Le coût du bonus/malus écologique: Que pouvait-on prédire?

Ex post evaluations are crucial to measure the empirical effects of public policies. At the same time, one can wonder how much of these effects could be anticipated ex ante. We address this question for the bonus/malus policy introduced in France to reduce co₂ emissions. This policy was, in particular, supposed to be financially neutral. At the end of the day, the co₂ emissions have been reduced by 5% at a cost of 225 M€. Using a structural approach, we show that this cost was not predictable ex ante. The underlying reason is that consumers appear to react more to this policy than to standard changes in prices.