Evaluating Public Policies with High Frequency Data: Evidence for Driving Restrictions in Mexico City Revisited

The evaluation of public policies is on the heart of the e cient management of public resources. As complex as it generally is, any reform should be assessed on its ability to achieve its preconceived goals. This research paper attempts to show the importance of the design of a public policy's empirical evaluation, considering the susceptibility that its conclusions might have to changes in the approach to the data. The work of Davis (2008), which nds that a driving restrictions program had no impact on air quality in Mexico City, is revisited showing that reasonable changes in the methodology used can dramatically alter its conclusions. Additionally, evidence is presented that shows the success of the restrictions program in reducing air pollution by 12 to 18% during the first months of its implementation followed by a gradual increase in pollutants concentration, consistent with more limited opportunities for adaptation to the policy in the short-run. Finally, an alternative and robust framework is proposed to carry out the policy evaluation con rming the reduction of pollution right after the program's implementation.