Paying Cash for Better Test Scores is no Silver Bullet

Wednesday June 2, 2010

One of the more zany ideas to improve student achievement is to pay students cash to improve their results. For economists, however, this is a no brainer – cash incentives always work.

ANU Economics Professor, Andrew Leigh, who was recently pre-selected as Labor’s candidate for the seat of Canberra in the next federal election, says that cash payments to improve test scores should be trialled in Australia. He says providing cash incentives for students to improve their results could be tried in any low-income school in the country, but that the most likely place to start would be for Indigenous students.

Several trials have been conducted in the United States over the past few years, including one in New York City which was whole-heartedly endorsed by Julia Gillard’s mentor, Joel Klein, the city’s schools chancellor.

The trials were organised by Professor Roland Fryer, an economist at Harvard University, and involved 38,000 students in 261 public schools in New York, Chicago, Dallas and Washington, DC. Over 80% of the students were from low income families and nearly 90% were black or Hispanic. Over $6 million was paid out in incentives to students.

Students in New York and Chicago were paid for better test scores or grades while those in Dallas and Washington, DC were paid for activities which may contribute to better test scores such as reading books, increasing attendance and better behaviour. In each city, about half the participating schools were randomly selected to be ones where students received money. The progress of these students was compared with that of their peers in the other participating schools.

In New York City, the program was trialled in over 60 schools. Students were given cash incentives for their performance on six computerized exams (three in reading and three in math) as well as four predictive assessments that were pencil and paper tests. For each test, fourth grade students earned $25 for a perfect score and could make up to $250. Seventh grade students received $50 for a perfect score and could earn up to $500. Fourth graders also received $5 for taking each test and seventh-graders got $10.

In Chicago, ninth grade students were rewarded on a sliding scale for good grades in five courses, including English, maths and science. Getting an “A” was worth $50; a “D” meant no money. In theory a student could earn up to $250 every five weeks and $2,000 in a year.

In Dallas, second grade students were paid $2 per book to read up to 20 books per semester and do a computer-based comprehension test to prove they had read the books. In Washington DC, sixth, seventh, and eighth grade students were paid for improved attendance, behaviour, wearing school uniform, handing in homework and for class work. Students were paid on a point system and could earn up to $100 a fortnight.

The results were completely surprising according to Professor Fryer. Remarkably, he says, in a paper published recently by the National Bureau of Economic Research, the cash incentives did not increase students’ test results in Chicago or New York. In each case, the effect of the incentives was statistically zero or very small. The paper concludes that paying students to increase test scores or grades does not work. It says that “incentives are by no means a silver bullet”.

Similarly, the incentives paid to students in Washington DC to increase attendance, behaviour and school work had a negligible effect on test scores. Attendance, behaviour and school work improved but there was only a marginal improvement in reading and no statistically significant effect on mathematics tests scores.

The payments to read more books in Dallas schools were more successful. Paying second grade students to read books had a relatively large effect on their reading comprehension, a more modest effect on their language scores, and a statistically insignificant effect on vocabulary.

From this, Professor Fryer concludes that providing cash incentives for “inputs” into student results is a better option “because students do not know the educational production function, and thus have little clue how to turn their excitement about rewards into achievement”. He further concludes that paying students to read more books is a more cost effective approach to improving reading than reducing class sizes in the early years, paying teachers more to work in high need schools and early childhood programs.

This is a very ambitious conclusion on the basis of one very small project involving just 22 schools. But, apart from this, it may not even be cost effective itself. Students can be encouraged to love reading with the right support from teachers and parents at no extra cost. Altruistic incentives seem to have positive results, as witnessed by the various schemes operating in several states in Australia whereby sponsors make monetary contributions to charities based on how many books children read.

Given these results, it can only be hoped that Professor Leigh does not take his proposal to Julia Gillard. Otherwise she might be inclined to take on another failed scheme from her mentor, Joel Klein.

Trevor Cobbold

Ronald Fryer. Financial Incentives and Student Achievement: Evidence from Randomized Trials, Working Paper No. 15898, National Bureau of Economic Research, April 2010.

1. The Andrew Leigh aspect to this story is interesting.

   During Julie Bishop’s push for ‘performance pay for teachers’ she produced a paper written by, yes, Andrew Leigh of ANU to support her policy. This paper was not published in the academic sense (E.g. had not been peer reviewed) but instead he had written it and put it on his web-site/blog.

   This ‘paper’ was then comprehensively trashed by Lawrence Ingvarson of ACER in an article in the melbourne Age. I’ve seen several of his ‘articles’ (none of which appear in the academic press) on education. His approach is to reduce education down to a monetary value.

   With this sort of history I would have been less surprised to hear he had been preselected for the Liberal Party, especially in it’s current state.

   Quite frankly, he is dangerous. He knows just enough to make a seemingly plausible argument, as he is an academic (not in education) therefore brings some creditability to the argument, but when you actually examine his argument you find only a facade.

   
   — David Hardie    Jun 3, 09:32 AM    #

Commenting is closed for this article.

Previous Next