Working Paper: CEPR ID: DP6106
Authors: Pietro Garibaldi; Francesco Giavazzi; Andrea Ichino; Enrico Rettore
Abstract: For many students throughout the world the time to obtain an academic degree extends beyond the normal completion time while college tuition is typically constant during the years of enrollment. In particular, it does not increase when a student remains in a program beyond the normal completion time. Using a Regression Discontinuity Design on data from Bocconi University in Italy, this paper shows that a tuition increase of 1,000 euro in the last year of studies would reduce the probability of late graduation by 6.1 percentage points with respect to a benchmark average probability of 80%. We conclude suggesting that an upward sloping tuition profile is efficient in situations in which effort is suboptimally supplied, for instance in the presence of public subsidies to education, congestion externalities and/or peer effects.
Keywords: regression discontinuity; students performance; tuition
JEL Codes: C31; I2
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Tuition increases (I22) | Probability of late graduation (C41) |
| Higher tuition raises effort in previous years (D29) | Probability of late graduation (C41) |
| Fourth-year tuition (A00) | Tuition incurred if student extends studies (I22) |
| Tuition increases (I22) | Dropout rates (I21) |
| Tuition increases (I22) | Students' performance (final graduation marks) (D29) |
| Upward sloping tuition profile (I21) | Educational efficiency (I21) |