Teacher incentives gone wrong

In an attempt to address low performance levels of Portuguese students, a new law came into effect in 2007 to improve teacher incentives.  Up until then, career progress in the public-school system was primarily based on seniority. The new law created two separate pay scales, with salary differences of 25% or more; and set out promotion rules based on teacher performance. In particular, student performance — measured by test scores at the school and national levels — featured as an important component of teacher performance. 

A recent paper assesses the effects of the 2007 incentive scheme on student performance. Methodologically, the study follows a difference-in-differences approach. Since the new law applied to public schools in continental Portugal, two control groups were considered: private schools and public schools in the Madeira and Azores regions.

At an aggregate level, the figure below shows the evolution of test scores following the introduction of the new law. In the legend, “treated” refers to public schools in the continent.

In order to get a more precise estimate of the effects of the 2007 law, the paper draws on matched student-school data covering the population of secondary school students who took national exams from 2002 to 2011, a total of about 1.5 million observations. Confirming the aggregate features present in the above figure, the authors find evidence of significant unintended consequences of the new incentive scheme. In particular, they estimate a significant decline in student achievement as measured by national scores. Although the point estimates vary somewhat from estimation to estimation, the diff-diff coefficient implied by the new law is always statistically significant at the 1% level. When private schools are used as control, the size of the treatment coefficient hovers at around 8 tenths of a point, which corresponds to about 7% of the national average score.

The authors also find evidence of grade inflation caused by the incentive scheme, as measured by the increased gap between school-level and national exams. Finally, the diff-diff coefficient is particularly high in schools where there is a limited number of slots for promotion (whereby promotion is effectively a contest).

All in all, the authors offer a cautionary tale on the unintended consequences of incentive provision. The results on grade inflation suggest that the problem is not so much the provision of incentives as the possible perverse effects of incentive provision.

Pedro S. Martins & João R. Ferreira

Effects of individual incentive reforms in the public sector: the case of teachers

Public Choice (2025)