:

(1) The joint development of industries which promotes the

attraction of auxiliary industries and increases the efficiency of

large scale production and utilization of special technical equipment.

(2) The development and growth of specialized labor due to the

greater opportunity for work in the area.

(3) The greater accessibility to raw material suppliers who can

provide material regularly and on short notice.

(4) The reduction in overhead costs, such as gas, water, elec-

tricity, roads, and communications.6




6Robert G. Turner, "General Theories of Plant Location: A
Survey," AIDC Journal, VI (October, 1971), pp. 25-26.










The Location of Economic Behavior7

The A.\nercan economist, Edtgar Iloover initially attempted to

improve i''eber's explanation of industrial location )by including the

consideration of such variables as the size of mIarket area and insti-

tutional forces in his theoretical analysis. Yet Hoover, although

critical of Weber's analysis, ultimately proceeded along the same

lines to present his location theory primarily in terms of costs,

Hoover, however, did expand Weber's agglomerative forces to include

the importance of banks, utilities, fire and police protection,
8
climate, property tax, and lower interest rates.

Hoover also developed some generalizations in the locational

habits of light industry versus heavy industry. Because of the

necessity of "handling of large quantities of goods either coming in

from elsewhere or being shipped out, heavier types of manufacturing,

warehousing, and wholesaling prefer locations in transshipment zones

along rail or waterways." "Manufactures, wholesalers and warehouses

of less bulky goods need not be located on railroads or waterfronts at

all, since they can be served by truck." They locate more in response

to "the attractions of labor supply, cheap land, and nearness to local

suppliers or customers. As a rule, they are found interspersed with

commercial and inferior residence uses."'




Edgar M. Hoover, The Location of Economic Behavior (New York:
McGraw-Hill, 1949).

Turner, op. cit., p. 27.

Hoover, op. cit., pp. 128-129.








The site considerations of industries of the 1930's and 40's have

changed in more recent times but several basic locational rules as

expressed by Hoover and Weber remain intact. Accordingly, several

variables such as neighborhood amenities, neighborhood compatibility,

proximity to labor force, availability of utilities, adjacent land

uses, and transportation accessibility have been included in this

analysis.


Imperfect Competition and the Duopoly Debate

A number of location theorists believed pure competition was not

a suitable theoretical structure for the study of plant locations, and

sought to explain locations in terms of the competition between two

firms attempting to capture the largest share of a market area.

Fetterl0 and Hotelling11 were among the earliest to expound on
12
duopoly location theory. They were followed by Lerner and Singer,2
13 14
Smithies,13 and Chamberlin4 who expanded the original concept.




10Frank A. Fetter, "The Economic Law of Market Areas," Quarterly
Journal of Economics, XXXVIII (May, 1924), pp. 520-529.

11Harold Hotelling, "Stability in Competition," Economic Journal,
XXXIV (March, 1929), pp. 41-57.

12A. P. Lerner and H. W. Singer, "Some Notes on Duopoly and
Spatial Competition," Journal of Political Economy, XLV (April,
1937), pp. 145-186.

13Arthur Smithies, "Optimal Location in Spatial Competition," The
Journal of Political Economy, XLIX (June, 1941), pp. 423-439.

1E. H. Chamberlin, The Theory of Monopolistic Competition
(Cambridge: Harvard University Press, 1936), pp. 194-196.









Devletogloul recently commented on the economic irrationalit) of

the approach and presented arguments against the theoretical base for

such location activities. Imrpo:rtant contributions, however, are still

to be noted. Duopoly theory sugg-sts consideration should be given to

the repelling or attracting properties of industries which depend upon

local markets.6


The Theories of David M. Smithl7 and Melvin L. Greenhut18

Smith's and Greenhut's contribution to the theory of industrial

location incorporated the minimum costs approach of Weber with the

maximum profit solutions of manufacturing location posited by August

Lisch.19 Smith calls this the maximin solution. The concept devel-
20
oped is illustrated in Figure 2.2 In (a), the costs of production

are permitted to vary over space (distance) and revenue obtained

(demand) is kept constant. This is essentially the Weber solution.

The basic concept of Lbsch's model is shown in (b), where revenue is




1Nicos E. Devletoglou, "A Dissenting View of Duopoly and Spatial
Competition," Economica (May, 1965), pp. 140-160.

1Turner, op. cit., p. 31.

1 DIvid M. Smith, "A Theoretical tFramework for Geographical
Studies oE Industrial Location," Economic Geography XLII (April,
1966), pp. 95-113.
18
18Melvin L. Greenhut, Plant Location in Theory and in Practice
(Chapel Hill: University of North Carolina Press, 1956).
19
William H. Weglom and Wolfgang F. Stalper (Translators), The
Economics of Location, by August Lbsch (New York: John Wiley and
Sons, 1957).
20Smith, op. cit., p. 96.
Smith, op. cit., p. 96.









COSTS
VARYING



/ LOSS
tossI


PROFIT


(a)





o
cr
aC
0
z
4
I-
u
0


DISTANCE -


-COSTS
VARYING


DISTANCE --a b


SMITH'S MODEL WITH COMPARISONS (AFTER SMITH, p. 96)
Figure 2.


WEBER


/


DISTANCE -p-



LOSCH


REVENUE
CONSTANT









permitted to vary over space and costs are held constant. Finally, in

(c) the combined solution suggested by Smith is offered with maximum

profit occurring at A, where costs are lowest and profit is highest.

Note that maximum revenue, however, is obtained at B.21

An interesting variation of Smith's model was the introduction of

noneconomic factors, in particular, the concept of psychic income.22

This innovation permitted social, psychological, or other personal

factors to be entered into the model, hence relaxing the assumption of

economic man. Such considerations according to Smith tend to divert

the location of a plant from the ideal site to locations closer to the

owner's home, a golf course, or perhaps a parochial school.

Smith suggests that stochastic procedures may ultimately have to

be used to simulate industrial location decisions as personal factors

cannot be accounted for by rigorous mathematical reasoning.23 This

research assumes that personal considerations may be accounted for by

noting neighborhood amenities near the potential site. The importance

of housing quality, proximity to churches, hospitals, schools, or

parks and personal services availability in the immediate vicinity of

the site are examined in this study.


Empirical Studies


The following empirical studies were significant resources in the

development of a tentative list of site-selection variables. Two




21Ibid., pp. 96-97.

22Ibid., p. 108.
23David M. Smith, Industrial Location: An Economic Geographical
Analysis (New York: John Wiley and Sons, Inc., 1971), pp. 269-273.









types of studies are presented: those studies directed primarily

toward the industrial developer who seeks new industry for a commu-

nity; and those studies which attempt to anAilyze industrial location

coraiderations by way of a large sample of new industries and to cate-

gorize the locational considerations by industry types. Only one

study is directed toward the development of a site-selection algorithm

for a land use model.

24 75
The Studies of Allen Pred24 and Richard Lonsdale2

Allen Pred has compiled a study of the history and present status

of industrial location decisions within a metropolitan region. These

patterns discussed pose interesting hypotheses for empirical analysis

but the interest of this research is directed primarily to the site

characteristics discussed by Pred. It should be noted that Pred's

analysis focused upon a single metropolitan area whereas this study

encompasses a region with a hierarchy of urban places.

In discussing location patterns, Pred identifies seven types:

1. Ubiquitous industries concentrated near the CBD The market

area of these industries is generally coincident with that of the

metropolis or city. Food processing industries, specifically bakery

goods, package foods, and fresh milk products, are Lome of the examples

of these types of industries.




24Allen R. Pred, "The Intrametropolitan Location of American
Manufacturing," Annals of the Association of American Geographers, LIV
(June, 1964), pp. 165-180.

25Richard E. Lonsdale, "Rural Labor as an Attraction for Indus-
try," AIDC Journal, IV (October, 1969), pp. 11-17.









2. Centrally located "communication-economy" industries Job-

printing industries, newspaper printing, and advertising printing

would be representative industries of this category.

3. Local market industries with local raw material sources -

These industries show a high degree of randomness in their locational

pattern but with some tendency toward CBD locations. Samples would be

ice plants, concrete brick and block industries or industries whose

raw materials are by-products of other large-scale industries such as

the pulp and paper products industry.

4. Non-local market industries with high value products Typi-

cally these industries provide a high value per unit weight product

and are insensitive to transport considerations within the local

region. The pattern is "at least superficially irrational." Computer

and related industries and chemical industries are typical examples.

5. Noncentrally located "communication-economy" industries The

subset of industries includes those which are not necessarily pulled

to any functional area of the city but rather tend to cluster together

in any suitable area primarily because of the necessity to "keep

abreast of the latest innovations or forthcoming contracts." Elec-

tronic, military equipment, and space age industries such as found in

Huntsville, Alabama, or Houston, Texas, are examples.

6. Non-local market industries on the waterfront Industries

where primary raw materials are imported by water or whose finished

product is often moved by water comprise this group. Petroleum

refining, coffee roasting, and sugar refining are prominent among

these types of industries.









7. Industries oriented toward nationLral markets These indus-

tries have extensive market areas and arc influenced by high transpor-

tation costs on a bulky fieiishcd product. A large percentages o-fr those

industries have a tendency to locate d:n t h side of the metrc olis

facing the most important market regii.. Pred identifies the Noway :

Lowland refining indust-ry anid th: i)e;. -c_ utomo(i.ie industry a.s
26
oyanp les.2

Richard Lonsdale in a study oi the locationa.1 habits of rural

industry notes that those industr~i-;. a .fected by transportation coAi.

tend to locate in urban areas whiJ.; ;.:o:;e highly affected by labor

cost, gravitate to rural areas. (t addition, Lonscda.ie note rural

firms tend to space themselves out; ij.t'. .irde- to assure a labor supply

industries wi-Lh tendencies toward i:rl 0oc.-tions are apiparc

food "rodC..ti's, textit, l J. umber :,d-w' .,*/ y r \, t;s,, p"p.cr 'p ourcG" ,;

chen.cal, and electrical machinery :L,-ecially routine assembly. I.o

profit margins, keen competition, and nigh petrc tageO2 of produce in

workers. are s5ome basicc character i.s;i.-. of thest :i J:Lus ries.

The patterns of in T.ustrial location detailed by Fred and Lon:',d:l

are rimari- ; identified with a spatl.. level slightAy higher tha-: j.11

;sit specific l e,'el which is the foc 'r:; this sty In relation i.

the ... jvelcpmen -. 1'e i. to-select i.o-. '.i rithim, '.a La ions in ::.

rerrn re signif. ic r it is obvi'.ort: sep.r t :tings wil

have -o be developeJd for each i ndi.sty type, but this is not a prjim',ry




6Pred, op, cit., pp. 175-1.78,

Lonsdale, :p. cit.









objective of this study. It was anticipated that after developing a

general site-selection algorithm additional research would permit the

matching of various types of manufacturing to specific types of sites.
28
Historical Studies of Industrial Location Factors8

Numerous empirical studies exist of the factors associated with

actual industrial location events. Although many of these studies

were undertaken by academicians, the viewpoint is that of the indus-

trial developer seeking to attract new industry or expand existing

industry within the community.

Characteristically, investigations of this type are not concerned

with a specified theoretical framework for approaching the problem




28Discussion within this section is based mostly upon a review of
the following articles: J. S. Bullington, "Utilization of State-Wide
Site Evaluation Committee to Aide in the Location or Relocation of
Plant Facilities," AIDC Journal, IV (October, 1969), pp. 27-42; James
E. Chapman and William H. Wells, "Factors in Industrial Location in
Atlanta, 1946-1955," Atlanta Economic Review, IX (September, 1959),
pp. 3-8; Ronald E. Carrier and William R. Schriver, "Location Theory:
An Empirical Model and Selected Findings," Land Economics, XLIC
(November, 1968), pp. 450-460, and a more complete explanation of the
study: Ronald E. Carrier and William R. Schriver, Plant Location
Analysis: An Investigation of Plant Location in Tennessee (Memphis:
Memphis State University, 1969); Melvin L. Greenhut and Marshall R.
Colberg, Factors in the Location of Florida Industry (Tallahassee:
The Florida State University, 1962); T. E. McMillan, "Why Manufacturers
Change Plant Location versus Determinants of Plant Location," Land
Economics, XLI (August, 1965), pp. 239-243; N. J. Stefaniak, Indus-
trial Location within the Urban Area: A Case Study of Locational
Characteristics of 950 Manufacturing Plants in Milwaukee County
(Milwaukee: Wisconsin Commerce Reports, 1962); Charles M. Tiebout,
"Location Theory, Empirical Evidence and Economic Evolution," Regional
Science Association, Papers, III (1957), pp. 74-86; U. S. Department
of Commerce, Industrial Location Determinants, 1971-1975 (Washington,
D. C.: U. S. Department of Commerce, Economic Development Administra-
tion, February, 1973); and D. C. Williams and Donnie L. Daniel,
"Industrial Sites for Small Communities," AIDC Journal, VI (April,
1971), pp. 33-39.









but rather to learn the reasons for the location decision from persons

acquainted with the location event. None of these studies were con-

cerned with the construction of an algorithm to simulate the process

of industrial land use development.

Greenhut and Colberg29 analyzed factors influencing the decision

of 400 manufacturers locating in the State of Florida between 1956 and

1957. Location considerations were divided into three groups: demand

(market) considerations, costs (assembly) considerations, and personal

(psychic) considerations. Access to markets and potential markets

(Table 1) rated the highest among the location factors with the

remaining factors surprisingly low. The study, however, was slanted

toward measuring regional and subregional factors and thus was of

limited value to this study.

The extensive study undertaken by Carrier and Schriver30 of

plant locations in Tennessee between 1955 and 1965 was conducted

within the framework of existing location theory and, in part, did

focus upon site-location factors. Many of the variables included

in this analysis are based upon the conclusions reached in this study.

Carrier and Schriver identified six classes of location factors

believed capable of affecting plant locations: (1) personal factors,

(2) procurement-cost factors, (3) processing-cost factors, (4) distri-

bution-cost factors, (5) location demand factors (including locational
31
interdependency considerations), and (6) certainty factors. (This




2Greenhut and Colberg, op. cit.

30Carrier and Schriver, op. cit.

Carrier and Schriver, op. cit., p. 451.




















Percentage of 400 Plant
Location Factors Listing as Primary Factor

Access to markets 51.9

Anticipation of market growth 12.8

Good labor relations 1.7

Lower wages 2.6

Ease of attracting out-of-state
personnel, including research 4.7

Low freight cost on obtaining raw
materials and components 7.7

Low cost on freight on shipping final
product 10.7

Climate as it affects operations 1.8

Community facilities (education, police,
medical, etc.) 2.9

All other factors 3.2


FACTORS MOST INFLUENTIAL IN THE LOCATION
DECISIONS OF FLORIDA INDUSTRIES, 1956-1957

TABLE 1









class was identified after the interviews.) Certainty factors were

defined as the confidence that the "prevailing and forecasted data used

to identify the site offering maximum profits would persist into the

future."32

Persons involved in the selection of sites for 307 manufacturing

plants were interviewed. Each respondent was asked to select six

factors from those listed in Table 2 and to distribute 100 points among

these six in order to indicate the relative importance each factor

contributed to the total plant location decision.

Of the 36 factors listed in Table 2, low cost and availability of

labor was mentioned most frequently as the primary factor affecting

the location decision (Table 3). Personal considerations without

economic advantages received the highest average number of points,

followed by low cost and availability of labor (Table 4).

On the basis of the interviews the authors grouped industries

according to the six factors previously listed:

(1) Personal Factors Miscellaneous manufacturing, furniture and

fixtures, and food and kindred products were highly sensitive to

personal factors with most of those firms being "home-grown."

(2) Procurement-Cost Factors Industries which need large

volumes of low-unit-value or perishable raw materials were character-

istically affected by this group of factors. Food and kindred prod-

ucts, stone, clay and glass products, and lumber and wood products

industries indicated greater sensitivity to these factors.



32bid.
Ibid.









1. Personal Factors;

Personal with economic
advantages
Personal without economic
advantages

2. Procurement-Cost Factors:

Better service from seller
of raw materials and
components
Low cost on raw materials
or components
Availability of low cost
raw materials

3. Processing-Cost Factors:

Low cost and availability
of labor
Low cost of fuel
Low cost of electric power
Low cost of financing project
through Area Redevelopment
Administration
Climate
Favorable labor-management
relations
Low cost of satisfactory
type of water
Adequate waste disposal
Low cost of building and
land
Low cost of financing plant
through revenue or
general obligation bonds
Favorable community and state
tax structure
Community concessions
Available existing plant
Available existing building
Particular characteristics
of building site


4. Distribution-Cost Factors:

Low freight cost,
finished product

5. Location-Demand Factors:

Greater demand in
the area
Greater demand poten-
tial in the area

6. Certainty Factors:

Nearness to metro-
politan city
Community facilities
Community planning
and zoning laws
Cultural qualities
of the town
Community leaders'
cooperation
Size of city
Data provided by
Chamber of Commerce,
community, etc.
Information provided
by local manufac-
turers
Recreation, a good
place to live, etc.
Nearness to corporate
headquarters
Local supporting
services
State administration
neutral in labor-
management relations
Progress in racial
adjustment
Data provided by the
state industrial
development agency


LIST OF POSSIBLE FACTORS INFLUENCING INDUSTRY
LOCATION AS UTILIZED IN THE CARRIER AND SCHRIVER SURVEY33

TABLE 2


33p. 453.
Ibid., p. 453.



















Percent
Factor of Firms Rank


Low cost and availability of labor 65.6 1

Low cost of electric power 36.0 2

Favorable labor management relations 35.7 3

Community leaders' cooperation 32.2 4

Low cost of building and land 19.8 5

Low freight cost, finished product 17.9 6

Available existing plant 17.5 7

Favorable community and state tax
structure 17.2 8

Low cost of financing plant through
revenue or general obligation bonds 16.9 9

Available existing building 16.6 10


TEN FACTORS MOST FREQUENTLY MENTIONED BY TENNESSEE
FIRMS AS AFFECTING THE LOCATION DECISION

TABLE 3



















Percent
Factor of Firms Rank


Personal without economic advantages 49.2 1

Low cost and availability of labor 38.0 2

Available existing plant 35.9 3

Personal with economic advantages 32.9 4

Availability of low cost raw materials 31.9 5

Greater demand in area 30.2 6

Greater demand potential in area 29.8 7

Low cost of financing project through
area Redevelopment Administration 29.6 8

Available existing building 27.6 9

Nearness to corporate headquarters 26.0 10


TEN LOCATION FACTORS WITH HIGHEST MEAN NUMBER
POINTS ASSIGNED BY TENNESSEE FIRMS INTERVIEWED

TABLE 4









(3) Processing-Cost Factors These factors are associated with

in-plant costs in assembling or processing the finished product,

e.g., labor, energy, external services, capital, land costs, etc.

Electrical machinery, apparel and related products, and textile mill

products industries were affected by these factors.

(4) Distribution-Cost Factors These factors reflect the costs

incurred in shipping the finished products to the buyer. Among the

most sensitive to these factors were food and kindred products,

miscellaneous manufacturing, and paper and allied products industries.

(5) Location-Demand Factors Industries affected by these

factors are highly sensitive to market-demand in terms of proximity.

Included in this category are paper and allied products, printing and

publishing, and primary metal industries.

(6) Certainty Factors The validity of existing and forecasted

data is considered to be highly important by industries affected by

these considerations. In other words, these industries want to know

the future stability of costs in production and the probable con-

tinuance of existing markets. Printing and publishing, leather and

leather products, and transportation industries were highly sensitive

to these factors.

It is obvious that the scope of the Carrier and Schriver study is

much broader than the objectives of this study. Its utility is

limited for this research purpose. The factors considered by Carrier

and Schriver span several spatial levels of locational decisions. The

result is that factors which may be very important at the site-

selection level are weighted low in comparison to the total list of

factors. Also, the disproportionate number of factors offered for









consideration under the six categories tends to skew the weightings.

Finally, the lack of a very large sample in specific SIC categories

tends to decrease the validity of the results of the weightings and,

therefore, the conclusions reached regarding the typical locational

patterns of specific industries.

Bullington34 offers a scheme to locate potential industrial

sites on a state-wide basis by suggesting the scoring of location

factors on an ordinal scale and aggregating them into an index to

determine the site potential for specific industries. Local and

state governments could then match the qualities of the industrial

sites available in the community to specific industries. The factors

which Bullington suggests are listed in Table 5.

The U. S. Department of Commerce recently published the partially

aggregated results of an extensive 5-digit industrial location survey

conducted by mail throughout the U.S. The purpose of the survey

was "to assist the nation's underdeveloped and declining areas in the

development of their economic resources and potentials."36 Only

manufacturing industries demonstrating "reasonable" growth between

1958 and 1967 were selected for inclusion in the survey. Survey

forms were mailed to a total of 2,950 companies in 254 different




34J. S. Bullington, "Utilization of a State-Wide Site Evalua-
tion Committee to Aide in the Location or Relocation of Plant
Facilities," AIDC Journal, IV (October, 1969), pp. 27-42.

35U. S. Department of Commerce, Industrial Location Determi-
nants, 1971-1975 (Washington, D. C.: U. S. Department of Commerce,
Economic Development Administration, February, 1973).

3Ibid, p. 1.





















Site Characteristics


a. Size of Parcel
b. Shape of Parcel
c. Topography
d. Drainage
e. Flood Record
f. Condition and
Appearance
g. Underground Water
h. Soil Bearing Capacity

Acceptability

(This referred to the
potential friction or
good-will prompted by
the location of industry)

Accessibility


a. Highway
b. Secondary Roads
c. Rail


d. City Water
e. City Sewer
f. Limitations of Site

Community Factors

a. Commercial Air Service
b. Water Transport
c. Location in State
d. Mileage Rate
e. Airport Facilities
f. Comprehensive Planning
and Zoning
g. Retail Accommodations
h. College
i. Community Appearance
a. retail
b. residential
j. Highways
k. Presentation of Facts
by Community
1. Sanitary Sewer and
Water Treatment
and Facilities


LOCATION FACTORS SUGGESTED BY BULLINGTON


TABLE 5









SIC categories. One form, Survey of Industrial Location Determiinants,

was to be completed by all companies to i.Jcntif'y the locational and

operating characteristics of existing plants.

Only the results of the Survey of Industrial Location Det-rmi-

nants have been published to date and its usefulness for this study

is limited because of the highly disaggregated form of the data.

Hopefully, these data have been digitized by now and can be utilized

for future analysis purposes.

The attributes measured or assessed in the survey are only

slightly coincident with those sought in this analysis. Also the

survey spans several spatial levels of locational decisions. Never-

theless, some of the variables utilized in the survey were included

in this analysis.


Land Use Modeling Studies


Site-selection algorithms are certainly not novel to land use

modeling methodologies. Numerous modeling efforts have utilized

various allocation systems to distribute projected change in land

use. Most of these approaches have been different from this approach

in framework and in terms of the spatial scale of allocation.

Previous land-use models usually operate at the census tract or

count) level simply because data (e.g., census materials) to calibrate

the models are more readily available at those levels. The following

studies, however, have approached the allocation problem similar to

this study and consequently are briefly discussed below.

Among the first land-use models to utilize a site-selection

algorithm to distribute projected land use change was one developed









at the University of North Carolina by Donelly, Chapin, and Weiss.37

The primary purpose of the model was to simulate the growth and spa-

tial spread of residential land use. Historical data (1948-1960) were

utilized to calculate an attractiveness probability to simulate the

conversion of rural land to urban residential use. Changes were

recorded by means of 1000-foot cells composed of nine 2.5-acre land

development units.

The model operated in the following manner:

1. All land within the city unsuitable for development is

eliminated from consideration at the beginning and the

supply of land remaining is identified as available for

residential use.

2. For each 1000 foot cell, a measure of relative value is

established, i.e., land value, as a measure of its attrac-

tiveness for residential development.

3. The effect that "priming" (expansion of municipal services,

commercial services, and industrial development) decisions

will have on modifying the value of the property is then

calculated for each cell. These are assumed to be exoge-

nously given but in this case the exact amount is known from

historical data from 1948 to 1960.

4. Land parcels are then "reassessed" to obtain a new attrac-

tiveness score.




37Thomas G. Donnelly, F. Stuart Chapin, and Shirley F. Weiss,
A Probabilistic Model for Residential Growth (Chapel ilill: University
of North Carolina, Institute for Research in Social Science, 1964).









5. Density conIstraints (numiibe.rs of units per acre/year) are

then introduced.

6. Finally, the known gro'H,;h of residential households between

1948 and 1960 is allocated on a probability basis.38

The drawbacks to the model are: (1) the need to have historical

data in order to develop the transitiGn probabilities; (2) the disre-

gard of the effect of changes in industrial location and employment;

and (3) the vast amount of programming needed to load and run the

model for a small region such as a single city. The concepts gene-

rated by this study, however, have become basic to many other modeling

efforts and were equally important in this study.

The Pittsburgh industrial location model, INIMP39 (Industrial

Impact Model), is similar to the hypothetical design of this modeling

approach; the major difference is that growth is distributed to

census tracts and consequently the variables are more aggregated than

those being considered in this study.

Four variables (attributes of census tracts) and one constraint

were adopted as being sufficiently discriminatory to determine site

locations. These are: weighted mean unit-assessed value of land;

weighted mean unit-assessed value of buildings; weighted mean struc-

tural density, and amount of industrial clustering. These measures

were determined by census tract. The constraint can either be imposed




3Ibid, p. 11.
39
3Steven H. Putman, "Intraurban Industrial Location Model Design
and Implementations," Regional Science Association, Papers, IXX
(1966), pp. 199-214.









artificially as in the case of zoning controls or non-existence of

services; or directly imposed by the model operator. On the basis of

the aggregated scores of the above indexes, the :nodel distributes a

portion of projected city-wide employment changes among exist ,ig

facilities and, upon reaching certain critical values of saturation,

switches to a separate routine to distribute new facilities to census

tracts having the highest scores for the remainder of the projected

industrial employment growth. The distribution to various census

tracts is accomplished by way of the maximum score. However, in the

event of a tie, the allocation algorithm switches to a Monte Carlo

routine. Similar but more elegant models of this type are being

developed at Harvard40 and the University of British Columbia.41


A Tentative List of Variables


Each theoretical study reviewed approached the problem of

explaining industrial location in terms of three components: demand,

cost, and personal factors. Carrier and Schriver subdivided the

process into six components: personal factors, procurement-cost

factors, processing-cost factors, distribution-cost factors, location-

demand factors, and certainty factors. Only processing costs, pro-

curement costs, distributing costs, personal costs, and certainty




40
4Carl Steintz and Peter Rogers. A System Analysis Model of
Urbanization and Change: An Experiment in Interdisciplinary Educa-
tion (Cambridge: M.I.T. Press, 1971).
41
4M. A. Goldberg, Quantitative Approaches to Land Management
(Vancouver, B. C.: University of British Columbia, The Resource
Science Center, 1970).









factors have any direct relationship to site location considerations.

Certainty factors may be considered a variation of location-demand

costs. At the site level, procurement costs and distribution costs

are sensitive to ,ne variable, accessibility. Considerations of

freight rates, transport modes, proximity to raw materials, supplies,

etc. are more related to locational considerations at the subregional

level.

In an effort to arrive at a tentative set of variables to be

utilized in this analysis, a list of variables mentioned in the

industrial survey literature as being related to the site selection

decision was compiled. This list is found in Table 6.

A number of variables may be eliminated as not related to the

site-selection process while others can be combined with other vari-

ables. After considerable study the following list of variables was

developed to be used in a survey of the site conditions of past

industrial location events. Selection was based upon the relevance of

the variable to industrial site-selection and the ability to obtain

measures of the variable. The variables are listed below:

I. Site Preparation Cost

a. Slope of land
b. Drainage
c. Clearing-Cover conditions

II. Market Price of Land

a. Distance to center of town
b. Distance to nearest major thoroughfare
c. Density of urban use in immediate vicinity
d. Overall rating of price of land from 1 to 10

III. Proximity to Work Force

a. Proportion of nearest city within 2-1/2 miles
b. Population of nearest community




















1. Industrial Park Space
2. Industrial Park Quality
3. Industry Nearby
4. Zoning and Building
Restrictions
5. Pollution Regulations
6. Land Costs
7. Site Preparation Costs
8. Topography
9. Drainage
10. Soil Conditions
11. Place to Dump Effluent
12. Processing Water
13. Utilities
14. Municipal Water
15. Sewage
16. Natural Gas Service
17. Proximity to Local
Markets
18. Proximity to Local
Raw Materials
19. Proximity to Sup-
porting Industry
20. Water Transport


21. Railroad Transport
22. Highway Transport
23. Commercial Airport
24. Distance to CBD
25. Community Transportation
26. Community Parking
27. Neighborhood Services
(Restaurants, Hospitals,
Parks, Gas Stations, etc.)
28. Nearby Housing
29. Nearby Labor
30. Community Cooperation
31. Community Stability
32. Community Wealth
33. Community Taxes
34. Community Progressiveness
35. Community Attractiveness
36. Community Labor Climate
37. Community University
38. Nearby Government and
Institutional Facilities
39. Community Wage Rates
40. Space for Expansion


TENTATIVE LIST OF VARIABLES


TABLE 6









IV. Transportation Accessibility

a. Distance to major highxi',-
1). Di stance to secoindaryl roa
c. Distance to rail
d. Distance to airport
e. Waterway service
f. Distance to nearest Interstate interchange
g. Overall quality of accessibility from 1 to 10

V. Utilities

a. Water available
b. Gas available
c. Sewerage available

VI. Compatibility with Existing Land Uses

a. Did community have zoning at time of location?
b. Was site zoned for industry?
c. Was zone changed to accept industry?
d. Was industry already in immediate area?
e. Overall rating of contiguous land use compatibility

VII. Neighborhood or Community Attractiveness and Amenities

a. Condition of neighborhood
b. Density of land use in immediate vicinity
c. Nearby community services

VIII. Industrial Park Space

a. Was the site in an industrial park?
b. Overall rating of the quality of park?

Additional Data Collected

a. Proximity of site to Knoxville
b. Amount of other industry located nearby at time
of event
c. Was building already there?

This list may omit variables which should be considered and,

therefore, should not be considered exhaustive. However, it is antic-

ipated that in measuring the importance of each variable some may be

eliminated thus simplifying the site-selection algorithm. The next

chapter explains the methodology and research procedures used to test

the importance of each variable.















CILI fER III

DESCRIPTION OF THE STUDY AREA
AND ANALYSIS PROCEDURES


This chapter describes the research methodologies utilized in

this study. The first part of the chapter describes the eastern

Tennessee study region, the second part the data collection proce-

dures, and the final part the analysis procedures.


The Study Region


The study region encompasses 16 counties surrounding and including

Knoxville, Tennessee, (Figure 3) and represents an administrative

entity called the East Tennessee Development District (ETDD). The

region spans 6,500 square miles and contains a population of approxi-

mately 750,000 people. Its selection for this study was based upon

the availability of data in the Oak Ridge National Laboratory (ORNL)

Data Base.1 ORNL selected the region on the basis of "the diversity

of the region, the availability of data, the presence of cooperative

and interested user groups, close proximity, etc."2




1Oak Ridge National Laboratory is currently developing a "holis-
tic" environmental model for the ETDD region. This research was
supported by this program. ORNL-NSF Environmental Program, Regional
Environmental Systems Analysis (A Research Proposal Submitted to the
National Science Foundation, February, 1972).

Ibid, p. 3.












The ETDD region is centered in the southern portion of the Ridge

and Valley Province (the "CGreat Vally"'), bordered to the northwest by

the Cumberland Plateau and Mountains, and to the southeast by the

Great Smoky Mountain complex (Figure 4). Both the Smokies and nhe

Cumberlands are characterized by steep slopes and forest cover, with

the Cumberlands distinguished by strip mining scars.

The area is drained by the Tennessee River system, the natural

flow of which has been vastly altered by the Tennessee Valley Author-

ity (TVA). By harnessing the power of the river system to produce low

cost electrical power and developing a navigable channel to Knoxville,

TVA has become the major development agency within the study region.

As a result much of the industrial development in the region has been

structured by TVA activities.

The largest urban center in the region is Knoxville which serves

as the major economic and transportation focus for the region. Sur-

rounding Knoxville are Oak Ridge, Maryville-Alcoa, and more distant

Morristown, each with between 20 to 35,000 in population. These

cities perform subregional functions. Remaining urban centers are

small in population and are mostly located within the confines of the

plateau and the Blue Ridge (Figure 5).


Industry within the Region

This analysis has concentrated upon secondary manufacturing (SIC

20 through 39) and excluded extraction industry (SIC 10 through 19).3

The location determinants of extraction (or primary) industry are




Industries are referred to by their Standard Industrial Code
(SIC) number throughout this study.












































.7~


/4


0


'I

























^1



I





to
b

















tr


t
,

k
^ I

fcl
,











u-1


b;

__


~-t









dictated more by the distribution of raw materials and consequently

the locational criteria of these industries will be different from

secondary industries. For this reason primary industry is not con-

sidered in this st-.dy.

Categorically one could state that most of the industry within

the region is concentrated in the Knoxville area. Of the 1002 indus-

tries in the region, approximately 46 percent are located in the

immediate vicinity of Knoxville. Knoxville also has the greatest

diversity of industry, whereas industry in other communities is

characterized by specific categories. (An example is the concen-

tration of furniture industry in Morristown.)

Figures 6 through 13 provide a visual overview of the dynamics of

industrial development in the region since 1943. Since 1953 a 48

percent turnover in industry has occurred. Of the 1002 industries

within the region in 1973, 488 have developed in the 16-county region

since 1953.

Table 7 provides a statistical summary of the new industry-new

employment expansion (or decline) from 1952 to 1973. The table does

not reflect the absolute employment expansion rate but rather the

amount of employment gained or lost by the birth or death of specific

two-digit SIC industries. A high average expansion rate indicates new

plants have developed in the region and have significantly increased

the total employment in that specific industry.

