0
$\begingroup$

I'm estimating this equation for the trade flows between various countries at time t: ($i$ and $j$ are countries)

enter image description here

The last 3 terms are control variables, one of which is oil prices. Can I add oil prices as a control variable if I already have a time fixed effects term, $ \beta_3 $. Further, can I have country pair fixed effects, $\beta_6$, when I'm also controlling for variables that are fixed for two countries across time (for example, if the two countries were in a colonial relationship in the past, this variable wouldn't change with time and would stay fixed for the two countries for the entire period).

$\endgroup$
1
  • $\begingroup$ Welcome to Economics:SE. Thank you for your question; please consider revising it to be more in line with our community expectations. Like many other stacks, we expect questions to provide evidence of prior research. That helps us to understand the question, and avoids our repeating work you've already done. Our help center, and other stacks provide additional resources to assist with revisions. $\endgroup$
    – 1muflon1
    Feb 27 at 10:13
3
$\begingroup$

Your model has $\beta_3 * t$, which is a linear time trend, not time dummies. If that's correct, you are controlling for only a linear trend. Because oil prices do not have a perfect linear trend, you can include them.

But I am not sure you really want the linear trend specification instead of time dummies (say, $\beta_{3t}$). For a model with common time effects (year dummies), time series variables are not allowed due to perfect collinearity. Thus, if the time series variables are the key variables, you have a trouble. But if you only want to control for them, you can just omit them because any common time effects will be eliminated by the time dummies.

Country-pair fixed effects are similar. You cannot use time-invariant variables if those fixed effects are included.

One possible solution is to take the correlated random effects approach, but there is the issue of RE $\ne$ FE. You will have to defend your RE model. That's a different issue.

$\endgroup$

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.