Performance-based management is increasingly common in public sectors. This trend is particularly evident within education where accountability has been the hallmark of recent reforms. In the United States, for example, the No Child Left Behind Act has markedly raised the costs of poor test results for administrators, principals, and teachers. One important feature of accountability reforms is the introduction of funding related to performance indicators.

Accountability reforms have been promoted by the notion that increasing incentives for principals and teachers improves the performance of schools. However, a concern for dysfunctional responses has been raised in relation to performance measurement. Schools may respond to incentives by “teaching-to-the-test”¯i.e., focusing on test-specific skills¯(Jacob 2005), “teaching-to-the-rating”¯i.e., focusing on students whose performance is important for funding¯(Reback 2008), or intentionally manipulating standardized test scores (Jacob & Levitt 2003).

This study examines whether introducing a performance funding system into public high schools¯that directly links school funding to enrollment and final passing rates¯causes schools to behave strategically by grading their students more generously. As admission to university programs is contingent on the student's high school grade point average (GPA), high school grades are of great importance to the student. Therefore, to make the school more attractive, teachers have an incentive to give better grades.

We use the gradual adoption of performance funding across different kinds of Danish high schools to identify the effect of performance funding on grading behavior. Furthermore, we exploit that Danish high school students get grades both internally and externally evaluated (within the same subject) that both count in the final GPA score. Written exams are standardized tests, evaluated by two independent external examiners, whereas the annual classroom performance is graded by the student’s teacher. Thus the school itself can only affect the grade of the classroom performance; not of the written exam. By calculating the difference between grades evaluated externally and internally¯and studying the change in the difference before and after the introduction of performance funding¯we test whether the introduction of the performance funding induced teachers to grade students more generously. We use data from Danish administrative registers containing information on all high school grades for all students graduating from a Danish high school in the period 2005-2013.

We find that schools responded to the introduction of performance funding by grading students more generously. Thus the difference between externally and internally given grades increases significantly after the introduction of performance funding. Furthermore, the more generous grading was particularly dominant among schools in which teachers on average reported that they felt a high degree of conflict between high quality teaching and achieving low dropout rates.