Difference in difference with multiple treatment times and time periods

[Zach Thompson]

Hi All

I'm trying to use a difference in difference method to estimate the effect of subsidies on the shipment levels of energy efficient goods between 2004-2015 but I'm getting a bit stuck. I have data from 4 provinces 1 which ran no subsidies, one which ran subsidies from 2008-2015, the other two Provinces ran subsidies but only from 2006-2009 and 2007 to 2009. Would this be suitable for a difference in difference estimation? and how would I perform this in Stata? I've run a DiD estimation between the province with no subsidies and the one with subsidies from 2008-2015 but was hoping to make my results a bit more robust by including the other two.

Thanks for you help,

Zach

[Clyde Schechter]

This data does not lend itself to classical DID estimation. But you can use generalized DID estimation here. You can find a good explanation and some examples at https://www.ipr.northwestern.edu/wor.../Day%204.2.pdf.

[Zach Thompson]

Hi Clyde,

Thanks so much for the help and quick response, that made things clearer.

Kind regards

Zach