Are there benefits to paying students for good performance in school?

I cheated that system pretty savagely - I would skim a book, look up cliff notes on the internet, and then pass the test. I can't say I really enjoyed many of the books I "read", nor remember much about them.

In my opinion, any sort of extrinsic "performance"-based reward system is just going to lead to optimizing to the system (rather than learning) and to the system being gamed.

There are a couple of options for students from homes with no money to make thousands of dollars at the price of significant stress. Would you prefer they try one of the other ones?

"Our children are too stressed" is a high-class problem in every sense of the word.

This entices the best and brightest to stick around and do some research for the university. Seemed to work pretty well too. I can think of one example off-hand who stuck around and did some amazing research for CSIRO and the CS department. After finishing he was hired by Google Australia as the lead for the first release of Google Maps.

It seems like scientific studies often ignore this factor because it's hard to measure. So it's no surprise there is conflicted data.