Research

Current project:

Do metro stations increase the safety of women? Evidence from Delhi

with Laila Ait Bihi Ouali and Daniel J. Graham

There is limited evidence on the safety of women in transport in the developing world, and all of the evidence uses small questionnaire samples or experiments. This paper uses a large, standardised dataset which is available for several years to estimate the impact of a new metro on women safety in public spaces. The contribution of this paper is two-fold: 1) the data enables us to use sophisticated panel regression and instrumental variables methodologies to ensure causality and robustness, and thus 2) we infer the causal impact of an urban public transportation project on women safety, which has not yet been done to the best of our knowledge.

Current project:

Do highways help post-communist transition? Evidence from Hungary

with Daniel J. Graham

Hungary increased its highway network from 269 km in 1992 to 1924 km in 2016. This paper is interested in how much of the GDP growth can be attributed to highways in Hungary between 1992 and 2016.

We are interested in whether the post-socialist transition led to a massive spatial redistribution of economic activity and whether regions close to Western-European markets benefitted from their location.

Besides, we estimate agglomeration elasticities for Hungary which have not been done before in an Eastern European setting even though it is essential for urban and transportation project appraisal.

Using Google Maps to calculate generalised cost and accessibility for London: a new methodology

with Zishu Liu and Daniel J. Graham

This paper develops a mode-specific accessibility methodology which uses generalised cost. The use of Google Maps directions service ensures inexpensive data gathering and ease of updating at a small geographical scale.

We calculate ward-level accessibility indicators for the Greater London area for 2018. After comparing these indicators based on overall performance, model sensitivity, consistency with land use and comparing them with widely-used methodologies, we propose the indicator which weights employment with the negative exponential of generalised cost.

This method differentiates wards with lower accessibility better than the current scheme used by Transport for London. We also suggest an adjustment to the currently used spatial decay parameter in market potential metrics.

Metros, agglomeration and firm productivity. Evidence from London

with Daniel J. Graham and Jose M. Carbo

This paper uses highly disaggregate data to study the impact of a new metro on firm productivity. The planned-route IV methodology ensures the causality of results and a fine spatial scale detects the geographical scale of the impact.

We find that within 750 meters to stations aggregate value added, mean firm productivity and the number of local units increase. Areas between 1250 and 2000 meters experience a decrease in the number of local units but mean firm productivity does not change, leading to decreasing aggregate value added.

This finding is interpreted as empirical proof for the agglomeration shadow phenomenon described by Fujita et al. (1999) and Redding and Turner (2015).

Growth or Displacement? A Metro Line's Causal Impact on the Spatial Distribution of Business Units and Employment: Evidence from London

with Daniel J. Graham and Jose M. Carbo

This paper uses data on the location of business units and employment in London to estimate the impact of a new metro line on the spatial distribution of economic activity.

Our unique business units level dataset allows us to track the movement of units across space and time, making it possible to estimate the growth and displacement component of the impact.

In addition to standard panel fixed effect methods, we employ a planned-route instrumental variables methodology which uses planned but abandoned metro alignments.

We find that areas within walking distance to stations experience a significant positive effect (3.6% more business units and 2.5% more employment), whereas areas further off but still within 2000 meters experience a significant negative impact (-1.3% for business units and -3% for employment).

Our result provides empirical evidence for the models of both Fujita et al. (1999) and Redding & Turner (2015) as areas close to the transport scheme but not subject to it are worse off than areas further away. The results suggest no growth, only displacement: the metro merely shifted economic activity closer to the stations. This paper contributes substantively to both the urban and transport economics literatures, and it also provides useful insights for the assessment of wider economic benefits in project appraisal, and thus, for the efficient use of public funds.

working paper

Economic growth and congestion - a new methodology for agglomeration elasticities

"Causal analysis so far has implicitly assumed that completing a transport investment necessarily leads to better accessibility and this better accessibility leads to higher agglomeration, which in turn makes firms more productive. However, in transport literature, it has been evident that an investment does not necessarily lead to better accessibility - due to the phenomenon of induced demand.

Using openly available traffic flow and journey time datasets, current effective density measures can be improved. The methodology presented in this paper makes it possible to evaluate if and how a transport investment leads to an actual change in traffic flows and journey times (thus change in accessibility). We create accessibility measures, which use actual observed journey times for every ward in England between 2009 and 2015. Especially peak-time accessibility captures negative agglomeration externalities arising from traffic congestion. We argue that peak-time effective density may provide a more realistic measure for agglomeration forces."

The wider economic benefits of reducing the cost of travel

"The aim of this paper is to estimate how increased accessibility affects the productivity of firms. We introduce a new measure for accessibility, which incorporates the cost of travel explicitly, thereby controls for negative externalities arising from congestion as well. This approach makes it possible to grasp the potential benefits of the sharp fall in transportation costs due to the ongoing technological revolution in transportation. After estimating total factor productivity for firms in England between 2009 and 2015, we provide several ways for the estimation of agglomeration elasticities. In addition to standard panel methods, we use a new instrumental variable for accessibility, traffic accidents. Our results show that changes in the cost of travel (including congestion costs) change the size of agglomeration for firms and thus lead to firm productivity increase."