?PLEMENTARY
SKAATERgég

INBACKOFBOO

PLACE IN RETURN BOX to remove thlo choekout from your rocord.
TO AVOID FINES rotun on or boforo duo duo.

 

DATE DUE DATE DUE DATE DUE

 

 

f .236 W
[#895 g

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU lo An Affirmdivo AdiorVEquol Opportunity Inflation
. m1

URBAN SOCIAL SPACE: RESIDENTIAL PATTERNS
IN TAIPEI, TAIWAN: 1980

By

Christopher Sutherland

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Geography

1991

ABSTRACT
URBAN SOCIAL SPACE: RESIDENTIAL PATTERNS
IN TAIPEI, TAIWAN: 1980
By

Christopher Sutherland

This study fills an existing gap in the geographic literature on
Taiwan. Taiwan has experienced dramatic success in economic development
over the past four decades. This economic success has had a significant
impact on the social fabric of the country. Within Taiwan the field of
geography is largely oriented toward the physical world. Little work has
been done within the realm of social geography. This study helps to fill
this gap by identifying the residential groupings of the city of Taipei
and examining their evolution.

Since the 1960's Factorial Ecology has been an accepted framework
for describing and investigating residential groupings in the city. This
study uses this approach to establish the spatial structure of
residential neighborhoods in Taipei at the time of the 1980 census. From
the census 138 variables were selected to describe the population.
Variables were gathered for the 1500 census tracts in the metropolitan
area. The data matrix was submitted to principal components analysis and
nine factors extracted that explain 80.6% of the common variance. While
the factors extracted showed little separation between the traditional
sorting elements of socio-economic status, stage in the family life
cycle and ethnicity; the mapped factor scores do provide an excellent
description of existing residential groupings. Historical analysis is

then used to explore the evolution of these groups.

ACKNOWLEDGEMENTS

This research could not have been accomplished without the support
of a variety of individuals and organizations. The most important and
consistent support has come from my family. My wife, Kathy, has
supported this effort fully, and paid the emotional price of being
involved with someone focussed on their research, often at the expense
of family. I want also to thank my children, Stephen, Erin and Kevin,
for their love and understanding. Beyond my family the person most
deserving of thanks is Dr. Jack F. Williams. As mentor and friend he has
put up with more than he deserves, but always been there when needed.

I would also like to extend my thanks to the Fulbright
organization for funding my field research. Thanks are also due the
Pacific Cultural Foundation in Taiwan for much needed financial support.

There have been many others who have helped me along the way.
While some such as National Taiwan University's Department of Geography
and the Institute of Economics at Academia Sinica in Taiwan are
deserving of special thanks; the list is too long for me to mention them
all. I want all those who did lend me their support and help to know

that I appreciate their efforts.

iii

TABLE OF CONTENTS

List of Figures ................................................... v
Chapter One ....................................................... 1
Statement of The Problem .................................... 1

The Geographic Approach to the Problem:

Theory and Analysis ................................... 1

Chapter Two ...................................................... 12
Taipei, Taiwan: The Historical Setting .................... 12
Introduction ............................................... 12
Historical Background ...................................... 13
Chapter Three .................................................... 41
The Theoretical and Analytical Framework ................... 41
Theory and the Classical Models ............................ 41
Behavioral Basis: The Residential Location Decision ....... 44
Analytical Framework- Social Area Analysis ................. 47

A Cross-Cultural Context: Theory and Framework ............ 52
General Evidence ..................................... 52

The Case of Taiwan ................................... 57

The Research Question: Residential Space In Taipei ........ 60
Chapter Four ..................................................... 68
The Research Experience and Results ........................ 68
Spatial Scale and Required Variables-The Data Base ......... 68

iv

Scale ................................................ 68

Variables ............................................ 70

The Search for a Data Base ........................... 73

Assembling the Data Base ............................. 81

Results—The Factors Extracted .............................. 83
Factor One: Socio-economic Status ................... 84

Factor Two: Socio-economic Status ................... 90

Factor Three: Stage in the Family Life Cycle ........ 98

Factor Four: Socio-economic Status ................. 103

Factor Five: Socio—economic Status ................. 109

Factor Six: Socio-economic Status .................. 115

Factor Seven: Socio-economic Status ................ 125

Factor Eight: Socio-economic Status ................ 130

Factor Nine: Stage in The Family Life Cycle ........ 136

Chapter Five .................................................... 143
Conclusions ............................................... 143
Evaluating the Hypothesis ........................... 143

Indications For Future Research ..................... 149

Appendix A ...................................................... 153
Appendix B ...................................................... 155
Appendix C ...................................................... 159
Appendix D ...................................................... 162
Appendix E ...................................................... 176
Bibliography .................................................... 183

Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure

Figure

LIST OF FIGURES

1: Taiwan .................................................. 14
II: Residential Location Decision Process .................. 45
III: Social Area Analysis .................................. 49
IV: Integrated Model of the Metropolitan Region ............ 51
V: Summary of Janet Abu-Lughod's Necessary Conditions ...... 54
VI: Possible Factorial Combinations ........................ 56
VII: Taipei Basin, Taiwan - Land Use: 1983 ............... 153
VIII: Taipei Metropolitan Region ........................... 78
IX: Factor One ............................................. 85
X: Factor One Scores ....................................... 87
XI: Factor Two ............................................. 91
XII: Factor Two Scores ..................................... 93
XIII: Factor Three ......................................... 99
XIV: Factor Three Scores .................................. 100
XV: Factor Four ........................................... 104
XVI: Factor Four Scores ................................... 106
XVII: Factor Five ......................................... 111
XVIII: Factor Five Scores ................................. 113
XIX: Factor Six ........................................... 116
XX: Factor Six Scores ..................................... 118
XXI: Factor Seven ......................................... 126

vi

Figure XXII: Factor Seven Scores ................................. 128

Figure XXIII: Factor Eight ....................................... 132
Figure XXIV: Factor Eight Scores ................................. 134
Figure XXV: Factor Nine .......................................... 137
Figure XXVI: Factor Nine Scores .................................. 139

Figure XXVII: Variables Used in Analysis of Chicago by

Berry/Kasarda ...................................... 155
Figure XXVIII: Variables Used in Hsu and Pannell's Study

of Taipei ......................................... 159
Figure XXIX: Variables Gathered in Taiwan Census ................. 162
Figure XXX: Variables Used in the Social Area Analysis

of Taipei ............................................ 176

vii

CHAPTER ONE

STATEMENT OF THE PROBLEM

Urban Geography has developed a significant body of theory that
attempts to identify patterns of residential location and explain the
processes which serve to create neighborhoods. Many of these theories
are embodied in an approach to studying residential patterns in urban
areas known as Factorial Ecology or Social Area Analysis. The theory
gives rise to a set of expectations concerning the spatial pattern of
residential arrangement that is tied to residential mobility and the
level of development. Broadly speaking, in countries with a reasonably
developed economy and a fair degree of residential mobility, there
should be discernible separation between arrangements based on
socio-economic status, stage in the family life cycle and, where
appropriate, ethnicity.

This study was structured to investigate the spatial arrangement
of residential groups in the post-World War II era, in Taipei, Taiwan.
Taipei was selected for the study for a number of reasons. It is a city
that functions as a capital city, and one that tends toward primacy. It
is a unique city in many ways, combining Chinese influence, Japanese
influence and aspects of western influence. It is a city that embodies

the entrepreneurial spirit that has enabled Taiwan to lift itself above

2
the ranks of the Lesser Developed Countries (LDC's) of the world. The
post 1949 growth has been dramatic, and that growth was expected to be
reflected in the emergence of residential groupings different from
historical patterns.

The study proposed investigating the residential structure of
Taipei over time. Data availability limited statistical investigation to
1980. Patterns for periods prior to that time were established using the
tools of historical geography. The patterns for 1980 were established
and mapped using as the framework of analysis Factorial Ecology. A
comparison of historical patterns with the 1980 pattern were expected to
indicate a changing residential pattern. A number of hypotheses were
established that predicted the expected patterns of change. In general,
it was felt that in 1980 social groupings similar to those in western
cities would be distinguishable.

As mentioned the available data bases limited the focus of the
study to 1980. While this limited the scope of the preposal it was
decided that the study had significant value. Taipei has attracted
relatively little attention by researchers, particularly those working
in English. Little has been published about the development of the human
landscape of the city. Consequently, the study offers a significant
contribution to the literature on Taipei. In particular, it establishes
the distinct residential groupings present in the city in 1980, and it
sets the stage for future research along these lines.

This study proposed the use of Factorial Ecology in examining the
residential structure of Taipei, Taiwan. Little has been published
concerning the social groupings and their spatial distribution in this

city. While a variety of qualitative assessments exist, only one article

3
has been published that attempts to systematically assess the area (Hsu
and Pannel, 1976) and one that attempts to establish the mechanisms
involved in residential moves (Wang S.C. 1981). Neither serves to
adequately address the issues involved. This is significant given the
growth experienced by Taipei and Taiwan since the end of World War II.
Taiwan is one of the few countries to have emerged from the colonial era
as an LDC and successfully transform itself into an economically
successful country. Taipei has played a significant role in this process
of development, and by examining the changes that have occurred the
potential exists to extend our understanding of the impact of
development on cities.

This study seeks to explore the current (1980) spatial pattern of
residential arrangement, and to trace the evolution of that pattern over
time. The remainder of the Introduction will give a brief overview of
the theoretical basis of the study. Chapter Two will examine the
selection of Taiwan as the study site, and of Taipei in particular.
Chapter Three of this paper will delve further into the background of
investigations into residential structure. It will explain the method to
be employed in this study for determining the residential structure of
Taipei and will set forth the expected results of such an analysis as it
applies to the city. Chapter Four will present the results of the
research. Chapter Five will draw together the various elements of the

research process and discuss those areas needing further investigation.

THE GEOGRAPHIC APPROACH TO THE PROBLEM: THEORY AND ANALYSIS
It has been recognized for some time that cities contain an

increasingly disproportionate share of national wealth in developed and

a

developing countries. More recently, awareness has extended to the
intimate but complex relationship that exists between urbanization and
national development. The urbanization process which was once seen as
only the inevitable result of development is now recognized as one of
the key elements of the development process itself. In short, the urban
scale is an extremely important one. The city, particularly the capital
or chief city of a nation, is often the focus of national life. This is
true in both developing and developed nations. It serves as the
principal innovator and diffusion agent. It provides the main point of
contact with the outside world. It is the center of governmental
activity and it is often the preferred location for industrial and
commercial activity. The central city of a nation, through its place in
the internal urban hierarchy, is often reflective of cities below it in
the hierarchy; and of core-periphery relationships both external and
internal to the nation. In the case of primate cities the hierarchy of
cities may be missing; but the role of the city is even more clearly
defined. It is usually the focus of the modern economy and a clear
example of core-periphery relationships. The varied and vital roles the
urban center plays within the nation's spatial structure makes it
essential to investigate cities in order to understand the nature of the
development process and the impact of such development on social
structure.

The role of cities within the development process has attracted a
great deal of attention. Initially, urbanization was simply viewed as a
byproduct of development, or more commonly, modernization. As more
attention began to be paid to the nature of the development process,

however, the role of the city within the process, as an agent of

5
change, was identified. These earlier studies of urban centers gave rise
to two major approaches for analyzing the city. The first involves
investigation into the propulsive or catalytic role of the city in
facilitating the development process. The second involves investigations
into the internal spatial structure of the city and investigation into
the way in which that structure is shaped by economic change and
development.

The first perspective is concerned with the manner in which
cities, and particularly systems of cities, serve to facilitate the
processes of competition, integration and diffusion. This perspective
springs from investigation into urban development in the More Developed
Countries (MDC's) of the world. In the 1960's, a number of works
appeared that classified cities in the MDC's on the basis of systems or
regions. Investigators such as Perloff (1960) and Duncan (1960),
building on the pioneering work of researchers such as Christaller and
Losch, argued that both the national economy and the national geography
of the United States could be succinctly described in terms of its urban
system. These initial works invigorated the field of urban geography. By
1961, Berry was proposing a formal link between urban population
distributions and the hierarchy of service centers, and linking these to
the language of general systems theory. In subsequent years the nature
and scope of the field of investigation was refined. Concepts such as
megalopolis (Gottman, 1961), the urban field (Friedmann and Miller,
1965), the functional economic area (Fox and Kumar, 1965), and the daily
urban system (Berry, 1973) extended the concept of the city within the

systems framework.

6

During the same period, other investigators were examining the
processes involved in economic development, particularly under free
market economic systems. Interest in such research had been stimulated
by the post-World War II breakup of colonial empires and the entry into
the global economy of a multitude of independent nations that were, by
Western measures, underdeveloped. Researchers such as Rostow(1960)
galvanized the field. What was clear to many of these researchers was
that the application of post-World War II development schemes to the
LDC's was not having the desired effects, and that regardless of their
success or failure the LDC's were experiencing increasing rates of
urbanization. In particular, the rise of the primate city phenomenon was
being noted. (A primate city being a city that is at least two times as
large as the next largest city in the nation, and one that dominates all
aspects of national life.) While the primate city was not a new
phenomenon (for example Rome, London, Amsterdam, etc.) it did indicate
problems with the spatial distribution of development. The inefficiency
of Western models of development in transforming the nation was leading
to changing definitions of development. Theorists and governments began
to define development in terms of goals such as balanced economic
development, regional equality and orderly population redistribution. A
number of theorists began to reexamine the role of the city within the
MDC's with the aim of extracting those elements seen as crucial in the
organization of urban systems in MDC's and applying them to the LDC's in
order to facilitate the spatial transformation and modernization of the
LDC's. The initial general concerns were expressed by researchers such
as Myrdal (1957) and led to more specific investigations by those such

as Thompson (1971) and Rodwin (1970). These individuals addressed the

7
universal aspects of development problems and, in part, were proposing
the necessity of urban systems in achieving transformation. Since then a
flood of literature has appeared that analyzes various components of
national urban systems along new or existing directions, and debates the
policy alternatives (Bourne 1978: 9-10). Examples of this type of
research can be found in the works of Cameron and Wingo 1973, Berry
1973, Hansen 1975, and Swain and Logan 1975. Basically these studies
involve investigation into the manner in which the city, acting as an
organizer of space, affects the development process.

The second approach to studying cities examines the way in which
the economic landscape affects the internal spatial structure of the
city. This type of research can be broadly categorized as falling into
two types. The first type is primarily concerned with the spatial form
of the city, the second with spatial processes. In both cases the search
is for a regularity of pattern, be it land use patterns, population
distributions or densities, density of functions, or whatever type of
pattern is being sought. Much of contemporary geography rests on the
assumption that there exists an inherent geographic order in human
society, a spatial anatomy of human behavior and societal organization
which has regular and discernible characteristics. The bulk of modern
geographical research has been associated with the search for order and
regularity in spatial systems (Soja and Tobin 1977: 155-156).

Much of this work, within the field of urban geography, has been
concentrated on the construction of models of internal spatial
structure, and have included the concentric zone (Burgess 1925), the
sector (Hoyt 1939), and the multiple nuclei models (Harris and Ullman

1945). These models have been criticized over the years on a number of

8
points (see for example Gist and Fava 1964, Alihan 1938, Davie 1937,
Gilmore 1953, or Anderson and Egeland 1961), but they are still
considered valuable tools. A more sophisticated model, integrating many
of the above elements was proposed in the 1970's (Hagget, Cliff and Frey
1977, Rees 1978).

One interesting point concerning these models is that they were
largely developed relative to the experience of the United States.
Investigations of cities in other cultures have been less extensive. The
model of the colonial city and the post-colonial city have been one
result of this type of investigation. These generalized models
concerning cities developed under colonial systems have arisen out of
various regional explorations of the structure of the city. While some
are applicable only to the region in which the city is located, a
significant number are relevant to other regions. These types of studies
include Griffin and Ford's (1980) study of Latin American urban
structure; the works of Abu-Lughod (1971) particularly those on Cairo,
Bonine (1980), Aderibigbe (1975) and Hance (1970) on Africa and the
Middle East; and those of Murphey (1966), Ginsberg (1965), Breese (1966,
1969) and Sit (1985) on Asian cities.

The second type of internal spatial studies has concentrated on
the processes at work in the city. In essence, these writers are
following a behavioralist approach. They "believe that the physical
elements of existing and past spatial systems represent manifestations
of a decision making behavior on the landscape, and they search for
geographic understanding by examining the processes that produce spatial
phenomena rather than by examining the phenomenon itself" (Amadeo and

Golledge, 1975). These studies have increasingly examined the impact

9
that development has on the urban structure. They often operate from the
perspective of a particular paradigm of developmental analysis. These
methods of analysis include dependency theory, core-periphery
relationships, world systems theory, Marxist theory as well as many
others. The range of literature is extensive and excellent examples can
be found in the works of such individuals as Harvey (1973), Slater
(1977), Lubeck (1977), Roberts (1975), and Smith (1979).

This study falls within the broad context of the second type of
perspective. In particular it is concerned with the internal spatial
structure of the city. It seeks to combine the traditional emphasis on
regional understanding and description with modern research approaches
and theories that sometimes have failed to appreciate the need for
regional scholarship. This is consistent with what is being called "
restructured regional geography” by among others Abler (1987), Gilbert
(1988), and Pudup (1988).

Systematic geographic investigation of the internal morphology of
cities is of relatively recent origin. It is generally attributed to
social applications of ecological work in the botanical sciences at the
University of Chicago in the early 1900's. Armed with an analytical
framework providing concepts such as invasion and dominance, geographers
and others began to propose theories and models aimed at explaining the
processes leading to the social/physical pattern of the city.
Investigators such as Burgess, Hoyt, Harris and Ullman (cited earlier)
and Wendell Bell (1953) were among the leaders in this type of research.
In the early 1950's statistical methodology began to make inroads into
the field. The advent of computers, in particular, gave rise to a wider

application of multi-variate analysis in the investigatory process.

10

Since the 1960's, one accepted framework for investigations into
the residential structure of cities has been Factorial Ecology. Urban
ecology, the philosophical foundation for the approach, has become a
cornerstone of the multi-dimensional research into urban structure.
"Urban ecology studies the interaction between human activity and the
space in which these activities are located" (Anderssen 1983: 153). The
Factorial part of the approach refers to the use of factor
analysis/principal components analysis as a tool to reduce a large
number of variables to a manageable number of factors or dimensions,
which describe various groups of residents and their associated dwelling
space. Through the application of such techniques, geographers have
developed a reasonably good tool for use in identifying and explaining
the social/residential pattern of the city.

The early work concentrated on developed countries, particularly
U.S. and Canadian cities. However a number of investigations began to
made into the structure of cities in developing nations. These
investigations were probably best described by Timms (1970) and Berry
(1973). These researchers linked the residential structure of the city
directly to the level of economic development in the society.

By the mid to late 1970's the use of Factorial Ecology had
virtually disappeared. Two major reasons seem to account for this fact.
First, it was criticized by geographers concerned with quantitative
methods for its use of principal components analysis. Secondly, and most
importantly, it was viewed as a highly useful descriptive device, but
one that seemed to have little use beyond description. It provided,
according to many 'an historically empty explanation'. Anderssen (1983)

has demonstrated that the description alone is valuable, and that by

11
linking the use of the tool to regional and historical geography, our
understanding of the study site, and of cities in general, can be
extended. In essence what Anderssen accomplished was to defuse the
arguments of those concerned with the use of principal components
analysis. To many geographers this statistical method is akin to
rummaging around in the data to see if any relationships emerge.
Anderssen has used historical and regional geography to provide a set of
expectations concerning the period under study. This ameliorates the
reservations expressed about use of the method. In addition by linking
the approaches the resulting explanation is no longer "historically

empty".

CHAPTER TWO

TAIPEI, TAIWAN: THE HISTORICAL SETTING
INTRODUCTION

Taiwan and the city of Taipei are ideal choices for this study.
Taiwan is, in the eyes of the R.0.C and the P.R.C. governments at least,
a province of China. Its history as a part of China is relatively long,
and as such it has developed largely within a cultural milieu that is
distinctly different from western culture. It had a colonial experience,
with Japan as the colonizer. Lastly, unlike the bulk of China, it has
developed in the last four decades within a capitalist economic system.
Within that system there has been both geographical and social mobility.
Western urban theory as it relates to residential structure assumes
residential mobility, and that residential location is closely
associated with class as defined by income level. Distortions in the
model, those not tied to physical limitations, are identified as the
result of social actions, such as racism, acting on potential, universal
behavior. Taiwan, having such residential mobility, meets the demands of
the theories being employed.

Taiwan was a nation that by 1980 (the target year of the research)
had achieved significant industrial and economic development. The per
capita income of the nation was more than $2100 in 1980, a rate placing
it fourth in Monsoon Asia. Real average annual increases in national
income have been among the highest of the LDC's, averaging nine percent

12

13

between 1963-1973, and reaching 11 percent in the 1976-1980 period. By
the late 1970's, Taiwan had become one of the top two dozen trading
nations in the world. Within Taiwan there were numerous indications that
the benefits of development had been distributed on a fairly wide basis.
Per capita caloric intake in 1980 reached 2800/day, while per capita
daily protein consumption increased from 49 grams in the 1960's to 78
grams by 1980. Statistics indicate that the nation is decreasing its
consumption of rice while increasing its intake of meat, fruits and
vegetables, dairy products, and cereals and pulses. Ownership of
motorcycles was one in four persons, automobiles one in 23. Average
living space had increased to over 128 square feet per person, and
virtually all homes had television, refrigerators and telephones.
(Directorate General of Budget, Accounting and Statistics (DGBAS), 1984)

In short, Taiwan and its capital city Taipei, make an excellent
choice for this study. The region combines a non-western culture and
colonial experience with an extremely successful market economy. The
economic system provides a commonality of experience, while the culture
and colonial experience provide the unique. With this in mind, attention

will now be turned to the historical development of the country.

HISTORICAL BACKGROUND

Taiwan is a subtropical island 240 miles long by 90 miles wide,
lying off the south China coast at 23.30N and 122.20E. It lies in the
East China Sea and is separated from the Mainland by the Taiwan Straits.

(see Figure I)

14

FIGURE I

TAIWAN

 

 

  

Population. 1980
o Iooothon 100.000
0 100.000-500000
0 500000- 1000000

moro than
1.000.000 .

 
 

Hsmchu

  
   
  
   
 

oChUIUno

  

O
Chunon - Totun

  
  
   

  
 

oMmolr

NORTHERN

   
   
   

CENTRAL
7.5199992?“ 0"
if? Iéichung

  
  
  

   
  

‘ Huolion : .1

s

YUCDII If ' 0?qu
OCT‘IN no ONIMOU

 
   
 
 
 

 

m
ofoulru

  
 

oCI'IIIyI

  
 
  

SOUTHERN

      

ammo
oChIoII - Mom
.. O" I EASTERN
’. 1‘ o
"o. ' .
._ :Tomon
JG"
$1 .. . common
‘T N51“}.
€33;-
ns \
- ‘1': OPingtung
Klnllqung ' O Loco: Conn:
Q RoglonoI Contor
. Political ond
Econormc Comor
1}; Motropolmn Aroo
‘1 — Rogoon Boundory

 

 

 

Sourco: WIIIIomo (1000)

15

Taiwan has been within the Chinese sphere of influence for over
one thousand years, and its culture and linguistic patterns have largely
been determined by this influence. However, four major migrations can be
identified in Taiwan. The earliest known migration to the island was by
a group that is identified today as the Taiwan Aborigines, or as they
are known in Taiwan, the Shan Ti Ren or Mountain People, a designation
linked to the geographical regions to which they have been forced to
retreat. The ancestors of these aborigines migrated to Taiwan from the
Philippines some 15,000 years ago (de Beauclaire, 1971:31). Permanent
settlement and culture building can only safely be placed as occurring
some 5,500 years ago.(Chang, K.C. 1977:85-91) While fishing appears to
have been the primary occupation, it is possible that agriculture was
occurring, some one thousand years earlier than in north China. (Gates
1988:28-29)

The second major wave of migration originated in south China,
particularly from the provinces of Fukien and Kwangtung. The migration
of Chinese to the island took place at a much later date. It is known
that migration began as early as the seventh century. Closer attention,
particularly by officials of the mainland government did not happen
until the sixteenth century. In 1564, in response to an increasing
number of attacks on shipping and coastal settlements by Taiwan based
pirates, the island was claimed as Chinese territory by the Emperor, and
a garrison established near the present day city of Tainan. An upsurge
in migration and eventually trade followed. (Gates 1988:28-32)

Taiwan itself offered relatively rich alluvial plains for
agriculture, as well as physical proximity to the mainland. More

importantly it offered the promise of land. "These motives brought a

l6

flood of immigrants to Taiwan during the seventeenth, eighteenth, and
early nineteenth centuries".(Wickberg 1981:212-213). Village building
was brought as a tradition, but was also a recognized necessity.
Conflicts with aboriginals grew as the Chinese influx did. Groups of
migrants banded together in fortified villages for the purposes of
common defense.

"From the beginning of large-scale Chinese immigration

in the seventeenth century, most peasant migrants were

firmly embedded in an embryonic capitalist economy run by a

class of state-aided entrepreneurs... Because Taiwan's

economy was built on land speculation and exports, the use

of money and the habit of calculating the costs of land,

labor, and goods permeated the way of life of settlers even

in remote valleys." (Gates 1988:35-36)

This sophistication continued throughout the colonial era and into the
present day.

While a capitalist economy was developing under this great wave of
immigration, spatial expansion and eventually city building was
occurring also. The remainder of this chapter focuses on a discussion of
the formation and history of urban life in the Taipei Basin. A primary
source is the work on the history of the Island by Chiang and Huang
(1985).

The spatial expansion of the population eventually led to the
northern section of the island. Much of this area is dominated by the
Taipei basin, which is an alluvial basin, triangular in shape, drained
by the Tamsui River and its tributaries the Keelung, the Hsintien and
the Tahan Rivers. Sometime between 1723 and 1735 the first small Chinese
settlement was established on the Tamsui River. This village was named
Mengchia and it was to become part of the core of the city of Taipei.

The village of Mengchia grew into a market town, aided by the

navigability of the Tamsui, downriver to the ocean, as well as to

17
interior markets. Originally much of the trade was with aboriginals. It
was supplemented by commerce with the Chinese, Dutch, British and
American traders located downriver in the village of Tamsui.

During the next century population in the basin continued to
grow. By the early 19th century the remaining aboriginal population had
been forced from the basin to the surrounding, and less desirable,
upland areas. During the period from 1821 to 1850 the basic form of the
city was established. Mengchia, with a population of 18,000 in 1853 was
the third largest city in Taiwan, after Tainan in the south and Lukang
in the center.(Chiang, Nora, 1985:191)

In the following decade conflict between rival immigrant groups
resulted in the defeated group to moving downstream to the north, where
they built a new market town known as Tataocheng. Meanwhile,
sedimentation of the Tamsui River reduced Mengchia's importance as a
port and it was superseded by Tataocheng. The opening of Tamsui port to
foreign traders, as required by the Tianjin Treaty of 1860, brought more
Chinese traders to the area and saw an increasing number of foreign
traders. Due to the rapid silting of the Tamsui River, with a short
period the whole course became unnavigable for large junks. Keelung, a
deep gulf with a rocky shore, 26 km. from the inner city was selected as
the trade outlet for the Taipei Basin.(Chiang, 1985:191 and Hsieh,
1964:234)

In 1887 Governor Liu Mingchuan decided to take advantage of the
site advantages of Tataocheng and to develop it as the commercial center
for Taipei. American, German and Dutch consulates were established, and

the Ching government's efforts to develop Taiwan accelerated its growth.

18
The outward-oriented character of Tataocheng would eventually help to
shape Taipei's development.

The third sub-city of the Taipei Basin was called Chengnei,
meaning ‘inside the walled city'. This area lies east of Mengchia and
south of Tataocheng. It was developed some 25 years later than
Tataocheng and some 70 years later than Mengchia. It basically
constitutes the area lying between the two sites, and was perhaps chosen
as a compromise between the existing rivals.

This area became the focus of the new provincial government

(a prefecture in 1882 and Provincial status in 1885) and in

1882 a new walled city was constructed as the administrative

center. This area inside the wall was constructed along

'modern' lines with government buildings, broad boulevards,

electric lighting and modern communications such as the

telegraph. The emerging city of Taipei was becoming not only

the political center of the province, but the communication

and trade focus as well." (Chiang and Huang, 1985:188-194,

Chiang, 1985:193, Hsieh, 1964:234-235)

The appearance of the walled city was significant, for this
represented the ideal form of the Chinese city. The idealized form of
the Chinese city is a well researched phenomena by such authors as Chang
Sen-dou (1963,1970), Elvin and Skinner (1974), Skinner (1977) and
Wheatley (1971). While the walled city was never fully attained in
Taipei, it is important to note its nascent development. The designation
as a prefecture (to be followed closely by the granting of Provincial
status) was important to Taipei, and the walled city was central to the
Chinese conception of urban space.

Much of the urban character of China and East Asia was
established centuries before Westerners came East. ...cities
began to evolve along precise administrative and ceremonial
lines. The spatial structure of the Chinese city evolved to
reflect the function such cities played in Chinese society,
and the spatial organization which created them. A
distinctive feature of the classic Chinese city was the city
wall. The wall surrounding the city was a primary defense

feature, but also served as a line of demarcation between
the urban and rural regions... in fact, the wall was a prime

19

feature that distinguished urban centers from lower order
urban places.

The wall however, was only one feature of the

idealized form of the Chinese city. The form of the city

aimed at reflecting the Chinese view of the universe and the

role of the emperor as the intermediary between heaven and

earth...in reality, such as in Taipei, the ideal was seldom

attained. Particularly in lower order cities this form

seldom existed in totality. Topography often prevented the

rectangular form. Population growth would produce spillover

into suburbs. Such suburbanization often dissipated the pure

form of the city, sometimes resulting in cities with several

walls each enclosing succeeding rings of suburbanization.

(Williams, 1983)

The walled city in Taiwan never achieved the development of major
mainland cities. The first walled city on Taiwan did not appear until
1704, and there remained relatively few years before the Island was no
longer under Chinese administration. Nevertheless some 43 walled cities
were built on Taiwan or in the Pescadores. In Taiwan the shape of most
walled towns were irregular and small in size. The largest wall, in
Tainan, was only 7 kilometers in circumference. However, the major
walled centers served as administrative, market, political and cultural
centers. Chengnei, with its status as Prefectural and then Provincial
capital, became the primary urban center. Chengnei, as was true of all
walled settlements in Taiwan, grew faster than settlements without a
wall. In most walled cities when the walled space was filled, the wall
became a limiting factor to growth, and began to exercise a significant
influence on street patterns, density and land use. This never was the
case with Chengnei, first because of the 'joining' of the original three
cities through growth, and secondly because of the change in ownership

of Taiwan and the subsequent influence of the Japanese. In short,

Chengnei never had the time to fully develop under Chinese influence,

20
and its siting between two existing cities was unique. (Chiang
Tao-Chang, 1966:53)

The elevation of Taiwan to provincial status, in 1885, was a
reflection of the growing economic importance of the province to both
China and ‘colony hungry' imperialistic powers. The move to keep Taiwan
within the domain of China ultimately failed. The Treaty of Shimonoseki,
which concluded the Sino-Japanese War in 1895, ceded Taiwan to Japan.
This marked the beginning of the third significant wave of immigration
to the island, that of the Japanese.

