Difference in difference without control group

[Flore Vandemoortele]

Hello everyone.

For my master thesis, I want to investigate if the impact of environmental taxes and environmental subsidies on the emissions per GDP differ after the implementation of the EU ETS. In my database, all EU ETS-covered countries are implemented and the variables vary from 1990-2020. In 2005, 27 countries implemented the EU ETS. In 2007, the last 3 countries implemented the EU ETS. Thus, from 2007 onwards, all countries in my dataset were threatened. I wanted to use a DID, but it is not possible because i have no control group (at the end, every country is covered by the EU ETS). How can i run this in Stata so I can see the impact of the environmental tax and environmental subsidy on emissions per GDP for the pre-EU ETS period on the one hand and post- EU ETS period on the other hand.

Thank you in advance!

[Jared Greathouse]

I wanted to use a DID, but it is not possible because i have no control group (at the end, every country is covered by the EU ETS)

You're right, it's not possible. Causal inference requires the availability of a control group (for DD anyways!). Even if you have staggered adoption, which you do, there must be some cohort of never-treated units. However, other approaches outside Stata are possible.