j.1435-5957.2010.00343.x the Impact of Urban Growth on Commuting Patterns in a Restructuring City_Evidence From Beijing

8/11/2019 j.1435-5957.2010.00343.x the Impact of Urban Growth on Commuting Patterns in a Restructuring City_Evidence F

1/21

The impact of urban growth on commuting patterns in a

restructuring city: Evidence from Beijing*

Pengjun Zhao1, Bin Lu1, Gert de Roo2

1 Department of Urban and Regional Planning, College of Urban and Environmental Sciences, Peking University,

5 Yiheyuan Road, Beijing, 100871, China (e-mail: [email protected], [email protected])2 Department of Urban and Environmental Planning, University of Groningen, Landleven 1, NL-9742 AD

Groningen, The Netherlands (e-mail: [email protected])

Received: 21 June 2009 / Accepted: 24 November 2010

Abstract. Our existing knowledge of the links between urban growth and commuting patterns

are dominated by cases from developed countries. This paper examines the impact of urban

growth on workers commutes using the case of Beijing, which is undergoing rapid economic

and spatial restructuring. The results of an analysis of household survey data show that clustered

and compact urban development in planned sub-centres is likely to reduce suburban workers

need for a long-distance commute to the city centre when the workers socio-economic char-

acteristics, the level of transport accessibility and household preferences for residential location

are taken into account. Workers employed in the manufacturing sector tend to have shorter

commutes and travel within the planned suburban sub-centres. This reveals that the decentrali-

zation of employment in the manufacturing sector provides more opportunities to enhance the

spatial matches between household residential and job location choices. Household preferences

for residential location have an effect on commuting patterns, and high-income workers are

likely to accept longer commutes in order to fulfil their residential preferences. Dramatic urban

restructuring, in conjunction with changes in lifestyle, is creating new commuting patterns in the

rapidly growing cities of China.

JEL classification: R14, R23

Key words: Urban growth, commuting, employment decentralization, Beijing

1 Introduction

In recent decades, one of the main arguments in relation to the reduction of energy consumption

and greenhouse gas emissions has focused on passenger transport, as this accounts for over 20

* The travel data used in this study came from a housing survey undertaken in 2006 by Professor Li Si-Ming fromHong Kong Baptist University and Professor John R. Logan from Brown University. The authors would like to thankthem for their data support.

doi:10.1111/j.1435-5957.2010.00343.x

2011 the author(s). Papers in Regional Science 2011 RSAI. Published by Blackwell Publishing, 9600 Garsington Road,

Oxford OX4 2DQ, UK and 350 Main Street, Malden MA 02148, USA.

Papers in Regional Science, Volume 90 Number 4 November 2011.


2/21

percent of the worlds primary energy use and 13 percent of energy-related CO2emissions (IEA

2006). Of the many factors influencing travel patterns at a city level, urban spatial change is

particularly important. Spatial changes affect travel features through patterns of urban growth

(see Crane 2000 for a literature review). In particular, employment decentralization and clus-

tered development in the suburbs are often believed to be associated with shorter commutes than

in areas of sprawling development and even monocentric spatial structures (Gordon et al. 1989b;Levinson and Kumar 1994). One major reason for this, according to economics research, is that

polycentric spatial structures provide more opportunities to enhance spatial matches between the

residential and job location choices of households living in the suburbs (White 1988; Yinger

1992).

Urban spatial change has been accompanied by changes in individuals housing and employ-

ment locations, which in turn causes changes in commuting patterns, such as commuting

distance and features of the commuting flow (Anas et al. 1998). Since the Second World War, the

pattern of urban growth in Western countries has been dominated by low density development

and employment decentralization, which was combined with a rapid increase in car ownership

and use (Glaeser and Kahn 2001; Knight and Stanback 1976). In particular, after the 1980s, newsuburbanization occurred in North American and European countries (Garreau 1991; Stanback

1991; Panebiaco and Kiehl 2003), with several common trends: the decentralization of employ-

ment in the service, finance, insurance and real estate sectors, after the decentralization of

employment in the manufacturing sector, occurs dramatically in the suburbs and forms new

sub-centres; a daily commute at city level is dominated by suburb-to-suburb flow and reverse

commuting (centre-to-suburb commuting) grows steadily; most high-income workers choose to

live and work in the suburbs, although some decide to reside in the city centre in order to have

access to amenities even if their jobs are located in the suburbs; spatial mismatch is often found

in the city centre with greater urban sprawl believed to be associated with higher spatial

mismatch.Due to these new trends in suburbanization and their influence on commuting, many

initiatives in relation to the management of patterns of urban growth have been introduced in

Western countries in order to mitigate the environmental effects of commuting, such as con-

tainment strategies, smart growth, compact city, new urbanism, transit-oriented development

(TOD). A recent report, Growing Cooler, based on a summary of a large body of literature,

concludes that it is realistic to assume a 30 percent cut in VMT [through] compact develop-

ment (Ewing et al. 2008, p. 9). However, current policies designed to manage urban growth on

the fringes of the city are often criticized for their limitations in terms of altering the course of

commuting patterns (Gordon and Richardson 1997; Giuliano 1999; Rodriguez et al. 2006). One

major reason for this is that there is still considerable uncertainty about the links betweenpatterns of urban growth and the travel behaviour of individuals.

In addition, our existing knowledge about the links between urban growth and commuting

is largely based on cases from North American and European countries. Cases from developing

countries are scarce. Research into the dynamics of urban growth in rapidly growing cities in

developing countries would enhance and expand our knowledge of the relationship between

urban growth and commuting. In the case of China, since the 1980s its large cities have been

undergoing rapid urban spatial change, which has been so dramatic that it is often called

restructuring (Ma 2004; Ma and Wu 2005; Lu and McCarthy 2008; Zacharias and Tang 2010).

Several main new trends in urban development can be described as follows: rapid expansion of

the urban space as a consequence of housing and industrial development on the fringes of thecity; suburbanization caused by the outward movement of the urban population and industries

from the city centre; and polynucleation in large cities, resulting from the growth of sub-centres

and industrial zones in the suburbs. Consequently, new relationships between jobs and housing

and new commuting patterns have developed in the suburbs of Chinas large cities during the

736 P. Zhao et al.



3/21

urban restructuring process (Gaubatz 1999a; Zhao et al. 2009c; Zhao et al. 2011). In particular,

a great deal of the new urban development on the fringes of the city has diverged quite

dramatically from the traditional compact urban form. Urban sprawl has already been identified

as one of the major traits of suburban development (Deng and Huang 2004; Wong and Tang

2005). Sprawling development, combined with an increase in car ownership, tends to worsen the

environmental effects of urban growth by promoting the need for long-distance commuting(Shen 1997; Wu 2002). These new trends in urban spatial change and their effects on travel

patterns in Chinas cities are likely to develop in a similar way to Western countries in the 1950s.

