Abstract
There has been a significant amount of research in the transportation area on the development of strategies that allow good comparisons between states, such that policy analysis are allowed to be carried out and informative policy-oriented questions are allowed to be answered. In this article, the use of Synthetic Control Methods (SCM) is proposed to overcome several identification problems present in previous studies when constructing comparison groups/counterfactuals. The SCM is used to analyse the effect of New York state's law prohibiting handheld cell phone use while driving on fatality rates. Results show that (i) a synthetic ‘peer state’ for NY when evaluating this specific policy is composed of a convex combination of the states of IL, MA and TX and (ii) that imposing the ban led to a decrease of about 9% in fatality rates.
Acknowledgements
I thank Kristine Brown, Juliana Guimaraes, Ron Laschever, Darren Lubotsky, Elizabeth Powers and Yony Sampaio for valuable discussions. I also thank the editor S.C. Wong and three anonymous referees for very helpful comments. I am responsible for any remaining errors.
Notes
1. The results are robust to the inclusions of each of these states.
2. In order to assess the robustness of the results, additional predictors are included among the variables used to construct the synthetic control, such as age and racial distributions, and unemployment rates. Results are consistent with the ones presented in Section 4 of this article.
3. The number of states receiving positive weights resembles that of Belot and Vandenberghe (Citation2011) and is grater than that obtained by Abadie and Gardeazabal (Citation2003), which had only two states with positive optimal weights.