Hello,
I am having trouble running pairwise DiD regressions for states. I am running Stata 15.1.
Thanks for all of your help so far.
I've downloaded my data from NCES, which is data by school year, and have merged all of my files.
After running the reshape long command, I have multiple variables, each from year 1 to 18, as defined by me.
I would like to begin DiD regressions. My dependent variable is the number of students in an IEP (variable is IEPS) for each year, and I am comparing states before and after charter school introduction. That is the dummy variable charterstate. charterstatus is a dummy variable if the district is a charter school district or not.
I am using Kentucky as my main state, and would like to run DiD with 7 neighboring states.
I want to run pairwise DiD regressions comparing districts in KY to districts in only one other state (like Ohio). Then I would like to repeat this for each other state in my sample (i.e., create separate samples with districts from KY and one other state).
So far, I have ran:
reg IEPS i.year i.charterstatus charterstate, vce(cluster charterstatus)
reg IEPS i.year i.charterstatus year#charterstatus, vce(cluster charterstatus)
However, I am not sure how to do this by districts in each state.
Thanks.
I am having trouble running pairwise DiD regressions for states. I am running Stata 15.1.
Thanks for all of your help so far.
I've downloaded my data from NCES, which is data by school year, and have merged all of my files.
After running the reshape long command, I have multiple variables, each from year 1 to 18, as defined by me.
I would like to begin DiD regressions. My dependent variable is the number of students in an IEP (variable is IEPS) for each year, and I am comparing states before and after charter school introduction. That is the dummy variable charterstate. charterstatus is a dummy variable if the district is a charter school district or not.
I am using Kentucky as my main state, and would like to run DiD with 7 neighboring states.
I want to run pairwise DiD regressions comparing districts in KY to districts in only one other state (like Ohio). Then I would like to repeat this for each other state in my sample (i.e., create separate samples with districts from KY and one other state).
So far, I have ran:
reg IEPS i.year i.charterstatus charterstate, vce(cluster charterstatus)
reg IEPS i.year i.charterstatus year#charterstatus, vce(cluster charterstatus)
However, I am not sure how to do this by districts in each state.
Thanks.
