Elsevier

Cities

Volume 67, July 2017, Pages 53-62
Cities

Metro-induced gentrification: A 17-year experience in Taipei

https://doi.org/10.1016/j.cities.2017.04.019Get rights and content

Highlights

  • Metro-induced gentrification occurred in emerging cities worldwide in the long term.

  • Metro-induced gentrification could happen in both inner and outer city areas but with different features.

  • Differences between inner and outer city areas could be diminished by elevating development densities in outer city areas.

  • Emerging cities are in a predicament of developing efficient transit systems and preventing gentrification.

Abstract

The present study explores the relationship between metro systems and gentrification. Three hypotheses are proposed as follows: access to metro stations induces gentrification, metro-induced gentrification occurs in both inner and outer city areas, and metro-induced gentrification differs between the inner and outer city areas. The present study tested these hypotheses by using population migration, college graduates, increased floor area, and house price as gentrification outcomes. Panel data and traditional linear regression analyses were conducted between 1996 and 2013 in Taipei City, Taiwan. Empirical evidence generally supports the hypotheses. Particularly, metro-induced gentrification revealed by highly educated residents and real-estate development was significantly stronger in the outer city areas, whereas metro-induced gentrification revealed by house prices was significantly stronger in the inner city areas. The current results broaden the understanding on the manner in which infrastructure investments were related to gentrification in an emerging city and imply that the local government should adopt sufficient means to deal with metro-induced gentrification when upgrading public transport systems.

Introduction

After World War II, many emerging cities have started developing their metro systems to enable efficient and environment-friendly urban progress. For instance, the metro systems in Eastern Asia have started their services sequentially in Beijing (1969), Pyongyang (1973), Seoul (1974), Hong Kong (1979), Singapore (1987), Shanghai (1993), Kuala Lumpur (1996), Taipei (1996), Guangzhou (1997), and Bangkok (2004). Empirical evidence and the literature generally agree that metro systems are associated with improved transportation efficiency (e.g., Cervero, 1994), reduced air pollution and energy consumption (e.g., Poudenx, 2008), elevated accessibility (e.g., Lewis-Workman & Brod, 1997) and land value (e.g., Lin & Hwang, 2004), and relocated populations and industries (e.g., Cervero & Landies, 1997) along metro corridors. However, the above changes also increase living expenses because they usually increase the livability and commercialization of a metro corridor. These changes can attract people and businesses who can afford land costs to move in the corridor and displace pre-existing low-income families and small businesses. This class-upward process along metro corridors can be considered metro-induced gentrification. Developing a metro system welcomes capital investment, and it is regarded as one of the major causes of gentrification (Zheng & Kahn, 2013); however, its effects on gentrification have not convincingly been recorded in the literature.

Previous research on the effects of capital investment on gentrification have mostly focused on urban redevelopment (e.g., the Shanghai study of Wang and Lau (2009s)) and housing renewal (e.g., the Seoul study of Ha (2004)); however, studies that have explored the relationships between metro system and gentrification are rare. The research by LeRoy and Sonstelie (1983) is the first in the literature to study the influence of transportation innovation on gentrification. The Alonso–Muth model (Alonso, 1964, Muth, 1969) was extended and used to explain how declining car costs influenced the residential locations of the rich and the poor. Their major argument is that the affluent resided at the city center before the era when cars were used as commuting modes, moved to suburbs when cars became affordable for only the rich, and returned to city center as cars were affordable to both the rich and the poor. They used socioeconomic descriptive statistics and commuting mode attributes between 1850 and 1977 in the US to support the above argument. Lin (2002) used the model developed by LeRoy and Sonstelie (1983) and hypothesized that metro station access spurs gentrification and empirically confirmed the hypothesis using the changes in residential property values between 1975 and 1991 in Chicago. Both of the above studies used a single indicator of gentrification (house price) while ignoring many of its other features (Hamnett, 1991, Lees, 1994, Zukin, 1987). The study by Kahn (2007) on 14 major cities in the US in 1970–2000 is a start of using multiple indicators to represent gentrification in transit and gentrification literature. He found that gentrification (denoted by house price and college graduates) is greater in communities with easy access to “Walk and Ride” stations compared to communities close to “Park and Ride” stations. From then, further research on transit and gentrification mostly used multiple indicators to represent gentrification and distance to station to represent transit access. These studies included those of Young (2007), Plevak (2010), Pollack, Bluestone, and Billingham (2011), Feinstein and Allen (2011), and Saldana and Wykowski (2012) in the USA, Pagliara and Papa (2011) in Italy, and Zheng and Kahn (2013) in China. However, none of the above studies determined whether their samples were from gentrifiable areas. A gentrifiable area reveals a below-average social status that could be measured through income, education, or percentage of residents in professional occupations. For an area to be susceptible to gentrification, it must be considered gentrifiable at the beginning of the analysis period according to the arguments by Freeman (2005), Hammel and Wyly (1996), and Walks and Maaranen (2008). Considering the above requirement, Grube-Cavers and Patterson (2015) analyzed a sample of gentrifiable census tracts in three major Canadian cities and reported that proximity to metro stations has a statistically significant effect on gentrification in Toronto and Montreal.

