Dynamic Estimation of the Incentive Schemes and Signalling Costs of Grade Inflation

Abstract

Higher education is a subject of continual interest. It is marked as an area to propel growth and equality, and public monies in the many billions of dollars are spent annually for its support. Over the past half century, there is evidence of grade infation within these universities and colleges. However, the potential and actual costs of this grade infation have been understudied, as well as the feedback effects of changes in the labor market. This paper ¯rst examines grading, enrollment, and quality trends in UCLA from 1980-2007. The data demonstrates that there is infation, even likely when controlling for improvement in student quality. The data also shows that there is a lot of movement in grading patterns over the period; some departments are choosing not to infate. I construct a stochastic dynamic model which demonstrates the tensions that encourages and discourages grade infation. It also shows that grade infation can cause higher variance in initial wages, which might induce some to pursue graduate degrees as an additional signal. The paper then provides suggestions of how it might be estimated with the proper data. Given that the UCLA data does not have all of the information necessary to estimate this model, the paper concludes by making suggestions for a model that can be constructed with the current data, with the intention of so doing.