In the following discussion each industry type is briefly

described. Maps are included to show the regional distribution of

specific industries. Graphs illustrating the historical employment





















15



tr-- 4- -,-- 1-

i:i
*-- 1
.. *. .
I. *. *






















S





ti
I \ !---L~;~,-.- t































!




k
-4 1



00


PI


I I I









49























-Ii


4;


2--i

4;


+M











51


















r I4r


____ I








r. ~







*~ Ii








5' 'I
Lr 6



I I i i i i i






52










i iCC
,
*. a -o g g
__^ / .~ m oo -
*6 S-^ e SA a-
I *





I I
oA

4. 6








S.J

c
^-- ^r--- 1r






S_ S ** I <
--------------Y--- .*~-













UI
\ ^____i i ..
S
*:. .. ;' 7 :'




S-~
S j i- -i- -- -Y -.
^_ ,_^ p -" lv ^ ; "i



,t~ P *!'-v;

S-- -^1)




"--
-~----^--*.^ry_.





a I 6 a i i











Average New
I industry Expansion
by Rate
SIC 1955 1958 1960 1963 1966 1969 1973 1952-1973


20
Food +9.4 +10.0 +0.5 +8.7 +6.5 -3.2 +11.8 +3.8

21
Tobacco

22


Textile


23
Apparel

24
Lumber

25
Furniture

26
Paper

27
Printing

28
Chemical

29
Petroleum

30
Rubber

31
Leather

32
Stone,
Glass, Etc.


Primary
Metals

34
Fabricated
Metals


-4.6 -2.7


+28.3


-40.4


+24.0


-4.7


+9.3


-.6


-8.2


-1.4


-9.7


-13.6


-3.2


+30.4


-16.4


-2.0


+13.4


+4.7


-11.6


-7.1


-7.5


+2.3


-2.2


+53.1


+12.3


+24.5


-4.3


+5.2


+9.8


+25.8


+1.4 +3.2 +7.6



-6.9 +10.1 +.5



+36.6 +12.0 +.8


+8.3 -3.9 +5.2


+20.1


+2.1


+13.8


+10.2


-4.1


+2.3


-41.9


+10.3


+13.1


+11.2


-0.3


+5.2


+4.2


+2.2


-14.7


+27.2


+79.5


+25.9


+40.3


-7.2


+5.9


+49.1


+21.0


-2.4


+43.0


+61.6


+62.6


-19.7


-12.4


+11.2


+7.8


+41.8


-3.3


-25.6


-5.0


-32.4


+63.5


+.3 -11.0 -4.4 +25.5


-22.8 -4.6 -6.9


+10.1


-2.4


-3.6


+10.9


-5.7


+20.0


+13.8


+6.8


+2.8


+3.6


+16.5


+24.6



+3.2



-4.7


+4.7 +36.8 +6.3 +15.3


EXPANSION OF ETDD INDUSTRIES BETWEEN 1952-1973


TABLE 7
















195' 1958 1960 1963


1966 1969 1973


Average New
Expansion
Rate
1952-1973


35
Machinery
Non-Electric

36
Electrical
Machinery

37
Transpor-
tation

38
Instrument

39
Miscellaneous


+6.7


-4/.7 -1.3 -6.9 4409.8 --21.1 -30.1 +19.1


+26.4 -17.2 +22.5


+37.3 -1.2 -3.9


-15.7 +2.6 -8.4 +


+3.7 +439.0 -13.2 +136.5


+,.5 -1.1 +9.5 -3.3


34.5 +4.4 +11.6 -26.1


+85.4


+3.3


+10.6


*This table is based upon ORNL industrial location data. Expans:
rates are determined on the basis of new employment prompted by
new plants (or death of plants) since the previous time period.
base year is 1952.


ion


The


TABLL 7 Contiiinued


Industry)
by
SIC


_ ~_I_


___ _I_____


_(__ ~_ _____~_ ___


-3.4 +1i.2 -8,6


,.3.1 -15., +55.3 +0.5









trends relative to total regional employment and the industry's share

of total regional manufacturing employment accompany each map.4


Food products industries (SIC 20)5

Of the 103 plants existing within the region, 66 are located in

the Knoxville region. Most of the remaining plants are located in the

southern half of the region bordering the Smoky Mountains near the

better crop producing areas along the margins of the Nolichucky,

Pigeon, Little Tennessee, and the French Broad Rivers (Figure 14).

Approximately 75 percent of these industries employ less than 100

persons. Employment growth has been gradual and in accord with total

regional growth. Proportionally, however, employment has been de-

clining (Figure 15).


Tobacco products industries (SIC 21)

No tobacco industries exist within the study region.


Textile mill products industries (SIC 22)

The textile industry is distributed broadly from northeast to

southwest through the center of the region with Knoxville having the

greatest concentration followed by Morristown and Sweetwater (Figure




4The maps and graphs utilized in this chapter were prepared by a
CALCOMP plotter driven by an IBM 360 computer. The data utilized were
originally collected by Osbin L. Ervin. For a complete presentation
of these data, see: C. R. Mcyers, Jr., 0. L. Ervin, D. L. Wilson, and
P. A. Lesslie, Spatial Distributions and Employment Trends of Manufac-
turing Industries in East Tennessee (19413-1973) (Oak Ridge, Tennessee,
Oak Ridge National Laboratory, June, 1974).

5Transparent overlays found in Appendix B may be superimposed
over each industry map for spatial referencing.





























E-4
16



















0
bb

















C,


H









I.
---?
i0
--*1















0




a)







M H
C )



/ (Dn 0 0 l





o





OS Sh Oh S 0
o ^ 4














a)



0
U-)


















C)





S0 4
\ -4














V










In -
















os ll l o h se oII I S sI
SO1 S 0t S L0t S zOt S -
un
^ t 'fl 2 z
I '2 M

















Sol s oi s cot 1 9O so S









16). Average employment in the industry is approxi-atcly 210 persons

per plant. Rockwood and Harriman have only four plants but share

nearly 50% of the total employment inl the region. The distribution

of plants reflects a tendency to gravitate to re1.Ltively small communi-

ties as suggested by Lonsdale6 and Pred.7 Textile industry has declined

in terms of the percent of total regional employment in recent years

perhaps resulting from the introduction of higher wage industry into

the region (Figure 17).


Clothing and related products (SIC 23)

Knoxville leads the area with the greatest concentration in both

plants and employment (Figure 18). Average employment is approxi-

mately 315 persons; however, the median is closer to 150 employees.

This type of industry though concentrated in Knoxville is scattered

throughout the region with numerous small communities having at least

one (sometimes two or three) small employment industry. In terms of

new industry expansion the clothing industry expanded at an average

rate of 10.9 percent between 1953 and 1973. Historically, however,

growth has been erratic (Figure 19).


Lumber and wood products except furniture (SIC 24)

This industry is widely scattered throughout the region mostly in

small communities (Figure 20). It .s a small employment industry

mainly utilizing local raw materials. In recent years, an overall




Richard E. Lonsdale, cited in Chapter 2.

Allen Pred, cited in Chapter 2.




















































cn

L)












rLr






L)L
1-4
'n






















































Io, s, ~ .... ,. .,... I I I... Ji
OS Sh Oh SE OE SZ O SI O S 2


I.......I .....iii 111111 I I


li 1 I 1 II I I I I - .. . .
SO ?U UL e


111111 i I I


u






CD
z z







uE




HZ






O[-


Ht
C"-





E-4






n


EU


sOI s vUI








61


































00


















~C'4




-4



























H




00










H
l-n


OS Sh Oh SE OE SZ OZ SI O S



D(
12


I ~ .. i i I 1 1 1


I i li i


I I I I I I I il lI I I A I I i i I r


cOI s ZoI s


I


Ii n n


sOl s h01





































t



iI




















* .

S-* .


S.Y




S

S--


a a k
a a


CL.














ho
) a,




ca
I-
< 4x0




E-i

UT -































a \ Pa a


1.









decline in the industry has occurred partly due to a decline in small-

operated lumbering in general for the region. The average employment

per plant is only 32 persons with the median closer to 15 persons per

pl'-.t.

The growth prospects for this industry are rated good by the TVA

Regional Planning Office but it is expected to be in terms of large,

permanent sawmills utilizing local oak, hickory, and poplar forest

resources (Figure 21). Those species are being increasingly utilized

by the furniture manufacturers.8


Furniture and fixtures industries (SIC 25)

The furniture and related products industries (e.g. woodworking

industry) is heavily concentrated in Morristown (with 20 of the 46

plants) even though Knoxville has 17 plants (Figure 22). The average

number of employees per plant in Knoxville, however, is 161 persons;

whereas the average in Morristown is approximately 310 persons.

The furniture industry located in Morristown not because of the

local timber sources but because of cheap labor and good rail connec-

tions to large mills in the South and to market regions in the North.

With increasing reliance upon local timber, however, the growth

prospects for this industry are seen to be significant (Figure 23).




Communication with Bill Ogden, TVA Regional Planning Office,
Knoxville, Tennessee, May 3, 1973.

Ibid.

























































I ... I I I I I I I I I i I I I I I II I ... I I I... It r
OS S O S SE OE S2 0 SI 01 S


C


A s. I A I I I I 1 I 1 I I 1 I I I I I I I


Z









H


0

M Q)
et'















E-4


0)















u




1-
u


~


I,,,,,~ ~ ,


I I I I I I I I I


I, ,,


hOT s


EOl s 201 s




















2'E







I,,
-4
'-
tjD









67




0




CO







H-4
0o '










Z


NJH
Ln z








V a)
=r
0 0










E-4






C1


H
1 1







\ t





ta E-4
0)
iD C






en z







O






III I I I I I Jl ll ) I I I I l llI II I I l ill lu I I
sOl S hO I S 01 S O S 2s









Paper and pulp products industries (SIC 26)

The paper and pulp industry is concentrated in the Knoxville

area both in terms of employment and number of plants (Figure 24).

Intuitively, one suspects market-Kemand to be a strong locational

factor except for the plants in Morristown which are locationally

related to the by-products produced by the furniture industry.

Employment averages 78 persons with the largest plant employment only

190 persons. Total employment in the paper and pulp products industry

is relatively low compared to other industries, only 1377 employees

in 1973. The average growth rate, however, has been significant

(Figure 25).


Printing and publishing industries (SIC 27)

Printing and publishing includes local newspaper printing and

for obvious reasons is widely dispersed throughout the 16-county

region (Figure 26). Knoxville has the greatest concentration with

over 75 percent of the employment and 2/3 of the total number of

establishments. Total employment is low for the region (1670 employees

in 1970) with an average of 19 workers per plant. Growth has increased

along with population with no great expansion predicted (Figure 27).


Chemical products industries (SIC 28)

The shunning of large urban places by chemical plants noted in

Chapter II at first appearance is not substantiated by the regional

pattern of chemical industries within ETDD (Figure 28). Industries

are found in the other communities (e.g., Morristown and Oak Ridge)

but Knoxville is the leader in number of plants with all but six of

the 33 plants in the region. However, the Knoxville industries are









































04











1C1

1-4
n,


r
































































os S O S O E se oO S O s 01 s So


,T C II I I I |
sl 1I L


- -
z z







0
o o









o.
E-4


I


111 1 I I I I 1! i i 1


Il l I i i


I .


IUL


EUL s


Su i




















r- -


0






I i l ;


-c------,













.. ^ _i ._-









+ i --_ --


I


(fl*
c=
.- a
C
=
a. --


6t,~







t2













E-4 r.I


P.4


H


4-L


6 b
E >


II






















































OS Sh Ot SE O0 SZ 0 S 1O SI 0


I I I I I I I I I


SI I I I I I I


II III I I I


SW z u


I l I I


u








z 3
Z H




HU




^^3


"


I


SUI S


201 s






73




















P P4




00
"il



cn-














C- ------ ----
T *~





tt
_ I __ ____ ___














_i I



I ft SI I









relatively small (the largest has 69 employees) wh;hle the ENKA plant

located at Lowland (six miles south of Mlorrito-..n) employees over

4500 persons. Large employment figures in Oak Ridge and M'orristown

(Lowland) skew the average to 308 persons per plant; therefore, the

median 10 per plant is more descriptive of the true employment

picture. Employment in the chemical industry has been erratic in the

last two decades (Figure 29) and future expansion is difficult to

predict.


Petroleum refining and paving and roofing products (SIC 291

Employment and plant number are relatively insignificant relative

to total industry within the region with all but one plant (seven

employees) located in Knoxville (Figure 30). Little change in employ-

ment has been noted in recent years (Figure 31).


Rubber and plastic products industries (SIC 30)

Ten of the 12 rubber and plastic products industries are found

in Knoxville presenting over 60 percent of the employment (Figure

32). The average number of employees is 68 people with 20 employees

as the median. Between 1952 and 1973, the new industry growth rate

has averaged 16.5 percent but the total employment in 1952 was only

514 persons (Figure 33).


Leather products industries (SIC 31)

The leather product industries are few in number (15), most

occurring in Knoxville, Dandridge (near Jefferson City), and Morris-

town (Figure 34). The average expansion rate has been significant,

24.6 percent even though the base year employment (1952) was only 546




























01


DI -
a)



.0


(0

Ur)

U')
a),



Lr)


CY)
t p









\es 0
II 001


1111 1 1 1 1 1


lilt I IO I I0 I IS


C.,
)"





1-
Cr


I\


00 4

z




0
E-4
zO
cn


:
-


.


,o


I I I I I I l l I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I


sOI s h01 s


cOI S 20I s
































































_________-T,________ /______\_ _____________ _______________ ___________________________

I.

L i
i^ k,-^r -~------ --- s
e- L L_________


























































os si Oh SE oE SZ OZ SI 0 S


I 1 l i l l i l i 1 I l i I 1


0
z



Z1-



















I


CN
Pri
1-4










(n
Oi


-""''' 'i


i
i
i
I r


I i I1 i


I I !


11i i i


sOl s 01 s


OI s 1O







78


















C-










000
0 Po






to













I IV










I i iI I I Ih SP

























































OS Sih A SE O S 0 S O[ S


11111l(1 I IiiilIi


liii1 I I I


. . .. I I I I3
SUL #"I I LL .


liii i


C.3
ZZ


SH
m g











.-I E-I
0
O


On

O^


0
W1





H
3




04















1-4
IM
fc






Ub
COi


U L


' "


SUl


hUL






































liz
Hr


k~









persons. Employment per plant averages 100 persons as does the

median, indicating a fairly symmetrical distribution. Expansion in

the future may be expected to continue but probably not at such a

high rate (Figure 35).


Stone, clay, and glass products industries (SIC 32)

These represent strong local raw material, local market indus-

tries, and thus their location is determined by the chance occurrence

of resources and the distribution of urban places. The spatial

pattern supports this observation (Figure 36). The greatest concen-

tration, as expected, is in the Knoxville vicinity with 2/3 of the

total number of plants (68). The average employment per plant is 32

persons. The median is 14 employees per plant. Expansion has kept

pace with population growth in recent years and probably will con-

tinue to expand in response to total regional growth (Figure 37).


Primary metal industries (SIC 33)

This industry type is distributed throughout the region simply

because of the large number of foundries which historically developed

in East Tennessee (Figure 38). The largest employment industry is

Alcoa Aluminum in Maryville-Alcoa with 5000 employees (70 percent of

the total employment). Median employment, however, is only 50 persons

per plant. Employment has dropped steadily since 1963 but is expected

to level out in the future (Figure 39).


Fabricated metal products industries (SIC 34)

This classification includes welding and machine tool industries

which usually develop locally and are local-market oriented. This


























o







o H












01
0 m


















In


r)
CVl













=r








H H
*o M
















0r)



o a













In
N E- 0

































Sol hol S co Z I







83



















'000




0 0e









1 CN
I ---~F -4





pC- __ __






F I,


~-: I I I+T




I ~ i I -I
























































IO S. Oh SI I SZ I OZ SI ... S I
OS Sh Ob SIE OE Sa 02 ST l s F!S


I I I Ii I I I 1 lI II I I I I I


Iii 1) ii i i1 i I


DI s s0i s Oi s s


cn


u
Z;














.z0
E-4

a^


u
I
v,







85


















---- -K
- -------------------










E-4~








CU,
A 1-41
11~- --1con

























































OS Sh Ot SE OE SZ o. ST 0o s


I


/11111 I I lii ii r H


1 c 1"V
s1L LU


ZZ













E -4
az
1.4
Q 5



o5
en F


. -. I=


*r-4
HM
I






2 -,
I


cj,



en


1ll1111 I


~
u


I .


hul


EU









explains the wide distribution of the industry Figure 40). Most of

the 73 plants are found in Knoxville (upproxi ,:itely 75 percent) indi-

cating some response to raw material availability which is greater in

the Knoxville area. The average number of persons employed is 45 but

the median is only 14 employees per plant. The industry has grown

significantly between 1952 and 1973 probably in response to national

growth in the transportation and recreational vehicle markets (Figure

41).


Machinery (except electrical) industries (SIC 35)

The machinery industry is heavily concentrated in the Knoxville

region and to a lesser extent in Oak Ridge and Morristown (Figure

42). Many of these industries perform supporting functions such as

the refurbishing of industrial tools (for example, the woodworking

tool industries in Morristown). The largest employer has 639 employees

but the average for the region is 27 persons per plant and the median

is only 7 persons per plant. Total employment has remained fairly

constant in the last decade (Figure 43).


Electrical machinery, equipment, and supplies industries (SIC 36)

The electrical machinery industry has grown from 578 employees

in 1952 to 2726 employees in 1973, averaging a growth rate of 60.9

percent. The largest employer is Magnavox in Morristown with 716

employees in 1970. Average employment is 120 persons; the median is

40 persons per plant. Most of the industry is concentrated in Knox-

ville and dispersed lightly elsewhere (Figure 44). In recent years,

employment has expanded significantly (Figure 45).




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E8CG7UW1G_411VS6 INGEST_TIME 2011-07-01T00:29:42Z PACKAGE UF00102837_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

THE EVOLUTION OF INDUSTRIAL LAND USE WITHIN THE KNOXVILLE METROPOLITAN REGIO N : AN ANALYSIS USING AERIAL PHOTOGRAPHY AND HISTORICAL DATA FOR THE PURPOSE OF LAND USE MODELING By ROBERT B. HONEA /\. DISSERTATIO N PRESENTED TO THE GRADUA TE COU NC IL O F THE UNIVERSITY OF FLORIDA I N P,\RTTAL FULFILLMENT O F THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1975

PAGE 2

DEDICATED TO: Dr. (Jim) James C. Wilkinson J

PAGE 3

ACKNOWLEDGEMEN TS Of the man; individu G. ls 1 ,ho have cont:.:ib uted to this s tudy, fore IT' ost is Dr. Clark I. Cross, Chairr.i an of ;n y Sup e rvisory Conur_tte e l aP1. appreciative for his professional and personal assistance His periodic e nco u ragement ha s been mos t opp o rtune. U r. Gary Shannon has contributed both to thi s study and at other stages of my graduate career I am grateful for his he lp D r James R. Anderson has been a source of encouragement before and during my graduate tenure. To him I extend my appreciation. To Dr. !)avid L Nid drie l extend my appre ciation for many engaging discussions. His intellectual vigor is a source of professional inspiration. The lectures of Dr. Clayton C urt.i _c_ f F O \ id e d m2 n.y j cl ca~ whicli are :n c-;u decl in t n:i. :, '. : tudy To h i m my sin c ere :lppr eciation is extended. Cra ti t ude i s a lso exten.Jed to m e n; bers of the f aculty for thejr 2 ssistanc~. who contributed to the research effort: Mr. C. T N ( Te o ra t.ory; 2n
PAGE 4

AC K.!\lO\I LELG EMl:N T S us~ OF TABLES LIST OF ILLUS TRA TIO NS ABSTRACT CHAPTER I. INTRODUCTION T ABLE OF CO XTE~rs Problem Background Stateme n t of Proble m P rob l em Operationali zat ion Model Operation ... Motivation for Stud y O rgani za tio n of Study. II. T H E ORETI CA L AND EMPIRICAL s n ; oIE S hELEVA>i T 1 0 'JiI [: lll V 1 J L< xii l 1 3 3 5 6 8 DEVEL OP MENT OF I N DUST RIAL SIT E -S Ll l LCTlCN A LGOF I T~ I 9 Se l e cted The ore t ica l Works 9 D er I s o l e Y le St c1 :it . 1 r, lJJ:-,er d en Stan d ort d c r I 1~ .Ju ::, tr i.cn 1 :2 T he Locati o n of E conoiilic l klia vio r 1-1 Imp e rfe c t f: ompetiti o n ::n r l th e D t 1 c 1. n l y n e bz~r:> lS 'fh E, ; heor .i.. e s o f lJ av1d \i. Sm j_ tl 1 -'.i,i iJ M e lv i n L Gr ee nh ut 1 6 E m p irica l St u di es lB Th e StJ d ies of A l l en Pr ~ d and Ri c hard Lon sd~ l e 1 9 His to rica l Studies of lP du stria l Lo cat i on Factors. 2 2 l.3.nd lJ se i'l odeling Stu d i e s . i2 A l' e nt<1tive L is t o f Vari o b:tes 35 I II. DES CRI P TIOi'J UF THE S TUJ 'f AR EA AN D ANALYSIS P KG CS DUR ES 39 Th e Study Regi on Indu s trf wi thin th e Reg~on Food products indust ries (SIC 2 0) Tobacco products indus tr ies (SIC 2 1 ) T ex t i l e mi ll produc ts indu s tr ~ es ( S I C 22). iv 39 55 S5 ss

PAGE 5

Paae __ o_ Clothin g and relat e d product s (SIC 2 3) . 58 Lu m ber and ~0o cla y and g la; ::: products indu5tri~ s (SIC 32) . . . . . . 81 Pri:nary met a l industri u~; (SIC 33). . . 81 Fabricated metal prod~cts i ndustries (SIC 34). 81 Machinery (except electrical) industries (SIC 35) . . . . . . . 87 Electri c al m a chinery, equipment, and s upplies industries (SIC 36) . . . . . . 87 Transportation equipment industries (SIC 3 7) 94 Instrument and related products industYies (SIC 38) . . . . . 9 4 Mi3cellaneous industries (SIC 39) 94 Data Collection. . 101 Statistical Procedures 106 IV. D!PT.RICA L A..NALYSIS . ne~criptiv e Statistic ~ Description of Sample. Des cr iption of Variable Meas ur e rriel!t S R es ults nf the Factor Analysi s V. SUMMARY AND CONCLUSIONS. 111 J. J. J lJ) 117. 148 158 Sum'. nary. . 153 Conclusions. 159 Conclu s ions Rega1 ding the Form of d w Site Se ] ec tion Algorithm. . . . . . . 159 Conclusions Regarding tha Proper Form of Model Operation. . . . . . . . 160 Concl usio ~s Regarding the Use of Ae ri al Photography in 1,:o
PAGE 6

R[SE/'J'.C H DIEL 10 G RAPHY. P er i odi c als . . Books. . . . Govern men t Doc1..i1c;nts and Agcnc, :i.nd ln ~tj tute Res ear ch R eports . .. Unpublished Sources. APPENDICES A. Sample Survey Forms B Map Overlays BIOGR AP HICAL SKETCH. 163 163 168 172 177 179 182 184 vi

PAGE 7

LI ST O F TABLES T a bl e 1. f-2.ctors ~ l os t Influ e ntial ~n th e Location Decisions of Flor i da Industri es 19 56 -1 957 . . . 24 2. List of Possible Factors Influen c i ng Industry Loc a tion as Utilized in the Carri e r and Schriver Survey. 26 3 Ten Factors Most Frequently Mentioned by Tenriessee Firms a s Affecting Location Decisions . . . 27 4. Ten Location Fact ors with Hi g i t es t M e a n ~umber Po i nts Assigned by Tennessee Firms Intervie1,ed. 28 5 Location Factors Suggested b y B ullington 31 Tentative J..ist of Vari ab l es. 7. Exp an s ion o f ETD D Industries Be tw een 195 2 -19 73 8 Distribut i on of S a mple by SIC C 2. t =?g ory 9. Variable Slope of La n d 10. Variable Draina g e . 11. Variable Clearing-Cover Co ndition 12 Variable Distan ce t o C e nter o f To wn. 13. Variable Distance to Major T hroughfare 14. Varia b le Density of Land Use in I1rnncdiate V : i d 1 :i t y 15. Variable Rating of Price of Land . . 16. Var iabk Proporti rm of Urb an Arer1 W i th i n T wo a n cl a tl"1-l f t-\ .: les 1 7. Var iable . D is tance to Hajor Highviay 18 Va.!"iable Distance to Secondary Road 19 Va.riable Distance to Railway 20. Variable Distance to W ater way 2). Var ia ble Distance to 1irport v :.ii 37 53 J ) 3 114 115 11 7 ) )8 119 120 1 22 12 3 J 2S 126 J 27 128 1 3 0

PAGE 8

T a bl e 22 2 :: 24. ~5 26 27 28. 29. 3G 3 1. Va r i ;:: ~ol e O ve 1 2. l i Q u, 11i.tv o f r \ cct: s sib i 1 i t y -Th c n 1/ariaule Clverall Qu;:ility of ~cce ss lbllit y -~ o w. V2..Tiable City Water Avci.i lab i. Ji ty Variable City Sewage Avail a bility. Variable Gas Ava.i l ahility . Variable Did Community I Jave Zoning Then? Variable Was Site Zoned for Industry? V ::i. r i a. ble \fa s Ind 1 s try A lrea d y Jn A r e a ? V a. riable O v cr al 1 Ra t:in g o f o n t i g : 1 0 u:: L and Use Co:npatabili ty . . . C o n --li t ion. of Ne igh b or h o o d \'a r~ .ablE: D 8 nsit y of Land Use .. 3'1. Variable Nearby Community Ser v ice. SS V a riable Wa s the Site i n an I ndustr ~ ~ l P ar k? :'. 6 )_" ..., .l I Proximity t o Knoxv : i ll e . ... . 3S. V ariabl e knount of Other In d us t r y Loc a t e d Nearby 39 C o r re ]ation M a t ri x /i. U 41 : J n rot a tecl Fa c to T M at ri x 42. Varimax Rotated Factor Matrix . 131 133 134 135 135 136 136 137 1 3 7 14 0 Hl 143 1 44 146 147 1 4 9 1 S 2 15 4 lSS viii

PAGE 9

LIST OF FIG (lfU:S Fi g. i re --1. lJrt Ja n L a ncl-U s e L o -:: at ion Sc h em , tJ s 6 . I 8. 9. 1 0 11. 12. 1 .) Smith I s ~ l o de 1 w ith Com par .i.sons. East Te nnessee D e velopm e nt D ~ :c tr i c t Physiography .. . Towns and Communitie s Plru1 t Loca t ion 19 4 3 Employment 1943 . P l a~ t Loc ation 1 953 Empl oym ent 1953 P lan t Loc2ti ons 1963 Employme;it l 963 Pl R nt Locations 197 3. F.mp l oymen! 1973 : j. SI C 7(\ fo o d P 1 od u c. ts (P l:-mt f -::, r a ti c,r r : 1 S SIC 20 Food Product s (Employment) . 16 SIC 22 Textile Produc t s (Pl an t Lo catio n s) J.7. S I C 2 2 Text i le Products ([mplc yment ) .. l S. JC 2 3 Ap aTt': ] ancl flA l a i.:e d ; r oduc i (f'l a n t. l.o c-1. ~ions) 19 SI C 23 A p p a Tel and Related. P :... o du-:::t s ( Em p J oy;r 1e nt) 20. S I C 24 Lumber and Wood Products ( Pl a nt Luca c ion s) 21. SI C 24 L um t er &nd Wood Produ c t s (E mp lo y me nt ) 22 SI C 25 Fu rniture an d Fixtures (P la nt Locati ons) 2 3 SI C 2 5 . F u r nitu re and F ix ture s ( Employment) 24. SIC 26 P1per and Pu lp Product s (P l a1 1 t Lo ca tions) ix 11 17 40 42 43 46 4 8 4 9 50 51 56 57 59 60 61 62 63 6 5 66 67 69

PAGE 10

25. SIC 26 P a p er and Pulp Produc ts ( E m~loyD e~t ) .. 2 6. SIC 27 Printin g and Publi s h i n g (Plant Lo c ations) r SIC 27 Pr i ntin g a n
PAGE 11

X l F i gur e ;7 --, I SI C 37 Transportation Equi.pm e nt ( En qloy m ent) .. 96 48 SJC 38 Instrum e nts and R : ~ Ltte d Pro d ucts (Pl a nt Locations) 97 49. SIC 38 I nstruments ar!d Related Produc t s (Employment). 98 5 0 SIC 3 9 M isc e J. lan e ou s l nd u ::, tr i es (P la nt Location s ) 99 i; .L SI C 39 M j s c ellaneous Ind u s tr ie s (E ; n ployrnent) HJO 52. 1958 S t ereo Image of Beta -To k Ind u stria l Si t e 102 53. 197 4 S t.ereo Image o f i3eta ," e k lndu '.: trial Site 103

PAGE 12

Abstract of Dissertation Presented to the Graduate Council in Parti al Fulfillment of the Requirements for th e De gree of Docto r of Philosophy THE EVOLUTION OF I~ DUSTR IAL LAND USE WITH!~ THE KNOXVILLE l'IETROPOLITAN REG ION: AN A~ALYS IS USING AERIAL PHOTOGRAPHY AND HISTORICAL DATA FOR Tll::. PURPOSE OF LAND USE MODELING Rohert B Honea March, 1975 Chairman: Dr. Clark I. Cross ~-laj or Department: Geography This research effor t repre se nts a port:ion of a mu ch larger research goal, the ide !~!_~ ication of the d e t erminants aff'"cting l:ind ~~~hange. This study focuses, however, upon only one aspect of this problem, the con vers ion of land t6 in
PAGE 13

were collected on th e location charact er istics of each industrial site. Tl1 esc data wer e then subjected to factor analysis to determine the interrelations of variables, to mininize the li s t of v aria bles needed to describe the industrial site-selection pro cess and to determine if the preconceived ideas concerning the factors affecting the process were valid. Seven factors accounting for 72 percent of the variance found in the original data were identified. The 30 variables studied did not group as previously conceived but nevertheless the factors presen t a logical arrang eme nt of vari ables. Four factors are almost singularly identified by accessibility chara~teristics. Reexamination of t he acc ess ibilj ty factors indicates they could be combined, thereby reducing the number of indexes to two or three. 1be results suggest that a zoned or de signate d place for / industry to locate, oven space suitable for industrial development, and accessibility to the sit e He ~ h e thr ee m ost i r.1 porta nt con s iderations in industrial si tese lect : ioil ~ lecisions. Any algorithm intended to simulate industrial land use cori~ e s ion should b e construc.ted with these parameters as major co ~ponents. xiii

PAGE 14

CHAPTER I INTRO DUCT ION Problem Background Most urban and regional planners seek to answer the question, "What would happen if ?" The ans1,er is not simply determined. "Even though equipped with the best information system, no decision maker, however adept, can anticipate or visualize the many relationships which must be consid e red in a comprehensive, regional plan. 111 If, however, we could specify the correct development pro cesses and represent them symbolically as a set of equations, the computer c a n approximat e throu g h simulation, years of r eg ion a l growth in a matter of minutes. B e fore we can accurately simulat e regional growth, th e development process e s must first be d e fined and under s tood, and by ob s erving the limit e d success of pr e viou s land use modelin g efforts we becom e acutely awa r e th a t we do not sufficiently und e rstand them. 2 1 Ch ar l es R M eyers Jr., New Tool s for Regional Plannin g ," AIA Jou rna l, Octob er 1971, p. 1. 2 Models should not b e jud ge d simp l y on a s u c c ess-failure b asis but ra ther in terms of th e m ode ls' ab il ity to answer qu es t ions about re a l 1 v0rld. In th e past, s om e m ode l s h ave be e n more s u cce ssful th a n oth ers For a review of previo u s mode lin g efforts one s hould c o nsult : Ira S. Lo wry Seve n Mo d els of Urb an Deve l opment : A Stru c tt1r a l C om p ar i s on ( Sa nta ~lonica, California: RAND Corporatio n, 1 967 ); B. W. ~br a nd IV T. Newe U, A ssess m e nt of Se l ected RANN Envjronmenta l ~lodelin g Efforts. A r epor t prepared for th e E' 1 1 vironrnen t a l Systems and Resources 1

PAGE 15

We would expect land use change or growth to be stimulated or brought about by certain "triggering" events or mechanis ms Many early land use modeling efforts assumed the "triggering" mechanism to be brought about by the location of new industry. We now recognize that other stimulating mechanisms such as improvements in transpor tation or expanding demand for personal services also may prompt regional growth. 3 This research problem, however, is predicated upon the assumption that industry (specifically export industry) is the major stimulus for urban and regional growth and thus deserves initial consideration in any modeling effort. Brown et al. recognized "the critical importance of industrial location ... yet land-use modelers have devoted surprisingly little effort to analyzing the determinants of industrial location. 114 Only a few research efforts have made serious attempts to model the basic determinants of industrial location choices and ''for the most part these attempts have been quite limited and crude. 115 The focus of Division, National Science Foundation (Seattle: University of Washing ton, June, 1973); and 0. Stradel and B. G. Hutchinson, No tes for a Short Course on Practical Applications of Regional Development Models (W a terlo o Ontario: University of Waterloo, The Transport Group, Department of Civil Engineering, November, 1971). 3 Homer Hoyt was among the first urban economists to study the growth and development of u rban land. His primary axiom i vas th a t "no city could grow by taking in its own wash.'' Homer Hoyt, Th e Structure and Growth of R esi dential Neighborhoods in American Cities (Washington, D. C.: U. S. Government Printing Office, 1939). 4 Jame s II. Brown and oth ers Empirical Models of Urban L a nd Use: Sugg es tion s on Research Objectives and Or ga niz a tions, Exploratory Report 6 (New York: Columbia Un iversi ty Press, 1 972) p. 82. 5 Ibid. 2

PAGE 16

th is stucly .ts on the conv ersion of lant.l from prior u ses to indu s trial u :c;c Stat cr.1? nt of Problem An initial requisite of any model purporting to replicate and predict industrial land use development is the identification of those land parcels having the greatest probability to change to industrial land use. It is necessary, therefore, to construct an algorithm \,hich calculates the suitability of some land parcel for a specific indus trial land use. The primary objective of this research effort is to develop a site-selection algorithr.1 which identifies land parcels having the highest probability of attracting industrial land use. More specifically this research attempts to identify and measure the importance of the intrinsic site characteristics which appear to control the conversion of land to industrial use within the 16-county, Knoxville metropolitan region and to formul ate a simulation algorithm to id e ntify potential industrial sites. Problem Operationalization Intrinsic site char ac teristics refer to the qualities possessed by so m e land parcel which id e ntify its sttitabi llty for o sp ec ific land u se In th P. ca s e of industr laJ l an d use, example s of s ome of th e attribut es which should b e examined are: pro ximi ty to rail service; acce ss to nearby hi g hw ays; sit e preparation costs; availability of city services ; or whether the site is \ vi thin a develop e d industrial pa r k. Th e importance of th ese attribu t es 1 nay be mea s ured and subse qu e ntly a gr;rcg ated to fo r m a co rn bine
PAGE 17

