Weighing observations for difference and difference

[Julianna Grato]

Hi everyone, I have a variable age of home. I am running a difference and difference regression and I would like to weight age since for the two zip codes I am using, the average age is 6 years apart. How could I weight the age variable to make the two zip codes more comparable? Thank you.

[Jeff Wooldridge]

I don’t think you need to weight. Just include age as a control variable in the usual DID regression. You might interact age with all of the dummies, too, to allow heterogenous treatment effect.

[Julianna Grato]

Thank you Mr. Wooldridge! Love your textbooks!