Sociology Mind
2014. Vol.4, No.1, 15-23
Published Online January 2014 in SciRes (http://www.scirp.org/journal/sm) http://dx.doi.org/10.4236/sm.2014.41002
15
Community Inhabitants’ Attitudes on the Partitioning of Urban
Space Derived from South-Eastern Asian Migrant Workers’
Gathering in Urban Commercial Area—A Case Study on Tainan
City, Kaohsiung City and Taichung City in Taiwan
Kung-Hung Chen
Graduate Institute of Taiwan Culture, National University of Tainan, Taiwan
Received October 5th, 2013; revised November 24th, 2013; accepted December 11th, 2013
Copyright © 2014 Kung-Hung Chen. This is an open ac cess article distrib uted und er the Cre ative Co mmons At-
tribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the
original work is properly cited. In accordance of the Creative Commons Attribution License all Copyrights ©
2014 are reserved f or SCIRP an d the owner of the intell ectu al pro pert y Kung-Hung Chen. All Copyright © 2014
are guarded by law and by SCIRP as a guardian.
This study is to explore south-eastern Asian migrant workers’ impacts and their influential factors of ga-
thering/consumption activities on divided cities in Tainan City, Kaohsiung City and Taichung City. We
will take on an empirical study in the cities’ commercial area or shopping street in three cities. Three ma-
jor research methods are applied, including questionnaire survey, participant observation and regression
analysis. Data will be obtained from the questionnaire answers of local inhabitants and analysed via
SPSS10.3 statistical methods. The main outcomes of this study are as follows: 1) In Tainan City, Kaoh-
siung City and Taichung City, we have found that the partitioning of urban space has begun to become.
Among this, the outcome of Taichung City is much stronger than that of the two other cities. 2) In the
impacts of migrant workers’ gathering on commercial streets, community inhabitants also have more to-
lerant attitude than shop keepers. The community inhabitants of Taichung City have more hostile than
those of the two other citi es . 3) In comparison of South-eastern Asian migrant workers’ gathering in urban
commercial area and its relations between the partitioning of urban space, from the viewpoint of local in-
habitants, we have found different influential factors among Tainan City, Kaohsiung City and Taichung
City. We can say that these factors will be valid predictable variables on divided cities resulted from mi-
grant workers’ gathering/consumption activities in the paper. In the meantime, we can obtain these in-
fluential factors among these three cities through two variables: 1) community inhabitants’ socio-eco-
nomic attributes and 2) community inhabitants’ responses to migrant workers’ gathering/consumption ac-
tivities in commercial area. Basically we can find these influential factors among the paper which are the
same as the Netherlands and Singapore, and not the same as Guangzhou, China.
Keywords: Gathering of South-Eastern Asian Migrant Workers; Partitioning of Urban Space; Tainan City;
Kaohsiung City, Taichung City
Introduction
According to the statistic data of Bureau of Employment and
Vocational Training, from at the end of September, 2011, it is
indicated that the total amount of migrant workers in Taiwan
and Fukien area is 417,844, including Tainan City with 30,237
people, Kaohsiung city with 30,394 people, and Taichung city
with 56,374 people. (These three cities that are mentioned
above are after merger upgraded to municipalities). Saskia Sas-
sen (1988, 1991, 1996) clearly analyzed in many of her writings
how migrant workers, in the era of globalization and economic
dynamism, created new global cities and how this kind of large-
scale flow of capital and human resources inevitably causes a
great impact on residential countries urban spaces. As a series
of researches about “Divided Cities” by Peter Marcuse (2002)
proposed, it is more obvious to notice that, in the global market
economy, the spatial structures which are the spaces for pro-
duction, consumption and living are formed by people in dif-
ferent economic classes. Thus, the effects on the migrant work-
ers versus the spatial development of commercial communi-
ties and the changes of the space axis are really interesting and
have significant academic meanings for the researches of met-
ropolitan spaces. In other foreign countries, the empirical re-
searches, which relate to the ideas of “Divided Cities”, are more
exuberant and complete, but less direct to the factors of the
space division of cities. In Taiwan, most of the empirical re-
searches are concerning the distribution of the gatherings and
consumption activities of migrant workers in the cities, as well
as the aspects of socialized spaces. Thus, to explore from
community residents’ points of view, how migrant workers
from Southeast Asia’s gathering cause an effect on urban space
division is not only the characteristic of this paper, but also the
research motivation of this paper.

K.-H. CHEN
16
In this paper, we use the concept called “divided cities”, pro-
posed by Peter Marcuse to see how local community residents’
attitude changes and their responses to the inflow of different
cultural groups impact on urban space division, and to investi-
gate, analyze and interpret how migrant workers’ consumption
activities cause an effect on urban community space. We use 3
of Taiwan’s biggest cities, Tainan city, Kaohsiung City and
Taichung city as our objects of research, which is also the pur-
pose of this study.
Literature Revi ew
In other countries, the empirical studies about the concept of
“divided cities” are richer and more complete, but however,
they rarely focus directly on urban space division’s influencing
factors’ studies. Therefore, we introduce the cases as following:
Netherlands
In his paper called “Towards Partitioned Cities in the Neth-
erlands” (2002), Ronald van Kempen targeted the changes of
patterns of urban space’s segment in a well-developed welfare
state—the Netherlands. He found that, in the Netherlands, so
called “partitioned cities” have become a more and more ob-
vious trend and they have been developing rapidly. Apart from
the changing role of the government, the change of economic,
population and socio-cultural factors must be taken into account
as well (Marcuse & Kempen, 2002). In the Netherlands, me-
tropolis of ethnic minorities’, segmentation and concentrated
phenomena are very obvious. In recent years, a large number of
migrants moved to the Netherlands, and they also entered its
labor market to look for jobs, this phenomenon has been con-
tinuing to expand, due to the supply and demand imbalance, it
leads to the result of persistent structural unemployment. Mi-
grants (Especially Turkish and Morocca n) also suffer from this
development, many low-skilled migrants must rely on state
support in order to survive, more and more phenomena of “par-
titioned cities” are resulting from the welfare state’s recession,
which leads to the results of social and spatial problems. In
short, the factors that made the Netherland’s “partitioned cities”
phenomenon so obvious are: the changing role of government,
economic, demograph i c, so cio-cultural changes, employment rate
and the level of income.