The following is a list of indexes which are often identified as affecting industrial iand use choices: Site Preparation Costs--The land to manufacturing use. into account are foundation slope of the land, and type factor does not include the cost of converting rural or open Matters which might be taken conditions, drainage conditions, of vegetation cover. This purchase price of the land. Market Price of Land--The per-acre cost of land to the manufacturer who wishes to build a new plant or expand his existing plant. This would include property taxes but not necessarily taxes on improvements to the property. Proximity to Suitable Labor Force--The number of suitable workers within easy access of the site. Transportation Accessibility--The ease with which people or materials can be moved from the plant site to roads, rail roads, airports, or waterways. In this factor, the concern is with the linkage between the plant and major transporta tion facilities within the immediate area. Utilities--The kinds and quality of utilities ava ilable at the potential plant site. Some utilities which might be considered are water, sewer, and gas. Compatibility w i th Existing Land U ses --The compatibility of general manufacturing activity with oth er existing land uses adjacent to the potential site. Nei gh borhood or Community Attractiveness and Amenities--The condition and density of dwelling units and busin ess estab lishments in the imm e diate area of the plant site, and proximity to hospitals, schools, parks, and churches. Industrial Par k Space--The availability of suitable buildin gs or l and in a n industrial park. The appeo.rance of th e park and th e qu a lit y of services produc e d are some of th e con sid e rations to be examined. This list i 1as compiled from both a r e view of lit era ture a nd prel i minary study. Each factor is a composite of sev e ral variables co ll ec tiv e l y id e ntifi e d a s an ind ex Particular combina t ions of varia bles are not unique bu t s imply r e pr ese nt lo gical groupings based u po n th e liter a tur e search. It i s co1weni e nt for th e purpos e of th is 4

PAGE 18

s t u 9, = I t = ]. LUS Attractiv e ness ::: N I i=l sc o r e W ]. J I. ]. Q, for l a nd u se (in th i s c ase = in d u s tri a l l a n d Number of Indic es Ind ex w ei g ht f o r t he .th i ndex and ]. u se category and th c a t eg o r y u se ), th e th 6 ]. ind ex for th e th l a n d use ca t egorr -, I M ode l O pera ti on Q, l an d I n n~ d e lin g in dus t ria l l an d u se developmen t attractiveness sco r es ar0 uti l ized i n th e fo ll m 1 ing mann er : Froj ec t e cl i ncl1 J s try ?, l'O\\'t h 6 o a k R id g e Na tio na l L aborat o ry R egiona l En v ironmental Systc,n An a l ysis A R esearc h Proposa l Submi t t e d to t h e \at.io na l Scie n ce Foun bt ion Research Applied to ~ationa l Needs ( R..\~:l) r eb 1, 1 973, p. 9 7 Land u se mode l s arc constantly c vu l\j ;ig anJ thu s on e can on l y sp e:1 k of th e fram e1.;ork of tli e model in a hyp o th e t.ic.:il r:nnner T h i s 1 1 : t y or may n ut become th e structur e of tlte r.iode l and r q.Jrese nt s on l y a tent a tiv e vic1, as to th e op e rat.i.n g mariner of th e n ole l 5

PAGE 19

provided via a socioeconomic model is distributed initially to various subregions through a subregional allocation algorithm. Industrial expansion within the subregion is distributed to available land par cels on the basis of attractiveness scores calculated in the above manner. The availability of various land parcels is determined from knowledge of the probability that specific land parcels will be offered for consumption. 8 The land use model as presently conceptualized is perhaps best characterized as a stochastic (Monte Carlo) model. Projected growth in land use will be awarded to various land parcels on the basis of the land use scores (LUS 1 ). Basically, however, the model is heuris tic in that as the land use conversion processes are better under stood, analytic-deterministic procedures will be substituted for simulation-stochastic procedures. Motivation for Study It has been noted that "urban spatial organization is the outcome of a process which allocates activities to sites. In our society, 8 The assumption that all land is potentially available is not a valid premis e in that not all property owners are willing to sell property. Cadastral data and ownership characteristics necessary to develop a site-availability algorithm are not readily available and, therefore, very few land use models have included this consideration. The sociopolitical modeling team at Oak Ridge National Laboratory is currently studying the methods whereby local opinion and knowledge of owner characteristics may be utilized to id e ntify land parcels which m ay be available for various land use activities. See: Osbin L. Ervin and Ch a rles R. Meyers, Jr., The Utilization of Local Opinion on Land Use Simulation M ode lin g : A Delphi Approach (Oak Ridge, Tenn essee : Oak Ridge National Laboratory, ORNLNSF Environmental Pro gra m, 1973). Olaf H e lmar, The Delphi M e thod for Systemizing Jud g ment s about th e Future (Los Angeles: University of California, April, 1966). 6

PAGE 20

the process is mainly one of transactions between owners of real estate and those who wish to rent or purchase space for homes and businesses. 119 The market process of "transactions between willing buyers and willing sellers determines the spatial organization of urban activities ... and thus should dictate the methodological structure of land use models. IO The market place however is not perfect. Individual land specu lation, family or corporate property gifts, over-building by contrac tors, land use planners over-estimating demand, governments exercising property rights, the tendency for land uses to remain intact and the perpetuation of mistakes rather than corrections, all spoil the simple modeling of market transactions. However, after investigating numerous approaches toward modeling land use development, the market place appears to be the most viable way to approach the simulation of urban activities. 11 The site-selection algorithm seeks m1ly to replicate the buyer's considerations in selecting an adequate industrial site. Seller considerations and market perturbations are not considered at this time. The results of this res ea rch effort are intended to integrate with a "holistic" environmental model to be developed later. The motivation which stimulated this research problem, however, was the 9 John P. Crecine, Computer Simulation in Urban Research (Santa Monica, California: The RAND Corporation, 1967), p. 2. 10 Ira S. Lowry, Seven Mod e ls of Urban Development: A Structural Comparison (Santa Monic a California: The RAND Corporation, 1967), p. s. 7

PAGE 21

desire to undcr s t J. nd th e lJ.nd u s e develo pr 1 e n t p roces:s After sofile preliminar y analy s is and r e vl ~ w of prc v i ot 1 s r ese a r ch it b e came appar e11t that tmtl c r s tandin g ai,d mode lj ng land u s e c h o.:1ge c on t a ined pro bl ems to occupy several dissertations. The decision was m a de, therefore, to focus upon one aspect of the land use evolutionary process, the develop ment of industrial land use, in the hope that by expanding the under standing of this process, rigor could be added to the eventual development of a comprehensive land use model. Organization of Study The organization of this study is: the present chapt e r attempts to introduce the reader to the research problem and to rationalize the need for the study. Chapter II discusses th e literature from which a tentative list of site-selection variables w as develop e d. In Chapt e r III, the test region is describ e d and the res e arch m e thodolo g ies and procedures explained. Chapter IV present s the r e sults of an empirical study conducted to identify and determine the relative i!ilportance of various location factors in the 16-county Knoxville r eg ion. Finally, Chapter V swnmarizes the conclusions d e rived fro m th e empirical study and presents recommendations for the construction o f a site-selection algorithm and r e co m mendations for furth e r r esear ch. 8

PAGE 22

CHAPTER II THEORETICAL AND HIPIRICAL STUDIE S RELEV:\~T TO THE DEVELOPMENT OF INDUSTRIAL SITE-SELECTION ALGORIT~W The purpose of this chapt e r is to relate to this rese ar ch effort ideas expressed within selected th e or e tical and empirical studies and to discuss the evolution of th e inde xe s li ste d in th e previous chapter. Di s cu ss ion of some works may be abbreviated depending upon their relevance to industrial site -se l ec t ion processes. Th eo ret ica l \ ;or k s involving cl a ssical location th eory are d isc u sse d fir s t, follow e d by a discussion of empiric a l studies of industrial l ocation f ac tors. The final sections discu ss appro ac hes util i z ed in ot her l and u se mod e l s and th e d e velopm ent of a tenta t iv e li s t of variab l es. Selected Th e ore t ica l \\'orks Th e location of industry i11 a n int ram e tro po l ita n area is th e r es ult of a complex interac ti on of variable s that ca n best b e und e s to od only th roug h careful examination of hi s tori c al even t s Cl ass i c d }o c ation t 1 L ory ho1~ e v e :-, abslrac t s fr or :1 r ea li ty uy u s8 of the prin cip l e cetcri s pa~ibus wh ere; on e o r t1. o vari.:ibl es a re p 8 rr.ri.tt e d to vary while a ll oth er vari a ble s are h e ld con s t a nt V on ThUn e n s a nd \\ 'eber s s tudi es repr esen t early contrib u ti ons to l o ::a t i o n theo; : y an d, thu s d c se rve fir :: t con s i.der:..i.tion j n th i s stu d y lore r ece nt th c orctlcal ~ ork s fo ll ow 9

PAGE 23

Der Isolierte Staat 1 Von Thlinen's work dealt with a hypothetical agricultural land use system whereby transportation costs and economic rent were used to explain the location of specific types of agricultural production as they were concentrically displaced around a market center. In a similar manner, one might expect industrial, commercial, and residential land uses to arrange concentrically around an urban center (Figure 1). For a specific land use the utility (accessibility) derived from any location would decline with increasing distance from the CBD as in (a). This, of course, assumes that the CBD is the most accessible point within the urban area. The optimal price one would pay for utility or accessibility is illustrated in (b). Cost incurred in obtaining utility (Y) is represented by the area (XYZA) but the profit 2 to be derived at utility (Y) is (ABCZ). The profit is similar to the land rent of van Thlinen's agricultural model. Thus, rent for various land uses can be expressed as a function of di s tance from the CBD as in (c). Based upon the comparativ e bid-rent capabilities of each land use, one would expect (Rf\) to repres en t the rent function for residential land use, (QQ 1 ) for wholesaling and industry l a nd use, 1 Joh a nn H ein rich van ThUnen, Der I so li erte St aa t in Bezichung Auf Land 1~ i rtschaft Un
PAGE 24

Page Missing or Unavailable

PAGE 25

and (PP 1 ) for commercial and service activities. The intersections at 3 Y and Z would form the boundaries between the various land uses. Commercial activities are the highest bidders for sites and usually occupy the most accessible places within the city. But commer cial land uses are also found in the industrial and residential zones. Similarly, residential uses may also be found in other zones. For example, a four-story apartment building may yield several times more rent as a one-story commercial operation on the same site and thus may 4 out-bid competitors for the property. Bid-rent functions are a very important real world phenomenon encountered in explaining industrial site selections. Industries seeking highly accessible sites upon which to build must compete against other bidders for those same sites. Consequently cost per unit acre, distance from CBD and proximity to transportation facili ties were included as variables to be examined in this study. Uber den Standort der Industrien 5 Alfred Weber was among the first economists to pose a general theory of industrial plant location. The optimal location for an industrial plant was seen to be a formation of three factors: 3 rbid. 4 Ronald Reed Boyce, The Ba ses of E conomic Geo g raph y : An Essay on the Spatial Char a cteristics o f M a n s E c o no m ic Activiti e s (Atlanta: Holt, Rinehart and 11/in s ton:-Inc., 1 974 ), p 2 6 4. 5 Alfred W e b e r, Uber d e n Standort d er Industri e n (Ch i c ag o: University of Chica g o Pre ss 192 8 ). 12

PAGE 26

transportation costs, labor costs, and agglomerative forces. Weber theorized that the optimal location would be found: 1. where total transportation costs per unit of output were at a minimum or, 2. where transportation diseconomies were offset by savings through agglomeration factors or access to labor. Of particular interest to this study are Weber's ideas concerning, agglomerative factors and proximity to labor. In general, the ideas posited by Weber are of greater importance at the regional or sub regional levels of industrial location than at the site level. Some of Weber's agglomerative factors, however. relate to industrial site characteristics and thus, are considered in this analysis Weber's agglomeration factors are: (1) The joint development of industries which promotes the attraction of auxiliary industries and increases the efficiency of large scale production and utilization of special technical equipment. (2) The development and growth of specialized labor due to the greater opportunity for work in the area. (3) The greater accessibility to raw material suppliers who can provide material regularly and on short notice. (4) The reduction in overhead costs, such as gas, water, elec. d d . 6 tr1c1ty, roa s, an commun1cat1ons. 6 h f Robert G. Turner, "General T eories o Plant Loc a tion: A Survey," AIDC Journal, VI (October, 1971), pp. 25-26. 13

PAGE 27

Th e L oc a tion of Econ o mic B e havi o r 7 The \ m c r ~can econo m ist, bl :-;a r ll oo Ye r i : -:.tio. lly attem pted t o improve ~cbe r s cxp l ~na. t i o n of inJu st r i~ l l o ca tion by i nc l udin g t l1e cons i d e r ac: io n of such v a r i ables dS the size of ra::i rket area and i n s tit ut iona l f o rc e s in hi s th e o re t ica l ana l ysis Y et Hoover a lthou g h c rit.i c a l of We ber' s analys is, ult imately procee d ed alon g th e same lJn es to p r e s e nt hi s l oca t io n th e o ry pri m arily in te r ms o f co s t s H oove r, h owe ver, d i d e x p a nd W e be r s agg l ome r a t ive fo r ces to i n c lud e t h e impor t ance o f b a nks, utilit ies f ire and. po l ic e prote ct io n, 8 r a tes c li ma t e pr op erty t ax a nd low e r in teres;: Ho ove r al s o d e v e lo pe d som e ge n e raliza t i o ns 1n the l ocati on a l h a bits o f li g ht industry ver s us h ea v y indu s t ry Because o f th e n ece ssit y of h a ndlin g o f larg e qu a ntit ies o f g o o d s e ither comin g in from els e wher e or b e in g shipped out, h eavie r ty pes of man u fa ctur i n g 1 ar e housing an d w h o l e sa l i n g p r e fer loc at i ons iE t rans hipme n t zones a lon~ r a il or wat e rw ays ." "Manu fa ctur es ;h olesalers an d wa r e h ouses o f le ss bulky go o ds n ee d not b e loc a t e d on r a ilro a d s o r wa t e rfronts at a ll, si n ce th ey c a n ti e ser v e d by t ruck." They l oca te rr.or e in res p o n se t o '' the a ttr ac tion s of labor sup pl y. c h ea p l and and nearness t o local s u pplie r s o r c u stomers As a r u le th ey c> r e found interspe rsed 1 it.h CJ c orn1rier ci ? .l and inf e r io r resi d ence u ses w 7 Ed ga r M. Hoov e r, The Lo ca t i on of Eco n orai c Behavio r ( Ne w Yo T k: ~c G raw -H i ll, 19 49 ). 8 cit., 27. Turn e r, op. p. 9 Ci t 1 28 -1 ~9 Hoover op pp 14

PAGE 28

The site considerations of industries of the 1930's and 40's have changed 1n m ore recent times but several basic locational rules as expressed by Hoover and Weber remain intact. Accordingly, s e ve ra l variables such as neighborhood amenities, neighborhood compatibility, proximity to labor force, availability of utilities, adjacent land uses, and transportation accessibility have been included in this analysis. Imperfect Competition and the Duopoly Debate A number of location theorists believed pure competition was not a suitable th e oretical structure for the study of plant locations, and sought to explain locat io ns in t erms of the competition between two f irms attempting to capture the lar ges t share of a market area. Fetter 10 and Hotelling 11 w ere among the earliest to expound on f d b d S 12 duopoly location th eory They were ollow e y Lerner an 1nger, Smithies, 13 a nd Chamberlin 14 who expanded th e original concept. lOFrank A. Fetter, "Th e Economic L a1-1 of Marke t Areas," Quarterly Journal of Economics, XXXVIII (May, 19 24) pp. 520-529. 11 Harold Hotelling, "St abili ty in Comp e tition," Economic Journal, XXXIV (March, 1929), pp. 41-57. 1 2 A. P. Lerner and H. W. Singer, "So me No t es on Duopo ly and Spatial Competition," Journal of Politic a l Economy, XLV (April, 1 937), pp. 1 45 -1 86. 13 Arthur Smithie s "Opt ima l Location in Spatia l Competition," The Journal of Political Economy, XLIX (Jun e 1 94 1), pp. 423-439 1 4 E. H. Chamberlin, The Th eory of Nonop olist ic Competition (Cambrid ge : Harvard University Press 19 36 ), pp. 19 4 -1 96 15

PAGE 29

D l S l evl c toglou recently c o; 11 rne nt. ec on t he eco110::1ic irrationalit:, of t h e :i.;1p r oach a nd pr ese nt e d art;u1: 1 ~ nts :: 1g ::1~r> st the theoretica l ba se f or s u ch l ocat i on acti.vitics J rnpp ,ta nt cont!ibrti u n s h ol,evc r, are s ti l l to b e noted Duop o ly th eor y st1 ggJ sts c on s iJcration shou ld b e gi~cn t o the r e p e lling or attr a cting properties of industries which depend u po n 16 loc a l m a rkets. The Th eor ies of David ~I. S m ith 17 an d t,le l vin L. Gree nhu t 18 Smith's and Greenhut's contribution to the the ory of industria l location incorpor a ted the minimum costs approach of Weber with the maximum profit solutions of manufacturing loc a tion posited by August L osch 19 Smith calls this the maximin so luti on. The concept devel oped is illustrated in Fi~1re 2. 20 In (a), the costs of production arc permitted to vary over space (distance) an d revenue obtained (demand) is kept constant. This is essentially the Weber solution. Th e basic Cvih>::p t of Los c h' s :riode l is s h01m in ( b) where r eve nu e i s 15 Nicos E. Oevletoglou, "A Dis se nting Vi e 1 of Duopo ly and Spatial Competition," Economica ( ~ l ay, 1965), pp. 140-160. 16 Turner, op. ci~., p. 31 l ? ll~ v i d i'--1. Sm~ th, 11 1 Th eoret i ca l f rarn ew o.ck ::: or Geo gra ph j ec, 1 Studies o[ Industrial Location," Econom ic Geog_rap h y XL1I ( 1 \ pri l, 1 966) pp. 95-113. 18 Nelvin L. Greenhut, Plant Location in Theory and in Practice (Chapel Hill: Univ ersity of North Carolina Press, 19 56) 19 william H. Weglom and l 'lolfgang F. Stalper (Translators), The Economics of Location, b y ; \ugust Lo sc h (N e w York: John Wiley and Sons, 1957). 20S I mit 1, ~.I?_cit., p. 9 6 16

PAGE 30

WEBER ( 0) ; 1 //11 er Q. 0 z C 1/) I/) 0 u ( b) 1/) I/) 0 u ( C) 1 UJ u er Q. DISTANCE _., LOSCH DISTANCE SMITH PROFIT DISTANCE b REVENUE CONSTANT // / ///: COSTS CONSTANT LOSS REVENUE VARYING REVENUE VARYING .COSTS VARYING SMITH'S MODEL WITH COMPARISONS (AFTER SMITH, p. 96) Fig ur e 2. 17

PAGE 31

permitted to vary over space and costs are held constant. Finally, in (c) the combined solution suggested by Smith is offered with maximum profit occurring at A, where costs are lowest and profit is highest. Note that maximum revenue, however, is obtained at B. 21 An interesting variation of Smith's model was the introduction of f h. ?2 noneconomic actors, in particular, the concept of psyc ic income.This innovation permitted social, psychological, or other personal factors to be entered into the model, hence relaxing the assumption of economic man. Such considerations according to Smith tend to divert the location of a plant from the ideal site to locations closer to the owner's home, a golf course, or perhaps a parochial school. Smith suggests that stochastic procedures may ultimately have to be used to simulate industrial location decisions as personal factors cannot be accounted for by rigorous mathematical reasoning. 23 This research assumes that personal considerations may be accounted for by noting neighborhood amenities near the potential site. The importance of housing quality, proximity to churches, hospitals, schools, or parks and personal services availability in the immediate vicinity of the site are examined in this study. Empirical Studies The following empirical studies were significant resources in the development of a tentative list of site-selection variables. Two 21 rbid-, pp. 96-97. 22 Ibid., p. 108. 23 oAvid M. S mi th, Industri a l Location: An Economic G e o gr aphical Analy s is (New York: John Wiley and Sons, Inc., 1971), pp. 269-273. 18

PAGE 32

typ es of s tudie s a r e pr ese nted: tho s e s t udi es dir ec t e d primarily to 1,.:1rd th e i n du s t ri:il cl evc lop t., r 1, ho seek-; n e 1.-. iridus tr y for a com m unit y ; a nd tho s ~ st u dies 1 v hich at: t e1:1 pt t o an d :,:e indu s t r ia l lo c:=l t i on cor Jidc:.:rat ions by way of a l ::irge s amp l e of n e1~ ind u s tri es a nd to c a t gorize the lo cationa l considerations b y ind us try t ypes Only on e study is directed toward th e dev e l o pment of a sitese lection a lgorithm for a land use model. lhe Studies of Allen Pr e d 24 and Richard Lon s dale 2 5 A ll en Pred has compiled a s tudy of th e history and present sta tu s of industrial location decisions 1,;i thin a m e tropolitan region. These patterns discussed pose interesting hypoth eses for empirica l analysis but the interest of this research is directed primarily to th e site characteristics discussed by Pred. It should be noted that Pred's analysis focused upon a single metropolitan area whereas this study encompasses a region with a hier orchy of ur b:i;-i pl aces In discussing location patterns, Pred id e ntifies seven type s: 1. Ubiquitous industries concentrated near the CBD Th e market area of these industries is generally coincident wi t h that of the metropolis or city. Food proces s ing indu s tries, s pec ifically bak e ry goods, pa ck age f o od s, an d fresh milk pr o du c ts o r e s ~e of th e examp l es of thes e types of industries 24 Allen R. Pred, "The Intrarnetropoli t an Loc ation of American Manufacturing," Annals of the Association of American Geographers, LIV (June, 1964), pp. 165-180. 25 Richard E. Lonsd a le, "Rural Labor as an Attrac tion for Indus try," AIDC Journal, IV (October, 196 9), pp. 11-17. 19

PAGE 33

2. Centrally located "communication-economy" industries Job printing industries, newspaper printing, and adverti s ing printing would be representative industries of this category. 3. Local market industries with local raw material sources These industries show a high degree of randomness in their locational pattern but with some tendency toward CBD locations. Samples would be ice plants, concrete brick and block industries or industries whose raw materials are by-products of other large-scale industries such as the pulp and paper products industry. 4. Non-local market industries with high value products Typi cally these industries provide a high value per unit weight product and are insensitive to transport considerations within the local region. The pattern is "at least superficially irrational." Computer and related industries and chemical industries are t y pical examples. 5. Noncentrally located "communication-economy" industries The subset of industries includes those which are not necessarily pulled to any functional area of the city but rather tend to cluster together in any suitable area primarily because of the necessity to "keep abreast of the latest innovations or forthcoming contracts.'' Elec tronic, military equipment, and space age industries such as found in Huntsville, Alabama, or Houston, Texas, are examples. 6. Non-local market industries on the waterfront Industries where primary raw materials are import ed by water or whose finished product is often moved by water comprise this group. Petroleum refining, coffee ro as tin g and sugar refinin g are prominent among the se types of industries. 20

PAGE 34

7 Indu s tries ori e lltr: : .J t o 1 a nl nati c, 1ul markets Th e se indu 5tri es h l'!C e:-,.;tcns i vc m~1.rke'.:. areas aaJ a.cc inf l ucn ec:c\ by hi. g l i t r:msp0r -t::, ci o7. c os t s on ct bu l ky f ini _<; h cd p roduct A l ar ge pe, : ce nt etg'.':: of thcsr~ ind u s tri es ii ov e: a tendency to lo c::_ 1 1:.c r:;i ~r,8 s id e of th e me tr r,1oli s 2(, oya:np l es R:ir.:lrnrd Lon < :cla.le i n a st ud y of' t ;, ~ -~o cat ion a L habit s of ru:ra1 i:1 J u s t T / uctes tha t those in d. ustT : ; ,-:; a~T:, ~ted by ti a n sp ort a t ioll c:o ,1.; co st : g:-: av i. t ate to nrral are;:i:; ( it adc.l i. tion, L on sd ':!. l.e n,J te s. ru rJ J fi rms t en d to s pace th em:; el v es 01..1 !~ i n )1-der to c1.~,su:i:-e a lab o :r supp 1 /. l llclust ries h'i th t er..dencie ::; Lmia c rJ. 1ci:c2.l loc: 0 t: i. 0ns e .re a p p ;ncJ c h e F 1 L ~ c1 't and e l ect r ica.l J1 1ac hi nery ,,s:)e c iatl y rou tin e assem bly J .r, 1 .-1 p ro f.i.t keen compe ti '?.:i.on, 2J pe~cen t age of 'd OT r-cr.:: ar~ s ome L asic c haract eri.;Lic j 0 th e se j :-i u.s":.ries s it e -s peci f5 c }. l 1 -ihich is t.lw i : o( ;w~ :,f this st ,_,d y. c j t p p l 7 5 J. 7 8 17 L onsdalE, ~ E: c it. 1H elati 011 1rl ?. l

PAGE 35

objective of this study. It was anticipated that after developing a general site-selection algorithm additional research would permit the matching of various types of manufacturing to specific types of sites. Historical Studies of Industrial Location Factors 28 Numerous empirical studies exist of the factors associated with actual industrial location events. Although many of these studies were undertaken by academicians, the viewpoint is that of the indus trial developer seeking to attract new industry or expand existing industry within the community. Characteristically, investigations of this type are not concerned with a specified theoretical framework for approaching the problem 28 Discussion within this section is based mostly upon a review of the following articles: J. S. Bullington, "Utilization of State-Wide Site Evaluation Committee to Aide in the Location or Relocation of Plant Facilities," AIDC Journal, IV (October, 1969), pp. 27-42; James E. Chapman and William H. Wells, "Factors in Industrial Location in Atlanta, 1946-1955," Atlanta Economic Review, IX (September, 1959), pp. 3-8; Ronald E. Carrier and William R. Schriver, "Location Theory: An Empirical Model and Selected Findings,'' Land Economics, XLIC (November, 1968), pp. 450-460, and a more complete explanation of the study: Ronald E. Carrier and William R. Schriver, Plant Location An a lysis: An Investigation of Pl ant Location in Tennessee (Memphis: Memphis State University, 1969); Melvin L. Greenhut and ~larshall R. Colberg, Factors in the Location of Florida Industry (Tallahassee: The Florida State University, 1962); T. E. McMillan, 11 1~ hy r lanufacturers Change Plant Location versus Determinants of Plant Location," Land Economics, XLI (August, 1965), pp. 239-243; N. J. Stefaniak, Indus trial Location within the Urban Area: A Case Study of Locational Characteristics of 950 M a nufacturing Plants in Milwaukee County (Milwaukee: Wisconsin Commerce Reports, 1962); Charl es M. Tiebout, "Location Theory, Empirical Evidence and Economic Evolution," Regional Scienc e Association, Papers, III (1957), pp. 74-86; U. S. Department of Commerce, Industrial Location Determinants, 1971-1975 (Washington, D. C:.: U. S. D epartme nt of Comm erce Economic Develop me nt Administra tion, February, 19 73 ); a nd D. C. Williams and Donnie L. Daniel, "Industri a l Sites for Small Communities," AIDC Journal, VI (April, 1971), pp. 33-39. 22

PAGE 36

but rather to l earn the reasons for th e lo catio n decision from persons acquainted with the location even t. None of th ese studies were con cerned with the construction of an a l gorithm to simulate the process of industrial land u se development. 29 Gre enh ut and Colb erg analyzed factors influencing the decision of 400 manufacturers locatin g in the State of Florida between 1956 and 1957. Location considerations were divided into three groups: demand (market) considerations, costs (assembly) considerations, and personal (psychic) considerations Access to markets and potential markets (Table 1) rated the highest among the location factors with the remaining factors surprisingly low. The study, howev er was s l anted toward measuring regional and subregional factors and thus was of limited value to this study The extensive study undertaken by Carrier and Schriver 30 of plant loc a tions in Tennessee between 1955 and 1965 was conducted within the framework of existing location theory and, in part, did focus upon site-location factors Many of the variable s included in this analysis are based upon the conclusions reached in this s tud y. Carrier and Schriv er identified six classes of loc a tion factors bel ieve d capable of affecting plant loc ations : (1) personal factors, (2) procurement cost factors, (3) processing-cos t factors, (4) distri bution-co s t factors, (5) location demand factors (including location al interdependency considerations), and ( 6) certainty factors. 31 (This 29 and Colb e rg cit. Greenhut op 30c arr1cr a nd Schriver, ~E. cit 31c and Schriver, cit p 45 1. arr1cr OP -23

PAGE 37

Location Factors Access to markets Anticipation of market growth Good labor relations Lower wages Ease of attracting out-of-state personnel, including research Low freight cost on obtaining raw materials and components Low cost on freight on shipping final product Climate as it affects operations Community facilities (education, police, medical, etc .) All other factors Percentage of 400 Plant Listing as Primary Factor 51.9 12.8 1. 7 2.6 4.7 7.7 10.7 1.8 2.9 3.2 FACTORS MOST INFLUENTIAL IN TJ-lE LOCATIO N DECISIONS OF FLORIDA INDUSTRIES, 1956-1957 TABLE 1 24

PAGE 38

class was identified after th e interviews.) Certainty factors were defined as the confidence that the "prevailing and forecasted data used to identify the site offering maximum profits would persist into the future. 1132 Persons involved in the selection of sites for 307 manufacturing plants were interviewed. Each respondent was asked to select six factors from those listed in Table 2 and to distribute 100 points among these six in order to indicate the relative importance each factor contributed to the total plant location decision. Of the 36 factors listed in Table 2, low cost and availability of labor was mentioned most frequently as the primary factor affecting the location decision (Table 3). Personal considerations without economic advantages received the highest average number of points, followed by low cost and availability of labor (Table 4). On the basis of the interviews the authors grouped industries according to the six factors previously listed: (1) Personal F ac tors Miscellaneous manufacturin g, furniture and fixtures, and food and kindred products were highly sensitive to person al factors with most of those firms being "home-grown." (2) Procurement-Cost Factors Industries which need large volumes of low-unit-valu e or perishable raw materials were character istically affected by this group of factors. Food and kindr e d prod ucts, stone, clay and g lass products, and lumber and wood products industries indicated greater sensitivity to these f ac tors. 25

PAGE 39

1. Personal Factors; Personal with economic advantages Personal without economic advantages 2. Procurement-Cost Factors: Better service from seller of raw materials and components Low cost on raw materials or components Availability of low cost raw materials 3. Processing-Cost Factors: Low cost and availability of labor Low cost of fuel Low cost of electric power Low cost of financing project through Area Redevelopment Administration Climate Favorable labor-management relations Low cost of satisfactory type of water Adequate waste di sp osal Low cost of building and land Low cost of financing plant throu g h revenue or general obligation bond s Favorable community and state tax structure Community concessions Avail a ble existing plant Available existin g buildin g Particular characteristics of building site 4. Distribution-Cost Factors: Low frei g ht cost. finished product 5. Location-Demand Factors: Greater demand in the area Greater demand poten tial in the area 6. Certainty Factors: Nearness to metro politan city Community facilities Community planning and zoning laws Cultural qualities of the town Community leaders' cooperation Size of city Data provided by Ch a mber of Commerce, community, etc. Information provided by local manufac turers Recreation, a good place to live, etc. Nearness to corporate headquarters Loc a l supporting services State administration neutral in l a bor management relations Progress in racial adjustment Data provided by the state industrial develop me nt agency LIST OF POS S IBLE FACTOR S INFLUENCING I ND USTRY LOCATIO N AS UTILI ZE D IN THE CARRIER AND SCHRIVER SURVEY 33 TABLE 2 33 Ibid., p 45 3. 26

PAGE 40

Pe rcent Factor of Fir ms Lo1-. 1 cost and availability of l abor 65.6 Low cost of electric power 36.0 Favorabl e l a bor management r e lati o ns 35 .7 Communi t y l ea d ers cooperation 32.2 Low cost of building and l and 1 9.8 Low freight cost, fin i shed product 1 7 9 Availab l e exis tin g p l ant 1 7.5 Favorable community and state t ax struc tu re 1 7.2 L Oh' cost reven u e Availa b le of financing plant throug h o r general obligat i on bonds 16 9 existing bui l ding 16.6 TEN FACTORS MOST FREQUE N TLY ~lENTIO \ E D BY TENNESSEE FIRMS AS AFFECTING THE LOCATI ON DE CISIO N TABLE 3 27 Rank 1 2 3 4 5 6 7 8 9 1 0

PAGE 41

Factor Personal without economic adv a nta ge s Low cost and availability of labor A v ailabl e existing plant Personal with economic advantages Availability of low cost raw material s Greater demand in area Greater demand potential in area Low cost of financing project through area R e development Administration Available existing building Nearness to corporate h e adquarters Perc e nt of Firms 49 2 3 8 .0 32 9 3 1. 9 30.2 29.8 2S.6 27 .6 26.0 TEN LOCATION FACTORS WITH HIGH E ST MEA N NUMBER POINTS ASSIGNED BY TE NNE SSEE FI@IB I NT ER V IEWED TAB LE 4 Rank l 2 3 4 5 6 7 8 9 2S

PAGE 42

(3) Processing-Cost Factors These factors are associated with in-plant costs in assembling or processing the finished product, e.g., labor, energy, external services, capital, land costs, etc. Electrical machinery, apparel and related products, and textile mill products industries were affected by these factors. (4) Distribution-Cost Factors These factors reflect the costs incurred in shipping the finished products to the buyer. Among the most sensitive to these factors were food and kindred products, miscellaneous manufacturing, and paper and allied products industries. (5) Location-Demand Factors Industries affected by these factors are highly sensitive to market-demand in terms of proximity. Included in this category are paper and allied products, printin g and publishing, and primary metal industries. (6) Certainty Factors The validity of existing and forecasted data is consider e d to be highly important by industries affected by these considerations. In other words, th ese industries want to know the future stability of costs in production an d the probable con tinuance of existing markets. Printing and publishing, leather and leath er product s, and transportation industries were highly sensitive to th ese factors. It is obvious that th e scope of th e Carrier and Sc hriver s tudy is much b roader than th e objectives of this study Its utility is limited for thi s r esearch purpose. The factors considered b y Carrier and Schriver span severa l spatial levels of locational d ecisions TI1e result is th at factors which may be very importan t a t the site selection l eve l are w e i g hted l ow in comparison to the total li s t of factors. Also, the disprop or tionat e nu mber of factor s offered for 29

