Rural-urban fringe--continuum or dichotomy?


Material Information

Rural-urban fringe--continuum or dichotomy? a study of the high-growth area, Seminole County, Florida
Physical Description:
xii, 173 leaves : ill. ; 28 cm.
Richards, Storm L., 1950-
Publication Date:


Subjects / Keywords:
Cities and towns -- Growth -- Florida -- Seminole County   ( lcsh )
Urbanization -- Florida -- Seminole County   ( lcsh )
Geography thesis Ph. D
Dissertations, Academic -- Geography -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1987.
Bibliography: leaves 157-165.
Statement of Responsibility:
by Storm L. Richards.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001039154
oclc - 18327545
notis - AFC1687
System ID:

This item is only available as the following downloads:

Full Text







To Jeanne and Emerson


My gratitude must first be expressed to the members

of my supervisory committee. I am indebted particularly to

my chairman, Dr. David Niddrie, who has been supportive of

my work for many years. Most important has been his

personal concern for my academic and professional careers.

Dr. Edward Malecki, who has served as cochairman, provided

me with the direction in making the necessary revisions of

this study. For this, and for his seemingly endless

patience, I thank him. Additionally, I would like to

express my gratitude to Dr. Louis Paganini, who taught me

the value of field geography, which has been important in my

professional career. I would also like to thank Dr. Earl

Starnes for his participation in review of my dissertation.

Finally, I should like to thank Dr. William Maples, whom I

have known and respected since I first came to the

University of Florida, for his direction, his support, and,

above all, his honesty.

I wish also to express my gratitude to the

individuals who have worked with and shared my concerns in

completion of this dissertation, including Dr. Kathy Ross,

Dr. Monte Blair, Dr. John Braaksma, and Dick Haury. My


sincere appreciation goes to Sofia Kohli for editing and

word processing this dissertation.

I wish most of all to express profound gratitude to

my wife, Jeanne Fillman-Richards, who always listened,

supported, and put up with the things necessary to complete

this work.


ACKNOWLEDGMENTS...................................... .

LIST OF TABLES................................... ......

LIST OF FIGURES........................................

ABSTRACT...... .........................................


I INTRODUCTION................................

Review of Census Definitions of
Urban, Urban Fringe, and Rural.........
The Urban Fringe....................
The Extended City ...................
Rural Areas..........................
Problem Statement and Purpose of
Present Research.......................
Purpose of Research..................
Study Area ..........................
Intent of the Research..............

II LITERATURE REVIEW............................

Urban and Rural Definitions.............
United States Definitions...........
International Definitions...........
Rural-Urban Definitions.............
Urban Form and Land Use: Models
and Analyses...........................
The Three Classical Models of
Urban Form.........................
Concentric-zone model..........
Sector model...................
Multiple-nuclei model..........
Criticisms of the classical
models. .....................














Social Area Analysis and
Factorial Ecology.................... 37
Social area analysis.......... 37
Criticisms of social area
analysis..................... 38
Factorial ecology............. 39
Criticism of factorial
ecology .................... 40
Von Thunen's Theory of Agricultural
Land Use............................ 40
Criticisms of Von Thinen's Theory... 41
Alonso's Theory of Urban Land
Market.................. ...... ..... 42
Criticisms of Alonso's Theory........ 43
The Rural-Urban Fringe................. 44
Models of Land Use in the Rural-
Urban Fringe....................... 45
Empirical Studies of the Rural-
Urban Fringe.............. ......... 46
General loss of farmland....... 48
Encroachment of urban
land uses onto the
rural-urban fringe.......... 51
Land-Use Conversion within the
Rural-Urban Fringe of
High-Growth Areas ................. 53
Seminole County as a High-
Growth Area................. ..... .. 54
Conclusion.................... ..... ... .. 55

III METHODOLOGY............... .................. 58

Traffic Zones as a Standard Areal
Measurement............................ 59
History of Traffic Zones............ 59
Use of Traffic Zones in
Seminole County ................... 63
Use of the Traffic Zone in
the Present Study ......... ......... 68
Definitions of the Variables............ 71
The Roles of Distance and Density........ 74
Regression Analysis................. 74
F-Values.................. ........... 76
Z-Score Statistic.......... ......... 77
Summary................... ........ .....** 77

IV ANALYSIS.................................. .. 79

Analysis and Data Sets................... 79
Presentation of Findings................ 80

Relationships...................... 81
Single-Family Housing--
Distance Relationships............ 86
Multi-Family Housing--
Distance Relationships............. 92
Persons per Household--
Distance Relationships............ 96
Resident Employment--Distance
Relationships...................... 97
Attendant Employment--Distance
Relationships ..................... 103
Retail Employment--Distance
Relationships...................... 103
Income--Distance Relationships....... 107
Conclusions ............................. 112

V SUMMARY AND CONCLUSIONS..... .................. 117

Summary of Findings..................... 118
Conclusion.............................. 124
Future Studies of the Rural-
Urban Fringe........................... 126

APPENDIX: SOCIOECONOMIC DATA--1980 ................... 132

REFERENCES..... ........... ............................. 157

SUPPLEMENTAL BIBLIOGRAPHY............................. 166

BIOGRAPHICAL SKETCH........................ ............ 172



Table Page

1. Seminole County Population, 1930-1980........ 17

2. Sources of Comprehensive Planning Data
for Cities in Seminole County, Florida...... 64

3. Population and Distance from Node............ 82

4. Single-Family Housing and Distance
from Node....................... ............. 88

5. Multi-Family Housing and Distance from
Node....................... .................. .. 93

6. Persons per Household and Distance
from Node.................... ..... ... ........ 98

7. Resident Employment and Distance from
Node...................... ................... 100

8. Attendant Employment and Distance
from Node........ ............................ 104

9. Retail Employment and Distance from
Node....................... .................. .. 108

10. Income and Distance from Node............... Ill

11. Comparison of Means of Density
Populations................... ............ .. 115

12. Comparisons of F-Value and Significance
Levels (95-Percent Confidence Level)......... 122

13. Comparisons of Values for Incorporated
and Unincorporated Intersections............. 123



Figure Page

1. Seminole and adjacent counties............. 13

2. Counties included in the Orlando
Metropolitan Statistical Area and
the Orlando Urban Area Metropolitan
Planning Organization........................ 14

3. Model describing land use in the
rural-urban fringe ........................... 47

4. Seminole County traffic zones............... 65

5. Seminole County census tracts............... 67

6. The incorporated cities and urban
node of Seminole County.................... 70

7. Regression lines for incorporated
and unincorporated population size
planning data............................... 83

8. Regression lines for incorporated
and unincorporated population density
data ................................... ..... 84

9. Major transportation arteries in
Seminole County.................... .......... 87

10. Regression lines for incorporated
and unincorporated single-family
housing planning data................ ...... 90

11. Regression lines for incorporated
and unincorporated single-family
housing density data......................... 91

12. Regression lines for incorporated
and unincorporated multi-family
housing planning data....................... 94

Figure Page

13. Regression lines for incorporated
and unincorporated multi-family
housing density data......................... 95

14. Regression lines for incorporated
and unincorporated persons-per-
household planning data...................... 99

15. Regression lines for incorporated
and unincorporated resident
employment planning data................... 101

16. Regression lines for incorporated
and unincorporated resident
employment density data.................... 102

17. Regression lines for incorporated
and unincorporated attendant
employment planning data................... 105

18. Regression lines for incorporated
and unincorporated attendant
employment density data...................... 106

19. Regression lines for incorporated
and unincorporated retail
employment planning data.................... 109

20. Regression lines for incorporated
and unincorporated retail
employment density data............... ...... 110

21. Regression lines for incorporated
and unincorporated income dollars
planning data..................... ........... 113

22. Regression lines for incorporated
and unincorporated income dollars
density data................... .............. 114

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Storm L. Richards

December 1987

Chairman: David L. Niddrie
Cochairman: Edward J. Malecki
Major Department: Geography

Despite a growing interest in urban, rural, and

rural-urban fringe, it has been difficult to define these

terms precisely. Within the United States, the criteria set

by the Bureau of the Census are most used as definitions.

These are set out in such a way, however, that they are

often inappropriate for local decision-making. Yet

economic, political, and administrative decisions often

depend on whether or not areas are labeled "urban" or


At present, decisions, particularly in high-growth

areas such as Seminole County, Florida, are often made on

the basis of a regional data set which may or may not

reflect characteristics of rural and urban. This study

examines, through various statistical analyses, the utility

of one such data set in differentiating urban and rural

populations and areas.

The analysis describes and compares the relationship

between distance from the urban node and eight socioeconomic

variables in each of six data sets. The F-tests show that x,

distance from the node, is useful in predicting y, increase

or decrease in the socioeconomic variables in many of the


Thus, the study supports a distance-decay function

of the variables in several of the data sets. Because of

the gradient within these data sets, the study also suggests

the existence of a rural-urban fringe which is a continuum

between rural and urban areas of the county.



Unfortunately, there is no universally accepted
distinction between urban and rural.
--Murphy 1966:1

In the last ten years, a large proportion of the
theoretical work in urban geography and other
disciplines in the urban realm has been devoted to
the problem of definition.
--Sayer 1984:279

Throughout the United States, and especially in

high-growth areas such as Florida, cities are expanding

beyond their incorporation boundaries. Near these cities,

areas which had once been rural countryside, farms,

pastures, and woods are now sprouting residential,

commercial, and industrial developments. The amount and

direction of urban growth are often unprecedented and

unpredicted, and the effects of the growth on the

surrounding rural areas are poorly understood. The greatest

amount of change, however, and therefore the greatest effect

on the countryside occurs within that nebulous area often

called the rural-urban fringe.

Despite a growing interest among geographers,

planners, and anthropologists in urban, rural, and rural-

urban fringe areas, it has been difficult to define these

terms precisely. Definitions become particularly difficult

within regions which have been subjected to rapid urban

development with their residential, commercial, and office-

complex growth. Within these rapid-growth regions, the

distinctions among urban, rural, and fringe areas may be

obscure and assessments are often based on inadequate


The terms urban and rural have wide and diverse

meanings both geographically and culturally. It is not

difficult, therefore, to understand that defining these

terms is among the most problematic area of any rural-urban

fringe study. In order to determine where rural and urban

areas interface to form the rural-urban fringe, it is

clearly necessary to know what constitutes a rural area and

an urban area. As Fesenmaier et al. (1979:255-256) point

out, "There seems,to be little prospect of a more adequate

Definition of the limits of the rural-urban fringe since the

question simply begs the far greater one of defining urban

and rural."