Singapore
Brenda Yeoh and Shirlena Huang (1998), in the study of
housework female migrant workers’ taken in strategies and
types in the Singapore’s public space, by viewing the social
map of these migrant workers, it can be seen how it has been
structured and compromised on public space and investigated
how these migrant workers, viewed as margins of society in
Singapore built their urban landscape. They found out, in capi-
talist, so-called “partitioned cities” reflect more clearly and
strengthen the gender division of labor as well as the characte-
ristics of daily live experience. It not only presented the space
by the traditional patriarchy, but also presented by the space of
racial factors and other means of isolation, to expand it. This
study indicated that these housework female migrant workers
aren’t entirely passive recipients of dominant practice and ideas,
in fact, they are capable of taking a variety of different ways
and strategies to use and compete the public space.
Their research based on public domain’s points of view is
based on the following: 1) deprived of space and 2) competition
of space, in order to analyze them. The former, use Filipino
housework labor’s gathering and all kinds of consumption ac-
tivities in the shopping mall—“Lucky Plaza”, as the research’s
main focus. The shopping mall—“Lucky Plaza”, has become
like “little Manila”. Over the weekend, there are around 2000
Filipino female migrant workers gathering in this area. Thus,
the Filipino migrant workers have caused the colonization trend
of this public space. Gradually, as for some Singaporeans, “Lu-
cky Plaza” has been a “partitioned city”, changing from an
“other type of space” to a “hesitant space”. The latter, by the
limitations of “other type of space”, the female migrant workers
can still make differences by managing their ways of living
their daily life, including understanding how to define and use
space well, we’ll use this concept as our entry point, in order to
analyze the so-called “competitive space”. In short, the influen-
cing factors that made Singapore’s “Lucky Plaza” become such
an obvious “partitioned city”, included not only a laissez-faire
government, but also the tolerance of the community residents,
and most importantly, how these Filipino migrant workers used
“competitive space” as means to colonize this pubic space.
Guangzhou, China
Huimin Du and Si-ming Li (2010), have been observing
China’s economic reform and the process of urbanization. In
recent years, millions of migrant workers came from the coun-
tryside to the big cities in China, and they have become an im-
portant part of the labor force of the cities, and eventually
shaped the so-called “urban villages”. In this development trend,
both of them focus on the “urban villages” laborers’ socioeco-
nomic background, life satisfaction, living environments etc., to
describe the situations in detail, and at the same time use “com-
munity satisfaction” and “community contacts” to explore the
emotions of these “urban villages” laborer’s emotions about the
community.
The research findings show that “urban villages” not only pro-
vide a place to live, but the laborers of the “urban villages” also
use various ways to integrate in the cities. Community senti-
ment is mainly based on labor’s cognition on neighborhood
environment quality and neighborhood relationships, in addi-
tion, the effect of local participation and migrants social net-
works also gradually unfolded. Participating in local affairs
bring residents a higher level of community satisfaction, how-
ever, participating in local affairs doesn’t have a statistically
significant difference on degrees of community contact; like-
wise, the migrant social network theory can also explain com-
munity sentiment, it can fully support community satisfaction,
rather than degrees of community contacts. In short, the reasons
why Guangzhou, China’s “urban villages” phenomenon seems
so similar to “partitioned cities” are mainly because of: good
cognition on neighborhood relationships, public participation,
migrant social network, etc. But the Netherlands and Singa-
pore’s cases’ migrant socio-economic attributes and roles of
government are comparably less important.
In Taiwan, the empirical researches about the concept of “di-
vided cities or partitioned cities” are relatively few. Most of the
very few researches are about migrant workers gathering, their
consumption activities’ distribution in the urban space and their
socialized space field kind of researches. For example: The first
paper in Taiwan that studied about migrant worker’s space
issue was written by Sheu, Horng-Yih in 2000. It is called “The

K.-H. CHEN
17
use of Space and Its Effect in a Filipino Labor Gathering Area
—Christopher Church Area on Chung-Shane North Road”. He
saw the migrant workers from an urban point of view, and use
“Filipinize” to interpret the Filipino migrant workers impact on
Chung-Shan North Road, focusing the study on Filipino mi-
grant workers space conflict with the city residents, and finally
concluded that the public sector should intervene to control it as
a policy recommendation. Wu Bi-nar’s thesis (2003), “Chung-
Shan—The Formation of a Filipino Migrant Workers’ Com-
munity Space in Taipe i”, basically compared with Sheu, Horng-
Yih’s paper, has a better insight on Filipino migrant workers
relationship with the space., By gathering around Christopher
church area in a common space-time and the entering of sur-
rounding shops, Chungshan is where displaced Filipino migrant
workers can re-territorialize the space. This space has given rise
to richer and more diversely activities, it not only satisfies the
Filipino migrant worker’s transnational life needs in Taiwan, it
also became Filipino labor community’s leisure, consumption
and cultural center. In one of Pei-Chia Lan’s paper (2002) call-
ed “A Transnational Topography for the Migration and Identi-
fication of Filipino Migrant Domestic Workers”, we found out
they perform multiple roles and shifting identities through the
front/backstage segregation in social space (working days vs.
rest day; home country vs. host country), isolating labor and
consumption activities, and the differentiation of temporal ho-
rizons (now vs. future) in order to build floating identification.
In Chou Cheng-Hung’s thesis called “Migrant labors: a ca talyst
for rebuilding urban space”, he found out that migrant workers
always appear at a certain time, which alienates the urban space,
the ethnic contact, cultural contact and the overlapping use of
objects’ process generates a “critical space”, by walking the
delicate balance between public, privacy, other people and our-
selves. This paper is noteworthy because it has some similari-
ties to our research of Qixian 3rd Road, Kaohsiung city’s ob-
served phenomenon, thus, it has a higher reference value.
Chih-Hung Wang (2006), used “Dis/placed Identification and
Politics of Space: The Consumptive Ethnoscape around Tao-
yuan Railroad Station” to explore what kinds of social tension
and conflicts are derived from the formation of Southeast Asian
consumptive ethnoscape around Tao-Yuan railroad station. And
how do these tensions and conflicts appear as contention about
uses and meanings of space? What the relation between the
formation of consumptive ethnoscape and the negotiation is of
dis/placed identities, as well as various stores and police control?