PAGE 43

consideration under th e six categories tends to skew the weightings. Finally, the lack of a very large s amp le rn specific SIC categories tends to decrease the validity of the results of the weightings and, therefore, the conclusions reached regarding the typical locational patterns of specific industries. Bullington 34 offers a scheme to locate potential industrial sites on a state-wide basis by suggesting the scoring of lo cation factors on an ordinal sca le and aggregating them into an index to determine the site potential for specific industries. Local and state governments could then match the qualities of the industrial sites available in the community to specific industries. The factors which 13ullington suggests are listed in Table 5. The U. S. Department of Commerce recently published the partially aggregated results of an extensive 5-digit industrial location survey 35 conducted by mail throughout the U.S. The purpose of the survey was "to assist the nation's underdeveloped and declinin g areas in the development of th eir economic resources and potentials." 36 Only manufacturing industries demonstrating "reasonable" growth between 1958 and 1967 were selected for inclusion in the survey. Survey forms were mailed to a total of 2,950 companies in 254 different 34 J. S. Bullington, "Utilization of a State-\Vide Sit e Evalua tion Committee to Aide in tl1 e Location or Relocation of Plant Facilities," AIDC Journal, IV (October, 1 969), pp. 27-42. 35 u. S Department of Commerce, nants, 1971-1975 (Washjn g ton, D C.: Economic D e velo pme nt Administration, 36 rbid, p. 1. Industrial Location Determi U. S. D e partment of Commerce, February, 1973). 30

PAGE 44

Site Characteristics a. Si ze of Parcel b. Shape of Parcel c. Topography d. Drainage e Flood Record f. Condition and Appearance g. Underground Water h. Soil Bearing Capacity Acceptability (This referred to the potential friction or good-wi ll prompted by th e location of industr y ) Accessibility a. Highway b. Secondary Roads c. Rail d. City Water e City Sewer f. Limitations of Site Crnnmunity Factors a. Commercial Air S e rvice b. Water Tran sp ort c. Location in State d. Mileage R a te e. Airport Facilities f. Comprehensive Planning and Zoning g. Retail Acconunodations h. College i. Conununi ty Appear a nce a retail b. resid e ntial J Hi E hways k. Presentation of Facts by Community 1. Sanitary Sewer and W a ter Treatment and Facilities LOCATIO>l FACTORS SUGGESrED BY BULLINGTON T A BLE 5 31

PAGE 45

S IC c at e gori e s On e for m S ur v e y of 1ncl u~ tr i tl L oca tion Det e r m i n;i.n t s :as t o be c om pl e t ed b_ v a ll c o, r,p u n ics to i ,l ntif y t he l oca ti o n a l a nJ op c r a tin ~ ch a r ac t c ri s t i cs of ex i s ti rig pL tn ts 01 1 l y t he re s ult s of the Surv e y o f In
PAGE 46

at the University of North Carolina by Don e lly, Chapin, and Weiss. 37 The primary purpose of the model was to simulate the g rowth and spa tial spread of residential land use. Historical data (1948-1960) were utilized to calculate an attractiveness prob a bility to simulate the conversion of rural land to urban residen~ial use. Changes were recorded by means of 1000-foot cells composed of nine 2.5-acre land development units. TI1e model operated 1n the following manner: 1. All land within the city unsuitable for dev e lopment is eliminated from consideration at the beginning and the supply of land remaining is identified as available for residential us e 2. For each 1000 foot cell, a measure of relative value is established, 1.e., land value, as a measu re of its attractiv e n ess for residential d eve lop men t. 3. The effect th a t "primin g (exp ansion of municipal scrvil:e s commercial services and industrial d eve lop men t) deci s ions will have on modifying th e valu e of the property is then calculated for each cell. Thes e are assumed to be exogenou s ly given but in thi s case the exact amount is known from hi s torical dat a from 1948 to 1960. 4. L a nd parcels are th en "r eassessed to obt ain a n c1-r attractiv eness scor e 37 -nwmas G. D onne ll y F. St u ar t Chc1pin, and Shirley F. W e iss, A Probabilistic ~1occl for R es identi a l Gro1vth (Chap e l Hi ll: Univcrsjty of Nor th Carolina, In s t i tute for l {c s ea1=-c;1 in Socia l Science, 1 964) 33

PAGE 47

5. D e n s i ty c on' ;t r ~i .i. nts (n u rn b ers c,f units p e r a c re/y e a r ) a r e th e n i ntroduce d. 6 F .i.n a ll y th e K.11 0i d l g r th uC rt '::-i d en tia l hou se h o l d s bctHee n 194 8 a nd 1 96 0 is ;:i ll ocat e d pro b abi lity b as i s 3 8 on a Th e dr a \vbucks to th e m odel a re: (1) the n ee d to have historic a l data in order to dev e lop th e tr ans i t.i e,n p r oba bil i ti es ; ( 2) the disr gard of the effect of ch a nges in indu s tri a l lo c ation and employment; and (3) the vast amount of pro g ramming ne e ded to load and run the model for a small region such a s a sin g le city. The conc e pts g e ne rated by this study, however, have b e come b as ic to many oth e r modeling efforts and were equally important in this study. The Pittsburgh industrial location model, INH!P 39 (.!__ndustrial Im_pact Model), is similar to the hypothetical design of this modeling approach; the major difference is that growth is distributed to cen s us tracts and co11sequently the variahl es ar e m ore a gg r eg ated than those being considered in this study. Four variables (attributes of census tracts) and on ~ constraint were adopted as bei ng sufficiently discriminatory t o determine site locations. rhese are: weighted m e an unit-as s es se d value o f l a nd; wei g hted mean unit assessed value of buildings; w e i g hted m e an structur a l d en sit y a n d amo unt o f i n
PAGE 48

ar ti f i cial l y as 1 n th e case of z onin g contro l s or n o i1-ex i s t e n cc of servic es ; or dir e ct l y imposed by t h e n1o d e l oper ato r. O n t he b asis of t h e aggre g ate d scores o f t h e aha v c ind exes the mode l dist r ibute s a p or tion of p r o je ct e d cit y -1 id e employme n t c hanges amo n g c xist~,i g facilities and, upon r e achin g c e rtain c r iti ca l v a lu e s of s atur a tion, s1-ih:ch e s to a s e par a te ro u tine to di str i bu t e n e 1 ~ f ac i liti es to c e n s u s tr a ct s h a vin g the highest scores fo r th e r emai nd er o f th e proj e ct e d industrial e mployment growth. The distribution to various census tract s is accompli s hed by way of th e m ax imu m scor e Howev e r, in th e ev e nt of a tie, the allocation al gor ith m s w i tch es to a M onte C a rl o routine. Similar but more e legant model s of thi s t y pe are being d 1 d 1--1 d 4 0 d h U . f B . C 1 b. 41 eve ope at arvar an t e n1vers1ty o r1t1sn o um 1a. A Tentative List of Variables 35 Each theoretical study revi e w e d appr oac h e d th e p ro blem of explaining industrial location in terms of three components: demand, cost, and personal factors. Carrier and S c hriver subdivided th e proces s into six components: p e rson a l f a ctors, p ro cur e m P, nt -c o s t factors, processing-co s t factor s distributi o n-c os t factors lo ca tion demand factors, and certainty factors. Onl y proc e ssing costs, procur eme nt cos t s di s trib ut in g c o st5 perso n a l costs, a nd cer ta i n ty 40 carl Steintz and P e t e r Rogers. A System Analysis Model of Urbanization and Change: An Exp e riment in Interdi s ci p linary Educa tion (Cambridge: M.I.T. Pr e ss, 1971). 41 M. A. Goldberg, Quantitative Appro a ches to Land Mana g ement (Vancou v er, B. C.: University of British Colu m bi a Th e Re s ourc e Scienc e Center, 1970).

PAGE 49

facto1 s !~ ave any
PAGE 50

1. Industrial Park Sp a ce 2. Industrial Park Quality 3. Industry Nearby 4. Zoning and Buildin g Restrictions 5. Pollution Regulations 6. Land Costs 7. Site Preparation Costs 8. Topography 9. Drainage 10. Soil Conditions 11. Place to Dump Effluent 12. Processing Water 13. Utilities 14. Municipal Water 15. Sewage 16. Natural Gas Service 17. Proximity to Local Markets 18. Proximity to Loc a l Rm~ Mat e rials 19. Proximity to Sup porting Industry 20. Water Transport 21. Ra il r o a d Tr a nsport 2 2. Highwa y Transport 23. Commercial Airport 24 Distance to CBD 25. Community Transportation 26. Community Parking 27. Ne ighborhood Services (Re s taurants, Hospitals, P arks, G a s Stations, etc.) 28. N earby Housing 29. Nearby Labor 30. Community Cooperation 31. Community Stability 32. Community Wealth 33. Community Taxes 34. Community Progressiveness 35. Community Attractiveness 36. Comrnuni ty Labor Climate 37. Community University 38. Nea rb y Gov e rnment and I n s titutional F a ciliti e s 39. Community Wage Rates 40. Space for Expansion TENTATIVE LIST OF VARIABLES TABLE 6 3 7

PAGE 51

a h. c cl f g Di st a ncc t o maj ) Y h i g h \, ;r: O j ~ran ee t o sec o nda ry n iJ,J Oi sta 1 1cc to r.1i.l Di stan ce t o ai r n or t ~ a t e rw ay s e rvic e Dist aace t o n earest In t e r s tate interch a ng e Ov e r a ll qualit y of accessibility fr om 1 to 1 0 V. Utiliti es a. Water a va il a bl e b. G as avai lable c. Sew e rag e availabl e VI Compatibility with Ex i s t ing L an d U ses a. Did community h ave zo ning at time of lo ca ti on? b. Was site z oned for indu st r y? c. Was zone c:hang e d to ac c e pt industry? cl. Was industry already in im me diate area? e. Overall rating of contiguous land u se compatibilit y VII. Ne ighborhood or Community Attr ac tiveness a nd Ame niti e s a Condition of neighborhood b. Density of land use in i mmed i a te vicinit y c. Nea rby community se rvice s VIII. Industrial Park Sp ace a. Was th e site in an industrial park ? b. Overall rating of the qu a lity of p ark ? Additional Data Coll ec ted a Proxi mi t y of si t e to Kno xvi ll e b Amount of other indu s try locat e d nearby at ti me of ev ent c Was bui ldin g already th ere? Thi s li s t may o m it variables which sh ou ld be co nsider ed a nd, th e refore, should not be consid e red exhaustive H oweve r, it is a ntic ip ate J th at in measuring the i mpor t a n ce of eac h variabl e so me ma y b e eliminated thus simplifying th e site-sel e ct io n a l g orithnt. The next ch ap t er exp lain s th e me thodolo gy an d res ea rch procedur es u se d to test t h e impor t ance of ea ch vari u bl c 38

PAGE 52

Cl L\ I fER II I DESCRIPTION OF THE STUDY AREA A ND A N ALYSIS PRO CEDURE S This chapter describes the research me t hodolo g i es utili z ed in this st ud y The first part of th e cha pt er d escribes th e eas t ern Tennessee study region, th e second p art the d a t a coll ection proce d ures and th e final p ar t the analy s is proc e dures. The Study Region The s tudy r egi on encompasses 16 counties surrounding and including Knoxville, Tennessee, (Fi gure 3) and r epresents an adminis t ra t ive entity c a ll e d the Eas t Tennessee Development District (ETDD). The region sp a ns 6 ,SOO squ are mil es an d contains a population of appr o xi mately 750,000 people. Its selection fo r this s tud y was b a sed upon the availability of d a ta in the O ak Ridg e Na t io n a l L abora tor y (ORNL) 1 Data Base. OR N L selec t e d th e re g ion on th e basis of "th e diversity of the r egi on, th e availability o f dat a the presence of co ope r at i ve and inter e st e d user gr ou p s, c l o s e proxi m it y e tc ." 2 1 oak Rid ge Na tional L a borator y is curr ent l y d eve lop i n g a "holi s tic" environmental model for th e ETDD r egio n. Th is research was supported b y this pro gram ORNL-NSF Environmental Program, Regional Environment a l S ys t ems A nal ys is ( A Research Proposal Submi tt ed to th e National Science Foundation, Fe bru ary 197 2). 2 Ibid p. 3. 39

PAGE 53

~ ,---,-------1\ -,-----------,----,---------~ -~ \ ---_ / ___ ------~-~ -~-~ \ Monroe EAST TENNESSEE DEVELOPMENT DISTRICT Figure 3. COUNTY BOUNDARIES I --1 -] J -~! __

PAGE 54

Th e E TDD reg ion i s c en t ered in t h e southern p or t i on of t he f { id ge an d Va ll ey P ro v ince (t h e "C cat Va ll ey '' ) b o rd er e d t o th e n o:t h e s t by t h e Cu m b er l an d Pl ateau and ~lou ntai n s, an
PAGE 55

' PHYSIOGRAPHY Figure 4 '":. ~ t.b .,. I ~ :_.. .... .. ~ ( ~ ~.... ,.,. --1,;:__..,__ ~ ;o,_= ~ ; ~ ~ \ t Z~Ji!:t;+~ ~ ti >, ~~:r11ree D,m en s1on o1 R e h el of th e fas, Te-nness ee Reg ion I I

PAGE 56

1-__J v 1-+--+ --+L.. 6..C,.,-" + --,---1 __ --+ -"" +1----l---+ ~ -l--+ --1----j e,_ ~ -~L/~r--------+--v A -" f----4 --1---l---'---' ---'---L..-.. ---+--,l-l -\ I / 1---------+---+ l-,J ~-; ... ~ ----~ ~LJ ~ _I Figure S. --1--~ --T OWNS ANO COMMUNITIES I I I r -1 _J

PAGE 57

d ict a t e d mo re by t he dis tr:i. bucion of r aw mi.ttcr:i.a ls and c o nsc q i. t c ntlr th e l o ca t.io1n l c r i. t e r .i a of the sr .; in c.l u st r i. e:s 1 -1 d 1 b e d i.f fi::!re n t fro m secondary iHd u s tric s F or th is re aso n prirn ir r indu st ry i s n o t con sid e r e d i n thi s s t dy. Cate g ori ca lly on e could st a t e th at most of th e industry within the region is c oncentr a t ed in t he Knoxv i ll e a rea. Of th e 1 00 2 indus tri es in the r eg ion, app roxi m at ely 46 perc e nt are lo ca t ed in th e immed ia te vicinity of Knoxville. Knoxvill e also has the greatest diver s ity of industry, whereas industry in oth e r co mm uniti es is charact e ri ze d b y s pecific categori es (An examp le is th e concentration of fu r nitu re industry in Morri s town.) Figures 6 through 13 provide a visu a l ov e rview of the dynamics of industrial development in the re gio n since 19 43 Since 1953 a 48 percent turnover in indu s try has occurr e d. Of the 100 2 industries within the re gi on in 1973, 488 ha ve d e velo pe d in th e 1 6 co u ~ty r egio n sinc e 1953. Table 7 provides a statistic a l s ummar y o f th e new industry-new employment expansion (or d ec lin e ) f rom 19 52 to 19 7.3. Th e t a bl e do es not reflect th e abs olut e employment expa n sion rate bu t ra th er th e a mount of employment gained or lo s t b y the birth or d ea th of specific t w o-d ig it SIC in dustr ies A hi gh aver ag e e xp ansion rate i ndicates n e w plant s ha ve d eveloped in th e r e gi on an d h ave sig n ificant l y i ncreas ed the total employment in that specific industry. In the following discussion each indu s tr y typ e is briefly de sc ribed. Maps are included to show th e r egi onal distribution of sp eci fic industries. Graphs illustr a tin g th e histori ca l employment 44

PAGE 58

, T--~' -~ ---iJ ,-I --,7 \ I / I \ I I ,i-I .. I-~ ,-_: . i <~ r --~-. A J . ;. ~ ~ I --.---~ ( I .<:I ; : ... ( . . I\ r i ~ . J -..":,': ,. \ : : 'r . : .-< .. .. v .... j : < : D .. .,. ... ,1-v'a : : l---1 lrJ hA ~ V \ IL/ PLANT LOCATIONS 1943 \ V : \ I I < ) I -~ ,r\. '----J I I I I I I I I __ _l_ __ .....J Figure 6

PAGE 59

. I ---,------, \ ~--+-1 \ \---+---'-+----+--____j____j__~ u f-------+-----\ ~ ~i-. ~ LJ .. .---,. -I II ; . . . .., I ~: . .. I . . . J . : .,_ 1 r. \__.., Figure 7 -7-/ I \ I I I I I EMPLOYMENT 1943 0 0 1 SI I E NO. Of [ M PL OY EES I I _L -" I 24 25 49 50 99 1 00 499 500 999 I 000 4 999 5000 O VER I I I ----1

PAGE 60

r ---,7 I T \ / \ : I\ . \ .,. I .. < r ... l ..s'~ A LJ \ . .. ( ;l, ,;;. \ -. / : . . .;c, ( .. . . . I"'\ : . I \ I .. ... .. .. . > I"-. . \ .. i,. --:: ~: l: _; : r. .. Ir t . v ... V < > .. : > D ,'l-. v------.:, ..... .... ,-.; l---1 .. -~ h,,__ , lrJ "'.r I / :~ I'\ /L/ PLANT LOCATIONS 1953 1 ~ \ V : \ l
PAGE 61

. .. \ \ I \ .J \ .. I e
PAGE 62

.,. i ----I \ / \ [\ \ I : . .. <~ lJ lr'~ . vi\ .. : \ . ( .. ~k1 / :0. .( ... J .. .. ,.. I ~ I\ .. I>. . .. . ,. ... .. r-1il: .. I',. I f .. .r I'-,. \ I ~ .. r: ~---.~ ~ .. I j .. .. v . ~-. ... < .. .: ... D a) .. v-------.... ,. ,-Lv .. l-.-1 lrJ \..__, .A .,, I / 1. I\ IL/ PLANT LOCATIONS 1963 1 : \ I / \ 1 2 ~ ) I -"' Figure 10.

PAGE 63

\ .. f\ \ -. .. LJ l[-' .. . . ( .. !". ill, I'"' I\ . i -. . tt.i : \ ...,. . .. 11 -, .. .. ~< : ) 4 .. ~~ --Lv .. L..,,, . i I\ L/ \ I/ \ I < j -~ Figur e 11. / \ I ,, .. ~I'" A I\ .. : . . ( / I ~ . -~ .. ,. -; . ,.L-J .. lrJ r-----. V EMPLOYMENT 1963 I I 001 S il E NO OF E HP L O>EES I 2 4 25 4 9 50 99 1 00 499 500 999 1000 4999 SOOD OVER V, 0

PAGE 64

,----, . --~ ----~ \ I\ -\ .; . .. lJ .. -. '( .. \ . i'" I i' I\. .. ). . ... -~ . . . / "!. < ) D Iv et> L.._,,__ . .i I\ ., \ '\ I C j /\ I / I . ... : j : ,, .. ..... \ "f . .,, .. .. .. , : f..... .. s .... IL/ V Figure 1 2 \ I .. ; I .. <~ ~ A :I\ / 1.:r \ ( .,. . . I \ I'.,_ .,. :-,. .)..~ / v------,L..-J V PLANT LOCATIONS 1973 I I \JI t--'

PAGE 65

\ \ \ . . _. J ., . . t I I\~ .,,; . \. < .. .. .. ,. I ii . .;. [/" II!' : < . .. D ., -~ . ... . Lv .. .~ h "" . s._ ; I\ . ,L/ \ I / \ I ~ ..,,. J r--. l __ '-~ __ l__ _l_ -~j__ F ig ur e 13. / _.., . .. \ I <~ LA I\ .:.. . V ~( C .. :-----.. ._ V ,c:r V EMPLOYMENT 1973 ' 001 SI ZE NO OF [ HP LOYEES I 2 4 25 49 so 9 JOO 499 50 0 999 4999 5000 OVER I I I I I I I I _j__ I -~ _j_ ~ V, N

PAGE 66

5 3 A vera ge Ne .; Jncl us t r y Expansio n b y R a te SI C 1 955 1~ 58 1 960 1 96:s 1 %6 1 969 E J 73 1 9:>2 -1 973 20 Food +9.4 +10 0 +0.5 +8 .7 +6 5 -3. 2 +11. 8 +3.8 21 Tobacco 22 Tex.tile -4 6 -2.7 7.5 + 8 .3 -3.9 +5 2 -19.7 -3.6 23 Apparel +28.3 -13 6 +2.3 +11.2 + 2 0.1 +40.3 -12.4 +10 9 24 Lumber -40 4 -3.2 -2.2 0.3 +2.1 -7 2 +11. 2 -5 7 25 Furniture +24 0 +30.4 +53 1 + 5 .2 +13 8 +5.9 +7 8 +20.0 26 Paper -4.7 -16.4 +12.3 +4.2 +10.2 +49 1 +41.8 +13.8 27 Printing +9 3 -2.0 +24.5 +2.2 -4.1 +21. 0 -3.3 +6 8 28 Ch e mical -.6 +13.4 -4.3 -14 7 +2.3 -2. 4 -25. 6 + 2 .8 2 9 Petroleum -8.2 +4.7 +5.2 +27 2 -4 1. 9 +43.0 -5.0 +3.6 30 Rubber 1.4 -11.6 +9 8 +79.5 +10.3 +61.6 -32.4 +16.5 31 L e ather -9.7 7.1 +25.8 +25 9 +13.1 +62 6 +63.5 +24 6 3 2 Stone Gl ass Etc. +] 4 +3.2 +7. 6 + 3 -] 1. 0 4 4 +25 .5 +3 2 3 .S Primar y Metals -6 9 +10.1 + 5 -22 .8 -4.6 6.9 2 4 -4 7 34 Fabricated Metals +36 6 +12 0 +.8 +10.1 +4. 7 +36.8 +6.3 +15 3 EXPANSION OF ETDD INDUSTRIES BETWEEN 1952-19 7 3 TABLE 7

PAGE 67

----------35 I ndu st ry b y SIC n a ct 1 i ne r y Non-Electric 36 Electrical Machiner y 37 Transpor tation 38 Instrument J SI 19Sr 1 958 1 960 1 963 1 96 6 Average Ne 1, Expa n sion Rate 1 96 9 1 973 1 95 2-1973 3.4 + l ~ 2 S 6 ~ .s .l 15.C + 5 5 3 +O .S +6 7 -4 'l .7 1 3 -6 .9 ~ 2 1.1 +30 1 +1 9 .1 +6U.9 +26.4 -17. 2 +22 5 +3 7 -1 3 2 +136 .5 +85 .4 +37.3 -1 _2 3 9 + I:. 5 -1.l +9.5 -3.3 +3 .3 t.liscellaneous -15.7 + 2.6 -8.4 +34 5 + 4 4 +11 .6 26.1 +1 0.6 *This t a bl e is b ased upon OR N L ind1 .. s trial loc .::tion data Expansion rates are determined on the basis of new employment prompted b y n ew p l ants ( o r death o f plan ts) sinc e th e pr ~ viou s t ime p e r iod Th e base y e ar i s 1 9 5 2 T AiJ L L 7 C o JL t iH u c ci. 54

PAGE 68

trends relative to total re gi onal employment and the industry's share of total regional manufacturing emp loy ment ac comp any each map. 4 Food products industri es (SIC 20) 5 Of the 103 plants exis ting within the region, 66 are located in the Knoxville re gi on. Most of the remainin g plants are located in the southern half of the region bordering the Smoky Mountains near the better crop producing areas along the margin s of the Nolichucky, Pigeon, Little Tennessee, and the French Broad Rivers (Figure 14). Approximately 75 percent of these indu s tries employ less than 100 persons. Employment growth has been gra dual and in accord with total regional growth. Proportionally, how ever employment has been de clin i ng (Figure 15). Tob ac co products industries (SIC 2 1) N o tobacco industries exist within the st udy r egio n. Text i le mill products indu s trie s (SIC 22) The textile industry is distributed bro ad ly from north eas t to southwest throu gh th e center of the region with Knoxville having the greatest concentration follmved by Morristown and S'eetwater (Fi gure 4 The m a ps and graphs ut ilized in tl1is chapter w e re prepared by a CALCOMP plotter driven b y an IB~I 360 computer Th e data utilized were originally collected by O s b in L. Ervin For a complete pr ese ntation of th ese d ata, se e : C. R M e yers, Jr., 0. L. Ervin, D. L. Wi l son, and P. A. L esslie, Spatial Distrihuti ons and Emp loym ent Trends of ~l a nufa tur ing Indu s tr : i es in East Tenn essee (19 13 -19 73) (O ak Ridg e T en n essee O ak Rid ge Nationa l L abo ratory, Jun e 1974) 5 Tra n sp:ne nt overlays found in Arp en di x B m ay b e superimposed o ver eac h i 11 clu s try rn::tp for spa ti a l r eferencing 55

PAGE 69

. ~r \ -t\ ~ --+---= ~ ~~-4------1--+---r--, ----+ f ---.s c-F--+--+---+----l----------+------+-------+--,:,__ f-----tt---t-----+---+---+---+---+---+---+---+---l--+ I\ t / \ I l <:'.~ ~ A -' ( "' -+-----+---< -+---+---1--------l---+---+-----+\ I 19 J ,;,.----1 vJ V / < -~-f--------4> --------+----+---+--~--+-i--7 PLANT LOCATIONS J 1973 e .. . t--t-----+---+-~ ,I -l-_L_-_ [--_. ____,__ -.: ~ . --~ ----'--~-~ ----SIC 20 FOOD PRODUCTS Fi g ur e 1 4 V, 0\

PAGE 70

:r 0 1940 ---R e gl o n S IC 20 1945 1950 1955 1960 1965 EMPLOYMENT IN SIC 20 AND IN TOTAL REGIONAL MANUFACTURING 1970 1975 1980 0 I/) 0 en I.fl (\J 0 (\J I.fl 1940 SIC 20 FOOD PRODUCTS Figure 15. 1945 1950 1955 1960 1965 1970 SIC 20's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) 1975 1980 V, -.J

PAGE 71

1C1). Ave rage em p l oyme n t 111 th e indu s tr y i _..; appro:-.:i r;:3.tc l ~2 1 0 pe rso ns p er p l a nt Ro c kwood a nd Harr ima n h a v e only fou r plants but shar e n c, i rl y 50 9 of th e to ta l emp l oym e nt in th e region T;1e dist ributi o n of pLrn ts r ef lect s a t en den cy to g ravit a t e t o r e 1 ..1ti vc l y sma ll c omrn uni.6 7 ti es as sugges ted by Lon sda l e a nd Pred. Textil e industr y h as declined in tP,rms of the percent of tot al regioJ1al emp l oyment in recent years perhaps resulting from the introduction of higher wage industry in t o th e r eg i o n (Figure 17). Cl othing an d re l a t ed products (SIC 23) Knoxville leads the area with the g r ea test c once ntr at ion in b o th pl an ts and employment (Fi g ure 18). Average employmen t is a pproxi mately 315 person s ; however, the median is clos e r to 15 0 emp lo yees Thi s type of industry thou g h concentrated in Knoxvi il e i s scatter ed throu g hout the region with num er ous small co m munities ha v in g at least one ( s om etimes t 1v0 or three) small em ployment ind us~ry In te rms of new industry expansion th e clothin g ind ustr y expanded a t a n average r a t e of 10.9 percent between 1953 and 1973. Historically, however, gro wt h has b ee n erratic (Figure 19). Lumber and wood products exc e pt furnitur e (SIC 24) This indu s tr y i s wi d e ly sca tt ere
PAGE 72

\ I\ \ ut \ rs. \ I/ < D Lv h,,,_ \_ I I \ \ / 0 .. .. I'------/________/ V \ I /"'\ j -SIC 22 TEXTILE MILL PRODUCTS Figu r e 1 6 \ I <~ ~r\ V ( I"-. l/ v----' 1,..---1 ,lrJ / PLANT LOCATIONS 1973 e = v ~ I)" U' 1" 1,r V, \0

PAGE 73

1940 _.. Reg l o n S I C 22 1945 1950 1955 1960 1965 EMPLOYMENT IN SIC 2 2 AND IN TOTAL REGIONAL MANUFACTURING 1970 1975 1980 0 l/) l/) :::J' 0 :::J' lf) (T') 0 (T') 0 N 1/) 0 l/) 1940 1945 1950 1955 1960 1965 1970 SIC 22's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 22 TEXTILE MILL PRODUCTS Fi g ure 1 7 1975 1980 0

PAGE 74

( \ \ l/ < !) .... lvo I\ ~___,,~ I ,; >---+-+---1~ +--\I \_._ r ------+-------+-------+--l / '..J-L/ ----1------1------1-------+-------+---+---+----1 \ \v 1 2 J f----+ ----+---1-------1---SIC 23 APPAREL AND RELATED PRODUCTS Figu r e 1 8. PLANT LOCATIONS 1973 e I ---~ ---"'---_ ___J___ -'---___ Q'O \~O" .... 10" II"' ~ 0-, ......

PAGE 75

"' 0 V, 1940 __. --Reg i.on S I C 2) 1945 1950 1955 1960 1965 EMPLOYMENT IN SIC 23 AND IN TOTAL REGIONAL MANUFACTURING 1970 1975 1980 0 U") U") (T) 0 (T) U") ("\J 0 ("\J U") 0 U") 1940 1945 1950 1955 1960 1965 1970 SIC 23 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 23 APPAREL AND RELATED PRODUCTS Fi g ur e 19 1975 1 9 80 N

PAGE 76

. ,, ~ r n ~ .. ,, I I .. ..,. I I I ~ ,. -' " W I ~ I r l j_ I I I I ., i I I I I I --I I .. ,. I -'t i I I 1' 1 I I I PLANT LOCATIONS 1973 e SIC 24 LUMBER AND WOOD PRODUCTS (except f urniture) Figure 20. 7 t 1" Ji ,. rr ,. J a~

PAGE 77

decLi.n .::: i n th e indu s try h as o;:: r.: urr e d p a rt.L y
PAGE 78

1940 __. 0 1/) ----1945 Reg l o n S I C 2 4 1950 1955 1960 196 5 1970 EMPLOYMENT IN SIC 24 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 1/) (T) 0 CT) U") N 0 N U") 0 U") 1940 1945 1950 1955 1960 1965 1970 1975 SIC 24's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 24 LUMBER AND WOOD PRODUCTS (except furniture) Figure 21. 1980

PAGE 79

V V t r \ / \ \ \ I <~ l,J ( v\ -. #,:. ( ,. I'I -) ,->.t. 'I/ < D ,lv' L..--1 I I~ I \ IL/ f-"---_._.__ V \ I / \ k: ) I /"'. __ ,___ -----~---'-SIC 25 FURNITURE AND FIXTURES Fi g ur e 22. '--PLANT LOCATIONS 1973 e I L I\ I / I ~ I'---. J '

PAGE 80

__. --S I C 2~ 0 I.fl If) "' 0 "' I.fl ('") 0 ('") I.fl (\J 0 (\J 1 9L40~....._.1_9L45..___._....L.....1,1-9s'=:o-'---'-....L.....l,1-='951:-:s~~1-='9se:-:o~--<---f-1 ~95;;-:s~--<---f-1 ~97;;-:o~--'---7-1 ~97:;;:sc-'--"---7-;:;'.19so 19:..40~~1:;.45~~1n95:c-;oz---'""--'-----:1-;;:9s~s~--'-----:1~9s;c;:o~----:-, ~9s:;-;,s~----:-1 ~97:;--;;0~--'---7-1 -=-197==5~_,_,_, -=-1 980 EMPLOYMENT IN SIC 25 AND IN TOTAL REGIONAL MANUFACTURING SIC 25 FURNITURE AND FIXTURES Figure 23. SIC 25's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) O'\ .......

PAGE 81

Paper a n J pu l p produ c ts i n d u3tr i e s (S I C 26 ) Th e pclp er a nd pu l p i n dust ry j s conc c n t r cl t e cl rn the K noxvi ll e ~ tr ea ]:-;ot h in t c n11s of c m p l oy-;nent and nu rnb ( ~r of p L1 1 1 t s (Fig ur e 24). Int u i t i ve l y on e s u s p ec t s m arket-~ ~ ma n d t o b e a st r o n g lo ca t i on a l f a ctor e x cept for the plants in Mor r i s to w n whi c h are lo ca tionally r e l a ted to th e b y -products produc e d b y t h e furnit u re i n dustry Employment averag e s 7 8 p e rsons with th e la rges t pl a nt employm e nt only 190 persons Total employment in the p a per and pulp products industry is relatively low compared to oth e r industri e s, only 1377 employees in 19 7 3. The averag e growth r a t e h o we ve r, h as be e n si g nific a nt (Figure 25). Printing and publishin g industries (SIC 27) Printing and publishing includes local n e wspaper printing and for obvious reasons is widely d i sp e rsed throu g hout the 16-county r eg i o n ( F i g u re 26). ~ 1 ox vill e h as th e gr e atest co n centra t io n with over 75 percent of the employment and 2/3 of the total number of establishments. Total employment is low for th e r eg ion (1670 employees in 1970) with an average of 19 workers per pl a nt. G rowth ha s increas e d al o ng w ith popul a tion wi t h no g r e at expa n s io n pr e d i c ted (Fi g ur e 27). C hemi c al p roduc ts indu s tr ies ( S IC 28 ) The shunnin g of 1 ar g e urban p J 2 c es by che m j ca ] pl a nt s not e d in Chapter II at fir s t appearance is not substantiated by the regional pattern of chemical industries with i n ETDD (Figure 28) Industries are found in the other communities (e g., Morr i stown and Oak Ridge) b~t Knox v ille is the le a der in numb e r of plants with all but si x of th e 33 plant s 1n th e r eg ion. H O\ve ve r t h e Kn ox v j J l e in d u s trj e s a r e 6 8

PAGE 82

r \ 1 I \ ,__ t--\ ,s~) i\ ) < V I) Lv h,,,_ \ ~ -/ \ I < v5\i\ V ( 4 I ~ ,. V v--------t ___,J V \ 1 2 J V I /"\ j . -i ,r SIC 26 PAPER AND PULP PRODUCTS Figure 24. PLANT LOCATIONS 1973 ,,. ., .IF' ~ .,.. ,\ ano ,u~ -. a, I.O

PAGE 83

1940 __. --1945 R e g i o n S I C 2 6 19 5 0 1955 1 960 1 965 1 970 EMPLOYMENT IN SIC 2 6 AN D IN TOTAL REGIONAL MANUFACTURING 1975 1 980 lf) en 0 en lf) N 0 N lf) 0 lf) 1 940 194 5 1950 1955 1960 1965 1970 19 75 SIC 26's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 26 PAPER AND PULP PRODUCTS Figure 2 5 1980 ....... 0

PAGE 84

-1 ----, I I I I 1 ~ SIC 27 PRINTING AND PUBLISHING Figure 26 PLANT LOCATIONS 1973 e -1 -...J I-'

PAGE 85

N 0 1940 --0 lJ') -----1945 R egion S I C 2 7 1950 1955 1960 1965 1970 EMPLOYMENT IN SIC 27 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 lJ') :::1' 0 :::1' lJ') ('I") 0 ('I") lJ') N 0 N 1/) 0 1/) 1940 1945 1950 1955 1960 1965 1970 1975 SIC 27's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 27 PRINTING AND PUBLISHING Figure 27. 1980 -...J N I __j

PAGE 86

~~ -\ I\ \ J lf( \ \ I/ < b Lv '"h -"' \ \ <( I / I 'f /L/ r---------. i / \ ,/\ J s 1c' ~ 2s .:: CHEMICAL-PRODUCTS Figure 28. \ ( l,/\i\ V ( v-----L..-1 / PLANT LOCATIONS 1973 I"-. V ....... l,.)