The above metro-induced gentrification research obtained interesting conclusions but left notable questions unanswered. The first question is whether metro-induced gentrification occurs in both inner and outer city areas. If so, then the next question is: what are the differences between metro-induced gentrification in the inner and outer city areas. Hackworth (2002) argued that gentrification extended from city centers to suburban areas and even rural areas by the 1990s; however, previous metro-induced gentrification research neglected this sprawling and provided very limited information about the outer areas of a city. The last question is whether metro-induced gentrification also occurs in regions other than North America in view of long-term experiences. The existing evidence of metro-induced gentrification outside North America are all short-term changes, which are three-year changes in China (Zheng & Kahn, 2013) and seven-year changes in Italy (Pagliara & Papa, 2011), and these changes were not confirmed that they were in gentrifiable areas. Gentrification is related to land development and household migration; thus, the changes that occurred over a period longer than a decade should be more convincing than those that occurred within a few years.

To answer these questions, the present study used data collected during a 17-year period in order to explore how access to metro stations is related to gentrification in Taipei, Taiwan. Sample data were from seven gentrifiable districts in both inner and outer areas in Taipei City between 1996 and 2013. Panel data and traditional linear regression analyses were used to examine the relationships between proximity to metro stations and gentrification. The present study contributes two novel arguments to the literature. First, in addition to the developed cities in North America, metro-induced gentrification could also occur in emerging cities worldwide. Second, metro-induced gentrification could happen in both inner and outer city areas, as well as reveal the different features between them. Analytical results imply that governments should adopt sufficient means to deal with metro-induced gentrification when upgrading public transport systems.

Section snippets

Method

This section specifies the gentrification outcomes and the factors that can potentially affect such outcomes. Furthermore, it describes the hypotheses and model specifications of the present study.

Data

The number of stations in the Taipei metro system increased from 12 stations in 1996 to 108 stations by the end of 2015. Fig. 1 shows that by the end of 2015, the network included5lines and was 131.1 km in length. The system currently serves 5million residents in the 12 Districts of Taipei City and 10 of the 29 Districts in New Taipei City. On average, the system has two million passengers per day. The Taipei metropolitan area is located in a basin, and the metro network covers the bottom of the

Results

Two regression methods were used to analyze the four gentrification outcomes: panel data and ordinary least-squares regressions. Hausman tests were used to determine whether the fixed effects model or the random effects model should be selected in the panel data analyses. Explanatory variables with a coefficient significance below the confidence level of 1-α = 90% were withdrawn. Table 3, Table 4, Table 5, Table 6 present the estimation results. For each outcome, models were estimated for the

Discussion

Surprisingly, population migration declined with decreasing distance to metro stations among the sample Lis in the inner city areas. Based on the sample data, we infer two possible reasons causing the results. The first reason is from an extremely outlier observation in the inner city sample. The distribution of the Migrateit and MetroDistit distributions among the inner city sample clearly reveals an outlier, i.e., Chengkong Li. This observation denotes a migration ratio of 0.42 and a travel

Limitations

Future studies should more closely examine the following issues, which represent the possible limitations of the present study, for a more thorough exploration of how metro systems affect gentrification. First, given that gentrification is related to multiple features, multiple outcomes must be analyzed to obtain reliable conclusions. Owing to data availability, the present study analyzed the four gentrification outcomes associated with neoclassical and Marxist perspectives. Study outcomes are

References (75)

  • M.E. Menon

    The demand for higher education in Cyprus: An educational policy perspective

    Higher Education Policy

    (1997)
  • F. Pagliara et al.

    Urban rail systems investments: An analysis of the impacts on property values and residents' location

    Journal of Transport Geography

    (2011)
  • P.E. Petrakis et al.

