We use a sequential model of school competition, featuring endogenous investment, grading, and evaluation, to study the interaction between a school’s investment decision and its choice of grading policy. When schools control information about student ability, they choose grading strategies that do not perfectly reveal student types to evaluators. However, they also invest more to increase education quality than in a benchmark case where all information about ability is observed by evaluators. The increased investment has the potential to outweigh the cost of noisy grading. Viewed from a broad perspective, our analysis considers competing institutions that make (real) investments to improve stochastic outcomes and also exert influence over the release of information about outcome realizations.
Our focus is on competition between schools, similar considerations are important in a variety of other settings. For example, a division within a firm may exert effort to develop a prototype and simultaneously control the way in which the prototype is tested prior to undertaking a development decision. Competing for promotion, a number of employees may exert effort to improve their output, while simultaneously controlling the flow of information to superiors within the firm. In finance, firms make investments and exert effort to generate profits for shareholders, but have the capability to influence the information available to investors about their performance by controlling the way in which earnings are disclosed.
In the context of education, our results have novel implications for both the policy discussion and common industry practices. We highlight a potential benefit of noisy grading—that it leads to greater investment in education quality—and show that the benefit can dominate the costs associated with information loss. This suggests that grade inflation may not be as bad for welfare as popular discussion suggests. In fact, it may lead to better outcomes than would arise if student ability is perfectly observed. By implication, the results suggest that policies or practices that convey information about graduate ability can reduce investment in the quality of education, decreasing student ability. Licensing and entrance exams provide employers and universities with information about applicant ability. To the extent that performance on these tests is an indicator of ability, graduate school entrance exams (such as the LSAT, GRE, MCAT, and GMAT) and state licensing exams in various industries may undermine the incentives of colleges and universities to invest in quality. College entrance exams like the SAT and ACT as well as state assessment exams may have similar unintended consequences for primary and secondary education. Paradoxically, the use of such exams may reduce investment in education quality and graduate ability.
Boleslavsky, Raphael and Christopher Cotton. 2015. “Grading Standards and Education Quality.” American Economic Journal: Microeconomics, 7(2): 248-79.