PAGE 87

r e l atively s:n al l (th e l ~irgest h t s C,9 e~1plo : e: cs ) \.hi. l e th e E \KA p J a nt 1 o ca t e d at L oid a nd (six mil es SL1 u t h o f i' l o rr i. s t ;n) en ;, lo y ee s c r 4500 p e r s on s L a rg e employm en t fi g ure s in Q 3.l ; R id g e a n d \lorristown (L o 1"1 : md) ske 1v the avera ge to 30~ p e r sous per p l a nt; th e r e for e the median 10 per plant is more descriptiv e of the true em ployment picture. Employment irt the chemical indus t Ty has be e n erra tic in the last two decades (Figure 29) and futur e expan s ion is di ffic ult to predict. Petrol e um re fin ing and paving a nd roofing products (SIC ~91 Employment and plant number are relatively insignificant relative to total industry within the regio n wit h all but one plant ( seve n employees) located in Knoxville (Figure 30). Little change in employ ment has been noted in recent years (Figure ~l). Rubber and plastic products industries (SIC 30 ) Ten of the 12 rubber and plastic products industries are found in Knoxville presenting over 60 percent of th e employment (Figure 32). The average number of employees is 68 people with 20 employees as the median. Between 1952 and 1973, the n ew industry grow th rate has averaged 16.5 percent but the total em p lo ym ent in 1952 was only 514 p e rsons (Fi g ure 33). Leather products industries (SIC 31) The leather product industries are fei1 in number (15), most occurring in Knoxville, Dandridge (near Jefferson Cit y ), and Morris town (Figure 34). The average expansion rate has be e n significant, 24.6 percent even though the base year emplo )~ ent (1 9 52) was only 546 74

PAGE 88

1940 1945 _. ---R e gi o n S I C 28 0 l/) 0 l/) (T) 0 (T) l/) N 0 N 0 l/) 1950 1955 1960 1965 1970 1975 1980 1940 1945 1950 1955 1960 196 5 1970 EMPLOYMENT IN SIC 28 AND IN TOTAL REGIONAL MANUFACTURING SIC 28 CHEMICAL PRODUCTS F i g ur e 29 SIC 28's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) 1975 1980 --.J V,

PAGE 89

. ... ir r p. ... . ( \ / \ \ \ l < V ~( ~f\ V ( \ I~ -~ "---. rs. ) l/ I < v----J t v ;lr-J fl---, A '-V i\ /l-/ \ V IC J PLANT LOCATIONS 1973 V'\ sic 29 :. PETROLEUM REFiNrNc AND -PA VING Amt ROOFING PRODUCTS Figure 30. ~ -...J "'

PAGE 90

1940 __. -R eg i on S I C 29 1945 1950 1955 1960 1965 1970 EMPLOYMENT IN SIC 29 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 0 l1) U1 l1) en 0 en l/} N 0 N l/} 0 l/} 1940 1945 1950 1955 1960 1965 1970 SIC 29 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 29 PETROLEUM REFINING AND PAVING AND ROOFING PRODUCTS Figur e 31. 1975 1 9 80 -..J -..J

PAGE 91

.--\ I\ \ .s ) t I'\ > / < ) 1-v' / \ I I ( v\l\ V ( I~ ,v---L.--1 h A lrJ / i\ IL-/ r"--------.... \ V \ 12 J PLANT LOCATIONS I/'\ 1973 e sic 30 RUBBER 0 0 Afm PLASTIC PRci'J:5"ucr 's Figure 32. "' V -...J 00

PAGE 92

1940 __. --R eg i o n S I C 3 0 1945 1950 1955 1960 1965 1970 EMPLOYMENT IN SIC 30 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 0 ll) lJ) ::1' 0 ::1' ll) en 0 en ll) N 0 N lJ) 0 1940 1945 1950 1955 1960 1965 1970 SIC 30 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 30 RUBBER AND PLASTIC PRODUCTS Figure 33. 1975 1980 ....... \0

PAGE 93

. .,. ,_ __ -r -\ v ~-. --( :\ r--. I ) < / ) --\ I ~ I ---/ >---,, __/' / / < /'\ j f----SI C 31 :: LEA T HER 0 PRODU CT S Fig ur e 34 -\ i <'.~ A [>f--( I'--. J ,-L---1 lrJ V -PLANT LOCATIONS 1973 e CX) 0

PAGE 94

persons. Employment per plant averages 100 per s ons as does the median, indicating a fairly s ym m e trical distribution. Exp a nsion in the future may be expected to continue but probably not at such a high rate (Figure 35) St one clay, and glass products industries (SIC 32) These represent strong local raw material, local market indus tries, and thus their location is determined by the chance occurrence of resources and the distribution of urban places. The spatial pattern supports this observation (Figure 36). The greatest concen tration, as expected, is in the Knoxville vicinity with 2/3 of the total number of plants (68). The average employment per plant is 32 persons. The median is 14 employees per plant. Expansion has kept pace with population growth in recent years and probably will con tinue to expand in response to total regional gr owth (Figure 37). Primary met a l industries (SIC 33) This industry type is distributed throu g hout the region simply becaus e of the large numb e r of foundries which historically dev e l ope d in East Tenn essee (Fi g ure 38) The l arges t emplo:,rment indu s tr y is Alcoa Aluminum in ~laryville-Alcoa with 5000 employees (70 percent of the total employment) Median employment, however, is only SO pe rs ons per plant. Emplo~nent h as dropp e d steadi l y since 19 63 but is expected to level out in th e fut ure (Fi g ure 39). Fa brica ted metal products i ndu stries (SIC 34) This classification includes welding and machine tool industries which u sua ll y d e velop loc a ll y and :ire l oca lmar k e t o rie nted. Thi s 81

PAGE 95

1940 ----Reg Io n S I C 3 1 1945 1950 1955 1960 1965 --1970 1975 1980 0 lf) lf) :::J' 0 :::J' lf) ('t') 0 ('t') lf) N 0 N lf) 0 lf) 1940 EMPLOYMENT IN SIC 31 AND IN TOTAL REGIONAL MANUFACTURING SIC 31 LEATHER PRODUCTS Figure 35. 1945 1950 1955 1960 1965 1970 SIC 31's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) 1975 1980 00 N

PAGE 96

--~ --\ \ r \ lf ( ) < v ) lv' h"' 1 I\ / \ I <~ l-Ai\ ( Ill le :, .. -~ti~ .... ,, ---1 / r"----.... /L/ V \ -~-~j PLANT LOCATIONS I /""\ 1973 e -SIC 32" :: STONE. CLAY, ~ AND GLASS PRODUC T S Fi g ur e 36 ,I / '\ I"J

PAGE 97

1940 0 l/) -1945 R e gi o n S IC )2 1950 1955 1960 1965 1970 EMPLOYMENT IN SIC 32 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 0 :::1' l/) (T') 0 (T') l/) (\J 0 (\J 0 1940 1945 1950 1955 1960 1965 1970 1975 SIC 32 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 32 STONE, CLAY, AND GLASS PRODUCTS Figure 37. 1980 CX> .i,

PAGE 98

r ~ l \ i r ~ -I~ ~ r -I\ ) .. l/ < D Lv h ----1-I i I \ -T '/ l I I I I ---I I I ---~ ( -, ,C-f 7~ ~-I\ \ 12 /_/ V I /'\ j '~ ... ,1 o ., ~c- -11r 1,-. l II" f p, i, ,aSIC 33 PRIMARY METAL INDUSTRIES Figure 38. PLANT LOCATIONS 1973 e 7 I I"'j 00 V,

PAGE 99

1940 --0 LI) _ _.... -----1945 Regl on S lC J J 1950 195 5 1960 1965 1970 EMPLOYMENT IN SIC 33 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 LI) (T) 0 (T) LI) N 0 N LI) 0 1940 1945 1950 1955 1960 196 5 1970 1975 SIC 33's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 33 PRIMARY METAL INDUSTRIES Figure 39. 1980 00

PAGE 100

ex p lo. i ns t he wid e d ist r i b u t ion of th e indu s try (f-igu r e iG). i'-los t of th e 73 plant s arc fo u nd i n Kuox\' i l l c ( o ppro .x i,~ t el : -;-5 p e rcen t) i n cli eat in g som e response t o ro.1v m ateria l avai l c Ll ii..l it:-' 1: hi(h is g reate 1 in th e K n oxv ill e a r ea T h e a v erage nu jl' ,b~ r of p e rson~ e ~p l oy2 d is 45 but the m e di a n is onl y 14 e m pl oyee s p e r pl a nt. T he in dus tr y ha s g ro1m significantly b e tween 19 5 2 and 19 73 p r ob a bl y in r espo n se to n a tional growth in the tr a nsport a tion and r ecrea ti on al ve h i cl e m arke t s ( F ig ure 41). Machin e ry (except electrical) industri es (SIC 35) The machinery industry is h eav ily conc en tr a ted in the K no x ville region and to a les s er extent in O a k Ridge and M orristown (Fi g ure 42). Many of these industries p e rform suppor.:ing fu n ctio n s such as the refurbishing of industrial to o l s (for e xa;;ip le, th e 1 wo dw o rking tool industries in Morristown). The lar g est emp loye r h as 6 3 9 employees but the av e rage for th e r eg ion i s 27 p e r so n s pe r pl an t and th e med ian is only 7 p e r s ons p e r plant. Total e mployment h as r em aine d f airl y constant in the last decade (Figure 43). Electrical machin e ry, equipment, a nd suppli es indu s tr ies ( SIC 36) The electr i cal m a chin e ry industr y h as g r o i-TI fro m 57 8 e mp lo y ees in 1 9 5 2 to 2 7 26 em pl oyees in 19 7 3, a v e ra gi n g a g r owth r ate of 60 9 p e rcent. The l a rg es t emplo ye r is M ag n av o x in M orr istown 1 ith 716 employees in 1970. Average employment is 120 per s ons; the m e dian is 40 persons per plant. Most of the industry i s conc e ntrat e d in Knox ville and dispersed lightly elsewhere (Figur e 4 4). In recent years, e;nplo y m e nt ha s e x pand e d s ignificantly (F ig ur e 4 5). 8 7

PAGE 101

I I -r ~ I ~ ( --1-+-----~ I\ r-\ < V ) L, .__ .... . .."V IU" ..... !r .,.., .,. I I I I I / I I -----, I I I I j_ --~I I I I I '0 1 I A I I I )I I I I 7 ( -1-'---._ --t I ~ t ~-~ I"-. ---I .. ,_ v-i / ,..... L--J v~ -I \ L/ r-------I \ V k: J PLANT LOCATIONS I ~ 1973 e ---~ SIC 34 FABRICATED METAL PRODUCTS Figure 40 co co

PAGE 102

1940 1945 -. R e gi o n 1950 1955 1960 1965 1970 EMPLOYMENT IN SIC 34 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 0 l/) l/) :::1' 0 :::1' l/) (\") 0 (\") l/) N 0 N l/) 0 1940 1945 1950 1955 1960 1965 1970 1975 SIC 34 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 34 FABRICATED METAL PRODUCTS Figure 41. 1980 CD \0

PAGE 103

,--. .--r . -:-. . \ I\ \ ~'\ > V < ) Lv / \ I .. <'. ~ .. v\i\ V .. .. ( iv~ .,,_. .,, 4 v-----' l------1 fl-,,,, lr-/ v I\ /',__/ -"'-----. \ V \ I< j PLANT LOCATIONS Lr\ 1973 e SIC 35 MACHINERY (EXCEPT ELECTRICAL) Figure 42 I"V 0

PAGE 104

"' 0 "' 1940 0 t.n --1945 Regi on src J5 1950 1955 1960 1965 1970 EMPLOYMENT IN SIC 35 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 t.n t.n t.n N 0 N t.n 0 1940 1945 1950 1955 1960 1965 1970 1975 SIC 35's SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 35 MACHINERY (EXCEPT ELECTRICAL) Figure 43. 1980 \.0 I-'

PAGE 105

r -r--r \ I I / \ . t---' r --r-t-~r-r--r I ~ l's v------V oo o ar \0 N

PAGE 106

1940 __.. 0 lf) -1945 Regi o n S I C 36 1 950 1 955 1960 196 5 1 970 EMPLOYMENT IN SIC 36 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 lf) ::::r lf) (T") 0 (T") lf) ('\J 0 ('\J lf) 0 1940 1945 1950 1955 1960 1965 1970 1975 SIC 36 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 36 ELECTRICAL MACHINERY, EQUIPMENT, AND SUPPLIES Figure 45. 1980

PAGE 107

T ra n s p o rt a t ion equipme nt jndu st r ~ e s (S l C ~_7) A l arge p or tion of the tra n s portation : industry i s fo u;1d in Kn oxvi ll e ; how e ver, S8v -ra l tr J. i.l c r tr a'.cd trai l er ancl tru ck ecmp e r manufactu r ers ar e foun d n ea r ~ e T azehe ll, Swee~~at e r an d LaFoll et t e (Fi gure 46). Tot a l employment is low (413 emp loye es) with th e average plant employing 34 p ersons 1he m e d i a n is 15 person s Because of th e low initial employment 1 n 1 952 (1 06 peo p le) th e g ro w th rat e see ra s phenomenal (85 4 percent) R egiona lly, how ever the industry is of minor importance when compared to total e ;np loyrnent except in the smaller communiti es (Fi g ure 47). Re ce nt problems with transportation fu e l availability may dam pen futur e growt h in this industry. In s trument and related products indu s tries (SIC 38) Th e se industries ar e concentrated mo st ly in Knoxville and O a k Ridge (Fi g ur e 48). Oak Ridge AEC pl a nts distort the picture consider ably with ove r 10,000 emp loy ees Excluding Oak Ridge t he averag~ employment is 93 p erso n s per plant while th e median emp lo y ment is only 10 persons. Little expansion in employment ha s oc c urred in recent years (Figure 49). Mi sce llaneous industries (SIC 39) Al r.10::. t a ll r : 1iscel lan eo us industries a re loc ated i n Knoxv illc, with 48 of th e 54 plants iT the region (Fi gure SO) Average e mploy ment is 44 persons, with the median 10 persons per plant. Among the more significant industries are National Cash Register in ~lorristO\-m and Southern Athletic Inc. in Knoxville with 300 and 780 employees respectively. Changes in employment h ave n ot been sig nificant in recent ye a rs a lthough growth is evid e nt (Fi g ur e 5 1 ). 9--1

PAGE 108

. ,--..-r--7 -, 1 \ / I I\ : -i I r I_ 4 < 4 V v5\ r-r---, --(7( ( --c I ~ r\ 4 I, ) 1 l/ ll < D v-------4 ,vu i..-----r lr/ s_ V \ LL,/ r-----.-.-...__ \ I/ \ <: ) PUNT LOCATIONS _L\ 1973 ['---/ ...._,.,. r _,, , .,. "'-,o r ~., , r ., ,.,. .,. or ~ m.,..., r _,. .,. ID"1II r a,..n..,. .,. ,, r SIC 37 TRANSPORTATION EQUIPMENT F i g u re 46 .

PAGE 109

1940 Regi o n ---. -__.. S I C ]7 1 945 19 50 19 55 1960 196 5 197 0 EMPLOYMENT IN SIC 37 AND IN TOTAL REGIONAL MANUFACTURING 197 5 198 0 D l/) D :::1' l/) (T) D (T) l/) ru 0 ru l/) D 1940 1945 1950 1955 1960 1965 1970 197 5 SIC 37 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 37 TRANSPORTATION EQUIPMENT Figure 47. 1980

PAGE 110

F----,\ / \ -I\ \ l <~ vl,/\i\ . t V ( 'r \ \ V < D v-----. ,Lvo l----1 h"' / \ /L/ \ / \ C ) PLANT LOCATIONS /\ 1973 . SIC 38 INSTRUMENTS 0 AND RELATED PRODUCTS Figure 48 "' V -..J

PAGE 111

N D 1940 R egion __.. D lf) _..M 1945 S I C )8 ,,,__ __ __ ..__ 1950 1955 1960 1965 1970 EMPLOYMENT IN SIC 38 AND IN TOTAL REGIONAL MANUFACTURING 1975 1980 lf) =>' D =>' lf) (Tl D (Tl l/) N D N lf) D 1940 1945 1950 1955 1960 1965 1970 1975 SIC 38 1 s SHARE OF REGIONAL MANUFACTURING EMPLOYMENT (PERCENT) SIC 38 INSTRUMENTS AND RELATED PRODUCTS Figure 49 1980 co

PAGE 112

\ ~ -\ : LJ ; ( I\ ._ 1t\. I ) -~ -~ . -~ -\ _/ 1// -I< ,.,....,_ j ..... -~ SIC 39 MISCELLANEOUS INDUSTRIES Figure SO. ( "~ J ,,.V c=7 1--PUNT LOCATIONS 1973 \!J \.0

PAGE 113

1940 --_... -. Re gi o n S I C 39 a l/') l/') ::I' a ::I' l/) (T") a (T") l/) ru a ru II) a 1945 1950 1955 1960 1965 1970 1975 1980 1940 1945 1950 1955 1960 1965 1970 EMPLOYMENT I N SIC 39 AND I N TOTAL REGIONAL MANUFACT UR ING S I C 39 MIS CE LLA NEO US INDUSTRIES Figure 51. SIC 39 1 S S H A R E OF REGIONAL MANUFACTURING EMP LOYM ENT (PERCE T) 1975 o 0

PAGE 114

Data Collection This research effort is unique in the use of historical aerial photography in th e analysis of industrial development processes. The use of aerial photography in the analysis of land use change in itself is not unusual; however, its use as an aid to develop a land use model is considered to be novel. Experience indicates that aerial photography provides a convenient and economical means to assess the processes effecting previous land use change and to assess the current potential of various land areas for fu ture development. Remote sensing has proven to be an efficien t data acquisi t ion technique capable of partially fulfilling many urban and regional mode lin g needs. To illustrate the use of the aerial photography in this analysis two stereo images have been included (Figures 52 and 53) illustrating the before and after scenes of an industrial location event. The industry site is located n e ar Harriman, Tennessee. The site was occupied 1n 1966 by the Beta-Tek Inc. which manufactures electrical machinery. Present employment is approximately 130 people. The first stereo image indicates the condition of the site and surroundin gs as of March 30, 1958 (TVA photo g raph y ) \,hile th e second st e reo ima g e indicates the conditions a s of M a rch 22, 1974 (NASA photography) Aerial photo g raphy is utili z ed in the follm.;in g m an ner: Th e OR N L dat a base contain s infor ma tion con ce rn i n g th e g en e ral location of e x i s tin g indu s try within the ETDD r egi on; th e
PAGE 115

--------1 958 STEREO I MAGE OF BETA TJ=:K INDUSTRIAL SITE (TVA 1 :30,000) Figure 52 lU L

PAGE 116

1974 STEREO IMAGE OF (NASA BETA-TEK I ~D USTRIAL SITE 1:24,00 0) Figur e 53

PAGE 117

each industry. Usin g a CALCOMP plotter these data were spatially plotted at a scale of 1:250,000 for initi a l analysis. Additional plots at a scale of 1:24,000 were obtained of those TVA-USCS topo graphic quadrangles having manufacturing plants. Using historical aerial photography, census materials and interviews with local citi zens, spatial conditions which existed prior to the time of industrial location event were reconstructed. Location events occurring before 1950 were not analyzed. However, industries which expanded or relo cated after 1950 were included. Most of the location events analyzed occurred after 1956, th e year Congress passed the National Defense Highway Act creating the Interstate Highway System. It is thou g ht th a t many of the locational decisions after 1956 (and perhaps before) were partially affected by knowledge of the location of Interstate hi g hw ays. It was initially intended to analyze each industrial location event for which d ata were available. Once data acquisition was begun, it became apparent that analysis of each event would take longer than anticipated. With over 1,000 industries in the ETDD region, it was impossible to compl e te the analysis in th e time allotted. After consultation with several ORNL statisticians, a minimum sample size of 150 industries was considered to be adequate to d ete rmine the statistical significance of the variables believed to be associated with indu s trial site selection processes. (This repr e sents a 15 percent samp le of the total number of industries presently within ETDD a nd a 33 percent sample of the industri es which have lo cated in ETDD since 1950.) Indu s trie s were selected in a manner t n maintain a homogenous mixture and to assure an adequate 104

PAGE 118

. re g ion a l sample Effo rt s were also made: t o i ,1ci i n t J.in Cl s-amp l i n g ba l :rnce b c t, ,iee n m et r oo lit a n a1td rur. ;il i. ndu;;tr:,acco rd ing t o pl an t di ::;tr ib u ti on As r ,:e ntio n ed a bove, a d.J i t i o r :.. 1 d;it a s o irces 1,e1: e ut i li z e d in th e a n a l ys i s along w i th hi s torical ae rial ph o to g raphy. It emize d below are some of the
PAGE 119

I NDEX S. Ut-i.liti cs 6. Compatibility with Existing L a nd Uses 7. 8. Neighborhood or Community Attrac tiv e ness and Amenities Industrial Park Space \'.ARI ABL E ::i. Kinds ::ind q u:: j t i. c s ; : 1.vajlJb J c n e: ir s:i ce a. Typ es of adja cent l a nd u s e b. T)'l )es of adj cent zo ning a. Condition and d e nsit y of adja cent l a nd uses b. Proximity to hospitals, school s parks, and churches a. Availability of suitable building or land in a n ex istin g industrial park b. Quality of park PR I:! AR Y D ATA S OU RCES L oc::i l intervi ews ; ,.1 til i ty and m un i cip 'i l service ma ps; topo graph ic ~a ps; ::i erla l photo g ra p hy Loc a l interviews; aerial photo g raph y; topo g raphic maps Zo ri ing m a ps; lo ca l interviews Field obs e rvati on; local interviews; aerial photography; censu s d a ta; topo gr::iphic map Topo graphi c maps; aerial photography; state hi g hway maps; field observation Local planning office; industri~l par k maps; aerial ph otog r aphy Field observation; loc al interviews; aerial photography In addition to these variables, observations were tabulated for : the position of the site relative to Knoxville to determine if orien tation to a rd Knoxvill e was important; th e existe n c e of a suit ab le building on the site; and wh8ther the site \ vas located w ithin a cluster of existing industries. Statjstical Proc ed ures It was expected that as analysi s of the determinants of indus trial lo ca tion progr e ssed, insi g ht would be acquired as to the proper 10 6

PAGE 120

methodo lo g ie s to b e e m pl oye d in r ati n g and ev alu a ting t h e importance of th e v n.r:i a ble s ex;1m i ned. /\ s t h ::) st ucl y pro g r essed and th e cornp l X Lti es o f t h e in dustr ia l s it e sel cc ti : : m p ro c es s \\ e r e reco g niz e d, it was d e c i ded th a t s im pl e s t atis ti c a l Jroce dur es ~oul d n o t b e suffici e nt to "unt a n g l e the inter a cting variabl e s to produce m ean in g fu l ans1-1ers 111erefore, a multivariat e proc e dur e (specifically fac tor analysis) was chos e n to d e scribe interconnection of th e v a ri ables th a t appear to be meaningfully related to the industri a l site selection process. Th e r e are several reasons for this d e cisi o r, Among m a ny multi variate procedures, factor analysi s is
PAGE 121

d e vic e the con s tru c ti o n o f :i.ncl:i. ces to b e u se d as n e1v v a ri ables in l ,, 1 3 l a t er a n a ys 1 s ~ : ost studi es utili zi n g facto-r :11 : a ly s i s emp lo r the t ec h niu.u\; :for exploratory purpo ses. It s u se in t his study was : 1 ) to re du c e the original set of variables, 2) to determine if the variables g roup as previously conceived, and 3) to d ef ine more succinct indexes which describe the site-se l ec ti on proc e ss "The beauty of factor analysis is that it takes thousands of measurements and qualitative m ea surements and res olv es t hem in to 14 distinct patterns of occurrence." For example, the data m3.trix in this study contains (150 industri es x 30 variables) 4500 pieces of information. Factor analysis permits one to identify patterns of relationships am ong these data which would be impossible for th e human mind alone. Fa c tor a n a ly s i s b eg in s with t h e con struction of a cor r el~ c ion matrix, usually throu g h th e use of Pearson product-mo m ent correl a tion coefficients. In setting up the correlation matrix, the user of factor analys i s has som e alternat ive s; h e may calcu l a te correl atio ns among vari a bles (or attr i butes), in which case the approach is called R-f ac tor analysis, or h e :-nay calcul a t e "a s c. ;oci a ti o n" b e tw een indi v icluals or object s which is kno~n as Q-f ac tor analysis We are primarily 13 Ibid. 14 R. J. Rummel, "Underst andi ng Factor Ana ly sis ," The Journal of Conflict Resolution, XI (December, 1967), p. 445. 1 08

PAGE 122

c o ncer n e d wi t l 1 R f ac t o r a n a l ys i s in t hi s s tud y as t he d e s i re i s t o l . b 1 5 g r o u p o r e .1I111nate var.La l e ::; T h e s eco nd s t ep i n f ac t or a ~ alysis i nvolves th e r e
PAGE 123

--------------T h e r es ults of th e factor ana .ly s _i s of m e asur eme nts ob tain ed on 30 v a riabl es for 1 5 7 ind 1 1stric s alo7 g with si~ple s tati st ic a l d esc ription s of th e vari :ilJ l cs ar e pr ese n ted in th t.' n ex t c hapter. 11 0

PAGE 124

Ci -t.\ PT ER IV EM PUUC A L ANALYS IS This chapter presents the r esults o f a n anal ysis o f c er ta i n variables believed to be r e lated to i ndustri a l site-selection eve nts which have occ urr ed in the ETDD re gi on since 1 95 0. Descriptive statistics of each vari a ble a re pr esente d first, followed by the results of the factor analysis. Descriptive Stati s tics This analysis made use o f a system of compu ter programs called the Statistical Pack age for th e So ci a l Sciences (SPSS) d es i gned b y Norman Nie and Dale Brent. l The se package programs permit a variety of analysis proc e dures a nd means f or pr es entatj on of results. D e scription of Sam p le This a nal ysis i s based upon a sample o f 1 57 industries, 15 percent of th e total numbe r of indust ries 1 n the r egio n. Fift y percent of these indu s tr ies l ocat e d i11 th e r e gion since 1 96~ and 25 perc e nt sin ce 1 96 8. A p pro xi matel y SO perc en t of t he i nd ust r ies 1 Norman Nie, Dale H. B e nt, and C H a dl ai Hull, SPSS: Statis tical Packa ge for th e Social Scienc es ( Ne w York: McGraw-Hill Book Company, 1 970). 111

PAGE 125

samp led had 50 e mployee or l ess w i tl1 SO p erc ent of the indu s t ries sL-L mp l e cl 1 "it h l ess th an 200 er.1ployees 0 1l y 2 per ce n t of th e sa ; 1 1p l e h ad over 1 000 emp l oyees Tabl e S pre se nt s a breakdo1vn o f th e sa mp l e by SIC nu rnbe -r r e l a t ive to the total number of industries in the ETDD region. The mixture is proportional to the tutal r e gional mix tur e of indu s tries \~hen one considers that some SIC cate g orie s have f e w n ew i~dustries. Description of Variable Measurem e nts For each of the 157 industrial sites election events analy z ed measurements were obtained for 30 variables. The tables which follow provide a statistical summary of these data. Appendix A contains a sample survey form used to record these measurements. Ordin a l scores from one to ten were used for variables which required measurements of absolute or rel a tive value: 1 = lowest value, and 10 = greatest va lu e ~l ea surements re q ui ring yes or n o an st v ers 1 ,ere sc ored: 1 = yes and 10 = no. Measurement of the slope and draina ge charact eris tics obtain ed for the industrial sample are presented in T a bles 9 and 10 r espec tively. Flat site s (l ess than 1 % s lope) we re scor e d 1 and steep sl ope sit es (30 % or grea t er) w e r e scored lC P refer~nce f or gentl e sloping t errain is obvious from th ~ data These sites, h o wever, were j not always located i n extensive flat areas. Many sm all local industries seem to prefer small flat sites which may be surrounded b y steeper sloping land. Drainage conditions were assessed on the b asis of (1-3) = no \ ater problem s ( 4-7) = som e problems, and (8-10) = 1 1:2

PAGE 126

STC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC SIC r;umbcr l.!1 Tot a l \u m b e r o f I n du stry Typ e Sample I nd 1 .1st r ies 1 n -2 0 Fo o d s 22 T ex t i l e s 23 Clothin g 9 24 Lumber 12 25 Furnitur e 25 26 Paper 4 27 Printing 6 28 Chemical 7 29 Petroleum 1 30 Rubber & Plastic 7 31 Leather 2 32 Stone, Clay 13 33 Prim. Metals 6 34 Fab. Metals 1S 35 Mach (Ex Elec ) 7 36 Elec. Mach. 8 37 Trans s 38 Instruments 1 39 Mi s c. .., I Totals 1S7 DISTRIBUTION OF SAMPLE BY SIC C.UEGORY TABLE 8 11 2 40 67 110 64 7 95 42 10 14 15 86 18 86 75 25 20 21 [" ""7 ..J I 975 ETD9 1 973 2 The total number of industries in 1974 i s 10 02 Thes e fi g ures are based upon a censu s co nduct e d i n 1 97 2-73. 113

PAGE 127

-R e l a ti ve Adju s ted Cu m ulative Ab so lute Fr eq u e ncy Frequency Adj Fr e q Va l ue Frequency (P rcent) (P e rcent) (Pe rcen t) 1.00 37 23.6 23.6 23 6 2.00 33 21. 0 21 0 44.6 3.00 35 22.3 22.3 66.9 4.00 26 16. 6 16.6 83 4 5.00 15 9.6 9 6 93.0 6.00 6 3.8 3.8 96. 8 7.00 1 0.6 0.6 97.5 8.00 2 1. 3 1. 3 98.7 9 00 2 1. 3 1. 3 100.0 0 0 0 0.0 Missing 100.0 Total 157 100.0 100.0 100 0 Statistics Mean 2 955 Std Dev 1. 715 Mode 1.000 Median 2.743 Ske 1,m ess l. 0 03 Kurtosis 1 1 86 Mi s sing Observation s 0 VARIABLE SLOPE OF LAND TABLE 9 VJ.r ia nce 2.9 --10

PAGE 128

Abso lu te V alue Frequency 1.00 113 2.00 24 3.00 1 2 4.00 2 5 00 3 6 00 1 7.00 1 8.00 1 0.0 0 Tot a l 157 Statistics Mean 1. 535 Mode 1.000 Std Dev 1.141 Mi ssi n g Ob se rvations 0 Relative Adj u sted Cumulative Freq u ency Frequency Adj Freq (Per cent) (Percent) (P erce nt) 72 0 72 0 72.0 1 5 3 15.3 87.3 7. 6 7.6 94 9 1. 3 1. 3 96 2 1. 9 1. 9 98 .l 0.6 0.6 98 7 0.6 0 .6 99. 4 0.6 0. 6 100.0 0.0 Missin 2 100.0 100.0 100.0 100.0 Median 0. 0 Kurtosis 10.739 Skewness 3 031 VARIABLE DRAI N AGE TA BLE 10 Varia nce 1 302 115

PAGE 129

freq ue nt pr ob l em s ~ l o st of th e industri;1l sit e s examLn c d appear'. ~ d to ha ve f e1 ,; or n o pr ob l e ms 1 :.i t h f l o o
PAGE 130

R e 1 c1 ti ve Absolut e Fr e qu '2 nc: y Valu e Frequency (Perc e nt) 1. 00 9 5 60 5 2 00 11 7 0 3.00 8 5 1 4.00 7 4 5 5 00 2 1. 3 6.00 7 4.S 7.00 7 4 5 8.00 5 3.2 9.00 5 3 2 10. 0(l 3 1. 9 0.0 7 4.5 Total 157 1 00.0 Statistics Mean 2.567 Mode 1 000 Median 0 0 Std Dev 2 .59 7 Skewn e ss 1 51G Missin g Obse rva tions 7 Adjusted Cu m ula ti \-e Fre c ,u ency Adj Freq (P e r ce nt) (Perc en t) 63 3 63. 3 7 3 70.7 5.3 76.0 4 .7 80.7 1. 3 92.0 4.7 86.7 4.7 91. 3 3 3 94.7 3.3 98.0 2 .0 100 .0 Missing 100 0 1 00 0 100 0 Kurtosis 0.9 2 7 Variance 6.7 44 VARIABLE CLEARI ~ G-COVER CO N DITIO N S TABLE 11 1 l 7

PAGE 131

Ab s olut e Valu e Frequency 1.00 2.00 3.00 4 00 5.00 6.00 7.00 8 00 9.00 10. 00 0 0 Total Statistics Mean 5 365 Std Dev 2 650 12 1 2 23 12 29 12 17 14 14 11 1 157 Mode 5.000 Missing Observations 1 R c lat i\ 'C ~ .dj usted Cu m u l a tiv e F r eq u e n cy Frequ ency Adj Freq (P e rce nt) (Per cent ) (P e rcent) 7.6 7.7 7.7 7 6 7. 7 15. 4 14.6 14.7 30.l 7.6 7.7 ::;7 S 18.5 18 6 56.4 7.6 7.7 64 .1 10.8 10.9 75.0 8.9 9.0 84.0 8.9 9.0 92 9 7 .0 7 J. 1 00 0 0.6 Missing 100.0 100.0 100.0 100.0 ~l e dian 5 155 Kurto sis 1 027 Variance 7.020 Ske1-1nes s O 107 VARIABLE DISTANCE TO CENTER OF TO\vN TABLE 12 11 S

PAGE 132

Absolute Valu e Fre q uency 1.00 49 2.00 83 3 00 18 5 00 2 6 00 1 10.00 1 0 0 3 Total 157 S tat i stics Mean 1.916 Mode 2.000 Std Dev 1 .035 Missing Observations 3 R e lativ e Ad just ed C um ulative Frequ e nc y Fr equency A dj Freq (Perc e nt) (Perc e nt) (Percent) 31. 2 52 9 11 5 1 ., .) 0 6 0 6 1. 9 100.0 Median 1.837 Skewness 3.789 31. 8 53.9 11. 7 1. 3 0.6 0.6 ~ is sing 1 00.0 31. 8 SS.7 97.4 93 7 99.4 100 0 100.0 --100 0 Kurtosis 24.461 Variance 1 071 VARIABLE DISTANCE TO NEAREST ~ IAJOR THROUGHFAR2 TA BLE 1 3 11 9

PAGE 133

------R e l ative A dj u s t e d C umu l a ti ve Ab s olut e F req u ency F r e qu e nc y Adj fr e q Val u e Fr e qu e ncy (Percent) (P e rc e nt) (P e r ce nt) 1.00 28 17 8 1 7 9 1 7 9 2.00 16 10 .2 10.3 28. 2 3.00 29 1 8.5 1 8.6 46.8 4 00 1 3 8 3 8 3 5 5 .1 5 00 6 3 8 3 8 59 0 6 .00 14 8 9 9 0 67 9 7.00 1 6 10 2 10.3 78.2 8.00 1 2 i".6 7 7 85 9 9.00 1 7 1 0 8 1 0.9 96.8 1 0 00 5 3 2 7 .) 100 0 0 .0 1 0 6 Missing 1 00.0 To ta l 1 57 1 0 0 .0 1 00.0 1 00 0 Sta t isti c s Mean 4.6 4 1 Std Dev 2.87 8 Mo de 3 .00 0 Med i an 3 835 K urt os i s 1 273 Sk e wne ss 0 299 V a :r i an c e 8 2 S3 M issi n g Obse r vatio n s l VARI A B L E D E N S I TY O F LA N D USE I N H!MEDIATE V I CINITY TA BLE 1 4 1 2 0

PAGE 134

Estim a t es of th e unit price o f l and i e re b ased u pon th e lo cation of ti1 e s i t e relative to the CG D and mJ._i o r tho r o u g hfa r e :-; Th e
PAGE 135

Rci a t i..ve .\a ju st e d Cu m ul a tiv e Abs o lute fr e qu e nc y Fr eq u ency Adj Freq Valu e Frequ e ncy (P e rcent) (P e rc e nt) (P i2 rcent) 1. 00 4 2.5 2.6 2.6 2 00 13 1 0. 8 8. 4 11. 0 3 00 17 10 8 11 0 22 1 4.00 17 10.8 11.0 33.l 5.00 16 1 0 2 10. 4 43.5 6.00 24 15 3 15.6 59.1 7.00 30 19.1 19 5 78.6 8 00 21 13 4 13.6 92.2 9.00 11 7 0 7 .1 99.4 10 00 1 0 6 0 6 100.0 0 0 3 1. 9 Missin g 1 00.0 Total 1 57 1 00 0 100.0 1 00.0 Sta t istics ~ 1 e a:n 5. 5 8 4 Std Dev 2. 210 Mod e 7 .0 00 M e d i a n 5 9 ] 7 K ur tos i s 0. 934 Skewnes s -0. 268 Missing O bservations 3 VARIABLE ~ATI ~ G OF PRICE CF LA N D TABLE 15 Vari an c e 4.885 1 2?.

