I have data where each observation is a region within a country, and each observation falls into one of two large groups (each group makes up half the dataset). Each group corresponds to a seven-year funding period for a fiscal transfer program, and my independent variable is the percentage of total allocations that a region was able to spend within a seven-year timeframe.

For the first group, I have data on all seven years of the funding program, but for the second, I have data just on the first five years. Thus, for the observations in the first group, my dependent variable is fund absorption after seven years; for the second group, I can only get fund absorption after five years as a DV. As expected, the dependent variables for the second funding period are all much lower than for the first.

Other than the fact that the DVs are measured differently and are calculated after different years, they are fairly comparable. Would I be able to run a single OLS regression with funding period fixed effects, or should I run two separate regressions?



Your Answer

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