|
Labor provision when individual incentives are barely predictable : An analysis of the behavior of taxi drivers| old_uid | 20278 |
|---|
| title | Labor provision when individual incentives are barely predictable : An analysis of the behavior of taxi drivers |
|---|
| start_date | 2022/03/25 |
|---|
| schedule | 11h15-12h30 |
|---|
| online | no |
|---|
| location_info | zoom |
|---|
| summary | A fundamental assumption of expected utility models is that agents make predictions by formulating rational expectations. However, such predictions might not always be feasible, exposing the agents to considerable uncertainty. Using data from 11 million taxi trips in Hamburg (Germany), we show that traditional econometric analyses as well as methods from machine learning show that incentives for labor provision at the individual level are barely predictable. Under such uncertainty, a competitive test between utility and heuristic satisficing models shows that heuristic models predict 99.5 percent of drivers best, utility models best predict behavior of 0.5 percent of drivers. Heuristic models do not require calculating expected earnings but terminate shifts when reaching an aspiration level on shift duration or earnings. Co-authors : Gerd Gigerenzer, Perke Jacobs |
|---|
| responsibles | Le Lec, Laslier |
|---|
| |
|