    Growth and educational levels: A comparative analysis

    Economics of Education Review

    (2002)
  • M.J. Potepan

    Intermetropolitan migration and housing prices: Simultaneously determined?

    Journal of Housing Economics

    (1994)
  • P. Poudenx

    The effect of transportation policies on energy consumption and greenhouse gas emission from urban passenger transportation

    Transportation Research Part A

    (2008)
  • A. Tansel

    Determinants of school attainment of boys and girls in Turkey: Individual, household and community factors

    Economics of Education Review

    (2002)
  • J. Wang et al.

    Gentrification and Shanghai's new middle-class: Another reflection on the cultural consumption thesis

    Cities

    (2009)
  • W. Alonso

    Location and land use

    (1964)
  • R. Atkinson

    The hidden costs of gentrification: Displacement in central London

    Journal of Housing and the Built Environment

    (2000)
  • B. Badcock

    An Australian view of the rent gap hypothesis

    Annals of the Association of American Geographers

    (1989)
  • D.M. Blackley

    The long-run elasticity of new housing supply in the United States: Empirical evidence for 1950 to 1994

    Journal of Real Estate Finance and Economics

    (1999)
  • G. Bridge et al.

    Microgeographies of retailing and gentrification

    Australian Geographer

    (2001)
  • A. Chen

    China's urban housing reform: Price-rent ratio and market equilibrium

    Urban Studies

    (1996)
  • M.C. Chen et al.

    An empirical analysis of determination of housing prices in the Taipei area

    Taiwan Economic Review

    (2002)
  • E. Clark

    The rent gap and transformation of the built environment: Case studies in Malmö 1860–1985

    Geografiska Annaler. Series B, Human Geography

    (1988)
  • E. Clark

    On blindness, centre pieces and complementarity in gentrification theory

    Transactions of the Institute of British Geographers

    (1992)
  • C. Colclough et al.

    Educating all the children: Strategies for primary schooling in the south

    (1993)
  • M.A. Davis et al.

    The rent-price ratio for the aggregate stock of owner-occupied housing

    Review of Income and Wealth

    (2008)
  • C. Dawkins et al.

    Promoting affordable housing near public transit: The role of Planning

    (2010)
  • C. Dawkins et al.

    Transit-induced gentrification: Who will stay, and who will go?

    Housing Policy Debate

    (2016)
  • B.D. Feinstein et al.

    Community benefits agreements with transit agencies: Neighborhood change along Boston's rail lines and a legal strategy for addressing gentrification

    Transportation Law Journal

    (2011)
  • L. Freeman

    Displacement or succession? Residential mobility in gentrifying neighborhoods

    Urban Affairs Review

    (2005)
  • M. Frenette

    Too far to go on? Distance to school and university participation

    Analytical Studies Research Paper Series

    (2002)
  • M. Frenette

    Access to college and university: Does distance to school matter?

    Canadian Public Policy/Analyse de Politiques

    (2004)
  • A. Grube-Cavers et al.

    Urban rapid rail transit and gentrification in Canadian urban centres: A survival analysis approach

    Urban Studies

    (2015)
  • J. Hackworth

    Postrecession gentrification in New York City

    Urban Affairs Review

    (2002)
  • D.J. Hammel et al.

    A model for identifying gentrified areas with census data

    Urban Geography

    (1996)
  • Cited by (33)

    • Does mass rapid transit reduce motorcycle travel? Evidence from Taipei, Taiwan

      2023, Transportation Research Part D: Transport and Environment
    • Can public transport improve accessibility for the poor over the long term? Empirical evidence in Paris, 1968–2010

      2023, Journal of Transport Geography
      Citation Excerpt :

      However, the longer-term impact of such measures is less clear, as changes in neighborhoods composition (e.g. gentrification, if low-income residents are displaced from the newly accessible neighborhoods by higher real estate prices) may progressively mitigate these effects (Long and Rice, 2019). Numerous studies have examined the impact of public transport improvements on population displacements, in many cities across the world (to cite a few: Foth et al., 2013; Rayle, 2015; Pathak et al., 2017; Lin and Chung, 2017; Heilmann, 2018). Padeiro et al. (2019) did a systematic review of quantitative papers presenting evidence on gentrification outcomes resulting from transit-based interventions, published between 2000 and 2018 (35 papers analyzed).

    View all citing articles on Scopus
    View full text