Such topics are worth being referenced. Chen Kung-Hung’s
(2011a-2011c, Chen, 2008) research is different from the pre-
vious studies because he advocates: the original existence of the
urban business district and its’ surrounding communities’ local
community residents and shop owners, their original life style
and shop’s mode of operation will change because immigration
of different cultural groups, will bring influence and change on
its culture and space, their way of presenting changes of lives,
culture and space are called the “result” of “transnational space”
or “divided space”. Under the concept of “transnational space”
or “divided space”, the paper focuses on the attitudes of local
community residents and shop owners toward the impacts of
immigration of different cultural groups, of urban space parti-
tioning and it also explains the degrees of “partitioning of urban
space”. It emphasis the main research orientation on “urban bu-
siness district’s migrant workers gathering caused partitioning
of urban space phenomenon”, it’s an innovative area of re-
search, which is worthy of attention.
Based on the literature, theories and related researches above,
we believe, there are some points in common:
1) Based on the theory of “migrant workers, in the era of
globalization and economic dynamism, created a large-scale
flow of capital and human resources, they inevitably cause a
great impact on residential countries’ urban space”. This is the
context that so-called “tra nsnational space” and “space partitio-
ning” form, it belongs to global macro kind of driving force.
2) Based on daily life, leisure, shopping and religion needs,
migrant workers naturally have to gather in the business dis-
tricts and do consumption activities, and their consumption ac-
tivities have become urban community space’s “outside invad-
er”, “influencer”, this is the reason why the “transnational space”
or “space partitioning” is formed. Furthermore, local commu-
nity residents living in the business districts as well as sur-
rounding communities’ l ocal residents have changed their ways
of living and the ways of using the space due to the impacts of
immigration of different cultural groups on culture and space.
Based on the impact of urban community space, they have be-
come migrant worker’s consumption activities’ “local insiders”,
“influence”, at the same time, their ways of presenting their
changes of life, culture and space are the results of “transna-
tional space” or “space partitioning”.
3) In view of the above cases, they’re in fact a ll the same re-
gardless of space, which can be the evidence of the “causes”
and “results” that were mentioned above. These “causes” and
“results” are the research motivations. This paper’s main pur-
pose is also based on the concept of “transnational space” or
“space partitioning”, to conduct investigations, analysis and in-
terpret the feelings of migrant workers’ space usage, the atti-
tude of the local community residents toward the impacts of
immigration of different cultural groups of urban space parti-
tioning, and the influencing factors of migrant workers’ con-
sumption activities to the community space.
Theoretical Framewo rk
According to Figure 1, we can propose some hypotheses of
the research as follows.
Hypothesis 1:
The attitude of the local community residents toward the im-
pacts of immigration of different cultural groups of urban space
partitioning is an important influencing factor of the level of
urban partitioning.
Hypothesis 2:
Community inhabitants’ socio-economic attributes are im-
portant influencing factors of the level of urban partitioning.
Figure 1.
Theoretical framework.
Under the global economy, large-scale flow of migrant
workers have made a great impact on resettlement
countries-transnational space or space partitioning.
The attitude of the local community residents toward
the impacts of immigration of different cultural gro ups
of urban space partitioningCommunity residents’ basi c
socio-economic attributes
Explor ing the i nfluenc i ng fact o rs o f migra nt workers ’ gathe r ing t o
urban space partitioning from community residents’ viewpoints

K.-H. CHEN
18
Research Methods and Content
Research Methods
The required information comes from local community resi-
dents of Tainan, Kaohsiung, and Taichung City. These mem-
bers of the community live around the area where Southeast
Asian migrant workers like to gather in the business districts.
Designing a questionnaire of “urban business district’s migrant
workers’ gathering in urban business districts and the its rela-
tions between the partitioning of urban space” (community
residents section) as one set, and we’ll go to the commercial
area for sampling, in order to study the relations between mi-
grant workers’ gathering and the partitioning of urban space.
The questionnaire’s content mainly included: 1) The attitude
reflection on migrant workers’ weekend gathering and its im-
pact of urban space partitioning, and 2) the basic information of
the respondents. About data analysis, we use version 10.3 of
SPSS as the s tatistic al software to do the statistical analysis.
Types of Stores in the Research Areas
In Tainan City, the so-called “Indonesian Migrant Worker
Street” refers to the shopping streets around Fu-Bei Street/Xi-
Hua Street/Bei-Zhong Street. It currently has a total of 43 shops,
including 2 Indonesian restaurants, ENCAR, Ming-Yue Viet-
namese restaurant, 2 foreign exchange banks, Yi-li Telecom
Service Co., Ya-Jiu food store, 3 Thai restaurants, TK store,
Charlie Brown restaurant, EEC store; When adding the 12
stores that are owned by local Tainan shop owners, it has a total
number of 26 shops. It contains more than half of the total
shops which include a tea shop, a supermarket, a telecom com-
pany, a cafeteria, complex restaurants, noodle restaurants, a ve-
getarian restaurant, the Qiao-Yi kitchen and the Ma-Na restau-
rant. They’ve become places where migrant workers (Indone-
sian-based) often consume at. The stores mainly opened in
2001 in a total number of 26 stores, and 36 rental stores.
In Kaohsiung City, the so-called “Southeast Asian Migrant
Worker Street” is mainly around the Kaohsiung Railway Sta-
tion and the Kaohsiung port area. It’s where migrant workers
like to gather and shop in Kaohsiung City nowdays. The former
includes: 1) Jianguo 3rd Road: a total of 117 stores, 2) Zhong-
Hua 3rd Road: a total of 100 stores, 3) Qi-Xian 1st Road: a
total of 26 shops, 4) Qi-Xian 2nd Road: a total of 88 shops, 5)
Ba-De 2rd Road: a total of 82 shops, the latter includes Qi-Xian
3rd Road and Bi-Zhong Street, it has a total of 99 shops, but
from the interviews of the bosses, there are around 30 shops
that Filipino migrant workers rarely go to. And the rest of the
60 shops’ main c ustomers are Southeast Asian migrant workers,
or the shops that Southeast Asian migrant workers often go to.
