corporate and think-tank researcher

Donald Gutstein

24 Jun '11

Debunking the Fraser Institute’s Latest Crusade: Teacher Merit Pay

Most studies show scant link between student achievement and financial rewards for instructors.

Fresh from the triumph of successfully promoting its fallacious school report card, this time in Alberta, the Fraser Institute is already scheming to peg teacher pay to student test scores and create a market for teachers.

We should remember that the institute’s success with school rankings would not be possible without over-the-top support from the corporate media. The day the institute released its Alberta high school report — June 12 — the Calgary Herald made this event its front page story with the headline, “How does your school stack up?” The Herald gave the rankings eight pages in the B section, interspersed with ads for private schools which, of course, topped the charts.

Will the media assist the institute with its merit pay campaign?

They already are. The article accompanying the rankings featured the headline “Teacher evaluations remain contentious idea,” and asked if there is a better way to evaluate and pay teachers than the way we do now, which is to base pay on years of experience and advanced educational achievement.

“[T]he rankings naturally open debate about how quality teachers impact a school’s score,” the story claimed.

But there’s nothing natural about this debate. It is being pushed by the institute, which published a piece on merit pay in the May/June 2011 issue of Fraser Forum, timed to coincide with the Alberta report card.

Praise from University of Chicago economist

The Fraser Forum article, by Peter Cowley, director of school performance studies, and Neils Veldhuis, director of fiscal studies and vice-president of Canadian policy research, argues that since merit pay is common in the private and public sectors — think sales commissions, bonuses, piece work, team incentives and pay raises based on past success — why should teachers be exempt?

They then claim “there is solid evidence that properly designed merit pay systems can have positive affects” (sic — they mean “effects”).

But the evidence is far from solid.

They cite one review by University of Chicago economist Derek Neal, who found “strong suggestive evidence that total teacher effort rises following the introduction of performance pay.”

But was Neal — who framed his research as if schools were business firms — looking only for the strong suggestive evidence? One clue is that his work was supported financially by the Searle Freedom Trust, which “fosters research… that promote[s] individual freedom and economic liberty, while at the same time advancing a commitment to personal responsibility and a respect for traditional American values.” The foundation “encourages the transfer of responsibility from government to the private sector.”

Neal’s work on merit pay is right up the Searle Freedom Trust’s alley.

Cowley and Veldhuis also cite the work of Victor Lavy, who studied the impact of monetary incentives on teachers of English and math in Israeli secondary schools. Lavy did find a positive correlation with student test-taking rates, conditional pass rates and test scores.

But Pedro Martins, an economist at the University of London, cautions that because of the way Lavy structured his study, his results may not be generalized beyond the group of individuals or the setting used in the study. The results from Israel may not be relevant to Canada or anywhere else, for that matter.

So the evidence Cowley and Veldhuis offer is far from convincing.

The wider research

There is also a large literature that finds little or no correlation between merit pay and increased student achievement. One famous study of teacher performance-related pay in Portugal’s public schools found that student achievement actually went down when greater emphasis was placed on individual teacher performance.

Closer to home, studies in Texas, New York City and Chicago found little or no correlation between merit pay and improved student performance.

The Texas program distributed $10 million to 99 schools that turned in high scores on state tests despite enrolling large numbers of students from low-income families. A study done after three years of the program found it had a “weakly positive, negative or negligible effect on student test-score gains.”

A Columbia University study of a school-wide bonus scheme in New York City public schools found that teacher incentive pay had little effect on student achievement, teacher absenteeism and teacher quality.

And an analysis of a performance pay program in Chicago by Mathematica Policy Research found no evidence that the program boosted student achievement on math and reading tests compared with a group of similar non-participating schools.

It is certainly true that critics could find problems with these studies, since the relationship between teacher merit pay and student achievement is notoriously difficult to isolate.

Pay grade?

But Cowley and Veldhuis do not refer to these analyses, as they should have done as responsible researchers. The studies are not difficult to find.

Instead, they divert our attention from the missing reports by raising the red herring of “B.C. union leader Jim Sinclair” who “will almost certainly trot out predictable objections to any merit pay proposals.”


Perhaps it’s time for the Fraser Institute to implement its own program of merit pay for performance. Or perhaps it already has one in place. In 2009, the last year for which figures are available, Cowley’s total compensation was $115,790 (US), while Veldhuis took home $134,078.

But, one has to ask, do they get rewarded for skill at propaganda or proper research?

First published in The Tyee

Buy Now