top of page

Public transport may offer a solution for healthy living. However, we do not have much scientific evidence to support a transport infrastructural modification. It might be because we cannot study public transport interventions experimentally, as its design is usually out of researchers’ control. 


Numerous metros have been implementing in developing countries. In 2010, five metro lines were approved in a medium-sized Chinese city in one package. We used these public transport intervention opportunities and conducted a first natural experimental study of active travel change before and after a new metro system, allowing a plausible causal inference.

Not as expected, we found a new metro does not promote active travel increase or car use reduction. People who adopted the metro are those who used to take the bus. The causal evidence calls for caution in making general assumptions about the effects of urban rail transit investments - not like how the local government has justified.

figures_natural experiment nanchang_fina

Planning knowledge plays a central role in natural experiment designWorking with the local transport planners, we verify that the government used the same rationale to decide where to place the station for all the new lines, such as neighbourhood density, job opportunities or (re)development potentials. Therefore, in our treatment assignment, we use the residents living around new stations as the treatment groups and those living around future stations as control groups. The treatment-control groups could have no systematic difference in the suitability to have a rail transit station, but it is a matter sooner or later to have one.

A natural experiment

Causal effects of new metro on active travel

Not as expected

mode-choice-combo-final1 (1).jpg

How to cite:

G. Sun, J. Zhao, C. Webster, H. Lin (2020). New metro system and active travel: a natural experiment. Environment International 138. doi: 10.1016/j.envint.2020.105605

bottom of page