pull down to refresh
I don’t doubt that the data quality is great. It’s just that you can’t infer from simple averages what the returns are.
In order to estimate returns, you also need a measure of each student’s counterfactual earnings, which is very difficult.
reply
Who is the reference group? The charts say "people who didn't enroll". The footnote says "data from public colleges". I can only assume that the reference group is "people who didn't enroll in a public college"?
That seems like a mish-mash of a lot of different kinds of people.
reply
reply
I thought they were just looking at earnings minus financial cost of college for those who attended, which has a great many shortcomings.
reply
Here is the full report, which covered roughly 1 million students for 10 years and goes into detail about how the benefits vary. It highlights the type of degree, major, quality of the institutions, demographics of the students, and whether students complete the programs.
It's by no means perfect, nothing like this ever will be, but it is another sign.