I'm looking at the effect of a specific registration requirement on turnout and have been trying to decide what form the dependent variable should take. My two main options are turnout levels (i.e. absolute values), whilst controlling for the size of the eligible population, and turnout rates (i.e. turnout levels/eligible population numbers).

Running my main specification with the two gives very different results, one being significant and the other producing insignificant results. Interpretation-wise, I am interested in how the requirement impacted turnout so I feel like turnout levels provide a more natural interpretation for this, however, all existing literature I've seen uses turnout rates, so I'm wondering if there's something I'm missing.


  • $\begingroup$ Welcome. Why the difference-in-difference tag? $\endgroup$ Commented Sep 19, 2020 at 23:22
  • $\begingroup$ Also, is this post helpful? $\endgroup$ Commented Sep 19, 2020 at 23:39


Your Answer

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

Browse other questions tagged or ask your own question.