Changes

So, despite the graphs, the main results are here! The diff-in-diff coeffs show that an LBO reduces patent flows, grants, and purchases, and increases term expirations. Only the results on R&D are unexpected, though the hadlbo coefficient is significant, negative and about the same magnitude as the diff-in-diff coeff for R&D, so this would suggest that PE firms target firms with low R&D and then increase it back to conventional levels. The pair fixed effects leave the coefficients unchanged (versus no controls) but reduce the standard error. We should check/prove that this is correct.