This migration occurred between 1895 and 1945, during the period
of Japanese colonial rule. The colony attracted several hundred thousand
Japanese immigrants. Although nearly all of these people returned to
Japan at the conclusion of WW-II, their influence on culture, language
and urban design and planning remains significant.

The Japan which controlled Taiwan was a radically different Japan
than had existed some thirty years earlier. The Mejii restoration in
Japan had come about as a result of a new alliance between the Samurai
class and that of the merchants. In the thirty odd years between the
restoration and the colonization of Taiwan, Japan had undergone dramatic
change. The society had transformed itself from a technologically
immature, feudal, agricultural nation, into a strongly centralized and
rapidly developing country that was successfully contesting with western
nations for influence in the Pacific basin. Japan was driven by many of
the same factors that drove western colonial nations, in particular the
need for both raw materials and markets. Taiwan became not only a source
of raw materials and goods and a market for Japanese produced items, but

a demonstration project aimed at showing western powers that Japan had

21
arrived. This attitude was clear in many of the writings of the day. For
example, Yosaburo Takekoshi, a leading historian of Meiji Japan wrote in
the introduction of his book about Taiwan, ”The white people have long
believed that it has been the white man's burden to cultivate the
uncivilized territories and bring to them the benefits of civilization.
The Japanese people now have risen in the Far East and want to
participate with the white people in this great mission." (Takekoshi,
l907:iv.)

The Japanese government and industrialists worked hand in hand in
developing a wide range of projects largely based on local raw
materials. "While business profits went mostly to Japan, much of the tax
collected in Taiwan was spent there on the infrastructure and
administration that had made increased productivity possible. For the
Taiwanese peasants, life became much safer, healthier, a bit more
comfortable." (Gates, 1988:41)

The Japanese made the emerging city of Taipei the colonial
capital. The Japanese were strongly influenced by the more
technologically advanced West, and envisioned a Showplace city in the
western tradition. (Chiang, 1985:195) In 1898 the Japanese government
created a city planning committee. It proceeded to demolish the city
wall (although the gates were kept as historical landmarks) replacing it
with a three lane highway, and expanded the old city of Cheng-nei
westward and connected it with Meng-chia (soon to be renamed Wanhua) and
Tataochen. Transportation links were also improved between the central
city and both Yang Ming Shan and Peitou to the north. Yang Ming Shan
functioned much like the hill stations or residences of the British in

India. It offered the Japanese elite an alternative residential location

22
for the hot, steamy summer months in the cooler higher elevations, but
still within reach of the commercial center. Peitou, with its hot
springs and sulphurous mud baths became a favored retreat.

In 1905 the Japanese established the first large scale city plan
for Taipei. The plan incorporated the original three villages and
surrounding territory into a 1,000 hectare city with a planned
population of 150,000. (Chiang and Huang, 1985:195) Although the city
plan envisioned a 1,000 hectare city, actual development was
concentrated in the old inner city area of Cheng-nei where the colonial
government and Japanese population were located. In this district were
built government and financial institutions (in Meiji style
architecture) and a new commercial center arose in the area fronting on
the central railroad station. To the east of the original core, high
class residential districts for the Japanese(largely consisting of
traditional Japanese style houses) were established. The native
Taiwanese were largely confined to the remaining districts, in a pattern
that saw complete mixture of social classes as well as land uses. All of
this activity was facilitated by the typhoon which struck the Island in
1911 and destroyed much of the native housing. This native housing,
built of brick and clay, was replaced with construction of concrete and
steel, in the western style. (Hsieh, 1964: 235)

"Population growth exceeded the estimate [made in

1905], and by 1913 the target population had already been

reached. In 1904 Taipei was the largest city in Taiwan with

a population of 85,890. In 1932 the population of Taipei had

reached 281,852 and a revised city plan was announced which

envisaged a population of 600,000 by 1955. The city limits

were expanded to include a total area of 6,676 hectares

(66.76 sq.km.), with the inner city as the focus. A modern

road system was drawn into the plan. The Taipei City Plan

Act was issued in 1936 in order to implement the above plan,

and this was the beginning of city planning legislation in

Taiwan. However, a lack of balance could be detected in the
plan: development efforts were concentrated in those areas

23

populated by Japanese, while Taiwanese populated areas were
more or less ignored." (Chiang and Huang, 1985:195)

By the end of the Japanese colonial period much of the character
of present day Taipei was set. On a macro-scale the gridiron pattern of
streets was in place. While on a micro-scale this is often not apparent,
with land use and space combined to an extent not apparent in the West,
and patterns often having no regularity as a result of the presence of
older sections of the city or areas where development was done with
little regard for plans, the overlying pattern is still the gridiron
pattern of Japanese colonialism. In large part the Japanese plans still
dictate growth and patterns as one author states "Even today, Japanese
colonial structural plans are being followed in the layout of road
networks, parks and public lands, and drainage systems in many towns and
cities". (Pannell, 1973:4) Beyond the physical layout of the city,
social institutions, law, education, in fact the full spectrum of life
had been influenced by the Japanese. Certainly the Meiji style of
architecture dominated the core regions of the city, and traditional
Japanese style houses the upper-class residential areas. But more
importantly, the city reflected a blending of styles and cultures that
still impacts today.

But it must be noted that the style, while having similarities to
Japan and to other cities in the region, was unique. Much of the
physical layout of the cities was apparent in Japanese cities of the
modern era. The newly developing parts of Tokyo, to the North, West and
South, reflected many of the same characteristics in building style and
gross pattern. However, the population of Japan has always been almost
entirely Nipponese. In Taipei the two ethnic groups were separated, both

socially and geographically (at least at the upper levels of society).

24
This, of course, was not evident at the same scale, in Japan. In
addition, by the 1930's Japan was exhibiting a residential pattern as
diverse and identifiable as the one in U.S. cities of the same period.
Social and residential groupings were clearly identifiable. (Ueno, 1985)
This was not the case in the ethnic Chinese neighborhoods in Taipei,
where indications are that residential groupings were largely
undifferentiated, as least as regards income or stage in the family life
cycle. (other mechanisms may have been at work. This is discussed later
in the paper;also see for example Gallin and Gallin, 1974)

The end of the Second World War in 1945 saw the beginning of the
last great migration to Taiwan. October 25, 1945, known in Taiwan as
Retrocession Day, marks the return of Taiwan to Chinese control. The
mainland of China was in transition at the time, from a war primarily
with the Japanese, to a resumption of the civil war between the
Nationalists (Kuo Min Tang or KMT) led by Chiang Kai-shek and the
Chinese Communist Party (CCP) under the leadership of Mao Tsetung. At
the time of Retrocession the Nationalists were the recognized legitimate
government of China. As such control of the Province passed to them. In
December of 1945 troops dispatched by Chiang, and led by Chen Yi arrived
to garrison the island. Analysis of the period immediately following the
war depends upon your point of view. It is generally agreed that the
arriving troops were initially welcomed. The welcome however was short
lived as the troops began to run rampant over the Province in an orgy of
looting, stealing and raping. Some historians feel that the Taiwanese
were treated, in the words of one recent historian, as ‘slaves without a
country'. (Shi, 1979: 706). Other historians view the actions, for the

most part, as justified, particularily as many of the largely peasant

25
army saw Taiwan, with its recent Japanese history, as hostile territory.
In any event, sharp divisions between the Taiwanese and the newcomer
Mainlanders soon occurred. Tensions became so great that on February
28,1947 (known in Taiwan as 2:28) the Taiwanese, led by the middle class
that had arisen under the Japanese, revolted against the Mainland
presence. The revolt was short lived, a fact that was the result of the
superior military might of the Mainlanders, as well as promises of
redress made to the Taiwanese. The cessation of hostilities by the
Taiwanese did not mark the end of the incident. To firmly establish
their control, the forces led by Chen Yi quickly rounded up between ten
and twenty thousand Taiwanese. Those arrested tended to be the liberal,
educated, and generally pro-Japanese middle class. These people were
executed by the Mainland forces. Divisions between the two groups were
now firmly established. They were further deepened in 1949 and through
the early 1950's. In 1949 the KMT were defeated by the CCP forces on the
mainland. Chiang Kai-shek and the remnants of his forces, (upwards of
one to two million individuals) retreated to Taiwan. The arrival of the
defeated KMT forces resulted in widespread confiscation of land and deep
social cleavages. The distance between the Taiwanese and Mainlanders was
further complicated as Chiang consolidated his position on the island by
insuring that any remaining potential threats to his political position
were either eliminated or rendered ineffective.

The arrival of the Mainlanders brought deep social divisions to
the island. Also at peril with their arrival was the economy, which was
already devastated by normal wartime inflation. This was further
complicated by the fact that a yen based economy was being forcibly

recast into the Chinese system. The conversion to the Chinese system was

26
complicated by the fact that little faith remained in currencies being
issued by the KMT which was notorious for the inflationary printing of
money to pay its debts. "Between the wartime inflation of Japanese money
and the unprecedented Nationalist expansion of currency, staple foods
that cost one yen in prosperous prewar 1937 cost 72,262 units of Chinese
currency by May 1949, and building materials inflated from one unit to
93,563 over the same period. Between January and June of 1949, the cost
of living rose ten times." (Gates, 1988, 51) If this was not enough
other complications existed. The retreat to Taiwan was seen as temporary
by the KMT, it was not expected to last any longer than it took to
regroup and then regain control over the mainland (still the official
policy of the Government). As a result of this policy, the KMT continued
to field an extraordinarily large military force. Much of the Government
revenue from both domestic and foreign aid sources went to support this
structure through at least the beginning of the 1960's. This put
extremely strong inflationary pressures on the economy. The state of the
economy was further complicated by the inefficient and sometimes crooked
management of state run enterprises (monopolies) in such commodities as
rice, sugar, oil, tobacco and liquor; as well as by the destruction of
the native Taiwanese middle class of entreprenuers.

Under pressure by the Americans, Chiang agreed in the early 1950's
to institute a land reform program long sought by some elements within
the KMT. This reform, completed by 1953, restricted farmland ownership
by individuals to three hectares (7.2 acres). Land over this amount was
bought from the owner and sold to families having less than the
prescribed amount of land. The former owners of the land were not paid

in cash but in government securities or stock in state run enterprises

27
(thus not only conserving capital, but also avoiding the potential for
problems if currency in which people had little faith were forced upon
them); new owners were allowed to finance the land by accepting ten year
mortgages at an annual rate of 37.5% of its annual crop value.(see Chen
1961, Yang, 1970) In the short term this move did improve the lives of
the rural dwellers. Production of agricultural products increased and
capital was created. Surplus capital was reinvested in the farm or in
off farm enterprises. By the end of the 1950's some improvement had been
seen in the economy. However, the land reform also had unanticipated
results. The small size of the farms restricted mechanization. This
meant that family labor continued to be the most important input into
the farm. This combined with improved health care and diet led to high
growth rates for the population. Much of this population soon became
surplus to the rural economy, constrained as it was by a finite amount
of land. While this surplus labor was to form the basis of Taiwan's
transformation to an industrial nation, it also created many new
problems as labor left the countryside to seek opportunities in Taiwan's
cities. When Retrocession occurred in 1945, Taipei first reassumed its
status as a provincial city. However the provincial capital was soon
moved to Tainan and the city became the ‘temporary' capital of the
Republic of China (R.0.C.).

At the time of the coming of the Mainlanders the city consisted of
ten administrative districts covering an area of 66.98 sq.km. Post war
city planning for Taipei began immediately after the war, in 1945.
”Based on the 1932 plan, with very few adjustments, five land-use types
were distinguished: residential, commercial, industrial, agricultural

and mixed. By 1954, 43.9% of the total planned land had been put into

28
use." (Chiang and Huang: 1985:196) Between 1954 and 1967 Taipei
continued to grow. Fueled by the surplus labor from the countryside, as
well as by natural increase, the city grew rapidly. Not only did the
city grow, but the surrounding cities and townships did also. The
territory lying west of the Tamshui River, cities such as Sanchung, also
experienced rapid growth. At the height of the Japanese period, in 1944
the population of Taipei had reached 401,497. The immediate postwar
period saw a dramatic decline in the population as some 100,000 Japanese
were repatriated. The influx of Mainlanders, however, soon made up for
this loss, and by 1950 the population had reached 503,450. From 1950,
until the late 1970's the population expanded at over 5% a year. Since
that time the rate has hovered around 3%. At any rate, by 1963 the
population of Taipei had reached one million in the city proper and
close to two million in the metropolitan area (Williams, 1988:177). This
doubled during the following ten years.

The growth of Taipei was reflective of a major residential shift
occurring nation wide. When Taipei became the 'temporary' capital of the
Nationalist government, Taiwan was primarily an agricultural society. In
1950, only about 25% of the population of the island lived in urban
places. Williams, 1988:177) During the following decade the population
of the island increased to 10.8 million (DGBAS, 1980). In the same time
period land reform began to raise rural incomes, and industrialization
began a sustained period of explosive growth. The labor to fuel the push
to industrialization came largely from rural to urban migration. This
migration process is well documented by researchers such as Tsai (1986),
Liu, P. (1982, 1983), Tsay (1982), Chiang (1984), and Speare (1973,

1974, 1988). The migration from rural areas in the 1950's on had as its

29
destination the large cities and their satellites. The Taipei
Metropolitan Region was the destination of many of these migrants,
leading to explosive growth. ”For example, Panchiao, Yungho, Sanchung,
Chungho, and Hsinchuang, all satellite cities of Taipei, experienced
explosive growth. Chungho alone increased 1300% from 20,000 in 1956 to
nearly 300,000 by 1982. Panchiao swelled 967%, Hsingchuang 851%.
(Williams, 1988:180) By 1980, 66% of the population was urban.

This growth in urban residents and in the residents of Taipei
however is not attributable only to migration or natural increase. In
1967 Taipei was elevated to the status of an Executive Yuan (Council)
administered city and its borders were redrawn to include sixteen
districts and a total area of 272.14 sq.km. At the time of the boundary
changes the population was 1,604,543. In the 1980 census the city had a
population of some 2,267,584 in the city itself and about twice that
number in the metropolitan region. (Chiang and Huang, 1985:195-199)

This growth in population is reflective of the economic
transformation of the country. The Taiwanese economy which had been
stagnant in the 1950's finally began to take off in the early to mid
1960's. This dramatic change within the economy was due to a variety of
factors. Changes in domestic leadership as well as fears of being
surpassed by the then plausible gains of the ‘Great Leap Forward" on the
mainland were certainly factors. Equally important was a boom in
foreign investment attracted by KMT assurances that Taiwan offered not
only cheap labor, but labor that was forbidden to strike. This foreign
investment was further attracted by the establishment of export
processing zones which served to maximize their profits while giving

Taiwan the benefit of the transformation of an agricultural workforce

30
into an industrial one and providing wages to stimulate the local
economy at the same time. In any event Taiwan soon established itself in
such export sectors as textiles, footwear, plastics and electronics,
particularily to the U.S. with which it had Most Favored Nation (MFN)
status. The final push that took Taiwan from the take off stage into a
developing economy came with the Vietnam War. Taiwan greatly benefitted
from production contracts for military goods, repair facilities,
building products sales, increased U.S. military presence, as well as
being a major R&R site for American troops. The money generated by these
activities created a tremendous pool of capital, which when combined
with the entrepreneurial skills and hard work of the Taiwanese, provided
the means for Taiwan to join the ranks of the Newly Industrialized
Countries (NIC's).

It has already been seen that population in Taipei grew
dramatically as a result of economic changes in the society. During the
postwar period Taipei also saw dramatic changes in other areas. The
ethnic makeup of the city also changed significantly during this period.
During the colonial period, immigration from the mainland of China to
Taiwan was severely restricted. Thus Chinese inhabitants of Taipei
largely consisted of those whose ancestors, or who themselves, had moved
to Taiwan prior to 1895 (this is the group that is today known as the
Taiwanese). This group accounted for some 65% of the city in the pre-war
period. Japanese accounted for 25-35% of the population, while the
remainder consisted of Chinese mainlanders, those who had moved to the
country after 1895 (and who tended to be urban dwellers, given the lack
of available land in the rural areas). The postwar period saw the

repatriation of the Japanese, whose place was soon taken by the

31
retreating mainlanders. By 1950, mainland Chinese made up one third of
the city's population. As the city expanded, a new group became a
significant sector of the population. The rural to urban migration
fueled by rural birth rates and labor surpluses, as well as by increased
urban opportunity, saw the rural migrants account for 35% of the
population by 1975. (Chiang and Huang, 1985:190-198) (DGBS, Various
years)

The population and areal expansion of the city were accompanied by
significant changes in population densities as well. In general there
has been an increasing outward movement, or suburbanization, of the
population. Aging housing in central districts, combined with increased
land values has prompted many to move to outlying districts or to
satellite cities within the metropolitan region. This movement has been
facilitated by improved transportation networks, as well as by the
dramatic increase in privately owned motorcycles and cars brought on by
economic success.

The outer suburbs and satellite districts are also the destination
of many migrants. Shut out of the center city by high land prices, they
are forced to compromise, often living far from work in order to obtain
reasonably priced housing. This urban sprawl with its changing density
gradients was first reported in a study that theorized that urban
population density gradients for Taipei exhibit patterns characteristic
of Western cities. This was a hypothesis that reflected the growing
transformation of the city. "Unlike some Western cities, Taipei...was
not structurally formed around a purely commercial central business
district. Instead a primary commercial-cum-residential area which

performs the function of the western central business district developed

32
with a juxtaposition of shop and residence ... primarily due to
traditional business organization...(which is)...small in scale and
owned, managed and operated largely by members of a single family" (Liu,
1986: 5) This traditionally led to extremely high densities in the
central city area. In any event, the study found that from 1950-1972 the
density gradient slopes declined continuously throughout the study
period, and the central densities first increased and then decreased.
(Graff, 1976) The trend towards suburbanization was reaffirmed by a
study that demonstrated that high growth rates occurred at a distance of
5 km. from the city center from 1963-1968, at 6.5 km. from 1969-1974,
and 10.5 km. from 1975-1979. (Chen, 1981)

Unfortunately, few urban or social geographies, or even a
comprehensive description of the modern metropolitan region exist. The
most current description of the city is contained in a map of land use
patterns for the city as of 1983. The map includes historical and
regional maps as well as a textual discussion of land use in the Taipei
Basin. (Williams, Sutherland and Chang, 1988). While Taipei is a major
metropolitan region, it has attracted relatively little notice in the
literature outside of Taiwan. The best sources on the city are Chinese
language sources published by scholars in Taiwan. Unfortunately, many of
these are uncirculated research or technical reports which do not reach
a large audience. Some of the more valuable ones are included in the
Bibliography. Beyond that one must depend on English language sources,
and luckily many Taiwanese do publish in English, and most Chinese
publications include an English language abstract. However, as stated,

the literature in English or Chinese is not that extensive.

33

Even though the range of the literature is limited, there are a
number of studies that do have bearing on this one. Clifton Pannell's
(1973) study of Taichung City in Taiwan provides a good overview of the
development of a major Taiwanese city; especially of the influence of
both traditional Chinese styles and Japanese colonialism. Pannel (1974)
has also contributed to the general understanding of urban land
consolidation and city growth in Taiwan. This aspect of urbanization was
also addressed by Lin (1980). The operation of the real estate markets
in Taiwan, and the development of urban housing has been addressed by
Lee (1979), and Hsung—Hsiung Tsai (1988). Population distribution,
structure and migration have been the focus of a number of researchers.
Among the more notable authors who have delved into these phenomena are
Speare (1973, 1974), J.L. Li (1983), and Li and Tsai (1988). Economic
development and labor migration are important elements in the growth of
urban areas. This process has been the focus of Paul Liu (1979, 1982,
1983), Speare (1974), Hang-chin Tsai (1986), and Sun and Tsai (1980).
Lin Tse-t'ien (1973) has written on the development of industry and
cities in Taiwan; while Li Jui-lin (1973) has contributed significantly
to understanding the structure and development of Taiwan's cities. The
recently published book by Speare, Liu and Tsay (1988), provides a good
overview of the rural to urban transition that Taiwan has undergone over
the past several decades. The development of urban and regional planning
has drawn the attention of Lin Tsu-yu (1973), C.Y. Wang (1988) and
Williams (1988).

Taipei itself, as mentioned, has drawn limited attention. Among
the studies that bear on this work are Li Hsun-feng's (1983) analysis of

industrial land in the Taipei Metropolitan Area, Chin-lung Tsay's (1982)

34
investigation into migration and population growth in the city and
Ya-mei Chiang's (1984) look at migration and residential decision making
in the suburban city of Sanchung. Several studies have a direct
application to this work, in helping to establish current patterns and
behavioral processes. Among these are Hsueh-tao Chien's (1976) study of
the structure and distribution of commercial areas in the city, Tou-chin
Lin's (1983) look at the process of industrial location in the city,
Jing Meng's (1982) research on the spatial variations of population
growth and distribution, and Hsu and Pannell's (1978) investigation into
the residential and social structure of the city.

The best overall source of material on all aspects of
Taiwanese history and life (in English) is the comprehensive
bibliography on English language publications compiled by Bruce Jacobs
(1984).

Within the broad context of urban research on Asian cities,
studies that apply to Taipei's situation are also sparse. The best and
most comprehensive study of traditional Chinese urban form and structure
is, without doubt, Paul Wheatley's (1971) landmark work on the origins
and character of these cities. Tao-Chang Chiang's (1966) study of walled
towns in Taiwan provides a geographic perspective on the distribution
and realization of such urban forms in Taiwan. Another valuable source
on gaining a broad understanding of Chinese urban form and experience is
the three volume study of Chinese cities by Skinner (1971,1974) and
Elvin and Skinner (1974).

In terms of systematic investigation into urban growth and
structure, and attempts at theory building, the best work has been

concentrated in South and Southeast Asia. The most extensive work with

35
applicability to Taiwan has been done by T.G. McGee (1967, 1971) who has
investigated both the urbanization process in the third world, as well
as in Southeast Asia in particular. Before examining the applicability
of Southeast Asian urban models it is important to reiterate that the
primary influence on urban development in Taiwan was (and to some extent
remains) Japanese. "Morphologically and in physical design, the larger
cities are in part Japanese. Even today, Japanese colonial structural
plans are being followed..." (Pannell, 1973:4). Pannell and others have
found McGee's Southeast Asian models of some use in understanding
Taiwan's cities, but the differences far outweigh the similarities. Up
until relatively recently the major cities of Southeast Asia developed
along the lines of what McGee termed "the colonial city in Southeast
Asia" (1967). These colonial cities emerged in the nineteenth century
and were for the most part, port cities having a variety of functions.
McGee describes them as the "...foci of colonial control and domination.
They were also administrative, processing, and transportation foci, but
it was largely their economic function as the focal point for the
collection of raw materials and the distribution of imports and exports
which led to their massive growth and dominance in the urban hierarchy
which persists today" (1967: 22). It is clear that no city in Taiwan has
dominated the economy and urban hierarchy of the country in the same
manner as does these Southeast Asian cities. (Pannell, 1973: 5)
"Although Taipei is without question the key urban center, there is in
fact a relatively balanced hierarchy of secondary and tertiary cities
below Taipei, in a pattern that is close to normal rank-size
distribution." (Williams, 1988: 179) In addition Taipei never really

functioned as a major port city, delegating that role to the city of

36
Chilung located about 15 miles to the northeast. While Taipei was the
center of Japanese colonial administration and an important commercial
and industrial center, it shared its economic functions with both
Chilung and Kaohsiung (in the south). In addition, the Japanese were
late entries into the colonial picture and never fully developed the
commercial aspects the way other countries did (Pannell, 1973: 5).
Finally, it must be noted that when the Island was returned to the KMT
at the end of WWII, the Provincial capital became Taichung. While Taipei
eventually became the focal point of the national government, all
governmental functions were not focused in one location. This
decentralization of functions is important, for administrative and
military functions appear to have been major determinants of urban
growth since the Japanese period. For example during the colonial era
the city of Taichung grew much more rapidly (and eventually surpassed)
its central Taiwan rival Changhua, the traditional commercial and
transportation center in the region (Pannell, 1973: 5).

While it is true that Japan exploited Taiwan for its primary
products, in much the same way that occurred in Southeast Asian cities,
and invested little in the industrial sector (Barclay, 1954: 18-42), the
situation is different than that described by McGee (1967). McGee sees
Third World countries, and Southeast Asian nations in particular, as
having "inherited their economic structure from an era of colonialism
characterized by an excessive specialization in the production of
materials for the industries of the metropolitan powers. (And) Despite
independence,... still closely linked and dangerously dependent upon the
industrial powers..." (McGee, 1967: 22). While it has been stated that

there are similarities, Barclay (1954), among others, makes it clear

37
that Japanese investment in the transportation and power sectors set the
stage for both industrialization and a more balanced regional growth.
(Barclay, 1954: 18-42), (Pannell, 1973: 4-7).

McGee (1967, 1971) has also described the explosive growth in
Third World urbanization. This growth is seen as occurring as a result
of both natural increase and in-migration. On the whole McGee felt that
natural increase was the largest contributor to the rapid growth of
urban areas. While rural to urban migration is occurring in very large
numbers; McGee felt that such migration merely raised the level of
growth in the city to the equal of that in rural areas (McGee, 1971:
35-58). In addition, the rapid growth of population created an ever
increasing under-class in the rural areas that in the future would push
rural dwellers toward the city as lack of opportunity in the countryside
increased. The failure of industrialization in the city would then mean
the city could not adequately absorb these residents and particularly
not in addition to those born in the city (McGee, 1971: 35-58). Lastly
he felt that any growth in the middle class would not serve to more
fairly distribute income, but rather serve to concentrate it even
further (McGee, 1971: 35-58).

The Taiwanese experience has been much different than that
described by McGee. Even with the colonial experience, and the influx of
the Mainland Chinese after 1949, only 25% of the population lived in
urban places (Williams, 1988: 177). "In the 1950"s, however, island-wide
urbanization began to take off as a result of industrialization..."
(Williams, 1988: 177). During the 1950's the island's population grew by
4.4%, but urbanization increased at an annual rate of 12.6%; obviously

rural to urban migration was playing a significant role in urban growth.

38
This is confirmed by the experience of the 1960's. During this decade
total population growth slowed to 3.2% a year, while urban growth
continued at a rate of 8.8% a year. By the 1970's the country's
population growth rate had slowed to 2.1% a year, while cities continued
a relatively high growth rate of 5.1% a year. By 1980 some 66% of the
total population lived in cities (Williams, 1988: 177-179). As cities
were growing much more rapidly than the nation as a whole it is obvious
that rural to urban migration has played a vital role in city growth. In
addition McGee cites Davis's (1965) contention that under-developed
countries gains in the proportion of urban population was 20%, compared
with an average gain of 15% for industrialized countries during the
decades of their most rapid urbanization (McGee, 1968: 17). The rates in
Taiwan place it much closer to the Western experience that the general
experience of the Third World. Taiwan, in fact, has experienced what has
been described as an 'orderly urbanization' (Myint 1981; Liu 1979, 1983;
Li 1983; M.C. Chang, 1984; Williams, 1988).

Unlike the situation described by McGee, Taiwan's rural sector has
not become a persistent under-class. In the earlier years of
urbanization Taiwan did experience a widening gap between rural and
urban incomes; an experience heightened by the fact that most migrants
tended to be young people. By the late 1970's, as a result of direct
government action, the outflow from rural areas had declined (Williams,
1988: 181). Even with the rural-urban gap in incomes it is opportunity,
not poverty, that has led to much of the migration. "...rural-to-urban
migration in Taiwan has not been the result of real poverty in rural
regions but rather a response to better employment opportunities in

cities. Taiwan thus differs significantly in this regard from most other

39
developing countries. The majority of migrants were better educated
young adults of both sexes" (Williams, 1988: 181).

Taiwan has also not experienced an ever increasing gap between a
large under-class and small middle and upper class as described by
McGee. In fact, "Taiwan has managed to achieve a fair degree of social,
as well as spatial, equity in the distribution of its wealth" (Williams,
1988: 175).

The last element of McGee's model that on the surface seems to
have applicability to Taiwan is the residential pattern. McGee describes
cities that generally were made up of an indigenous minority, a small
Western colonial upper class of administrators, and a large middle group
of laborers, commercial people, and petty administrators from other
regions of Asia. He describes these groups as clustering in ethnic
quarters, with interaction determined by commercial linkages (McGee,
1967: 52-75). This is a view that was upheld by Berry and Rees (1973) in
their investigation into Calcutta. Taiwan's experience was quite
different. As Pannell notes:

"...in Japanese dominated Taiwan, while there existed
enclaves of mainland born Chinese or other nationalities,
the colonial cities and towns were basically composed of two
groups-native Taiwanese... and Japanese colonials most of
whom were expatriate administrators, professional, or
commercial types. Between the two groups distinct social
cleavages emerged mainly based on wealth, status, and
privilege,... Undoubtedly, while certain exclusive enclaves
of Japanese citizens existed, there was also considerable
contact and interchange among lower income Japanese who
could not afford that luxury of living in the Japanese
quarter and the indigenous Taiwanese. Indeed the evidence
from Taipei, which indicated that many of the city's
administrative districts had a Japanese population of
between 10% and 50% of the occupants would suggest more
association among the groups than is generally conceded.
and there occurred much mixing at a lower income level among
the various national groups. On the other hand, in the
Southeast Asian city, the Asian groups were economically
associated to some extent, but the colonial European group
was sharply segregated...The conclusion may be that the

40

ethnic element was less conspicuous in Taiwan...whereas in

Southeast Asia, not only were several national groups

involved, but several racial groups as well, and the

percentage of purely colonial Europeans was indeed a very

small percentage of the total" (Pannell, 1973: 43-45).

In short then, the experience of Taiwan is very different than
that experienced by Southeast Asia and much of the rest of the
developing world. While McGee's model offers a useful foil and
comparative device, it does not explain the development of Taiwanese
cities. At the same time the view of Reissman that "industrial urban
development in the West and in the underdeveloped countries today is the
same process although greatly separated in time and place" (Reissman,
1964: 165-166) does not fit either. While Western experience can shed
light on the development of Taiwanese cities, the colonial experience

and Chinese culture add elements that make the experience different and

perhaps unique.

CHAPTER THREE

THE THEORETICAL AND ANALYTICAL FRAMEWORK
R E SS AL ODEL

Investigation into the spatial patterns and processes of urban
areas has a relatively long tradition. In the mid-nineteenth century,
Kohl (1841; in Berry and Kasarda, 1977:108) devoted an entire chapter to
the internal structure of the city. Urban structure, he felt, might be
viewed vertically as a series of layers with the ground floor containing
businessmen, the first floor, areas of wealth and pleasure, and the
upper and subterranean levels having the lower income residents.

Kohl's observations were a precursor to a whole body of work
dealing with the spatial structure of the city. However, it was not
until the 1920's that significant work in this area began to be
produced. Inspired by the work being done in plant ecology at the
University of Chicago, a number of researchers began to examine urban
spatial structure from an "ecological" approach. Robert Park in his
essay "Human Ecology (1936) illustrated the nature of this approach. He
talks about "competition" between various populations in the metropolis,
the ”dominance" of one group or another in the functional areas of the
city, and of “invasion" of the area by competing groups and "succession
and dominance" by the new group. These urban ecologies, Park felt,
result from population increase, and were expressed in a series of

concentric rings around the city that formed over time. Based on

41

42
empirical observation of the city of Chicago, he identified a process of
social mobility involving geographic mobility, wherein the longest term
residents would move from core areas to the periphery of the city, and
new in-migrants would fill the abandoned areas.

