Working Paper: NBER ID: w20184
Authors: James Andreoni; Andy Brownback
Abstract: We model contests with a fixed proportion of prizes, such as a grading curve, as all-pay auctions where higher effort weakly increases the likelihood of a prize. We find theoretical predictions for the effect of contest size on effort and test our predictions in a laboratory experiment that compares two-bidder auctions with one prize and 20-bidder auctions with ten prizes. Our results demonstrate that larger contests elicit lower effort by low-skilled students, but higher effort by high-skilled. Large contests also generate more accurate rankings of students and more accurate assignment of high grades to the high-skilled.
Keywords: No keywords provided
JEL Codes: C91; C92; D47; H52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Student Ability (G53) | Effort in Contests of Different Sizes (C72) |
| Contest Size (L25) | Effort of Low-Skilled Students (D29) |
| Contest Size (L25) | Effort of High-Skilled Students (D29) |
| Contest Size (L25) | Accurate Rankings of Students (A14) |
| Contest Size (L25) | Accurate Assignment of High Grades to High-Skilled Students (D29) |