However, in the restructuring of Chinas cities, the relationship between urban growth and

commuting is more complex than in other countries (Zhao and Lu 2010). First, urban growth in

China has been fundamentally driven by the transformation from a centrally planned system to

a market system. The country has witnessed dramatic political decentralization, globalization

and marketization since the 1980s (Wei 2001). Market-oriented real estate and commercial

development, following government-led industrial development, is shaping the suburbanization

of China (Zhao et al. 2009a). Compared with employment decentralization in Western countries,

Chinas current suburbanization is characterized by residential decentralization due to old cityregeneration and the development of housing in the suburbs (Zhou 1997). Employment decen-

tralization develops slowly and is dominated by the manufacturing and construction sectors,

while the finance, insurance and real estate sectors are still concentrated and growing rapidly in

the central areas of cities. Second, compared to Western cities, most high-income households in

China are likely to choose to live in central areas of the city, where jobs for those who are highly

qualified and high-price housing are located (Zheng et al. 2005). Traditional cultural factors are

an important reason for this, as well as the high level of accessibility to urban services and

amenities in central areas. For example, living in the city centre, where the privilegentsia

traditionally lived, usually indicates higher social status in China (Gaubatz 1999b). While some

high-income households move to the suburbs because of changes in their lifestyle (Wang and Li2004), most low-income workers reside in the suburbs because they cannot afford the high-price

housing in the central urban areas, even if their job is located in the city centre. That means that

employment decentralization and urban sprawl would bring benefits to low-income households.

Third, in terms of policy-making, the supply of various types of housing and residential land use

in China is still more centrally planned and tightly regulated in the present reform era than in

Western countries (Zhao et al. 2009b). As a result, the role of residential preference in the actual

choice of residential location may be weaker than in Western countries where there is a higher

degree of market-oriented housing.

Therefore, an investigation into urban growth in China is of great importance to freshen and

enhance our understanding of the impact it may have on an individuals travel behaviour.However, relatively few studies have examined Chinas situation. By looking at the case of

Beijing, this paper examines the relationships between patterns of urban growth and commuting

in Chinas cities. Two research questions will be answered. One is how patterns of urban growth

affect a worker commuting on the fringes of the city of Beijing. The other is whether existing

regional policies designed to form a polycentric spatial structure in order to reduce crowding in

the city centre also have implications for the reduction of long-distance suburb-to-centre com-

muting. Two aspects of commuting patterns are studied: commuting destination choice and

commuting trip length. This paper also takes into account the individual workers socio-

economic features, household characteristics and household preference for residential location.

The remainder of this paper is organized as follows: Section 2 introduces the theory andreviews the existing literature concerning the relationship between urban growth and commut-

ing; Section 3 presents an analysis of the city context of Beijing; Section 4 presents the data and

methodology used in this study; Section 5 explores the impact of patterns of urban growth on

individual workers commuting features and Section 6 presents the conclusions reached.

737The impact of urban growth on commuting patterns



4/21

2 The relationship between urban growth and commuting patterns

2.1 Theory

Search theory has been widely used to explain individual travel behaviour (see reviews by

McFadden 2001; Waddell 2001). According to search theory, a workers choice of residentialand job location determines their commuting distance, and a worker will make a location choice

which has a maximum degree of utility. A workers commute length is therefore understood as

a function of the utility of their home and job location choice. In a monocentric city model, all

employment is located in the central business district (CBD), and workers are often only faced

with a choice in terms of housing location (Alonso 1964; Muth 1969). However, in a polycentric

city model, employment is located throughout the city, for example, in the CBD, at sub-centres

and in rural areas; workers are facing not only a set of alternative housing locations but also

employment locations (White 1988). A workers commuting distance is determined by the

combination of housing and workplace location choice (Anas et al. 1998; Van Ommeren et al.

2000). The utility of a workers choice of location is constrained by various factors such as theworkers personal characteristics, household characteristics, housing and job features as well as

aspects of the zone in which the worker resides and works (Richardson 1977). A workers

commuting distance in a polycentric city model can be expressed as:

D f U P F H L W Sij i i i i j j= ( )[ ], , , , , , (1)

where,Dijis the workers commuting distance between residential locationiand workplacej,Piis a vector representing the workers personal characteristics, such as age and gender, Fi is a

vector in which the workers family characteristics, such as income and the number of employed

family members, are represented, Hiis a vector that shows the features of the housing, such assize and price,Liis a vector that represents the characteristics of the location in which the worker

resides, such as neighbourhood land use and transport services, Wjis a vector that represents the

features of the location in which the worker works, such as proximity to the CBD or sub-centres,

land use pattern and transport accessibility,Sjis a vector that indicates the characteristics of the

job, such as employment sector, and eis random utility. A workers commuting distance is also

subject to household budget constraints and his/her time constraints.

However, traditional explanations from the above model usually ignore the influence of

household attitudes to location on the choice of residential location. Location attitudes, meaning

household preferences for certain attributes of housing, location and neighbourhood, may affect

individuals residential location and thus commuting patterns (Handy et al. 2005; Scheiner andHolz-Rau 2007). When the variable Ni, representing a households preferences for residential

location is added, Equation (1) can be rewritten as:

D g U P F H L N W Sij i i i i i j j= ( )[ ], , , , , , , (2)

A hypothesis can be derived from the theory above: for workers residing on the fringes of the

city, a polycentric model would entail a shorter commute than a monocentric model. The main

reason for this is that the polycentric model permits workers to choose between several job

locations within a given metropolitan area, allowing greater average proximity between home

and work (White 1988; Yinger 1992).A workers commute length can be indicated by the degree of spatial separation between the

location of the commuting destination (workplace) and the location of its origin (housing

location). Therefore, a discrete variable can be used to reflect a workers choice of commute

length and direction if the workers housing location is fixed.

738 P. Zhao et al.



5/21

Equation (2) can be rewritten as:

C k V P F H L N W S j i i i i i j j= ( )[ ], , , , , , , (3)

whereCjis the workers job location. For those who live on the fringes of the city, when the city

centre is studied as a commuting destination, the discrete variable reflects a long-distancecommute from suburb to centre. When the sub-centre is studied as a commuting destination, the

discrete variables can indicate the suburb-to-suburb commute. This discrete method has the

advantage of simultaneously revealing both the commuting trip length and commuting flow at

city level, compared with the traditional method of measuring distance with a continuous value.

2.2 Literature review of previous empirical studies

Many empirical studies have provided evidence to support the hypothesis introduced above, that

is, in the process of employment decentralization, polycentric development can reduce theamount of suburb-to-centre commutes and overall commuting distances. For example, Gordon

et al. (1989a) argued that decentralized employment centres tend to shorten the overall length of

commutes. A similar finding was reported by Cervero and Wu (1997), who found that employees

who live in suburban centres in the San Francisco Bay area have shorter commuting trips than

those who reside in larger, denser urban centres. Song (1992) found that compared to mono-

centric and omnicentric urban spatial structures, the polycentric spatial pattern leads to less

commuting and shorter commuting distances. The results can be largely explained by the

co-location mechanism, according to which, employment and population follow each other in

the process of suburbanization, decentralizing into sub-centres to optimize commuting costs.

Consequently, this mutual relocation enhances the jobs-housing balance in the suburbs,thereby reducing commuting distance and the need to commute to the city centre (Dubin 1991;

Gordon et al. 1991; Downs 1992; Kim 2008).

