The Effect of Transport Policies on Car Use: A Bundling Model with Applications

In an effort to reduce pollution and congestion, Latin American cities have experimented with different policies to persuade drivers to give up their cars in favor of public transport. Borrowing from the bundling literature, the paper presents a novel model of vertical and horizontal differentiation applied to transport decisions: households differ in their preferences for transportation modes -cars vs public transport- and in the amount of travel. The model captures in a simple way a household's response to a policy shock, i.e., how to allocate existing car capacity, if any, to competing uses (peak vs off-peak hours) and how to adjust such capacity overtime. Using few observables, the model is then used to analyze the effects of two major transport policies: the driving restriction program introduced in Mexico-City in November of 1989 -Hoy-No-Circula (HNC)- and the public transport reform carried out in Santiago in February of 2007 -Transantiago (TS). The model's simulated effects are not only consistent with the econometric estimates in Gallego et al (2013) but also help understand the mechanisms that explain them.