PAGE 136

--Re l c1tive Adjusted Cu m ul at iv e Abso lu te FT e qu ency Freq u ency Adj Freq Valu e Fr eq u ency (P e rc e nt) (P e rc en t) (P e r cen t) 1.00 46 29.:S 2 9.3 29.3 2.00 15 9 6 9 6 38.9 3.00 6 3 8 3.8 42.7 4 00 22 1 4.0 1 4 0 56.7 5.00 7 4.5 4.5 61.1 6.00 4 2 5 2.5 63.7 7.00 5 3 .2 3. 2 66.9 8.00 15 9.6 9.6 76.4 9.00 13 8.3 8.3 84.7 1 0 .0 0 24 1 5 3 1 5 .3 100. 0 0.0 0 0.0 ~ lis si n g 100.0 Total 157 100.0 100 .0 100.0 Statist ics M e an 4. 796 M o d e 1. 000 ~!edi a n 4 0 23 Kurtosis 1.4 90 Std Dev 3.469 Sk e wne ss 0 .3 10 Missing Observations 0 VARIABLE PROPORTIO N OF URBAN AREA WITHIN TWO A N D A HALF f, IILES TABLE 16 Variance 12.035

PAGE 137

remainder of th e sample includ e d sites near small communities and frequently the 2-1/2 mil e circle \ iou ld include most of th e urbanized area (score= 8-10). Distance to m a jor highway was measured differ e ntly from distance to major thorou g hfare. This mea s ure was included to determine if proximity to federal and state highways functioning at the time of the location event affect site choices differently th a n proximity to major thoroughfares. Most of the sites were located close to major highways (Table 17); however, 10 sites were located in a reas 10 or more miles from a major highway. Distan ce to secondary road was intended to measure the importance of egress and in gress to the plant site. It was believed that most industri es would prefer sites which did not requir e plant tr a ffic to interact dir e ctly with a major highway. The distribution in Table 18 indicates re s ponse to this factor. Althou g h 60 percent of th e sites examined were adjacent to rail roads, the importance of rail accessibility in curr e nt location deci sions is questionable. M any of th e sites occupied in the last decade were inaccessible to rail service (score= 10) or they could be served only with great difficulty (scor e = 5-9) (T ab le 19). Distanc e to waterway was considered to h ave pot e nti a l site importance to only a few industry t ypes This study, how ever, is conc e rn e d wit h th e development of a ge ner a l site selection a l go rithm. Cons e quently di stance to waterway was included. Sixty perc e n t of the site s included in the s tud y were in a ccessib le to TVA wat e rways (Table 20) indicating th e impor t ance of th is factor t o b e relatively lo w for most industry t ypes Some of the sites near Knoxville, Lenoir City, 124

PAGE 138

R e lative Adjusted Cumulative Absolute Frequency Fr e quenc y Adj Freq Value Frequency (Percent) (P e rc e nt) (Percent) 1.00 36 22 9 22.9 22.9 2 00 101 64.3 6 4.3 87.3 3.00 10 6.4 6. 4 93.6 5.00 2 1. 3 1. 3 9 4 .9 10.00 8 5.1 5.1 100.0 0.0 0 0.0 Missin o C, 100.0 Total 157 100.0 100.0 100.0 Statistics Mean 2.280 ~lode 2. 000 Median 1 9 2 1 Kurtosis 11 .143 Std Dev 1.901 Skewnes s 3. 4 1 4 Missing Observations 0 V AR IA BLE DI S TA NCE T O MA J O R HI GH\'/AY TA B LE 17 V ar iance 3.613 1 2 5

PAGE 139

Relative Adjusted Cumulative Absolute Frequency Frequency Adj Freq Value Frequency (Percent) (Percent) (Percent) 1.00 146 93 0 93.0 93. 0 2.00 8 5 .1 5 .1 98.1 3 00 2 1. 3 1. 3 99.4 5.00 1 0 6 0.6 100 0 0 0 0 0 0 Missing 100.0 To t al 157 100 0 100 0 100.0 St a tistics M ea n 1.102 Mode 1.000 Medi a n 1.0 Kurtosi s 41. 611 Std Dev 0 441 Skewness 5. 8 81 Missing O bservations 0 VARIABL E DISTA N C E TO S E CO N DARY RO A D TABL E 18 Varianc e O 195 126

PAGE 140

Absolute Value Frequency 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0.0 Total Statistics Mean 3.682 Std Dev 3.563 80 17 6 5 2 5 3 12 1 26 0 157 Mode 1.000 Missing Observations 0 Relative Frequency (Percent) 51.0 10.8 3.8 3.2 1.3 3.2 1.9 7.6 0.6 16.6 0.0 100.0 Median 1. 3 Skewness 0.886 Adjusted Cumulative Frequency Adj Freq (Percent) (Percent) 51.0 51. 0 10.8 61.8 3.8 65.6 3.2 68.8 1. 3 70.1 3.2 73.2 1.9 75.2 7.6 82.8 0.6 83.4 16.6 100.0 Missing 100.0 100.0 100.0 Kurtosis -0.950 Variance 1 2 .693 VARIABLE DISTAi\JCE TO RAILWAY TABLE 19 127

PAGE 141

Absolute Value Frequency 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0.0 Tot a l St atis tics Mean 8.391 Std D e v 2.646 3 8 5 5 6 3 5 14 10 97 1 157 Mode 10.000 Mis si ng Observations 1 Relative Frequency (Percent) 1. 9 5.1 3.2 3.2 3.8 1. 9 3.2 8.9 6 4 61.8 0.6 100.0 ~l e di a n 9. 3 Skewness -1.53 4 Adjusted Cumulative Frequency Adj Freq (Percent) (Percent) 1.9 1.9 5.1 7.1 3.2 10.3 3.2 13.5 3.8 17.3 1.9 19.2 3.2 22.4 9.0 31.4 6.4 37.8 62.2 100.0 Missing 100.0 1 00.0 100.0 Kurtosis 0.9 67 V aria nce 7.001 VARIABLE DISTA N CE TO \1/ATERl~AY TABLE 2 0 128

PAGE 142

and L o ud o n 1 e r e clos e to publi ... cl oc ks an d, th e r e for e r ece i ve d l o 1~ e r acc ess ib ili ty v a lu es hut o nl y three s i te-; surveyed had do ck si t e a ccess at th ~ pl ant fo mm e rc i a l ai rport acccssitil; .! r ma y !J e consid e r e d a s both a re g ional influence and a local influen c e in industrial site sel e ction decisions particularly in the cas e o f small to medium-sized industries. The effect of small airports in th e decision process was not consid e red due to the widespread distribution of private and municipal airports throu g hout the r e gion. The frequ e nc y distribution probably r e flect s the re g ional spread of industry more than the importance of the factor in the site selection process (Table 21) Distance to Interstate highway interchange measures the impor tance that proximity to the Interstate system may have on the site selection process. Sites within two miles were scored al; sites f a rth e r aw a y w ere s c or e d on th e b as i s of 2-m i l e in c r e men t s f o r eac h number assigned. Approximately 90 percent of the industries surveyed have located or relocat e d since the Interstate highway system w a s created in 1956. Eighty-eight percent of the sites we r e within eight miles of an interchange (Table 22 ) Two measures of the overall quality of accessibility were included i n thi s a nal y sis to dete r mine if a collecti v e m e asure of a cc e s s ibility would be signifi c antly different from individual m easures of accessi bility and to determine if the overall quality of the accessibility has significantly changed since the location event. The latter may indicate whether industry location may influence future transportation improvem e nts Th e m eas ures w e r e o ve r a ll qu a lity of acces s ibility then 129

PAGE 143

Abso lut e Value Frequency 1.00 2.00 3 .0 0 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0.0 Total Statistics Mean 6 968 Std Dev 3. 7. 36 1 8 32 14 12 3 2 6 2 76 1 157 Mode 10.000 Missing Observations l R e l a tiv e \d j1 _.:; t ed Cu m ulative Fr eq,1 c ncy Frequency Adj Freq (P e rc en t) (Perc e nt) (P e rcent) 0 6 0.6 0.6 S. l S 1 5.8 20 4 20 S 26 3 8.9 9 .0 35.3 7.6 7.7 42.9 1.9 1. 9 44 9 1.3 1. 3 46 2 3 8 3 8 50.0 1.3 1. 3 51. 3 48.4 48 7 100.0 0.6 ~ l issing 100.0 10 0 .0 100.0 100.0 ~l e dian 8 500 Kurtosis -1.708 Varian c e 10. 4 70 Skewnes s -0 289 VARIABLE DISTANCE TO AIRPORT TABLE 21 1 3 0

PAGE 144

Ab s olute Value Frequency 1.00 2.00 3.00 4.00 6.00 8.00 10.00 0.0 Total Statistics Mean 2.813 Std Dev 2.352 58 21 48 10 6 2 10 2 157 Mode 1.000 Mi s sing Observation s 2 R e l a t.i.v e \ d j u s t e cl Ct :m ul a ti ve Fr e qu e ncy Fr e qu e nc y Adj Freq (Percent) (Perc e nt) (Percent) 36.9 37. 4 37 4 13 4 13.5 51.0 30.6 31.0 81. 9 6.4 6.5 88 4 3 8 3 9 92.3 1.3 1. 3 93 5 6.4 6.5 100 0 1. 3 Missing 100.0 100.0 100.0 100.0 Median 2 4~9 Kurtosis 3.285 Variance 5.530 Skewness 1.9 2 7 VARI A BLE DIST AN CE TO N EAR ES T L 'iTER STAT E TAB LE 22 1 3 1

PAGE 145

(T ab l e 23) a nd overall qu a lity of acc ess ibili.ty n ow (Table 24 ). So me i m nro, eme nt ca n b e not e d (t h e m e an di s t a n ce c h a n ge d fro m 4 55 t o i .10 ) Ov e ra ll accessibility 110: J 1 : as n o t i n:::: orpora ted .into t h e factcr ana l ysis T a bles 25, 26, and 27 refl ec t measu r emen ts o f th e ava ilab i lit y of utilities at th e site b e fo re th e ind u s tr i a l lo ca t io n event. In 90 p e rcent of the cases, both water a0.d gas 1 1e r e ava ilabl e a t t he site and in approximately 70 percent of th e c ases mun i cip a l sewage was a vai labl e It w as not e d that man y industri es loc a te beyond th e city limits to avoid r es trictions but e nj oy oth er city b e n e fit s In s u c h instanc es only mini mum sew age tre a t me nt is required an d th e indu st ries will install their own faciliti es This i s r e flected by th e lower percentage of sites sel e cted having preexisting sewage treatment facilities. T able 28, 29, and 30 refl ec t measurements of compa t ibility associated with the sites surv e y e d. Attributes which were measured include: Did the community have zoning at th e time of the event? Was th e site zoned for industry? and Was indL1stry already in th e immediat e area? Indust:ri a l zonir. g affec t s site sel e ct ion signifi cantly (60 percent of the sites examined were locat e d in ar ea s z oned for i1 1 dustry) T his i s i n ter es tin g considering T e nne ssee does not r e qui re co u nt ies to exe rcise zo nin g control a nd consequently ind u s try is not restrict e d by public policy in site se lection. Count y govern ment s how eve r, can exercise qua si zo nin g contr o l by limitin g coopera tion in road con s truction and other servi c es Th e existence of other industri es nearby th e potential si t es a l so ap p ea r s to b e a st ron g 13 2

PAGE 146

---R c la tivv Adju sted Cu m ulativ e Absolut e Fr eque ncy Frequenc y Adj Freq Valu e Frequency (P e rcent) (Per ce nt) (P e rc e nt ) 1. 00 1 O G 0. 6 0 6 2 00 1 4 8 9 9 1 9.7 3 00 24 15.3 1 5 6 25 . 3 4.00 24 15. 3 1 5. 6 40 9 5 00 60 3 8 2 39.0 79.9 6 .0 0 20 1 2 7 13.0 92.9 7 00 4 2.5 2.6 95 5 8.00 7 4.5 4 5 1 00.0 0 0 3 1.9 Missin g 100.0 -Total 157 100.0 100.0 100 0 S t atistics M ea n 4.552 Mode 5.000 Med :i an 4. 7 33 Kurtosi s 0 007 S t d Dev 1.4 69 Ske-,,ness O 08 4 Vari ance 2.157 ~lissin g Observa t ions 3 VAR IABLE OV E RALL QUALITY OF A CCES SI BJ ~ I T YTH E1 TABLE 23 13 3

PAGE 147

Absolute Value Frequency 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 0.0 Total Statistic s Mean 4 .104 Std Dev 1 505 5 17 26 53 30 13 6 3 1 3 157 Mode 4.000 Missing Observations 3 R e l a tiv e Adjusted Cumul ative Frequenc y Frequ e nc y Adj Freq (Perc e nt) (Per.::ent) (P ercen t) 3.2 3 2 3.2 10.8 11.0 14 3 16.6 16.6 31. 2 33.8 34 4 65.6 19.l 1 9.5 85.l 8.3 8. 4 93.5 3.8 3.9 97.4 1.9 1. 9 99.4 0.6 0. 6 100.0 1. 9 ~ li ssing 100.0 -100.0 100.0 100.0 Median 4.047 Kur to sis 0.481 Skewness 0.400 Va r ia nce 2.264 VARIABLE OV ERA LL QUALITY OF ACCESSIBILITY-NOW TABLE 24 1 34

PAGE 148

Value Y es 1.00 N o 10 00 0.0 Total Value Yes 1.00 No 10.0 0 0.0 Tot a l ------! ~ C .l Qt j_ \' C Absolut e Frc. quency Fr equencr (P e rcen' ) 1 4 1 89.8 16 10. 2 0 0.0 157 100 0 VARIABLE CIT Y WATER AVAILABILITY T AB LE 2 5 R e lative Ab s olute Frequenc y Frequ e nc y (P eT c e nt) 106 67.5 47 2 9.9 4 2.5 157 100.0 VARIABLE CITY SEWAGE A\'AILA!3ILITY T ABLE 26 1.35 Adj L S ted Fr e q c1 -2 ncy (P e rcent) 89 8 1 0 2 Missi n g 100. 0 Adjusted Frequency (P ercen t) 69 3 3 0. 7 Miss i ng 100 .0

PAGE 149

Va l u e Y es 1.0 0 No 10.0 0 0.0 Value Yes 1. 00 No 10 00 0.0 Total R e l a ti ve ; bso lu t e Frequ e nc y Frcquenc: y (Pncent) 137 87.3 2 0 12.7 0 0.0 -157 100.0 VARIABLE GAS AVAIL AB ILITY TABL E 27 R e lative Absolute Fr e que n cy Frequenc y (Percent) 93 62 4 57 36 3 2 1. 3 157 100.0 VA R IABL E DID COM MUN IT i' 1-L A.VC ZONIN G THEN? TABLE 28 1.3 6 i\djustc
PAGE 150

i 37 1 -:21a ti ve Ad j usted A bsolu te r-r e: q u e nc y Frequency V a lu e fr e quc:nc :,; ( P c:::-cen.,.) (P e r ce nt) Y es 1.00 87 ss. 4 65.9 i'J o 1 0 .00 45 28 .7 3 4 .1 0.0 2 5 1 5 9 ~ i s sin g Tot 2. l 157 1 0 0 .0 100.0 V A RIABL E WAS SIT E Z O N ED F O R I ND USTR Y? TABLE 29 R e l a tive Adjusted Absolut e Frequenc y Frequency Valu e Frequency ( Per c e nt ) (Percent) Y e s 1.0 0 llO 70. 1 7 1. 4 No 10 00 4 4 28 .0 2 8 .6 0.0 3 1. 9 Mis s ing Total 157 1 00 .0 100.0 V ARIA B LE WAS I N DU ST R Y AL R E .\ D Y l:'-; AREA? T AB LE 3 0

PAGE 151

influence in the site selection process. This applies in both urban fringe and near CBD cases. In lieu of recording each type of land use adjacent to the sites surveyed, an aggregate measure of land us e compatibility (the overall rating of continuous land use compatibilitz) was included. Measure ments varied from 1 = no problem with adjoining land use (e.g., situa tions with industry all around or open farm land or forest land all around) to 10 = significant compatibility problems (e.g., situations where the site was adjacent to a wealthy neighborhood, a hospital, or recreation area Very few sites were objectionably located and most of those cases w ere in small communities where complainants would be few (Table 31) Condition of nei g hborhood was included to measure the effect of value of housing on site selection proc esses This was also con sidered a surrogate measure of family income and worker occupation types nearby the sites. Difficulty was incurred in applying th e measure to sites in rural or open land. As can b e see n from Tab le 32, the distribution is platykurtic with son1e skewness toward low quality nei gh borhoods. The number of community services n ea rb y the site was assess e d to determine if proximity to gas stations, restaurants, parks, golf cour ses clubs, bank s and oth er commercial development affected industrial site se l ection T ab le 33 indicates th a t 70 percent of the sites h a d a t l east two services n earby an
PAGE 152

-Absolute V a lu e f-requency 1.00 5 8 2.00 30 3.00 26 4 .0 0 1 9 5.00 11 6 00 5 7.00 3 8.00 2 0.0 3 Tot. ee l 1 57 Statistics Mean 2 .558 Mode 1.000 Std Dev 1.69 2 Mis si n g Observat io n s 3 R e l a tiv e Adju s t ed Cumul ative Fr e qu ency Frequency A dj Fr e q (P e rc e nt) (P erce nt) (P erce nt) 36.9 19 1 16.6 12 1 7.0 3. 2 1. 9 1. 3 1. 9 1 00 0 M e dian 2.133 Ske\ v ness 1. 061 37 7 19.S 16. 9 1 2 3 7 1 3.2 1. 9 1. 3 ~lissing 10 0 0 37.7 57 l 74 0 86 4 93.5 96. 8 98.7 100.0 100.0 100 0 Kur to s i s 0.566 V ari a nce 2.863 VAR I ABLE OV ERALL RATING OF CONTIG U OUS LA N D U SE COMPA TI BILITY TABL E 31 139

PAGE 153

Relativ e r\ dj :.ste d Cumulativ e Absolute Fr e qu ency Fr e quen cy Adj Freq Value Frequency (P e rc e nt) (P e rcent) (Per cent ) 2.00 3 1.9 2 .0 2.0 3.00 22 14 .0 1 4 7 16.7 4.00 26 16.6 17.3 34.0 5.00 35 22.3 23.3 57.3 6.00 26 16.6 17.3 74.7 7 .00 17 10.8 11. 3 86.0 8.00 17 10.8 11. 3 97.3 9.00 4 2.5 2.7 100.0 0.0 7 4.5 ~!i s sing 100.0 Total 15 7 100.0 100 .0 100.0 Statistics Me a n S.320 Mode 5.000 Median 5.186 Kurtosis 0. 771 Std Dev 1.712 Skewn ess 0 .228 Missing Observ a tions 7 VARIABLE CONDITIO N OF NE I GHBORHOOD TABLE 32 Var i ance 2.930 1 40

PAGE 154

l ~e l cujve Adj u s t e d Cu m ul a tive Ab s olute Frequ ency Fr e quenq Ad' .l Fr e q Valu e Frequency (Perc e nt) (P e rcent) (P e rc e nt) 1.00 34 21. 7 22 7 22.7 2 .00 20 1 2.7 13 3 36.0 3.00 24 15. 3 16.0 52.0 4.00 7 4.5 4.7 56.7 5 .00 14 S.9 9.3 66.0 6.00 6 3 8 4 0 70 .0 7.00 11 7.0 7.3 77 .3 8.00 13 8.3 8 7 86.0 9.00 19 12.1 12.7 98.7 10 00 2 l. 3 l. =~ 10 0 0 0.0 7 4 5 Missing 100.0 Tot al 157 100.0 100.0 100.0 Statistics M ea n 4 347 Std Dev 2.936 ~lode 1. 000 Median 3.375 Kurtosis -1.2 88 Skewness 0.414 Missing Ob ser v a tion s 7 VARIABLE DE N SITY OF LAND US E T AB LE 33 Variance 8 617 1 4 1

PAGE 155

studi es it is beli e ved th at som e co n tro l shou l d be int r o Ju ced i f indu s tr y pre se ntly ex i s t s i n th l ~ v icinit y ot th e site s ur v ey e d I t 1 as po s i. t c cl th a t cl es i gn ~ it ccl p n .1 s t1 1:i l p a r k sp ac e h o uld b e a stron g v a r iab l e i1 1 indu s trial s i re s e l ect i o n Onl y 37 pe rc en t of th e industry l ocated in industri.al park s (T able 3 4 ) H owever many recent industries have locat e d i n indu s trial parks indic a tin g a trend tow a rd such lo cations. Additional d a t a concerning industri a l p a rk qu a lit y was collected but not included in th e ana l ysis b e cause of th e hi g h p erce nta ge of missin g observations (63.1 percent) Tabl e 35, how e v er in dic ated th e di str ibution of th e data collect e d. Finall y three a ddition a l variabl es w e r e includ e d in the analy sis. These were: (1) Proximity of sit e to Knoxvi ll e; (2) Amount of oth er industry loc ated nearby; a nd (3) W as a suitab l e building a lr e~ d y th ere? (Tables ~ 6, 37, and 38) Proximity to Knoxville was included to de t e r mine if sites ori ented to w ard Kno xvi lle had an impact on the p rocess. Some eff ec t is indicated (low values= ori entation t ow a rd Knoxvi l l e) but a l ar ge pe rc e nt of th e s a mpl e c a me from th e Kno xv ille \ici ni t y In any sub sequen t stud y the Kno x vi ll e sa m pl e wil l be eliminated fro m th e analysi s The ex iste nce of a suitable bu i ldin g a t the site se em s t o have influenced th e d ec ision process to s o me extent ( 30 p e rcent). Also the amou nt of industry n ea rby seems to h ave influ e nc e d l oca t ion decisions. 1 42

PAGE 156

R e l ative Adjusted Cumul a tive Absolut e Freq u e nc y F r equency Adj Freq V a lu e Frequency (Perc e nt) (Percent) (P e rcent) 1.00 18 11.5 1 2 2 12.2 2.00 8 5 1 5 4 17.6 3 00 8 5.1 5 .4 23.0 <-l 00 7 4.5 4 .7 2 7.7 5.00 11 7.0 7.4 35 1 6.00 11 7.0 7 4 42 6 7.00 15 9.6 10.1 52.7 8 00 21 13.4 14. 2 66.9 9.00 22 14.0 14 9 81.8 1 0 0 0 27 17 2 1 8.2 1 00 0 0.0 9 5 7 Missin g 1 00.0 Total 157 100 0 100 0 100.0 Statistics Mean 6.405 Std D ev 3.079 Mode 1 0.000 M e di a n 7.2 3 3 Kurtosis 1. 066 Skewness 0.5 2 5 Mi ss in g Obser0ation s 9 VARIABLE NEARBY COMMU N ITY SERVICES TABLE 34 Variance 9 481 1 43

PAGE 157

Absolute Value Fr e quency 1.00 10.00 0 0 Total Stati st ics Me a n 6 654 Std Dev 4.364 58 98 1 157 Mode 10.00 0 Mi ss ing Observations 1 Relativ e Frequenc y (P e rcent) 36.9 62 4 0 .6 100.0 ~le d ian O 0 S kewn ess -0.53 1 Ad ju s t e d Cumulative Frequency Adj Freq (Perce nt) (Percent) 37.2 37.2 62.8 100.0 Missing 100.0 1 0 0.0 100 0 Kurtosis 1. 7 1 9 Variance 19 041 VARIABLE \VAS THE SIT E I;-; A N 1 1';0USTRIAL PARK? T AB L E 35

PAGE 158

Absolute Value Frequency 1.00 2.00 3.00 4.00 5.00 6.00 7.00 0.0 Total Statistics Mean 3.534 Std Dev 1.667 8 9 11 14 9 4 3 99 157 Mode 4.000 Missing Observations 99 Relative \dj usted Cumulative Frequency Frequency Adj Freq (Percent) (Percent) (Percent) 5.1 5.7 7.0 8.9 C: ;> I 2.5 l. 9 63.1 100.0 M e dian 3.5 71 Skewness O .191 13.8 15.5 19.0 24.1 1 :i. 5 6.9 5.2 ~!issing 1 0 0.0 13.8 29.3 48.3 72.4 87.9 94.8 100.0 100.0 100.0 K urto s i s -0.681 Variance 2. 779 VARIA BLE WHAT WAS TH[ QU A LITY OF T HE PARK ? TABLE 36 145

PAGE 159

-------Relativ e Adj u s t e d Cwnulative Abso lut e Fr e -: ,t1ency Frequency i\
PAGE 160

R cl:c tiv e i\dju s t e d Cu mulative Absolute 1-'r e q --'e ncy Fr equen cy A dj freq V a lu e F req u e nc y (P e rcent ) (P e rc e nt) (P erce nt) --------1. 00 30 1 9 l 1 9. 4 19. 4 2.00 9 5. 7 5 8 2 5 2 3.00 1 0 6 4 6 5 3 1. 6 4.00 6 3.8 3.9 35 5 5 00 1 4 8.9 9.0 44 5 6.00 8 5.1 5 .2 49.7 7 00 1 2 7.6 7 .7 57 4 8.00 21 13.4 13. 5 71. 0 9 00 13 8.3 8.4 79. 4 10.00 32 20 4 20.6 100. 0 0 0 2 1. 3 Missin g 1 00 0 Total 157 iOO O 1 0 0. 0 100 .0 Statistics Mean 5 865 Mod e 1 0 000 ~ l e dian 6 54 2 Kurtosis -1 4 39 Std D e v 3 36 5 Ske1 m ess -0 234 Variance 11. 326 Mi s sing Obser vn tions 2 VARIABLE AMOUNT OF OTHER I N DUSTRY LO CATE D NEAR BY T A BLE 38 14 7

PAGE 161

Results of the Factor Analysis The SPSS factor analysis procedures were discuss e d in the pre ceding chapter and will not be repeated her e T a bl e 39 presents the correlation coefficient matrix for 27 variables utilized in the factor analysis. Correlation values greater than ~-50 are considered to be significant. In general, the values associated with each pair of variables agrees with theoretical expectations. Eigenvalues associated with the initial factor matrix are found in Table 40. The eigen-values represent the proportion of the stan dardized total variance (27) accounted for by each factor. The SPSS factor analysis routine automatically stops ex tractin g factors when the eigen-value for a factor fall s b e low on e This assures that only factors accounting for at least th e average total variance (1/27) will be treated as s i g nificant. In this case seven factors were extracted accounting for 72 percent of th e original variance. Some explanation of the "loadin gs or sc or es found in the factor matri x is necessary b efore discussing th e r es ults of thi s analysis. Scores found in th e factor matrix are evaluated similar to corr e lation or regression coefficients. Valu es may r a nge from +l to -1. The greater th e ab s olute value of the score, th e greater the relati6nship b e tw ee n th e factor and tl1 e variable The lev e l at which factor l oadings m a y be considered as significant is op e n to ques tion. Most statisticians suggest that l o.i din gs grea ter th an .50 s hould b e consid e red significant. Ideally, only a sma ll numb er of variabl es s h ould lo ad significc1ntly on more th an one column in th e matrix. Variabl es l oadi n g on mor e than one fac.tor c omp li cate the int er pretat i on 148

PAGE 162

Columns R.o w 1 2 3 4 5 6 7 9 J. ti Slope of land 1 1 0 0 0 -. 247 .51 9 .13 2 -.007 -.1 26 0 88 .13 0 055 D raina ge 2 24 7 1.000 -.0 27 .070 .0 88 .093 .065 013 013 Clearing-cover conditions 3 5 19 -.0 27 1.000 .1 46 .0 48 .1 43 -.143 -. 141 0 3 8 Dis tance to center of town 4 132 .070 .1 46 1. 000 .206 -.814 -. 5 29 -.1 80 223 Distance to nearest major throughf a re 5 -. 007 .088 048 206 1.000 -.223 -.ll O 218 345 D e nsity of l an d use in i m m e dia te v icinity 6 -. 126 -.093 -.143 -. 814 -. 223 1.000 .500 244 -. 214 Ra tin g of price of l and 7 -. 088 -.065 -.143 -.529 -.11 0 500 1. 00 0 0 1 17 .315 Per c en t of urban area within 2 -1/ 2 mi l es 8 -. 130 .013 -.1 41 -.180 .218 244 04 7 1 000 169 Dis t :mc e t o major hi g hway 9 0 55 .013 .0 38 .223 .3 45 214 -. 31 5 .1 6 9 1. 000 D ist a nce to secondary ro ad 10 074 -. 033 -.0 75 .0 83 .208 -.15 8 -. 2 48 -. 078 -.15 7 Dis t ance to railway 11 .150 .11 4 .15 8 .082 .074 028 -. 312 -. 023 055 Distance to airport 12 -. 054 138 -.032 399 016 -. 4 13 -. 338 3S O .15 4 Di st a nce to wa terway 13 0 18 .058 .135 .315 -.Oll -. 232 -. 25 6 222 .1 21 D i sta nce to nearest Int e rst ate 14 04 9 -.066 -.031 390 .188 -.373 -.4 26 275 702 O verall quality of a cc essibi l i t y then 15 03 6 .076 087 427 .0 8 1 -.363 -. 5 03 2 77 4 72 Ci t y w:i t er availability 16 3 50 .053 .203 401 070 -. 3 47 -. 37 0 -. 0 8 5 122 Gas avai l abi lity 1 7 25 0 028 .1 so .388 203 -.29 9 .521 148 624 C ity s ewage avai l abilit y 1 8 .1 8 4 088 .140 493 .143 -. 495 6 4 7 -. 079 1 4 1 Did commu nity hav e zonin g tl w,1 ? 19 -. 00 1 .0 84 .031 225 -.01 4 .1 62 5 03 3 6 0 205 \ \'as it zoned for i ndu s tr y ? 20 113 .034 .12 8 .351 060 .248 -. 6.3 8 149 2 t ,S \Va s industry already in area? 21 147 .078 .1 67 .350 4 7 223 -. 4 9 1 .1] 4 249 \'l as a buildin g already ther e? 22 120 -.073 .1 38 .372 -. 061 449 107 240 105 Rati n g of l and use com pa t ibil ity 2 3 -. 07 0 -.168 -.013 .2 46 24 7 298 -.0 38 299 -. 040 Nea rby community servic es 2 4 -. 225 -.033 261 .650 -. 093 6 37 6 08 0 6 3 .11 7 Was th e site in an i ndustrial park '? 2 5 -. 038 072 .0 15 -.12 9 -.037 .201 -. 335 336 -. OSJ. Pro xi mi t y to Knoxvi lle 26 12 1 .026 .013 409 100 .4 0 5 -. 300 1 S9 203 /\mo unt of oth e r ind u s t r y lo c:i te d n ea rby 27 137 -.0 46 .227 486 -. 093 434 6 96 13() 2 05 CORRE f.l\ TlON MJ\T lHX TA BLE 39 ,-..., ..i:,. ID