However, most of this research ignores the differences in commuting patterns between

economic sectors. A few recent studies provide evidence for this, for example, Crane and

Chatman (2003), by using American housing survey data, found that the average commute

length became shorter as employment decentralized to the suburbs. In particular, they found that

there are differences in terms of industry, for example, the decentralization of service, wholesale

and construction employment is associated with shorter commutes, while finance, insurance and

real estate, manufacturing and government employment decentralization is associated with

longer commutes. The main reason is that service, wholesale and construction employment issubstantially more clustered in suburbs. Lee et al. (2006) found that the average commuting time

did not increase considerably in twelve of the largest metropolitan areas of the US in the 1990s

when they witnessed a dramatic decentralization of employment. Travel times were shortest in

the retail sector, followed by services, as employment in retail and consumer services is largely

local. The results imply that the extent to which employment decentralization can affect

workers commuting patterns largely depends on the type of decentralized employment. The

relationship between various types of employment and housing in suburbs is a precondition for

a spatial match between an individuals housing and job location.

As well as the jobs-housing balance, land use at a local level also has an effect on workers

commutes. Sprawling development on the fringes of the city is often criticized for giving rise tohigher trip distances and promoting peoples dependency on cars (for example, Crane 2000;

Camagni et al. 2002; Travisi et al. 2006). Compact urban development on the urban fringe,

which is characterized by high levels of density and self-contained development, would reduce

the trip distance to work and the need for driving (for example, Cervero 1995; Kenworthy and




6/21

Laube 1999; Schwanen et al. 2004; Boussauw et al. 2010). A recent study by Moilanen (2010)

shows that denser settlements have a greater capacity to meet the needs of job seekers in

Norway. This means that high-density development in the suburbs would reduce the need for

long-distance commuting.

It is undeniable that a few studies have presented contrary findings. For example, Cervero

and Wu (1998) found that commuting distances in the San Francisco Bay area had increasedrapidly since 1980, when it had been assumed that employment decentralization would increase

the jobs-housing balance in the suburbs. Baccaini (1997) showed that employment decentrali-

zation was not associated with shorter average commuting distances in Paris. Some research has

concluded that monocentric metropolitan areas have shorter commuting distances than poly-

centric metropolitan areas (Schwanen et al. 2001; Van de Coevering and Schwanen 2006).

Therefore, Giuliano and Small (1993) argued that other factors must be more important to

individuals location decisions and thus commuting patterns.

In particular, some literature has reported that the residential self-selection factor, which

concerns a households attitudes to housing location, can explain a large portion of the variations

in workers travel behaviour (see review by Bohte et al. 2009; Cao et al. 2009). For example,Schwanen and Mokhtarian (2005) found that workers who prefer to live in suburban areas were

substantially more likely to commute over longer distances and by car than those living in other

urban regions. Kitamura et al. (1997) reported that residents attitudes to their residential

environment was more important in explaining the variation in travel behaviour. These results

imply that differences in commuting patterns at the city level may be more a matter of residential

location choice than a travel choice. Undoubtedly, when considering self-selection, land use

patterns must not be overlooked, as many previous studies have concluded that land use patterns

still have a significant impact on commuting distances and mode choice even when residential

self-selection is taken into account (Pinjari et al. 2007; Circella et al. 2008).

A few empirical studies have been undertaken in order to investigate Chinas situation. Forexample, Yang (2006) found that residential relocation to the citys periphery, which has a

low-level local jobs-housing balance, increased commuting times in Beijing. Wang and Chai

(2009) found that a decrease in the jobs-housing balance caused by the dismantling of the

danwei (work unit) system increased the demand for longer travel. A recent study by Pan et al.

(2009) revealed that the demand for longer journeys and greater car usage was caused by new

forms of land development at the neighbourhood level those non-pedestrian/non-cyclist-

friendly urban forms of development which emerged after the 1990s. At least two shortcomings

of these previous studies can be identified: the lack of an in-depth analysis of the impacts of

urban growth on commuting at city level and the little attention paid to individuals residential

preferences.

3 The context of Beijing

The city of Beijing has existed for more than 3,000 years and is now Chinas capital. Beijing has

a land area of 16,410 square kilometres and its total population was 16.95 million in 2008 (BSB

[Beijing Statistic Bureau], various years). The city is divided into four zones the central urban

areas, the inter-suburban areas, the outer suburban areas and the ecological conservation

areas (see Figure 1). This study focuses on the central urban areas and inter-suburban areas

because in 2008 these two zones had 69 percent of the total number of households and 74percent of the total number of employees in Beijing (BSB various years). Furthermore, it is

mainly in the inter-suburban areas that urban expansion has occurred in Beijing since the 1980s.

Beijing has been undergoing rapid suburbanization since the 1980s, stimulated by urban

expansion. Between 1982 and 1990, the population in the inter-suburban area grew, while the

740 P. Zhao et al.



7/21

old city centre experienced a significant decline in population (Zhou 1997). This suburbaniza-

tion of the population, as well as employment, continued after 1990 (BSB various years). The

economic census data shows that employment and population levels increased rapidly in the

inter-suburban areas, while the city centre experienced a continuous decline in population and

a rapid decline in employment after 2004 (see Figure 2). In particular, 10 planned peripheral

constellations in the inter-suburb areas started to shape Beijing into a polycentric urban form

(Ma 2004). Figure 2 also shows that employment in both the outer suburban areas and ecologi-

cal conservation areas decreased after 2004. This means that the urban growth of Beijing tendsto be concentrated in the inter-suburban areas. However, urban sprawl still occurred across the

suburbs of Beijing (Deng and Huang 2004; Jiang et al. 2007; Zhao et al. 2009b). Urban sprawl

is characterized by low density levels, dispersed and scattered physical patterns, a low degree of

local mixed land use and, most importantly, a low jobs-housing balance.

Fig. 1. Beijing urban spatial structure and the areas studied in this paperSource: the authors, edited from BIUPD (1993).




8/21

Spatial restructuring in Beijing has been accompanied by economic restructuring. Figure 3

shows that employment in the manufacturing sector decreased greatly between 1991 and 2008

in all areas except the ecological conservation areas. This resulted in an overall decrease in

manufacturing employment as the economy became increasingly service-sector based.

However, in the outer suburban areas, manufacturing still accounted for 40 percent of total

employment. Construction employment declined dramatically in all areas, although there was

some slight growth in 2004 in the outer suburban and ecological conservation areas. In the

central urban areas, employment in the finance, insurance, real estate and commercial services

sectors increased rapidly. This has been attributed to dramatic development in the old city centre,which caused manufacturing factories to move out of the city centre and make space for the

high-revenue yielding service sector. Employment in the communication and logistical service

sectors grew rapidly in both the inter-suburban and outer suburban areas due to these areas

experiencing large increases in the number of residents and firms, which provided local cus-

tomers for the communication and logistical sectors. Figure 3 shows that employment in urban

services (research, education, sport and civil services), declined in both the central urban areas

and inter-suburban areas, while it increased slightly in the outer-suburban areas and ecological

conservation areas. This largely resulted in urban growth in the outer-suburban areas and

ecological conservation areas.

Engaged as it is in a rapid process of urban growth, Beijing is experiencing changes incommuting patterns which include, in particular, increases in long-distance commuting trips

from the citys fringes to the centre. Data from Beijings second (2000) and third (2005) travel

surveys reveal that the number of commuters who entered the central urban areas in rush hour

increased by 46.2 percent between 2000 and 2005 and had reached three times the number of

residents living in the centre by 2005 (BTC [Beijing Transport Committee] 2000; 2005). During

the same period, the travel distance per person increased on average from 8.0 kilometres to 9.3

kilometres. The lengthening of the commute between the city centre and the suburbs in Beijing

would have a sizeable impact on both the local and global environment, especially in the context

of rising levels of motorization. In Beijing, the number of private cars increased two fold in the

period between 2000 and 2005. The number had reached 3.25 million in 2008 and will continueto increase by 15 percent annually over the coming years. Transport-related air pollution makes

up 23 percent of all air pollution, followed closely by industrial pollution (Huang 2009).