A better knowledge and, ultimately, the definitions

of urban, rural, and rural-urban fringe require the

differentiation of characteristics, or gradations of

characteristics, which are urban or rural. The

identification, precise description, and quantification of

these characteristics, or variables, is essential, not only

to academicians studying urban areas, but also to planners

and policy-makers who are involved in urban growth and rural


In historical terms, areas designated as rural or

urban have been treated as opposites with few, if any,

characteristics in common. The reason for the urban or

rural designation may be as arbitrary as the placement of an

incorporation line. In such instances, the incorporation

boundary is analogous to the wall of a medieval "walled

city." The area not within the "wall" was not considered

part of the city even when construction and population moved

beyond the wall. It was not until the wall was moved to

encompass the newly built-up area that the rights and

protection of the city dwellers were conferred on that area.

In the United States, incorporation boundaries,

population, and, to a lesser extent, population density and

economic development, have been the major factors used by

the Bureau of the Census to define urban. A review of the

various definitions used by the Census since the turn of the

century illustrates the emphasis placed on these few factors

in the designation of urban, rural, and fringe areas. The

Census criteria are particularly important because

. most people follow the lead of the Bureau of the

Census which defines certain types of areas as urban"

(Murphy 1966:1).

Review of Census Definitions
of Urban, Urban Fringe, and Rural

The 1910 Census established the most important

element in all subsequent Census definitions of urban. In

that year all incorporated places with 2,500 or more

inhabitants were designated urban (U.S. Bureau of the Census


From 1910 until 1950, the basic definition of urban

remained unchanged. Each Census, however, also included

special rules to cover certain areas which were not

incorporated, but contained a substantial population. In

1910 and 1920, all towns (townships) in Massachusetts, Rhode

Island, and New Hampshire with populations of 2,500 or more

were included whether or not they were incorporated because

in these states incorporation is not granted until the

population reaches 10,000. In 1930, the Census special

rules stated that all townships in Massachusetts, Rhode

Island, and New Hampshire were included in the urban

designation if they had a population of 2,500 and "certain

urban characteristics" which were not specified in the

Census. The category was also expanded to include "a few

large townships in other states." Again, the criteria for

inclusion were nebulous.

In 1940, the special rules became more exacting.

Townships in the three previously mentioned New England

states were included if they contained one village with a

population of 2,500 or several villages whose combined

population was greater than 50 percent of the total

population of the township. In other states, townships, or

other political subdivisions, were included if they had a

population of 10,000 and a density greater than 1,000

persons per square mile (U.S. Bureau of the Census 1913,

1923, 1933, 1942). Thus, the 1910, 1920, 1930, and 1940

Census definitions of urban were based almost entirely on a

combination of incorporation lines and a minimum population

standard. There was only peripheral recognition of some

areas which were equally populous or densely settled, but

remained unincorporated.

By 1950, growing concern with a more precise

separation of urban from rural fostered major additions to

the basic Census definition of urban. For the first time,

the term "urban territory" was used and the concepts of

"places," "urbanized areas," and "urban fringes" were

included. These terms have to be described in detail

because they continue through the 1980 Census, with only

minor revisions, to be the basis for most urban, rural, and

urban-rural fringe determinations in the United States.

The 1950 Census defines the urban territory as

incorporated places with populations of 2,500 or more, the

urban fringe around cities with 50,000 or more inhabitants,

and unincorporated places outside of the urban fringe with

populations of 2,500 or more. Since the territories

described in the latter two categories are not legally

defined, the Bureau of the Census set up boundaries prior to

enumeration (U.S. Bureau of the Census 1952). These

boundaries were based on the concepts of "places,"

"urbanized areas," and "urban fringe" described in the


Places are concentrations "of population regardless

of legally prescribed limits, powers, or functions" (U.S.

Bureau of the Census 1962:xiv). Two types of places,

incorporated and unincorporated, are recognized by the

Bureau of the Census. Places are considered urban only when

their population meets or exceeds the minimum limit of 2,500

(U.S. Bureau of the Census 1952, 1962, 1972, 1982).

An urbanized area (UA) consists of one or more

central city or cities and the urban fringe. The central

city continues to be defined as the largest city in the

area, but exact population requirements for a central city

have changed over time. In 1950, the largest city was

required to have a population of 50,000 or more for central

city designation. The second- and third-largest cities

could also be central cities if their populations were one-

third that of the largest city and at least 25,000 (U.S.

Bureau of the Census 1952). By 1980, the minimum population

requirement for central city designation had been abandoned.

Instead, a minimum population of 50,000 for the entire

urbanized areas was substituted (U.S. Bureau of the Census


The Urban Fringe

The second element of an urbanized area is the urban

fringe which lies without, but contiguous to, the central

city or cities. The following criteria define those areas

included in the urban fringe:

1. Incorporated places with a population of

2,500 or more

2. Incorporated places with a population of

less than 2,500 if the area has at least 100

dwelling units and a density of 500 or more

units per square mile which is equivalent to

a density of 2,000 persons per square mile

(in 1980 this requirement was dropped to

1,000 persons per square mile)

3. Unincorporated areas with a density of 500

dwelling units per square mile

4. Commercial, recreational, industrial, and

other functionally related areas

5. Outlying noncontiguous areas where the

required dwelling unit density located

within 1.5 miles of the primary contiguous

urbanized portion, as measured by the

shortest connecting highway, as well as by

other outlying areas within 0.5 mile of said

noncontiguous areas "which meet the minimum

residential density rule" (U.S. Bureau of

the Census 1952:xiv).

The Extended City

In 1970, another concept, that of the "extended

city," was recognized. An extended city is a city whose

incorporation boundary has been expanded to include

substantial territory which is rural in character. Only the

urban parts of an extended city are considered the central

city if such a designation applies (U.S. Bureau of the

Census 1972, 1982).

As has been shown, recent Census definitions of

urban, which include the urban fringe, have become more

complex and more specific than previous definitions. While

there is a continued reliance on corporate boundaries and

minimum population, other factors, such as population

density and dwelling-unit density, are now considered,

especially in the definition of the urban fringe.

Rural Areas

By contrast, the Census definition of rural remains

the same as it was at the turn of the century. With one

exception, no specific characteristics have been assigned to

describe rural areas. The exception is that of an extended

city. In 1970, when the concept of the extended city was

recognized, the "Bureau of the Census examined patterns of

population density and classified a portion or portions of

each such city as rural" (U.S. Bureau of the Census 1972:v).

The rural classification required that the population

density be less than 100 individuals per square mile in an

area of at least five square miles or 25 percent of the

incorporated area (U.S. Bureau of the Census 1972). Aside

from this one instance, the Census definition of rural

simply states that any area not classified as urban is


The criteria for urban, urban fringe, and rural

which are established by the Census have been delineated

because these are the most widely accepted and utilized

definitions currently available in the United States. The

criteria also determine the various categories of Census

data. Thus, if the criteria and, therefore, the categories,

are flawed, the data are flawed. This could significantly

affect research into urban, rural, and rural-urban fringe

because virtually all of that research is based on Census


Several problems with the Census definitions and

data are apparent. Even the 1980 Census acknowledges that

the absolute lower limit for urban designation has caused

consternation among those who inhabit smaller places.

"Within small counties, measurements of urban and rural

populations over time may be significantly affected by the

increase or decrease of a place's population across the

2,500 population threshold, e.g., the increase of 1 person

to a place of 2,499 results in an increase of 2,500 to the

county's urban population" (U.S. Bureau of the Census

1982:51). By the same token, the place itself wavers back

and forth between a rural and an urban designation depending

on the 2,500th individual.

Other problems are encountered in the urban fringe

criteria for urban areas. Boundaries for the urban fringe

are based on population density and contiguity or proximity

to areas with certain population densities. For example,

industrial and office parks are included in the fringe only

if they are within densely settled areas. Since land use,

even if it would seem typically urban, is not considered, it

is possible to conceive a situation, especially in a rapidly

developing area, where large segments of land devoted to

office, research, or industrial parks are relegated to rural

status simply because they are beyond areas with certain

population density limits.

For large-scale, generalized research, the Census

criteria may be adequate. Local research and planning,

especially in rapidly developing areas, however, requires

more precise definitions of urban, rural, and rural-urban

fringe. Unfortunately, such definitions do not exist.

Problem Statement
and Purpose of Present Research

In much of the United States, development is

spilling beyond the incorporation boundaries of cities into

what were regarded as rural areas. The consequences of

rapid urbanization on the countryside are dramatic and, in

some cases, seem almost instantaneous. A recent Wall Street

Journal article (March 26, 1987) described the phenomenon in

the following terms:

From Plano, Texas, to Middlesex County, New
Jersey, from Aurora, Colorado, to Gwinnett County,
Georgia--places once considered rural and idyllic
and far from the central-city blues--the scenario is
the same. Offices and shopping centers shoot up,
subdivisions follow, and it gets harder to tell
urban from suburban from rural. Mini-cities seem to
be everywhere. (Morris 1987:1)

In high-growth areas across the country, urban

researchers, planners, and governmental decision-makers face

the problem of distinguishing urban, rural, and rural-urban

fringe areas. Such information is essential for

comprehensive plans which outline growth and development

guidelines, transportation documents that give direction to

federal, state, and local roadway plans, grants for housing,

capital improvements, property acquisition, and local

governmental policy.

Purpose of Research

Efforts to locate precisely and predict urban,

rural, and fringe areas are often hindered because the

process and characteristics of urbanization in rapidly

developing rural areas are poorly understood. The purpose,

then, of the present research is to analyze urbanization in

a rapidly developing rural area, to improve the

understanding of the characteristics associated with rapid

urbanization, and to contribute to a better definition of

the rural-urban fringe.

Study Area

The geographic focus of this study, Seminole County,

is located in Central Florida adjacent to Orange, Lake, and

Volusia Counties (see Figure 1). It is included, along with

Orange and Osceola Counties, in the Orlando Metropolitan

Statistical Area (MSA) and the Orlando Urban Area

Metropolitan Planning Organization (OUAMPO) (see Figure 2).

Seminole County was selected as the study area

because it is exhibiting several signs of rapid

Figure 1. Seminole and adjacent counties

Figure 2. Counties included in the Orlando Metropolitan
Statistical Area and the Orlando Urban Area
Metropolitan Planning Organization

urbanization. Agricultural employment and rural residence,

according to Census criteria, are dropping. Total

population and urban residence, again according to Census

criteria, are rising rapidly and population density is among

the highest in the state.

