LAUSR

We provide new evidence on the impact of recessions on traffic accidents, by exploiting the case of Spain, where the effects of the 2008 economic crisis have been among the strongest in the developed world. We exploit differences in the incidence of the recession across Spanish provinces due to the unequal evolution of the real estate bubble across the territory. We use a unique dataset on the universe of traffic accidents in Spain between 2004 and 2011. We first follow the literature on the topic and examine the impact of the economic crisis on the probability of having a traffic accident. However, we also go one step further, as we are able to identify any changes in the composition of both victims and driving behaviors as a result of the crisis. First, our results show that the Great Recession reduced traffic accidents in Spain. Second, for the compositional effects, we report decreased probabilities of dying or reporting a serious injury. More importantly, we also detect an increase in the probability that people involved in an accident abuse alcohol and drugs. Our results are robust to different measures of the treatment (i.e., employment in the construction sector) and the use of a spatial fixed effects model and are not biased by anticipatory effects. Finally, we show that our findings are driven by less-populated areas. Thus, we suggest that alcohol and drug control measures be reinforced during recessions and more attention should be devoted to rural areas to to strengthen the reduction of road traffic accidents.