Since the 1980s, the central government and municipal government have been making an

effort to create a polycentric spatial structure for Beijing in order to move residents from the

Fig. 2. Growth in employment and population in Beijing

Source: The authors, edited from data sets from BSB (1997; 2002; 2005; 2009).

742 P. Zhao et al.



9/21

crowded city centre to the suburbs (BIUPD 1982, 1993, 2004). According to Beijings urban

master plan, the central urban areas and inter-suburban areas comprise five different zones: the

old city centre, a mixed urban area, the greenbelt, peripheral constellations and the outskirts ofthe inter-suburban area (BIUPD 2004) (see Figure 1). The old city centre and mixed urban area

make up the central urban areas. Ten peripheral constellations are located outside the mixed

urban area and are connected to the old city centre by a radial road system. These peripheral

constellations were designed to be sub-centres, attracting new firms and residents from other

Fig. 3. Growth of employment in economic sectors in BeijingSource: The authors, edited from data sets from BSB (1997; 2002; 2005; 2009).




10/21

cities and encouraging existing factories and the population to move away from the city centre.

In these peripheral constellations, the plan encourages compact development, characterized by

self-contained, high-density and physically concentrated land use and the strict control of

sprawling development. The greenbelt is located between the mixed urban area and the periph-

eral industrial areas to act as a buffer, restricting further growth and development in the mixed

urban areas and curbing sprawling development. Outside the peripheral constellations, theoutskirts of the inter-suburban area are designed to serve agricultural needs and future devel-

opment. In this study, the fringes of the city of Beijing refers to the peripheral constellations

and the outskirts of the inter-suburban area.

It has been claimed that patterns of urban growth are one of the major factors influencing the

changes in commuting in Beijing, with others including rising incomes, skyrocketing levels of

motorization, a shortage of transport infrastructure, weak traffic management and poorly devel-

oped municipal, fiscal and regulatory institutions (Li 2006; Zhao et al. 2007). In particular,

urban sprawl on the fringes of the city is seen as one important factor leading to the growth in

long-distance commuting from the suburbs to the city centre and congestion in the city centre

(Zhao and Lu 2009; Zhao 2010). Therefore, the current policies designed to promote theformation of sub-centres and compact urban development in the suburbs should also play an

important role in reducing long-distance suburb-to-centre commuting. However, policy-makers

in Beijing often refute this point and current policies designed to address transportation prob-

lems frequently focus predominantly on huge and expensive transport infrastructures. One

major reason for this is that previous empirical studies provide scarce evidence of the impact of

urban growth on commuting in Chinas cities.

4 Data and variables

The commuting data used in this study are derived from a household interview survey conducted

in Beijing in 2006. Among other questions, employed respondents were asked to state the

location of their homes and workplaces as well as their preferences for residential location.

During the data-collection process, a multilevel probability proportional-to-size sampling strat-

egy (PPS) was used to provide a sample of all urban districts in the central and inter-suburban

areas of Beijing. A number of communities were selected from each district according to the

distribution of urban residents in Beijing, calculated on the basis of the Beijing Statistical

Yearbook2005 (BSB various years), with 60 communities being selected. A systematic sam-

pling approach was used to select households within the communities. In each of the 60

communities, 25 local households were selected. Face-to-face interviews were conducted, withthe survey successfully interviewing 1,500 residents. In the analysis within this paper, 370

employed respondents from the citys fringe were used. Of the 163 workers living in the

peripheral constellations, 104 commuted within the peripheral constellations and only 42

commuted to the central urban areas (Tables 1 and 2). Among the 208 workers who resided on

the outskirts of the inter-suburban areas, 131 commuted to the central urban areas. Generally, for

quantitative analysis, the more cases that are observed, the better the models goodness-of-fit. A

total of 370 samples is small compared to the large number of residents living on the fringes of

the city of Beijing. However, in terms of this statistical method, a mid-range number of observed

cases (N =100) is acceptable (Aldrich and Nelson 1984).

In this paper, the impact of patterns of urban growth on workers commuting patterns isexamined in a quantitative way. The variables indicating patterns of urban growth reflect

regional policies adopted in Beijing. Patterns of urban growth are measured by land use forms

at a local level and the location where a worker lives. According to the regional policies

introduced above, different policies in relation to patterns of urban growth are applied in the

744 P. Zhao et al.



11/21

planned peripheral constellations and the outskirts of the inter-suburban area. Therefore, this

study uses a location variable to indicate different spatial policies employed in Beijing. There

are many methods that can be used to measure land use forms (see, Ewing 1997; Galster et al.

2001; Song and Knaap 2004a, 2004b; Tsai 2005). In Beijing, regional policies promote compact

urban development in the suburbs, which is characterized by high-density levels, high levels of

jobs-housing balance and a high level of public transport services. This study measures land use

forms by the local jobs-housing balance, the net density of the population and transport

accessibility (see Table 3). In addition, the jobs-housing balance also reveals the patterns of job

distribution at a local level.The variables of jobs-housing balance and density are measured at a sub-district level. In

Beijing, the census data are collected at a sub-district level, at the level of the street (jiedao) or

rural town (xiangzhen). The sub-district is the basic administrative unit in Chinese cities.

Generally speaking, a precise definition or universal standard is lacking with respect to the term

Table 1. Data summary of workers residential location and commuting destination

Commuting destination Residential location in the city fringe

The peripheral

constellations

The outskirts of the

inter-suburban areas

Commuting to the central urban areas 42 131

Commuting to the peripheral constellations 104 49

Commuting to other areas in the city fringe 17 27

Total 163 207

Table 2. Land use pattern, socioeconomic and preferences variables analysed

Variables Classification Cases Percentage

Household

annualincome

Less than 20000 (RBM) (2920 USD) 88 23.8

Between 20000 and 50000 (RBM) (between 2920 and 7300 USD) 208 56.2Equal or above 50000 (RBM) (7300 USD) 74 20.0

Employment Employed in agriculture, geology, mining or construction sectors 36 9.7

Employed in manufacturing sector 95 25.7

Employed in real estate or finance sectors 33 8.9

Employed in wholesale, retail or other commercial sectors 102 27.6

Employed in national government, party or research & development

sectors

49 13.2

Employed in social services and other sectors 55 14.9

Gender Male worker 245 66.2

Female worker 125 33.8

Travel mode By car 68 18.4

By public transport (bus, metro) 115 31.1Other (walking, bicycling, other transport mode) 187 50.5

Family composition Two earners 269 72.7

Single earner 101 27.3

Housing

location

preferences

Prefer higher quality of community living environment (low-pricing,

large space housing, open space, etc.)

99 26.8

Prefer higher proximity to workplace 57 15.4

Prefer higher accessibility to transport 88 23.8

Prefer other amenities in the community in which they live (schools,

shops, etc.)