Historically, the county has been predominantly

agricultural, producing truck crops, such as celery,1

cabbage, and watercress, and the agricultural sector has

provided the major employment. In recent decades, however,

the percentage of agricultural employment has dropped

dramatically. In 1940, 3,290 individuals, or 36 percent, of

the total labor force of 9,134 were employed in agriculture

(U.S. Bureau of the Census 1942). By 1980, less than 2

percent, or 1,622, of the 82,316 employed individuals in the

county worked in the agricultural sector (U.S. Bureau of the

Census 1982).

Rural residence, based on the Census criteria for

rural, has also decreased sharply. Between 1970 and 1980,

rural residence dropped from 31,709 to 15,578 within the

county. This represents a -50.9 percent change within one

decade. By contrast, urban residence has risen from 51,983

in 1970 to 164,174 in 1980, representing a 215.8 percent

IThe nickname of the county seat, Sanford, is still
Celery City even though celery is no longer grown in the

increase in the urban population (U.S. Bureau of the Census

1982). The percentage of rural residence decrease is second

only to Pinellas County, while the percentage of urban

residence increase is the third greatest in the state.

With 603.2 persons per square mile, Seminole County

was the sixth most densely settled county in Florida in

1980. It also experienced the seventh greatest percentage

of population increase--114.8 percent, between 1970 and 1980

when the population rose from 83,692 to 179,752 (see

Table 1).

Despite the county's seemingly dense population and

high percentage of population increase, the population is

not evenly distributed. There are some sections of the

county which appear clearly urban, others which seem rural,

and still others which may not fit in either category. Yet,

the Census has separated the population, and thus the land

area, into rural and urban contingents based solely on

incorporation lines, population size, and/or population

density. These may not be the most important and diagnostic

criteria for such classifications. While the "city wall"

incorporation limits may differentiate urban from rural for

carrying out governmental policy, other definitions of urban

and rural must now be considered. A functional definition,

particularly in high-growth areas such as Central Florida,

is required. Areas well beyond incorporation limits often










o o


0 co

0 C0

-4 C0
o I

0 O

) r-




>) 0
>- &14










exhibit characteristics similar to those in the city, while

areas which appear rural according to a functional

definition and governmental policy may be viewed quite

differently by developers, who base their perspective on

profitability. The present research, therefore, utilizes

alternative types of data which may lead to a clearer

distinction between, and ultimately understanding of, urban,

rural, and rural-urban fringe areas in rapidly developing

sections of the country.

Intent of the Research

The intent, then, of the presen- research is first

to characterize the various areas of Seminole County,

Florida, using a standard socioeconomic data set derived

from a variety of sources (East Central Florida Regional

Planning Council 1984:v) and which is divided into "traffic

zones" (TZs). Such information is used by regional, county,

and municipal governments as well as planning agencies in

Central Florida. The data include population

characteristics, housing types, employment, income, and

education. (See page 71 in Chapter III.)

Secondly, through the manipulation of the standard

data set, it is possible for areas in Seminole County to be

more clearly defined and designated as rural, urban, or

rural-urban fringe. At present, for planning purposes,

areas of the county which are incorporated are simply

considered urban and those which are unincorporated are

considered rural, but these simple designations of urban and

rural are unacceptable. Even the greatly criticized Census

designations no longer depend solely on incorporation

boundaries to distinguish urban from rural.

Thirdly, areas defined as urban, rural, and rural-

urban fringe as a result of this study are mapped and

compared with areas which have been designated similarly by

the Census. Based on these comparisons, recommendations for

criteria which may more accurately depict urban, rural, and

rural-urban fringe areas are enumerated.

It may be argued that the conversion of rural lands

to urban has little to do initially with population data.

Thus, a single calculation of population change, especially

without determining which characteristics represent urban

and which represent rural, does little to explain, or

accurately describe, rural-to-urban conversion either from a

geographic or planning perspective. Yet, the most widely

accepted and used definition of urban, and, therefore,

rural-urban fringe and rural, in the United States is based

largely on the single characteristic of population. As a

result, any studies of urban expansion, rural conversion, or

rural-urban fringe development in an area rely almost

entirely on population change. Alternative characteristics


must be, and are, explored in the present study of an area

experiencing phenomenal growth.


The urban-rural boundary becomes so blurred that
the dichotomy becomes arbitrary and
essentially meaningless. No one has been able to
demarcate the boundary of a city in a consistent
--Bourne and Simmons 1978:22

[Yet, while] abstractt social theory can .
abstract from the contingencies of spatial form,
research on concrete effects must take them
into account.
--Sayer 1984:282

The urban-rural boundary has not been, and may never

be, demarcated in a manner which is satisfactory for many

types of urban and rural research. While the Bureau of the

Census definitions and criteria are used for much of the

urban and rural research conducted in the United States,

these designations are not adequate for local research and

planning. They are also unacceptable for studies in other

countries which depend on their own census definitions of

urban and rural.

Despite the fact that there is no universally

accepted differentiation of rural and urban, many

disciplines, including geography, sociology, planning, and

anthropology, have developed subfields of study based on an

urban-rural dichotomy. Each subfield has its own


considerable and specialized body of literature.

Nevertheless, because the boundaries between urban and rural

are often ambiguous and contrived, areas of research within

the separate subfields occasionally overlap. This is

particularly true in the case of rural-urban fringe studies.

The literature reviewed for the present study is

drawn mainly from urban and rural geography; however,

sources from other, closely related disciplines are also

discussed. Sociology, anthropology, and planning each have

urban and rural contingents in which problems of definition

similar to those encountered within geography must be

confronted. The literature review, then, is set out,

regardless of discipline, within the broad themes of urban

and rural.

Urban and Rural Definitions

One technical aspect that causes some concern is
the actual delimitation of geographic areas that can
be regarded as urban. Such delimitation is
necessary, particularly in studies focusing on .
rural-urban land conversion.
--Yeates 1987:64

Because there is no consensus among researchers on

the exact definition of urban, it is difficult to pinpoint

areas, particularly peripheral areas, which are, or may

become, urban. For some urban investigators, this

difficulty in defining specific areas as urban has become a

moot question. In fact, it has been suggested "that the

concept of a separate urban phenomenon, defined spatially,

is irrelevant to understanding society" (Johnston 1986:103).

To bolster this position, the researchers argue that, in

more developed countries, factors such as mass media,

migration, and transportation networks have spread the

"urban way of life" to such an extent that rural-urban

distinctions can no longer be studied (Dunleavy 1982, Lang

1986). Others, however, contend that there are urban-rural

distinctions and issues which may be studied only through

the bounding of discrete areas into urban and rural (Sayer

1984, Johnston 1986, Lang 1986, Yeates 1987).

If the position that urban areas must be delimited

in order to address urban issues and questions adequately is

accepted, then the problem of urban definition must be

confronted and specific characteristics of urban must be

explored. In this context, several characteristics

attributed to urban areas are commonly used to discriminate

between urban and rural places. One of the most widely used

criteria is "high" population density (Cadwallader 1985,

Yeates 1987). The determination of exactly what level of

density constitutes "high," however, falls upon the

individual researcher (Yeates 1987). Another popularly

employed characteristic is nonagrarian-related occupations

(Cadwallader 1985, Johnston 1986, Lang 1986). Johnston

(1986) used this criterion in his study of political

attitudes and location by designating British constituencies

with less than 3 percent agricultural employment as urban.

United States Definitions

Within the United States, the criteria set by the

Bureau of the Census are most often used to define urban and

rural areas (Murphy 1966, Lang 1986). The major factors

employed by the Census to determine urban areas are

population size, incorporation boundaries, and, to a lesser

extent, population density and economic development. The

major factor determining designation of rural areas is that

the areas are not urban by census definition.1

Unfortunately, the census criteria and, therefore,

the census definitions, are not always sufficient. In a

recent study of the adequacy of census definitions, Lang

found that policyiy makers and census data users across the

United States are calling for revisions of the definitions

or urban and rural population and places currently used by

the U.S. Census Bureau. They contend that the definitions

are outdated and do not reflect the current realities of

population distribution, lifestyles, and settlement

patterns" (Lang 1986:118). Despite these criteria, the

census data remain at present the most comprehensive and

1Census definitions of urban and rural have been
discussed in greater detail in Chapter I, pages 4-11.

immediately available source of urban and rural information.

Therefore, whether or not they are based on adequate

definitions, these data will continue to be used for much of

the recent research in the United States.

International Definitions

If comprehensive definitions of urban and rural for

the United States have proven elusive, such definitions at

an international level have been impossible. Numerous

problems have been encountered which negate an international

consensus of criteria for urban and rural. In most cases,

comparable data are simply not available. Many countries

lack the skills necessary to collect detailed information,

while others do not deem detailed census data a national

priority either for economic or cultural reasons. Also,

where the data are available, there is often inconsistency

of nomenclature and of geographic bounding of urban areas

between countries. The result is a "bewildering variety of

definitions of 'urban' and 'rural'" (United Nations

Secretariat 1975:18).

By describing just one of the problems, that of data

comparability, it becomes easy to understand why the United

Nations has allowed that standard international definitions

of urban and rural are not, and may not be, possible (Lang

1986). The problem of comparable data is illustrated in two

ways. The first is a discussion of which general census

criteria are used to establish urban areas in different

countries and the second is to describe how the criteria are

interpreted and applied to the delimitation of urban areas

in specific, selected countries around the world.

The ways in which different countries determine

urban areas range from simple to complex to bewildering. In

a survey of the characteristics used in various national

censuses to indicate urban areas, the United Nations

Secretariat (1975) found great diversity. Some countries

used only one characteristic, others used two or more, and

still others (56) did not indicate what they used.

Population size, the most common criterion, was employed

singly in 23 cases and in combination with other criteria in

49. Housing or population density, while used as the

indicator in only one census, was combined with other

characteristics in 11 censuses. Major economic activity was

also used as the criterion in only one instance, but was

used in concert with others in eight censuses. Type or

structure of government appeared alone three times and

combined with other characteristics three times. Finally,

various urban indicators not included in the other

categories were used by themselves in three censuses and in

combination with other characteristics in 16 others (United

Nations Secretariat 1975:18-19).

The astonishing array of characteristics and

combinations of characteristics used to define urban in the

various censuses is only part of the problem faced by the

urban researcher who is trying to compare international

data. Because each country sets its own values for, and

interpretations of, the criteria it chooses, comparisons

become even more difficult. The following descriptions of

census criteria for urban and rural in several countries

illustrate these points.