This process was given formal statement by Burgess (1925) in his
concentric zone model. "The model is crude and unrefined, but it
provided a set of ideas about urban spatial structure which could be
empirically tested, and a framework for more detailed study of natural
areas within the city" (Berry and Horton, 1970:307). The studies that
were produced in initial reaction to Burgess' work, focused on the
"disorganized" communities of the city and the real and immediate
problems of poverty and deviant behavior posed by such residential
areas. These studies postulated a number of reasons for the persistence
of certain neighborhoods in producing social problems. Burgess himself
was led to postulate a correlation between mobility and individual
behavior as being of central concern. "But he missed the point of his
own model of the city. The movement of people from one residence to
another as the city grows is the very mechanism by which the zones and
natural areas are created" (Berry and Horton, 1970:307). An alternative
explanation was provided by Louis Wirth (1938). Wirth felt that size,
density, and heterogeneity were the key elements needed to explain
social disorganization. He theorized that secondary and tertiary modes
of interaction are substituted for primary modes, in large, densely
populated, heterogenous cities. However, later researchers have shown
little tendency toward the withering away of family life in an urban

environment (Berry and Kasarda, 1977:128).

43

The urban ecology models were later refined by Firey (1947) who in
a study of Boston, added a cluster element based on ethnicity; and by
Harris and Ullman (1945) who added multiple nuclei to the model. The
model became to be regarded as being of limited practical use, however,
by the late 1940's. Berry and Horton (1970:307) theorize that this was
because the "relationship of the urban environment to the behavior of
the urban population remained obscure in this classical stream of
research”.

Homer Hoyt (1939) put forth an alternative model to explain the
changing spatial structure of the city. Based on a statistical study of
142 cities, Hoyt theorized that high and low rent neighborhoods occupied
distinct areas of the city, and were not distributed concentrically, but
sectorally, in the urban area. The major criticism of Hoyt's model is
that it is at best a partial view of the city, constrained by his narrow
focus on housing and rent. He gave little consideration to the
characteristics of the inhabitants who occupied the structure (Berry and
Kasarda, 1977:126).

“There has been considerable discussion of the

relative merits of these classical models of the spatial

structure of the city's population. There is an emerging

consensus that, in Western metropoli born in the industrial

age and populated by a variety of races or national groups,

the models are independent, additive contributors to the

total socio-economic structuring of city neighborhoods.

Those indices which measure the socio-economic status of

individuals or groups, vary principally by sector; those

which measure the familial characteristics and age of the

population vary principally by concentric zone; and those

which isolate a minority group within the city population

show a tendency for that group to cluster" (Berry and

Horton, 1970:309).

This emerging consensus is largely the result of the

investigations in recent years into the behavioral basis for the

residential location decision.

44

BEHAVIORAL BASIS: THE RESIDENTIAL LOCATION DECISION

One attraction of the ecological view of the industrial city is
that it postulated a process of group competition and mobility that
produced the spatial structure of the city's population. In Hoyt's
model, on the other hand, the operations of the real estate market are
not spelled out, although he recognized the importance of the decisions
of prestigious individuals in the location of the high-grade rental
sector (Berry and Kasarda, 1977:126).

For North American cities an alternative view has been proposed,
based on the behavior of individuals and institutions. Residents of a
city are faced with a choice of where to live. Theoretically the
principal determinants of such a choice are the price of the housing,
its type and its location relative to both neighborhood environment and
place of work. These determinants have cognates in the attributes of the
individuals making the choice. These would be the price willing to be
paid for housing which is dependent on income, housing need which is
based on marital status and family size, life style preferences and
location of the job. When the values of the two sets of characteristics
match, a decision to locate will be made (Berry and Kasarda, 1977: 126).

Of all these characteristics, income has been considered the most
important. Income determines, to a great extent, the ability to meet the
preferences of the individual. The identification, importance and
operation of these factors have been shown by a number of researchers.
Key among these was Cans (1967) in his participant-observer study of
residents of Levittown.

The conclusions of the variety of research that has been conducted

is summarized in Figure 11. ( This is a graphical representation of this

45

FIGURE II

RESIDENTIAL LOCATION DECISION PROCESS

A SOCIAL SPACE Unite: Individual or FemiIios

ProIoeeioneI
Collep Educetion
Menogenol

Hi"! SEMI CluicoI
Educerion m

CreIumen

Ooeretive
Grede School
Educeiion
Lehorce

Underemoioyed
Little Schooiin.
Unnamed

 

 

 

 

8. HOUSING SPACE Unite: Dudlh'

 

 

 

Low Income

 

Poet Child-baring Age:
endlor
EorIv AduIthood

Pie-School. SchooI
end Child-booting Agee

Si - individuel i'i position in eociel niece

 

 

Shel. Maui hiee Love Iiee run a... 3...,
Sinol ForniIv Lone FemiIv Anomnont Aeenment Mm OooIoe It...
ChiIdIeee Coulee Merried with Children _ Mon Seecioue Loeurioue v.“ “U Price
Socloocgiormc Sieiue H.” 'mm' W MW or a.“
. Mil-Medium
Specioue Comiortehle .
High Middle Income '7'“ 0' ROM
fl.
Si Avere'e Aw"; ' Au... m.
- siege on Lil. Cycl. Averege Income or Rent
Crowded Could be improved Low-Medium
Love Middle Income Pnce or Rent
Overcrovvded swim mi"

 

 

Ne individuel lot SineiIer Let Smell Lot Ler'e Let

er orivete verd end Yerd

end Yerd end Yerd

Him Deneity Medium Deneity Lew Demity

Hi I WI i't home in houdn. W

C. COMMUNITY SPACE UnIte: Trecte or Lerner Sub-once O. LOCATIONAL OR PHYSICAL SPACE
Unite: Trecte or Lerner

 

 

ESIVUI Hflm
I

2 4 '

flavor: O-CI .
I: 7
O
9 H
Loam

 

 

 

 

O-teeelioiuonyinohichl'ehomehleceud

Source: Berry and Kaoerde (1077)

 

 

u-Iheeeiolotheeommeehyleohlmrehomeleleened

46
model as it operates at four varying levels of spatial organization. The
discussion centers on these models.)

"The individual and the family occupy a position,
s(i), in social space, determined by the economic status and
family status. The household matches this position with that
of a dwelling located in an analogous position, h(i), in
housing space, and of housing in a similar location, c(i),
in an equivalent community space, whose axes comprise
socioeconomic status on the ordinate and familial
characteristics on the abscissa. From a range of possible
communities found in the same zone of community space, one
dwelling in one community is selected on the basis of
proximity to job location or on the basis of other important
neighborhood characteristics, thus fixing the choices in
physical space.

An orderly social ecology results through like
individuals making like choices, through regularities in the
operation of the land and housing markets, and through the
collaboration of similar individuals in excluding those of
dissimilar characteristics from their neighborhood or in
restricting certain minority groups to particular areas. The
autonomous suburb is the prime example of the process of
exclusion, and the ghetto the most glaring illustration of
the process of restriction. Sectoral patterning of such
attributes of the neighborhood residents as education,
occupation, and income, and of neighborhood structural
characteristics such as rent or value and quality of
housing, is a product of the differing abilities of various
income groups to bear the costs of the journey to work.
Lower income workers, because of their restricted budgets,
must live close to their work. The higher incomes of upper
status workers give them the freedom to locate their homes
in areas of higher residential amenity, away from their
places of work, away from the smoke and dirt of industry,
and close to amenity features such as lakeshore and open
space. The age structure of the population, average family
size, and female labor force participation change as
distance from the city center increases; young families
locate farther from the center than do older families. This
pattern is a response to the change in house age and type as
distance from the center becomes greater, the houses newer,
and single family homes predominate as the city center is
left behind. It is the lower land values toward the urban
periphery that make possible this land-voracious
construction, and the increasing real income of home buyers
makes possible the purchase of such newer houses. Finally,
minority groups find themselves segregated from the rest of
the population to a greater or lesser degree as a result of
recent arrival in the city, discrimination in the housing
market, or through choice of home in congenial communities"
(Berry and Horton, 1970: 311-313; Berry and Kasarda,
1977:130-131).

47
ANALYTICAL FRAMEWORK- SOCIAL AREA ANALYSIS

The behavioral basis for analyzing the residential location
decision, and the resulting spatial basis for the differentiation of the
city's population, as expressed above, provides a theoretical base for
examining spatial structure. In the late 1940's and the 1950's, attempts
began to be made to provide a holistic framework of analysis, based on
theory, that would make clear the linkages between the social,
structural, and locational spaces of the city.

The initial attempt to provide a better analytical framework was
outlined by Shevky, Williams and Bell in their studies of Los Angeles
and San Francisco (1949, 1953, 1955). From a number of postulates, they
derived three constructs which, they felt, described the way in which
urban populations are differentiated in industrial societies. The three
constructs were called social rank (economic status), urbanization
(family status), and segregation (ethnic status). Based on these
constructs they proposed three indexes, one for each of the constructs,
made up of from one to three census variables, and designed to measure
the position of census tract populations on scales of economic, family
and ethnic status. The analysis also made possible the classification of
census tracts into social areas based upon their scores on the indexes
(Hartshorn, 1980:232).

This initial attempt at creating an analytical framework was
criticized on both theoretical grounds that is the theory underlying the
constructs, and for empirical reasons or the method of dimensioning the
constructs (Hawley and Duncan, 1957; Duncan, 1955). Bell (1953)
attempted to meet the empirical objections that the social area analysts

selected measures on the assumption that the constructs were correct,

48
but failed to provide a test of their validity. He used factor analysis
to show that, in the case of Los Angeles and San Francisco, the
variables selected from the census to construct the indexes, did indeed
conform to Shevky's formulation. Bell's work was extended by Van Arsdol,
Camilleri and Schmid (1958) who tested ten cities along these lines. Of
the ten, six conformed to the Shevky constructs. Berry and Kasarda
(1977:123) feel that the presence of four cities that did not conform
suggests that the existence of the constructs should be left as an
empirical question to be determined by the patterns in the variables,
rather than one to be assumed correct a priori. The logical extension of
this argument is that many more variables detailing the way in which
census-tract populations vary according to socioeconomic characteristics
should be included in any study and that factor analysis should be used
to isolate the fundamental patterns of variation in the data. This
involves the use of another type of factor analysis, principal component
analysis. Kendall emphasized the difference between factor analysis and
principal component analysis in a manner that makes clear the difference
in the approaches. He states: "In component analysis we begin with the
observations and look for the components in the hope that we may be able
to reduce the dimensions of variation and also that our components may,
in some cases, be given a physical meaning. In factor analysis we work
the other way around; that is to say, we begin with a model and require
to see whether it agrees with the data and, if so, to estimate its
parameters" (Kendall, 1958: 122). The evolution of this type of
analysis, formally called factorial ecology, is traced in Figure III.

(This figure depicts the two approaches to Social Area Analysis and the

Typo

Mottuxd
Ernp I eyed

‘49

FIGURE III

SOCIAL AREA ANALYSIS

SOCIAL AREA ANALYSIS, broadly defined

SOCIAL AREA
ANALYSIS.
STRICTLY DEFINED

Construction of
Shevky-Boil
indices

FACTOR ANALYSIS
OF SOCIAL AREA
STRICTLY DEFINED

VARIABLES

Factor analysis

of Shevky-Bell
index variables

FACTORIAL
ECOLOGY

Factor analysis

of a wider set

of socio-economic
variables, Including
the Shevky-Boil sot

Social area analysis, strictly defined
SHEVKY and WILLIAMS; SHEVKY and BELL

Criticism
HARLEY and DUNCAN, )UNCAN

Continuing applications
of social area analysis,
strictly defined

o.9.. HERBERT;McELRATH

Tests of social area
constructs using
factor analysis
BELL; VAN ARSDOL.
CAMILLERI and SCHHID

Cluster analysis
otwflrnwmm
data tor census
tracts

TRYON

Factor analysis of socio-
economic data for census
tracts deriving basic
dimensions which are
compared with aveilabie
theory.

Works reviewed:
SCHMID and TAGASHIRA
SHEETSER

GOHEEN

MUROIE

PEOERSON

BERRY

ABU-LUGHOO

Source: Berry and Horton,

I970

50
criticisms and refinements of the initial proposal that have been
accomplished) .

In factorial ecology, a data matrix is analyzed containing
measurements on m variables for each of n units of observation (census
tracts, wards, etc.), with the intent of (1) identifying and summarizing
the common patterns of variability of the m variables in a smaller
number of independent dimensions, r, that additively reproduce this
common variance; and (2) examining the patterns of scores of each of the
n observational units on each of the r dimensions. The dimensions
isolated are an objective outcome of the analysis. Interpretation of the
dimensions (factors) depends on the nature of the variables used in the
analysis and the body of concept or theory that is brought to bear.
Theory provides the investigator with a set of expectations regarding
the factor structure which can then be compared to the actual set of
ffiletors produced. This comparison was made formally by Van Arsdol,
Camilleri and Schmid (1958). What is important, however, is that studies
of American cities have, by and large, have identified the three social
area dimensions originally proposed by Shevky (Berry and Kasarda,
1977:123). (See Figure IV, Illustrations A, B, and C )

Moreover, as mentioned earlier, it has become increasingly evident
that each of these dimensions captures the essential features of one of
the classical spatial models: socioeconomic status (Hoyt); family status
(Bul'gess); ethnic status and segregation studies (Firey). This
Observation was clearly made by Anderson and Egeland (1961) in their
analY$is of the spatial variance of a number of socioeconomic variables

for Akron, Dayton, Indianapolis, and Syracuse.

 

 

51

FIGURE IV Integrated Spatial Modei of the Metropolis

 

 

 

 

 

 

 

I - SUIURBANIZED INDUSTRIAL
WORKPLACES

2 - INDUSTRIAL SATELLITES

3 - HEAVY INDUSTRIAL AREA
NEAR THE LAKE

 

52

This basic triad of spatially arranged social dimensions can be
superimposed to form, at the intersections of sectors, zones, and
segregated areas, communities of similar social, family, and ethnic
status. Figure IV provides an idealized picture into which distortions
can be successively introduced to approximate reality more closely.
However, there is a further complication. The zones within the
segregated area occupied by the minority group do not correspond to the
general life cycle zones of the metropolis; the segregated area is a
microcosm of the whole, compressed spatially, reproducing in miniature,
the metropolitan-wide pattern (Figure IV, E). This modified pattern is
then further distorted by city growth (Figure IV, F). ‘Tear faults'
develop as zones cross sectoral boundaries, with displacement of zones
Ontward in the early-growth sectors. Finally, introduction of secondary
w(Dirk-place nodes- a heavy industrial area in the southern part of the
City, industrial satellites in a crescent- further changes the form of

the model for the metropolitan region (Berry and Rees, 1969:115).

A CROSS-CULTURAL CONTEXT: THEORY AND FRAMEWORK

W

It has been shown that from the bases of residential choice, the
ecological pattern of American cities is dimensioned cumulatively by
socioeconomic status, family status, and the constraints of race and
ethnicity. What of non-American cities? Do they exhibit the same
patterns of variation in census tract populations?

One of the earliest, and most comprehensive studies of a
non‘Alnerican city, was Pedersen's (1967, in Berry and Horton, 1970:317)

Study of Copenhagen. Pedersen constructed a matrix of 14 socioeconomic

 

53
variables and 76 zones of the city, for both 1950 and 1960. Pedersen
identified three factors, (1) urbanization or family status; (2)
socioeconomic status; and (3) population growth and mobility. The first
factor, when mapped, displayed a Burgess concentric ring pattern. The
second factor displayed a sectoral pattern, with the exception of the
central zone, which was of uniformly low status. The third factor
(:1115tered in the central zone and the urban periphery.

Sweetser (1965) conducted a cross cultural comparison between
Iiefilsinki and Boston in which the Shevky-Bell constructs were identified.
These early studies highlighted the cultural context of the factor
structure. This limiting factor was addressed by Abu-Lughod (1969) and
Berry and Rees (1969). These investigators aimed at isolating those
bas 1c conditions in the urban system, social and spatial, which are
necessary to produce the observed factor structure. Abu-Lughod in her
StUdy of Cairo found that "no factorial separation between indicators of
SOCial rank and the indicators of family cycle stage could be obtained"
(1969: 21). While this was not unexpected given the limited
industrialization of Cairo; this result prompted her to outline the
conditions that were necessary and sufficient to produce the dimensions
of Socioeconomic status and family status that are found to have
independent existence in the studies of American cities (Berry and
Holi'ton, 1970:319). These conditions are set forth in Figure V.
CO“cerning these conditions, Abu-Lughod writes:

"The disassociation between social rank and familism
variables found in contemporary western cities in societies

at the terminal stages of the demographic transition can be

attributed to the reinforcing and cumulative effects of

several conditions that define the nature of urban

organization in such cities: (1) residential segregation

according to modern ranking systems; (2) relatively low

correlations between social rank and differences in
fertility and family styles: and (3) high differentiation of

54.

FIGURE V

Summary of Janet Abu-Lughod's 'Neccessary conditions.

 

Factor Conditions Types of Variables Used

Soclo-economlc Education, Occupation,

status factor income

Family status Family size, Portions

factor of the age pyramid.
Fertility

Disassociatlon between
between soc lo-economic
status and family
status dimensions

Source: in Berry and Kasarda (1977)

Necessary Conditions

1.

2.

2.

"That the effective ranking system

in a city be related to the operational
definition of social status";

"That the ranking systul in a city be
manifested In residential segregation
of persons of different rank at a scale
capable of being Identified by the
areal units of observation used in the
analysis."

"That family types vary, either due

to 'naturai' causes such as those
associated with sequential stages in
the family cycle, or to 'social' causes
such as those associated with other
divisions in society, whether ethnic,
socio-economlc or other";

"That subareas within the city are
differentiated in their attractiveness
to families of different types” at a
scale capable of being identified by
the areal units of observation used in
the analysis.

Either:

1.

2.

That there exists little or no

association between social class

and family type; or

if there is some association between

social class and family type,

a. there is a clear distinction
between stages in the family cycle,
each stage being associated with a
change of residence;

b. "subareas within the city offer, at
all economic levels, highly
specialized housing accommodations
especially suitable to families at
particular points in their natural
cycle of growth and decline" at a
scale capable of being identified
by the areal units of observation
used in the analysis; and

c. "Cultural values permitting and
favoring mobility to maximize
housing efficiency, unencumbered by
the 'unnatural' frictions of
sentiment, local attachments or
restrltlve regulations.”

55

residential sub-areas by housing types. To the extent that

these conditions are not perfectly fulfilled, the vectors

will not be totally disassociated" (Abu-Lughod, 1969, 30).

In their study of Calcutta, Berry and Rees addressed the issues
raised by Abu-Lughod. The conditions outlined by Abu-Lughod were
expressed in several alternative factor structures representing
permutations of the three basic sets of variables. (SES: socioeconomic
status set; LC: stage in life cycle of family status set; MG: minority
group set) (Berry and Rees, 1969:468). These permutations are presented
in Figure VI. These combinations constituted the frame within which
Berry and Rees evaluated the factorial ecology of Calcutta. Within this
frame, Berry and Rees were also able to draw upon a qualitative study of
Calcutta's spatial structure by Bose (1960). Berry and Rees were also
working within the context of Sjoberg's (1960) contrast of preindustrial
and industrial urbanization. (Berry and Rees saw a strong similarity
between Sjoberg's classification and that of Chatterjee (1960) for
Indian cities.) Sjoberg's view was that:

”The feudal city's land use configuration is in many

ways the reverse of that in the highly industrialized

communities. The latter's advanced technology fosters, and

is in turn furthered by, a high degree of social and spatial

mobility that is inimical to any rigid social structure

assigning persons, socially and ecologically, to special

niches...There are three patterns of land use wherein the

pre-industrial city contrasts sharply with the industrial

type: 1) the pre-eminence of the central area over the

periphery, especially as portrayed in the distribution of

social classes, 2) certain finer spatial differences

according to ethnic, occupational, and family ties, 3) the

low incidence of functional differentiation in other land

use patterns” (in Berry and Kasarda, 1977:110).

The inclusion of this additional theoretical base is of particular

note. It moves the study out of a pure American context, and places it

in the context of a pre-industrial to industrial society continuum.

Combination i (Shevky-Bell)

Variable Set Correspondence Factor Variable Set Correspondence
SES 1 SES
LC 2 LC
MG 3 MG
Combination 3
Combination 4
Variable Set Correspondence Factor Variable Set Correspondence
SES 1 SES
LC 2 LC
HG MG
combination 5
Combination 6
Variable Set Correspondence Factor Variable Set Correspondence
SES 1 SES
LC LC
HG MG
Combination 7
Variable Set Correspondence l Factor
SES i Family Status
2
LC
3 Urbanization
as 4
Source: Berry and Rees, 1969

56

FIGURE VI

FACTMIAL MINATIONS

Combination 2

Factor
i
2
3

Factor

Factor
i
2
3

57
In their conclusions, Berry and Rees state:

"In terms of the factor models then, Calcutta conforms
most closely to Combination 3, in which there is a separate
family-status dimension, but socioeconomic status and
minority group membership are linked ..... This particular
factor combination raises serious questions about relative
emphasis placed on different dimensions in the choice model.

(However) the resulting ethnic base of much decision making

can be embodied in the residential choice model...To the two
choice dimensions of individual social space discussed in
the American context can be added a third dimension of
ethnicity. The dimensions can be reordered, with ethnicity
placed at the head and altered in length; this makes
ethnicity the most important element in the choice process

(for Calcutta)." (Berry and Rees, 1969:490).

Beyond this, Berry and Rees did find the same dimensions outlined
by Shevky and Bell, and as modified by the conditions set forth by
Abu-Lughod. They also concluded that the invasion and succession
process, and a continuing transition from a pre-industrial to industrial
system was in effect. Concerning this, they concluded "As the
transitional process continues, however, one might reasonably expect the
socioeconomic and ethnic bases of differentiation to separate. The
expectation is, therefore, that differing urban ecologies related to
differing factor combinations can be arranged along a scale of urban
development from pre- to post-industrial forms" (Berry and Rees,
1969:491).

THE CASE OE TAIWAN

When the literature relating to Chinese urban areas is examined,
two points are immediately apparent. First, there is a relative lack of
literature. Research on China has tended to center on rural areas and
villages, not unexpected given the rural character of the region.
Secondly, no literature has been uncovered in this effort that deals

directly with the urban ecology . The work on Chinese cities in the

pre-1949 period centered on the structural morphology of the city, the

58
distribution of cities, and the hierarchial arrangement of cities. The
literature relevant to this study on Chinese and other Asian cities was
reviewed in Chapter Two.

The research that has been done on the traditional city structure
in China has primarily focused on the form of the city. Researchers such
as Chang Sen-dou (1961, 1963, 1970) and Eberhard (1955) concentrated on
the structural morphology of the city and its relationship to cultural
and cosmological values. In addition, Chang has devoted sections of his
work to the distribution and hierarchial arrangement of cities. Thus,
while a clear picture of the form of the city is available, little has
been presented on the arrangement of people within the city.

This also holds true for the studies that have centered on the
treaty ports. A clear example is that of Murphey's (1953) work on
Shanghai. While a clear picture of land use is drawn, delineating
residential areas, little mention is made of the social arrangement of
those areas.

In the post-1949 period certain distinctions in the nature of
research on cities must be made clear. While work on the spatial
structure of the PRC cities has been done, this work is of little value
in terms of the theoretical approach being outlined here. This theory
rests on concepts of social and spatial mobility, and processes of
invasion and succession. Given the social constraints on mobility in the
PRC, and the significant intervention of the government in restructuring
the cities, their utility to this study rests on their function as foils
or counterpoints to the process occurring in Taiwan. This is not,
however, to say that such an approach can not be utilized in the PRC,

given the proper theoretical reformulation. In terms of Taiwan, research

59
on cities has tended to concentrate on economic base and transformation,
migration, and the hierarchial structure. While these have value in
explaining the factorial space of Taiwanese cities, they provide little
information on the social geography of such cities. However, the
migration studies of Liu (1979) and Speare (1974) are of particular use
in establishing the presence of a process of invasion in the urban
regions. The economic works of such researchers as Galenson (1979) and
Ho (1978) are valuable in establishing the increasing transformation in
Taiwan to an industrial based society.

In establishing the residential groupings that did exist, reliance
must be on passing or oblique references to this issue contained in
works dealing with some other aspect of urban life in China. A large
number of authors have at least some small measure to contribute. Among
the more significant contributors were Deglopper (1977), Gallin and
Gallin (1974), Baker (1977), Eberhard (1955,1964), Fei (1953),and
Feuchtwang (1974). The factors that have been identified are as follows:

1. Ethnicity Factor - Based on regional and linguistic
associations.

2. Family Structure Factor - Based on kinship, age, and stage in
the family life cycle. To a lesser degree this grouping could have a
lineage, clan or surname dimension.

3. Socioeconomic Factor- Based primarily on occupation and
education. Income seems to have been expressed very weakly.

4. Religious Factor- Based on neighborhood temples and differing
religions.

5. Military Factor- Based on groupings of military units in the

urban environment.

6O

6. Land Use Factor- Based in particular on commercial sectors,
agricultural land in the city, and "parkland" such as that associated
with Bhuddist temples.

Based on these factors, and what is known about Chinese cities,
some general hypotheses concerning the nature of the urban ecology of
traditional cities can be made, at least as they were structured before
1911 in general, and before 1895 in Taiwan. In terms of social groups, a
broad division existed between the gentry class characterized by
administrative occupation and education levels and the rest of the
residents. Within the non—gentry sector of the population, several
things can be postulated. First, evidence from researchers such as
Eberhard indicates that cities were organized into distinctive areal
units based on occupation. Secondly, it can be postulated that these
occupations were strongly linked to ethnicity, particularly place of
origin. Thus, it can be postulated that there was little separation
between the ethnic and socioeconomic factors. However, ethnicity was not
necessarily tied strongly to the income dimension. In addition, one can
theorize that there existed clusters of residents identified by the

other factors listed.

THE RESEARCH QUESTION: RESIDENTIAL SPACE IN TAIPEI

In studying Taiwan, the theory and framework, discussed earlier
appears to be applicable, with some modification along the lines
suggested by Berry and Rees (1969). Taiwan has an economy that can be
characterized as industrial in nature and within a market type economy.
In addition, residents have both spatial and social mobility, a key

element in the behavioral basis theory of residential location. However,

61

given the potential of the ethnicity factor to play a key role in the
decision making process, modifications along the lines suggested by
Berry and Rees must be included. In other words, to the choice
dimensions of individual social space discussed in the American context,
must be added the dimension of ethnicity. While researchers such as
Gallin and Gallin (1974) indicate that this dimension played a key role
in Taiwan, it is not a proven fact. Remember, in the Taiwanese context
'ethnicity' is narrowly defined as differences in place of origin,
dialect and surname. Thus, the reordering of the dimensions, suggested
by Berry and Rees, should not be accomplished at the expense of the
original ordering. In practical terms, this means that the dimensions
should be left as is, with forewarning that a situation not unlike that
found in Calcutta may be exhibited; that is a situation that does not
fit the standard model, based upon differences unique to the country.

Based on the preceding discussion and the area of investigation,
it can be argued that Taipei has, over the last one hundred years, been
the scene of confrontation between traditional Chinese institutions and
industrial urbanization. If this is in fact the case, then the urban
ecology should not only reflect the diversity of Taiwan and its culture,
but also the two further ecological themes of the preindustrial city
categorized by Sjoberg (pre-eminence of center over periphery, low
incidence of functional differentiation in land use), and an increasing
admixture of the ecological patterns of the industrial city. (after
Berry and Rees, 1969) In addition, as the city moves from a pre-to a
post-industrial form over time, it can be argued that a different urban

ecology, related to differing factor combinations, will be realized.

62

The research question will be addressed by assuming the analytic
stance of factorial ecology. The study is patterned on that conducted by
Berry and Rees (1969) for Calcutta, and can be considered as modeled on
that work; although more properly it is an extension of the social area
analysis of Taipei conducted by Hsu and Pannell (1978).

In reviewing previous studies of this type, it is evident that the
region primarily selected is the politically defined city. Berry and
Horton (1970) argue for a wider area, so as not to underbound the labor
market. By the same token, they also argue against overbounding the
labor market by analyzing too large of an expanse. Cullingford and
Openshaw (1984) argue for a regional approach. Working within a
development context, and the identification of deprived areas in the
rural sector, they proposed an area approach given the area basis of
census data in England. Basically they propose analyzing urban and rural
data separately. This would allow the interpretation of the factors to
be set in either a rural or urban context.

This study proposes to examine the Taipei Metropolitan Region as
defined by Liu (1979) and depicted on Figure VII. (located in Appendix
A. It is the 1983 land use map of the Taipei Basin mentioned earlier.
The area covered corresponds to the large map on the sheet.)
Classification of rural and urban areas will not be done. The area is
highly urbanized and limiting the study to the Basin region not only
encompasses the functional area of the city, but eliminates the need to
construct separate factor structures for rural and urban areas. The
area, for all intents and purposes,is urban.

In selecting the variables to be used in the social area analysis,

the body of theory that is brought to bear directs the dimensions within

63
which the variables are selected. That is, the theory creates a set of
expectations regarding the factor structure, and these expectations
provide the dimensions from which specific variables are selected. In
this study the characteristics of Taipei initially deemed significant in
explaining residential patterns are based on the dimensions drawn from
the literature on Taiwan (cited in Chapter Two). They are:

l. Demography

2. Race and Ethnicity

3. Religion

4. Income

5. Occupation

6. Educational Levels

7. Mobility

8. Housing

9. Employment

10. Land use characteristics

The actual variables that are selected must, in sum, depict the
various dimensions. In this context it seems necessary to provide
further definition to two of the dimensions. Racial and ethnic
characteristics must reflect regional origin and language to capture the
dimension. Housing characteristics includes data on structures, as well
as on household organization.

The actual variables selected to depict the various dimensions was
done in the field. The field work was conducted in 1985 and 1986. The
variables were drawn from secondary data sources, particularly census
data, information which is unavailable in the U.S. at the scale needed.

By and large all available data from the target years, at the

64
appropriate scale was used. The attempt was to insure that the perceived
dimensions were represented, but not to influence the outcome by
eliminating variables that may have great explanatory power. The process
of selecting the actual variables selected is discussed in Chapter Four.

Most factorial ecologies have been confined to one time period.
However, researchers such as Goheen (1971) have illustrated the benefits
to be gained by comparison over time. From these comparisons, and a
factor analysis of the change variables, trends can be identified, as
can the significant variables undergoing change. In terms of this study,
one time period, 1980, was selected, based upon data availability. By
the year 1980 the industrial economy was firmly established and rural to
urban migration was low. By this time some 66% of the nation was urban,
and much of the transition from a rural to urban nation was
accomplished.

The method employed is a factorial ecology. The factor analysis
used the Statistical Package for the Social Sciences (SPSS) program with
varimax rotation.

The factors were interpreted based upon the theoretical
considerations expressed in this paper. Factor score intervals will be
mapped and analyzed. The study, while having tremendous value simply as
a descriptive tool, does expect to discover change occurring over time.
While research of any kind concerning Taipei is limited, historical
analysis in conjunction with the theories outlined, establish the
opportunity to theorize about the changing nature of residential space
in the city.

