Net average treatment effect of blacklisting.

Note: The table reports first difference estimates. Columns 1–5 use the change in log of yearly newly deforested area as the dependent variable. Columns 6 and 7 use the change in log of counterfactual deforestation in the blacklisted observations. Counterfactual deforestation in column 6 is constructed form paired control matches. Counterfactual deforestation in column 7 is estimated based on the set paired control matches. Standard errors, clustered at district level, are reported in parentheses. Time-invariant and time-variant controls include first differences of the variables reported in S2 Table. Observations are selected by a 1:1 closest neighbor matching using inverse-variance weights, with replacement.*,** denote significance at the 10/5% level.Net average treatment effect of blacklisting.