126 34.1

Total 370 100




12/21

jobs-housing balance. In Beijing, the average value of the jobs/housing ratio in 2006 was 1.45

in the central area of the city and 0.62 in the suburbs. In the 10 peripheral constellations, the

average value of the jobs/housing ratio ranged from 0.80 to 1.15. We used this range as an ideal

value for the jobs-housing balance because regional policies in Beijing are attempting to develop

these peripheral constellations as sub-centres and the land use patterns of these planned periph-

eral constellations are viewed as the optimal model, which should be promoted in other areas

of the suburbs.

Many studies have reported that individual and household socio-economic characteristics

can largely explain variations in commuting patterns (Hanson 1982; Hanson and Schwab 1995;Shen 2000; Dieleman et al. 2002). In this study, a method of multiple regression analysis is

employed to examine the effects of urban growth on the commuting pattern of individuals,

allowing for the socio-economic factors included in Equation (3). Four variables have been

selected to indicate individual and household socio-economic characteristics. They are: the

Table 3. Patterns of urban growth, socioeconomic and preferences variables analysed

Variable name Value and description

Patterns of urban growth

Jobs housing balance =1 if worker resides inside the sub-district where the value of jobs-housing

balance is between 0.85 and 1.15Density the net density of population in the sub-district where a worker lives (persons

per hectare)

Car transport accessibility =1/ the distance from the community where the household is located to the

nearest main road

Public transport accessibility =1/ the distance from the community where the household is located to the

nearest metro or bus station

Reside in the peripheral constellation =1 if worker lives in one of the peripheral constellations

Workers socioeconomic features

Primary and construction =1 if worker is employed in agriculture, geology, mining or construction

sectors

Manufacturing =1 if worker is employed in manufacturing sector

Advanced tertiary =1 if worker is employed in advanced tertiary sectors (e.g. real estate orfinancial sectors)

Commercial sectors =1 if worker is employed in wholesale, retail or other commercial sectors

Government =1 if worker is employed in national government, party or research &

development sectors

Social services reference variable for employment

Male =1 if worker is male

Travel by car =1 if worker travels by car

Travel by public transport =1 if worker travels by public transport

Households characteristics

Low-income household =1 if worker is from a household with annual income equal to or less than

20,000 (RBM)

Middle-income household =1 if worker is from a household with annual income between 20,000 and50,000 (RBM)

High-income household reference variable for income

Two-earners household =1 if household has two or more earners

Preferences for residential location

Prefer community environment =1 if households preference is a community with higher quality of

community living environment

Prefer job proximity =1 if households preferences is a community with higher proximity to

workplace

Prefer transport accessibility =1 if households preference is to live in a community with higher

accessibility to transport

746 P. Zhao et al.



13/21

annual household income, the number of earners, the workers employment sector and their sex.

A workers employment sector also reflects the features of his/her job. Household preferences

for location are also indicated and examined (see Table 3).

5 Analysis

The multinomial logit (MNL) model is applied in order to estimate workers travel distances in

Equation (3). Table 4 presents the results of the logistic regression analysis. The goodness-of-fit

of the model is measured by the minus two log-likelihood (-2LL), Pearsons index and the

percentage correctly predicted (PCP). Table 4 reveals that the model chosen has a satisfactory

fit: Pearsons index and -2LL with a chi-squared distribution both have a significance level of

p


14/21

commuting within the peripheral constellations is higher than that of commuting to the central

urban areas. For those who reside in a sub-district with a higher jobs-housing balance or

population density, there is a higher likelihood that they will choose to commute within the

fringes of the city (within the peripheral constellations or to other areas in the suburbs), rather

than choosing to commute to the central urban areas. The Exp (B) means that the probability of

choosing to commute within the peripheral constellations is almost five (4.8351) times higherthan the probability of choosing to commute to the central urban area when the jobs-housing

balance is high. The ratio of the two odds ratios of jobs-housing balance within the peripheral

constellations and in other areas in the suburbs is 4.8351/3.5805 = 1.35, giving the odds of

choosing to commute within the peripheral constellations over commuting to other areas in the

suburbs. This can be interpreted to mean that when the jobs-housing balance is high, a worker

residing in the peripheral constellations has a probability of choosing to commute within the

peripheral constellations that is 1.35 times higher than that of commuting to other areas in the

suburbs. One major reason for this is that patterns of urban growth in the planned peripheral

constellations are characterized by higher self-containment and jobs-housing balance, which

would provide opportunities to increase the spatial match between the residential and job-relatedlocation choices of households within the planned constellations. That means that the current

policies designed to encourage the formation of sub-centres on the fringes of the city would also

play a positive role in reducing suburb-to-centre commuting.

According to the results shown in Table 4, employment sector characteristics have an impact

on a workers commuting pattern. When a worker is employed in the manufacturing sector, there

is an increased probability of commuting within the peripheral constellations or to other areas

in the suburbs, compared with workers employed in social services or other sectors. This

probability is higher than that of commuting to the central urban area. However, a worker in

advanced tertiary employment has a higher likelihood of commuting to the central urban areas.

One reason for this is that since the 1980s, industrial development in the suburbs of Beijing hasbeen characterized by dramatic increases in manufacturing, while the finance and other

advanced service sectors grew slowly in the suburbs. This differs from suburbanization in

Western countries where advanced service sectors have grown with considerable speed in the

suburbs since the 1980s (Lee et al. 2006). As a result, in Beijing, an increase in the manufac-

turing sector in the suburbs would reduce the need of those who are employed in manufacturing

to travel to the central urban areas. The results of the analysis in Table 4 show that a worker

employed in the commercial sector tends to commute to the suburbs rather than to the central

urban areas. This has resulted in the rapid growth of the commercial sector in the inter-suburban

areas, promoted by a growth in population size since the end of the 1990s.

The results in Table 4 suggest that the probability of commuting within the peripheralconstellations or to other areas in the suburbs decreases when a worker is employed by the

national government, by the party or in research and development. The values of the odds ratios

show that this probability is higher than that of commuting to the central urban area. A major

reason for this is that the government administrative sector is still concentrated in the central

area of the city for political reasons, despite there being an obvious process of decentralization

of the non-government-related sectors and population (Zhou 1997).

Table 4 shows that when a worker is from a low-income household, the probability of

choosing to commute within the peripheral constellations or to other areas in the suburbs

increases. The odds ratios indicate that this probability is higher than that of choosing to

commute to the central urban area. The results suggest that workers from low-income house-holds tend to have shorter commuting distances within the fringes of the city rather than

engaging in long-distance commuting to the city centre. One reason for this may be that workers

from low-income households need to save on transport costs and thus would deliberately refrain

from long-distance commuting. Another reason may be that housing prices are lower on the

748 P. Zhao et al.



15/21

fringes of the city. This finding is contrary to that reported by Dubin (1991), who found that

low-income workers had longer commutes while higher income workers had shorter commutes

during the employment decentralization process in the US. This is largely a result of the

differences between the US and China in terms of their suburban labour and housing markets.

In most of Chinas large cities, Beijing for example, the labour market in the suburbs is

dominated by manufacturing, which is one of the main sources of jobs for low-income workers,while in the US, the finance, insurance and real estate sectors along with advanced tertiary

employment account for a large share of the labour market in the suburbs (Garreau 1991;

Stanback 1991; Panebiaco and Kiehl 2003). Compared to the US, the housing market in the

suburbs of Chinas cities is dominated by low-priced housing, although a few luxurious gated

villas are now appearing in the suburbs.