Among them, Qi-Xian 3rd Road has a total of 46 shops, in-
cluded 6 shops which owned by Taiwanese boss and hired South-
east Asian as employees, they’re Qi-Xian restaurant , Sally mini
supermarket, MARINERS eatery, PINOY BISTRO BAR
(opened two), and CEBU CITY Korean restaurant; and there
are 23 shops on Bi-Zhong Street, including 2 shops owned by
Taiwanese bosses which hired Southeast Asian as employees,
namely “We llcome” and “Annie’s boracay”. Most of the shops
have opened in less than 10 years, and half of them rented the
place to open up the stores. So, it can be seen that same as in
Tainan City, because of migrant workers gathering and con-
sumption in recent years, it’s common for non-local residents to
come all the way to Kaohsiung City to open up stores.
In Taichung City, it’s mainly around the business district
opposite the Taichung Railway Station, includes where the
migrant workers like to gather and shop most in Taichung City:
Taichung First Square/Lu-Chuan Western Street and two shop-
ping streets: Cheng-Kung Road, Chi-Kuang Street: 1) First
Square/Lu-Chuan Western Street: They mainly distributed on
1st to 3nd floor, it has a total of around 120 shops, and there are
around 1/4 shops owned by foreigners (Mainly from Southeast
Asian countries), the types of shops are mainly restaurants,
eatery, shopping malls, and telecom service companies, such as
“BIG KING Shopping Mall”, “Indah Counter”, “Tai-Luo 77”,
Thailand restaurants, “Vinh Ha Long snack bar”, Vietnamese
Restaurants, et al.; 2) Cheng-Kung Road: it has a total of 27
shops, and there are around 6 shops owned by foreigners, such
as BOSS STORE, the branch of BOSS STORE, Café Batavia,
JASON STORE, 3) Chi-Kuang Street: it has a total of 53 shops,
and there are around 2 shops owned by foreigners and Viet-
namese snack bar .
Statistical Respondents and Sample Size
The survey respondents are mainly community residents. In
Tainan City, the communities around Fu-Bei Street/Xi-Hua
Street/Bei-Zhong Street have a total number of about 380
households, using household as a sampling unit, it has 100 ef-
fective samples and thus the sampling ratio is 26.3%. In Kaoh-
siung City, the “Southeast Asian Migrant Worker Stree” con-
tains 6 migrant worker streets, which have a total number of
about 1150 households. Using household as a sampling unit, it
has 421 effective samples which results into a sampling ratio of
36.6%. In Taichung City, the “Migrant Worker Street” is
around the business district, on the opposite of the Taichung
Railway Station, including Taichung First Square/Lu-Chuan
Western Street and two shopping streets: Cheng-Kung Road,
Chi-Kuang Street, which has a total number of about 1200
households. Using household as a sampling unit, it has 400
effective samples which results into a sampling ratio of 33.3%.
Discussion of Survey Results
Discussions about the Regression Analysis of
Influencing Factors of Migrant Workers’ Strong
Movement towards Urban Spaces Are Partitioning
1) A. Tainan City
Seeing from Table 1 and the overall regression model, the
overall regression model has an explanatory power (F = 1.968,
P < .05), it can effectively explain why migrant workers have
39.9% variance as positive impact on consumption activities.
Among all these, residence time has reached statistically signi-
ficance; it’s an effective forecasting variable, because β is neg-
ative. It shows that the longer the residents live in the commun-
ities, the less they think migrant workers have positive impact
on consumption activities. Moreover, for respondents that work
in agriculture, forestry, fishery, and animal husbandry field, it
has also reached a statistically significance. It indicates that
comparing people who are working in agriculture, forestry,
fishery, and animal husbandry field with middle class people,
people who work in those fields tend to think that migrant
workers have better positive impacts on consumption activities.
At the same time, the standardized regression equation is as
following:
Y = −.053 Gender (male ) + .201 Age + .221 Educat iona l Level

K.-H. CHEN
19
Table 1.
Regression coefficient table.
Predictor Standardized Regression
Coefficient β t Significance
(Constant) 4.154 .000
Gender (male) −.053 −.466 .643
Age .201 1.360 .178
Education al Level .221 1.761 .082
Residence Time −.246 −2.056 .043*
Monthly Salary .167 1.516 .134
Family Structure −.073 −.659 .512
Marriage (unmarried) .058 .435 .665
Marriage (divorced) .107 .992 .324
Religion (Catholic) −.122 −1.140 .258
Religion (Christian) −.147 −1.339 .185
Religion (Taoism) .134 1.193 .237
Religion (Other) .058 .553 .582
Activity time (noon) .114 .792 .431
Activity time (evening) −.168 −1.225 .225
Activity time (other) −.235 −1.747 .085
Contact time with the
migrant workers
(weekends or holiday) −.039 −.267 .790
Contact time with
the migrant
workers (other) −.028 −.193 .847
Occupations (agriculture,
forestry, f i shery, animal
husbandry) .290 2.875 .005**
Occupations (industries) −.056 −.468 .641
Occupations
(low-level of white-collar
office labors and low-level
military personnel)
.027 .255 .799
Occupations
(high-leve l of white -collar
labors, middle class) .028 .240 .811
Occupations
(homemaker) −.031 −.253 .801
Occupation (none) .025 .202 .840
Occupation (other) .108 .857 .394
Social-identity from
community residents to
migrant workers .011 .107 .915
Note: The dependent variable Y is the total score of migrant workers positive
impacts on commercial activities.
+ −.246 Residence Time + .167 monthly salary + −.073 Family
Structure, .058 Marriage (Unmarried) + .107 Marriage (Di-
vorced) + −.122 Religion (Catholic) + −.147 Religion (Chris-
tian) + .134 Religion (Taoism) + .058 Religion (Other) + .114
Activity time (noon) + −.168 Activity time (evening), −.235
Activity time (other) + −.039 Contact time with the migrant
workers (weekends or holiday) + −.028 Contact time with the
migrant workers (other) + .290 Occupations (agriculture, fo-
restry, fishery, animal husbandry) + −.056 Occupations (indus-
tries) + .027 Occupations (low-level of white-collar office la-
bors and low-level military personnel) +.028 Occupations (high-
level of white-collar labors, middle class) + −.031 Occupations
(homemaker) +.025 Occupations (none) + .108 Occupations
(other) + .011 Social-identity from community residents to mi-
grant workers.