In some countries, designations of urban and rural

are based on concepts similar to the "places" designations

in the U.S. Census. Legal boundaries and powers are

disregarded and concentrations of population are used

instead. In Ghana, population size of places is the major

determinant of urban and rural. Concentrations of 5,000 or

more inhabitants are considered urban. The Swedish urban

and rural designation system combines population size with

housing density. Places with 200 or more residents in

dwellings that are 200 or less meters apart are considered

urban (Wander 1975). The Canadian Census is similar to the

U.S. Census in that it uses both population size and

population density as urban designators. Areas with a

population of 1,000 or more and a density of at least 400

per square kilometer are considered urban (Yeates 1987).

By contrast, some countries require official

bounding for urban designation. Argentina, for example,

considers an area urban if its population is 2,000 or

greater and it has official boundaries. All other areas,

regardless of population concentration, are rural (Wander

1975). India classifies legally recognized cities and towns

with populations of 5,000 or more as urban if they also have

a population density of 1,000 per square mile and three-

quarters of those employed work in other than the

agricultural sector of the economy (Bose 1975, Wander 1975,

Jordan and Rowntree 1982).

Still other countries define urban areas "in terms

of minor civil divisions which have fixed boundaries

and some local government status" (Wander 1975). Until the

1960 Census, only "shi," cities with populations of 30,000

or more, were recognized as urban in Japan. The rest of the

country was considered rural, or "gun." However, as in

Extended Cities in the U.S. Census, "shi" often contained

extensive areas which might better be classified as rural so

the Japanese Bureau of Statistics devised the concept of

Densely Inhabited District (DID). A DID is defined as any

area of contiguous enumeration districts with a population

of 5,000 or more and a density of at least 4,000 persons per

square kilometer. The grouping of enumeration districts

must also be within the boundaries of a "shi," "machi"

(town), or "mura" (village) (Wander 1975, Yamaguchi 1984).

The Dutch system divides municipalities into urban,

urbanized rural, and rural using a variety of

characteristics. Urban municipalities are those with

overall population densities of 300 persons per square

kilometer. These areas must contain at least one nucleus in

which 70 percent of the area's population (a minimum of

2,000 individuals) lives. This settlement must also have a

population density of 2,000 per square kilometer and at

least 90 percent of the male population must be

nonagriculturally employed. In urbanized rural areas, these

characteristics are considered. The most important is that

less than 20 percent of the male population is employed in

agriculture. Also, the major settled area has a population

less than 20,000. The third characteristic employed is

commuter patterns. Rural municipalities are those with more

than 20 percent of the male population employed in the

agricultural sector. These areas have less than 5,000

inhabitants in the major settlement and less than 300

persons per square kilometer overall density (Wander 1975,

Borchert 1984).

While only a few, selected examples of the criteria

used in different countries to determine urban or rural

status of areas have been described, it is apparent that

cross-cultural comparisons of urban and rural data are

difficult at best. It is also evident that none of the

combinations of criteria, even the relatively complex Dutch

system, is adequate for differentiating urban, rural, and

rural-urban fringe at the local level.

Rural-Urban Definitions

The most recent U.S. Census definition (1980) (see

Chapter I, pp. 8-9, of this study) of the rural-urban

fringe, also known as the "urban fringe" or the urbann

fringe" (Champion 1983:30), differs only slightly from the

earliest academic descriptions. In 1937, T. L. Smith

described the rural-urban fringe as "the built-up area just

outside the corporate limits of the city" (Pryor 1971:59).

In an article entitled "Urban-Rural Fringe" written

in 1942, Wehrwein outlined the inevitable growth which must

occur when there is an absence of land-use control within

the zone of transition which lies between areas of well-

recognized urban development and lands devoted to

agriculture. In his summary, Wehrwein predicted that the

availability of transportation, cheap land, lower taxes, and

fewer land-use controls in rural areas would become the

incentives for industrial relocation from the city to these

transition areas (Wehrwein 1942).

In 1946, Rodehauer defined the rural-urban fringe as

"that area in which the land is utilized in an urban manner,

while at the same time certain attributes of the rural area

are present" (Rodehauer 1946:50). This general description

continues to be quoted in the literature (Fesenmaier et al.


Russwurm's study (cited in Fesenmaier et al. 1979)

differentiated a rural-urban fringe based on density of

population and the presence or absence of farming. He

specified that 50 percent or more of the fringe population

should be nonfarming individuals and that in the large

Canadian urban areas with populations of 100,000 or more,

the fringe should be more than 10 miles wide.

During the same year, 1971, Pryor suggested the

following detailed definition:

The rural-urban fringe is the zone of transition in
land use, social and demographic characteristics, lying
between (a) the continuously built-up urban and suburban
areas of the central city, and (b) the rural hinterland,
characterized by the almost complete absence of nonfarm
dwellings, occupations and land use, and of urban and
rural social orientation; an incomplete range and
penetration of urban utility services; uncoordinated
zoning or planning regulations; areal extension beyond
although contiguous with the political boundary of the
central city; and an actual and potential increase in
population density, with the current density above that
of surrounding rural districts but lower than the
central city. These characteristics may differ both
zonally and sectorally, and will be modified through
time. (Pryor 1971:62)

After elaborating on his definition of the fringe

area, Pryor states that the characteristics he describes are

not based on empirical evidence. He then calls for further

investigation by other researchers.

Urban Form and Land Use:
Models and Analyses

S. geographers have been chiefly associated with
the description and analysis of urban form. What
theory there is of urban form in geography, however,
is largely derived from other fields. One is urban
sociology, especially the human ecology school of
Chicago (Park, Burgess, and others). Another is
land economics .
--Agnew, Mercer, and Sopher 1984:12

In their pursuit of generalizations and laws which

explain and predict urban form and urban and rural land use,

geographers have used a number of models and analyses.

Certain of these have had considerable impact on geographic

thought and continue to be discussed and criticized in the

literature. They include the classical models of urban

form: concentric zone, sector, and multiple nuclei; social

area analysis and factorial ecology; and von Thunen's

agricultural land-use theory and Alonso's land-rent theory.

While none of these studies has dealt directly with the

problem of urban and rural definitions, each has contributed

to a better understanding of the characteristics of urban or

rural areas.

The Three Classical Models of Urban Form

The three classical models of urban form--concentric

zone, sector, and multiple nuclei--were developed in

response to the desire to understand the type and direction

of urban growth (Light 1983). While all three were based on

the ecological theory, each presents a different description

of city form (Jordan and Rowntree 1982, Light 1983).

Concentric-zone model

Burgess (1925), University of Chicago sociologist,

advanced the concept of concentric zones for urban

development. His theory suggested that urban land use and

growth could be described in terms of five concentric, and

internally homogeneous, zones. At the center of his model

is the central business district, the core of the city.

Surrounding the central business district is the area he

termed the "zone of transition." The urban characteristics

of this area include a mixture of land uses and diversity of

residential areas, often in a state of deterioration. The

third zone consists largely of "blue-collar" homes. The

fourth zone is another residential area catering mainly to

"white-collar" workers and middle-class families. The

outermost zone, the commuter zone, lies outside the

corporate boundary and contains higher-income residential

areas (Palm 1981, Jordan and Rowntree 1982, Light 1983,

Cadwallader 1985). The fifth zone probably corresponds with

what has been called the rural-urban fringe, although

Burgess did not explicitly refer to it.

Sector model

In 1939, approximately 10 years after Burgess

published his work on the morphology of the city, Hoyt

provided fresh insights into the patterning of residential

land use. He postulated that residential land uses "tend to

be arranged in wedges or sectors radiating from the center

of the city along the lines of transportation" (Murphy

1966:211). These sectors were divisible and Hoyt elaborated

on both their physical composition and evolution. He stated

that the gradient or outward progression of residential

properties moved "downward" from high-rent2 areas to less

affluent sectors. The "low"-rent areas occupy different

sectors which may occur in a variety of urban areas (Hoyt

1939, Palm 1981, Jordan and Rowntree 1982, Light 1983).

The dynamics of the sector concept are associated

with high-rent areas. According to Hoyt, high-price areas

exert an influence on the direction of residential growth

2Rent is defined as monies paid in the form of
purchase or lease for occupying space (Jordan and Rowntree

and real-estate agents may influence the direction of high-

rent growth (Palm 1981).

Multiple-nuclei model

The concept of the multiple-nuclei forms of the city

evolved from studies by R. D. McKenzie (1939), author of The

Metropolitan Community. It was elaborated by Harris and

Ullman in 1945. In this model, a separate function for each

nucleus from one metropolitan area to another is specified.

The relationship of urban ares is derived from a clear

distinction between the central business district and other

identifiable areas (Palm 1981, Jordan and Rowntree 1982,

Light 1983).

Harris and Ullman identified four factors which they

believed caused the emergence of separate nuclei in urban

land use:

1. The interdependence of certain types of

activities and their need to be located in

close proximity

2. A natural clustering tendency among certain

types of activities

3. A repetition of certain types of activities,


4. The related factor of high rents or high

land costs which can be an attractive or

repelling force

Criticisms of the classical models

The classical models have been the subjects of much

criticism. Both the theoretical underpinnings of the models

and the models themselves have come under fire.

Urban ecology, the theory which is the basis for the

concentric ring, sector, and multiple-nuclei models,

describes a process of invasion-succession land-use change

in which the mechanism is economic competition (Light 1983).

Critics suggest that there is too much emphasis on

mechanistic causal forces (Ley 1983). Unlike competition in

the animal world, human actions are subject to laws,

institutions, and conventions (Cadwallader 1985). The

theory also fails to account for sentiment and social values

which may be important in the determination of land use

(Murphy 1966, Ley 1983, Light 1983).

Other criticisms are directed at the models. It is

charged that they are too simplistic and lack universality

(Ley 1983). They exaggerate rigid land-use boundaries which

cannot be proven empirically (Murphy 1966, Ley 1983). Also,

the homogeneity of the various zones and sectors has been

questioned (Ley 1983). "But the very fact that controversy

continues and the debate remains current underlines [their]

immense value .. .in an initial examination of urban

social areas" (Ley 1983:74).

Social Area Analysis and Factorial Ecology

Unlike the classical models, social area analysis

and factorial ecology do not rely on the assumption that

distance from the center of the city is a major factor in

residential differences. Instead, these analyses are based

on the idea that residential areas within a city can be

grouped on the grounds of demographic and social

characteristics (Palm 1981).