Two studies done on the city, with applicability, include one

article that attempts to systematically assess the area, and one other

65
that attempts to establish a foundation for describing the motivations
and processes involved in residential moves in Taipei.

Hsu and Pannell (1978) used the technique of factorial ecology to
investigate a portion of the politically defined city. The study, while
providing some valuable descriptions and signposts for future
researchers, fell short in its execution. Its weaknesses were basically
twofold. First, the range of variables was extremely limited. Secondly,
the geographic unit (scale), known as a ‘traffic zone' was too large and
the total area studied seriously underbounded the functional region.
This is not to denigrate the study. Basically, Hsu was working in the
field of transportation geography, and the paper was an attempt to
stretch the gathered data. While the work had its problems, it is the
only attempt found that tries to systematically describe the residential
space of the city.

Wang Su Chang in her dissertation (1981) dealt with the process of
residential moves. Wang is a phenomenologist by training, and seeks to
apply the approach to understanding the push/pull factors involved in
intra-urban moves; and particularly why certain chosen locations (the
study areas) were selected as destinations. While one can question
Wang's approach and methodology, she has provided the only information
generally available on movement behavior and perception of the relative
status or desirability of various destination sites in the city.

As mentioned, this study builds on the larger body of theory in
the field, as well as the above two studies. In particular this work is
an extension of the Hsu and Pannell study. In any event, both the
previous study and the larger body of theory give evidence that the

residential pattern of the city should change over time. That is, in the

66
traditional city the residential pattern was relatively simple. The
major residential group was an educated or ‘gentry' class, either
Japanese or Chinese, depending on the time frame. The remainder of the
city was characterized by a mixture of income and social classes with
neighborhoods delineated by ethnicity as expressed in family surname or
village of origin. As development came to the city in the post-World War
II initial patterns continued. This study, based on the available
evidence would hypothesize that over time changes did begin to occur as
Taipei began to establish a wider range of neighborhoods characterized
by a similarity of socio-economic status, ethnicity or stage in the
family life cycle. In Chapter Two the traditional pattern has been
shown. In Chapter Four changes in specific neighborhoods will be shown
by establishing the current pattern and tracing the historical record
for the area.

Abu-Lughod in her work on Cairo cited earlier was basically saying
that for these elements to become apparent as significant contributors
to neighborhood formation and separation along the Western model, then
development must occur. Conversely it can be argued that if development
occurs then it should be accompanied by a change within the residential
structure of the city, and that such change will be reflected in the
factor structure of variables describing the residential population.
More specifically, it is argued that over time there should be an
increasing separation in the four major factors of socio-economic
status, stage in the family life cycle, ethnicity and mobility. In
precise terms this study hypothesized that:

1. In 1980, the city had progressed to the industrialized stage

(described by Timms,l970) and separation among the factors is found.

67

2. In 1980, the factor separation that is present will be
expressed with its associated pattern.

In plain English, the hypotheses are stating that by 1980, a
factorial ecology of the city will indicate that distinct neighborhoods
will be able to be detected along the lines of those formed in other
industrial cities. Each factor for the residential population that is
found to be separate and distinct will present its associated pattern.
For example if a socio-economic sorting factor is found to be separate
and distinct, it will present a concentric zone pattern.

Again, the true value of the project is that even if the
hypotheses must be rejected, the factorial ecologies will still yield
sufficient information to identify and map social distributions in the

city.

CHAPTER FOUR

THE RESEARCH EXPERIENCE AND RESULTS
ca e a e re V a es- e Data Base

mic

With the research problem formulated and the method of analysis
selected, it became time to answer the proposed questions for Taipei.
The first major step involved was to select the appropriate scale at
which data would be gathered and analyzed, and to define the variables
needed. In the United States the tendency of studies of this type is to
use the data generated by the Census Bureau. This data base provides the
types of variables desired by the researchers. This is advantageous as
it saves researchers from the expense of time and effort that would be
involved in generation of the data. However, this data base would have
been unusable if it was not gathered at the scale required by the
researcher.

While not privy to the decisions of earlier researchers using
factorial ecology and with no significant discussions of the problem of
scale in their published works, it seems likely that the goal was to
find a spatial unit that was at some level of aggregation (as the
studies were concerned with gross patterns and not individual
locations), but not a level so large as to subsume significant
differences within the general urban pattern. Census data offered a
viable data base, as the information that is gathered can be accessed

68

69
at a variety of scales. The problem for researchers, once they determine
the census offers the range of required variables, is to determine if
any of the scales are appropriate to the type of study being undertaken.
The scale chosen by most researchers as appropriate is the tract. This
subsumes the individual choices, as well as the smaller block unit, but
it is not so large as to overshadow smaller areas of potential
significance. Understand that this may not be the ideal, desired choice.
Logically, the level of individual household is probably the most
desirable. By treating each household as a separate unit the researcher
can be sure of identifying areas of significant difference. With any
level of aggregation, and the accompanying averaging of statistics, one
introduces some potential error. However, the level of individual
household simply results in too many spatial units. For example, if the
household level was used in this study, and combined with 50 descriptive
variables, this would result in an algebraic matrix that was four
million by 50 and contained some 200 million individual cells to be
analyzed. It is clear that some compromise concerning scale must be
made. Earlier researchers selected the tract level, understanding that
it represents a compromise.

For this study it was decided that a similar size spatial unit was
appropriate. The ideal unit would encompass somewhere between five to
ten square blocks or approximately 1000 family units. This would keep
the number of spatial units to a manageable size, while still permitting
the identification of smaller areas of significant difference. Ideally,
the goal was to find an existing data base which met the same demands
facing earlier researchers in the U.S., that is, a data base that had an

acceptable variable range at the desired scale. If such a data base did

70

not exist it would create serious problems in terms of both time and
money.

Variables

An underlying assumption of the method of analysis employed in
this study is that the patterns of residential arrangement within urban
areas of a free market economy are universal. That is, the theories that
have been developed, and that are being applied in this study to Taipei,
are universal (provided free market forces are in effect) and by and
large transcend a specific cultural context. Thus the intent of this
study is not only descriptive, but aimed at linking the behavioral basis
of the approach and the associated urban patterns, to the Chinese
context. Studies in the western world have been relatively consistent in
selecting variables that, by and large, describe the constructs ( social
status, family status, and ethnicity) and thus are likely to produce
factors which can be described in those terms. This of course was one of
the original criticisms of Williams, Shevky and Bell's initial work
establishing the ecological theory and approach. Later studies, using a
wider range of variables, upheld the validity of these constructs, thus
making it acceptable to limit variable input. This, combined with the
use of principal component analysis (which some describe as rummaging
around in the data in order to find out if it is related in any way),
continue to be consistent criticisms. Both have a certain amount of
validity, but the first, that is the limiting of variable input, is
probably the more serious of the two. It does seem that to make an
assumption that residential space will be described by the three
Constructs, and thus to limit variable input to those relating to these

constructs, is a weak point of the traditional approach. This is even

71
more critical in a non-Western cultural context where other, unknown,
factors may be significant in residential sorting.

While numerous studies have supported the efficacy of the
constructs in explaining much of the residential space of the city ( as
perceived by the researcher), the utility of the constructs is by no
means agreed to universally. The framework which supports social area
analysis is a theory. By definition it is something that should be
viewed with some degree of skepticism. Given the range of criticisms and
the fact that the urban experience in China is different from that of
the West in a number of potentially significant ways, the universality
of the constructs and approach becomes slightly shaky. It becomes
incumbent, then, to insure expansion of the variables selected to
include a wider range than is often employed in the U.S..

This is not necessarily inconsistent with the framework of
analysis provided by factorial ecology. Not only did researchers
subsequent to Williams, Shevky and Bell initially employ an expanded
variable selection, but Bell himself, in his initial response to early
critics (1955) used a wider range of variables. Berry and Kasarda (1973)
in their study of Calcutta used all available variables at the scale
selected. (See Appendix A) They recognized the difficulty of working in
a Lesser Developed Country and the problems involved in obtaining
reliable statistical data. Thus, when they found a data base that fit
the scale required, and that had many of the standard variables, they
used the data base. However, aware of the possibility that the Indian
city may be arranged in a different manner than a Western city, they
expanded the variable base to include all (potentially relevant)

variables at the desired scale. They recognized, as others have, that

72
the constructs have proven to be of explanatory value and thus variables
that relate to them should be included; but that to be sound in
approach, one must (if possible) include a range of variables that may
describe the residential space of the city in question.

To this end the variables that were considered to be of potential
value to this study fell into a number of distinct categories. In
Chapter Three were detailed a number of potential residential groupings
as identified from the qualitative research of others. These groupings
included the three constructs that were detailed by Shevky, Williams and
Bell, as well as the potential for a religious, military, governmental
housing, and a land use factor (particularly given the lack of land use
differentiation in some districts of the city). At the least then, the
study sought to include variables that described these potential
dimensions. Ideally the study should include a range of variables that
describe not only these dimensions, but any potential dimension that the
researcher can justify as potentially existing. From a more practical
viewpoint there is a limit to the number of variables that can be
economically manipulated. In addition, inherent weakness of principal
component analysis demands that the constructs (and thus the variables)
be justified to some degree before 'rummaging' around in the data. That
is, the model construction must be sound. Thus the researcher must
select, prior to variable selection, the dimensions it is felt will best
explain the majority of the residential pattern, and select the
variables based on these expectations. For this study, given the
expectations concerning the residential pattern of Taipei, the desired
variables fell into a number of distinct categories. These categories

satisfied the demands of strict construction of expected spatial

73
arrangement and still maintained some degree of flexibility. The
categories chosen were drawn from the literature on Taiwan. (See Chapter
Two for a discussion of these categories.) They include data on the
demography, race/ethnicity, religion, income, occupation, education,
mobility, housing and employment of the city's residents.

These categories formed the desired data base. It was not meant to
necessarily exclude additional variables, rather it was meant to provide
a guideline for an acceptable data base that was linked to both the
larger body of theory and to the expectations concerning the residential
patterning in Taipei. With these expectations in mind, the search began
for an existing data base that would meet the requirements of the study
both in terms of diversity of variables and availability of the data at
an appropriate spatial scale.

The Searc o a a ase

Previous research of this type centering on Taipei was limited to
one study. Hsu and Pannel (1978) used the technique of factorial ecology
to investigate a portion of the politically defined city. As the only
study of this type on Taipei it provided the logical starting point for
this research. The study itself did exhibit several weaknesses. First,
the range of variables was extremely limited. Secondly, the spatial unit
employed, something called a ‘traffic zone' was too large. That is, the
spatial unit covered so much territory that potential significant
differences were likely subsumed. Lastly, the total area investigated,
limited to a portion of the political city, seriously underbounded the
functional metropolitan region. Not to be unduly critical, the study was
a subsidiary study to the project's main focus on transportation in

Taipei. None the less, the study is the only attempt extant to try to

74
systematically describe the residential space of Taipei. As such it
provided valuable signposts for further research efforts. Appendix B
lists the variables employed in the study. The use of these variables
indicated that at least a portion of the desired variables were
available.

Using the Hsu and Pannel study as a starting point it was possible
to identify one potential source of data. The Taiwan government requires
each household/individual to register on a yearly basis with the
government. During this registration procedure information is gathered
on demographics and education. Unfortunately the information does not
include information on economic status, mobility, housing stock, or any
of the other desired categories. The lack of range in the variable base
made this source less than desirable, but did at least provide a
minimally viable alternative if nothing else could be found.

Attention was next turned to the Census of Population that Taiwan
conducted in 1960, 1966, 1970 and 1980. As can be seen in Appendix C the
range of information covered by the Census is much broader than that
covered by the Household Registration Survey. While most major
categories that were deemed desirable were covered to one degree or
another, it did exhibit one serious lack of data that was of concern.
The census data base offers no information concerning economic status,
that is, there is no breakdown of the population by income categories as
in the U.S. census. This was of vital concern, as class structure and
its spatial expression in neighborhoods is vital in understanding the
spatial arrangement of cities evolving within a market economy. Given
the fact that social status/class are largely associated with economic

status the lack of economic data was serious.

75

Investigation indicated that there existed at least two major
reasons for the lack of income related data. First, Chinese social
structure is based on extended family units. It is often difficult to
set a standard for what constitutes the ‘household income', given the
pattern of income contributions and disbursements within a varied and
complex household system (See Gates, 1982 for one discussion of this
problem). Secondly, knowledgeable sources within the country indicate
that the residents consistently misrepresent their income when asked (a
phenomenon that is hardly unique to Taiwan), thus making such
information unreliable and the gathering of it an exercise in futility.
Further, inquiries of this nature are highly resented both for cultural
reasons and because if nothing else, high incomes attract not only the
attention of the tax bureau, but also of thieves. Misrepresentation is
possible given the cash nature of the economy. Personal credit, and to a
large extent business credit, is not present, at least as we are
familiar with it in the United States. Consumer and business credit can
come from the ‘black market' (sometimes referred to as ‘curbside credit'
Operations in Asia) or more commonly from friends, relatives and
employees. Further, checks are usually distrusted, most likely stemming
from the widespread practice of issuing post-dated checks, with the
inherent gamble on the recipients part that said check may or may not be
good when the valid date arrives. As a consequence the economy runs on
huge amounts of cash, which makes accurate tracking of the economy
difficult. It also makes it nearly impossible for the government to
confirm reported figures. Although a personal income tax system is in
place, and the government most likely can validate the primary

source/amount of income, the issue is further complicated by the large

76
percentage of the population that has secondary sources of income,
usually cash and apparently often unreported. In any event the Tax
Bureau is not about to release specific information about households,
and it does not aggregate its data for publication purposes. Even if
they did publish such aggregate data local economists feel that it could
not be used with a large enough degree of confidence as to accuracy.

It was learned that the Institute of Economics at the Academia
Sinica in Taiwan has, in recent years, conducted a confidential sample
survey that is thought to be relatively accurate in assessing household
income. (The survey results are not released and the information was
provided on a personal basis and then only in the form of conclusions.)
While the sample size is statistically significant for the country as a
whole, it does not provide complete or adequate coverage of smaller
scale spatial units. However, Academia Sinica does feel that a number of
valid observations can be drawn from the survey. The conclusion germane
to this study is that, at the present time, income and social status
exhibit a positive correlation with education. That is, the higher the
educational level the higher the status and income. Conversely, the
lower the educational level the lower the social status and income is
likely to be. This is not an absolute correlation. For example, some of
the richer residents became rich as a result of geographical accident;
they happened to be lucky enough to own farmland in a high urban growth
area. However, the correlation between education and income/status is
the best indicator available. It was vital to this study, because it
allowed either of the two data bases to be employed.

The next step then was to examine the spatial units by which the

gathered data was aggregated to see if any of the levels were

77

appropriate to the study. It was clear from examination of the published
(book form) data that the readily available levels of aggregation were
unsuitable. Unlike the U.S. the lowest level of aggregation available
for urban areas was at a level known as the ‘district'. The districts
varied in size (see Figure VIII) but were uniformly too large in
population, usually ranging in the hundreds of thousands. A district in
Taipei would be the rough equivalent of perhaps a ‘ward' in a U.S. city,
or perhaps akin to our more generic descriptive terms such as downtown,
east side, south side etc. In any event they were too large for the
purpose of the study.

Through contact with various governmental and educational agencies
it was determined that urban regions were politically organized in a
typical pyramidal fashion. The functional metropolitan region of Taipei
is organized under two major governmental units. (See Appendix A) The
territory lying to the east of the Tamshui River comprises the Taipei
Municipality. This city enjoys equal status with county level
governmental units. The city is broken down into a number of distinct
districts, such as Shihlin. Each district is broken down into a number
of tggn or 11, depending on whether the unit in question is rural or
urban, respectively, in character. The EggQ/li are at the same level of
organization and the major difference is based on land use character.
The tggg/li are a geographic unit that is roughly equivalent to either a
precinct or census tract in the U.S.. Ideally, the Eggg/li are roughly
equal in the number of households, although not in population, housing

units or area. Districts vary greatly in size (both population and

 

 

78

FIGURE VIII

 

Taipei Metropolitan Region

 

 

District Boundaries

/

Census Tract Boundaries

Kilometers

2

 

 

 

k; .

AV

\xwx
\
b, ,

Ronald D. T/eienbach i988

 

 

 

 

area) a

._a

H
as
F“
ti
RI

County .
heavily
breakdc
aipei.
nature

directl

gatheri:
there i;
that eq:
data has
the dist

Ir
SEIIES (
IGSearC}
requeSte
Was gene
SCale f‘
50%me
appeare:
Plated ,
leamed

 

thus V6

 

79
area) and in the number of tggg/li (from 12-15 to as many as 110). The
_i sized units are further subdivided into blocks.

The area west of the Tamshui River is largely made up of Taipei
County. Within the county, the situation is more complex. Several
heavily urbanized areas, such as Sanchung, have city-level status. The
breakdown within the cities follows the same pattern as in the city of
Taipei. Other areas of the county, which are more suburban or rural in
nature omit the equivalent of the urban level of government and move
directly to the Eggg/li level.

The census takers in Taiwan utilize this political organization in
gathering their data. However two significant points must be made. First
there is no equivalent level of aggregation or presentation of results
that equates to the SMSA in the U.S.. Thus there is no easily available
data base. Secondly, and most importantly, the data is published with
the district being the lowest level of aggregation.

Inquiries to the appropriate government agencies presented a new
series of problems. From what could be learned, very few, if any,

researchers, government agencies, or other users of census data had ever

requested data at less than the district unit. It was not a level that

 

was generally available. However, the tsun and l; which met the desired
scale for this study, was the level of aggregation used by the
government in creating working files for computer use. By chance it
appeared that the data required might be available. Hapes had to be
placed on the availability of the computer records in as much as it was
learned that the paper records of individual responses, block and
Linn/11 level aggregations had been seriously damaged in a flood, and

thus were not available in any central location. The possibility existed

80
that district offices had maintained records, but there was no
guarantee, and it was advised that obtaining such records from the
various offices, particularly by a Westerner, was virtually impossible.
Additional research at the district level also indicated that retention
of 5533/11 level data for the district was more a matter of chance and
available storage space than policy.

By this point it was clear that the only potentially useable data
base that covered the area required, and the range of variables needed,
was the census. In addition, it was the only data base that was
consistently aggregated at the tggg/li level which had been identified
as the desired spatial unit. Use of the data base depended then, on the
availability of the computer tapes.

Through a series of fortuitous circumstances arrangements were
made to obtain the desired data for 1970 and 1980. Earlier data was not
computerized and not saved in any but the published form. Another glitch
soon arose. Computerization of census data began in 1970. A master tape
was made from the original data which had been aggregated at the gggg/li
level. Unfortunately, the original tape failed, with some 60-70 percent
of the tape becoming unreadable, and a backup tape was not found. The
tape was never reconstructed, most likely, it was explained, due to the
lack of demand for information at that level of aggregation, as well as
the unavailability of the original source material. In any case this
meant that the project was only able to obtain records for the 1980
census year. While this met the needs of the project, it would have been
desirable to have the data from both years. This would have allowed a
systematic investigation of change over time. The project then,

continued its original proposal of examining the pattern of residential

81

location in 1980 and comparing it to historical data in order to
theorize about any change that might have occurred. While two data bases
from widely separate periods may be more desirable, the lack of any base
description of the social geography of the city in itself makes the
production of such a geography extremely valuable. This, combined with
the historical data does provide a picture of any change that is
occurring. In addition, the construction of a valid social geography for
1980, paves the way for a more systematic investigation of change when
the 1990 census data becomes available. Thus the project is also part of
an ongoing investigation into the social geography of Taipei; and the
changes that are occurring over time.
A t e a 5

Having determined the appropriate scale, variables and available
information, it was time to assemble the needed data base. The computer
files for the defined study area were obtained. Investigation showed
that there were 1,458 spatial units (tggn/li) within the defined study
area. The complete variable base for these units were separated from the
main file and a sub-file created. For each spatial unit the census
listed 380 variables. (A complete list of these variables is included in
Appendix D.) These variables were examined to determine which should be
included in the study. The conventional wisdom in selecting variables
suggests a limited number of variables that represent the expected
dimensions. This study, as explained earlier, sought to include any
potentially significant variable at the spatial unit selected. The
variable base available was, by and large, used in its entirety. The
major change was to shrink the data base to a more manageable size by

expanding (combining variables) data categories. For example, the census

82
listed full employment data in two separate categories, each broken down
by type of employment and sex. The division within the census was for
the benefit of other users of the gathered data and reflected technical
differences in the employment picture that were not germane to this
study. Therefore the categories were combined to reflect total
employment in a particular economic sector.

In addition, figures were converted in one of two ways. Where
appropriate raw numbers were converted to percentage figures. This did
not in any way affect the relative position of any spatial unit to any
other. Secondly, some new variables were created, the most significant
being a number of Locational Quotients (LQ). (The concept of an LQ was
developed and used by Berry and Rees in their 1973 study of Chicago) An
LQ is a statistical measure that expresses the percentage of a
particular sub-group of the population as a function of the total area
wide population of the sub-group, and as of a function of the
relationship of the spatial unit's population to the total population of

the study area. An LQ is expressed as follows:

i'
LQ- Eélj Where: LQ -the location quotient
p for population group i in
J
H
5?; area j.
T'I

gij-the total population of
group i in area j
256 - the total population of
J" group i in the region
Pj- the total population in

area j

m
33} - the total area population
4'!

83
In short, a location quotient provides a measure of the concentration of
a particular sub-group in the spatial unit.

These changes resulted in the selection of 137 variables to
describe the residential population of the study area. The variables
selected are listed in Appendix E. The data base consisted of a data
matrix having 1,458 spatial units x 137 variables. The variables
selected fell into a number of distinct categories. These categories
are:

Total Families

Population (by age and sex)

Type of Household

Marital Status

Labor Force/Occupation

Place of Origin

Mobility

Births/Fertility

Literacy

Housing Type/Year Built/Vacancy Rate

Housing Ownership

Housing Size

Housing Facilities (Baths, Kitchens, etc.)

Rggults-The Factors Extrgctgg

The data matrix having been created was then submitted to
principal component analysis using the SAS statistical package available
at the Taiwan Ministry of Education's Computer Center. The center
employs a number of mainframe computers, and this study utilized one of

their IBM 4341 mainframes. In order to facilitate interpretation of the

84
meanings of the common factors extracted, the factor structure was
simplified by use of varimax rotation. In addition, in order to examine
the characteristics of each spatial unit, the factor scores of common
factors (eigenvalues > 1) in each unit were calculated.

From the matrix, nine factors were selected that explain 80.6% of
the total common or shared variance within the data base. Selection was
based on the amount of total variance explained by the factor as well as
the break point in the scree plot of the factors. Factor number ten and
beyond explain less than 4% of the shared variance; and fell beyond the
natural break point on the scree plot. The nature of the factors was
interpreted using the factor loadings with variables being included that
loaded at .5 or above.

Fagtgr One; Sogio-economig Status-Upwardly Mobile Blue Collar
Workers/Illiterate and Poor Agricultural Labor

Factor One explained 18.6% of the common variance. Figure IX lists
those variables with loadings over 0.5. These variables combine to make
up the dimension or factor. In addition the factor scores for Factor One
are mapped in Figure X. The map shows how the factor is distributed
throughout the region. For example, on the map, locations having a
Factor Score higher than 2.0 on the positive scale would indicate a
place that exhibits a strong concentration of people described by the
positive Factor structure (Categories are based on Berry and Rees,
1973). On the other hand, a location with a high negative score, greater
than -2.0, would indicate a location that exhibits a strong
concentration of people associated with the negative side of the Factor

Structure.

85

FIGURE IX

Factor 1 - Total Variance Explained - 18.6%

High Scores - Positive

Mobilityzgrigin
Concentration of people born in: Changhua
Chiayi
Nantou
Taichung
Tainan
Yunlin

Miaoli
Taoyuan

Ilan
Taichung City
Hsin Chu
Kaoshiung

Pingtung

% of total population with an intra-urban move within last 5 years.

Housing

% Total
% Total
% Total
% Total
Employment
% Labor
8 Labor
% Total

% Total

housing having modern toilet
housing having piped water

housing having 4 stories or less

housing having rental - private sector

force - blue collar
force sales/service/guard - labor
population employed

population employers

86
Edgcatign
% Total population with vocational education
% Total population with senior vocational education
- e ve
liming
% Total housing with traditional toilet
% Total housing with no piped water
% Total housing with traditional farm style
% Total housing with other than established categories
% Total housing with villa
% Total housing with built before 1945
% Total housing with no toilet, bath
MobilitxZOrigin
% Total population born in Taipei City
% Total population born in County
W
% Total population - agricultural labor
% Total population - agricultural foreman
% Labor force in mining
Edugatign
% Total population - illiterate
Concentration in population - illiterate
Ma a tatus Dem ra hic
% Total population - widowed

% Total population male-female 65 + years

 

 

87

FIGURE X

 

FACTOR 1, 1980

Factor Score intervals

greater than 2.0 -
1.0 to 1.9
0.0 to 0.9 i:

0.0 to -0.9
—1.0 to -1.9
less than -2.0

 

 

District Boundaries ---«-:—

Principal Roads

 

 

 

 

  
 
 
 
 
 
  
 
   

Kilometers

O 1 2 3 4 5

      
  

 
 

>""“‘~.—-_‘

 

Ronald D. 710/0nl).1<‘h I988

 

 

 

CC
PE
SE

511'

de:

had
VHS
to I

SQtt

88

The positive scores describe an immigrant population of blue
collar workers. These blue collar workers are employed in a a variety of
largely service type industries. These individuals are relatively recent
migrants to the city, having arrived within the 1975 to 1980 period.
Upon reaching the city they moved into the type of four story apartment
building that was the predominant type of housing being constructed in
the 1960's and 1970's. The ownership pattern shows a strong tendency
toward rental rather than ownership. This indicates either a strong
investment market in real estate, or (as scholarly observation would
lead one to theorize) the housing represents the continued investment,
through ownership of, what was previously, primary housing units by an
upwardly mobile population. The concentration of vocational school
graduates combined with the employment pattern in service, guards and
laborers indicates a blue collar segment of the population that could
probably be described as either lower middle class,or upper lower class.
The strong element of the population that is described as employers
could be misleading. This sector does not describe employers as we
perceive them, but rather small scale entreprenuers providing such
services as night market stalls, street vendors or small scale
sub-contractors to larger industrial concerns.

What is significant about the segment of the population being
described is that it correlates with the observations of Gallin and
Gallin (1974) in their follow-up study of residents of Hsin Hsing who
had migrated to Taipei. In that study Gallin and Gallin theorized that
village associations served as a strong control mechanism in migration
to the city. They felt, based on gathered evidence, that new in-migrants

settled in the immediate vicinity of those from the area of their

89
origin, who had moved to the city at an earlier time. In addition, they
utilized these connections to obtain employment, if not in the same
shop, in the same sector of the economy and in the same general
geographic area. For example, the migrants under study lived in a very
small geographic area and were largely employed in the Central Farmer’s
Market. In this instance the same type of mechanism is at work, with
in-migrants from a number of areas congregating in the manner described
and being employed in the same general sector of the economy.

The population described by the positive side of Factor One have
tended to concentrate west of the Tamshui River in the cities and
suburbs that make up the bulk of the industrial region of the Taipei
Metropolitan Region. In addition they exhibit a large presence in the
industrial suburbs to the southwest of the central city area of Taipei.
The mapped pattern conforms to the perceived pattern of residential
location by the author and a large number of city residents interviewed.
It can also be confirmed by examining the population chart on the map in
Appendix A. This chart shows the growth of the region over time, and
confirms the relatively recent and rapid growth of the area.

The negative loadings in the factor structure describe a segment
of the population that are native to the region, and still engage in
traditional agricultural pursuits. The population described lack
mobility, largely having Taipei or Taipei County as their place of
origin. The type of housing is older, often pre-World War II, and either
lacking sanitary facilities, or having traditional style facilities.
Many of the units lack piped water. The residents are engaged in
agriculture or mining, but are not usually owners of the means of

production, rather they are laborers. The population exhibits a high

90
degree of illiteracy. In addition, the population tends to be an aging
population; a population that is 65 years old or older and one in which
the spouse is likely to have died.

From Figure X we can see that the population described by this
dimension tends to inhabit either the fringe areas of the region, such
as in the southeast near Hsintien; or the more marginal lands (often in
flood plains) such as the peninsula where the Keelung and Tamshui rivers
come together. A third example would be the narrow valleys and slopeland
near both Sanhsia in the south and the area between Neihu and Shihlin in
the northeast. A final example would be in the far eastern section of
the metropolitan region near Nankang.

The population described by this dimension most likely represents
those residents of long standing who have not adapted to the changing
'urban environment. Market forces have not been sufficient to drive them
from their traditional life style, but changing circumstances has
'resulted in the children of these families migrating to other places in
the urban region and following a more urban style of life. The
'population segment described is likely to disappear as the city further
expands, or the older individuals die. Again, as in the positive side,
the statistical pattern matches perceptions.

Factor Ego; §ocio-economic Status- Upwardly Mobile, Educated
Professionals/Indigenous Urban Poor

Factor Two explains 17.7% of the shared or common variance. The
variables which make up the Factor, both the negative and positive
aspects, are presented in Figure XI. The factor scores for each
observation site are presented in Figure XII. In general the factor

describes two additional segments of the population. The positive scores

91

FIGURE XI
Factor 11 - Variance Explained - 17.7%
Hi e se - Positive

Mobilitlerigig

Concentration of Population born in Lianchang

Epstein;
% Total housing 30-39 ping

Economic Activity
% Labor force in High Tech./Private Administration
% Labor force in Government Construction
% Labor force in Sales/Service/Guard-labor
% Labor force in White Collar
Edgcation
% Total population M-F 2-3 year college or more
% Total population M-F high school graduates
Dem a ic
% Total Population - Density
Hi es onse - Ne ative
MobilityZOrigin
% Total population born Taipei City
% Total population born Taipei County
E o tiv t
% Labor force production labor
% Male-Female self employed - no help

% Labor force private sector construction

92
Educa o
% Total population - primary school education
% Total population - junior high school education
% Total population - illiterate
Demo ra c
% Total population 0-14 years old
Concentration of live births/surviving infants in last 12 months

Ewing

% Total housing - Residence-Manufacturing combined

 

 

93

FIGURE XII

FACTOR 2, 1980

Factor Score intervals
greater than 2.0 -

 

0.0 to

-1.0 to
less than

 

District Boundaries “-.. ------

Principal Roads

 

 

 

 
 
 
  
  
 
    
     

 

1

Kilometers

2 3

Ronald D. Tlclenbach 7988

4 5

 

 

 

 

 

de'.

the

POP

has

IESC

educ

in t’.

the :

in ti

the 3‘

admin

94
delineate a middle/upper class segment of the total population, while
the negative scores describe a lower class segment.

The positive, high response variables, in Factor Two describe a
population that is highly educated. The major share of the population
has a minimum of a high school education and a significant share has at
least some college. This is a significant point in a society that values
education, and in which education is highly competitive. Scarce
resources combined with social and governmental policy restrict
education to a progressively smaller percentage of children at each step
in the educational ladder. The ...population described here represents
the intellectual and social elite of society. This is further reflected
in the employment pattern. The population is overwhelmingly engaged in
the white collar sector of the economy in such things as high tech,
administration and government construction agencies. Those in the
sales/service/guard sector of the economy represent higher level
administrative jobs in that category of employment.