2) Kaohsiung City
Examining Table 2 and the overall regression model, you
can see that the overall regression model has an explanatory
power (F = 1.812, P < .01), it can effectively explain why mi-
grant workers have 12.2% variance on positive impact on con-
sumption activities. Amongst all these, Religion (Islam)’s β =
−.109, (t = −2.241, p < .05) has reached statistically signific-
ance, it’s an effective forecasting variable, because β is nega-
tive, it shows that non-Muslim residents think migrant workers’
gathering makes urban space partitioning more serious. More-
over, marriage (Divorced)’s β = −.117, (t = −2.363, p < .05) has
also reached a statistically significance. It indicates that non-
divorced residents tend to think migrant workers’ gathering
makes urban space partitioning more serious. And as for the
reason why “Married residents have lower degree of satisfac-
tion to community spatial links” we can find explanations from
resident’s marriage status and their current family structure.
According to the surveyed residents, the proportion of married
people is much higher than unmarried and divorced people,
because married couples with kids, married couples with kids
plus the parents, four generations of families… all these above
in total make the proportion higher than others, considering the
cultural and spatial impacts due to immigration of different
cultural groups, people have to consider the negative impacts
that family members will possibly encounter. Usually married
people tend to tolerate the situation and even accept it. In other
words, it creates a lower degree of satisfaction to community
spatial links, and they tend to think it can make urban space
partitioning more serious. In the end, residents attitude on mi-
grant workers’ β = .121, (t = 2.444, p < .05) has reached a sta-
tistically significance, so it's an effective forecasting variable. It
indicates the more positive attitude of residents towards migrant
workers gathering, the easier they think it leads to urban space
partitioning.
At the same time, the standardized regression equation is as
follows:
Y = −.070 Gender + −.149 Age + −.021 Educational Level +
−.027 Monthly Salary + −.013 Family Structure, + .043 with or
without Migrant Maids Marriage + −.069 Religion (Catholic) +
−.041 Religion (Christian) + .002 Religion (Buddhism) + −.109
Religion (Islam) + −.038 Re ligion (Other) + −.026 Occupations
(agriculture, forestry, fishery, animal husbandry) + −.027 Oc-
cupations (industries) + .028 Occupations (low-level of white-
collar office labors and low-level military personnel) + −.019
Occupations (high-level of white-collar labors, middle class,
high-end military personnel) + −.005 Occupations (Bourgeois,
freelancers) + .011 Occupations (big capitalists) + −.021 Occu-
pations (homemaker) + .048 Occupations (none) + −.012 Mar-
riage (Unmarried) + −.117 Marriage (Divorced) + −.060 Activ-
ity time (morning) + .077 Activity time (noon) + −.006 Activity
time (afternoon) + −.047 Activity time (midnight) + −.100
Contact time (morning) + .002 Contact time (Noon) + .015
Contact time (Afternoon) + −.009 Activity time (midnight)
+ .121 Community residents’ attitude on migrant workers.
3) Taichung City

K.-H. CHEN
20
Table 2.
Regression coefficient table.
Predictor Standardized regression
coefficient β t Significance
(Constant) 11.098 .000
Gender −.070 −1.268 .206
Age −.149 −1.967 .050
Education al Level −.021 −.390 .697
Monthly Salary −.027 −.460 .646
Family Structure −.013 −.219 .827
With or Without
Migrant Maids .043 .848 .397
Religion (Catholic) −.069 −1.351 .178
Religion (Christian) −.041 −.781 .435
Religion (Buddhism) .002 .037 .970
Religion (Islam) −.109 −2.241 .026*
Religion (Other) −.038 −.707 .480
Occupations
(agriculture, forestry,
fishery, a ni mal
husbandry)
−.026 −.495 .621
Occupations
(industries) −.027 −.430 .667
Occupations (low-level
of white-collar office
labors and low-level
military personnel)
.028 .494 .621
Occupations (high-level
of white-collar labors,
middle class, high-end
military personnel)
−.019 −.308 .759
Occupations
(Bourgeois, freelancers) −.005 −.072 .943
Occupations
(big capitalists) .011 .222 .824
Occupations
(homemaker) −.021 −.339 .734
Occupations (none) .048 .807 .420
Marriage (unmarried) −.012 −.147 .883
Marriage (divorced) −.117 −2.363 .019*
Activity Time
(Morning) −.060 −1.028 .305
Activity Time (noon) .077 1.333 .183
Activity Time
(Afternoon) −.006 −.095 .925
Activity Time
(Midnight) −.047 −.831 .407
Contact Time
(Morning) −.100 −1.795 .073
Contact Time (Noon) .002 .037 .970
Contact Time
(Afternoon) −.015 .236 .814
Contact Time
(Midnight) −.009 −.156 .876
Community r esidents’
attitude on migrant
workers .121 2.444 .015*
Note: The dependent variable Y is the total score of urban space partitioning.
(Please name Y on your own).
Seeing from Table 3 and the overall regression model, the
overall regression model has an explanatory power (F = 6.437,
P < .001), it can effectively explain why migrant workers have
37.3% variance as migrant workers’ impact on urban space
partitioning. Among all these, industrial labors’ β = −.162, (t =
−3.147, p < .01) have reached a statistically significance, thus
it’s an effective forecasting variable. β being negative indicates
that non-industrial Labors think migrant workers’ consumption
gathering makes a great impact on urban space partitioning.
Bourgeois/freelancers’ β = −.181, (t = −2.939, p < .01) have
reached a statistically significance, thus, it indicates that people
who aren’t bourgeois/freelancers tend to think migrant workers’
consumption gathering makes urban space partitioning easier.
