1
$\begingroup$

I am planing a panel data analysis evaluating a continuous policy intervention. I do have individual level data. I would like to account for individual fixed-effects. However, the explanatory variable does not vary on individual level, but only on state level. The response variable is on individual level. Is that a problem?

$\endgroup$
  • $\begingroup$ So to be clear, I'm guessing this is a policy introduced by the state that affects individuals differentially? $\endgroup$ – hrrrrrr5602 Jan 18 '19 at 11:50
  • $\begingroup$ Yes, that is correct. $\endgroup$ – R-User Jan 18 '19 at 14:12
1
$\begingroup$

If you want to do unit-fixed effects (so eliminating any unobserved variable that varies cross-sectionally but remains the same over time), this should not be a problem, at least assuming that your explanatory variable varies over time (be it a dummy that is 1 in the year the policy was administered and 0 otherwise, or a continuous variable that changes over time according to the intensity of the policy treatment).

It does not matter that the variable is on state level, as long as the policy varies over time. However, if you only have individual observations for the one state that introduced the policy, you will not be able to interpret your results counterfactually. For that you would need similar individuals observed over the same period of time in a state that did not introduce the policy (and apply difference-in-difference or something of the sort).

| improve this answer | |
$\endgroup$
  • $\begingroup$ Thank you for your answer. Every state in my dataset introduced the policy, but to varying degrees. The policy varies over time and state. The aim of my analysis is to find out the effect of the extent of policy intervention on the behavior of individuals. $\endgroup$ – R-User Jan 18 '19 at 14:51
  • $\begingroup$ Then it should not be a problem to introduce unit-fixed effects. You can code an ordinal variable that reflects the varying degrees of the policy introduction. $\endgroup$ – hrrrrrr5602 Jan 18 '19 at 18:24

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.