The factor also describes a housing pattern in which house size
averages 30-39 212g (a ping is a unit of measurement equal to
approximately six square feet). This would indicate that the residents,
or more properly the area in which they reside, has a significant
percentage of the more spacious housing available in the city. The fact
that the regions associated with the factor are also characterized by
high population densities is not necessarily counter to the trend being
described. The map of the factor scores for Factor Two indicates that
the group being described is largely clustered in a roughly L shaped
area that corresponds to the perceived (by the author and interviewed

residents) upscale region of the central city. This area contains such

95
major academic institutions as National Taiwan University and Taiwan
Normal University; as well as the area locally named "Ding Hao" which
roughly translates as "the best” or ”number one” and which runs the
length of Tun Hua Street (north-south). This is an area that is serving
as a growth pole in the city, particularly around such nodes as the
Hsin-11 project and the World Trade Center, and its development in the
late 1970's and into the 1980's reflects the growing wealth of the
country. The area is sprinkled with high-rise office and apartment
buildings; and even the slightly older and lower profile buildings are
usually more than the four stories common in earlier times. The combined
perception of the region as a highly desirable location (confirmed by
Wang, 1984) and the high rise architectural style has led to much higher
population densities than can be achieved in less desirable areas
usually having lower-rise architecture.

The final element in the positive loadings, not yet explored,
would be the concentration of people born in Lianchang. This could have
two explanations, both of which require further research. First,
residents who have moved from Lianchang to Taipei could be either
brighter than the average, or they could have developed a better support
system for helping residents from the home area adapt to, and succeed
in, the big city. Secondly, and perhaps more likely, the number of
individuals who have moved to Taipei could be relatively small and have
moved to attend college or high school and then remained. The relatively
small number of residents in Taipei from Lianchang could be concentrated
in one area creating a statistical anomaly.

Much of the distribution of this population segment has already

been described. However, there remains two significant concentrations.

96
Both of these are in the Shihlin District to the north of the city
center and north of the Keelung River. Both areas represent a
combination of enclaves of government provided housing for senior
executives, housing associated with two universities, and higher class
housing in the private sector that has come about in association with
existing housing. In addition, the far northern area, known as Tien Mou,
is an area that was once American military, and is now populated by a
large number of foreigners and upscale locals. Lastly, the western tip
of the northern area corresponds to an area known as Peitou which has
been an upscale area since Japanese colonial times. (At that time it was
a resort area for senior members of the government and for leading
citizens). Again, the patterns exhibited correspond to the perceived
residential pattern of the city.

The negative loadings describe the opposite end of the social
scale. The population described is native to the region. It is a
population that is poorly educated, with a primary or junior high
educational level being common, and a high rate of illiteracy being
exhibited. The population, reflective of educational levels, is largely
engaged as either production or construction labor. Those not so
employed largely function as sub-contractors to larger industrial
concerns. This is confirmed by the high score in the housing variable
that describes a combined residential-manufacturing housing unit.
Lastly, the factor describes a population that exhibits families with a
high concentration of children under the age of 15; as well as births in
the 12 months preceding the census.

The last points described, the number of children and the birth

rate, present an interesting sidelight. As can be found in most

97
countries of the world, Taipei residents exhibit birth rates that are
correlated to social status and income. The more educated, more
economically successful segment of the population described in the
positive sector of Factor Two do not have family sizes or birth rates
that deviate from the mean. However, the poorer, less educated segment
described in the negative side of the dimension exhibits a significantly
higher rate than the average. The population described has a
disproportionate share of both children and births. This is particularly
significant given the traditional social structure in Chinese society
that promoted extended and large families; particularly the drive to
have a number of male heirs. Male heirs represented security in old age,
and insured that the family name and line would continue. Even today the
drive to have male children remains, and while the average number of
children per family in Taiwan is 2.2, the ideal family has three, two
boys and one girl.It can be theorized that the higher educated segment
of the population has experienced demographic transformation, while the
lower educated element has not. This would indicate that
industrialization and economic success is transforming the traditional
cultural patterns.

The segment of the population described by the negative loadings
in Factor Two tend to be concentrated in the area west of the city
proper, and west of the Tamshui River (see Figure XII). The largest
concentrations of the described population are located in and around the
cities of Sanchung, Hsinchuang and Shulin. Additional areas are located
to the south of the center city around the city of Panchiao. Lastly,
significant concentrations of the described population are found in the

older sections of the central city. In particular large concentrations

98
can be found in the central city districts of Tatung and Yenping. This
is not surprising as these districts were the heart of the Taiwanese
districts during the Japanese colonial period. Again, the mapped pattern
reflects a high degree of correlation to the perceived pattern.
Fagtgr_1hree: Stage in the Family Life Cycle-New Housing and
Growing Families/Older Housing and Single Women

Factor Three explained 8.3 percent of the total shared variance.
The variables which make up the positive and negative dimensions of this
factor are presented in Figure XIII. The spatial distribution of the
Factor is presented in Figure XIV. In general the positive loadings
describe areas within the city that are undergoing significant growth
and development. The negative scores describe those areas that are
characterized by single females.

The positive variables reflect a mixed population. A major share
of the population described by this factor is new to the areas scoring
high on the Factor. The population has ...been mobile with either an
intra-urban or inter-urban move within the five years preceding the
census. It is a population that is moving into houses built after 1976
and living in areas characterized by low rise apartment buildings built
after 1976. The areas are high in population density, total population
and households. Birth rates are high. In addition there is a segment of
the population that has not been mobile within the past five years.

The mapped pattern for Factor Three indicates that the areas
scoring high on this factor tend to be on the developing margins of the
city. The areas scoring highly on the positive side include such
‘outlying' districts as Neihu, Shihlin, Hsinchuang, Wuko, Hsintien and

Shenkeng. This would be consistent with observation as well as expected

99

FIGURE XIII

Factor Three - Total Common Variance Explained-8.3%

We

Will

Percent of

move

Percent of

move

Percent of
5223.121

Percent of

Percent of

Demographic

Total Population with intra—urban
in the past 5 years.
Total Population with inter-urban
in the past 5 years.

Total Population not moving in the past 5 years.

Total Housing Built Since 1976

Total Housing Units 4 story or less.

Total Households

Percent of

Total Population

Population Density

Concentration (LQ) of Live Birth Within the Past 12 Months.

Hi s-Ne ative

Housing

Percent of

Percent of

Percent of
Demo r c

Percent of

Total Housing 10-19 Ping.
Total Housing Japanese Style

Total Housing Built from 1946-1975

Total Population-Single Females

 

 

FIGURE XIV

 

FACTOR 3, 1980

Factor Score Intervals
greater than 2.0 -

 

0.0 to -0.9 '"
—1.0 to —1.9 ,1'
less than —2.0

 

r

District Boundaries ———«::—_:—_j -----

Principal Roads 7—4

 

 

 
  
 
 
   
      

Kilometers

l—Ffi—T—‘fl—i—fi
O 1 2 3 4 5

Ronald D, Tiefenbmh 1988

 

   

 

 

 

 

 

101
results in a developing city. Taipei is a city undergoing tremendous
growth. As the chief city of the country and the political and cultural
center it is, in fact, attracting an almost disproportionate share of
growth. Graff (1976) showed that the older central areas of the city had
extremely high densities, and that the peak density center of the city
was moving outward. In addition, in Chapter Three it was shown that the
city had expanded outward over time. The mass of people fueling the
growth of the city cannot all be absorbed in the older confines of the
urban area. To accommodate an expanding population the city is moving
outward; and land that was once farmland is being converted to urban
uses. This clearly consistent with the positive side of the factor. The
areas scoring highly on the positive side are regions that are either
being converted from rural to urban uses, or are older urban areas being
redeveloped in response to market demands. While there are no socially
descriptive variables contributing to the factor, it can be theorized
that the segment of the population being described is comprised of
younger families ( as illustrated by the high birth rates) and
successful enough to move into newer housing, but not well enough off to
afford more desirable locations or the high rise apartments favored by
the wealthy. This theory does hold up based on field checks and informal
interviews in the areas described.

The negative side of Factor Three describes a radically different
segment of the population. Simply put, it is a population of single
females. The female population group inhabits areas characterized by
small houses (10-19 pings) and older housing stock, housing built before

1975 and much of it in the Japanese style.

102

The presence of a distinct population group of single females was
anticipated. This group was identified in Hsu and Pannel's (1976) study
and is described in the works of many anthropologists such as Gates
(1982). A number of factors account for the spatial concentration of
single females. Some of the concentrations, such as the area in Shihlin
district just north of the Keelung River and Expressway intersection, or
the areas in the districts of Kuting and Chingmei on the southern edge
of the city proper, can be traced to the locations of Universities. The
Shihlin district is home to Minchuan Women's College and the southern
concentrations are adjacent to Taiwan Normal University and National
Taiwan University. The concentrations sprinkled throughout the urban
area usually reflect the employment structure in Taiwan. As in many
developing countries, young, female labor is in high demand. Industries
such as electronics assembly and textiles often prefer women employees.
This can be traced to the comparative advantage exhibited by this group
in manual dexterity, and the fact that employers can often pay
significantly lower wages. Young woman are recruited from throughout the
country to fill these types of positions. Housing is often provided by
the employer in a "dormitory district". Such areas serve as magnets for
other young women who are migrating to the city or who are city
residents looking to establish residency separate from their families. A
final element that one can speculate plays a role in the pattern (albeit
a small one) would help explain the presence of concentrations in parts
of Sungshan district and in the general area of Lungshan. Both districts
have neighborhoods that could be described as ”red light" entertainment
areas. Further, it was mentioned to the author on numerous occasions

that certain neighborhoods within Sungshan district had high

103
concentrations of "hsiao tai tai's" or mistresses of successful
businessmen and government officials. This was confirmed by field checks
and informal interviews of residents in the area, and with business
people. While these checks did not provide absolute proof, indications
were that this was true; and it certainly coincides with the perceptions
of many of the city's residents.
Faggg: £921: Sgcig-ggonomig Status-Employers and
Entreprenuers/Laborers

Factor Four described 5.7 percent of the total shared variance.
The variables that combine to form Factor Four are set forth in Figure
XV. The spatial distribution of the factor is presented in Figure XVI.
In general the positive factor loadings describe a middle class segment
of the population that is native to the city. The negative scores
describe a working class segment of the population and areas of the city
that were developed in the high growth period of the 1960's to 1970's.

The population group and areas described by the positive loadings
in Factor Four represent a traditional Taiwanese segment of the
population that is native to the City. An examination of the mapped
factor scores show that the factor is particularly strong in areas such
as Lungshan, Chengchung, and Yenping. These are the older areas of the
city, the core areas around which much of the present day City evolved.
Further, they are areas that during the Japanese colonial period, were
reinforced as Chinese areas. Field checks of the area indicate a
profusion of traditional style Chinese shops selling everything from
herbs and spices and other traditional foodstuffs to furniture and

computers. The land

104

FIGURE XV

Factor Four - Total Common Variance Explained-5.7%

n

WELD

Percent
m
Percent
Percent
Percent
Percent
Economic
Percent
Percent

Percent

of

of

of

of

of

of

of

of

P tive

Total

Total

Total

Total

Total

Total

Total

Total

Population Born in Taipei City

Housing From 50 to Over 100 Ping in Size
Housing More Than 4 Stories
Housing in Japanese Style

Housing-House/Commercial Combined

Labor Force Employed
Labor Force-Blue Collar

Labor Force-Sales/Service Labor

Concentration (LQ) of Employers

d n

- e at ve

MobilitleEigin

Concentration (LQ) of Individuals Born in Hualien

Hogsing

Percent of Total Housing Units-Residential

" " -Built 1961-75
" " -20-29 ping
" " -Without Toilet

" " -Type Other

105
Than Listed
Ms
Percent of Labor Force in Production Labor
De 0 1
Percent of Total Households

Percent of Total Population

 

106

FIGURE XVI

 

 

FACTOR 4, 1980

Factor Score Intervals
greater than 2.0 -
1.0 to 1.9

0.0 to 0.9 C]

0.0 to —O.9
-1.0 to -1.9
less than —2.0

 

 

4 .....

District Boundaries --—«-.:—_-_—_—_-

Principal Roads 7—4

 

 

  
 
 
 
 
 
  
  
  
 

Kilometers

O 1 2 3 4 5

 

 

 

 

Ronald D. Tie/'enbat‘h 7988

 

 

di~

av;

on

Ca;

107
use pattern is dominated by very large structures that combine
residential space with both warehousing and commercial/retail functions.
This is not surprising given ...the proximity to the Tamshui River.
Goods moving up and down river by boat would, traditionally, be unloaded
adjacent to these districts and put in storage. The physical mixing of
land uses not only reduced costs but provided security against thieves.

The variables combining in this Factor describe the observed
reality. Housing is described by variables relating to extremely large
size, height, style and function. The areas described contain a
significant share of the region's housing larger than 50 piggg, greater
than four stories, of housing that is in the Japanese style and
functions as both housing and commercial sites. The population is
dominated by a larger than normal share of the labor force, the
concentration of employers, and by those in the blue collar occupations;
particularly sales and service.

The tggg/li scoring high on this factor do, as stated, exhibit the
generated characteristics. The areas mentioned, as well as the
scattering of such areas throughout the metropolitan region, are,
essentially, the remaining areas developed in pre-colonial or colonial
periods. They are the nodal points around which much of the early growth
occurred. The ambiance of the districts is distinctly colonial, with
Japanese/Victorian (Meiji) architectural styles prominent. The streets
teem with people as the areas act as magnets for those seeking the
diversity and range of traditional goods, particularly foodstuffs, not
available in the more modern sectors of the city. The shops spill over
on to the sidewalks and streets, yet a look inside will often show a

cavernous space that serves (or perhaps served) as warehousing. Family

108
activity takes place in and among the commercial enterprises. In short,
the areas described by this factor provide a cultural oasis in a rapidly
homogenizing cityscape, an oasis that provides one with a glimpse of the
past.

The negative loadings, on the other hand, describe a population
and areas that are much more representative of the industrial growth of
the 1950's, 1960's and 1970's. The residential structure described
represents the type of housing built to provide space to a city that was
undergoing rapid expansion as the economy boomed and rural migrants
flooded to the City. The factor describes housing built in the boom
period from 1961 to 1975, and being of moderate size, ranging from 20-29
ping. The areas are densely populated having an abnormal share of both
population and households. The factor also describes areas and a
population group dominated by those engaged in production labor.

This factor describes a large segment of the population and area
of the metropolitan region. A perusal of the map shows that large areas
of the city are associated with this factor. This is not unexpected. The
factor describes a housing structure erected in the initial boom period
of the economy. Certainly, that type of housing stock is prevalent, and
the visitor or researcher is confronted by examples at every turn, and
in almost every area of the city.The style of the housing is dominated
by low-rise apartment blocks, almost uniformly constructed out of
prestressed, poured concrete; and exhibiting a functional look that
could almost be described as lacking in any architectural style. Simply
put, the housing tends to be strictly functional.

The population described is also expected. Taiwan can be

characterized as having an industrial economy. While jobs exist, and are

109
created, in the primary and tertiary sectors, the major share of
existing and newly created jobs are in the secondary sector. The
majority of people are dependent on this sector for their livelihood.
The population described migrated to Taipei from other areas of the
Island during the 1950's, 1960's and 1970's in response to the exploding
industrial economy; and pushed out of rural areas by the declining need
for agricultural labor and the explosion in population growth-and thus
in the agricultural labor force. This group forms, it can be theorized,
the largest single sub-group within the Taipei population. The broad
regions scoring well on this factor are reflective of the development
history of the city, being located in those areas that before 1950 were
agricultural suburbs, as well as the continuing importance of the
industrial sector. At the same time the areas which scored highest on
the negative side are removed from the city center and are on the
periphery of the city in locations such as Chungho, Hsintien and
Chingmei. This reflects the geographic dispersal of industrial
enterprises to outlying districts in the latter part of the development
cycle. These areas are geographically proximate to a great number of
large scale industrial enterprises and reflect residency patterns of
industrial labor. They have not been favored sites for other population
sub—groups.
Fagtg; Five: Socio-economic Status- Blue Collar Company
Employees/ Agricultural and Day Laborers
Factor Five describes 5.5 percent of the common or shared
variance. The variables that combine to form Factor Five are set forth
in Figure XVII. The factor scores for the observation sites are

presented in Figure XVIII. In general, the positive Factor loadings

110
describe a lower class segment of the population, while the negative
ones describe an upper/lower class or lower/middle class segment of the
population.

The positive variable contributors to Factor Five describe a
population that is self-employed without help or who may work for a
relative for no compensation. It is a labor force that is collectively
either blue collar or agricultural in nature. The housing described is
either the traditional farm style or in a category other than one
described by the census form. The farm style housing is most likely
associated with the segment of the sub-group engaged as agricultural
labor. The other category of housing is most likely associated with the
blue collar segment of the population which is either self-employed or
working for ...relatives. The type of housing that does not fit the
published category can be diverse in nature. It can range from a knocked
together ‘shanty' in one of the few remaining areas of this style of
housing, to sleeping space provided in a commercial enterprise. The
space provided in stores would be exchanged for labor, particularly in
guarding the enterprise in off hours. The limited use of the commercial
space for housing purposes would not be sufficient to cause such space
to be classified as combined residential/commercial housing units.

The factor scores for the positive dimension of the Factor are
presented in Figure XVIII. The strongest areas of response are
peripheral areas where the norm is farm style housing and agricultural
labor. The more urban side of the dimension can be found in areas such

as Chengchung near the original CBD of Taipei. The description offered

111

FIGURE XVII

Factor Five - Total Common Variance Explained - 5.5%
High Rgsponse - Positive
Emerita
Concentration (LQ) of Self Employed Persons with no help
Concentration (LQ) of Those Working For Relatives For Free
Percent of Labor Force-Blue Collar
Percent of Labor Force-Agricultural Labor
Reusing
Percent of Total Housing-Traditional Farm Style

Percent of Total Housing-Type Other Than Listed

High Response - Negative

Economic

Concentration (LQ) of Self Employed With No Help
Concentration (LQ) of Those Working For Relatives For Free
Percent of Labor Force-Blue Collar
Percent of Labor Force-Agricultural Labor

Housing
Percent of Total Housing-Traditional Farm Style

Percent of Total Housing-Type Other Than Listed

b l t ri

i

112

Concentration (LQ) of Individuals Born In:

Egonpmig

Percent
Percent
Percent
Percent

Housing

Percent

Educapion

Percent

Percent

of

of

of

of

of

of

of

Labor

Labor

Labor

Labor

Total

Total

Total

Ilan, Hualien, Taitung, Keelung

Force in Private Construction
Force in Mining
Force in Production Labor

Force-Blue Collar

Housing-Supplied by Private Company

Population With Vocational Education

Population With Senior Vocational Education

 

 

FIGURE XVIII

 

FACTOR 5, 1980

Factor Score Intervals

greater than 2.0 -
1.0 to 1.9

0.0 to 0.9 E:
0.0 to —0.9

—1.0 to -1.9
less than —2.0

 

 

District Boundaries “attirii

 

Principal Roads 7—4 *'

  
 
 
 
 
  
 
 
  

Kilometers
fi—T—fi—Tfi

O 1 2 3 4 5

     

      

       

 

Ronald D, frets/1b." h 1988

 

 

 

 

 

 

gr.

ad.-

COI

114
by the Factor does conform to the perceived reality. For example, the
area of Chengchung being described is an area of intense commercialism,
combining shopping and entertainment, particularly nightclubs and bars.
It is an area that is alive with food stands, street peddlers and an
extensive night market.

The variables that contribute to the negative side of this factor
describe a blue collar segment of the population, engaged in such
occupations as mining and construction. The population being described
is characterized by having achieved some level of vocational education.
As such it is likely to be a population that is engaged in crafts or
trades and likely to be higher paid than a population sub-group engaged
in similar occupations.( Factor Three described a similar group, but
having less education. The difference can be thought of as the
difference between an apprentice and a journeyman.) The population
sub-group described lives in areas that have a disproportionately large
share of the region's housing that is provided by the company. This
would tend to reinforce the theory that the sub-group is in a
specialized craft or trade, and thus form a valuable employee resource
group. Also at work here, it can be theorized, is the same in-migration
adaptation described in Factor One. The Factor indicates an abnormal
concentration of the region's migrants from Ilan, Hualien, Taitung and
Keelung. It would appear that the process of using previous migrants to
Taipei from the home area as adaptive resources is in force. While at
this point it can only be theorized, it does conform to past theory and
provides an interesting area for further research.

The areas exhibiting the strongest negative factor scores for

Factor Five are presented in Figure XVIII. Two strong response areas

115
exist to the south of the city proper in Yungho and Panchiao. An
additional area of particularly strong response is to the west of the
city proper near Nankang. Field checks of these areas indicate adjacent
slopeland that is heavily mined, as well as numerous large scale
factories with associated housing blocks.
Facpo; §in; Socio-economic Status- Government Employees/
Urban Poor

Factor Six explains 5.3 percent of the common or shared variance
of the data base. The variables that combine to form the positive and
negative dimensions of the factor are set forth in Figure XIX. The
factor scores for the spatial units are presented in Figure XX. In
general the variables that form the positive element of the factor
describe a population that is middle to upper class and heavily
representative of government official, both civilian and military. The
variables that combine to form the negative dimension of the factor show
the opposite end of the social spectrum, that is a population that is
poor and illiterate.

The population outlined in the positive dimension of the Factor is
a population that is white collar, in the military, or in the civilian
government. Much of the population exhibits advanced vocational
education. The areas contributing to the positive dimension show a
disproportionate share of the region's government owned housing, as well
as housing built in the period from 1946-1960. The areas in question

also exhibit a higher than normal share of total population.

116

FIGURE XIX

Factor Six - Total Variance Explained - 5.3%

Hi S 0 es - Positive

Mobilipy

Concentration (LQ) of Total Population Born In Lianchang

Mains
Percent
Percent

Employnent
Percent
Percent
Percent

Emmerich
Percent
Percent

Demographic

Percent

of Total

of Total

of Labor

of Labor

of Labor

of Total

of Total

of Total

H S are - Ne at ve

MobilityZOrigin

Percent

of Total

Housing Government Supplied

Housing Built 1946-60

Force-Military

Force-Government Construction or Service

Force-White Collar

Population With Vocational Education

Population With Senior Vocational Education

Population-15 - 65 Years Old

Population Born in Taichung

Tainan City or Hualien

Honsing

117

Percent of Total Housing-Privately Owned

Econonig
Percent
n
n
n
n
n
Edugntipn
Percent
De 0 r c
Percent
Percent
Percent

-No Piped Water

-Traditional Toilet

-Traditional Farm
Style

-Built Before 1945

-40-59 Ping

of the Labor Force-Agricultural Labor

of Total

of Total

of Total

of Total

-Private Construction
-Employers
-Employed
-Self Employed No Help

-Working For Relatives for Free

Population-Illiterate

Population- 0-14 years old

Population-Female 65+

Population-Female Handicapped

 

 

 

118

FIGURE XX

 

FACTOR 6, 1980

Factor Score Intervals

greater than 2.0 -

 

0.0 to —O.9 .-

-1.0 to -1.9
less than —2.0

 

District Boundaries ""'~""f":;"-

Principal Roads

   
 
 
   
  
        

1

Kilometers

2 3 4 5

Ronald /) freienbarh 1‘!le

 

 

 

119

This dimension is also an expected dimension, and its emergence
can be found in the developmental history of the City. In the period
following the defeat of the ...Kuomnintang on the Mainland of China and
the retreat to Taiwan in 1949, the island was not an economic
powerhouse. The Japanese had developed the primary sector of the economy
and put in place a reasonably good transportation system in the years
before 1945. In addition small commercial, industrial and service
sectors of the economy were in place. The political and social upheavals
experienced by Taiwan after the end of WWII and Retrocession inhibited
growth of the domestic economy. In 1949 more than a million people moved
en masse to the Island. A good part of those people were the remnants of
the KMT armies led by Chiang Kai-shek. Within the city of Taipei
elements of that migration occupied the areas of the city vacated by the
colonial rulers. During the 1950's significant attempts were being made
to reorganize agriculture through land reform, and to stimulate other
sectors of the economy. Remember though, this period in history is a
period of high tension, particularly in Asia. Not only did the KMT
arrive with a siege mentality and a desire to rebuild the shattered
military and governmental structure, but the Korean Conflict with its
associated domino theory, fanned the flames. Within Taipei, although the
groundwork was being laid for an economic miracle; the expanding sectors
of the economy were the military and the government. Both offered a
level of economic security unavailable in the general economy. The
development of the military and governmental sectors in terms of
employment, was accompanied by development of enclaves to house those
employed. Many of those sites that were being developed were within the

Taipei Metropolitan Region which was fast emerging as the principal city

120
of the Republic of China. The white collar, middle class that began to
expand in this period had strong ties to the ruling party and new middle
class housing developments began to sprout up in the geographical
vicinity of the government enclaves.

Factor Six identifies the remaining elements of this distinct
group. It is the (presumably) aging elements of the middle class of the
1950's whose fortunes were tied to the military and governmental
expansion by the KMT in Taiwan. Although information concerning the
social identification as ‘Mainlander' and ‘Taiwanese' could not be
obtained; one can theorize that the population subgroup identified by
this factor represents the remaining elements of first generation
immigrants from the mainland, as well as those in succeeding generations
who continue to socially identify as ‘mainlanders', a social
identification that is spatially expressed. In earlier studies of Taiwan
researchers often used variables that clearly identified the point of
origin of city residents. The researchers had variable categories that
included, as points of origin, all provinces on the Chinese mainland.
From this data base they were able to establish which groups were recent
immigrants from the mainland. This was important as the
Mainlander-Taiwanese split was generally accepted as both real, and as
an integral part of the social, economic and governmental matrices. This
split was actualized in Taiwanese society in a manner similar to ethnic
splits in the population in other countries. In Taiwan it was generally
believed that the Mainlanders had control of the political process,
while the Taiwanese controlled much of the economy. These categories
have been dropped from the published census data. While no official

reason could be found for their disappearance, informed speculation,by a

121
variety of interviewees, held that these categories had been eliminated
in order to downplay domestic divisions in the population. Whatever the
reason is, it does represent a fundamental shift within a government
that prior to this time had used the diverse regional base of the
population as a major underpinning in its claim to being the sole,
legitimate government of China.

The scores mapped for Factor Six on Figure XX, would tend to
validate the belief that this group represents the remnants of the
Mainlanders. The concentration of strong response sites in the central
city area corresponds to the area extending out Jen Ai Boulevard that
was the heart of the Japanese residential district. It is an area that
is, today, dominated by the massive Chiang Kai-shek memorial, and
extending eastward from there along the Boulevard. This is an area that
formed one of the initial residential districts for government officials
in the 1950's. The map shows that various areas of strong response
extend quite a distance eastward from the Chiang Kai-shek memorial. This
reflects not only additional government/military compounds, but middle
class housing that arose in geographic proximity to these (in the
31950's) centers of social prominence and political power.

In more recent times many of the pure military establishments in
tflne city proper regions have been converted to other uses. A prime
el'lample would be the new World Trade Center complex in Sungshan district
that is built on the site of a former military arsenal. Nevertheless
Scattered sites, although smaller, do still exist in scattered locations
1J1 the city proper. However, many of the high factor score sites in the
‘31ty proper are reflective of enclaves of government owned civilian

lRousing. For example the strong response in the southeast portion of the

122
map in the Kuting district corresponds with the housing complexes
associated with the National Taiwan University. A similar complex, and
area of high factor scores can be found in the extreme eastern portion
of the map, to the southeast of the district center of Nankang. This
concentration corresponds to the extensive complex associated with
Academia Sinica.

Conversely, the areas to the north of the city exhibit different
associations. The areas immediately north and south of the expressway
are government complexes associated with the military. The large, rather
odd shaped area, lying north of the expressway and along the banks of
the Keelung River reflect a particularly large and extensive military
complex. The area has a diversity of military agencies as well as
extensive housing projects. The centerpiece of the mini-district is the
‘Tomb to the Martyrs of the Revolution' a combination museum and
memorial similar to a combination Tomb of the Unknown Soldier, Iwo Jima
Memorial and other cultural monuments in Washington D.C.. The area that
shows a strong response in the far northeast corresponds to a
governmental enclave for high officials. Chiang Kai-shek maintained one
of his homes in this area, and it is generally acknowledged that one of
the many unofficial presidential homes is still in the vicinity. It is
an area of large homes provided to senior government officials and
extensive middle class housing in the immediate proximity.

In short, the pattern identified in the positive scores of Factor
Six conforms to observations by the researcher, and the mental maps of
those city residents interviewed for this project. It is felt that the

regions established by the analysis do exist, largely in the locations

123
identified, and can be established with some degree of historical
accuracy.

The variables which serve to describe the negative dimension of
the factor describe the opposite end of the social spectrum. The
population described contains a disproportionate share of the
illiterate, the elderly and handicapped females, and those employed in
agriculture and construction. It is a segment of the population that
exhibits a significant share of small entreprenuers working alone, using
the free help of relatives, or employing very few individuals. It is a
segment of the population that has an extremely high birth rate as
reflected by the presence of a statistically high share of children. The
areas associated with the population contain a large share of the
region's older housing, built before 1945, lacking piped water, having
traditional bathroom facilities, built in traditional farm styles, and
units privately owned and relatively large.

A perusal of the mapped scores for Factor Six presented in Figure
XX does not present any significant surprises. The population and
housing stock defined by the negative scores is spread across the entire
metropolitan region with a few significant concentrations. It must be
remembered that Taipei is a city that is experiencing significant growth
and change. Spatial units, even though as small as the pppn/ i, are
aggregate levels. They can often contain a number of radically different
population groups and housing types. The presence of up-scale, high rise
housing in one part of the spatial unit does not preclude an area of the
spatial unit characterized by traditional style housing. The presence of
both types in a particular unit can, further, represent a significant

variation from the normal amount of that type of housing for the

124
statistically average spatial unit. Thus, a given spatial unit, can
exhibit strong responses (or factor scores) on a number of diverse
factors. The broad pattern of spatial association with the factor can,
in a sense, be thought of as a continuum.

The city of Taipei emerged and developed many years ago. At the
height of the Japanese colonial period it was a fair sized city.
Physical elements of that city do not disappear overnight, anymore than
do the illiterate in an increasingly educated society. If this is
recognized one can view the pattern as a continuum in which many
locations have a weak association with the factor (that is, some
elements of the population/housing structure being defined), other
locations have a stronger tie and some locations a strong
identification. It must also be remembered that not every household or
individual exhibits every characteristic. The factor is a combination of
variables and must also be thought of as a dimension.