Jobless people’s β = −.125, (t = −2.234, p < .05) has reached a
statistically significance, thus, it indicated residents who have
jobs think migrant workers’ consumption gathering makes ur-
ban space partitioning easier. We can find the explanation of
the results from community residents’ jobs: low-level of white-
collar office labors, high-level of white-collar office labors,
housekeepers and other jobs, these 4 types of jobs fields have a
total of 48.2% in the communities. It’s the composition of the
bottom, and they have more direct feelings on changes of daily
life. So, they think migrant workers’ consumption gathering
makes urban space partitioning easier. Additionally, the com-
munity residents think the phenomenon has brought changes on
its cultural basis, which has reached a statistically significance,
thus, it’s an effective forecasting variable, β = .276, (t = 4.960,
p < .001). Because β is positive, it indicates that the community
residents think the phenomenon has brought changes on its
cultural basis, at the same time, it also makes urban space parti-
tioning easier. The community residents who think the pheno-
menon has brought changes on its activity events is β = .344, (t
= 5.168, p < .001). Since β is positive, it indicates that commu-
nity residents think things that make changes on activity events,
at the same time, also makes urban space partitioning easier. As
a result we can see from community resident’s attitude reflec-
tions, migrant workers’ cultural impacts and the impacts on
activities, as for urban space partitioning, it’s an effective fore-
casting variable, so the findings are noteworthy.
At the same time, the standardized regression equation is as
following:
Y = .001 Gender + .121 Age + (−.067) Educational Level +
(−.016) Religion (Catholic) + (−.079) Religion (Christian)
+ .023 Religion (Buddhism) + .023 Religion (Taoism) + (−.033)
Religion (Islam) + (−.076) Occupations (agriculture, forestry,
fishery, animal husbandry) + (−.162) Occupations (industries) +
(−.020) Occupations (low-level of white-collar office labors
and low-level military personnel) + (−.097) Occupations (high-
level of white-collar labors, middle class, high-end military
personnel) + (−.181) Occupations (Bourgeois, freelancers)
+ .076 Occupations (big capitalists) + (−.033) Occupations
(homemaker) + (−.125) Occupations (none) + (−.091) Marriage
(Unmarried) + (−.172) Marriage (Married) + .082 Monthly
Salary + (−.088) Family Structure + (−.056) With or Without
Migrant Maids + .010 Activity Time + .036 Contact Time
+ .037 X1 + .276 X2 (cultural changes) + .033 X3 + .11 X4
+ .344 X5 (changes on activities).
Tainan Ci t y , K a oh siung City and Taichung City’s
Influencing Factor’s Comparison and E xplanati on
As we can see from Table 4, in the resident’s point of view,

K.-H. CHEN
21
Table 3.
Regression coefficient table.
Predictor Standardized Regression
Coefficient β t Significance
Gender .001 .024 .981
Age .121 1.600 .111
Education al Level −.067 −1.182 .238
Religion (Catholic) −.016 −.347 .729
Religion (Christian) −.079 −1.517 .130
Religion (Buddhism) .023 .391 .696
Religion (Taoism) .023 .417 .677
Religion (Islam) −.033 −.712 .477
Occupations
(agriculture, forestry,
fishery, a ni mal
husbandry)
−.076 −1.613 .108
Occupations (industries) −.162 −3.147 .002**
Occupations
(low-level of white-collar
office labors and
low-level military
personnel)
−.020 −.357 .721
Occupations
(high-level of
white-collar labors,
middle class, high-end
military personnel )
−.097 −1.608 .109
Occupations
(Bourgeois, freelancers) −.181 −2.939 .004**
Occupations
(big capitalists) .076 1.561 .120
Occupations
(homemaker) −.033 −.564 .573
Occupations (none) −.125 −2.234 .026*
Marriage (unmarried) −.091 −.648 .517
Marriage (married −.172 −1.298 .195
Personal Month ly Sal ar y .082 1.504 .134
Family Structure −.088 −1.747 .082
With or Without
Migrant Maids −.056 −1.187 .236
Activity time .010 .208 .835
Contact time .036 .744 .457
X1’s changes
on population .037 .752 .453
X2’s changes on culture .276 4.960 .000***
X3’s changes on
spatial function .033 .583 .560
X4’s changes on
socioecon omic status .011 .164 .870
X5’s changes
on activities .344 5.168 .000***
*p < .05, **p < .01, ***p < .001, Y: The dependent variable.
the influencing factors between Southeast Asian migrant work-
ers’ gathering in the business district and urban cities differ.
Furthermore we can see that Tainan City, Kaohsiung City and
Taichung City all have obvious differences on the impact va-
riables. The reasons are as mentioned above in the first para-
graph (Section 5. 1).
The Results of the Study’s Dialogue with Existing
Theoretical Research
The survey’s responder’s are Filipino and Indonesian tran-
sient migrant workers. It’s quite different from Peter Marcuse’s
(2002) study which focuses on how the “Ghetto” area in the US
formed “divided cities”; and Ronald van Kempen (2002) in the
Netherlands, using “segregation index, SI” “to prove so-called
partitioned cities” have become more obvious with time and
have developed rapidly. Comparing it with our research method,
it is completely different. Even though this survey didn’t use
the “segregation index, SI” to prove the degree of urban space
partitioning in Tainan City, Kaohsiung City and Taichung City,
it still has similarities with Brenda Yeoh and Shirlena Huang’s
(1998) studies about the subjects, which are strategies and types
thereof of housework female migrant worker in public spaces in
Singapore and it is found that the phenomena of “divided cities”
−especially for the gathering and all kinds of consumption ac-
tivities in shopping malls like “Lucky Plaza”. As for some Sin-
gaporeans, “Lucky Plaza” has been a “divided city”, changing
from an “other type of space” to a “hesitant space”. I, as the
author of this paper, speculate that, when more and more mi-
grant workers gather in the research areas in the future, and
community residents have lower links of satisfaction on cultural
and spatial influences, then the citizens of these three cities in
Taiwan will also view those areas as “divided cities”. Thus, it
proves Ronald van Kempen’s (2002) studies’ notion on “social
polarization and urban space partitioning” is correct, including
1) The growing influence of ethnic groups, race and migration,
2) Exclusive of race and class. These two factors can fully de-
scribe the “divided cities” phenomena in these three cities.
