Estimating effect of merger activity

[Will Spencer]

Dear All,

I am attempting to produce an estimation of the effect of merger activity on a variety of outcome variables. I have produced a control group using propensity score matching, but I am uncertain as to how to proceed. I am looking to use an event study method, as the mergers I am considering take place at different points over a 10 year period (I have data from 2 years prior and 2 years post the merger). I am unsure how to generate an estimation that de-trends the treatment outcomes.

Any guidance or advice to get me started would be greatly appreciated.

Kind Regards.

[Jared Greathouse]

One command that uses propensity score matching and (I THINK) Coarsened Exact Matching is the user written flexpaneldid. It performs DD within the context of matching, working with treatments that come and go, as well as staggered implementation designs where units are once treated and always treated.

Theoretically, you COULD do this yourself if you want to, but this command gives a good place to start with.