This is not to downgrade the weak pattern of differentiation
exhibited in much of the city. This demonstrates that the city is
undergoing change. Even with the growth and development experienced in
the past forty years, traditional elements of housing are still present
on a widespread basis. Further, even with a high degree of social
development, ‘lower class' groups or groups such as elderly and
handicapped females are still fairly widely distributed over a large
geographic area. It can be theorized from the above information, that to
some degree traditional patterns of social arrangement are still in
force, particularly in the middle and lower class areas. In these areas
we find significantly less social sorting based on age, class etc. It

can also be theorized that the extended family structure may retain more

125

strength and flexibility in these areas than in others; something that
is traditionally characteristic of Chinese society. At the same time,
areas such as the one in the northeast of the city proper, east of
Sungshan airport, and extending south into the Taan district exhibit a
much stronger association with the Factor structure. This could indicate
areas resistant to change, not yet redeveloped, or perhaps areas of
emerging distinctiveness related to western style social sorting. The
overall pattern is interesting, and is one that demands further study
particularly in tracking the changes that occur by the time of the 1990
census.
Facto; Spvgn; Socio-economic Stapug- Poor Blue Collar and

Service Workers/Poor Elderly and Handicapped

Factor Seven accounts for 5.3 percent of the total shared
variance. The variables which combine to form the positive and negative
elements of the dimension are set forth in Figure XXI. The factor scores
of the spatial units is presented in Figure XXII. In general the
positive elements of the dimension describe a blue collar population
group inhabiting poorer regions of the city. It is a group that is
largely native to the City. The negative scores describe an older
segment of the population living in reduced circumstances and in areas
with a disproportionate share of the city's vacant housing.

The positive loadings indicate a population group that is native
to Taipei. It is a blue collar labor force and heavily represented in
the sales and service fields. It is a group that has a spatial
association with housing situations in which essential kitchen and bath

facilities are shared by households, by housing built before 1945

126

FIGURE XXI

Factor Seven - Total Variance Explained - 5.3%

W
Mobilityzoiigin
Concentration (LQ) of Population Born in Taipei City
mam;
Percent of Total Housing-Share Bath
Share Kitchen
Share Piped Water
Share Modern Toilet
Share Trad. Toilet
Built Before 1945
Occupied
Japanese Style
Rental-Private
House/MFG Combined
House/Commercial
Combined
Enploxnenp
Percent of Labor Force-Blue Collar

Percent of Labor Force-Sales/Service Labor

127

c e - e a ve
MopilityZOEigin

Concentration (LQ) Population Born In Taipei City

Concentration (LQ) Population Born In Taipei County

Housing
Percent
mm
Percent
Percent
Education
Percent
De a
Percent
Percent

of Total Housing-Vacant
No Piped Water
Without Kitchen

More Than 1 Person

of Labor Force-Agricultural Labor

of Labor Force-Self Employed-No Help

of Total Population-Self Educated

of Total Population- Over 65

of Total Population-Handicapped

 

 

FIGURE XXII

 

   
 
 
 
   
     

FACTOR 7, 1980

Factor Score Intervals
greater than 2.0 -

 

0.0 to —O.9
4.0 to —1.9i
less than -2.0

 
 

Kilometers
i—I—r—r—r—i

O 1 2 3 4 5

 

District Boundaries -—-«-: _<

Principal Roads —7—-4

t-fa.’

 

 

 

 

Ronald D. Tieienbach 1988

 

Ii

di

38

129
and often in the Japanese style, or housing situations that combine
residences with either commercial or manufacturing concerns. It is a
group that by and large rents its housing, with ownership rates being
low. Lastly, it is a group that is associated with areas in which
vacancy rates are extremely low and occupancy rates among the highest in
the city.

A review of the mapped scores for Factor Seven as presented in
Figure XXII indicates that the described population is heavily
represented in areas adjacent to the traditional CBD of the city. In
addition, there are significant concentrations in locations such as
northern Sungshan and western Neihu and in northern Shihlin. These
concentrations are associated with older urban nodes, as well as major
institutional complexes such as the Veterans Hospital complex in
northern Shihlin. The Veterans Hospital region is home not only to a
very large hospital complex but a number of associated institutions such
as rehabilitation institutes and mental health facilities. These
complexes require large amounts of labor, usually low paid, and draw
large numbers of users. Many locate in the immediate vicinity and
require the type of lower cost housing described. Another example of
this type would be the heavy concentration in the near south of the city
proper. The location in question is just south and east of where the
rail line takes a 90 degree turn from a north-south orientation to an
east-west one. This area is in the immediate vicinity of National Taiwan
University Hospital and Medical School.

The variables which combine to form the negative segment of this
dimension describe a population native to the region, that is

self-educated, older and often handicapped, and either self-employed or

130
in agricultural labor. It is a population group that is associated with
housing that lacks piped water and kitchens yet houses more than one
person. It is, not surprisingly by western standards, associated with
areas having abnormally high vacancy rates.

A review of the mapped scores in Figure XXII indicates that the
population described is strongly associated with areas peripheral to the
city proper, although scattered concentrations are present in the city.
A comparison of the factor score map with the map of land use patterns
of the Taipei Basin (see Appendix A ) indicates that the areas
associated with this dimension show a strong correlation with the
slopeland land use category. The associated sites correspond to areas of
slopeland. This slopeland is land lying 100 meters or more above seal
level, and which is used for a variety of agricultural purposes. The
described dimension dovetails with the observed land use pattern. Areas
showing a strong association with the dimension within the city proper
can be expected to a small degree in any city.

Eantn;_§ign§: Sopip-gconomip Status- Female Labor Force/
Poor, Low Education, Mixed Land Use, Labor
Force

Factor Eight explains 5.1 percent of the total shared variance.
The variables that combine to form the positive and negative elements of
the dimension are set forth in Figure XXIII. The factor scores for the
observation sites are presented in Figure XXIV. In general the positive
loadings describe a younger population that is economically very active
and which has large numbers of single females, females in the work force
and few children. The variables combining to form the negative side of

the dimension describe a population which is dominated by self-educated

131
senior citizens and children. It is a population group which earns its
wages as production labor or in self-employed small scale enterprises,
often manufacturing concerns within the home.

The variables which form the positive segment of the dimension
describe a non-native population, a population that is characterized by
people from 15 to 64 years of age, by single females and females from
15-29, and a population with vocational educations. It is a population
that, economically, is characterized by females in the work force, by
sales, service and construction labor, by those working for relatives
for free; and in general by the high numbers of residents simply
working. The population sub-group lives in areas with a disproportionate
share of housing units combined with businesses.

The map of the factor scores, Figure XXIV indicates that the
heaviest association or concentration of those described by this
dimension live in Chungshan district, an area immediately adjacent, (to
the northeast), of the CBD. Field checks indicate an area that is
dominated by mixed land uses although having a significantly large
presence of high rise office buildings and tourist hotels. It is
perceived as a desirable urban neighborhood, and in fact served as an
upscale area before the emergence of a secondary upscale growth pole in
central Sungshan district. That district has served as a secondary CBD
and replaced the Chungshan district as the primary upscale area for the
Taiwanese. The Chungshan district is, however, still perceived as a
highly desirable residential district. The offices and hotels in the
area, along with the Tatung Industrial Complex and vocational school
serve as magnets to the younger elements of society, particularly

unmarried females. The concentration of economically active people,

132

FIGURE XXIII
Factor Eight - Total Variance Explained - 5.1%
- ve
ob t
Concentration (LQ) Population Born In Taichung and Tainan
Housing
Percent of Total Housing-Residence/Business Combined
Economic
Concentration (LQ) Labor Force-Females
Working
-Construction
Labor
-Sales/Service
Labor
~Work For Free
For Relative
-Employed
Demographic
Percent of Total Population
Percent of Total Population-Single Females
-Females 15-29

~M/F 15-64 Years Old

Edupatipn

133

Percent of Total Population With Vocational Education

Percent of Total Population With Senior Vocational Education

H c e - Ne at ve
Housing
Percent of Total Housing-Residence/Mfg. Combined
Percent of Total Housing-More Than 1 Person
Mien
Percent of Population- Self Educated
E om
Percent of Labor Force-Production Labor
Percent of Labor Force- Self Employed No Help
emo
Percent of Total Population-O to 14 Years Old
Percent of Total Population- 65 or older

 

 

134

FIGURE XXIV

   
 
 
 
   
 
    

FACTOR 8, I980

Factor Score Intervals

greater than 2.0 -
1.0 to 1.9

 

0.0 to —0.9 '5
-10 to —1.9.
less than —2.0

 

 

Kilometers

District Boundaries "-3: ------

Principal Roads 7-4

2 3 4 5

 

 

Ronald I). Tieienhat'h 1988

 

 

 

Sui
the

CE!)

135
particularly females from age 15 to age 29, combined with the absence of
any statistically significant variation in the number of children, could
lead one to theorize that the area represents a region that functions as
a residential district for those in the household formation stage, who
may move once more children are born. Lastly, the high percentage of
housing that serves both residential and business purposes is consistent
with the area. Being a desirable district, and one that is heavily
populated by the economically active means that rents tend to be high.
Significant savings can be achieved by combining uses within the housing
unit. In addition, the presence of a significant sector of the
economically active population, particularly the single female
population, would indicate a population that, taken as a whole, can
demonstrate significant purchasing power and disposable income. This
serves to increase the number of businesses present in the area and
serve to facilitate the combined usage.

The variables that combine to form the negative element of the
dimension describe a population that is self educated, engaged in
production labor or is self-employed with no help. It is a population
that is characterized by children and those over 65 years of age. It is
a population that is associated with housing that combines the residence
with a manufacturing operation; and houses more than one person.

A review of Figure XXIV indicates that observation units that show
a strong correlation with the dimension tend to be in the peripheral
suburbs in such districts as Taishan and Wuku that lie to the west of
the city proper, as well as in scattered neighborhoods in the industrial

centers such as Sanchung, Hsinchuang and Panchiao.

136

Field checks in these areas indicate that the described housing
type is indeed common. Many apartment units, particularly at street
level, are open to the street, and have had the front half or more of
the apartment converted to a manufacturing use. The manufacturing space
will often contain such machinery as a mechanical loom, or an injection
molding machine. The housing space serves as the workplace and family
life takes place around and within the factory setting. The visual
impression is one of poverty, but of a poor element that is attempting
to succeed. The poorer nature of the group tends to be confirmed by the
self-educated nature of the population and by combining that with the
knowledge of the relationship that exists between education and income.
Given the lower income and educational levels the significant number of
children present is not unexpected. Worldwide there is a demonstratable
correlation between birth rates and education/income. The presence of
significant numbers of elders would indicate the continued persistence
of the extended family, and most likely (based on observation) the use
of the elders in child care so that the parents can both be freed up to
bring in income.
anpp; fling; Stage in The Ennily Lifp Cycie- Elderly Poor/

Lower to Middle Class Young Families

Factor Nine explains 4.9 percent of the total shared variance. The
variables which combine to form the positive and negative elements of
the dimension are set forth in Figure XXV, and the scores for the
observation sites are presented in Figure XXVI. In general the positive
element of the dimension describes a population sub-group that can be
characterized as a self-educated, elderly, living in single family homes

that are small, substandard and rented. The negative elements of

137

FIGURE XXV
Factor Nine - Total Variance Explained - 4.9%

High Scores - Positive

timing
Percent of Total Housing-No Kitchen
-No Toilet
-No Bath

-Share Traditional Toilet
-Built Before 1945
-Built 1946-1960
-Rental-Government Owned
-5 to 19 Ping
-Japanese Style
-Type Other Than Listed
-Other Than Listed Owner
e a 1
Percent of Total Households-Single Person Family
Percent of Total Population-65 Years or Older

Mica

Percent of Total Population-Self Educated

138

High Spoges - Negative
margarine
Concentration (LQ) of Population With Inter-
or Intra-Urban Move Within Past 5 Years
Concentration (LQ) of Population Not Moved
In Past 5 Years
Imam;
Percent of Total Housing- Without Bath
-With Kitchen
~20-29 Ping
-Rental-Private
-Built 1961-1975
-More Than 1 Person Household
Econonip
Percent of Labor Force-Private Sector Construction Labor
Percent of Labor Force-Sales/Service Labor
emo c

Percent of Total Population-O to 14 Years Old

 

fir

. FIGURE XXVI

 

 

  
  
 
 
   
      

FACTOR 9, 1980

Factor Score Intervals
greater than 2.0 -

 

0.0 to -0.9

—1.0 to -1.9
less than —2.0

Kilometers

2 3 4 5

 

District Boundaries "“iti'.i’_“ """

Principal Roads 7—4

 

 

Ronald D Tielenbarh 1988

 

 

01
19
in

in

ob,

C01

adj

alo

140
the dimension describe a population sub-group that can be characterized
as a combination of mobile and settled with both living in multiple
person households that are rented and that are sub-standard. It is a
poor population that is in low paying labor situations and that tends to
have large families.

The population described in the positive element of the factor is
dominated by individuals over the age of 65 and living in single family
households. This would indicate the absence of an extended family
structure, at least in a residential sense. The population is largely
self-educated; hardly surprising given the age characteristics and the
relative lack of educational opportunities when they were children.
Given the relationship between education and income it can be theorized
that income is low, a circumstance further complicated by the apparent
lack of extended family structure. This low income status is reinforced
by the variables contributing to the housing segment of the positive
dimension. The housing described lacks kitchen and bath facilities and
require the sharing of traditional style toilets. The housing stock is
older having been constructed prior to 1960. It is small, averaging 5 to
19 ping. Much of the housing stock is from the colonial period and built
in the Japanese style. It is either rented from the government, or held
in an undefined ownership structure.

A review of Figure XXVI, which shows the factor scores for the
observation sites relative to Factor Nine, indicates a significant
concentration of this style of housing in Chengchung in neighborhoods
adjacent to the administrative center of the city. Much of the housing
lies on the fringe of the neighborhood known as Hsi Men Ting, which lies

along Chung Hua Road south of the intersection with Chung Hsiao Road.

141

This neighborhood spreads back westward from the indicated sites that
lie parallel to the rail line. This is significant in that it is an area
that in the 1950's was redeveloped as a neighborhood for retiring
soldiers, in particular those soldiers who accompanied Chiang on the
retreat from the mainland. The neighborhood included the development of
a three story shopping mall parallel to the rail lines. The shopping
mall was meant to provide economic opportunity for the veterans, and
housing was associated. This would dovetail with the group being
described by the dimension. The veterans, particularly the lower ranks,
were many times forced by circumstances to abandon their natal families,
and often their spouse and children in the Mainland. They often found
themselves in later years as an aging population, forming single person
households, and dependent on the government. Their housing would
correspond to that requisitioned or built with limited resources in the
1950's in order to care for this population group. The scattered sites
in other locations are most likely associated with government enclaves
set aside for this population sub-group. Further research is indicated.

The variables which combine to form the negative aspect of the
dimension describe lower class population sub-group. It is a sub-group
that is similar to others described. It is a group made up of large
numbers who have moved, either inter or intra-urban, within the past
five years; as well as those who have not, but who have not moved up the
economic ladder. It is a population group that exhibits the large
families usually associated with the less well off; and a group that is
engaged in labor in the construction, sales and service fields. The
population forms more than one person households, but is forced to

occupy substandard housing without kitchens or baths and of moderate

142
size, due to economic circumstances. While the housing is relatively
recent, it is more reminiscent of the type of railroad flats created to
service our own immigrant population in earlier years. The reduced
economic circumstances, as well as the exploitive nature of the
particular housing type, is confirmed by the overall rental pattern from
private interests.

Figure XXVI shows the factor scores for observation sites relative
to this dimension. Significant areas of concentration can be found in
the area of Hsuangyuan, to the south of the CBD, in the areas along the
Tamshui River. Addition significant concentrations can be found in the
industrial suburb of Panchiao, in the industrial city of Sanchung, and
in the industrial suburb to the north of Sanchung, Luchou, and in the
older regions of Shihlin district. Other sites exist both in peripheral
regions and within the city proper as well. It can be theorized that the
style of housing, combined with the years built and the rental pattern
have a possible explanation in the trickle down theory of housing. This
housing may indicate the emergence of western style housing mechanisms.
The housing stock was constructed during the initial boom period of
post-WWII Taiwan. One can theorize that while some of the original
inhabitants never progressed significantly in an economic sense, others
did. Those that did may have moved onward and upward, retaining
ownership of the original housing site for investment purposes, and
providing a housing market for recent in-migrants to the city. It is
certainly a factor which deserves to be singled out and investigated

more thoroughly.

CHAPTER FIVE

CONCLUSIONS
Hvsinnping phe Hypothesis

The hypothesis for this study stated, in essence, that separation
between the factors of socio-economic status, stage in the family life
cycle, and ethnicity would be present in the Taipei of 1980. The
question then is do we accept the hypothesis or reject it?

Factor One can be characterized as a socio-economic dimension. It
describes a population group that ranges from the lower to lower middle
class. This is confirmed by the variables that contribute to the
dimension. In education they range from illiteracy to senior vocational
education. In light of the survey by the Institute of Economics at the
Academia Sinica in Taiwan mentioned earlier, this would correlate to the
lower end of the economic scale. In housing, the variables range from
pre-WWII housing with no facilities to functional housing built more
recently. The employment variables depict a population that is blue
collar in its makeup. At the same time, a significant contribution to
the dimension is made by a segment of the population age 65 and over.
The question arises as to whether this indicates their inclusion due to
age (stage of the life cycle) or income? Before assessing the answer to
this question it is instructive to review the second factor extracted.

On the surface it would appear that Factor Two is also a

socio-economic dimension. The positive and negative scores describe both

143

144
an upper class, highly educated population segment, and a poorly
educated, lower class group. However, the negative scores in this factor
include high concentrations of live births and children. Thus, the same
type of situation exists as in Factor One.

This pattern is repeated throughout the entire Factor structure.
It appears that there is no separation between the socio-economic and
stage in the family life cycle dimensions. At the same time, Factor
Three describes a population of single females, and Factor Six a group
that displays the characteristics of being in the household formation
stage.

The third expected dimension, ethnicity, was eliminated early in
the study. The lack of data concerning places of origin outside of
Taiwan made it impossible to authoritatively establish its existence.
Although aspects of Factors Six and Nine can be interpreted as
indicating mainland backgrounds in the population no absolute proof is
present. Thus testing of the hypothesis must rest on the separation
between socio-economic status and stage in the family life cycle.

At the same time, if we define ethnicity as place of origin, then
indications are that this is still a factor in residential location.
Earlier in this study the work of Gallin and Gallin (1974) and others
was cited, showing that one of the mechanisms determining urban
residential location traditionally, was place of origin. Many of the
factors have as a component the variables identifying various
origination points from around the island. In each case where they play
a role it indicates that there exists in the associated locations a
strong deviation from mean; or a greater than average concentration of

individuals from those places of origin. In terms of the original

145

hypothesis, the ethnicity factor (if defined in this manner) is not
separate from the others.

Given these conflicting patterns the hypothesis must be rejected.
There is no clearly defined separation between the two dimensions that
usually dominate the social landscape in other cities that have been
studied. The pattern in Taipei would still seem to tend toward the
pre-industrial model set forth by Sjoberg. The rich are clustered in the
city center along the Tun Hua Road and Nanjing Boulevard axes. The
middle class form the next ring and the poor and recent in-migrants form
the outer boundaries. At the same time, significant deviations to this
pattern exist. Factors Three and Six indicate groupings that are based
on stage in the family life cycle, young females and household formation
stage respectively. In addition, Factor Seven would indicate an elderly
dimension within the social structure. Based on these the theory can be
put forth that Taipei is in the process of change. That is, while the
city does not yet exhibit a pattern similar to industrial cities in the
West, it appears to be moving in that direction. At the least it points
out the need to continue research of this type. If the city is indeed
evolving into the Western pattern, geographers have the opportunity to
significantly expand the disciplines knowledge concerning urban social
processes.

However, the fact that the factor structure for Taipei does not
meet the western model is only part of the evaluation. Abu-Lughod (1968)
and Berry and Rees (1969) investigated the structure of non-western
cities and formulated a set of 'necessary conditions' for factor
separation, as well as a fuller set of expectations concerning the

factors extracted. The residential structure for Taipei could fit either

146
of two 'factorial combinations' set forth by Berry and Rees (see Chapter
Three for a full discussion of this theory). Combination Five in Berry
and Rees's scheme indicates one type of factor being extracted, that of
Socio-economic status. Combination seven, in the scheme, indicates a
mixture of socio-economic status and family status; with further factors
lumped under a general category of urbanization that includes elements
of stage in the life cycle and ethnicity; but primarily consists of jobs
associated with an urban environment, such as industrial.

On the whole, the first case (Combination Five) fits best, but
with significant modification, or at least explanation. If one refers
back to Abu-Lughod's (1968) 'necessary conditions' the situation in
Taipei meets the conditions for all three factors. The reason that clear
separation or 'disassociation' between the factors is not evident is
explained by Abu-Lughod's set of conditions. In the case of
disassociation between Socio-economic Status (SES) and Stage in the
Family Life Cycle (LC), Abu-Lughod proposed that either of two
conditions must be met. The first condition was that there was little or
no association between social class and family type. Given that it was
indicated that social class does exist in the residential market, as
shown by the factor structure for Taipei, the question is, does an
association exist between such class and family type. At this juncture
the answer is not clear, given the lack of income data for the region.
However, indications are that associations do exist. Earlier in this
study the problems of obtaining accurate income data was discussed. One
major problem related to the contributions to household incomes from the
extended family structure. If social status is linked to income, and

expressed in residential choice, then the fact that total income

147
reflects more than nuclear family contributions must be taken into
account. In the research process for this project numerous young married
couples were interviewed (on an informal basis). A common practice that
was noted by the researcher, and acknowledged by many in society, was
the practice of remitting a portion of the family income to parents
(usually the husband's parents). The young marrieds consequently had
reduced 'real' incomes that did affect residential choice. On the other
hand, the relatives benefitting from the income transfers could afford
housing that would appear to be beyond their means. This practice was
not limited to remittances to parents. Under Confucian ethics the oldest
son of a family bears special responsibility. It is not uncommon for
the oldest son of the family to help other family members to purchase
housing. In any event there does appear to be a link between family
status and socio-economic status.
In Abu-Lughod's scheme, if association does exist between SES and

LC then three criteria must be met for disassociation between the two
factors to occur. The three criteria are:

a. clear distinction between stages in the family cycle,
each stage being associated with a change of residence

b. that subareas within the city offer, at all economic
levels, the full range of housing suitable to families at particular
points in the natural cycle of growth and decline.

c. cultural values permitting and favoring mobility to
maximize housing efficiency, unencumbered by the 'unnatural' frictions
of sentiment, local attachments or restrictive regulations (Abu-Lughod,

1968:132-33).

148
The social structure of Taipei does not appear to meet these criteria.
First, there does not appear to be clear distinctions between stages in
the family cycle, and certainly no absolute association of stages with a
change in residence. Wang Su Chang (1981) in her investigation into
residential choice mentions a trend in this direction, often motivated
by the reputations of certain schools, but income often limits such
mobility. It is not uncommon to see families domicile children with
relatives in other parts of the city in order to take advantage of
schools perceived as ’good', rather than moving themselves. Such moves
are limited by income, the length of the journey to work, or in many
cases by the sentiment and local attachments mentioned in Abu-Lughod's
third condition. Secondly, not all subareas of the city offer a full
range of housing at all economic levels to families at various points in
the life cycle. For example, the upscale areas along Tun-hua North Road,
Jen-ai Road, or the Tien-mu area offer virtually no housing to the poor
or even lower middle-class family. Conversely, it does not appear as if
any 'suitable' housing for upper-class families exists in the older
districts of the city such as Yen-ping or Cheng-chung. In any event, the
social structure of Taipei is not such, at the present time, that would
necessarily result in the separation of the factors.

At the same time, it would appear that such separation might be
present in the future. This study revealed significant groupings of
elderly, students, young female workers and others. Wang Su Chang (1981)
has identified a growing desire to change residence associated with the
change in the family life cycle. Improved incomes, the development of

more efficient transportation networks, and continuing urban development

149
and redevelopment should extend the options of families in the
metropolitan area.

Thus, if we evaluate the study on the basis of Abu-Lughod's theory
then Taipei does meet the criteria established. While Taipei's social
structure does not show the same separation of factors as cities in the
developed Western nations, it does fit urban theory as modified by
Abu-Lughod, Berry and others as it applies to other regions of the
world. While the hypothesis must be rejected, as it stated that in
Taipei the same type of separation would exist as is present in cities
located in developed countries of the West, urban theory is still
applicable to Taipei. Taipei currently does not show the same separation
of factors, but indications do exist that such separation might be
present in the future. It is not the separation of factors in itself
that is important. A similarity of pattern would, however, indicate that
similar processes are at work in creating the residential pattern. In
Taipei socio-economic class is the primary sorting factor that is
currently in operation determining residential patterns. However,
indications exist that stage in the family life cycle; as well as
traditional sorting mechanisms such as place of origin, play a role as
well.

Innigspipns En; Entnre Reseagph

As discussed above future research of this type is indicated. In
assessing the project at least three areas of investigation are
suggested. These include examining specific social groups, conducting a
time series analysis, and establishing the mechanisms involved in the

residential process.

150

The Factorial Ecology that was conducted for Taipei in 1980
pointed up the presence of many unique social groups within the city. An
example is the group identified in Factor Nine, the former soldiers. It
is an aging group that is disappearing due to mortality. It is also in
danger of losing its spatial concentration in the rebuilding process
occurring around the central railroad station. It is an interesting
group and one that is part of the living history of China. If
investigation is to occur is must be done soon. Another group that would
appear threatened is the one associated with the older neighborhoods in
the city, and described in Factor Four. This area is immediately
adjacent to the original CBD of the city. Its prime location makes it a
logical candidate for renewal in a city that exhibits a tremendous
dynamism in its growth process. The population of young women
identified in Factor Three also bears attention. For example, much could
be learned about the role of women in the development process, or
perhaps about the changing role and status of women in a society in
which traditional roles are being challenged by changing economic power.
In any event Chapter Four tried to indicate the rich diversity of
opportunity for future study that exists.

The second type of research that is indicated is a time series
analysis. This research indicates that change in social groupings is
occurring. Continued research in 1990, of the type envisioned by this
work in its original formulation, has the potential for isolating those
areas and groups most subject to change as the city develops. By
concentrating on these groups, as well as the larger pattern, our
knowledge of cities and how they develop, can be expanded. Many of the

problems that faced this study can be avoided. For example, the district

151
and nsnn/ii boundaries are now known. Even if they are changed for the
1990 census, the baseline is established. The potential for significant
research is there, and only awaits the new census.

Finally, more research needs to be done on the residential
selection process in Taipei. This study made the assumption that the
model developed and used as the standard in the West, was operative in
Taiwan. The field research conducted for this study brought home the
realization that such assumptions for the non-Western world are
inappropriate. It may be that the process is similar in Taiwan, but to
make the assumption without both regional knowledge and hard evidence
reflects a kind of academic imperialism. In addition, if the patterns
that are found are to be accurately explained such knowledge is of

paramount importance.

APPENDIX A

153

APPENDIX A

FIGURE VII

TAIPEI BASIN, TAIWAN: LAND USE, 1983

APPENDIX B

155

APPENDIX B

FIGURE XVII

VARIABLES USED IN THE ANALYSIS OF CHICAGO BY BERRY/KASARDA

e a C

Population-1960

Population Under 18 Years Old

Population Over 65 Years Old

Population Density

Percent of Total Population

Hsgiai and Ethnic Characteristics

Percent of Population-Non-White

Percent of Population With at Least One Foreign Born Parent

Concentration (LQ) Polish

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)

German
Italian
Russian
Irish

Czech
Swedish
British
Canadian
Foreign Born

Black

156
Concentration (LQ) Native White
Concentration (LQ) Other Non-Whites
Heligious Characteristics
Percent of Population-Protestant
Percent of Population-Catholic
Percent of Population-Jewish
Income aracteristics
Percentage of Labor Force-White Collar
Percentage of Labor Force-Professional/Managerial
Percentage of Labor Force-Clerical/Sales
Percentage of Labor Force-Craftsmen/Operatives
Percentage of Labor Force-Laborers
Percentage of Labor Force-Service Occupations
Ed cat h acte st
Median School Years Completed
Percentage of Population 25 and Older With 4 years High
School or More
Mobility
Percentage of Families Who Moved In Past Five Years
Liaising
Home Value
Percentage of Housing Units In 1 Unit Structures
Percentage of Housing Units In 2 Unit Structures
Percentage of Housing Units In 3 Unit or More Structures
Population per Household
Percent Owner Occupied Housing

Percentage of Housing Built Before 1940

Percentage

Percentage

Median Rent

157
of Housing Built 1940-1949

of Housing Built After 1950

Percentage of Housing Units-Sound (Meet Codes)

Percentage of Housing Units-Overcrowded

Employnent Characteristics

Percentage of Labor Force Unemployed

Total Unemployed

Percentage
Percentage
Percentage
Percentage

Percentage

of Women 14 and Older in the Labor Force
of Workers Commuting By Car

of Workers Commuting By Bus

of Workers Commuting By Train

of Workers Commuting By Foot

Source: Berry and Kasarda, 1976

APPENDIX C

159

APPENDIX C

FIGURE XXVIII
VARIABLES USED IN HSU AND PANNELL'S STUDY OF TAIPEI

e r c ara teristics

Total Population

Percent of Population Younger Than 15 Years Old
Percent of Population Over 65 Years Old
Percent of Population 15-65 Years Old
Percent Population Age Over 15 and Single
Percent Population Age Over 15 and Single
Native Hlace Characterisgics

Percent Native Population

Percent Taiwanese Population

Percent Mainland Population

Educational Characperistics

Percent Population With High Education
Percent Population With Secondary Education
Percent Population With Primary Education
Percent Population Self Educated

Percent Population Illiterate
Employnent Characteristics
Percent Population-Employers

Percent Population-Own Account Workers

Percent

Percent

Percent

160
Population-Wage and Salary Workers (Government)
Population-Wage and Salary Workers (Private)

Population-Unpaid Family Workers

industrial Characteristics

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Oc t

Percent

Percent

Percent

Percent

Percent

Percent

Source:

of Population in Agriculture
of Population in Mining
of Population in Manufacturing
of Population in Utility Work
of Population in Construction
of Population in Commerce
of Population in Transportation/Storage/Communications
of Population in Banking/Insurance/Real Estate/Service
of Population in Social and Personal Service
aracte 3 cs
of Population-Professional/Technical and Related Workers
of Population-Administrative Workers
of Population-Clerical Workers
of Population-Service Workers
of Population-Farmers, Fishermen, Related Workers
of Population-Operators of Production and Communication
Equipment and Physical Laborers

Hsu and Pannell, 1976

APPENDIX D

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

162

APPENDIX D

FIGURE XXIX

VARIABLES GATHERED IN THE TAIWAN AREA CENSUS

Families

Population

Population-Male

Population-Female

Housing Units-One Person

Population-One Person Household
Population-Male-One Person Household
Population-Female-One Person Household
Households-More Than One Person
POpulation-More Than One Person
Population-Male-More Than One Person
Population-Female-More Than One Person
Households-House/Business Combined
Population-House/Business Combined
Population-Male-House/Business Combined
Population-Female-House/Business Combined
Population-Male-O to 4 Years Old
Population-Female-O to 4 Years Old

Population-Male-S to 9 Years Old

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

163
Population-Female-S to 9 Years Old
Population-Male-IO to 14 Years Old
Population-Female-IO to 14 Years Old
Population-Male-lS to 19 Years Old
Population-Female-IS to 19 Years Old
Population—Male-20 to 24 Years Old
Population-Female-20 to 24 Years Old
Population-Male-25 to 29 Years Old
Population-Female 25 to 29 Years Old
Population-Male 30 to 34 Years Old
Population-Female 30 to 34 Years Old
Population-Male-35 to 39 Years Old
Population-Female-35 to 39 Years Old
Population-Male-40 to 44 Years Old
Population-Female-40 to 44 Years Old
Population-Male-45 to 49 Years Old
Population-Female 45 to 49 Years Old
Population-Male-SO to 54 Years Old
Population-Female-SO to 54 Years Old
Population-Male-SS to 59 Years Old
Population-Female-SS to 59 Years Old
Population-Male-60 to 64 Years Old
Population-Female-60 to 64 Years Old
Population-Male- 65 or Over
Population-Female-65 or Over
Population-Single-Male