Basically, the influencing factors of these three cities’
Southeast migrant workers’ gathering to urban space partition-
ing, have some factors in common as in the Netherlands and
Singapore, which include: The Netherland’s economics, popu-
lation and changes in the socio-cultural factors; Singaporean
government’s laissez-faire attitude as the role of the govern-
ment, the tolerance of the community residents and shops, but
however, it doesn’t have common factors as in Guangzhou,
China. If we try to interpret the “de-territorialization” and “re-
territorialization” points of views to describe the relation be-
tween transnational migration and local spaces, according to the
survey’s results, the same thing happened which was written in
Wu Bi-nar’s (2003) thesis. In Chung Shan’s research results—
displaced Filipino migrant workers used the entering of Chris-
topher church and its surrounding shops to re-territorialize the
space. And, as in Pei-Chia Lan’s paper (2002), we found out
Filipina housework migrant workers perform multiple roles and
shifting identities through the front/backstage segregation in
social space (working days vs. rest day; home country vs. host
country), isolating labor and consumption activities, and the
differentiation of temporal horizons (now vs. future) in order to
build floating identification. Based on the results of these two
research papers, the research of this paper added two more
viewpoints: under the concepts of “transnational space” or
“space partitioning”, we focused on the attitudes of local com-
munity residents toward the impact of immigration of different

K.-H. CHEN
22
Table 4.
The attitudes of community residents in Tainan City, Kaohsiung City and Taichung City toward migrant workers’ gathering in the business districts to
urban space partitioning and comparison am ong the influencing factors.
Tainan Cit y Kaohsiung Ci ty Taichun g City
Impact variables 1) Short residence time
2) First level of occupational background
1) Non-Muslim
2) Non-divorced
3) Attitudes of residents toward
migrant wor kers
1) Non-industrial labors
2) Non-Bourgeois
3) Non-freelancers
4) Labors
5) Thinking changes of culture exist.
6) Thinking changes of activities e xi st .
cultural groups, in order to describe the “degree of urban space
partitioning” better, and at the same time, observing from the
local residents points of view, we can explore the influencing
factors between Southeast Asian migrant workers’ gathering to
urban space partitioning.
Psychological Explanations
Represent ational Space and Third Space Theory
We also can interpret these consumption patterns and the
partitioning of urban space derived from migrant workers’ ga-
thering in urban commercial areas from the viewpoint of “re-
presentational space of migrant workers in other homeland
(Taiwan)”. For migrant workers, Taiwan acts as a relative loca-
tion, not only in a geographic space of map but also in a social
space. Taiwan has been a subject as compared to “other” (mi-
grant workers). In other words, dining room acts as a space “the
family take a dinner together”—that is “representation of space”
called by Lefebvre also is “second space” called by Soja. This
space is not existent for migrant workers. Therefore, “represen-
tational space” acts as a bridge linking social life with art and
imaginary, has been a struggling space to strive for freedom
and disarm. Soja think that “third space” by Lefebvre (1991)
has strengthened “governance”, “obedience” and “resistance”
and he call it “thirds pace” (1996). This is reason why migrant
workers usually favored excursion, shopping out and dining out
with their friends on weekend. Because they can compete for
dignity and human rights via “third space”.
Migrant Network
The result of this research has to be a dialog with the con-
sumption theory, has to extend to all consumption theories, then
attempt to create a new consumption theory. From the expe-
rience of this research we can discover, through Philippine,
Thai, Indonesian, Vietnamese nationality labourers in Tainan
county, Tainan city, Kao-hsiung county, Kao-hsiung city com-
mercial district or shopping street’s consumption patterns creat-
ed innovative consumption theory, has to excede Veblen or
Simmel’s classificational determining consumption viewpoint,
neither using Bourdieu stated habitual (habitus) or distinction
causing consumption. Rather using Massey’s “migrant network”
(1996) stand, or Erickson (1996) proposed “network variety”,
“cultural variety”. To create a migrant network, consumption is
a very important system, through different consumption events,
the whole network becomes operated, in the whole migration
process framework, “class” will influence “consumption” and
“network”, in the meantime consumption and network also mu-
tually influence each other, the relationship between the three is
just like above.
Just like this research discovers, the four countries labourers
are looking forward to have cell-phone, because they are yearn-
ing for “language intimacy”, a familiar language can have a
kind of emotionally relieving effect, also can be a helping me-
dia in the migrant network to individually get acquainted with
Taiwan’s unfamiliar environment. In the aspect of “Consump-
tion in a specific store”, except for the Vietnamese, the Philip-
pine, the Thai and the Indonesian labourers all generally have
their specific consumptional behaviour. The Philippine labour-
ers prefer to consume in [shopping mall], [sport-dress store],
[department store], the Thai labourers usually like to choose
[Thailand store], [supermarket], [communication/cell phone
store], and the Indonesian labourers also usually like to choose
[Indonesian store]. There have been influenced by their mother
nations’ socio-economic and cultural context, traditional con-
sumer s’ habits, consumption ability and unique lifestyle and so
on. For example, under Western/colonial education and consu-
merism, the Philippine labourers are used to patronize mod-
ern/western-style commercial facilities. Beca use the national ity
of collective life and dinning-in-home, the Thai and the Indo-
nesian labourers are used to patronize the stores which selling
homeland commodity. If we explain the differences between
the Philippine, the Thai and the Indonesian labourers and the
Vietnamese, we can say that when compared to the Vietnamese,
the former three have needs to establish their social network
and have adequate migrant network. From the interviewee’s
answer, we know the Vietnamese have not strong need to meet
and chat with their friends. And they consider few Vietnamese
store are less than Thai or Indonesian store on shopping streets
in Taiwan. Featherstone (1991) had the same point of view,
thinking in the modern world class-based consumption’s influ-
ence gets less and less determining, there are more and more
facts to confer consumption behaviour, including gender, race,
religion, country, migrant network, human relationships, ethnic
identity, etc.
Limitations of This Research
Although we are going to trace back the research to the ini-
tial theoretical framework: “The era of globalization and eco-
nomic dynamism, has created a large-scale flow of migrant
workers, it has made a great impact on the urban space in the
residential countries—transnational space and partitioning
space”. By thinking of this theoretical framework, We found
out, that the result of this research not only can provide expla-
natory arguments in the theoretical framework of “global in-
dustrial movements—migrant workers gathering and preference
choice of space-cross-cultural diversity—the transformation of
local consumption spaces—the segmentation of the functions of
urban space—the restructuring process of local space”, but it
can also be used as a foothold in order to develop more values

K.-H. CHEN
23
of academic research. Especially the induced outcomes of di-
vided space in the three city’s spatial structure, not be consi-
dered in this paper. This is the limitation of the paper.
Conclusion
The following points are the outcomes of this paper:
1) The activity spaces of local residents show that a signifi-
cant segregation of migrant workers activity spaces, the so-call-
ed “independent, divided space” has obviously taken shape.
