A Sufficient Statistics Approach for Aggregating Firm-Level Experiments

Abstract

We propose a method to estimate the effect of firm policies (e.g., bankruptcy laws or subsidized credit) on allocative efficiency using (quasi-) experimental evidence. Our approach takes general equilibrium effects into account and requires neither a structural estimation nor a precise assumption on how the experiment affects firms. Our aggregation formula relies on treatment effects of the policy on the distribution of output-to-capital ratios, which are easily estimated in (quasi-) experimental data. We show that this method is valid as long as the true data-generating process belongs to a large class of commonly-used models in macro-finance. Finally, we apply this method to the French banking deregulation episode of the mid-1980s and find that this reform led to an increase in aggregate TFP of 2.7%.