Hello everyone,
I am wondering if it possible to perform a difference-in-difference analysis for a time series dataset. I have a cross section of data observed over multiple periods on a single country.
I intend to perform a DiD to see if the occurrence of certain events ('shocks') altered my y variable (GDP per capita). I created a dummy variable and coded the year of the 'shock' as 1 and the same in subsequent years, and in years prior as 0. I then intend to regress my y variable against that x variable (the dummy/DiD estimator).
Problem is, I am unsure if this is workable given that since it is a time series dataset, I do not have the counterfactual (i.e. other countries) to compare with. So essentially, I am just testing a pre-post sort of scenario.
Wondering if this is workable. Thanks.
Ashvinder
I am wondering if it possible to perform a difference-in-difference analysis for a time series dataset. I have a cross section of data observed over multiple periods on a single country.
I intend to perform a DiD to see if the occurrence of certain events ('shocks') altered my y variable (GDP per capita). I created a dummy variable and coded the year of the 'shock' as 1 and the same in subsequent years, and in years prior as 0. I then intend to regress my y variable against that x variable (the dummy/DiD estimator).
Problem is, I am unsure if this is workable given that since it is a time series dataset, I do not have the counterfactual (i.e. other countries) to compare with. So essentially, I am just testing a pre-post sort of scenario.
Wondering if this is workable. Thanks.
Ashvinder