As introduced in the theory section, in a polycentric model, a workers commuting pattern

is a result of a combination of workplace and housing location choice. For low-income suburban

workers in Beijing, there are two possible situations relating to their job and home location

choices. In the first situation, some residents move to live in the suburbs willingly to obtain

jobs. Most of these residents are employed in the manufacturing sector and move out of thecentral urban areas to be near the factories they work in. This point is supported by the results

of the analysis above. In the second situation, some residents are forced to be relocated to the

suburbs due to urban regeneration in the city centre (Zhou 1997). Most of them are unwilling

to move to the suburbs. Their job mobility often lags behind their housing mobility because their

job is still in the city centre. Consequently, most of them have a longer commute from the

suburbs to the city centre.

The results in Table 4 show that a worker from a household with residential location

preferences for a community environment tends to commute longer distances to the city centre

rather than to the peripheral constellations or other areas in the suburbs. One major reason for

the results is that a worker who makes the decision to live on the fringes of the city may do sobecause of low price and spacious housing but also because neighbourhoods with large, green

open spaces satisfy their preferences concerning their community environment, even if the

workers job is located in the city centre. This implies that a worker is likely to accept a longer

suburb-to-centre commute for the sake of their housing location preferences. In other words, as

with external land use pattern factors, the internal factor of residential location attitudes also has

a significant impact on a workers commute. This finding is consistent with what has been

reported in previous studies using Western cases (Rouwendal and Meijer 2001; Handy et al.

2005).

The ratio of the two odds ratios (location preferences for other areas in the suburbs and

within the peripheral constellations), 0.3481/0.2980 =1.17, indicates that when a householdspreferences for residential location are considered, a worker has a higher probability of choosing

to commute to other areas in the suburbs, compared with that of choosing to commute within the

peripheral constellations. This result confirms the differences in actual residential location

choices between the planned sub-centres and other areas in the suburbs. One major reason for

this is the supply of various types of housing and residential land use in the planned sub-centres,

which are more centrally planned and tightly regulated than other areas of the suburbs (Zhao

et al. 2009a). As a result, the role of residential preferences in the actual choice of residential

location may be constrained and weaker in these planned sub-centres than in other areas of the

suburbs.

The results above also imply that sprawling development outside planned sub-centres couldoffer benefits to some individuals. These benefits include, for example, low-priced housing,

large living space, auto use, etc. This is consistent with the findings reported from North

American and European countries (Gordon and Richardson 1996; Brueckner 2000). In Beijing,

this trend may be strengthened by changes in lifestyle. It has been reported that due to a desire




16/21


17/21

workers employed in finance and other advanced services tend to have a longer commute to the

city centre. This is different from the present situation in Western countries where workers

employed in the advanced service sectors have shorter commuting distances to the suburbs. One

main reason for this is that the suburbanization of Beijing is characterized by rapid growth in

manufacturing in the suburbs while the advanced service sectors are still concentrated and

growing rapidly in the city centre.If commuting patterns are analysed in terms of income, it appears that a low-income worker

tends to commute within the suburbs and have a shorter commute to work. This is attributed to

the limited budget in a low-income household, soaring housing prices in the city centre,

manufacturing employment decentralization and travel costs caused by long-distance commut-

ing.

Workers are likely to accept a longer suburb-to-centre commute for the sake of fulfilling

their housing location preferences concerning their community environment, such as inexpen-

sive and spacious housing and large green spaces. This implies that lifestyle, in conjunction with

other factors, has a growing ability to affect workers choice of residential location and their

commuting patterns. However, compared with patterns of urban growth, preferences for acommunity environment carry less power to influence the variation in commuting activities.

This may be because housing supply and land use in communities in Beijing are still tightly

regulated by the remaining centrally planned system in the current transformation era.

The results above are significant for at least four reasons. First, they confirm the argument

that patterns of urban growth have significant associations with individuals choice of residential

and job location and thus with their commuting patterns. Second, employment decentralization

and the formation of sub-centres would provide opportunities to enhance the spatial match

between a households choice of residential and job location in the suburbs. The effects of

employment decentralization on workers commutes depend on the course of the suburbaniza-

tion of economic sectors. In Chinas cities at the moment, there is rapid growth in the manu-facturing sector in the suburbs which will greatly affect workers employed in manufacturing.

Third, regional policies can affect individuals commuting patterns through the supply of

various types of housing, the distribution of industry and the delivery of transport services. In

Chinas cities, the regional policies designed to promote the formation of polycentric spatial

structures and control sprawling development can play a positive role in reducing long-distance

suburb-to-centre commuting. Fourth and finally, individuals commuting patterns can partly be

explained by their preferences concerning residential location. In the case of China, the changes

in lifestyle that are taking place at present are having an increasing influence on individuals

travel behaviour.

References

Aldrich JH, Nelson FD (1984) Linear probability, logit, and probit models. Sage Publications, Newbury Park, CA

Alonso W (1964) Location and land use. Harvard University Press, Cambridge, MA

Anas A, Arnott R, Small KA (1998) Urban spatial structure. Journal of Economic Literature 36: 14261464

Baccaini B (1997) Commuting and residential strategies in the Ile-de-France: Individual behavior and spatial con-

straints.Environment and Planning A 29: 18011829

BIUPD (1982)Beijing city construction master plan (19822000). Beijing Institute Urban Planning and Design, Beijing

(in Chinese)

BIUPD (1993) Beijing city master plan (19912010). Beijing Institute Urban Planning and Design, Beijing (in

Chinese)

BIUPD (2004) Beijing city master plan (20042020). Beijing Institute Urban Planning and Design, Beijing (in

Chinese)

Bohte W, Maat K, van Wee B (2009) Measuring attitudes in research on residential self-selection and travel behaviour:

A review of theories and empirical research. Transport Reviews29: 325357




18/21

Boussauw K, Neutens T, Witlox F (2010) Minimum commuting distance as a spatial characteristic in a non-monocentric

urban system: The case of Flanders. Papers in Regional Science. Published online: doi: 10.1111/j.1435-5957.

2010.00295.x

Brueckner JK (2000) Urban sprawl: Diagnosis and remedies. International Regional Science Review 23: 160171

BSB (1997) Report of the first (1996) national basic unit census of Beijing. Beijing Statistic Bureau, Beijing (in

Chinese)

BSB (2002) Report of the second (2001) national basic unit census of Beijing. Beijing Statistic Bureau, Beijing (inChinese)

BSB (2005)Report of the first (2004) national economic census of Beijing. Beijing Statistic Bureau, Beijing (in Chinese)

BSB (2009) Report of the second (2004) national economic census of Beijing. Beijing Statistic Bureau, Beijing (in

Chinese)

BSB (various years)Beijing statistic yearbook. Beijing Statistic Bureau, China Statistical Publishing House, Beijing (in

Chinese)

BTC (2000) Beijing second (2000) travel survey. Beijing Transport Committee and Beijing Committee of Urban

Planning, Beijing (in Chinese)

BTC (2005)Beijing third (2005) travel survey. Beijing Transport Committee and Beijing Committee of Urban Planning,

Beijing (in Chinese)