Social area analysis

Social area analysis as a method of social

differentiation of urban areas was developed by Shevky and

Bell during the early 1950s (Cutter 1985). The analysis

groups areas of like socioeconomic status based on three

constructs: economic status (social rank); family status

(urbanization); and ethnic status (segregation). Using

various criteria for each theme, census tracts are

classified and grouped into social areas (Palm 1981, Ley

1983, Cadwallader 1985, Cutter 1985). Shevky and Bell

(1955), thus, provided a scheme for the analysis of changes

in social areas based on statistical differences between

census tracts. The results of their studies provided

information about what actually existed in the city rather

than preconceived notions derived from earlier descriptive


Criticisms of social area analysis

Much of the criticism of social area analysis hinges

on its theoretical background. The theory espoused by

Shevky and Bell suggests that, since society produces the

city, any understanding of the social forms of the city must

be "within the context of the changing character of the

larger containing society" (Shevky and Bell 1955:3). The

three constructs represent broad characterizations of that

society (Ley 1983).

The theory does little to explain differences and

similarities between residential areas (Cadwallader 1985).

Nor, does it address the understanding of the "process of

residential or social patterning of a city" (Cutter

1985:21). It also does not justify the use of the three

particular constructs (Palm 1981, Cadwallader 1985).

The methodology is described as unsophisticated and

the selection of variables which are used for the constructs

has been questioned. It is also suggested that the

variables do not test the empirical validity of the

constructs (Palm 1981, Cadwallader 1985).

Factorial ecology

As a reaction to the criticisms of social area

analysis, the factorial ecology methodology was developed in

the 1960s (Cutter 1985). This analytical process differed

from earlier types of analyses in the number of variables

used and its greater emphasis on spatial patterns associated

with the variables. Using the statistical technique, factor

analysis, a large number of variables is distilled into

"factors" which describe patterns of correlation among the

data (Palm 1981). The factors are then used to cluster

similar census tracts (Ley 1983). Analysis of geographic

distribution of census tracts attempts to explain patterning

in terms of existing conditions rather than accepting

preconceived model predictions.

Because factorial ecology studies have focused on

predominantly urban residential areas, such studies have

contributed little to urban-rural differentiation. The

approach of wide-ranging empirical analysis, however, is

applicable to rural-urban fringe areas as in the study by

Fesenmaier et al. (1979), in which urbanizing areas were

divided into subzones on the basis of statistical analysis,


Criticism of factorial ecology

While factorial ecology is widely used and

considered a "highly refined and technically elegant method

of clustering like census tracts" (Palm 1981), it has not

escaped criticism. It has been argued that the technique

fails to identify processes that cause residential

differentiation so that it is of limited use because it only

describes the patterns (Cadwallader 1985). The validity of

the variables, and thus the factors, in discriminating

census patterns has also been questioned (Ley 1983,

Cadwallader 1985). It suggested that the census tract, as

the geographic unit, is too large and too diverse to be of

value (Palm 1981, Cadwallader 1985). Additionally, some

researchers have shown that by using several variations of

factor analysis, different results are obtained for the same

areas (Ley 1983, Cadwallader 1985). None of these

criticisms, however, negate the value of using wide-ranging

empirical studies in describing and differentiating discrete

geographic areas.

Von Thunen's Theory of
Agricultural Land Use

While von Thunen's well-known model is an

agricultural land-use model, and might more appropriately be

discussed in a rural section of the literature review, it is

the basis for the neo-classical urban land-rent models. It

is, therefore, included briefly in this section as a

background for Alonso's classic theory of urban land rent

(Alonso 1971, 1983).

In the early nineteenth century, von Thunen

developed a model of agricultural land use based on his

observations in and around his estate in Germany. He

believed that agricultural land-use patterns were determined

by the value of the land and the crops produced for market

and the distance to that market. Based on a simplifying set

of assumptions including an "isolated" and "uniform" plane

and "economic man," he described the processes that caused

the patterns (Palm 1981). Competition between different

types of land use was controlled by "Economic Rent, defined

. as return from investment in the land" (Sinclair

1967:73). Transportation costs which rose with distance

from the city were important in determining rent. The

further the land was from the city, the more the farmer

would have to invest in transporting his product to market.

Therefore, to receive a reasonable return on his land, the

farmer had to invest in products which would either cost

less to transport or to produce (Sinclair 1967, Palm 1981).

Criticisms of Von Thunen's Theory

While the patterns and processes of von Thunen's

agricultural land-use model may still occur in

lesser-developed countries, the utility of the model in

developed areas is questioned. In such areas, new modes of

transportation, declines in cost of transportation, and the

advent of refrigeration have lessened the effect of

transportation on siting of agricultural production.

Markets have expanded from the single central market to

regional, national, or international markets. Production

techniques favor large-scale agricultural enterprises.

Finally, and probably most important, land used for urban

purposes has become far more valuable than land used for

agriculture. Thus, the competition, especially in rapidly

developing areas, is no longer between different

agricultural uses, but between agricultural and urban uses

(Sinclair 1967).

Alonso's Theory of Urban Land Market

The foundations of the formal spatial analysis
of agricultural rent and location are found in the
work of J. von Thunen, who said, without going into
detail, that the urban land market operated under
the same principles.
--Alonso 1983:1

Based on von Thunen's ideas of economic rent and

land use, Alonso developed a model of urban land use. He

began with a set of assumptions similar to those of von

Thtnen. The physical environment of the urban area is

homogenous so that no area has any particular physical

advantage. There is only one central business district in

the urban area. Transportation is easily available and

costs increase with distances from the urban center. Also,

the land, which is sold for maximum profit, may be developed

in any manner, unencumbered by governmental, environmental,

or zoning restrictions (Dennis and Clout 1980, Palm 1981,

Cadwallader 1985). By using these assumptions, Alonso

calculated bid-rent curves and isolated the bid-rent

function involved in the urban land market (Palm 1981).

"Thus, each plot of land is sold to the highest and best

use: highest in the sense of being the highest bidder, and

best in the sense of being the type of land use that is best

able to take economic advantage of that particular plot"

(Cadwallader 1985:35-36).

One important concept relating to urban land use is

derived from Alonso's model when it is applied to

residential location. THe so-called "distance-decay

function," that is, decrease with distance from the center

of the urban area, should describe both land values and

population density of the urban area (Dennis and Clout


Criticisms of Alonso's Theory

Alonso's theory has not escaped criticism. Several

of the problems are related to applicability of the

underlying assumptions to urban areas, especially in

developed countries. Few urban areas now have one central

area utilized by all inhabitants (Palm 1981). By assuming a

totally free market and optimum usage of land, the model

fails to address a modern fact of life. Governmental

restrictions often preclude "highest" and "best" usage

(Dennis and Clout 1980). Nor does the model take into

account the involvement of large development and real-estate

corporations which undoubtedly influence the urban land

market (Cadwallader 1985). In addition, as with all models

which assume "economic man," there is the objection that man

is never "all knowing" and rarely acts in the most rational

manner (Palm 1981).

Despite its simplicity, or because of it, as Palm

(1981) suggests, Alonso's model persists. Even some of its

most determined critics, Dennis and Clout, admit that "the

model has proven remarkably, perhaps uncomfortably,

successful in predicting patterns of land use and population

distribution" (1980:102), even though, as Thrall points out,

it does not explain "the underlying forces for behavior of

the system" (1987:9).

The Rural-Urban Fringe

What has become known as the rural-urban fringe has

been central to few urban geographic models and studies.

Emphasis instead has been on the central district (CBD) and

industrial, residential, or commercial areas and on suburban

development and does not include urban expansion into rural

lands and the resulting rural land-use changes. Early

studies of suburban development, which would have been

likely to include rural and agricultural conversion,

probably did not because the development was not far enough

removed form the city to compete with rural land uses.

Within the past two decades, however, the rural-urban fringe

has received more attention in the literature.

Models of Land Use in
the Rural-Urban Fringe

The urbanization of rural lands and the influence of

urban development on nearby rural areas in "high-growth"

areas were discussed by Sinclair in his seminal article "Von

Thunen and Urban Sprawl" (1967). After outlining von

Thunen's model, Sinclair said that while it had been

appropriate in the past and continued to apply to less-

developed areas, it no longer describes processes occurring

in areas of rapid urban expansion. He suggested that there

is still a decline in rent with distance from the urban

area, but that this is not related to the urban market as it

was in von Thunen's model. New factors now influence

patterns of agricultural land use near cities and the most

important of these is what has been called "the anticipation

factor," which is defined as the effects of the perception

of encroaching urban development. He stated that,

As the urbanized area is approached from a distance, the
degree of anticipation of urbanization increases. As
this happens, the ratio of urban to rural land values
increases. Hence, although the absolute value of the
land increases, the relative value for the agricultural
utilization decreases. (1967:78)

Another model describing land use in the rural-urban

fringe was set out by Pryor (1971) (see Figure 3).

Described as a "process-response model," the model shows

relationships between percentage of urban and rural land use

and distance from areas which are 100 percent urban at one

extreme and 100 percent rural at the other. The process

involved is urbanization and the response is land-use

change. Pryor (1971:62) suggested that this model be used

in conjunction with other models to study urban encroachment

into rural areas.

Empirical Studies of
the Rural-Urban Fringe

Urban encroachment, ingress of urban influences to

areas which have been traditionally rural, is a major theme

of empirical studies of the rural-urban fringe. Such

studies focus on land-use change, particularly loss of farm

lands and encroachment of urban land uses (Furuseth and

Pierce 1982, Champion 1983).

Percentage Distance Urban to Rural
3 25 50 75 1C

x Boundary of Totally Urban
y Boundary of Totally Rural

Adapted from Pryor 1971

Figure 3. Model describing land use in the rural-urban

General loss of farmland

Geographers, using both qualitative and quantitative

analyses, have tried to explain the decline of the

countryside. While specific studies are varied in depth and

focus, a recurring theme is the loss of farmland (Bryant

1976, 1981, Rodd 1976, Troughton 1976, White and Silverwood


The reasons for loss of farmlands are complex and

numerous and range from economic to political to

environmental (Johnson 1970, Lindeman 1976, Brown and

Roberts 1978, Brown et al. 1981, Healy 1985, Koenig 1985).

The types and importance of factors which cause decline of

the countryside vary from region to region and from country

to country (Noble 1962, Patterson 1968, Bryant and Russwurm

1979, Bryant et al. 1981, Warren and Rump 1981, Crewson and

Reed 1982). In some cases, land-use change results from

expansion of urban land uses into the rural-urban fringe;

however, this is not the only factor.

To examine agricultural land-use change in an area

less subject to urban influences, Crewson and Reed (1982)

studied an area in south-central Ontario which was well away

from large urban areas. Six independent variables including

farm-capitalization, size of farms, occurrence of part-time

farming, the extent of developable shoreline, and the age of

the farm operators were analyzed for a period of 20 years.

