Abstract
Using high frequency data for Israel it is found that trainees exit unemployment at a much faster rate than comparable unemployed who did not receive training. It is argued that this powerful training effect is unlikely to be simply the result of positive self-selection of trainees. The training effect is detected when using nonparametric estimation methods. Parametric methods fail to pick it up because of a very large exit spike in the hazard function which occurs when the training is completed. Other findings include the following: the duration of unemployment varies directly with the duration of entitlement to unemployment benefit as well as the replacement ratio. Also, the exit hazard from unemployment rises as the benefit entitlement period draws to a close.
Original language | English |
---|---|
Pages (from-to) | 935-946 |
Number of pages | 12 |
Journal | Applied Economics |
Volume | 28 |
Issue number | 8 |
DOIs | |
State | Published - Aug 1996 |