A recent NBER working paper by Roland Fryer describes school-based randomized trials in 261 urban public schools in Chicago, Dallas, New York City, and Washington DC that distributed financial incentives worth $6.3 mn to roughly 38000 students so as to test the impact of financial incentives for various inputs and outputs on student achievement and finds that
"In stark contrast to simple economic models, our results suggest that student incentives increase achievement when the rewards are given for inputs to the educational production function, but incentives tied to output are not effective. Relative to popular education reforms of the past few decades, student incentives based on inputs produce similar gains in achievement at lower costs. Qualitative data suggest that incentives for inputs may be more effective because students do not know the educational production function, and thus have little clue how to turn their excitement about rewards into achievement."
The experiments varied from city to city on several dimensions - what was rewarded, how often students were given incentives, the grade levels that participated, and the magnitude of the rewards. The key features of each experiment consisted of monetary payments to students (directly deposited into bank accounts opened for each student or paid by check to the student) for performance in school according to a simple incentive scheme. In all cities except Dallas, where students were paid three times a year, payments were disseminated to students within days of verifying their achievement. Students were paid for inputs like reading a book (or lesson), attendance, good behavior, wearing their uniforms, and turning in their homework.
Prof Fryer finds that providing incentives for achievement test scores has no effect on any form of achievement measured; there is scant evidence that total effort increased in response to the programs, though there may be substitution between tasks; also no evidence that incentives decrease intrinsic motivation; and that input experiments seem positively associated with motivation and output experiments seem negative. However, the point estimates on all these are too small and the standard errors are too large to conclude that financial incentives are a game-changer in educational reforms. He concludes
"Relative to achievement-increasing education reform in the past few decades – Head Start, lowering class size, bonuses for effective teachers to teach in high need schools – student incentives for certain inputs provide similar results at lower cost. Yet, incentives alone, like these other reforms, are not powerful enough to close the achievement gap."
About the theory that students do not understand the educational production function and, thus, lack the know-how to translate their excitement about the incentive structure into measurable output, he writes,
"Students who were paid to read books, attend class, or behave well did not need to know how the vector of potential inputs relates to output, they simply needed to know how to read, make it to class, or sit still long enough to collect their short-term incentive...
There are three pieces of evidence that support this theory. First, evidence from our qualitative team found consistent narratives suggesting that the typical student was elated by the incentive but did not know how to turn that excitement into achievement. Second, focus groups in Chicago confirmed this result; students had only a vague idea how to increase their grades. Third, there is evidence to suggest that some students – especially those who are in the bottom of the performance distribution – do not understand the production function well enough to properly assess their own performance, let alone know how to improve it.
Three other theories are also consistent with the experimental data. It is plausible that students know the production function, but that they lack self-control or have other behavioral tendencies that prevent them from planning ahead and taking the intermediate steps necessary to increase the likelihood of a high test score in the future. A second competing theory is that the educational production function is very noisy and students are sufficiently risk averse to make the investment not worthwhile. A final theory that fits our set of facts is one in which complementary inputs (effective parents, e.g.) are responsible for the differences across experiments."