PAGE 163

Columns Ro w 1 0 11 12 13 14 15 1 6 17 J .8 Slope of land 1 0 74 .150 -.054 .01 8 .049 .036 .350 .25 0 .1 84 Drainage 2 -.033 -.114 .138 .058 -.066 .076 05.) .0 23 0 88 Clearing-cover conditions 3 -. 07 5 .158 -.032 .135 -.031 .087 .2 03 1 5 '.) .1 40 D istance to center of town 4 08 3 .082 .399 .315 .390 427 .401 3S& 493 Distance to nearest m~jor throughfare 5 -. 20 8 .074 .016 -.011 .188 .081 .070 2 0 3 .1 43 Den sity of land use in imm edia te vicinity 6 -.158 .028 -.413 -.232 -.373 -.363 -.3 47 -. 299 -. 495 R ating of price of land 7 -. 248 -.312 -.338 -.256 -.426 -.503 -.370 .521 64 'l Pe rcent of urban area within 2-1/2 miles 8 -. 07 8 -.023 .350 .222 275 277 -.0 85 .1 48 -.0 7 9 Dis t ance to major high'.vay 9 -.157 .055 .154 .121 .702 .472 .1 22 62 ,1 .Ul Dis tance to secondary road 10 1.000 .0 45 .124 .081 .075 .119 .0 22 O:IG 23 9 D istance to railway 11 045 1.000 -.144 .188 -.010 .388 .0 78 .ll E .30 4 Distance to airport 12 .1 2 4 .144 1.000 .646 .465 .701 .1 24 2:cJS .292 Dista nl.: e to wa terway 13 08 1 .188 .646 1.000 .278 .736 -. 052 231 .2 07 Dis tanc e to nearest int er st at e 14 07 5 -.010 .465 .278 1.000 .599 .17 5 710 264 Ov erall quality of accessibility then 15 .119 .388 .70 1 .736 .5 9 9 1 .00 0 .1 32 .S l3 407 Ci ty water availability 16 022 .078 .124 -.052 .175 .132 1.000 .4 79 4 72 Gas availability 17 .046 .116 .295 .231 .710 .513 .4 79 1. 000 .. ,90 City sewage availability 18 289 .304 .29 2 .207 .26 4 4 07 .472 .3 90 1. 000 Did community have zoni n g tl-:cn? 19 246 .003 .434 .220 .355 .376 243 .41_ )1\ S25 \ fas it zoned for industry? 2 0 26 4 .106 .346 .225 .446 .465 4 53 500 60 5 \ fas industry a lready in ar ea? 21 .0 3 9 .324 .178 .243 .311 .392 .2 7 2 41-1 4 44 Was a building already there? 22 .020 -.061 .183 .196 .108 .122 .177 11 2 080 R atin g of land use compa t ibility 23 -. 0 4 3 .354 -.066 .0 89 -.0 36 .165 -.1 80 -. 074 -. 090 N ea rby communit y services 24 -.109 -.032 -.351 -.302 -.210 -.288 .4 4 1 -. 27i -. 494 Wasthe site in an industri a l park? 25 .179 .107 .143 .009 .047 .141 .151 .1 56 365 P ro x imity to Knoxville 26 .1 8 0 -.040 .752 .544 .472 612 .180 329 251 Amoi .. mt of oth er industry loc at ed nearby 27 -.1 95 -.353 -.363 -.358 -.353 -.527 -.352 -. 44 9 -. 65 1 TABLE 39 Continued

PAGE 164

Slope of l and Dn.inage Clearing -co ver con d it i o ns Distance t o c e nt er of t own Dis t ance t o neares t m<'l.j o r through fare D ensity o f land u s e in immediate vi c i n i t y Ra tin g of price o f land P ercent of urban area within 2 1/2 mi l es D i s t;incc to mJ.jor hi gh\vay Dis t ance t o secondary road D ist ance to ra il way Di stance to airp o r t Distance t o waterway Dis t ance to neares t i nt ersta te Overa ll quality o f acce s sib ili ty th e n C ity wa ter avai l abi l ity Gas av a ilability Ci ty se w age ava i lab i lit y Did c o mm unity have z o nin g then ? fas it zoned for indus try? \fas industry alrea d y in a r e :1? \ as a bu i ld i n p; a l rea d y there? R Lt in g of l an
PAGE 165

Factor Eigen Value 1 7 79292 2 3 02117 3 2.02117 4 1.87910 s 1 75885 6 1.39755 7 1.17964 8 0.94247 9 0.84998 10 0 75369 11 0.67677 12 0.56666 13 0 53530 14 0 42438 1 5 0 37718 16 0 35815 17 0.339 5 0 18 0.32075 19 0.28415 20 0. 2359 1 21 0.18443 22 0.17080 23 0 152 23 24 0 12058 25 0.10901 26 0 07168 27 0 013 7.2 Pct of Var. 28.9 11. 2 EIGEN-VALUES TABLE 40 9.2 7.0 6.5 5.2 4 4 3 5 3.1 2 8 2 5 2.1 2.0 1.6 1.4 1. 3 1.3 1. 2 1 1 0.9 0.7 0 6 0.6 0 .4 0.4 0.3 0.0 152 Cum. Pct. 28.9 40.1 49.3 56 2 62 7 67.9 72.3 75 8 78 9 81. 7 84.2 86.3 88 3 89.9 91. 3 9 2 6 93.8 95.0 96.1 97.0 97.6 9 8 .3 98.8 99.3 99 7 100 0 100 0

PAGE 166

o f fa c tors. Colu m ns 1\ith a h igh n ur.ib cr of si;n ifica:1t l oa ding s shou ld ha 1 1e ::i. t. l ea s t as many n ea r z ero l o :i din gs a s t;1e nu mbe r of fa c tors cl e rLvcd 4 Attention is dir ec ted to a c ompar i s o n b e t 1,e~ n t he u n rot a ted factor ma tr ix (T able 41) and th e rotated factor ma tri x (Table 42). Note th a t rotation of th e axis simp lif ies interp r e tation of th e f actors In interpretin g th e rot a t ed ma tr ix, the ori g in a l 27 v ar ables have b ee n collapsed to s e v en fa ct ors acco unt i n g for 72 percent of the o rigi nal variance. Approxi mate ly 20 var iabl es c a n be identi fied as signific a nt. Th e followin g list s u mma rizes the va r ia bles which l oa d significantly on each fa c to r. FACTOR I 5 ZONED INDUSTRY ZONING PRESENT AMOUNT O F OTHER INDUSTRY LOCATED NEARBY SEWAGE AVAILABLE FRTCE OF LA N D SITE IN I N DUSTRI A L PAR K (I N DUSTRY IN AREA) FACTOR II DISTA NCE TO AIRPORT DISTA N CE TO WATERWAY OVERALL Q UA LI TY O F ACCESSIBILITY-THE N PROXIMITY TO K N OXV I LL E F A CTOR I Il DENSITY OF LAND U SE DI STAl~CE TO CENTER OF TOWN BUILDING PRESENT (SITE IN INDUSTRIAL PARK) (PERCE N T O F URB Al\J AREA WITHIN T WO Al'~D A PL' \ LF mLES) 4 Nic and oth ers p. 223 5 var i a ble s a re l isted from hi ghes t hctor sc o r e to 1 01,cs t. Parenth eses ind ic at e variabl e s with q u e sti onab l e s i g nific a nt l oa d ing s 1 53

PAGE 167

Slope of la-:1d Drainage Clea ring-cover conditions Distance to center of to w n Dist a nce to nearest rnaj or throughfare De nsity of land use in i m m e diate vicinity Ra tin g of price of land Pe rcent of urban areo. within 2-1/2 mil es D is t a nc e to majo r hi ghwo. y Dis t ance to secondary r oa d Dis t ance to railwa y D istance to airport Dist a nce to waterway D ist a nce to nearest Interstate Ov e r a ll quality of a cce ssib ility-then Ci t y water ava ilabilit y G;is ;:iva i l a bil i ty Ci ty s e wage av;_iilabilit Di d com m unity have zo nin g t he n? \fas j t zo n ed for inJust ry? \ fas industry already 1n a r e a.? W _,-;.s :1 buildin g alrc:idy t her e ? l ~ J. t ing of l and USC comr ::i til .Ji l i t y Nea rby community services \ fas the site in an ind ust r ial park? P oximity to Knoxvill e /\mo unt uf oth er industry locate d nearby 1 2 3 -.20 67 2232 4081 -.07 36 .0004 -.0584 -.19 70 .14 83 .3035 -.67 82 4880 .0749 -.17 30 .20 49 .0288 .62 29 -.6059 -. 0571 779 6 .0 663 -.2166 -.151 0 -.5074 -.3812 -.4 427 .0672 -.2789 -.210 2 -.0816 .1052 -.22 6 0 -.2168 .2428 -.63 74 .0250 -.5 8 57 -.50 26 -.0003 -.4337 -.6 462 .0167 -.3 867 76 H -.1117 -.4355 -. 4629 .1 492 .3 64 7 -. 6793 .0012 .039 3 -.70 34 -.030 2 .3539 -.63 03 -.4140 .0 404 -. 74 S8 -.3 44 1 .2116 -.61 ?. G -. 244 0 .2 24 5 -.17 6 1 .5503 0764 03 72 -.5 217 -,O L!.0 5 .63 36 3 111 -. 223 2 -. 29 80 -.69 28 24 29 -.5 761 .1559 -.40 78 .78 73 2211 -. 2532 U NR OTA TED FACTOR MATRIX TA BLE 41 F a ctors 4 -.2240 .1 243 -. 0923 1088 -.3079 -.1 64 0 -.0212 -.0 2 12 .694 1 2768 -.153 2 .378 7 2364 -. 4301 -.0 42 '1 -. 0729 t !660 .1357 .1291 O.iS7 -. 0 833 0748 .0538 -. 222 8 ltl-32 176 6 -.0 800 5 6 7 .3973 .6317 C849 -.2121 -.1 658 -. 2 3% .3060 .2 306 -.1 78 1 -.0269 -.1 692 O~ -ilS -. 1480 -. 2576 3 660 0894 .1 659 -. 0844 .0331 .1 725 136 2 -.0148 .1 727 0253 -.1252 -.0 8 14 Olll8 -.0355 077 2 2752 .5634 -. 3284 0 710 0210 .1 872 l 563 .3990 0 001 J 0 52 -.1635 0 667 1 7 3 9 .343 5 10 96 00 75 1296 1 805 -. 0520 1118 1 364 C28 :i -.06 60 -.1 02 1 03 :,3 -. 3 1 7 1 .1 80 1 J Ll l 3 -.,_ n o 1 2 l S I 77 .0712 -.1 2 o 5 l :,in 0769 .O n4 2 (J:--,1 ) 3853 1 0 1 1 l .1 .
PAGE 168

1 2 3 S l op e o f land .06 80 -. 0152 -. 0 85 7 D r ::ii n a g e .10 6 6 .0 8 5 4 -. 0042 Clea r i n g -co v er c ondition s .10 0 4 030 2 0999 D is t an c e to c e nt er of to w n .37 8 9 279 1 .6 82 0 D i s t ance to n ea r e st ma j o r throu g h fa re 02 63 0 4 87 1359 De nsi t y of l a nd u se i n i mm e dia t e vicinit y -.3 2 1 8 -. 2 540 7 8 8 4 Ra t i n g o f pr ice of l a.nd -. 7 0 79 .1708 26 98 P erc e nt of u r b a n a r ea wi th i n 2 1 /2 m il e s 1 0 87 3 47 5 :5 0 06 Di s t 3nce t o maj or h ighw.2 y D i s t a nce to se c o nd a r y ro a d Dis ta n ce to r a ilw a y D is t a n ce t o ai r p ort D istance t o 1 va t e r w a y Di s t ance to n eares t Int erst a t e Ove r a ll qu a lit y of a cc ess i b i li t y-th e n Ci t y wate r ava il a bili ty G as avai lability Ci t y sewag e a v ai l a bili ty Did co m m un i t y h ave z o ni n g t he n? \ fas it z on e d for indu s t ry ? \ fas i n d u s tr y a lr ea d y in a re a ? \ fas a b uildi ng a lr ea d y t he re ? 1<. a ti n 11 -~ of l a n cl u se c omp clt ibili t y i ca rby co:nm un i t y se r v i c es i\'as t h e s it e in a n ind us tr i::1. l p ark? Pro xomi t y to K no xvi ll e Amo un t o f oth e r indu s t ry l oc a t e d n ea rb y 051 5 .0 92 0 -.0500 31 2 7 0697 .1 0 5 2 1 6 3 0 -.0 04 1 00 9 1 247 4 .9025 -.0 83 5 .0 8 1 8 77 15 -. 0 87 9 23 88 3335 -. 0 99 1 2737 735 4 0 776 461 8 0 4 74 2 30 0 .39 7 0 1632 077 2 7 3 1 7 10 8 1 2 7 0 7 .7 43 6 2299 .1 76 5 8 3 3 5 .138 2 05 11 5 867 .1 488 .0 33 3 10 4 2 1 39 9 5 80 1 .0 29 S 0 4 7 4 3902 -. 489 5 -. 257 4 48 0 6 .654 0 .0 2 5 7 5290 .16 82 .6 957 -.1 90 5 .7 3 : 4 .278 1 0 938 VA IUMAX I WTA TED F AC TOR MATI UX TAB L E 42 F ac tors 4 s 07 9 8 9 102 0 774 17 4 0 -. 0 2 6 3 548 1 .1 6 1 7 075 6 2 4 5 5 03 0 8 .1 2 84 03 1 4 23 3 2 .0151 1 897 .1 362 84 69 0 0 8 1 1105 0 5 7 5 0 1 79 .1 4 73 1 0 36 .09 09 04 11 0383 7939 -. 0546 3938 0 1 32 .1 449 34 10 .6853 214 6 0559 .10 39 2003 -. 07 4 2 2 71 1 045 5 1 6 9 9 .1 395 0959 0948 -. 0 099 -. 070 0 019 3 2198 1 044 -. 0 2 S 3 1 8 1 8 .0692 -. 0 S 83 .1 2 6 5 6 0 036 -. 1 2 1 0 .11 76 .0 295 0 1 07 .0 5 84 2 2 51 020 4 07 42 -. 0 1 97 74 97 .2 272 23 1 9 03 98 .3 7 58 -.1 211 00 2 1 .1 287 .1 62( 1 0 1 62 2 8 0 4 -. 0 838 4 954 0 32 7 .0 079 .11 10 -. 3 664 7 l 62 5 2903 064. i J 634 -. S540 1319 024 2 1:) 6 1 21 iS .3 0 3 1 OOGS oou: OL67 0296 004 1 -. 0 8 09 .0 782 0630 1.is2 l :24i -.1 951 i S9S .2S54 1 ~-; ::;o 05~ ) 9 O.L70 11 75 ....... U7 v

PAGE 169

FAC TOR IV DJ STANCE TO ill i \.JLW II IGlll',' 1 \Y DJ ST A,-,;c c TO Nt::.i\lU:ST H-:l T l:.SL\T [ GAS / l.. VA I L AB 1 L l l' Y F ACTOR V SLOPE OF LAND (CLE A RING-COVER CONDITIONS) FACTOR VI DISTANCE TO RAILWAY (LA.i\JD USE COMPATIBILITY) FACTOR VII (DISTA N CE TO NEAREST MAJOR THOROUGH FARE) Several observations regarding the results of the factor analysis should be noted. Firstl y the variables did not group as previously conceived in Chapter II This suggests th a t a n alternate grouping of variables to form indexes should be considered. This is not particul arly disturb i ng s ince th e original g r ouping was base d upon a lit eratur e search of and not upon quantit a tiv e analysis. Secondly, the fir s t three factors account f o r n ear l y h a lf of the original variance in th e data (Table 40). Th e se factors, therefore, may be sufficient to capture the in du s trial land con v er s ion process provid e appropria te wei g hts can b e found t o r e flect vary in g lo ca tional preferences of specific i ndustr ies This would re du ce the number of variables to be measured to approxi ma t e ly 13. Thirdly, a designated or zoned place f o r industry with some industry already nearby, appea rs to b e th e most important considera tion in the site selection process. This is statistically verifi e d 15 6

PAGE 170

only for t h e 1 6 coun ty me t ropolitan r eg ion but such a n obs e rvati o n 1s n ot in..:on .:;is t c nt: with emp i ri c al ~tu di e:; n ot e d 111 C h ap t er II. F in; d ly, th e r ema inin g factor s alt.lwu ~h ide nt ifi e d as s cati. st ically distin ct hdVC a d eg r ee of conununal.i. t y in that each re lat es t o some aspe ct of accessi bilit y to th e site This sugg es ts th at a co !:1 bin cd ind ex o f ac c essib ility 5hould b e con sidered Fu r th er co mme nts and recommendations co ncerning th e development o f a s it e s e lection a lgo r ithm a r e found in the final ch ap t er 1 5 7

PAGE 171

C,t/\PT[R V SU MMARY AND CONCLUSIO:'\S The objectjves of this research effort r ep r esent only a portion of a muc h l a r g er rese arc h goa l, th e identification of the determi n a nt s a ff ec tin g land u se ch a n ge It w as d e cided to focus upon only one asrect of this problem, th e co n versio11 of l and t o industria l use. The approach \v as based upon the assumption th a t th e d e terminants of this conversion proce ss would be found jn th e "m a rk e t pl ace ," where land transactions among buy er s a nd s e ll ers o ccu r. This research direct e d t owar d only one s id e of the market transaction p cocess, namely that of the purch as er's desir es (in thi s cas e, th e industrial developer) in securing an ideal or suitable site. The probl em was to id e ntif y the ideal qualities, quantities or a ttribut es desire d in an indu s trial site a nd to formulate a gen eral algori t h mi c statement to identif y potential industrial sites. Research procedures involved th e
PAGE 172

of v ar iab l es t o 111-i_nimi:oe th e h st of \ :l r iab l cs n e ed ed to
PAGE 173

greater importance of specific accessibility variables, weights might be used in bu i ldin g si tin g potential scores. When this co mb inat io n is effected, four indexes are identified: (1) a protected or planned area for industrial development; (2) space for industry to develop; (3) accessibility to transportation media; and (4) site preparation costs. Conclusions Regarding the Proper Form of Model Oper a t io n As suggested previously, additional research would produce a more complete algorithm to simul a te th e industrial site-selection process. The factors identified in this study, however, should form major com ponents of the al g orithm. Ther e is some question as to whether it is possibl e to develop a compl e tely deterministic algorithm. It is doubtful th a t we will ever be able to predict the exact location of industrial development; therefor e, the se l ectio n of s uitable indus trial si tes ultimately should be accomplished throu g h stochastic procedures. It is lik e ly that other types of land u se d eve lop me nt will have to be simulated in a similar manner. The algorithm form suggested in Chapter I which utili zes an aggregation of indexes lends itself to stochastic simulation. This study demonstrate s a way in which p a ramet ers relevant to the loc a tion proce ss can be d e termined a nd thus incorporated into the algorithm. Conclusions Regarding the Us e of A e rial Photography in J\lodel Design A n anci ll ary objective of this ana ly sis wa s that of d emo n stra t i n g th e utility of aeria l photo g raphy in compil i n g land us e data and det er mining mea s urements on r elevant variable s A major i nit ia l 160

PAGE 174

problem in analyzing industri a l sites is finding the site and locating them relativ e to specific grid coordinates. Aerial photography was found to be most useful for locating industries a nd, consequently, reduces field work time considerably. The imagery utili z ed in this study was provid e d by NASA-Marshall Space Flight Center in Huntsvill e Alabama. It was flown at a scale of 1:24,000. This permitted superimposition directly onto 1:24,000 TVA maps and greatly simplifying location a l analysis procedures. The determination of neighborhood quality characteristics, slope charac teristies, accessibility characteristics, and size of the site was also facilitated by air photo interpretation. In fact most of th e variables identified by th e fac tor analysis procedures may be d eter min ed and measured from aeria l photography. It s hould be s tr essed th a t the u se of aerial photography was not a panacean solution to data acquisition problems. Field work and gr ound-truth su rv eys are always necessary. Analyses of this type, howe ver can ben e fit greatly from the use of aeria l p hotogr aphy. Conclusions Regarding Recommendations for Future R e s e arc h The d ata collection proced ur e s and analy s is procedur e s utili ze d in thi s r esearc h t ask have be e n prov e n to be useful in attempting to understand th e industrial l:i.nd conversi o n proc e s s Th e re is no re a son th :i. t similar procedure s might not b e ap p li ed toward an a ly z in g oth e r types of l a nd u se con ve rsion proce sses One mi g ht al s o c o 11 s idcr v a riation s of th e f ac tor an a l y s is proc e dur e s to g roup indu s tr ie s h a vin g simil a r l o c a tion a l pr e fer e n ces rath er th a 11 v a ri. a bl c s In s i g ht ac qu i r e d in th is study s u gge st s that th e 161

PAGE 175

l oc a t ion a l crit e ri a are rcg i o,nl l y d e p e nd e n t ::;u ch th at d irferi n _g l oc a tj on pa r ame t e r s 1 r.a y b e c ome .i rn t h J rta nt 1 n t,t.iic ii'C t r o p u ~ itan r egions an ;:i :. ys.i._, Th e fin a l conc l us i on is crr, p h a s i z8 d : t h e proce dur es
PAGE 176

R ESi:ARCH .J3 ~ BL.I O Cf( -'I.Pr!Y Peri o di c a l s ---A l chian, Alan A. "Unc e rtainty, [v a luation, and Econo m ic 1-Iistory," Jo_urnal of Politjcal Ec<2nomy LVII (June, 19 50), pp 211-221. The America n F actory Tod ay and Tomorrow ," Dun's Review a nd Modern In dustry ( Apri l 1 958 ). AEderson, T. R. "S ocial and Econo mi c Fac tors Affec t ing the Location of Residential Neighborhoods," Papers of th e R egiona l Science Association, IX ( 196 2) pp. 161-170. Ball, C., a nd M T er t z 11 E xpre ssw ay a nd Industrial Loca t ion,1 1 Traffic Quarterl y, XII (1 958), pp. 589 60 1. B a rlcon, Ma rvin J. 11 Th e In terre l a tion ship of th e Changing Struc t me of Ame r ica n Tr ansp o r tati o n and Changes in Indu s tri a l Loca tion, 11 L a n d Economics, XLI (1 965), pp. 169-179. !3 o ur;. .; L ar rr S 11 L: : rnd Use S ucce ss i o n in Urban Ar e as : A Stu d y of S tr u ct ure an d Ch ange 1 Proceedin gs of the Associa t io n o f Ame i ca n Geographers, I (19 69) pp. 1215. Rro1m, R. C. 11 Th e U se a nd Mis Use of Di s tance Variables in Land Use An a lysis, 11 Th e Professional Geograph e r, X X (September, 1 968) pp. 337 341. Bullington, J. S. "U tilization o f a St ate -Wide Site Evalua ti on Co mmi tt ee to Aid in th e Lo ca tion or Relocation of Plant Faci li ti es ,11 A lDC Journ.'.i l, IV (Oct uber 1 96 9), pp. 22-42. Blll'gcss, E. W. 1 1 The Deterr:1.i.nation of Gr a dients i n th e Growth of th e Ci t )' ," Amer i ca n Soci olo g ical S o c ~ e ty r> ub h e at io n s XXJ (1 921 ), pp. 1 78~ 1 84 ---Burnstall, R. M., R. A Leaver, and J E. Sussams. 1 E vaJuat i oi1 of T ranspor t Costs for Al tc rna ti ve Factory Sites: A Case Study, 11 Op e rational Research Quarterly, X II I (December 1 962), pp. .34 53.S4. Carr ier Ronald E. and Wil 1 iam R. Schri vc r. 11 Location Th e ory: An Empirica l Mode l a nd Se 1 e cted Findin g s ,1' Land Econo1:1i cs \LIV (No vembe r, 1 968) pp. 450-460. 1 65

PAGE 177

Chain 1 F. Stua-rt, Jr. \ /,lod c l. for S j1: 1 1 tlo.t ing Rc si, lcnti a l D e velop ment, J ournal of i\m L: r(c1n J_ 11 stitui.e 0Fr 1 a nn er s XXXI (~l ay J.%S), pp. 1 20 -1: 5 Citaj'man, James E., anL l \'.'jLJi. ;c 1111 / \. \\'el]s. 1-'ac to s jn Lnd u::;trial Lo cation in AtlantJ., 19 4619 55, ~ ~-~} _:rn ~ ~ Economic Revie1 ,1 I X (S eptember 1959 ), ~P3-8 Com, ay H. M. "700 P 1 a nt Location fac tor s ,' Industrial D eve lo pme!.1_!_ and M an uf ac turers Record, (Oc to ber, 1957). Czamanski, S. "Indust rial Location and Urbo.n Growth," Tmm Plannin g Revi ew, XXXVI (1 965), pp. 1 65-180 Devletoglou, Nicos E. "A Dissenting View of Duopoly and Spatial Competition," ~conomics (May, 1965), pp. 1 40-160 Dyckman, J o hn H. "Plannin g a nd Dec.i.s i m Th e ory,'' Jou rnal of t he Am e ri c
PAGE 178

H a rr.i.:; C h a un cy U., and l : d: -1a1.d L. Ul l n! ; ,n "The Na tur e o f Ci ti e s," \ nn : 1l s of th e A m.:: ric:1a Acadc : n y of i"oJjt i a l a nd S ocia l S e:r vi ce c:cn 1 -r (19 45) p p ::s 1 s ::.::; 1 s_ : --li erbe::: t .J D., a11J 8 H. Stc\ ens. ''i \ qodd for th e D istri) J ut i c. :t of Resid e n t ia l l c t iv it)' i n U!h J. n .\ r e : i s J ou rn a l of Region a l c i e ~ ~ ~ II (May, 1 960) pp. 235 2 S9. Hill, Don a ld t,I. ,Journ ;. tl o f 196 5), p p. "A Grow th Allocation itode l for the B os ton R eg ion," the A merican I ns t i tut e of P l a n ners, XXX I, No 2 (May 111 120. -------, D a ni el Brand, and W i lli ar d 1-! a nsoTJ. "Pro toty p e Developme n t of a Statistical Land U se Pr edi ction '-!ode l for th e G r ea t er Boston Region," Hi g hway Research R ecor d, N o 11 4 (J.96 6 ), pp. 5170. H ote llin g Harold. "St a bilit y in Compe tit ion ," Econo mi c Journa l, XXXIX ( Mar ch 19 29 ), pp. 4 1 -57 Hoyt, Hom er "R ece nt Di stortions of the Classical M o d e l s of Urba:1 Structur e ," L and Economics, XL (May, 1 964 ), pp. 19 9 2 1 2 Hudson, G. Don a ld. "The UnitArea r-! e thod of L a nd Cl assif ic a tio n ," Annals of th e Associ a tion of Ame ri can Geo g r a ph ers, XXVI (19 36 ;, -------pp. 9 9 -11 2 Huff, David L. A Probabilistic An a l y sis of Sho ppi ng Center Trade A r eas, 11 L a nd Economic s LII (] 963 ), pp 81 90 K ai n, John F., and James R. ~ eye r. C o mputer Si mu lations, Physio Ec onom i cs Systems, a nd Int ra Regiona l Mo del s ' A mer ican Eco nomic Review, LVIII (M ay 1968 ) pp. 1 7 1-181. Koopmans, T C., and Ma r tin B e c km a n "A ssig n me nt P roblems and th e Location of Eco n om ic A c t ivi t ies ," Econ ome trica, XXV (1957), pp. 53-7 6 Kr e ig, R. A. A e rial Photo g raphic Interpretation for Land Us e Classi f j c ation i n th e N e v; Yor k S t a te l.;:mcl U se an ci Na t 1 Jr a l R e source s Inve n tv ::-y 11 Pho t ogramme t ria Vo l. 26 (1970), pp 10 1-11 1. t e rner, A. P. c1 n d 1-l. W. S jnger Some N otes on D uop o l y and Spatial Competit io n," J ou rnal of Politic.ai Eco:--.omy, Y. LV ( A pr i l, 1937), pp. 145-186. L ons dal e Richard E "Rur a l L abor a s A n Attract ion for lndu s tr y 11 AIOC J o urn a l, IV (Oct o b er 196 9) pp 11-17. L owry I, a S. "A Short Cour se in ~ o d e l D esig n," Journa l of the Am eri c an Inst i tut e of Pl:.ln ners XXX (May 1 965 )--:pp. 15 816 6. l GS

PAGE 179

McMillan, T. E. "Why Manufacturers Chan ge Plant Locations Versus Determinants of Plant Location," Land Economics, XLI (August, 1965), pp. 239-243. Meyers, Charles R., Jr. "New Tools for Regional Planning," AIA Journal (October, 1971). Morrill, Richard L. "The Development of Spatial Distributions of Towns in Sweden: An (SIC) Historical-Predictive Approach," Annals of the Association of American Geographers, LII (January, 1963), pp. 1-14. Morrill, Richard L. "Expansion of the Urban Fringe: A Simulation Experiment," Papers of the Re gi onal Science Association, XV (1965), pp. 185-202. Mores, L., and H. F. Williamson, Jr. "The Location of Economic Activity in Cities," The American Economic Review, LVI (1967), pp. 211-222. Muncy, Dorothy A. "Influence of the New Highway Program on Indus trial Location," Industrial D eve lo p ment, IV (1957), IJP 187 3. ------"Land for Industry A Neglected Problem,'' Harvard Busi ness R ev i e w, XXXII (March-April, 1954), pp. 51-63. Nunnally, Nelson R., and Richard E. Witmer. "Remote Sen s ing for Land Use Studies," Photo gr arnmetric Engineering (~lay, 1970), pp. 449453. Olsson, Gunnar. "Trends and Spatial Model Building: An O verview," Analy s is, I, No. 3 (July, 1969), pp. 219-224 ------and Stephen Gale. "Spatial Theory and Human Behavior, The R egio nal Science Association Pap e rs, XVIII (1967), pp. 229242. Persons, B. S. "Soil Mechanic s A Tool in Plant Site Selection, Th e Pap e r Industry, XL (1958), pp. 97-98. Pred, Allan R. "The Intra Metropolitan Location of American Manu facturing," Annals of th e Association of American Geographers, LIV (June, 1964), pp. 165-180. Putman, Stephen H. "Intr a urb a n Indu s trial Location ~lodel Design and Implem e nt a tion," Papers of the Re g ional Science Association, IXX (1966), pp 199-214. Roi v A., and E. Jurkat "The Economic Forces Shapin g Land Use Pat t er ns," Journal of American Institut e of Planner s Special I ss ue (May, 1959), pp. 17 81. 166

PAGE 180

R1u11 !! 1el R J. "Un clersta ncl i n ;; F a ..::t or r \n;.il ys i s ," The J our n a l of Confl: i c t R e o l ution, \I (D e,: eml) e r, 1 967} pp 444 -4 80 S-: h r i.\ -c '. r \\' i lli a ::1 lL "Th e h du : ; t ri :11 i. : .:1 ti o0 of th e S outh ea s t sin ce 1950: S or~e C auses of : 2 nufa c turing P.d o c a ti on 1 ,itJ 1 Speculation abolit th e Eff e c t s ," T he \s-i-,2 ri.c : rn Jou:rn ::t l o f Econo m ics and Soc olo g y ~ XXX (J an uary, 1 97l); i:i: 47 -70. Se i in e r, I. "Random Walk a nd Random Rou g hn ess Models of Drainage Net1 o rks," lfa t cr R es ourc es l< ese arc h, V (1969), pp S91-607. Sh e nk e l, William M "The Economic Consequences of Industrial Zoning," I. a nd Economics, XL (1 964), pp 25S -265. Smith, David M "A Theoretical Framework for Geographical Studies of fndustrial Location," Economic Geography, XLII (April, 1966) pp. 95-113. Sr,iithies, Ar thur. "Optim a l Lo ca tio n in Spatial Co mpet ition," The Journal cf Political Econom r_ X LIX (Jun e 1941), pp. 423-4 3 9. Spie ge lman, R. G. "Activity Analysis Model s in Regio nal Development Planning,'' Regional Scienc e Association Papers, XVII (1966), pp. 143-1S9. St affor d, H A An I ndus tri ai Location Decision Model," Proce e din gs 0 the Association o f Ame ric an G e ographers, I (1969), pp. 177-1 88. St eg er, Wilker A. "Th e Pittsbur g h Urban R enewa l Simu lation t 1od el, 11 J o urn a l of the ~ ~ ericnn Institute of Pla~ n er s X XXI, No. 4 (Ma y 1965), pp. 1 44 -150. Steven s Benjamin H. "Linear Programming and Loc a tion Rent," Journ a l cf Re gi onal Science, III (Fall 1961), pp. 15-26. ------"A Revie1v of th e Literature on Linear Methods an d Models for Spatial Analysis," Journal of th e America n Institute o f Planner s, XXVI (May, 19 60), pp 253-259 T h o mps.: m, James H AJ[' C Jc.u.:-na l, ~ ----"Th e Decision -Makin g P r o ce ss in PLint Loc at ion," II (O c to b ~ r ) 1 966 ), ; 1 ~1 1 -20 T ie bout, Charle s M. ''Loc:itio n Theory Emp i ri ca l E v idence a nd J: c o no mic Evolution, 1 P ap12 r s and P:::-oceedings of th e Re g i o n a l Science ~ssoc i~t ion, III (19S7), pp. 74-86. Turner, Robert G. "Empi r ic a l Studies of Plant Location: A Surv ey ," AIDC Jo ur n a l, VI (April, 19 7 1), pp 1330 ------" Genera l Theor.ies of Plant Location: A Survey," AIDC Journal, VI (October, 1971), pp 2 1-36 1 67

PAGE 181

von Boventer, Edwin. "Towards a U nited Theory of Spatial Economic Structure," Papers and Proc ee dings of the Regional Science Association, III (1957), pp. 74-86. White, Gilbert F. "Industrial Water Use: A Review," Geographical Review, L (1960), pp. 412-430. Williams, D. C., Jr., and Donnie L. Daniel. "Industrial Sites for Small Communities," AIDC Journal, VI (April, 1971), pp. 33-39. Yeates, Maurice H. "Some Factors Affecting the Spatial Distribution of Chicago Land Values, 1910-1960," Economic Geo g raphy, LX (January, 1965), pp. 57-85. "You Gotta H a ve a Golf Course to Attract Industry," Business Week, June 25, 1955, pp. 86-87. Books Abl er Ron a ld, John S. Adams, and P ete r Gould S pa tial Or ganizatio n. Englewood Cliffs, New Jers e y: Prentice-Hall, Inc.; 1971. Ackoff, Russ e ll L. Scientific Method: Optimizing Applied Research Decisions. New York: John Wiley and Son s Inc., 1962. Alexanderson, Runn ar Geo g raph y of Manufacturing. En g l e wood Cliffs: Prentic eHall, Inc., 1967. Alon so William. Location and L a nd U se : To wa rd a Formal Th e ory of Land Rent. Cambrid ge : Harv ard Univ ers ity Press, 1964. Barlo wa Raleigh. Land Resource Ec onomic s : The Economics of R ea l Property Englewood Cliffs : Pr en tic eHall, Inc., 1972. Bartholomew, D J. St oc h as tic Models f or Social Pro cesses. Wiley and Sons, 1 967 Bartholomew, Harold. L and Use s in American Cities. Cambridge : Harvard University Press, 1955. B e rry, Brian J. L., and Frank E Horton. G eogra phic P erspect iv es on Urban Systems with Int egrated Readin gs Engle~ood Cliffs: Prentice-Hall, Inc., 1970. Birch, J. W L 2 nd U se Emb l e ton raphcrs, "Rural Land U se : A C e ntral Th e me in G eoz raphy," in and R es our ces : Studi e s in App li ed G e ography, Eds. C. and J. T. Coppock, Lo ndon: In stit ut e of British G cog 19 68 pp. 13 -28 168