The data show the "best predictor" of loss of farmlands,

based on statistical analysis, was the frequency of part-

time farmers. The fact that large-scale farms were not

being managed and operated by full-time farmers appeared, in

this area, to be most indicative of the decline of the rural

farming business. Crewson and Reed (1982:359) conclude that

"(t]he percentage of part-time farmers is increasing

steadily because off-farm income is vital to the maintenance

of the economic viability of farming today." Thus, the loss

of farmland in this particular study area is attributed

ultimately to an economic factor.

Certainly economic factors were important in Hart's

study of rural 'and-use change in the southeastern United

States (Hart 1980). In describing the demise of a major

cash crop--cotton--Hart outlined the evolution of once-

great agricultural holdings in the Piedmont counties of

Georgia and South Carolina. A pronounced change in land use

occurred between 1939 and 1974. During that period, nearly

4-1/2 million acres of productive farmland and 1-1/3 million

acres of cleared land were sold by farmers to nonfarmers

(Hart,1980:492). One reason for the transfers of ownership,

particularly those which involved subdivision for

development, was demonstrated in the answers Hart received

when he asked about the value of land in Carroll County,

Georgia. He was told that "old, worn-out cotton land was

worth about $150 to $200 an acre for forestry. Farmers were

paying $600 to $700 an acre for such lands. Realtors .

[said] that the price of five-acre lots in the northeastern

half of the county--the side toward Atlanta--begin at around

$7,000 ." (Hart 1980:525).

A national study of farmland loss conducted by the

U.S. Department of Agriculture (Zeimetz et al. 1976) found

that, in general, factors other than urbanization accounted

for the majority of cropland loss. In particular, marginal

croplands were converted to pasture and other agricultural

lands were idled as new technology made farming of those

areas uneconomical (Zeimetz et al. 1976:ii).

Plaut (1976) states that many researchers agree that

the direct conversion of agricultural land to urban uses

does not have a significant impact on amount of land devoted

to farming at a national level. Despite this commonly held

belief, it is well to keep in mind Hart's observation (1980)

that land which is fallow may be returned to production at

some time in the future, but that subdivision and

development of rural lands results in irreversible change.

There are indications that such change is occurring in the

rural-urban fringe of high-growth areas.

Encroachment of urban land uses
onto the rural-urban fringe

While direct conversion of rural lands to urban

purposes is considered of little consequence to the national

agricultural production of the United States, such

conversion along with other effects of urbanization may

produce considerable change in land use within the rural-

urban fringe of rapidly developing urban areas (Plaut 1976).

In fact, urbanization is often considered the dominant

process by which change occurs within the rural-urban fringe

(Plaut 1976, Champion 1983, Everitt 1984).

In 1976, Plaut explored the idea that "urbanization

[has] a substantial impact on the loss of farmland on the

rural-urban fringe" (1976:27). Using regression analysis,

he analyzed the changes in land use in Standard Metropolitan

Statistical Areas in the United States. The variables

included in the analysis were number of housing units built

in the county between 1960 and 1970, class of soil

productivity, age of the farmer, and productivity of

farmlands. Based on this analysis, he determined that there

were strong relationships between land-use change and

urbanization in the rural-urban fringe areas of the Midwest

and Northeast (Plaut 1976).

Another author (Pierce 1981) investigated the rate

of conversion of rural land to urban uses within the

rural-urban fringe of Canadian cities. Seven variables,

"population change, dominant economic function, city size,

average residential land values, population density,

geographic region, and agricultural capability of the land"

(Pierce 1981:164), were used. Data, gleaned from various

sources, were subjected to several statistical analyses to

measure both collective and relative impacts of the

variables on urban expansion. Only three variables,

population change, economic function, and agricultural

capability, proved useful. Pierce states that the large

degree of unexplained variation results from the complexity

of the problem and the lack of "more precise measures of

urban form and process and land-use data ."


Zeimetz et al. (1976) also analyzed rate of

conversion of rural lands in fringe areas. In their study

of 53 rapidly developing counties throughout the United

States, they found that, while there was considerable

variation between areas, an average of 0.173 acres of rural

land was converted for each person increase in population.

They also found that the overall rate of conversion of land

to urban uses in the 53 counties between 1960 and 1970 was

3.4 percent, which they did not consider high (Zeimetz et

al. 1976). However, even though the Department of

Agriculture does not consider the rate of land-use

conversion "high," the results of the conversion are

receiving considerable attention, especially in high-growth


Land-Use Conversion within the
Rural-Urban Fringe of High-Growth Areas

The Sunday real-estate ads make Fairfax County
sound like one big hunt club. .But,
increasingly, chrome and glass office towers sprout
from northern Virginia woods.
--Morris 1987:1

In high-growth areas across the country, more and

more attention is focused on the movement of urban land uses

and population from the urban area into the rural-urban

fringe. Newspapers and magazine articles describe and

lament the passing of the countryside and planners,

politicians, businessmen, and "just ordinary" people discuss

the positive and negative affects of the changes. Some

suggest that their quality of life is being damaged (Morris

1987), while others see great business opportunities

(Leinberger 1987).

A major contributor to the movement of population

and business into the fringe areas has been the growth of

the service and information industry (Brotchie et al. 1985,

Institute of Traffic Engineers 1985, Leinberger 1987, Morris

1987). This sector of the economy does not require a

central city location and businesses are free to move to

cheaper, more attractive locations away from the urban area.

As the jobs relocate, so do the employees. In some areas,

the relocations of these businesses have produced "suburban

megacenters" which rival cities in employment and size. The

Coastal Corridor in West Los Angeles with 40 million square

feet of office space and 186,000 daytime employees and City

Post Oak near Houston with 3.3 million square feet of retail

and 16 million square feet of office space employing 60,000

people are but two examples (Institute of Traffic Engineers

1985). In other cases, whole new "urban villages," low-

density housing surrounding a central core of businesses,

appear in the midst of previously rural land (Long 1987).

Near St. Louis, Chesterfield Village grew out of a cow

pasture and Fair Lakes, south of Washington, D.C., appeared

on fallow farmland (Leinberger 1987).

Seminole County as a
High-Growth Area

Similar types of urban expansion are occurring in

Seminole County, Florida. A recent Department of

Transportation study of satellite data showed that the

county had lost 21.63 percent of its agricultural land

between 1973 and 1984. This loss in Seminole County was the

fifth greatest percentage loss among the 67 counties in the

state (Jean 1987).

An average of 1,000 persons per month move into the

county. Most of these are white-collar workers with

relatively high incomes. In fact, the median family income,

$20,873, was the highest in the state in 1980 (Kemp 1985).

Most of the new construction is in smaller

developments located generally in the southern part of the

county. Two larger "urban villages," however, are in the

first phases of growth. Earlier this year, 1987, South

Seminole Corporation paid $12 million for a 783-acre tract

of long-idle rural land in south central Seminole County.

The area, Alafaya, will contain 3,835 homes and 65 acres of

office and retail space (Snyder 1984).

The second urban village, Heathrow, adjacent to

Interstate-4 in the northwest quadrant of the county, covers

1,248 rural acres. In addition to 4,000 houses, the

development will have a shopping center and several hundred

thousand square feet of office space (Snyder 1984). This

development has already attracted the national headquarters

of the American Automobile Association and others are

expected to follow.


There exists at present no satisfactory definition

of either rural or urban. In the United States, the

definitions set out by the Bureau of the Census are most

frequently used, but other countries define rural and urban

by their own standards so that definitions are rarely

compatible. Since data are collected based on the various

census definitions, it is difficult to make cross-cultural

comparison of the data. Even within the United States, the

census definitions and data are often not adequate,

especially for studies at the local level where finer

distinctions are required. Without adequate definitions of

urban and rural, a definition of the rural-urban fringe is

difficult, although several have been advanced.

Various models and analyses have contributed to the

understanding of the characteristics of urban, rural, or

fringe areas. None of the models, however, address the

problems of urban and rural differentiation.

Empirical studies of the rural-urban fringe

concentrate on loss of farmlands and encroachment of urban

land uses. While these changes seem to have little effect

on land use at a national level, they receive considerable

attention at the local level.

As urban land uses and population move into the

countryside in rapidly developing areas, change becomes more

evident. Depending on the point of view of the observer,

the changes may appear positive or negative. They may also

seem subtle or dramatic. In other words, the amount and

direction of change depends oftentimes on one's perception

and on whether or not one will benefit or lose as a result

of the change.

Attempts to measure change at the local level are

often stymied by lack of definition, data, and technique.

Census definitions are vague. Census data based on those

definitions are not available in the format necessary for

detailed local planning. For example, while detailed data

in census block form are available in urbanized areas, only

census tract data, which encompass a much larger areal unit

of measurement, exist for nonurban areas. Thus, if

comparisons are to be made using census data, only the

larger census tract material can be used. Technique depends

on time, money, data, and knowledge available to the

investigation. Local planning efforts often suffer as the

result of these limitations.

Planning staffs in rapidly growing Seminole County

have attempted to overcome data discrepancies by using

Regional Planning Council (RPC) data which are divided into

traffic zone units instead of census tracts or blocks. The

present study of urban, rural, and rural-urban fringe in

Seminole County will also use RPC traffic zone data, both in

original form and with modifications.


The review of literature in Chapter II shows that

current definitions of urban, rural, and rural-urban fringe

are inadequate and many of the models and analyses upon

which current research is based do not address the

fundamental issue of distinguishing characteristics of rural

and urban. Yet economic, political, and administrative

decisions often depend on whether areas are "defined" as

rural or urban.

At present, decisions, particularly in high-growth

areas such-as Seminole County, are often made on the basis

of a regional data set which may or may not reflect

characteristics of rural and urban. This study examines,

through various statistical analyses, the utility of one

such data set in differentiating urban and rural populations

and areas.

The data set selected for study is generated by the

East Central Florida Regional Planning Council (1984).

Instead of the more widely used census tract, the geographic

base for the data set is the traffic zone, a unit which was

originally set up for transportation planning.

Nevertheless, these data are the basis for all regional,

county, and virtually all municipal planning in Seminole

County and therefore become the basis for subsequent

definitions of urban, rural,and the rural-urban fringe.

Traffic Zones as a
Standard Areal Measurement

A city is first divided into subareas, using
spatial units such as blocks, census tracts, or
traffic zones.
--Cadwallader 1985:2

Decision-making at the local level requires some

basis for division of the area, city, county, or region into

urban or rural contingents. In Central Florida the traffic

zone has been chosen as the standard unit of areal


History of Traffic Zones

Although traffic zones did not come into common

usage as geographic units for transportation studies until

the early 1960s, they were first used in the 1916 Chicago

Transit Study, which analyzed home-to-work ridership. They

were also used in an origin-destination study in 1947-48

called the McGuire Report for the Boston Area. Then, in the

early 1950s, transportation studies in Detroit and Chicago

brought both traffic zones and computer usage to national

attention (Roger Creighton 1987, Roger Creighton Associates,

consulting engineers, Delmar, NY, personal interview).