Population-Single-Female

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

164
Population-Married-Male-Live With Spouse
Population-Married-Female-Live With Spouse
Population-Male-Separated or Divorced
Population-Female-Separated or Divorced
Population-Male-Widower
Population-Female-Widow
Population-Male-Literate
Population-Female-Literate
Population-Male-Graduate of Graduate School
Population-Female-Graduate of Graduate School
Population-Male-Attending Graduate School
Population-Female-Attending Graduate School
Population-Male-Have BA
Population-Female-Have BA
Population-Male-Attending University
Population-Female-Attending University
Population-Male-Graduate of 2/3 Year College
Population-Female-Graduate of 2/3 Year College
Population-Male-Attending 2/3 Year College
Population-Female-Attending 2/3 Year College
Population-Male-Graduate of 5 Year Junior Coll.
Population-Female-Grad. of 5 Year Junior College
Population-Male-Attending 5 Yr. JC-In Last 2 Yrs.
Population-Female-Attending 5 Yr. JC-Last 2 Years
Population-Male-Attending 5 Yr. JC-First 3 Years
Population-Female~Attending 5 Yr. JC-First 3 Years

Population-Male-High School Graduate

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

165
Population-Female-High School Graduate
Population-Male-Attending High School
Population-Female-Attending High School
Population-Male-Graduate-Senior Vocational School
Population-Female-Graduate-Senior Vocational School
Population-Male-Attending Senior Vocational School
Population-Female-Attending Senior Vocational School
Population-Male-Graduate Junior Vocational School
Population-Female-Graduate Junior Vocational School
Population-Male-Attending Junior Vocational School
Population—Female-Attending Junior Vocational School
Population-Male-Graduate of Junior High
Population-Female-Graduate of Junior High
Population-Male-Attending Junior High
Population-Female-Attending Junior High
Population-Male-Graduate of Primary School
Population-Female-Graduate of Primary School
Population-Male-Attending Primary School
Population-Female-Attending Primary School
Population-Male-Self Educated
Population-Female-Self Educated
Population-Male-Illiterate
Population-Female-Illiterate
Population-Male-Economically Active
Population-Female-Economically Active
Population-Male-Currently Employed

Population-Female-Currently Employed

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

166
Population-Male-Looking for Work
Population-Female~Looking for Work
Population-Male-Looking for 1st Job
Population-Female-Looking for 1st Job
Population-Male-Not Economically Active
Population-Female-Not Economically Active
Population-Male-Housekeeper
Population-Female-Housekeeper
Population-Male-Student
Population-Female~Student Total Population-Male-Old/Handicapped
Population-Female-Old/Handicapped
Population-Male-Not Described by Econ. Categories
Population-Female-Not Described by Econ.Categories
Population-Male-Job in Agriculture/Forestry
Population-Female-Agriculture/Forestry
Population-Male-Fishing
Population-Female-Fishing
Population-Male-Mining
Population-Female-Mining
Population-Male-Job in Manufacturing
Population-Female-Job in Manufacturing
Population-Male-Job in Utilities
Population-Female-Job in Utilities
Population-Male-Building or Architecture
Population-Female-Building or Architecture
Population-Male-Commercial-Self Employed

Population-Female-Commercial-Self Employed

Total
Total
Total
Total

Total

Total

Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total

Total

167

Population-Male-Job in Transportation
Population-Female—Job in Transportation
Population-Male-Banking,Accounting-Finance
Population-Female-Banking,Accounting,Finance
Population-Male-Public Administration,

National Defense
Population-Female-Public Administration

or National Defense

Population-Male-Other Service Job
Population-Female-Other Service Job
Population-Male-Special Technician
Population-Female-Special Technician
Population-Male-Private Co. Administration
Population-Female-Private Co. Administration
Population-Male~Guide/Guard
Population-Female-Guide/Guard
Population-Male~Commercial Sales,Buyer,Labor
Population-Female-Commercial Sales, Buyer
Population-Male-Service Labor
Population-Female-Service Labor
Population-Male-Primary Sector Labor
Population-Female-Primary Sector Labor
Population-Male-Production Labor
Population-Female-Production Labor
Population-Male-Military
Population-Female-Military

Population-Male-Employer

168
Total Population—Female-Employer
Total Population-Male-Self Employed-No Help
Total Population-Female-Self Employed-No Help
Total Population-Male-Government Construction
Total Population-Female-Government Construction
Total Population-Male-Private Sector Construction
Total Population-Female-Private Construction
Total Population-Male-Work for Relative-No Wage

Total Population-Female-Work for Relative-No Wage

Total Population-Born In- Tapei County
" " " Ilan
" " " Taoyuan
" " ” Hsinchu

" " Miaoli
" " " Taichung
" " " Changhua
" " " Nantou
" " " Yunlin
" " " Chiayi
" " " Tainan
" " " Kaohsiung
" " " Pingtung
" " " Taitung
" " " Hualien
" " ” Pinhu
" " " Keelung

" " " Taichung City

169

Tainan City
Taipei City
Kaohsiung City
Kinmen
Lianchang

Pong Sya Island
Nan Sya Island
Other Provincial City
Foreign Country
Taipei

Ilan

Taoyuan
Hsinchu

Mialoi

Taichung
Chunghua

Nantou

Yunlin

Chiayi

Tainan
Kaoshiung
Pingtung
Taitung
Hualien

Pinhu

Keelung City

Taichung City

II II II
II II II
II II II
II II II
II II II
II II II
II II

II II II
II II II

Total Population
Total Population
Total Population
Total Population

Total Population

170

Living in
Living in
Living in
Living in

Living 5 Years Ago

Tainan City
Taipei City
Kaoshiung City
Kinmen
Lianchang

Dong Sya Island
Nan Sya Island
Other Prov./City

Foreign Country

Same Place 5 Years Ago
Same District Different LI
Same City-Different District

Different County/City

Taipei
Ilan
Taoyuan
Hsinchu
Miaoli
Taichung
Chunghua
Nantou
Yunlin
Chiayi
Tainan
Kaoshiung
Pingtung

Taitung

171

" " " Hualien

" " " Pinhu

" " " Keelung City

" " " Taichung City
" " " Tainan City

" " " Taipei City

" " " Kaoshiung City
" " " Kinmen

" " " Lianchang

" " " Dong Sya Island
" " " Nan Sya Island

TOtal Population First Married at Less Than 15

Total Population First Married at 15

.. .. .. 16

" n n 17

" n n 18

.. .. .. 19

" .. .. 20

“ .. .. 21

" .. .. 22

" .. .. 23

" .. .. 24

" .. .. 25

" .. .. 26

" u a 27-29
" .. .. Over 30

Total Number of Live Male Births

 

172
Total Number Female Live Births
Total Number Surviving Male Infants
Total Number Surviving Female Infants
Number of Male Live Births in Last 12 Months
Number of Female Live Births in Last 12 Months
Surviving Males Born in Past 12 Months

Surviving Females Born in Past 12 Months

Variables 269 to 320 provide information in all educational categories,

and in marital status for those less than 15 years old.

Total Housing Units

Total Occupied Housing Units

Total Vacant Housing Units

Total Non-Residential Housing Units

Total Residential Units

Total Residence/Manufacturing Combined Units
Total Residence/Commercial Combined Units
Total Residence/Other Use Combined Units
Total Units Built Before 1945

Total Units Built 1946-1960

Total Units Built 1961-1975

Total Units Built 1976-1977

Total Units Built 1978-1979

Total Units Built 1980-

Total Units-Traditional Farm Style

Total Units-Villas

Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total

Total

173
Units-Japanese Style Units
Units-Apartments-4 Story and Less
Units-Apartments- More Than 4 Story
Units-Other Types
Units-Under 5 ping
Units-5-9 ping
Units-10-14 ping
Units-15-19 ping
Units-20-24 ping
Units-25-29 ping
Units-30-34 ping
Units-35-39 ping
Units-40-44 ping
Units-45-49 ping
Units-50-54 ping
Units-55-59 ping
Units-60-99 ping
Units-Over 100 ping
Units With Kitchen
Units Share Kitchen
Units No Kitchen
Units With Bath
Units Share Bath
Units No Bath
Units With Modern Toilet
Units Share Modern Toilet

Units Traditional Toilet

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Units

Units

Units

Units

Units

Units

Units

Units

Units

Units

Units

Units

Units

174
Share Traditional Toilet
No Toilet
Own Piped Water
Share Piped Water
No Piped Water
Private House
Public House
Other Type
Government Owned Rental
Privately Owned Rental
Government Provided
Private Company Provided

Other Type Ownership

APPENDIX E

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Concentration

Percent Total

Percent

Percent

Percent

Percent

Percent

of

of

of

of

of

of

of

of

of

of

of

of

of

176

APPENDIX E

FIGURE XXX

VARIABLES USED IN SOCIAL AREA ANALYSIS OF TAIPEI

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Housing-Single Person Family
Housing-More Than One Person Household
Housing-House/Business Combined
Population-Male/Female- 0 to 14 Years
Population-Male 15-64
Population-Female 15-64
Population-Singe Male
Population-Single Female
Population-Male-Widower
Population-Male Over 65
Population-Female Over 65
Metropolitan Population

Population-Male/Female 2-3 Years+ College

(LQ) of Males/Females 2-3 Years or More Coll.

Population-Male/Female High School Education

Total

Total

Total

Total

Total

Population-Male/Female Junior High School Ed.
Population-Male/Female Vocational Education
Population-Male/Female Primary Education
Population-Male/Female Self Educated

Population-Male/Female Illiterate

177

Percent Total Population-Economically Active

Percent Economically Active Population-Now Working

Percent Economically Active Population-Looking for Work

Percent Total

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Concentration

Population-Female-Economically Active

(LQ)
(LQ)
(IJQ)
(LQ)
(LQ)
(LQ)
(LQ)
(1Q)
(LQ)
(112)
(112)
(LQ)
(112)
(LQ)
(LQ)
(LQ)
(U2)
(LQ)
(LQ

(112)
(LQ)
(LQ)
(LQ)

of High School Graduates

of Male/Female Vocational Education

of Male/Female Junior High School Ed.

of Male/Female Primary School Ed.

of Male/Female Illiterates

Female Economically Active Population

Male/Female Previously Working Now Looking

Male/Female Employers

Male/Female Self Employed-No Help

Male/Female Private Sector Construction

M/F Government Construction

M/F Work For Relative For Free

M/F Military

Born

Born

Born

Born

Born

Born

Born

Born

Born

Born

in Taipei
in Ilan

in Taoyuan
in Hsinchu
in Miaoli
in Taichung
in Changhua
in Nantou
in Yunlin

in Chiayi

Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Concentration
Percent Econ.
Percent Econ.
Percent Econ.
Percent Econ.
Percent Econ.
Percent Econ.
Percent Econ.
Percent Econ.
Percent Econ.
Percent Econ.
Econ.

Percent

Percent Econ.

(LQ)
(112)
(LQ)
(LQ)
(U2)
(IQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)
(LQ)

Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active

Active

Born

Born

Born

Born

Born

Born

Born

Born

Born

Born

Born

Born

Born

Born

178

in Tainan

in Kaoshiung

in Pingtung

in Taitung

in Hualien

in Pinhu

in Keelung
in Taichung City
in Tainan City
in Taipei City
in Kaoshiung City
in Kinmen
in Lianchang

in Foreign Country

Male/Female in Agriculture/Forestry

Population M/F Blue Collar
Population M/F White Collar
Population M/F High Tech/Admin
Population-Labor-Sales/Service/Guard
Population-Labor-Agriculture
Population-Military
Population-Employers

POpulation-Self Employed No Help
Population-Production Labor
Population-Government Construction
Population-Private Sector Construction

Population-Work for Relative for Free

179

Percent Econ. Active Population-Fishing

Percent Econ. Active Population-Mining

Total Housing Units

Total Households

Concentration

Concentration

Concentration

Concentration

Concentration

Percent of Total

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

(LQ) of Population Not Moved in Past 5 Years
of Population-Intra-urban Move in Past 5 Years
of Population-Inter-urban Move in Past 5 Years
(LQ) of Live Births in Past 12 Months

(LQ) of Surviving Infants Past 12 Months

Population Moved Within Past 5 Years

Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing

Housing

Units-Occupied

Units-Vacant

Units-Non-residential
Units-Residential Units
Units-Residence/Manufacturing Comb.
Units-Residence/Commercial Combined
Units-Other Uses

Units-Built Before 1945

Units-Built 1946-1960

Units-Built 1961-1975
Units-Built 1976-1977
Units-Built 1978-1979
Units-Built 1980 on
Units-Traditional Farm Style
Units-Villa

Units-Japanese Style

Units-Apartments-4 Story/Less

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

of

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Total

Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing
Housing

Housing

180
Units-Apartments-More Than 4 Story
Units-Other Type
Units-Under 5 ping
Units-5-9 ping
Units-lO-l9 ping
Units-20-29 ping
Units-30-39 ping
Units-40-49 ping
Units-50-59 ping
Units-60-99 ping
Units-100 or more ping
Units-With Kitchen
Units-Share Kitchen
Units-No Kitchen
Units-With Bath
Units-Share Bath
Units-No Bath
Units-Modern Toilet
Units-Share Modern Toilet
Units-Traditional Toilet
Units-Share Traditional Toilet
Units-No Toilet
Units-With Piped Water
Units-Share Piped Water
Units-No Piped Water
Units-Private House

Units-Public Housing

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

Percent

of

of

of

of

of

of

of

of

of

Median Age

181
Total Housing Units-Other Type Unit
Total Housing Units-Rental-Government Owned
Total Housing Units-Rental-Private Supply
Total Housing Units-Government Supplied
Total Housing Units-Private Company Supplied
Total Housing Units-Other Type Ownership
Total Population-Male-15-29
Total Population-Female-15-29
Total Population-Female-65 and over/Handicapped

of Population

BIBLIOGRAPHY

183

BIBLIOGRAPHY

Abler, Ronald. "What Shall We Say? To Whom Shall We Speak", Annals of

§he Associstion of American geoggaphers, Vol. 77, No. 44, 1987 pp.
511-523.

Abu-Lughod, Janet; "The Factorial Ecology of Cairo," American Journal of
Sopiology, Vol. 74, No. 5 (March, 1969) pp. 198-212.

Abu-Lughod,Janet; Cnirp; 1001 Teaps pi she Sity Vicporious, Princeton:

Princeton University Press, 1971.

Abu-Lughod, Janet and Richard Hay eds.; Third Wprld Urbanization,
Chicago: Maaroufa Press, 1977.

Aderibigbe, A.; Lagos: The Development of an African City, London:
Congman, 1975.

Ahern, Emily and Hill Gates. The Anthropology of Taiwanese Society,
Stanford, CA: Stanford University Press, 1981.

Alihan, Milla; Social Ecology, New York: Columbia University Press,
1938.

Amedeo, D. and Golledge, R.; An Introduction to Scientific Reasoning in
Geogsaphy, New York: Wiley, 1975.

Anderson, T. and J. Egeland, "Spatial Aspects of Social Area Analysis",
Aperican Sociological Heviey, Vol. 26, (June 1961) pp. 392-98.

Anderssen, Harri. ”Urban Structural Dynamics In the City of Turku,
Finland" Fennis, Vol. 161, No.2., 1983. pp. 145-261.

Baker, Hugh, "Extended Kinship in the Traditional City," C. W. Skinner,

ed., The City in Lste lnpsrisl China, Stanford: Stanford University
Press, 1977. pp.

Ballantine, Joseph, Eggnpsn; A Egoblsn Ho; United States Foreign Policy,
Washington D.C.: The Brookings Institution, 1952.

Barclay, George W., Solpnisl Develppnent and Population in Taiwan,

Princeton: Princeton University Press, 1954.

 

Bell, Wendell, "The Social Areas of the San Francisco Bay Region",
Anerican Sociolpgigal Heview, Vol. 28, (February, 1953). pp. 29-47.

Bell, Wendell, "Economic, Family and Ethnic Status: An Empirical Test",
Anegican Sociological Hevisw, Vol.22, 1955. pp. 45-52.

184

Berry, Brian, "City Size Distribution and Economic Development",

Espnpmic Hevelopmen; and Cultnzsl Qhange, Vol.9,No.4, 1961. pp.
573-88.

Berry, Brian, ”Cities as Systems Within Systems of Cities," Papers and

Hgogeedings of she Regional Science Association, Vol. 13., 1964.
pp.147-63.

Berry, Brian, Esssys pn Qonnpdipy Elpws ans the Spatial Stgucture of the
lndisn Economy, Chicago: University of Chicago, Department of

Geography Research Paper, 1966.

Berry, Brian and Philip Rees, "The Factorial Ecology of Calcutta",

Anezican Journal of Sociology, Vol. 74, No. 5, (March, 1969).
pp.445-91.

Berry, Brian and F. Horton, Geo a i a ers ectives on Urban S stems,
Englewood Cliffs, NJ: Prentice Hall, 1970.

Berry, Brian ed., t C a t Ha dboo ' ethod and
Applications, New York: Wiley, 1972.

Berry, Brian, Hnnan Consegusnpss pf Urhsnization, London: Methuen,
1973.

Berry, Brian ed., szsnizapipn nnd QpnntsI-thanigation; Urban Affairs
Annusl, Vol. 11, Los Angeles: Sage Publications, 1976.

Berry, Brian and D. Kasarda, Qpnpsnppgs;y_yyhnn_§pplpgy, New York:

Macmillan, 1977.

Bourne, Larry. The lngernal Stgnctuge oi the City, New York: Oxford
University Press, 1971.

Bourne, Larry and J. Simmons eds., Systems of gities; Readings in

ucture G ow and 01 c , New York: Oxford University Press,
1978.

Bonine, Michael, "The Urbanization of the Persian Gulf Nations," A.
Cottrell ed., e e s a u State3° A General Surve , Baltimore:
John Hopkins University Press, 1980. pp. 225-278.

Boyd, Andrew, ese c tect o a in 500B C -l9ll
Alps, Chicago: University of Chicago Press, 1962.
Breese, Gerald, thanigstipn in Hewly stslpping gguntries, Englewood

Cliffs, NJ: Prentice Hall, 1966.

Breese, Gerald ed., e Cit ew eve 0 Countries, Englewood
Cliffs, NJ: Prentice Hall, 1969.

Burgess, E.W., "The Growth of the City," R. Parr, E. Burgess and R.
McKenzie eds., The Sipy, Chicago: University of Chicago Press,
1925. pp. 47-62.

185

Cameron, G. and L. Wingo eds., Cities, Regions, and Public Policy,
Edinburgh: Oliver and Boyd, 1973.

Carpenter, E.H. "The Potential for Population Dispersal: A Closer Look

at Residential Locational Preference", Rural Sociology, Vol. 42,
1977.

Chang, Han-yu and R. Myers. "Japanese Colonial Development Policy in
Taiwan, 1895-1906: A Case of Bureaucratic Entrepreneurship",

qupnal oi Asisn Studies, Vol. 22, 1963. pp. 433-449.

Chang. Kwang- -chih WW 3d ed New Haven:
Yale University Press, 1977.

Chang, Lung-sheng. "Regional Disparity and Regional Development Policies

in Taiwan", Egoceedings of php Qpnierence pn Urban Growth and

Econpmic Development in the Pacific Region, Taipei: Institute of
Economics, Academia Sinica, January, 1984. PP. 545-572.

Chang, Ming-cheng. "Economic Adjustment of Migrants in Taiwan: Recent

Findings and Policy Implications", Erpceedings pi the Conferenpe on

ba owt a d co v e cific Re ion,
Taipei: Institute of Economics, Academia Sinica, January, 1984. pp.
619-652.

Chang, Shih-tien, Taipei Sity; Activity Stpuctuge and Related
Eapilities, Taipei: Research Report for the Research Institute of

Urban Planning, Tamkang University, 1979. (In Chinese)

Chang, Sen-dou, "Land Use and Intra-Urban Travel in Taipei", Erpcpedings
pi the Association of American Geographers, No.II, 1970. pp. 40-45.

Chang, Sen-dou, "Some Aspects of the Urban Geography of the Chinese

Hsien Capital", Annals pf thp Assopiation of Amprican Geographers,
Vol. XLI, March 1961. pp.23-45.

Chang, Sen-dou, "The Historical Trends of Chinese Urbanization," Annals

pf ghe Asspciapion of Anerican Geographers, Vol. 53, (June 1963).
pp.109-43.

 

Chang, Sen-dou, "Some Observations on the Morphology of Chinese Walled

Cities", Annals of the Association pf Anerican geographers, Vol.
60, (March 1970). pp. 63-91.

Chang, T. S. and Chen Sun. "Land Use Planning in Taiwan, Republic of
China". Report: WWW.
Taipei: National Science Council, 1978.

Chang, Ying-Hwa, "Harvesting High-Rises From Rice Paddies", Free China
Heview, Vol. 38, No. 3, March,1988. pp. 28-31.

Chen, Cheng-siang, Thp_§i§y_pfi_Tnippi, Taipei: Fu-Min Geographica
Institute of Economic Development, Research Report No. 71, 1956.

Chen, Cheng-hsiang, Tsiwan; An Eppnonip anp Social Geography, Taipei:

186
Fu-Min Geographical Institute, Research Report No. 96, 1963.
Chen, Cheng. Lsnd Reiprm in Tsiwan, Taipei: China Publishing Co., 1961.

Chen, Kuan-jeng, "Suburbinization in the Taipei Metropolitan Area,
1962-1979", ou a P at o tudies, , No.5, 1981.

Cheng, Yueh-sui. "Taiwan's Comprehensive Development and Regional

Economic Development Planning”, Hsnh of Taiwan Quarterly, No. 34,
1983.

Chiang, Lan-hung, S t al nd hav ora ect f Female S atia
Hohility in Tsiwan, Taipei: Occassional Paper of The Population
Studies Center, National Taiwan University, 1983.

Chiang, Nora Huang and Michael Hsiao Hsin-Huang, "Taipei: History and
Problems of Development" in Victor Sit ed., Chinese Citips; Thp

Growth of the Metropolis Since 1949, Hong Kong: Oxford University
Press, 1985. pp.188-205.

Chiang, Tao-Chang, ”Walled Towns of Taiwan: A Study in Historical

Geography" Qppgzsphipsl_Spppips, Taipei: Department of Geography,
Taiwan Normal University, No.1, 1966. (In Chinese) pp. 53-80.

Chiang, Ya-mei, A Stndy pn php Higznpipn nnd Decision Making of
Hopulation in Sanchung, A snpurb City pf Taipei, Taipei: National

Taiwan Normal University Occasional Paper, 1983.

Ch'ien, Hsueh-tao, "The Establishment of the Commercial Areas in Taipei,
and Their Distributional Structure", Sanh of Taiwan Quarterly, Vol.
27, No. 3, 1976. pp. 79-115.

Chu, P'u-ying. "Housing Construction in Cities of Taiwan", Bank of
Tniwsn Quagtegly, Vol. 24, No.3, 1973. (In Chinese)

Chuang, Ch'ang-sheng. "Financial Analysis of Counties, Cities and

Townships in Taiwan", Bsnh of Tpiwsn Qnagtegly, Vol. 25, No.3,
1974. (In Chinese) pp. 33-65.

Clough, Ralph, lslsnp_ghins, Cambride: Harvard University Press, 1978.

Cohen, M.L., 0 se United Hou e iv ded’ e C in se am 1 in Taiwan,
New York: Columbia University Press, 1976.

Cressey, Paul, "The Ecological Organization of Rangoon, Burma",

WW. Vol. XI. January. 1956.

Cullingford, K. and S. Openshaw, ”Identifying Areas of Rural Deprivation
Using Social Area Analysis", Hegipnal Spupies, Vol. 16, No. 6 ,
1984.

Davidson, James, Thp Island of Formosa; Historical Visy Fronfl430-l900,
New York: MacMillan, 1903.

 

Davie, Maurice, "The Patterns of Urban Growth," G. Murdock ed., Studies

187

in thp Science pi Sopiepy, New York: MacMillan, 1937.

Davis, Kingsley. HpIlp_thsnignpipn;_lgpg;lglg, Berkeley, CA: University
of California Press, 1969.

Davis, Kingsley. "The Urbanization of the Human Population”, Scientific
Anpgipnn, Vol. 213, No. 3, September, 1965 pp.4l-S3.

De Beauclair, Inez. Stpdies pn Bptel Tobagp snd Ysp, Taipei: Orient
Cultural Service, 1971.

De Jong, G.F., "Residential Preferences and Migration", Hemoggaphy, Vol.
14, 1977.

Deglopper, R., "Social Structure in a 19th Century Taiwanese Port City,"

C. Skinner, ed., The Cipy in Lntp lnpenial Shina, Stanford:
Stanford University Press, 1977. pp. 633-650.

Diamond, Nora. "Women and Industry in Taiwan", Mopern Chins, Vol. 5,
1979.

Duncan, 0.D., W. Scott, S. Lieberson, B. Duncan, and H. Winsborough,

Hetpppolis and Region, Baltimore: John Hopkins University Press,
1960.

Duncan, Otis, "Review of Social Area Analysis by Shevky and Bell,"
Anegipan qugnal pi Spciology, Vol. 61, No. 2, (July, 1955).

Dwyer, D.J. ed., The Qipy ss 5 Cpnpe; of Shange in Asis, Hong Kong: Hong
Kong University Press, 1972.

Dwyer, D.J. ed., Thp_Qipy_in_php_Thi;p_flp;lp, London: Macmillan Press,
1974.

Eberhard, Wolfram, "Data On The Structure Of the Chinese City In The
Pre-Industrial Period", 0 o ev e Cu tu 1 han e,
No. 4, 1964.

Ellen, Roy. "Persistence and Change in The Relationship Between

Anthropology and Human Geography", Pppgrpss in Human Geography,
Vol. 12, No. 2, 1988.

Elvin, Mark and G.W. Skinner, eds., Thp Shinese Sipy Setween Two Worlds,

Stanford: Stanford University Press, 1974.

Eto, Shinkichi. "An Outline of Formosan History”, in M. Mancall, ed.,
Epznpsn_Tppsy, New York: Praeger, 1964.

Fann, Tseng-huei, "Social Adjustment of Rural Migrants in Taipei City",

Spn;npl_p£_§ppiplpgy, Taipei: National Taiwan University, No. 10,
1974. (In Chinese).

Fava, Sylvia ed., Hphsnisn_in_Hlep_£p;sppppiyp, New York: Crowell and
Co., 1968.

188

Fei, Hsiao-tung, "Village, Town and City," Shina's Gentry, New York:
Columbia University Press, 1953.

Feuchtwang, Stephan, "City Temples in Taipei Under Three Regimes", Elvin

and Skinner, eds., thnese Sity Hetween Twp Worlds, Stanford:
Stanford University Press, 1974. pp. 263-302.

Firey, Walter, Lsnp_Hsp_in_§pnp;nl_Hpsppn, Boston: Harvard University
Press, 1947.

Fox, K. A., Kumar, T.K., "The Functional Economic Area: Delimitations
and Implications for Economic Analysis and Policy", Eapers and

Exopepdings pf the Regipnsl Spiencp Assopiatipn, Vol. 15, 1965. pp.
57-85.

Franklin, George. ”Physical Planning and the Third World", Third World
Elanning Heview, Vol. 1, No.1, 1979.

Freedman, Deborah. "Consumption of Modern Goods and Services and Their

Relation to Fertility", Spngnal of Dpvelppment Studies, Vol. 12,
No. l, 1975.

Friedman, John and J. Miller, "The Urban Field,” Soupnal of the American
lnstitnte of Plsnnegs, Vol. 31 (November), 1965. pp. 312-320.

Friedman, John, "A General Theory of Polarized Development,",Hansen,
Growth Ce te s n e i a co m eve o ment, New York:

Free Press, 1972. pp. 82-107.

Friedman, John, "Urbanization and National Development: A Comparitive

Analysis”, Friedman, ed., U bani at annin and National
Development, Los Angeles: Sage Publications, 1973.

Freeman, Maurice. ghinpsp_Linpsgp_snp_Sppipny, London: Athlone, 1966.

Fuchs, Roland and John Street. "Land Constraints and Development
Planning in Taiwan", f evel eas, Vol.14, No.3,
1980.

Galenson, Walter, co 0 w an St c 1 an e in Taiwan,

Ithaca, NY: Cornell University Press, 1979.

Gallin, Bernard, Hsin Hsing, Taiwan; A Chinese Village in Change,
Berkeley: University of California Press, 1966.

Gallin, Bernard, "Mediation in Changing Chinese Society in Rural

Taiwan". WW. Vol. II. No. 1.2
(Jan. and April), 1967. pp. 77-90.

Gallin, Bernard, "Political Factionalism and Its Impact on Chinese
Village Social Organization in Taiwan", M. Swartz, ed.,

chal-vael Eolitips, Chicago: Aldine 1968. pp. 377-400.

Gallin, Bernard, "Rural-to-Urban Migration in Taiwan: Its Impact on

189

Chinese Family and Kinship," D. Buxbaum, ed., Chinese anily Law,
Seattle: University of Washington Press, 1978.

Gallin, Bernard, "The Integration of Village Migrants Into Urban

Society", Thomas and Hunter, eds., lntesnal Higrapion Systems in
php Devploping Wozld, New York: Schenkman, 1980.

Gallin, B. and R. Gallin, ”The Rural-Urban Migration of an
Anthropologist in Taiwan", Foster and Kemper, eds., Anthropologists
in Qipies, New York: Little Brown, 1974.

Gallin, B. and R. Gallin, "Socioeconomic Life in Rural Taiwan: Twenty
Years of Development and Change", Hodepn China, Vol, 8, No. 2,
(April), 1982. pp. 205-245.

Gallin, B. and R. Gallin. "The Integration of Village Migrants in

Taipei", in Elvin and Skinner, eds., The Shinese City Between Two
szlps, Stanford, CA: Stanford University Press, 1974.

Gallin, Rita, The Impact of Develppment on Women's Roles and Status; A

gsse Spudy from Taiwan, East Lansing: Working Papers on Women in
Development, No. 9, 1982.

Gallin, Rita, "Women, Family, and the Political Economy of Taiwan",
Jonrnal of Peasanp Spudips, Vol. 12, No. l,(October), 1984. pp.
76-92.

Gans, Herbert, The vaittownegs, New York: Pantheon, 1967.

Gates, Hill, "Dependency and the Part Time Proletariat In Taiwan",
prp;n_§hins, Vol. 5, No. 3, 1979. pp. 381-407.

Gates, Hill, "Ethnicity and Social Class", Gates and Ahern, eds., The

Anthropology of Tsiwanese Sopippy, Stanford: Stanford University
Press, 1981.
Gates, Hill. e e Wo C s L ve3° Ge n aiw n, Ithaca,

NY: Cornell University Press, 1988.

Gilbert, Anne. "The New Regional Geography in English and French

Speaking Countries". WW. Vol. 12. No. 2.
1988.
Gilmore, Harlan, Tgsnsppgtspion snp php Gpowph of Qities, Glencoe, IL:

Free Press, 1953.

