Dasgupta,2019 documented that

the treated group comprises all firms that are headquartered in countries that have passed a leniency law by year t. The control group comprises firms in countries that never adopted a leniency law in our sample period and firms headquartered in countries that adopted a leniency law at some later point of time.

I am wondering "sample period" here is the whole period of the study or the window [-2;+5] around the law implementation of a specific country. For example, Korea passed the law in 1997, so "sample period" means "1990-2012"(the whole dataset Dasgupta,2019 consider in their analysis) or "1995-2002"?


1 Answer 1


It is definitely hard to interpret, and I wish the authors had been more clear.

However I think "sample period" refers to the whole dataset. There is actually nothing in the identification or data sections that suggests that the authors are using the [-2,+5] window around law implementation in the main regression results. I could be missing something but I only see this window mentioned describing the graphical analysis at the very beginning of the results section. It is not very clear to have a graphical analysis that doesn't actually correspond to the regression being run, but I think maybe that is what is going on. If the authors were running a regression based on offsetting the data around the law passed I think they would have described the "t" index in the regression equations as referring to "years since passage of the lenience law" rather than simply "years".

A way to confirm with certainty that the authors are not using the seven year window in the main regression analysis would be to check the sample size under the regression tables. Seven years of data per country will generate a very different N to 22 years of data per country.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.