Employment and Rail Access: An Evaluation of Transport Improvement Benefits

Evaluation Of The Wider Benefits Of Transport Improvements

Stephen Gibbons and Stephen Machin ¹ November 2003

Executive Summary

Traditional systems of transport policy evaluation are built around the valuation of commuting time-savings. Travel time is valued at some fixed price, usually some pre-determined proportion of the market wage. Policy makers and researchers are, however, increasingly concerned that these traditional methods understate the true gains from transport infrastructure improvements, because there are other, wider benefits that are not fully captured. Some of these 'wider benefits' accrue directly to commuters. ²

Examples include better wages resulting from more-productive firmworker matches or 'thicker', non-monopsonistic labour markets, reduced job-search costs and lower durations of unemployment, and reduced uncertainty in travel times. Firms may benefit directly in terms through the easing of constraints on the supply of labour, from lower cost of transportation of output and intermediate goods, or because lower transport costs to major markets allows firms to increase output and reduce costs through economies of scale. If these factors encourage firms and workers to relocate near points of transport access, then industrial localisation and more general agglomeration economies and communications externalities could become important. ³

The Department of Transport is interested in appraising some of these wider economic benefits of transport improvements, particularly in the context of Crossrail, the proposed rail link to run across London. This, if introduced, is anticipated to have a significant impact on commuting travel. The evaluation entails an assessment of its potential impact on productivity and on the labour market. In this report we consider one set of empirical evidence that relates to this. We look at the association between employment density and rail access in London and the South East based upon disaggregated, ward-level data through time. We go on to examine the effects of recent transport improvements that had potentially strong effects on access to and from South East London - the opening of the Jubilee Line extension, Docklands Light Railway extension to Lewisham, and the reopening of the East London line between Shoreditch and New Cross.

The linkages between rail accessibility and employment are complex and there is no unified theoretical framework. So that it is clear what we are doing at the outset, let us efine some terms. We can think of two main concepts of employment effects from transport innovations. Firstly, a transport improvement may change conditions for firms that locate near that site, so that existing firms are more willing to employ more workers, or there are new firm start-ups that need a workforce. Secondly, a transport innovation may change conditions for commuters who live near that site, so that they can access places with jobs more easily and are more willing to work. The first firm-related effect primarily increases work-place related employment density near the site of the innovation. The density of firms or the number of workers per firm increases near the site of the transport improvement. The second, commuter-related effect increases residence-related employment density. The number of employed people in a residential area increases, or the number of unemployed goes down. The first, firm-related effect also increases residence-related employment density near the site of the improvement if local firms employ many local workers. The second, commuter-related effect cannot increase work-place related employment density near the site of the improvement, since, by definition, commuters do not work locally. Of course, the commuter-effects at all the origin nodes of a commuting network and the firm-effects at the destination nodes are all interrelated.

A firm related increase in employment density in one place must be related to increases in residence related employment density elsewhere - better transport access for commuters living in Uxbridge, Croydon and Enfield makes for a thicker labour marker for firms in the City and the West End. The problem for empirical analysis is that it only feasible, with the data available, to link local employment outcomes to local points of transport access. This leads us on to stress an important aspect of this research: we look only at quite localised, micro-geographic effects on employment. Our research question is whether transport access points act as magnets for employment, either through new firm start-ups or through more workers per firm. We examine work-place related employment density effects as defined above, but our research at this stage has to remain silent about the mechanisms by which these occur. Our target is to get estimates of these effects that are as robust as possible using highly disaggregated geographical data. This is not to say that we think this is the whole story; it is just one piece of the jigsaw.

Clearly, a transport innovation that improves access to the City of London could have far-reaching implications for employment across the whole of the capital. But there is no feasible empirical analysis that can identify the effects of transport innovations in the City on employment in, say, Uxbridge, without very detailed knowledge of the patterns of inter-firm linkages, intra-urban trade patterns and commuting behaviour that could link outcomes in these two places. The local focus of our study is necessary because we want to make links in our data between employment patterns and transport infrastructure, whilst not clouding the story with more general regional and urban effects that would arise from doing this at a more aggregated geographical level. The structure of the report is as follows. Section 2 describes our methods and Section 3 details the data we use to implement them. Sections 4 and 5 describe and discuss the empirical results, firstly for our cross-sectional analysis, and then for our methods based on transport innovations. We present some brief Conclusions and remarks on policy evaluation in Section 6.

The full report in PDF format is available as a download at the foot of this page.

¹ 1 Stephen Gibbons is Lecturer in Economic Geography at the London School of Economics. Stephen Machin Professor of Economics at University College London. Both are members of the Centre for Economic Performance at the London School of Economics.

² 2 The benefits to households of places with good rail access are the subject of our companion report (Gibbons and Machin (2003))

³ 3 For studies of these issues, see papers by Manning (2003), Brueckner and Zenou (2003), Coulson et al (2001), and the texts by Fujta, Krugman and Venables (1999), Ciccone and Hall (1996), Ciccone (2002), Henderson (2003)