Ginsburg, Norton, ed., The Hattprn of Asia, Englewood Cliffs, NJ:
Prentice Hall, 1958.

Ginsburg, Norton, "Urban Geography in non-Western Areas", Hauser and
Schnore, eds., The Spudy pf nganigation, New York: Wiley, 1965.
pp. 311-346.

Gist, N. and S. Fava, prsnpSppippy, New York: Crowell, 1964.

190

Goddard, W.G., Foznpss; A Spndy in Shinese History, London: Macmillan,
1966.

Goheen, Peter G., Victorian Toronto, 1850 to 1900: Pattern and Process
pf ngwph, Chicago: University of Chicago Geography Research Paper
No. 127, 1971.

Gordon, Leonard, Ta wa ° S udie cal stor , New York: Columbia
University Press, 1970.

Gordon, Peter and Jacques Ident. "A Demoeconomics Model of Urbanization:

The Case of Taiwan", Egoceepings of the Conference on Urban Growth

snp Ecpnomip Devplppnent in php Eapific Hegion, Taipei: Institute
of Economics, Academia Sinica, January,l984.

Gottman, Jean, Hpgslppplis, New York: Twentieth Century Fund, 1961.

Graff, Michael, C i rba u a Den Gradients n Tai ei,
East Lansing: Unpublished PhD Dissertation, Michigan State
University, 1976.

Greenhalgh, Susan. "Income Units: The Ethnographic Alternative to
Standardization", in Yoram Ben-Porath, ed., Inponp Dispgibution and
She Family, New York: Population Council, 1982.

Greenhalgh, Susan, "Networks and Their Nodes: Urban Society on Taiwan",

Shinn_9ns;pp;ly, September, 1984. pp. 529-552.

Griffin, E. and L. Ford, "A Model of Latin American City Structure",

Spogrsphicsl priew, Vol. 70, 1980. pp. 397-422.
Haggett, P., A. Cliff, A. Frey, Lppational Models, New York: Halstead,
1977.

Hance, William, 0 ation Mi atio and banizatio in Africa, New
York: Columbia University Press, 1970.

Hansen, N.M., w ' e s 0 cm cs of

Cipy Size and Sppnctuge, Lexington, MA: Heath, 1975.

Harris, C. and E. Ullman, "The Nature of Cities”, Annals of the American
WWW. V01. 242. 1945- PP- 7-17-

Hartshorn, Truman, lnterpgpping phe Cipy; An Urban Gepgraphy, New York:
Wiley, 1980.

Harvey, David, Sopial Justice nnd the Qity, London: Edward Arnold,
1973.

Harvey, David, prlanstipn in geography, London: Edward Arnold, 1976.

Hauser, P. and L. Schnore eds., Thp Sgnpy pf Hibsnigation, New York:
John Wiley and Sons, 1965.

Hawley, Amos and Otis Duncan, "Social Area Analysis: A Critical

191

Appraisal", Lnnp_fippnpnips, Vol. 33, No. 4, (November, 1957).
Herbert, David, Unban geogznphy, New York: Praeger, 1972.
Herbert, John. Urban Development in the Third World, New York: Praeger,
1979.
Ho, Samuel, Hppnpnip_Dpyplppnpnp_pfi_Tniysn;_lSpD;lSZD, New Haven: Yale

University Press, 1978.

Hoyt, Homer, The Stznptngp nnp ngwph pf Residential Neighborhoods in

Anppipnn_§ipips, Washington, DC: General Housing Administration,

1939.

Hsieh, Chiao-min, "The Physical Setting of Taiwin", Sih, Paul ed.,
Tpiynn_inanppzn_Tinps, New York: St. John's University Press,
1973.

Hsieh, Chiao-min, Taiwan: Ilha Fopnosa, A Geography in Perspective,
Washington D.C.: Butterworths, 1964.

Hsiung, James, ed., The Taiwan Expepience; 1950-1980, New York:
Praeger, 1981.

Hsu, Yi-Rong Ann and Clifton Pannell, "Urbanization and Residential

Spatial Structure in Taiwan", £npi§ip_yipyppinp, Vol. 12, No. 3.
pp.22-52.

Hughes, James, "Social Area Analysis", M. Greenberg, ed., Dghsn
W. New Brunswick. NJ: Rutgers

University Center for Urban Policy Research, 1974.

Hughes, James, ed., Suburbini a io lo and he Future of the Cit ,

New Brunswick, NJ: Rutgers University Center for Urban Policy
Research, 1974.

Institute of Economics, Academia Sinica, Qpnfepence pn ngan Ggpwth snd

prnonic Development in the 2pcifip Hegipn, Taipei: Academia
Sinica, 1984.

Jacobs, J. Bruce and J. Hagger and A. Sedgley. Taiwan; A Cpmprehensive
En 1 n a e i t , A Joint Publication

of The Borchardt Library, La Trobe University, Bundoora, Victoria
and the East Asian Institute, Columbia University, New York, 1984.

 

Kerr, George, 0 sa° censed evo utio a d e Home Rule Movement:
l82§-l24§, Honolulu: University of Hawaii Press, 1974.

Knapp, Ronald, ed., C ' ' s torical
Gpoggaphy pi Tsiwsn, Honolulu: University of Hawaii Press, 1980.

Kohl, Johann. Der Verkehr und die Asiedlung der Menschen in ihrer
Abhangigkeit von der Gestaltung der Erdoberflache, in Berry and

Kasarda, Spnppnppps;y_D;psn_Epplpgy, New York: Macmillan, 1977.

192

Koo, Anthony, "Economic Consequences of Land Reform in Taiwan”, Asian

Shrggy, March, 1966.

 

Koo, Anthony, a d o o i eve o ent: A Case Study of
Ipighh, New York: Praeger, 1968.

Koo, Anthony, "Economic Development of Taiwan: 1945—1970", Taiwan;

F 0 re- s o o M de T mes, New York: St. John's University
Press, 1973.

Krzynzaniak, Marian. ”Savings Behavior of Poor and Rich in Taiwan: 1964,
1966, 1968,and 1970", Jppppai of Developing Areas, No. 11, July,
1977.

Kublin, Hyman. "Taiwan's japanese Interlude: 1895-1945", in P. Sih, ed.,
Thighp_ih_hpg§;n_lihg§, New York: St John's University Press, 1973.

Kuo, Shirley- 132W. Boulder. CO: Westview
Press, 1983.

Kuo, K'un-mo and Liu Shui-shen. "A Study of Water Supply Improvements in
the Taipei Area", thk pf Igiwah thrtegiy, Vol.31, No.4, 1980.
pp.211-233.

Lamley, Harry. "The Formations of Cities:Initiative and Motivation in
Building Three Walled Cities in Taiwan", in G. William Skinner, ed.

The City in Late impegiai thhg, Stanford, CA: Stanford University
Press, 1977.

Lau, Lawrence, ed., ode s 0 ve o e ° om a ative Stud of

Econppip Ggowth ih South hope; ahd Igiwap, San Francisco: Institute
for Contemporary Studies, 1986.

Lee, Z.N. (Ju-nan). "Land Replotting as an Instrument of Land Assemblage
for Housing Development in Taiwan", Indpstgy oi Free China, Vol.
53, No.2, February, 1980. pp. 7-16

Lee, Z.N. (Ju-nan). "Land Problems of Housing Development in Taiwan",

ihgh§§;1_gf_£;gp_§hih§, Vol. 51, No.1, January, 1979. pp.16-23.
Leung, C.K. and N. Ginsburg, eds., a° tio an ational

Development, Chicago: University of Chicago, Department of
Geography Research Paper No. 196, 1980.

Lewis, John, ed., The Qigy ih Qppmunist Ching, Stanford: Stanford
University Press, 1971.

Li, Hsun-feng. "Quantitative Study of the Structure of Manufacturing in

Taipei City", Bhpk pf Igiwhh Qpaggerly, Vol.28, No.3, 1977. (In
Chinese) pp. 120-136.

Li, Hsun-feng. "A Survey and Analysis of Metal Product Industries in

Taipei", Qeographicgi Researph, Taipei: Institute of Geography,
National Taiwan Normal University, No. 7, January, 1981. (In

Chinese) pp. 129-152.

193

Li, Hsun-feng, "An Analysis of Industrial Land in the Taipei

Metropolitan Area", Bank pi Iaiwah Qparieily, Vol. 34, No. l, 1983.
pp. 230-259.

Li, Jui-lin, "The Structure and Development of Cities in Taiwan", Bank

pi Iaiwah Qpaiigiiy, Vol. 24, No.3, 1973. (In Chinese) pp. 1-29.

Li, K.T., "Population Distribution and Quality of Life in the Taiwan

Area", Industpy oi Eiee China, Part I: Vol. 60, No.3, September,
1983 pp.1-24; Part II: Vol. 60, No.4, October, 1983.pp. 17-31.

Li, K.T. and H.H. Tsai, ”Urban and Population Decentralization Policies:

The Experience of Taiwan, R.0.C.", ihdpstry oi Free China, Vol. 69,
No. 1, January, 1988. pp.1-14.

Liang, Kuo-chang. "The Spatial Structure of Manufacturing in Northern

Lin,

Lin,

Lin,

Lin,

Lin,

Liu,

Liu,

Liu,

Liu,

Liu,

Taiwan", Bank of Iaiwah ghaiteiiy, Vol.29, No. 2, 1978. (In
Chinese) pp. 140-177.

Ch'ing-lung. "An Analysis of Taipei's Income and Expenses", Bank of
Igiwgp Qpaiteiiy, Vol. 33, No. l, 1982. (In Chinese) pp. 70-114.

Tou-ch'in. ”Behavior of Industrial Plant Location in the Taipei
Metropolitan Area", a k 0 aiwan arterl , Vol. 34, No. 4, 1983.
(In Chinese) pp. 123-164.

Tse-tien. "A Study of Industrial Location in Cities in Taiwan",

Bhph oi Taiwah Qphiieiiy, Vol.24, No. 3, 1973. (In Chinese) pp.
11-67.

Tzu-yu. "Regional and Urban Planning in Taiwan", fighh oi Iaiwan
thiipily, Vol.24, No. 3, 1973. (In Chinese) pp. 30-40.

Ying-yen. "Urban Land Consolidation in Taiwan", hank oi Taiwan
ghagteiiy, Vol. 31, No. 1, 1980 (In Chinese) pp.98-136.

Hung-Hsi, "A Comparitive Study of Sanchung and Rei-Fong",

Qeographigal Resegrph, No. 4, Jan. 1979, Taipei: Institute of
Geography, National Taiwan Normal University. (In Chinese)

Hung-hsi. "An Analysis of Land Use in the Lower Valley of the Hsin

Tien River", Bhhh_pi_1hip§p_gp§;ig;iy, Vol. 23, No. 3, 1972. (In
Chinese) pp. 267-274.

Paul, "Economic Aspect of Rapid Urbanization in Taipei", Academica
Ecohpgic Ehpeig, Vol. 7, No. 1, March, 1979, Taipei: Academica
Sinica. pp.151-187.

Paul, "Toward a Closer Integration of Population in Development
Policies in Taiwan", 12g, Vol. 56, No. 2, August 1981. pp. 9-28.

Paul, "Labor Mobility and Utilization in Relation to Urbanization
in Taiwan", 1E9, Parts I-II, Vol. 57, Nos. 5,6, (May-June), 1982.
pp. 1-12 and pp. 7-14.

194

Liu, Paul, ”Factors and Policies Contributing to Urbanization and Labor
Mobility in Taiwan", iFC, Vol. 59, No. S, (May) 1983. pp. 1-20.

Liu, Paul. "Economic Aspects of Rapid Urbanization in Taipei", Academica
Rpsearph Papers, Taipei: Academia Sinica, Vol. 7, No. 1, March
1979. pp.151-l88.

Liu, Paul. "Urbanization and Employment Growth in the Taipei

Metropolitan Area", Ipdpfirry pi Eree Chipa, Vol 65, No. 5, May,
1986. pp.1-23.

Lubeck, Paul, "Contrasts and Continuity in a Dependent City: Kano,

Nigeria", Abu-Lughod and Hays', eds., Third World Development,
Chicago: Maaroufa, 1977. pp. 281-290.

Ma, Laurence and E. Hinten, Urban Development ip Modern China, Boulder,
CO.: Westview, 1981.

Ma, Laurence and C. Pannell, eds., a' e eo a of evelo ment

hpg_hpggrpig§iiph, New York: Halstead - Scripta Series in
Geography, 1982.

Mabogunje, Akin, The Deveipppeht Process; A Spatiai Perspective, New
York: Holmes and Meirer, 1981.

Mancall, Mark ed., Formosa Tpday, New York: Praeger, 1964

Marsh, Robert M.. "The Taiwanese of Taipei: Some Major Aspects of Their

Social Structure and Attitudes", ipprpg1_pi_5§ihh_§rpgig§, Vol. 27,
1968.

McFadden, D., "hpgeiing the thice pf Residential chation, Berkely, CA:
University of California Press, 1977.

McGee, T. G., The Urbghizgtiph Prpcess in the Third World, London: Bell
and Sons, 1971.

McGee, T.G., The Sputheasi Asiap City, New York: Praeger, 1967.

Meng, Jing, "The Spatial Variations of Population Distributions and
Growth in Taipei City", Geographical Research, Taipei:Institute of
Geography, National Taiwan Normal University, No. 8, January,
1982.(In Chinese) pp. 219-230.

Misra, R.P.,ed., Miliioh Cities pf Indip, New Delhi: Vikas Publishing
House, 1978.

Murphey, Rhoads, Shppggir__K_y_rp_Mpggrh_§hip_, Cambridge, MA: Harvard
University Press, 1953.

Murphey, Rhoads, "The City as a Center of Change: Western Europe and
China", ' e._ ‘ u: . : __; _ Vol.
XLIV,19S4.

 

195

Murphey, Rhoads, "New Capitals of Asia”, Econppic Development and
Cuitprai Change, Vol. V, No.3, April, 1957.

Murphey, Rhoads, ”Urbanization in Asia", Ekistics, Vol. 21,1966. pp.
8-17.

Murphey, Rhoads, F i f so 5 V s 0 ° nd C unt in
Chihp_§_flg__l_pp§hr, New York: Methuen, 1980.

Myers, R.H., "The Economic Transformation of the Republic of China",
W1. Vol.99. 1984.

Myint, H. "Comparitive Analysis of Taiwan's Economic Development With
Other Countries", Cohferehge 9h Experiences and Lessons of Economip

Dgyglppmgnr_in_T§ipph, Taipei: Academia Sinica, 1981.
Myrdal, Gunnar, o o c o derdevel ed 0 5, London:

Duckworth, 1957.

Myrdal, Gunnar, s a ° u t h ove t Nation , New
York: Pantheon, 1968.
Pannell, Clifton, T'hi-Chphg Thighp; Strpgiprg apd Ephction, Chicago:

University of Chicago Geography Research Paper No. 144, 1973.

Pannell, Clifton, "Past and Present City Structure in China", Town

Bianhipg Review, Vol. 48, No. 2, 1977. pp. 157-171.
Pannell, Clifton, a' eo a c nd H s o cal A oaches to

Eoreigh Area Stugies, Washington, D.C.: National Council for
Geographic Education, 1983.

Pannell, Clifton, "Urban Land Consolidation and City Growth in Taiwan",
Pacific Vieppoiht, Vol. 15, No. 2, September, 1974. pp. 111-122.

Pannell, Clifton. ”The Internal Structure of Chinese Cities: Nanking",

in Leung and Ginsburg eds., China; Urbapigaiion and National
Dgyglppmgpr, Chicago: Department of Geography, University of
Chicago, Research Paper No. 196, 1980.

Park. Robert. "Human Ecology". MW. Vol. 42.
No. 2, (July, 1936).

Peet, Richard, ed. e t o

Q_asemn_rarx_fiosisl_1§__es. Chicago Aldine. 1977.

Perloff, H., E. Dunn, E. Lampard, R. Muth, Rggipp§r_gg§pprp§§r_hhg
Econpmig Growth, Baltimore: John Hopkins University Press, 1960.

Pudup, Mary Beth. "Arguments Within Regional Geography", Progress in
Human Ceography, Vol. 12, No. 3, 1988.

Raper, Arthur ed., Crhgp ghg nghgirihl Taiwgh; Crpwded and Resourceful,
Taipei: Mutual Security Mission to China and The National Taiwan

University, 1954.

196
Reissman, Leonard. Thp_Drhpn_£rppp§§, Glencoe, IL: The Free Press, 1964.

Republic of China, Taipei City Government Information Office. "The City
of Taipei", Rppphlip pi Ching; A Rpierence Bopk, Taipei, 1983.

Republic of China, Taipei City Government Information Office. The New
ipage of Taipei, Taipei, January, 1982.

Republic of China, Taipe City Government Bureau of Budget, Accounting

and Statistics. Sipiigrippi Ahpirppi pi Taipei Municipality,

Taipei, Various Years.

Republic of China, Executive Yuan: Council for Economic Planning and
Development, Housing and Urban Development Department. Drban and

Regionai Deveippmehi Staiipiips, Repphiic of China. Taipei, Various

Years.

Republic of China, Council of Economic Planning and Development (CEPD),
Urban and Regional Development Statistics, Taipei, 1982.

Republic of China, Council of Economic Planning and Development, A

Deiinitioh of Drbah Epppiptioh ip Taiwpp,(in Chinese), Taipei,
1975.

Republic of China, Executive Yuan, Directorate General of Budget

Accounting and Statistics, Siatistipai Ahptract of the Republic of
Ching, Annual.

Republic of China, Ministry of the Interior, Tpiwgh-Ephien Demographic
Eppi Dooh, Annual.

Republic of China, Bureau of the Census, Cepprpl Cphsus Report on the
B,pphhij,p_pi_Ch_ipp_L Various Years (Particularly 1956, 1966, 1970 and

1980).

 

Republic of China, Taiwan Provincial Government, Governor's Office,

Eifty-one Years of Statistical Abstrapts of Taiwan (1895-1945),
Taipei, 1946.

 

Republic of China, Taipei Municipal Government, Repprt oi The Survey of
Famiiy Income and Expehditure, 1984.

Rees, Philip, si e t at e s in eric Citie , Chicago:
University of Chicago, Department of Geography Research Paper No.
189, 1978.

Roberts, Bryan, "Center and Periphery in the Development Process: The
Case of Peru", Cornelius and Trueblood eds., Lptih Aheripah
Research, Los Angeles: Sage Publications, 1975. pp. 77-106.

Rodwin, Lloyd, Npiipp§_ppp_Ciiip§, New York: Houghton Mifflin, 1970.
Rostow. W.W.. W. New York: Cambridge

University Press, 1960.

197

Rutter, Owen, Thrpugh Eorppaa; Ah Accppht oi Sapah'a ialand Coiony, New
York: H.W. Wilson Co., 1924.

Santos, Milton, "Society and Space: Social Formation as Theory and

Method", Antipode; A Radical Sourpai oi Geography, Vol. 9, No. l,
1977.

Schipper, Kristofer, "Neighborhood Cult Associations in Traditional

Taiwan", G.W. Skinner ed., The City ih Late Imperial China,
Stanford: Stanford University Press, 1977. pp. 651-676.

Shabad, Theodore, Chipa'a Chahging Map; National and Regiohal
Davpiopment i942-i921, New York: Praeger, 1972.

Sheu, Bor-shiou, tud o the r w h o c un Cit in the Last Ten

Tpars, Taipei: National Taiwan University, Department of Geography
Science Report No. 9, Feb. 1977.

Shevky, Eshref and Marianne Williams, The Sopial Area pf Lps Angeles;

Ahaiysis and Typpiogy, Los Angeles: University of California
Press, 1949.

Shevky, Eshref and Wendell Bell, Social Area Analysis; Theory,

illuatrative Applipatioh ahd Compptaiiohal Propedures, Stanford:
Stanford University Press, 1955.

Shi, Ming. ou d ed ea 0 iw Pe e' istor , San Jose, CA:
Paradise Cultural Association, 1979.

Shih, Chi-tseng. "Decentralized Industrialization and Rural Non-farm

Employment in Taiwan", Inpusiry pi Eree Chiha, Vol. 60, No.2,
August, 1983. pp. 1-20.

Shih, Timothy Tien-fu. "Spatial Patterns and Changes in the Level of
Urbanization and Their Implications for Regional Development in
Taiwan", Capgraphipa1_Raaaarph, Taipei: Institute of Geography,
National Taiwan Normal University, No. 7, January, 1981. (In
Chinese) pp. 153-173.

Sih, Paul, ed., Taipap_ip_upparn_Timaa, New York: St. Johns University
Press, 1973.

Sit. Victor ed .WLWis—m
i259 , Hong Kong: Oxford University Press, 1985.

Sjorberg, Gideon, The Era-inapprriai Ciry, Rapt ahd Rrepent, Glencoe:
Free Press, 1960.

Skinner. G. William ed.. WW. Stanford:
Stanford University Press, 1977.

Slater, David, "Colonialism and the Spatial Structure of
Underdevelopment: Outlines of an Alternative Approach With Special
Reference to Tanzania”, Abu-Lughod and Hays eds., Third Worip
Drhapiaariph, Chicago: Maaroufa, 1977. pp. 165-175.

198

Slater, David, "Geography and Underdevelopment", Ahiippde; A Radical

Warm. Vol. 9. No. 3. 1977. pp- 1-31-
Smith, David, e e G e e ' v n Une ual World,

Baltimore: John Hopkins University Press, 1979.

Soja, Edward and Richard Tobin, ”The Geography of Modernization: Paths,
Patterns, and Processes of Spatial Change In Developing Areas",

Abu-Lughod and Hays ed., Thirp_flprip_Drhapigariph, Chicago:
Maaroufa, 1977.

Speare, Alden, ”Urbanization and Migration in Taiwan", c o 0

MW. Vol. 22. No. 2. 1974. PP- 303-319-

Speare, Alden, Mary Speare and Lin Hue-sheng. "Urbanization,
Non-Familiar Work, Education and Fertility in Taiwan", Population
Stppies, Vol. 27, No. 2, July, 1973. pp. 323-333.

Speare, Alden, Paul Liu and Tsay Ching- lung. Drhahizariph_ahp
- - - ; Boulder, CO:

 

Westview Press, 1988.

Speare, Alden. "Migration and Family Change in Central Taiwan”, in Elvin

and Skinner eds., Thp Chipaap Ciiy Betweep Tgp Worlds, Stanford,
CA: Stanford University Press, 1974. pp.303-330.

Speare, Alden, Paul Liu, K.S. Hwang, C.L. Tsay and Mary Speare. "A
Measurement of the Accuracy of Data in the Taiwan Household

Register", Academia Sophomic Raperp, Taipei: Academia Sinica, Vol.
3, No. 2, 1975. pp.35-74.

Stites, Ronald, ”Small Scale Industry in Yingge, Taiwan", hoderh Chipa,
1982. pp. 247-279.

Sun, C. and Tsai H.H., "Urbanization, Regional Growth and Economic

Development in Taiwan", Tapparry_pi_Erpa_Chiha, Vol. 56, No. 4,
October, 1980. pp. 1-13.

Swain, H. and M. Logan, "Urban Systems: A Policy Perspective",
WW. Vol. 7. 1975- PP- 743-755-

Sweetser, Frank, "Factor Structure as Ecological Structure in Helsinki

and Boston". WW. Vol. 70. No. 3. 1965.

Takekoshi, YOsaburo, Sapapaaa_Rpia_ih_£prppaa, London: Green and Co.,
1907.

Tang, Mie-Chun. ' nth o 0 ca ield Stud
in_Taipai_Ciryr_Taixap, Taipei: National Taiwan University Press,
1978.

Tauber, Irene. "Migrants and Cities in Japan, Taiwan, and Northeast

China", in Elvin and Skinner, eds., Tha Chinese City Detween Two
Woripa, Stanford: Stanford University Press, 1974.

199

Thompson, Wilbur, "Urban Economic Growth and Development in a National
System of Cities," Hauser and Schnore eds., The Stupy pi
Drhahiaaiiph, New York: Wiley, 1971. pp. 431-490.

Thornton, Arland, Chang Ming-Cheng and Sun Ti-hsiung. "Social and
Economic Change, Intergenerational Relationships and Family

Formation in Taiwan", Demography, Vol. 21, No.4, 1984.
Timms, D. G., e a ' e esidentia
Diiiarphriaiiph, Cambridge: Cambridge University Press, 1971.
Trewartha, G.T., "Chinese Cities: Origins and Functions", Anpaia of the
Asaopiaiioh pi Aperipah Geographera, Vol.XLII, No.1, March, 1952.

Tsai, H. H., Urba G w h a d t e e of at a t ucture n Taiwan,
Taipei: Report of Land Use Planning Conference: Academica Sinica,
1979.

Tsai, Hsung-Hsiung, "Real Estate Development in Taiwan, Republic of

China”, Tapaarry_pi_Rrap_Chiha, Vol.70, No.10, 1988. pp. 15-25.

Tsai, Hang-chin. "Rural Industrialization in Taiwan: Its Structure and

Impact on the Rural Economy", ipdpatry oi Eree Chiha, Vol. 56, No.
6, December, 1981. pp. 17-32.

Tsai, Hang-chin, "Rural-Urban Employment Linkage in Taiwan's
Urbanization Process", Eppppmip_Rayiap, No. 231, May-June, 1986.
pp. 8-29.

Tsai, Hong-chin. "The Relationships of Population Growth and
Socio-Economic Changes Between Satellite Areas and The Inner

City-Taipei Metropolitan Case", Jpprhai pf Ropplatioh Studies,
Taipei: National Taiwan University, No. 8, 1985.

Tsay, Ching-lung, o e t d i t -W d an 8' ud
, Taipei: The
Institute of Economics Monograph Series, Academi Sinica, 1985.

Tsay, Ching-lung, "Migration and Population Growth in Taipei
Municipality", TRC, Vol. 57, No. 3, March, 1982. pp.9-25.

Ueno, Ken'ichi. The Rpsidehtial Structure oi Tokyo Darihg the Pre-World

War ii Raripp oi Showa (i230), Tokyo: Science Report of the
Institute of Geoscience, University of Tsukuba, Vol.6, 1985.

Van Arsdol, Maurice, Santo Camilleri and Calvin Schmid, "The Generality
of Urban Social Area Indices", Aperipan Socioiogicai Review, Vol.
23, 1958.

Vielle, Paul, ”The Global Space of Capitalist Organization", Space and
Society, No. 12, May, 1974.

Wan, Henry, "Manpower, Industrialization and Export Led Growth - The

200

Taiwan Experience", Hsiung ed., e iwa er ence. 1950-1980,
NewhYork: Praeger, 1981. pp. 131-141.

Wang, C.Y., "A Study of Regional Development Theory and Policy- The
Interaction of Cities and Regional Development”, Spience Report,
Taipei: Department of Geography, National Taiwan University, No.
13, 1988. (In Chinese) pp. 205-214.

Wang, Su-chang. Rreipranca anp Chpipe pf Residencea in Southeastern
Taippi, Unpublished Dissertation, University of Minnesota, 1981.

Wheatley, Paul, Tha Riv vpr pi iha Ehu ur Charge era; A Rraliminary Enquiry

lnio ihe Cr igins anp Cha arapie ro of ihe Anpient Chinese City,
Chicago: Aldine, 1971.

Wickberg, Edgar. "Continuities in Land Tenure, 1900-1940” in E. Ahern

and H. Gates, eds., Tha Anrhropplpgy pi Taiwanese Society, pp
212-38. Stanford: Stanford University Press, 1981.

Wilber, C.L., "Urbanization in Taiwan: 1964-1975", Journal oi Population
Studies, Vol. 4, No. 1, 1980.

Wilber, C.L., "Migration and Occupational Achievement in Taiwan",

Jpnrnal_pi_Sppiplpgy, Taipei: National Taiwan University, No. 6,
1982.

Williams, Jack F., "Urban and Regional Planning in Taiwan: The Quest for

Balanced Regional Development” Tijpaphriii Vopr Econpmische En
Sopiale Geografie, Vol.79, No. 3, 1988, Amsterdam: Royal Dutch

Geographical Society.

Williams, Jack F., "Toward a Model of Land Reclamation/Resettlement", Ma
and Pannell eds., v ° ese an erica Views, New
York: Methuen, 1981.

Williams, Jack, F. and S. Brunn eds., Ciripa pi rha Wprlp; Worlp
Regional urban Develppmeni, New York: Harper and Row, 1983.

Williams, Jack F., ”Taiwan's Development Strategy Since 1949: Prospect

and Problems", e Ta wan Issue° r ceed n o the S osium at
W. Nov. 6. 1975. East Lansing: Asian
Studies Center, Michigan State University, Occasional Paper No. 5,
May 1976.

Williams, Jack F. and C. Sutherland and C.Y. Chang. "Land Use in the
Taipei Basin”, na 3 f the ssociation of erican Geo ra hers,
Vol. 78, No.2, 1988.

Willmott, W. E., Eppnpnip Drganizatipn in Chinese Sppiety, Stanford:

Stanford University Press, 1972.

Wirth, Louis, "Urbanism as a Way of Life", Aneripan Journal of
Sociology, Vol. 44, No. 2, (July), 1938.

World Bank. City Size and Rational Spatial Strategies in Developing

201
Countries, World Bank Working Paper No. 252. Washington D.C., 1977.

World Bank. a i a a iz i 0 ve o n Cou tries,
World Bank Working Paper No. 374. Washington D.C., 1979.

Wu, T. S., "A Framework For The Study of Urban Adaptation in The Process
of Rural Migration", Journal of Sociology, Taipei: National Taiwan
University, No. 7, 1971.

Wu, Yuan-11. Inppna Diairihntipn in ihe Rrppesa pi Economic Growth oi

thp Repnblip pi China, College Park, MD: University of Maryland,
Occassional Papers Series in Contemporary Asian Studies, No. 2,

1977.

Yang, Chung-hsin, "Urban Agglomeration Economics in the Manufacturing

Industries in Taiwan", Rropeedings and Rapers oi the Conference on

U owt d ev o e t e c fi e 10 ,
Taipei: Institute of Economics, Academia Sinica, Jan. 9-11, 1984.

Yang, Martin. Socio-economic Results oi Land Reform in Taiwan, Honolulu:

East-West Center Press, 1970.

Yang, Wan-chin. "A Study of the Underground Water Supply and Its Use in
the Taipei Basin", ank o a wa arterl , Vol. 23, No. 4, 1972.
(In Chinese) pp. 235-253.

Yeh, Chang-Chou and Chiang Yu-sheng. "Transportation Problems in

Taiwan's Cities", Dank oi Taiwan Qaartarly, Vol. 20, No. 3, 1969.
(In Chinese) pp. 215-258.

Yen, Sheng-hsiang. "A Study of Industrialization and Its Structural

Change in Northern Taiwan”, Dank pi Taiwan Qaarterly, Vol. 24, No.
3, 1973. (In Chinese) pp. 1-15.

Yi, Chin-Chun. "Urban Housing Satisfaction in a Transitional Society: A
Case Study in Taichung, Taiwan", Drhan_Srnpia§, Vol. 22, 1985.

Yuan, D.Y., "The Rural-Urban Continuum: A Case Study in Taiwan", Rural

Sopiology, Vol. 26, 1964.

Zuiches, J.J. and G. Fuguitt, "Residential Preferences and Population
Distribution", Danpgraphy, Vol. 12, 1975.