The recent evolution of planning in the United

States as it applies to federal funding of transportation

projects corresponds roughly with the application of

advanced computer technology, an increased awareness of the

need for more precise urban and rural socioeconomic data for

modeling transportation, land use, and housing, and a

greater demand by private citizens for involvement in

governmental decisions. In the early 1960s, the Federal

Highway Administration began developing alternative models

for transportation planning. During the period traffic

zones (TZs) came into general use as an areal measurement.

The impetus for the growth of transportation

planning was the Federal-Aid Highway Act of 1962. The Act

required that after July 1, 1965, all federally aided

highway projects in metropolitan areas with populations over

50,000 be based on transportation modeling. A major

requirement of the Act was that "the Secretary [of

Transportation] shall not approve any program .

unless he finds that such projects are based on a continuing

comprehensive transportation planning process carried on

cooperatively by states and local communities .

(Morehouse 1969:160). This "Three C" (cooperative,

comprehensive, and continuing) process endures today and is

the basis for the establishment of Florida's Metropolitan

Planning Organizations (MPOs).

One of the major goals of the MPOs was to

demonstrate the transportation needs of their regions to

both the state and federal transportation agencies.1

Justification for new or improved roads was based on a

priority system, which was shown by development and use of

transportation models. The models projected, usually for a

20-year time period, average daily traffic and volume

capacity ratio, that is, the number of vehicles using the

road and the number of vehicles for which the road was


One of the problems of transportation modeling

identified early on was a need for accurate, consistent

data. As early as 1961, data requirements were articulated

in a joint policy statement by the American Institute of

Planners and the Institute of Traffic Engineers. The

organizations agreed that planning responsibilities included

collection of all land use and socioeconomic data on a zone-

by-zone basis (Joint Policy Statement of the American

Institute of Planners and the Institute of Traffic Engineers

1961:70). In this 1961 report, however, the term "zone" was

not defined.

-Roads considered were federal, state, and county.
Municipal roads which were not part of those systems were
not modeled and were maintained by local funding.

The U.S. Census provided much of the data for the

transportation planning models, but a number of problems

came to light. For example, Census data are collected only

every 10 years, but transportation plans are based on 5-

year intervals and required updating annually. The models

also required very detailed information on small geographic

areas. Such data were not available in the appropriate

format or within an appropriate time frame from the Census


One solution to the problem of standardizing data

for traffic models and revising the data routinely was the

use of traffic zones (TZs). While traffic zones had been

used on a routine basis in some areas, it was not until 1977

that the Federal Highway Administration outlined the

criteria for standardizing the zones in areas with

populations of 50,000 or greater:

The transportation analysis units are known as
zones. These zones vary in size depending on density or
nature of urban development. In the central business
district (CBD), zones may be as small as a single block
and in the undeveloped area they may be as large as 10
or more square miles. An area with a million people
might have 600 to 800 zones and an area of 200,000
people might have 150 to 200 zones. These zones attempt
to bound homogeneous urban activities; that is, a zone
may be all residential, all commercial, all industrial,
etc. Zones also should consider natural boundaries and
census designations.
An important consideration in establishing zones is
their compatibility with the transportation network to
be used. As a general rule, the network should form the
boundaries of the zones (U.S. Federal Highway
Administration 1977:2-3)

Since county participation was required if federal and state

transportation funds were to be allocated, directions were

given for creating TZs for both "developed land [and]

the undeveloped land that the urban area will encompass in

the next 20 to 30 years" (U.S. Federal Highway

Administration 1977:2-2).

As early as 1963, the East Central Florida Regional

Planning Council staff converted all census tract data into

traffic zones. The conversion of data was based on

allocation of Census data to smaller areas and while traffic

zones could be made up of block data, more often it

corresponded with road networks and not neighborhoods.

These data were based on collective county/city policy, but

did not change the state population figures. Thus,

socioeconomic data from the 1960 Census were converted into

TZs, which became the standard geographic base. These data,

initially, applied only to the Urban Transportation Planning

Process (UTPS) (Heaton 1987).

Use of Traffic Zones in
Seminole County

While traffic zones were developed initially for

federal and state transportation planning, they have been

adapted for use in comprehensive planning in Seminole County

(see Table 2 and Figure 4). They are used instead of census

tracts as a geographic base because

Table 2
Sources of Comprehensive Planning Data
for Cities in Seminole County, Florida

Transportation Density
City Zones Census Other Measurement

Oviedo Yes No No No

Lake Mary Yes No No No

Maitland No No Yes No

Longwood Yes Yes No No

Sanford Yes No No No

Springs No Yes No No

Springs Yes Yes No Yes

Source: Cohen 1987, personal interview; Delk 1987, personal
interview; Koch 1987, personal interview; Marder
1987, personal interview; Nagle 1987, personal
interview; Weaver 1987, personal interview; Wells
1987, personal interview.




r 8'

n 3


\^Ni4 t o
t rr

. i-. (-

\ n

1. Census tracts are generally too large to use

for detailed data analysis, especially in

the unincorporated areas (see Figure 5), and

2. The Census is based on a decennial count and

detailed information becomes available only

several years after each enumeration. The

TZ data, by contrast, is completed and

reviewed every five years and has a 20-year

planning horizon (Price 1987)

While traffic zones are smaller units than census

tracts and some of the TZ data are derived from census

materials, aggregations of TZs do not correspond necessarily

with census tracts (see Figures 4 and 5). This is because

of a difference in philosophy between the U.S. Census Bureau

and the Federal and State Departments of Transportation.

Census tracts, blocks, and enumeration districts do not

split neighborhoods and have a primary function of grouping

like areas. On the other hand, transportation data are

based primarily on street networks. The splitting of

neighborhoods is a common feature of traffic zones in

Seminole County (Price 1987).

Another common feature of the TZs in Seminole County

is that they are not based on density, even though the

Federal Highway Administration Study (1977) defining traffic

zones suggests that density be considered (see Table 2). In


iC^ i ( ', E E m
t\ Lu



7 c0
\ 1.--

C) "; *4

/Y, ~ ~ ~ P ^ ^ l^^s^>
f~~ ~~ ^^s^2^^^
/-Jo ^^s cs~ne
--*^\j '-^ p _0iK?0 ^r15'*a
's'^) *~ "I^ XiXSiS~tL 1-

i u3 P u c c
s; 04 !
i *r1
~ij~ 2

fact, according to regional, county, and city planners, no

calculations of area in various traffic zones in Seminole

County have ever been made (Delk 1987, personal interview;

Gilbrook 1987, personal interview; Grovdahl 1987, personal

interview; Heaton 1987, personal interview; Marder 1987,

personal interview; Nagle 1987, personal interview; Price

1987, personal interview; Ross 1987, personal interview;

Thompson 1987, personal interview; Weaver 1987, personal

interview; Wells 1987, personal interview).

There will be one exception in the future. The

planner in Winter Springs, one of the seven cities in the

county, intends to use Census data and density figures in

the city's Comprehensive Plan Update. The Winter Springs

Planning Department, which in 1987 is just beginning to

receive detailed 1980 Census data, plans to use block data

cross-referenced with Seminole County Traffic Zone data. It

also intends to calculate densities for population, housing,

and employment all to be used in future planning (Koch


Use of the Traffic Zone
in the Present Study

The traffic zone was chosen as the areal unit in

this study because it is the unit used most often by

planning agencies in Seminole County. In addition, it is

the unit on which the socioeconomic variables generated by

the RPC and used by the county to describe urban and rural

areas is based.

Despite the obvious disparity in traffic zone sizes

in the county (see Figure 4), neither area nor density data

have been calculated by the planning agencies. In the past,

comparisons of socioeconomic data have been based solely on

absolute data values. These values comprise the first major

data set of the study and are called the planning data.

In order to standardize comparison of traffic zones,

each of the 181 TZs in the research area has been

planimetered and the area calculated. Using these areas,

the planning data are converted to a second major data set

called density data (see Appendix).

For planning purposes in Seminole County,

incorporated areas are considered urban while the

unincorporated areas are designated rural. For this reason,

each of the major data sets is divided into two subsets:

1. Incorporated traffic zones which lie either

totally or partially within the incorporated

city limits of Altamonte Springs,

Casselberry, Winter Springs, Longwood, Lake

Mary, and Sanford (see Figure 6), and

2. Unincorporated TZs which correspond to the

rural portions of the county




For each of the traffic zones a centroid has been

estimated, which is used to calculate the distance to a

common point located at State Road 436 and Interstate 4 in

the major urban center of Altamonte Springs. This area was

chosen as the urban node because it is the economic and

transportation focal point for Seminole County and is

adjacent to Orange County and in close proximity to the City

of Orlando (see Figure 6).

Of the 181 traffic zones in Seminole County, 175

zones were used in this research. Six zones were deleted

because population data were recorded as zero and they would

have produced questionable analysis results.2

Definitions of the Variables

The variables used in the present study include

(1) population, (2) single-family housing, (3) multi-family

housing, (4) resident employment, (5) attendant employment,

(6) retail employment, (7) median family income, and

(8) persons per household. The data for the first

seven of these variables were obtained from the

20f the traffic zones which were omitted, three--82,
85, and 160--were agricultural or vacant land in 1980. The
other three--42, 124, and 137--were totally industrial
retail or commercial retail areas. These traffic zones are
small and scattered throughout Seminole County.

Orange-Seminole-Osceola Statistical Data, 1980-20053 (East

Central Florida Regional Planning Council 1984). Persons

per household was calculated by the author because it is

commonly used for planning purposes.

The variables are defined by the East Central

Florida Regional Planning Council (1984:vi-x) as

1. Population is defined as the total number of

civilian and military individuals whose

principal residence is in a particular

traffic zone. Excluded from these figures

are seasonal and transient residents.

Population allocations to traffic zones are

based primarily on 1980 U.S. Census data and

the total figures are allocated by Regional

Planning Council, County and City

governmental staff.

