Transport infrastructures and employment growth in the Paris metropolitan margins

Highlights

• Employment growth is assessed in small municipalities (<1000 jobs).

• A logistic regression model relates employment growth to transport infrastructures.

• Motorways, airports and railway differently affect growth.

• The role of transport differs according to economic sectors.

• Agglomeration replaces motorways as a key factor beyond a given mass of jobs.

Abstract

The role played by transport infrastructures such as motorways, railway stations and airports on local employment growth remains debated and partially inconclusive, especially for small municipalities located in metropolitan margins. This paper examines the case of the Paris region. Based on a logistic regression model using municipal-level employment data (1993–2008), it aims at assessing whether municipalities with small amounts of employment are likely to experience high employment growth depending on their location vis-à-vis transport infrastructures. Motorways are found to have an influence for very small areas and airports for bigger ones. However, the role of transport infrastructures is rather limited when the analysis is differentiated according to economic sectors.

Introduction

The role of transport in local employment growth has been widely discussed in recent decades (Blumenberg and Ong, 2001, Boarnet, 1995, de Vor and de Groot, 2010, Giuliano et al., 2011, Giuliano et al., 2007, Hoogstra and van Dijk, 2004, Redfearn and Giuliano, 2008, Zhang and Guldmann, 2010). Jobs have become more decentralised, cities have undergone a polycentric-based expansion rather than a random sprawling, with face-to-face contacts and agglomeration externalities remaining predominant. Beyond some main determinants such as specialisation (Henderson, 2003, Marshall, 1920) and diversity (Glaeser, 1999, Jacobs, 1969), the role played by transport infrastructures has remained controversial, largely because it raises problems of endogeneity in urban areas (Chandra and Thompson, 2000, Turner, 2009). Endogeneity is a major concern when a predictor (like the existence of an infrastructure) is a consequence of the phenomenon it is supposed to explain.

This debate, however, does not affect metropolitan margins, which have been rather overlooked by the research community. Most metropolitan jobs are usually located in the major centres, and the most common concern of metropolitan authorities is not to let employment scatter in sparsely populated areas. Yet there are several reasons to focus on this subject, even though small numbers of jobs are involved. First, understanding employment growth in metropolitan margins may allow for a better knowledge of the suburbanisation process, and for more adequate decisions in terms of local economic development (Baum-Snow, 2007). Second, it is useful to know whether the existing infrastructure (motorways, regional express trains, airports) provides opportunities for job creation in unlikely geographical areas. On this matter, a lively debate has emerged recently in Paris about the location of the future bypass mass rapid transit. The future stations might partly be built in almost empty areas: the main opposing arguments have focused on the possible creation of new real estate opportunities or, conversely, to the futility of stations located in “potatoes fields”.¹ However, the debate has remained fairly ideological and poorly informed by scientific findings. Third, metropolitan margins are an excellent study area to clarify the role of transport infrastructures in an environment where endogeneity cannot be highlighted, for infrastructure is usually received as an exogenous event (Chandra and Thompson, 2000), rather than a response to local previous growth. Moreover, externalities related to the concentration of activities, to their diversity or to specialisation are not likely to be explanatory factors, since few firms (if any) are located there. As such, measuring the benefits of transport infrastructures in places where urban advantages are missing may provide a clearer view on specific transport effects. A fourth and specifically European reason involves the diversity of existing transport modes: while the North-American literature focuses largely on motorways (Baum-Snow, 2007, Boarnet, 1995, Boarnet and Haughwout, 2000, Ewing, 2008a, Rothenberg, 2009, Ryan, 2005), it is worth questioning the role of regional rail networks that have been highly developed in European conurbations.

Focusing on peripheral areas with small numbers of jobs, this paper aims to test the assumption that the areas located near transportation infrastructures were more likely to undergo employment growth over a 15-year period (1993–2008). Section 2 presents a brief literature review on this issue. Section 3 includes a presentation of data and method. Results are shown in Section 4. Section 5 concludes.