PAGE 182

Boyce l~ o nald P.ee d The 3:7. '.:>CS of Eco n '.)1: i ~ Ceo g r aph: : .\n Essa, on t.h e .Sp:1.ti:1 1 Char ac t e ris r.:i cs of ~b.n 1 ::; Economic A c tivities Atl: : n t .i: Holt Rin e har_t .. :-01J \'i i nst~)rl~ Co 1r 1 9 7 ..i~[.; r oonc r, \ ~.ilJ i a; n C., and D a v id 1\. \ i chch. ''C o r> s id c ra t jon s and T e c l111iquc s fo r Inc orpo r atj n g Rer;.o t e l y Sens eel I magery i n to th e L a nd l ~cso un :c M a nagc 1:!e nt i-)roce :;s ," i n Remote S e n .,; in g of Earth R es ou rces Vo l. I Se l ec ted techni cal p ae rs from th e Conf er ence on Ea rth Resources Observation and I11 form ation A. n aly~is Syst em s TulJ.ahorna Ten ne ss e e: The Unive r s ity of Tennesse e Space Inst i tut e 1 9 7 2 p p 12 5. Bro,m, 1-1. J ame s and others. Empirjc a l Models o f U r ban Land U se : Su gg es t io ns on Re sea rch O b_JectTves anJ Organ i za t ion. Explora tory Report 6. New York: Colu mb ia Univ e rsit y P ress, 1972. Bun g e \Villiam. Th e oretical G e o g r ap hy. Lund, Sweden: Royal University of Lun d 1 9 62. C arrier Donald E ., and William R. Schriver. P l ant L oca tion Analr sis: An Investi ga tion of Pl a nt Loc a ti on in Tennessee ~ler.i p h is : Memp his Sta te Univ ers ity, 1 969. Cha ; 1berlin, E. H. The Theory of ~ l o nopelistic Co r.ipe tition. Cambr i dge: Har va rd Univ ersity Press 1936, pp. 1 94 196. Chapin, F. St uart Jr., and Shirle y F Weiss. Facto rs Influencin g Land D ev elopment: Evaluati on of Inputs for a Fo r ec:"Ls t t,!odel. Urb a n Studies Mono g ra p h. Chc:. p el Hill: In sti tutr : fo r Re search in Socia l Scie :n ce Un i v ers it y of ,'\orth C n roLi.n a 1 96 2 Claw so n, M ario n, and Ch ar le s L. St e wa rt. L an d U se Infor ma tion. Baltimore: The Johns Hopk i n s Press, 1 ~6 9 Cox, Ian H. (ed ). New Po ssi bilities and T ech ni ques for Land Us e ancl Related Survey s Papers present e d at an int e rnational sy mpos l um London, Apri l 21-23, 19 70; O c casion a l Papers N o 9, World Land U se Surv ey Commiss i on; To nbri J ge ~en t: Tonbrid ge Printers Limit e d, 19 7 0. F l aw n, Pe tPr T Environmenta l G eo lo gy : P l an n i n g a nd Re s our ce Manag eraen t. Rm~ubiishers, 19 70 Con se rv ation Lan d Use New York: H a r pe r a nd Forrester, J. W. Principles of S ys t ems Ca mb ridg e: Wrig ht All en Pre ss In c ., 1968. Garri s on, William L. Difficult Decisions in L an d U se ~Io d e l Con struction. Chic ago : Nortl11, e stcrn Un ive rsity, 1 966 Greenhut ~lelvin L., and Marshall R. C olb e r g t i.or1 of Florid a Industry. Tall a h assee : Un i vers ity, 1 962. Fact or s in t he Loc a The F lori d a St a t e 16 9

PAGE 183

----Plan t Loca ti on rn ni _<:_'::r)~a nJ _:0_1 Pr a ctic e Ch ape l H ill: Univ e r s ity of :fort!-. C 2roJ ina Pr ess 1 9 :) (, r ; r i t'fin .b h 11 I. Inc.l~ i:; t, i .~ J L oc at io n '.: i i, -r L c.) \ e i: Y o rk \re a Kew -----Y o r i--. ~: i ty : Th e City Co l le ge Pr ess 1 9::i:'i H a~:te tt, r ,'' c r. Location Analysis in H tunu. n Ge ogr ap hy \ eh' York: St Martin's Pre ss 1 966 H ami lton, F. 1:. Ja n. 11 i 1 l ode l s of Industriol Location/' in So_i_~ -::__ Economic Models in Geography, E ds Richard J. Chorley and Peter H agge tt. New York: Barnes and Nob l e 1967. H a milton, H. R., and oth e rs. fu1 Application to River ~l.T. Press, 1969. Systems Simulat io n for Regional Analysis Basin Planning. Cambridge: The Ho ove r, Edgar M. Th e Location of Economic B e havior. New York: ~!cGraw-Hill, 194 9 Location Theory and Th e Shce and Leather Industries. Cambridge: Harvard Univer s ity Press, 1937. f-br s t, Michael E. Eliot. A G eog raphy o f Economic Behavior: An Jntroduction. BelmonC"california: Dux b ury Press, 1972. Isard, Walter, and oth e rs. Ecologic-Economic Anaiysis for Regional lJ ev elopment. New York: The Free Press, 197 2 ls3 r d Waler Location and Sp ace-Econo my \ew York: The il.I.T. Press and John Wiley and Sons, 1956. ------Methods of Regional Analysis: An Introduction to Regional Scien-ce. Cambridge: Massachusetts In s titut e of Te c hnology Press ,-1960. Kariel, H. C., and P. E Karie!. Explorations in Social Geography. Reading, Massachusetts : Addison -Wes ley Publishing C ompan y 1972. Kenn y J. G., an d J. L. Snell. ~b th ematic a l '-lode ls in t '. 1e Scc iai Sci e nces. W al tham Massachusetts : Bl aes de l l Pub li ca tion s co-:, 1%2 King, Leslie J. Statistical Analy si s in G eogra phy. Englewood Cliffs, New J e r sey : Prentic e -Hall, Inc., 1969. Ka raska Gerald J., and David F. Bramhall. Locational An alysis for Manufacturing: A Selection of Readings. Cambrid g e: The t-r.I.T. Press, 1969. Kv e nne, Rob e rt E. The Th eo ry of General E c onomic Equilibrium. P r inc e ton: Pr i nceton Univer s ity, 1 96 3 1 70

PAGE 184

Malinowski, Zenon S., and William N. Kinnard. Highways as a Factor in Small Manufacturing Plant Location Decisions. Storrs: University of Connecticut, 1961. ------Persorial Factors Influencing Small Manufacturing Plant Location. Storre: The University of Connecticut, 1961. Marble, Duane F., and Frank E Horton. "Remote Sensing: A New Test for Urb an Data Acquisition," Urban and Regional Information System: Federal Activities and Speciali z ed Systems, Ed. John E. Ricket. Akron, Ohio: Hirey Printing Co. and Kent State University, 1969, pp. 252-257. McClellan, Grant S. Land Use in The United States. New York: The H. W. Wilson Company, 1971. McCracken, Daniel D. "The Monte Carlo Method," Mathematical Thinking in Behavioral Sciences. San Francisco: W. H. Freeman and Company, 196 8. McHarg, Ian L. Design with Nature. Garden City, New York: The Natural History Press, 19~ NcLoughlin, J. B. Urban and Regional Planning--A Systems Approach. Lond on : Faber and Faber, 19 69. Miller, E. Willard. A Geography of Industrial Location. Dubuque, Iowa: Wm. C. Bro1m Company Publishers, 1970. Mize, J. H., and J. C. Cox. Essentials of Simulation. Englewood Cliffs: Prentice-Hall, Inc., 1968. Nie, Norman, Dale H. Bent, and C. Hadlai Hull. Package for the Social Sciences. New York: Book Company, 1970, pp. 208-244. SPSS: Statistical McGraw-Hill Ray, D. M. Market Potential and Econ o mic Shadow: A Quantitative Analysis of Industrial Location in Southern Ontario. Chicago: University of Chicago Press, 1965. Simon, H. A. ~lodels of ~.Ian New York: John-Wiley and Sons, 1957. Smith, David M. Industrial Location: An Economic Geo g raphical Analy sis. New York: John Wiley and Sons, Inc., 1Y71. St ei nlt z Carl, and Peter Rogers. A Sy stems Analysis Model of Urbani zation and Ch ange: An Exp eri m e nt in Interdisciplinary E
PAGE 185

Stogcl ill, R a lph ; ,!. (c J ). Th :~ Pro ce ss o[ i \ r 1d e l Building i n the Behav ior c1 l Sc i. ence Co lum bu;-; : Ohio SL.1tc hiversitr Press 1 970 ~1\l'C1. ~ t Dav j cl C L~xin r.,ton :Lssachusc:tts : U. C. l! c;:ith anJ Company l 9i' 2 Thol!lpson, l hlber R. /1.. Preface to Udnn E c ono mir.., B a l ti rno r ~ : Th e John s H op kins Pr ess 19 65 Tryon, R. C., and E Bailey. Clu ste r An a l ys i s. Ne w Y o rk : NcGraw Hill, 1 97 0. V a n de Geer John P. Social Sciences. 1971. Introducticn to Multivariat e Analysis for th e San Francis c o: W H. Freeman and Company, von ThUnen, Johann Heinrich. Der isolicrte STAAT in B eziehu ng Auf Land. 1 ,iTtschaft Lind Nationalokonom ie s. Ber l in : Schu:nacher Zarchin, 1975 (in K. W. K app and L. L. Kapp, eels., Read in g s in Economics. New York: Barness N oble, 1949, p. 299 \lew York : .American Re sea rch Council, 1 96 0 Government Documents and Agency and Institute Research Reports funerican Waterwa ys Op era tors. A Study in Economic Growth : Waterside Sit e Plant Loc at ions and Expansions Sinc e 1952. 7. nc.l Edition Washington, D. C.: American \faterways Operators, 19 62 Arv a nitidi s N V., and oth~rs A Co mp ut e r S imu l at i o n ~ l o de l for Flood Plain Development. Part 1: L and Use Pl ann in g and B e nefit Evaluation A r epo rt sub mit t ed to U. S. Army Engi n e er In s titute for Wat e r R es ources by INTASA, Menlo Park, California, February, 1972. Bo s selman, Fred, anc. David Callie s Th e Quiet Revolution in Land Use Control: Su mm ary R e port Washington, D. C.: U. S. Govern ment Printing Office, Council on Environmental Quality, 1971. 1 72

PAGE 186

B o y ce D r an J R. W. Cot V e r i f i c a tion nf L a n d U s e F o r ecas tin g ~ l o d e l s : Pr ocec lt 1T es and D :1t a R e qui r e; ne ~ i ts -Co lu mbu s : B a tt e ll e fre m o ri a l I n s t i tu : e 19 6 6 Cc-1r rol l. J l l ., r Ch icaso \ r c i.?. Tr a n ::;porr c.1. t i o n St u dy V o l. ?. Data P r oj cc t.i on s ; \ fash i n g t o n, D C. : U. S !) 2 p ar t ne n t o f Co mm erce l3 1T e a u o:f Pu r. : i c R oa d s, 1% 0. Cha pi n, F. S t u a rt, and Shirley F. W e i ss Som e In p ut Re f inements f o r a R csi d 8 nti a l t! o del. Ch a pel Hill: Univer s ity o f : ~ or th Carolina, C e nter fo r Urb a n a nd R e g i o n a l S tudie s I r t s t i tu t e f o 1 R e search in Social Science, 196 4 Co r1 a n, Peter !-l. E N V1RO COU '.'! TY : A G am in g Si m ulation of R e g ional Plannin g Proc e s s Oak Rid z e, Tennessee: Oak Ridge N ational L a boratory ;-o RJ \JLN SF Environmental Pro g ram, 1973. Co n kli n, H. E. "The C o r n e ll S :-,s t e m of [ co nomic Land Cl assifi c a tion," Farm Econo m ics N o. 198. Itha c a, Ne w Y or k: Cornell U n iv e r s ity, J a nua r y, 1955. Cons a d Research Corporation. A Time-Orient e d Metropolitan ~ ~del for Urban Land Use Analysis. Research R e port No. 3; S e atr.le, W a sh ington: University of Washington, Urban Data C e nter, 1967. Crawford, Roger J. Utilit y of an Autom a ted G e ocoding Syst e m for Urban Land Use Analy s ~ Research Report N o. 3; Seattle, W as hin g ton: University of Washington, Urban Data C e nter, 1967. C r\' ; cine .J oh n P. C o m p u t er Sim: .d a t: ion in U rbr t1 1 R es ea r c h. S a nt a r o n .i. c. a : The RJ\,',JD Corporation, 196 7 ----Spatial Location Decisions and Urban Structure: A Ti m e Oriented Model. Discussion Paper No. 4, Ann Arbor: University of Michig a n, Institute of Public Polic y Studies, M a rch, 19 5 9. -A. Time-Oriented M e tropolitan Mod e l for Spatial Location. Dep a rt me nt of Cit y CR P Technical Bull e tin N o. 6. P i tts b ur g h: Plannin g 1964. D::i.r n ton D m a l d C. Lo e a t i on F ac t ors Af f e ct i ng L abor -1 n t e ns i ve Ind u s tries i n the Up pe r O hio Ei v e r Va ll e y Me mo g raph No. 1. :'\ th ens Ohio: Ohio Univ e r s ity, G l v i sio n of Re se a rch, Coll ege of 3 usin e s s Administration, Jul y 1962. Donnelly, Thomas G., F. Stuart Chapin, a n d Shirley F. \ 'le i s s. A Prob abilistic i lodel for R e sidenti a l GrO\ ; th. Chapel Hill: University of N o r th Carolin a Institut e for Res e a rch ir Social S c i e nc e 1964. Du :-fe e, R. C. Modelin g S e pt e mb e r Th e U s e of Factor a n
PAGE 187

J:;rv i n, O sbin L ., and Ch:ir l cs R. 1-l e y e r s .J r Th e Util.i za tion of Loc:i l Opi ni on in L ~nd U sc Simu l at i on \l c, d e lin g : /\Delph i. Approac h. Oat Ricl : ,c, T c nnc :ssce : O a k hcl g c \ ;_,t Tur ia l L 2bo rat ory OR~lL-\S r l::.m._i rnn ; n cn tal P o r rc tm l J73 (Ir l-ho;t s~ :.\~ : mo ) F e .1.Jt, A C No. 3 1 968 Th e Co m unity L :rncl l 1 sc G;111 1e (C: L U G ). Misce lJ ancous P a per I t h acci : C o rn (; ll Un.:. ve:r s i t 1 Ui vision of U rban Studi es Frazee, C J., !. I. Mye r s and West. R e mo t e S e n s in g fo r De t e ction of Soil L im itations in A g r i. cultur a l Ar ea s," Pr ocee dings o f th e Sev en th Int e rnational S ympo siu m on R emo te S ens in g of Environ men t, Vol. I, A nn Arbo,, Michi g an: Unive r si ty of Mi c higan ., 1l ay 1971, pp. 327 343 Gardner, S. S. A Stud y of Environment a l M o nitorin g and Inform a tion. Iow a City: I owa Uni ve rsit y Iow a City I ns titut e of Urb a n and Regional Resea rch, J a nuary, 19 72 Gib s on, J Sulliv an N B. Gu yo l, and G. Donald Hudson. At l as o f the Tennessee V a lle y R eg ion The Nat u ra S e ri e s Ch a tt a noo ga: Tenness ee Va ll ey Authority L an d Classification S e ction, Divi sion of Land Plannin g and Housin g Jun e 19 36 Go e hring, Darryl R. Monitorin g th e Evolving L a nd Use P a tt e rns on th e Lo s Ang e ies Metropolitan Re g ion. Rivers i d e : Uni versity of Cal i.fornia, Department of Geo g rap.hy, October 1 971 G oldoerg, M A Intr amet:-opoli t a ;1 I nclus t ::j o. l Lo cat i on : Pl an t S i .z.~ ~nd t]i ~ Theo ry of Pro c luctiun. B e r kclr::y: University o-f Califor nia, Center for Rea l Estate and Urban Ec onomics, Institute of Ur ba n and Re g ional D ev e l opme nt' 1 969 ------Quantitative Approac h es t o L a nd ~l a nag eme nt. Vancouver, B. C.: Uni ve r s it y of British Col umb i a, The Resource S cience C e nt er 19 70 Goldner, W. Proj ect i ve Land U se i', l ode l (PLLJ !', l): A i',lode l for th e Sp a tial Al loc a tion of Activities a nd L and Uses in a Metropo l i t a n R egio n. T ech nic a l R e po r t 2 1 9 S a nta ~!o nica: B ay A r ea Tra n s=portatio;-;. Stu
PAGE 188

----Org:rni z ing th <:.U s e of ~lodcls in ~l c tropolita r : Planni n g Fhilad c lphi. a : lln 1-. c r si ty of Fhilad t:' 1 pbia l nstitutefor 1 )wiron m e ntaJ Studies 1 9( 5. i-1 e J 1 :12 r 0 l a f FutL;r c The Delphi M e th o d for :-:iy .s ter 1 i z .i. n g J t, cl g.r 1<. ::! nts ::i.bout the Lo s 1 \ : 0 g clc; : ; : -U:1 T , crsity oi :--C i .. lifornia , \ p i:il, 1 966 ----an:l others. D e v e l opm8nt of Long Ran ge For ecas tin g Me th o d s for Connecticut: A Sum r. iar :1 I FfR e port R -5. ~tidd l etow n, Corm.: In stit ut e for t he F:.1tur e, Se pt em ber, 19 6 9. Hoyt, Homer. The Stru c tur e ancl Growth of Resi, l c nti a l Nei g hborhoods in Am eri can Citi es Washin g ton, D. C.: U S. Government Printing Office, 1939. Hudson, G. Donald The Rural L an d Cl ass if i cation Pro g ram: A Sum mary of Techniqu es and U ses Chattanoo ga : L an d Classification s ecticn, Division of Land P l a nning c md !l o usin g T en n essee Vail ey Autho ri ty, D ecember 19 3 5 Jon es Clar e nce F Th e Rural L an d Classifi ca tion Program of Pu e rto Rico. Evanston, Illinoi s : Northwestern Dniversity, 1952. Knox, Du a r.e S. The Distribution of Land Values in Top e ka, Kansas. Lawr e nce: University of Kansa s Bureau of Busin ess and Economic Research (n d.). L e opold, Lena B., Frank E. Clarke, B B e rve e Han s haw, and James R. Bo lsl ey A Proc e dur e for Eva luatin g E n v ironmental Im pac t. Gt:olog i ca. l Survey C:i.rcul.:r.r G is ; Tia:shi 1 1gton D. C .: U S G:)Ve:;:-n m e nt Printing Office, 1971. Lowry Ira S. Seven Mod e ls of Urban D eve lopm e nt: A Structural Com parison. Santa Monica, California : RA N D Corporation, 1 967. Mar, B. W., a nd W. T. Newell Assessment of Selected RANN Environ ment a l Modellin g Efforts. A Report Pr epare d for Environmenta l Systems ~ nd Re s ourc e s Divi sio n National Scien ce F ound a tion. Seattle: University of Wa s hin g ton, June, 1973. Mo.rb l e Du ane f., and Fra n k E. Horto n. "E xtrac tio n fo U rha n Da t a from Hi g h and Low R eso lut ion I mag es," Proceedin gs of the Sixth Intern a tional Sy mposium on Rem o te Se nsin g of tl1 e En v i ronm e nt. Vols. I, II, and III; Ann Arbor: The Univers i t y of ~li chigan 1969, pp. 807 817 McCarthy, J. F. Hi g hways, Truck s and New Industry, A Study of Chan gi n g P at terns-fr!Plant Location. Chicago: American Truckin g Asso ciation, 1963 ~IcLau~ 1 hlin, Glenn E., and Scef a n Roback Wh y Indu s t ry ~loves South. W as hin gto n, 0. C.: Na ti o n a l Pl a nning A sso ciat io n, 19 49 175

PAGE 189

~-l e ~ h e nb erg t-lich::tcl J. bnir o1, 1 ,:e n ta l r ar :n i n g : J\ Se l ected Ann ot ~ted Biblio r, r ap h _v ~-l c2g o: /\rn e ;:ri~: i n Soci e ty of Plan ni: n g 0 ff i c i 2. (s 1 9 7 Cl M 0 y cr::; C R. Jr ., /\ l-1 \-'oc ll:c, c1nd R C L 1 urf cc i.. a n
PAGE 190

Stefaniak, N. J. Industrial Location within the Urban Area: A Case Study of the Locational Characteristics of 950 Manufacturing Plants in Milwaukee County. Milwaukee: Wisconsin Commerce Reports, 1962. Stradel, 0., and B. G. Hutchinson. Notes for a Short Course on Prac tical Applications of Regional Development Models. Waterl00, Ontario: University of Waterloo, The Transport Group, Department of Civil Engineering, November, 1971. Swerdloff, C. N., and J. R. Stowers. A Test of Some First Generation Residential Land Use Models. Paper prepared for presentation at the 45th annual meeting of the Highway Research Board, 1966. Toebes, Gerritt H. (ed.). Natural Resource Systems Models in Decision Making. Proceedings of a 1969 Water Resources Seminar; Lafayette, Indiana: Purdue University, Water Resources Research Center, 1969. Tomazinis, Anthony. 11 An Introduction to Mathematical Models," PJ Paper No. 8, Harrisburg: Pennsylvania D ep artment of Highways, Penn Jersey Transportation Study, 1964, p. 5. U. S. Department of Commerce. How to Make an Industrial Site Survey. Washington, D. C.: U. S. Government Printing Office (n. d.). U. S. Department of Co~nerce. Industrial Loc a tion Det e rminants 19711975. Washington, D. C.: U. S. Department of Commerce, Economic Development Administration, F e bruary, 1973. U. S. Department of Commerce. Weather: A Factor in Plant Location. Washington, D. C.: U. S. Government Printing Office, 1961. US GS Conference on Land Us e Information f e rence Papers. Washington, D. C.: U. S. Geological Survey, June 28-30, and Classification: Con D e p ar tment of the Interior, 1971. Vegas, Paul L. A Proc e dur e for th e Use of Small Scale Photography in Land U se Mapping. N ASA T ec hnical Report, E ar tt1 Resources L a boratory at Mis s is si ppi Test Facility, ~larch 1, 1972. Well a r, Barry S. Generation of Hou si n g Quality Data from Multiband Aerial Photograph s NASA, bntract No 14-0S-0001-10 654 Task 160-75-01-44-10; Int erage ncy Report NASA-119. Washin g ton, D. C. : Departm e nt of the Interior, U. S. Geological Survey, r--Iay, 1968. Oth er Sources Dickerhoff, H. E. "F actors In f lu e n cing Industri a l Plant Lo cation in East Tenness ee ] 95:J t o 1962," Unp ub Ji s h e
PAGE 191

Eperson, Terry E. "Geographic Factors Influencing the Manufacturing Industries of Upp e r East Tennessee," PhD Dissertation, University of Tennessee, 1960. Marble, Duane F. "Transport Inputs at Urb a n R e sid e ntial Sites: A Study in th e Transportation Geography of Urban Areas," PhD Dis sertation, University of Washington, 1959. Meier, M. L. "Factors Influencing the Location of Electronic Manu facturing Plants within the Boston Metropolitan Area: A Study in Micro-Location," Masters Thesis, Harvard University, 1961. Northam, Raymond M. "An Analysis of Rec e nt 1nd.ustrialization in Northeastern Georgia.'' PhD Dissertation, Northwestern University, 1966. Stegman, l\1ichael Allen. "An Analysis ::ind Evaluation of Urban Residen tial Models and Their Potential Role in City Planning." PhD Dissertation, University of Pennsylvania, 1966. Walker, David Frank. "An Adaptive Framework for the Study of Indus trial Location Decisions,'' PhD Di s sertation, University of Toronto, 1971. 178

PAGE 192

A PPEN DI X A SAMPL E S URV E Y FO R M

PAGE 193

Sfu\lPL E SURVEY fO Rl'-1 L atitude L on g itud e -Indu s tr y th m e ORNL Industry N o. LJ a l1;; o f Ent c y SIC No. Quadrangle County Employment Beginn i ng City Employment N ow 1 , 3, i r; 6 7, 8, 9 1 0 S lo p e of J.and ~ J. 2, ~) 4, s' 6, 7, 8 9, 10 Dr ai 1 1 age l 2, 3' 4, 5, 6, 7, 8 9, 10 Clear i ng-cover conditions 1, 2 3, 4' s, 6, 7, 8 9, 10 Di st anc e to center of to w n J 2} .) 4, 5 6 7 8, 9, 10 Di s ta nce t o n e a rest ma j or t h r ou g h fare 1' 2, 3 } 4 s 6, 7, 8 9 10 Den s it y of l a ri d u s e :i n i mm e d ja t E vicinity I ? 3' 4, 5 6, 7, 8 9, 10 Rating of pri~e of l a nd 4 1' 3, 4 s 6, 7, 8' 9 iO Percent of urban are a within 2 1/ 2 11 ,-i L .: s 1, 2 3' 4, 5 6, 7 8' 9 10 Di s t, : 111ce to major highwa y 1' ? 3, 4 , (), 7, 8, 9, 10 Di s t ance to s econd ar y ro acl .) l 2, 3 4 5, l>. 7 8 9 10 D i s t a n c e to ra i lwa y 1 .) ) 4 5 6 8 9 1 0 Di s t n ee l o c : 1 r p 0 rt , L I> ] /_ ; 3 4' s 6, 7 8 9, 10 D .i. s 1 anc ,~ t o , a t e n, a y 1 C 2' 3, 4 5, 6, 7, 8, 9, 10 Di s tance to neares t lnt e r s t.ate 1 2 3 4, 5' 6 7) 8 9, 10 0vnall quali ty o i ac c es~ibi J ity (th e n ) 1, 2 3, 4 s, 6, 7 8' 9, 10 Gverall quality of acces s ibility (now) 1 80

PAGE 194

1 8 1 1 / 3' 4 s, 6, 7 s, 9 1 0 C i t y \\ ::l. t e r av a i l ab i. li t y 1 ?. 3 4 r 6, 7' 8 9, lt) Gas avai l abi h ty ,J 1 L, 3, 4 :, 6' 7, 8 l' ::i, 10 City s eivage a v a il ab ilit y 1 '") 3, ,1 5 6, 7 8, 9, 10 \ fas i t zoned for indu stry? ..I. ,.. r l, 2, 3, 4, 5, 6, 7, 8 9, 10 \ fas i n dustry a lready in a re a'? l '> L. .) 4' 5 6, 7 8 9 10 Over al l ratin g of cont:iguo u s l a nd u s e 1 '") 3, s' 6, 7' 8' 9 10 Cond .L tion of r1eigh b orhoo d ,.. '+, l ? { 4 5 6 7 8' 9, 10 N e ,L by com w n1 t y s e p ; i c es .. ., ' 1 ") 5 4 5 6 7, 8 9 10 \fa s t he site :i.n an indu s tri a l park ? 1, ) 3, 4 5 6 7 8 10 \\ 11a t v a s the LJ. U 3 lit )' o J th e si t e? ) > 1 ? 3, 4 s 6 7 ) s 9 10 p o ; u n it y t ' r n o xvi l J c : 1, 2' 3, 4 5, 6 7, 8, 9 10 Amount of other i ndu s try located nearb 1 2' ., LI. s' 6' 7' 8 9 10 W a s a bu 1ldjn g a] r.e a dy t h ere ? :., '

PAGE 195

APPE N DIX B iv'~t\P OVERT. _t \Y S

PAGE 197

.,. \ / \ I\ \ I c~ ~( yi\ i\ r / ( \ iv ~ I'_} ) Vt l < ~ .... / -Jt l~ V < ) fl I n n cu...r _C_ lJv -u lrJ I L, A ...... c;_ V \ IL-/ \ V \ T V A NAVIGATION < j 0 25 0 000 i n '----' Tr o nsv n Mrcalo r P r o i.cti on /"'-. '~

PAGE 198

a_____,r' ..... \ \ Jl' l o# tQ.O ,.,... ,...,, /KR' JUJltu \ "~ \ .,o, I "" ,'WNU /0~ ,.,,.. .. ,,_ <'.~ JUflSYIUL ,..., ,_ .. J., z tA J J UI ,-.,w -,, """ --~ fi\-"" ,,_,. I / ( i.~II" -.,,.,.._ ,., < .,. l(['fll p,c I ,.. ,.,. ,.._ kf''flll50o Clff J"" II I C I\ """' ,. JWl(41 .,, ... ~, .. P,.l'f.111 <'1 Jll'lllf:ii, ,.. ,p,. 11 a: JO,,., J ,, .,_..., "' .,_.,., ..u ) ,,..., ..... ... ,, ~ pc.111111 .. ,u ,. __ .,,._ / )C'flill'WIIC J J.III IS-IU ni,,~< .,.. ,~ ,._ 1i,,c,n ,._ ,f lCU,O BfU: ) IOCl$>1UL .,.... ,w ,-lvi p.u ~ ,1 IIIOOJ'l, I \...,.,__ --./ \ _/ "'---,r[ID liOff1ut \ / \ 2 ,_, =~ j TOWNS AND COMMUNITIES --~ .,.,,.,..

PAGE 199

V \ / \ l \_ I\ _) \ ,. \ / \ Claibo1 ne l ( Scott i \,. Cam p bell i'---) ~ ?.\.,. ...t',...J_ \" f~ \ ? < '"'_/'r/-~-\J':. ,-',v( I ( v5\_11 \ _j I\.,__ j "\ Union ) Grainger .. ,,I i.. .r L--....__ >---r---..j \ i ( Hamblen V "\ (I. / __ .,...."\ :v j ,..J ~ < Morran y / ---/u --\ __ / \ ("/ ----\ l.. ., Anders a n ./ / cf~ ). '\. 'l, ., Jefferson \ / \ ,,,.) \ ', \ \_, ,-. ......-\ r---Knox ,\ ( ) \ v'('-... I rv "= ")_ Cocke i-, ~ -y r---0 ,/ / j /\JJ ..... -J/ I., \ I ) < Roane \j \ i-'J D ;--.c l. ~Sevier s rv) / Loudon ( (_~ I ___...I j \ Blount \., _r '-.. i._,__ .__, \__('~ <, c'-.. ) i \ V '"'--"\ --.,..I / 'j -"-.. "'--.... \ \. __ ...,,. _/ ( \ Monroe r \ j COUNTY BOUNDARIES ---

PAGE 200

' "' { }. ... t ._ __ r,....-{ _ ... \ --. --' .. \ _.I ... .-""' .i ( i\ -' ,-t:.. -"'-\._ J . ... ) r ___ ... ~-, r, . i -r ~ ...... i r' .... ~-[7.t "\... -<~ "' lf~ .... ) l.r-. "i .5 /" l-}\ ,~ r J r I 1 fl '-~.. ,.. ...-~ .. ,../ y-~ --}. ,.c? \ ( ,.,.-1I ,....__, l.. // J ~h J' _.,..--./ } : \ \~ /..' ,-.-; .,/ '"'i \ ____ \I\ LL; ~r -~-;, ,..._..__..--"' i \ .. A.,_ .. ___ /~ v--' r l's \ r-.. ["-. \ ~ X p,::: .. ..._l i .. '----. ""--.. ---=--"'-r ,.., _.r (1 .. --....,...r VY ,,...,---l/ / ~ / .----rI / f-D / ) -=:--v-------,lvo V [ 1.----~) L..-,J h_._ I y lr/ ')_ .J v _........., V V J r-----\ ..,L/ '/ \ \ RAIL < j 0 250 000 in '---I Tron1ti1en Mer c ator Proict i on ...._ '--' -

PAGE 201

...., ,,. LAKES -+----+----1 0 250 000 in '--------' T,ansven Merc:oto, Projection

PAGE 202

Robert B e n Pon~a wa s born in ~ L h e n s G eo rgia, o n Au g ust 23, 1 941 ~k received both his e le mentary and s econdary education in Athens He enter ~ rtti rn c ln Ju) y, 19 7 4 l i e a ~ c~r ~ e J a po s ition as Researc h A s socl a l .e int ~~ D e p art m ent of : i s -;: ,:r esently em pl-Jyed During h i.s ,.; cad e rnic career h e h as been a r" _;ea r :::h 2ss oci at.c for the N a t iona 1. Ae ronautic s a nd S pa c _; Ad~1ini s ~l'3.H o :;: rnarrieJ and h e a nd his wife h a ve three c hildren. 1 8 4

PAGE 203

I certify that I have read this s tud y an d th at in my op1n1on it conforms t o acce pt a ble standards of scho l arly p re se ntation a nd is full y adequat e in scope and quality, as a di sserta tion for the degree of Doctor of Philosophy Clark I. Cro ss C hair ma n Associ ate Profe ss or of G eog raphy I certify that I h ave r ea d thi s study and that in m y opinion it conform s to acceptable standards of scl1olarl y presentation and i s full y adequat e in sco p e a nd quality, as a dis se rt atio n for the d eg r ee of Doctor o f Phi l osophy I c e rt i fy that J h ave r ea d this study ,:md th at in my op1n1on it conforms to acceptable standards of sc hol arly pr esenta ti o n an d i s full y adequate in scope and quality, as a dis se rtation for the degr e e of Doctor of Phi l osophy Associat e Profes s or of Geography I certif y th a t I have r ea d this st ud y and th a t in my op1111on it conform s to acceptable standards of scholarly pre se ntation and is fully adequate in scope a n d quality, a s a dissertation for the degree of Doctor of Philosophy David L N iddri e Prof ess or of G eogra phy

PAGE 204

I certify th at T h ave read this study and that in m y opinion it conforms t o acceptab l e s tand ar d s o f sc h o l ar l y presentation a nd is fully adequat e j n scope a nd qua lit y, as a d i s se rt a ti o n for the d eg r e e of Doctor of Phil osop h y \ Cl ayto n ~ Curti s A ss o cia t d Professor of R e al Estate \ / Thi s di sser tat ion wa s s ubmitt e d to th e D e partment of G e ography in the Coll ege of Arts a nd Sc i e nces and to th e Cradu a te Council and 1vas accepted as po.r t i a l fulfjllm e nt of th e requir e m e nts for th e d e g ree of Do cto r of Philo s ophy D ean, Gradu at e Sch o ol

PAGE 205

1 11111 1 i 1 l lii m !liilWl 1 1 1 11 1 3 1262 08285 178 2