causal

Paying for inputs works, outputs don't

When students were paid for outputs (test scores, grades), the incentive experiments produced no measurable effect, but when students were paid for inputs—behaviors known to drive achievement like reading books or doing homework—all three experiments produced statistically significant gains.

causalpending

Speaker

Roland Fryer

Evidence Quote

“when you pay for outputs, we didn't get any results at all. Pure zeros. Okay, when we pay for inputs, those experiments, all three of them worked and were statistically significant.”

Source

Roland Fryer on Educational Reform— EconTalk
Created: 6/13/2026, 12:20:02 AM

Your notes

Loading notes...