The result is a common phenomenon of Tainan City, Kaohsi-
ung City and Taichung City. However, Taichung City’s propor-
tion of this has a higher percentage than that in Tainan City and
Kaohsiung City, which is a major discovery of this paper.
2) In Tainan City and Kaohsiung City, community residents
have more tolerant attitudes towards the negative impacts of
migrant workers gathering in the shopping streets. Comparing
the community residents’ attitude in Taichung City with Tainan
City and Kaohsiung City, their attitudes to migrant workers’
gathering are relatively unfriendly. They seem to have a more
intense opinion in this issue, and thus they are l ess friendly and
hospitable.
3) The influencing factors’ comparison of Southeast Asian
migrant workers’ gathering in the business districts to urban
space partitioning, in regard to community residents, Tainan
City, Kaohsiung City and Taichung City all have significant
differences. They’re all effective forecasting variables which
are worthy of attention. Thus we can make the important con-
clusions that: in community residents’ points of view, we can
look for residents’ opinions from different socio-economic back-
grounds, and their attitude toward migrant workers’ gathering
in the business district to find the influencing factors of South-
east Asian migrant workers’ gathering to urban space partition-
ing. Basically, we can find that these influential factors among
the paper are the same as the Netherlands and Singapore, and
not the same as Guangzhou, China.
Acknowledgements
1) The author would like to acknowledge the support of the
National Science Council（NSC）of TAIWAN in funding the
research on which this paper is based (Project No.: NSC96-
2415-H-024-001-SS2, NSC98-2410-H-024-013-SSS, NSC99-
2410-H-024-029-SSS).
2) This paper has been published in The 2012 International
Conference on Asia-Pacific Studies and revised according to
the reviews of the consultant.
REFERENCES
Bureau of Employment and Vocational Training of Council of Labor
Affairs (2011). Foreign labors statistics.
http://labor.kcg.gov.tw/labor.htm
Chen Kung-Hung (2008). The Formation and Meanings of Southeast
Asian Migrant Workers’ Ethno-Consumptive Space on Tainan City/
County and Kao-hsiung City/County in Taiwan. Proceedings of In-
ternational Symposium on City Planning 2008, Chonbuk National
University.
Chen Kung-Hung (2011a). Study on south-eastern Asian migrant wor-
kers’ gathering in urban commercial area and its relations between
the partitioning of urban space in Kaohsiung City. Journal of Archi-
tecture and Pl anning, 12, 47-74.
Chen Kung-Hung (2011b). Study on Indonesian migrant workers’ ga-
thering and consumptive activities in urban commercial area and its
impacts on urban community spaces in Tainan City. Asia-Pacific Fo-
rum, 51, 66-93.
Chen Kung-Hung (2011c). Community residents’ attitudes to the im-
pact of migrant workers’ gathering and its relations between the par-
titioning of urban space—Taichung City Commercial Area Case.
2011 Conference of China Geographical Society, National Changhua
University of Education, Taiwan.
Chou Cheng-Hung (2003). Migrant labors: A catalyst for rebuilding ur-
ban space. Master Thesis, Graduate Institute of Architecture of Tung-
hai University.
Du, H., & Li, S. M. (2010). Migrants, urban villages, and community
sentiments: A case of Guangzhou, China. Asian Geographer, 27, 93-
108. http://dx.doi.org/10.1080/10225706.2010.9684155
Erickson, B. H. (1996). Culture, class and connections. The American
Journal of Sociology, 102, 217-250.
http://dx.doi.org/10.1086/230912
Featherstone, M. (1991). Consumer culture and postmodernism. Lon-
don: Sage.
Kempen, Ronald van (2002). Towards partitioned cities in the Nether-
lands? Changing patterns of segregation in a highly developed wel-
fare state. In P. Marcuse, & K. Ronald van (Eds.), Of states and cities:
The partitioning of urban space (Chapter 5, pp. 88-108). Oxford:
Oxford University Press.
Lan Pei-Chia (2002). A transnational top ography for the migration and
identification of Filipino migrant domestic workers. Taiwan: A Rad-
ical Quartwely in Social Studies, 48, 169-219.
Lefebvre, H. (1991). The production of space. Oxford: Blackwell.
Marcuse, P. (2002). The shifting meaning of the black Ghetto in the
United States. In P. Marcuse, & K. Ronald van (Eds.), Of states and
cities: The partitioning of urban space (Chapter 6, pp. 109-142). Ox-
ford: Oxford University Press.
Marcuse, P., & Ronald van, K. (2002). Of states and cities: The parti-
tioning of urban space. Oxford: Oxford University Press.
Massey, D. (1996). What drivin g Mexico-US migration? A theoretical,
empirical, and policy analysis. The American Journal of Sociology,
102, 939-999. http://dx.doi.org/10.1086/231037
Sassen, S. (1988). The mobility of labor and capital: A study in in terna-
tional investment and labor flow. Cambridge: Cambridge University
Press. http://dx.doi.org/10.1017/CBO9780511598296
Sassen, S. (1991). The global city: New York, Lond o n, To kyo. Princeton,
NJ: Princeton University Press.
Sassen, S. (1996). Losing control: Sovereignty in an age of globaliza-
tion. New York: Columbia University Press.
Sheu Horng-Yih (2000). The use of space and its effect in a Filipino
labor gathering area—Christopher Church Area on Chung-Shane
North Road. Master Thesis, Graduate Institute of Architecture of
Tamkang University.
Soja, E. (1996). Thirdspace: Journeys to Los Angeles and other real-
and-imagined places. Oxford: Blackwell.
Wang Chih-Hung (2006). Dis/placed identification and politics of space:
The consumptive ethnoscape around Taoyuan Railroad Station. Tai-
wan: A Radical Quart w ely in Social Studies, 61, 149-203.
Wu Bi-Nar (2003). Chung Shan—The formation of a Filipino migrant
workers’ community space in Taipei. Master Thesis, Institute of Buil-
ding and Planning of National Taiwan University.
Yeoh, B. S. A., & Huang, S. (1998). Negotiating public space: Strate-
gies and styles of migrant female domestic workers in Singapore.
Urban Studies, 35, 583-602.
http://dx.doi.org/10.1080/0042098984925