Camagni R, Gibelli MC, Rigamonti P (2002) Urban mobility and urban form: The social and environmental costs of

different patterns of urban expansion. Ecological Economics40: 199216Cao X, Mokhtarian PL, Handy SL (2009) Examining the impacts of residential self-selection on travel behaviour: A

focus on empirical findings. Transport Reviews29: 359395

Cervero R (1995) Planned communities, self-containment and commuting: A cross-national perspective Urban Studies

32: 11351161

Cervero R, Wu K-L (1997) Polycentrism, commuting, and residential location in the San Francisco Bay area. Environ-

ment and Planning A 29: 865886

Cervero R, Wu K-L (1998) Sub-centring and commuting: Evidence from the San Francisco Bay area, 1980-90. Urban

Studies35: 1059-1076

Circella G, Mokhtarian PL, Handy SL (2008) Land use, attitudes, and travel behavior relationships: A cross-sectional

structural equations model for Northern California. Paper presented at the Transportation Research Board 87th

Annual Meeting Washington, DC

van de Coevering P, Schwanen T (2006) Re-evaluating the impact of urban form on travel patterns in Europe andNorth-America.Transport Policy 13: 229239

Crane R (2000) The influence of urban form on travel: An interpretive review. Journal of Planning Literature 15:

323

Crane R, Chatman DG (2003) Traffic and sprawl: Evidence from US commuting, 1985 to 1997. Planning and Markets

6: 1422

Deng FF, Huang Y (2004) Uneven land reform and urban sprawl in China: The case of Beijing. Progress in Planning

61: 211236

Dieleman F, Dijst M, Burghouwt G (2002) Urban form and travel behaviour: Micro-level household attributes and

residential context. Urban Studies39: 507527

Downs A (1992)Stuck in traffic: Coping with peak-hour traffic congestion. The Brookings Institution, Washington, DC

Dubin R (1991) Commuting patterns and firm decentralization. Land Economics 67: 1529

Ewing R (1997) Is Los Angeles-style sprawl desirable? Journal of the American Planning Association 63: 107126Ewing R, Bartholomew K, Winkelman S, Walters J, Chen D (2008) Growing cooler: The evidence on urban develop-

ment and climate change. Urban Land Institute, Washington DC

Galster G, Hanson R, Ratcliffe MR, Wolman H, Coleman S, Freihage J (2001) Wrestling sprawl to the ground: Defining

and measuring an elusive concept. Housing Policy Debate 12: 681717

Garreau J (1991) Edge city: Life on the new frontier. Doubleday, New York

Gaubatz P (1999a) Chinas urban transformation: Patterns and process of morphological change in Beijing, Shanghai

and Guangzhou. Urban Studies 36: 14951521

Gaubatz P (1999b) Understanding Chinese urban form: Context for interpreting continuity and change.Built Environ-

ment24: 251270

Giuliano G (1999) Land use policy and transportation: Why we wont get there from here. Paper presented at the

Conference on Policies for Fostering Sustainable Transportation Technologies Pacific Grove, California

Giuliano G, Small K (1993) Is the journey to work explained by urban structure? Urban Studies30: 14851500

Glaeser E, Kahn ME (2001) Decentralized employment and the transformation of the American City, NBER Working

Paper 8117

Gordon P, Richardson HW (1996) Beyond polycentricity: The dispersed metropolis, Los Angeles, 19701990.Journal

of the American Planning Association 62: 289295

752 P. Zhao et al.



19/21

Gordon P, Richardson HW (1997) Are compact cities a desirable planning goal? Journal of the American Planning

Association63: 95106

Gordon P, Kumar A, Richardson HW (1989a) Congestion, changing metropolitan structure and city size in the United

States. International Regional Science Review 12: 4556

Gordon P, Kumar A, Richardson HW (1989b) The influence of metropolitan spatial structure on commuting time.

Journal of Urban Economics26: 138151

Gordon P, Richardson HW, Jun MJ (1991) The commuting paradox: Evidence from the top twenty. Journal of theAmerican Planning Association57: 416420

Handy SL, Cao X, Mokhtarian PL (2005) Correlation or causality between the built environment and travel behavior?

Evidence from Northern California. Transportation Research Part D 10: 427444

Hanson S (1982) The determinants of daily travel-activity patterns: Relative location and sociodemographic factors.

Urban Geography3: 179202

Hanson S, Schwab M (1995) Describing disaggregate flows: Individual and household activity patterns In: Hanson S,

(ed) The geography of urban transportation, 2nd edn. The Guilford Press, New York

Huang J (2009) Beijing will levy emissions fee for motorized vehicles. Beijing Youth Daily (in Chinese) 8. 31-01-2009

IEA (2006) World energy outlook. International Energy Agency, Paris

Jiang F, Liu S, Yuan H, Zhang Q (2007) Measuring urban sprawl in Beijing with geo-spatial indices. Journal of

Geographical Sciences 17: 469478

Kenworthy JR, Laube FB (1999) Patterns of automobile dependence in cities: An international overview of key physicaland economic dimensions with some implications for urban policy. Transportation Research A 33: 691723

Kim C (2008) Commuting time stability: A test of a co-location hypothesisTransportation Research Part A42: 524544

Kitamura R, Mokhtarian PL, Laidet L (1997) A micro-analysis of land use and travel in five neighborhoods in the San

Francisco Bay area. Transportation24: 125158

Knight RV, Stanback Jr TM (1976) Suburbanization and the city. Allenheld Osmun, Osmun & Universe Books, New

York

Lee S, Seo JG, Webster C (2006) The decentralising metropolis: Economic diversity and commuting in the US suburbs.

Urban Studies 43: 25252549

Levinson DM, Kumar A (1994) The rational locator: Why travel times have remained stable. Journal of the American

Planning Association 60: 319332

Li H (2006)Study on the urban form, traffic pattern and resident traffic mode . PhD, Southeast University Nanjing (in

Chinese)Lu L, McCarthy L (2008) Spatial restructuring in urban China amid globalization and economic reform: The case of

Guangzhou. Built Environment34: 444463

Ma LJC (2004) Economic reforms, urban spatial restructuring, and planning in China. Progress in Planning 61:

237260

Ma LJC, Wu F (2005) Restructuring the Chinese city: Changing society, economy and space. Routledge, Abingdon

McFadden D (2001) Disaggregate behavioural travel demands RUM side: A 30 years retrospective. In: Hensher D (ed)

Travel behaviour research: The leading edge. Elsevier Science, Oxford

Moilanen M (2010) Matching and settlement patterns: The case of Norway. Papers in Regional Science 89: 607623

Muth RF (1969)Cities and housing: The spatial pattern of urban residential land use. The University of Chicago Press,

Chicago, IL

van Ommeren J, Rietveld P, Nijkamp P (2000) Job mobility, residential mobility and commuting: A theoretical analysis

using search theory. Annals of Regional Science 34: 213232Pan H, Shen Q, Zhang M (2009) Influence of urban form on travel behavior in four neighborhoods of Shanghai. Urban

studies46: 275294

Panebiaco S, Kiehl M (2003) Suburbanisation, counterurbanisation, reurbanisation? An empirical analysis of recent

employment and population trends in Western Europe, paper presented at the ERSA 2003 conference, Jyvskyl

Pinjari AR, Pendyala RM, Bhat CR, Waddell PA (2007) Modeling residential sorting effects to understand the impact

of the built environment on commute mode choice. Transportation 34: 557573

Richardson HW (1977) A generalization of residential location theory. Regional Science and Urban Economics 7:

251266

Rodriguez DA, Targa F, Aytur SA (2006) Transport implications of urban containment policies: A study of the largest

twenty-five US metropolitan areas. Urban Studies 43: 18791897

Rouwendal J, Meijer E (2001) Preferences for housing, jobs and commuting: A mixed logit analysis. Journal of

Regional Science41: 475505

Scheiner J, Holz-Rau C (2007) Travel mode choice: Affected by objective or subjective determinants? Transportation

Quarterly34: 487511

Schwanen T, Dieleman FM, Dijst M (2001) Travel behaviour in Dutch monocentric and polycentric urban systems.