2. Single-family housing is defined as the

total number of completed single-family

dwellings and mobile homes, whether they are

occupied or vacant. Not included in the

3Two variables listed in the document, hotel/motel
units and school enrollment, grades 1-12, were not used in
this research. Hotel/motel units was excluded because only
20 TZs had more than one unit. While 39 TZs had values for
school enrollment, this variable would only indicate where
schools are since student population is assigned to the
location of the school and not the home.

count are seasonal or migratory mobile-home


3. Multi-family housing includes completed,

occupied, or vacant units where two or more

units occur in a single structure. Multi-

family housing does not, however, include

units which are considered group quarters.

4. Resident employment counts full- and part-

time employed individuals by place of

residence. If a person holds more than one

job, he/she is counted only once.

5. Attendant employment counts full- and part-

time employed persons by place of work. In

this case, if an individual holds more than

one job, he/she is counted at each place of


6. Retail employment is a subset of attendant

employment and counts only those individuals

employed in retail trade. A person is

considered employed in retail activities if

his employer has a two-digit Standard

Industrial Classification between 52 and 59


7. Median family income describes the income

level in each traffic zone for which half

the families in a particular area have

incomes below the median and half the

families have incomes above the median.

The data for the preceding variables have been

compiled by the East Central Florida Regional Planning

Council and represent the most comprehensive computation

available to the planning offices in Central Florida. As a

result, these are the data used in the area from day to day

for planning, policy, and land-development decisions.

The Roles of Distance and Density

The concept of the "distance-decay function" is

derived from Alonso's theory of urban land market (Dennis

and Clout 1980). This concept was applied to the rural-

urban fringe in Pryor's model (Pryor 1971) and is the basis

for using the variable "distance from the node" in the

present study. Pryor's model (see page 47 of this study)

shows the relationship between percentage of urban and rural

land use and population characteristics and distance from

areas which are 100 percent urban at one extreme and

100 percent rural at the other.

Regression Analysis

The statistical method of least-squares regression

is employed to describe the relationships between distance

from the node and each socioeconomic variable in the study

(Sincich 1982). For each variable, six regressions are run,

regressing the socioeconomic variable against the distance

from the traffic zone nodes to the urban node at the

intersection of State Road 436 and Interstate 4. These six

regression analyses include (a) combined planning data,

(b) incorporated planning data, (c) unincorporated planning

data, (d) combined density data, (e) incorporated density

data, and (f) unincorporated density data.

Two scattergrams are plotted for each variable. The

first describes the combined planning data and the second

shows the combined density data. On each, the "best fit"

regression lines for the incorporated and unincorporated

subsets are plotted.

It is assumed throughout that the characteristics of

the urban node are 100 percent urban. It is also assumed

that at some unspecified distance from the urban node there

exists an area which is totally rural. Each of the

variables analyzed represents an "urban"characteristic which

will be at its peak at a city center Orlando for Orange

County and perhaps Altamonte Springs in Seminole County at

the county urban node. As the regression line for a

variable's incorporated data increases in distance from the

urban node, the TZs become less urban in terms of that

particular characteristic. By the same token, the

regression line for the variable's unincorporated data

should indicate increasingly urbanized characteristics as

one approaches the urban node. Thus, the intersection of

the two regression lines on each graph represents the

juxtaposition of the characteristics of urban and of rural

traffic zones, or the rural-urban fringe, for the variable

exhibited in the graph.

Then, as a final step, the distance from the

intersection of the regression lines to the urban node has

been calculated for both planning and density data and the

two distances compared. The data are represented in charts

and graphs which illustrate the similarities and differences

in "urban" and "rural" as they apply to zones at various

distances from the urban center of Seminole County.


The F-values were calculated to test whether or not

the regression analysis was useful in predicting y. The

formula used is that devised by Sincich in his text


F =R 2/k
(1 R2)/[n (k + 1)]


R2 = multiple coefficient of determination

n = number of observations

k = number of B parameters in the model

(excluding B0)

Z-Score Statistic

The Z-score statistic was employed to determine

whether the incorporated and unincorporated populations are

statistically similar or dissimilar. The formula for

comparing the means of the two populations is derived from a

study by Walker and Lev (1958:233)

(XI X2 (P P2)
1 2 ^1 -- -- 2-
2 2
1 + s2
/ N^ N2
1 N2


p = population means

X = mean

s = sample standard deviation

N = sample size


The terms urban, rural, and rural-urban fringe have

never been adequately defined, yet developmental and

planning decisions must be based on these concepts. Local

governments often set their own criteria for differentiating

such areas. In Seminole County, the bases for

distinguishing urban from rural are the incorporation

boundaries and the characteristics described in the data set

referred to in this study as the planning data set.

The analyses to follow are intended to determine the

viability of the characteristics in (1) determining urban

and rural and (2) distinguishing the rural-urban fringe.

Toward these ends, the data, in original form (and after

modification to account for density), are subjected to

several statistical examinations.


The Census data are set out in such a way that they

seem difficult, and often inappropriate, for local decision-

making. For this reason, smaller governmental units, such

as cities and counties often bypass census data and choose

alternative data sets which seem more appropriate in areal

unit size, collection interval, and detail. Such is the

case in Seminole County, where rapid growth requires

continual monitoring and re-evaluation of land use.

The data set chosen by the county and most of the

city governments is produced by the East Central Florida

Regional Planning Council and based on traffic zones. These

data are in general use for daily governmental and private

planning decisions and, while the effectiveness of the data

in differentiating urban, rural, and rural-urban fringe

areas is assumed, it has not been proved. The present

analysis seeks to test the validity of this assumption.

Analysis and Data Sets

The analysis presented in this chapter uses the

statistical technique, regression analysis, to describe and

compare two major data sets. These sets are called the

combined data in the analysis.

The first set, planning data, is the one in general

use in the county. It consists simply of total counts for

each traffic zone.

The second data set, density data, was calculated by

the author. Because the area of each traffic zone had not

been previously determined, density figures for traffic zone

variables have not been available to local governments.

These data, however, would be useful for planning purposes

because they better reflect the distribution of population

and population characteristics than do the planning data.

Each of the major, or combined, data sets is

subdivided into incorporated and unincorporated data sets.

Thus, a total of six data sets is used in the analysis.

Presentation of Findings

The results of the analysis are presented as a

series of table and graphs. Relationships between each

variable and distance are described in two ways:

1. Tabular--A table lists analysis results of

planning and density data within

incorporated, unincorporated, and combined

traffic zones, and

2. Graphical--Two graphs in the form of

scattergrams are presented for each

variable. The first is based on planning

data, while the second is based on density


Population--Distance Relationships

The statistical findings in Table 3 describe the

relationships between population and distance to the urban

node for each of the six traffic zone data sets. Several of

these are of particular interest.

The R2 values suggest little systematic association

between traffic zone population and distance from the node

in any data set. It is clear, however, that by adjusting

the planning data for traffic zone size--that is, converting

the planning data to density data--higher values for R2 are


The slope of the regression line is negative, as

expected, indicating that the greater the distance from the

node, the smaller the population. This relationships exists

for all six data sets, but it is more pronounced in the

three density data sets (see Table 3 and Figures 7 and 8).

The ranges shown in the planning data results

represent the differences between the highest and lowest

absolute traffic zone populations, while the ranges within

,-1 rN Co CN
O t ct Co co
D0 *

- 00 -
S-l I M r-a

4- (U






c o



D 0






M 0

cr) t

r 0l C

r-. C%)
U( 0

10 0)

O r

*-1 0*

10 M



R a


04 4J
S0 a
u c u)
oC 0 0)

4-O 0 > [ c
C -11 7
1 m Fw; ft





















o o
0 CM



C 0O

- N

- (


t 1 I

10 C) l)
0) to

(Y' LO -4



- 4


0 40 80 120 160 200

Distnce to Nod tenthss of mile)
D 0 X
Incorporated Unincorporated Incorporated


Figure 7. Regression lines for incorporated and
unincorporated population size planning data


0 40 0 120 160 200

Dstancs to Ide (tenfts of ma)
So0 x
Incorporated Unincorporated Incorporated

nco ated

Figure 8. Regression lines for incorporated and
unincorporated population density data

the density results show the differences between the highest

and lowest population density figures. Thus, ranges for

absolute population in incorporated and unincorporated

populations for the traffic zones are 3,923 and 4,681,

respectively. The population density, however, has a range

of 5,923 persons per square mile for incorporated traffic

zones and 4,984 persons per square mile for unincorporated

traffic zones.

Although the distribution is scattered, the expected

intersection of the urban and rural planning data regression

lines indeed takes place at 11.84 miles from the node (see

Table 3 and Figure 7). When the data are adjusted to

reflect density, the intersection of the regression lines

moves outward to 15.59 miles from the node (see Table 3 and

Figure 8). The approximately 30 percent difference between

the two intersections could make a difference in

governmental and private planning decisions. The planning

data suggest a break between rural and urban at 11.84 miles

from the node. Since these are the data used to make

governmental decisions, this would typically be where

governmental services abate. The density data, however,

suggest that development has advanced beyond the 11.84-mile

limit, as far as 15.59 miles. Therefore, the outermost

development may not receive essential services.

There are several factors which could influence

population values in both planning and density data sets.

Large portions of the county are protected from development

by federal, state, and county environmental regulations.

Most developable areas of the county are serviced by at

least one major transportation artery. These roads include

Interstate 4, U.S. 17-92, S.R. 436, S.R. 434, S.R. 419,

C.R. 46, and Lake Mary Boulevard (see Figure 9).

Furthermore, the influence of Orange County and Orlando to

the south should not be discounted. Seminole County has

been historically a "bedroom community" (Osinski 1985)and

provided residential housing for Orange County while

employment opportunities are more prevalent in Orange

County. Thus, the traffic zones in the southwest and south

central part of the county are more densely populated than

the rest of Seminole County.

Single-Family Housing--Distance Relationships

Based on R2 there appears to be only a weak

relationship between single-family housing and distances to

the node. The planning R2 values are .009 for the

incorporated areas and .150 for unincorporated areas of the

county. These values are raised only slightly to .079 and

.209, respectively, when the planning data are adjusted for

density (see Table 4).


o oc





o rl


. 5a)




















0) (



(0 Q


*'-I 4-I


0. Q




*< ro





; Q



10 01
C o 0

.44 0

. 4

r-1 O0

.' 01
O o
to o

rr r
Ic c

4-' (U

la 4

S40 0 ) 0

@ 0 Qa (U m (
4J o > 0
14 C I I >

(4 wD (0 t0 r-

>o -4 CM
I -4

10' 10 (0 ( (0
S C) C )

S O tO r- ,D


I -4 -O

r- 0
C O (0


O 0

0 0

r l

co 0

cr) CV

























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 EAW4O8WX2_ITQNVX INGEST_TIME 2012-08-30T16:01:30Z PACKAGE AA00011771_00001