Happy Birthday? Manipulation and Selection in Unemployment Insurance

About this event

This paper documents strategic delays in the timing of layoffs around an age-at-layoff threshold entitling workers to a four months increase in potential UI benefit duration in Italy. Manipulation is quantitatively important and of far-reaching consequences: 12.5\% of layoffs in the manipulation region are strategically delayed with affected workers increasing their average UI benefit receipt and non-employment duration by 3.0 and 7.1 months, respectively. Manipulators are highly selected on both observables and unobservables. Manipulation is most prevalent among female, part-time, white-collar workers with longer job tenure at small firms. We show that compared to non-manipulators: (i) manipulators would have had higher exit rates from unemployment at each point during the unemployment spell, and (ii) manipulators' benefit receipt duration responds 25\% stronger to the increase in potential benefit duration. We study the welfare consequences of our results and discuss implications for standard RD-estimates.