Journal of Transport Geography 9: 173186




20/21

Schwanen T, Dijst M, Dieleman FM (2004) Policies for urban form and their impact on travel: The Netherlands

experiences.Urban Studies41: 579603

Schwanen T, Mokhtarian PL (2005) What affects commute mode choice: Neighborhood physical structure or prefer-

ences toward neighborhoods? Journal of Transport Geography 13: 8399

Shen Q (1997) Urban transportation in Shanghai, China: Problems and planning implications. International Journal of

Urban and Regional Research 21: 589606

Shen Q (2000) Spatial and social dimensions of commmuting.Journal of the American Planning Association66: 6882Song S (1992) Spatial structure and urban commuting, PhD, University of California, Irvine

Song Y, Knaap G-J (2004a) Measuring the effects of mixed land uses on housing values.Regional Science and Urban

Economics34: 663680

Song Y, Knaap G-J (2004b) Measuring urban form: Is Portland winning the war on sprawl? Journal of American

Planning Association 70: 210225

Stanback Jr TM (1991) The new suburbanization challenge to the central city. Westview Press, Boulder, CO

Travisi CM, Camagni R, Nijkamp P (2006) Analysis of environmental costs of mobility due to urban sprawl: A

modelling study on Italian cities. Tinbergen Institute, Rotterdam

Tsai Y-H (2005) Qualifying Urban form: Compactness versus sprawl. Urban Studies42: 141161

Waddell P (2001) Towards a behavioural integration of land use and transportation modeling. In: Hensher D (ed) Travel

behaviour research: The leading edge. Elsevier Science, Oxford

Wang D, Chai Y (2009) The jobs-housing relationship and commuting in Beijing, China: The legacy of danwei.Journalof Transport Geography 17: 3038

Wang D, Li SM (2004) Housing choice in a transitional economy: The case of Beijing, China. Environment and

Planning A 36: 6987

Wei YD (2001) Dencentralization, marketization and globalization: The triple process underlying regional development

in China. Asian Geographer20: 723

White MJ (1988) Location choice and commuting behavior in cities with decentralized employment. Journal of Urban

Economics24: 129152

Wong S, Tang B (2005) Challenges to the sustainability of development zones: A case study of Guangzhou Devel-

opment District, China. Cities22: 303316

Wu F, Webber K (2004) The rise of foreign gated communities in Beijing: Between economic globalization and local

institutions.Cities21: 203213

Wu L (2002) Beijing-Tianjin-Hebei regional spatial development plan study. Tsinghua University Press, Beijing (inChinese)

Yang J (2006) Transportation implications of land development in a transitional economy: Evidence from housing

relocation in Beijing. Transportation Research Record1954: 714

Yinger J (1992) City and suburb: Urban models with more than one employment center. Journal of Urban Economics

31: 181205

Zacharias J, Tang Y (2010) Restructuring and repositioning Shenzhen, Chinas new mega city.Progress in Planning73:

209249

Zhao P (2010) Sustainable urban expansion and transportation in a growing megacity: Consequences of urban sprawl

for mobility on the urban fringe of Beijing. Habitat International 34: 236243

Zhao P, Lu B (2009) Transportation implications of the metropolitan spatial planning in megacity Beijing.International

Development and Planning Review31: 235261

Zhao P, Lu B (2010) Exploring job accessibility in the transformation context: An institutionalist approach and itsapplication in Beijing. Journal of Transport Geography 18: 393401

Zhao P, Lu B, Woltjer J (2009a) Conflicts in urban fringe in the transformation era: An examination of performance of

the metropolitan growth management in Beijing. Habitat International 33: 347356

Zhao P, Lu B, Woltjer J (2009b) Growth management and decentralisation: An assessment of urban containment policies

in the transformation era in Beijing. International Development Planning Review 31: 5580

Zhao P, Linden GJ, Lu B (2009c) The effects of transport accessibility and jobs-housing balance on commuting time:

Evidence from Beijing. International Planning Studies14: 6583

Zhao P, Lu B, de Roo G (2011) Impact of the jobs-housing balance on urban commuting in Beijing in the transformation

era.Journal of Transport Geography19: 5969. doi:10.1016/j.jtrangeo.2009.09.008

Zhao Y, Zhang G, Wang Y, Zhou L (2007) Analyzing impacts of land-use patterns on travel characteristics. Urban

Transport of China 1: 4650 (in Chinese)

Zheng S, Fu Y, Liu H (2005) An emperical study on the willingness to pay for residential location and location choice

of urban residents in China. Economic Geography 25: 213 (in Chinese)

Zhou Y (1997) On the suburbanization of Beijing. Chinese Geographical Science 7: 208219

754 P. Zhao et al.



21/21

Resumen. El estado actual del conocimiento sobre los vnculos entre el crecimiento urbanstico

y los patrones de desplazamiento al trabajo est dominado por ejemplos de pases desarrollados.

Este artculo examina el impacto del crecimiento urbano en los desplazamientos al trabajo en

una ciudad como Pekn, la cual est siendo objeto de una rpida reestructuracin econmica y

espacial. Los resultados del anlisis de datos de un muestreo de hogares mostr que es probable

que un desarrollo urbano compacto y agrupado en torno a sub-centros planificados a propsito

de este modo reduzca la necesidad de largos desplazamientos al trabajo hasta el centro de la

ciudad, si se tienen en cuenta las caractersticas socioeconmicas de los trabajadores, la facil-

idad de acceso a transporte y las preferencias de cada hogar en cuanto a la zona residencial. Lostrabajadores empleados en el sector manufacturero tienden a realizar desplazamientos ms

cortos y a viajar dentro de los sub-centros suburbanos planificados. Esto revela que la descen-

tralizacin del empleo en el sector manufacturero proporciona ms oportunidades de mejorar la

correspondencia espacial entre la eleccin de la zona residencial y la del puesto de trabajo. Las

preferencias familiares en cuanto a la zona residencial tienen un efecto en los patrones de

desplazamiento al trabajo, siendo ms probable que los empleados con ingresos ms altos estn

dispuestos a realizar desplazamientos ms largos a cambio de satisfacer sus preferencias

residenciales. Una dramtica reestructuracin urbana, junto con cambios en el estilo de vida,

est creando nuevos patrones de desplazamiento al trabajo en las ciudades chinas de rpido

crecimiento.

doi:10.1111/j.1435-5957.2011.00343.x

2011 h h ( ) P i R i l S i 2011 RSAI P bli h d b Bl k ll P bli hi 9600 G i R d

Documents

j.1435-5957.2010.00343.x the Impact of Urban Growth on Commuting Patterns in a Restructuring City_Evidence From Beijing