Posted by Howard 

Do teachers love to play PowerBall? Or do teachers prefer to gamble on the small chance of a high financial reward? While it doesn’t fit the risk-averse, altruistic personality archetype of teachers, that, in a nutshell, is the conclusion of the paper, "Individual Teacher Incentives And Student Performance" by David Figlio and Lawrence Kenny.  The first draft of the paper was completed in 2003, and the revised version, available at the National Bureau of Economic Research's Web site since October, is just now starting to get media interest.

Figlio is an alumnus of my graduate school, the Economics Department at the University of Wisconsin, where other economists with an interest in education--such as Larry Mishel, Leanna Stiefel, Robert Berne and, alas, Mike Podgursky--have emerged.  He is not a partisan, so his paper merits a close reading.  The paper is:

• About old-fashioned merit pay (which we sometimes call crony-pay at the AFT), not paying teachers for student achievement growth.
• Not a sophisticated panel study of student achievement data in Florida of the sort we have been seeing from Figlio.
• Focused on high schools.

The paper also shows that:

• Schools that did not have sufficiently high financial rewards restricted to a low percentage of teachers were not counted as merit pay schools and became pat of the statistical control group.
• Catholic schools did not use financial incentives any more than public schools.
• As has been adequately covered in the media, the data are old and patched together, ranging from NELS eighth grade achievement data in 1989 to the author's merit-pay survey instrument in 2000 that had a low response rate.
• Private schools, which are included in the study, responded in disproportionately high rates to the survey and are responsible for the few favorable results in the study.
• In models using public school data only, merit pay policies did not have a statistically significant effect.
• In the combined public and private school sample, merit pay resulted in only a 1 to 2 percent higher test score--not annually but after four years of high school.
• The effect of merit pay was smaller in the public school sample and not statistically significant.
• Public schools are a little more likely than non-Catholic private schools to evaluate teachers annually.
• One in four public schools reported dismissing at least one probationary teacher in a three-year period, and 15 percent of public schools reported dismissing one or more experienced teachers in the previous 3 years, which contrasts with popular claims that it is impossible to fire a poorly-performing teacher.

The teachers I know who work in districts with performance pay say that their work with challenging kids cannot be fairly measured within  performance pay plans and that, consequently, they don't expect to reap the financial rewards of such plans.  However, they remain committed to teaching these students--they don't sound like PowerBall players to me.