I'm considering using difference in difference techniques to write my last two undergraduate papers relating minimum wages and employment/poverty. I've been told issues arise with using aggregate data like you would find on major statistical sources. I have notice most DID papers use survey data, however i do not have access to this or time to get it.

So, my question is: what are the issues with not using micro-level data? how big of a deal is it? and are there possible solutions?


  • $\begingroup$ Can you please give an example of a DID paper using survey data. $\endgroup$ – BB King Nov 2 '17 at 22:38
  • $\begingroup$ Card and Krueger (1994) minimum wages in new jersey and pennsyvania $\endgroup$ – Cole Nov 2 '17 at 22:51
  • $\begingroup$ Thanks, now I understand better what time of papers you are referring to. One issue you woud face with aggregate data is getting enough observations in the treatment and control group. Another is that for DID you need some sort of experiment, which usually does not apply to aggregate units of observation. Another issue would be that so many factors affect aggregate variables, that it is very hard to control for everything else even if you have a natural experiment to apply DID. However, in general, DID can be applied to aggregate data. $\endgroup$ – BB King Nov 2 '17 at 23:12
  • $\begingroup$ People can self-select in the Treatment : what if all pennsylvanian workers moved to New-Jersey to get a higher minimum wage. This would lead to increase unemployement and thus potentially over-estimate the effect of the treatment. If you have micro-level data, you can deal with that issue. I suggest that you take a look at "Mostly Harmless Econometrics" book on DiD. It's easy to read and quite clear on the subject. You also have Bertran & al paper "How much should we trust differences-in-differences estimates?" if you are working with "many years of data". $\endgroup$ – Adrien Pacifico May 4 '18 at 7:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.