Ebook Long-term impact of active labour market policy: evidence from a social experiment in Denmark
Most empirical evaluations of ALMPs look at short run effects on earnings, employment status some time after program start, or unemployment duration, and most of them report small and in some cases even negative effects on transition rates into employment, on employment status some time after the programme participation, or on subsequent earnings, see e.g. the reviews by Heckman et al. [1999], Kluve and Schmidt [2002], and Kluve [2006]. From the studies analysing effects on unemployment duration, the general finding is that for most programmes, the positive post-participation effect (increasing transition rates into employment) is dominated by a negative locking-in effect (reduced transition rates into employment during programme participation), such that the net effect of programme participation on unemployment duration is insignificant and in some cases even negative (Crépon, Ferracci and Foug`ere [2005], Munch and Skipper [2005]).
Recently, however, it has been acknowledged that compulsory ALMPs may affect job search behaviour before actual participation: the moment an unemployed worker realizes that there is a positive risk of having to participate in a programme in the future, the job seeker may increase his job search intensity or reduce his reservation wage in order to avoid programme participation (Black et al. (2003), Rosholm and Svarer (2008), Geerdsen and Holm (2007), Geerdsen (2006), Cockx and Dejemeppe (2007)).
It has also become increasing obvious that successful active labour market policy regimes contain more elements than just a set of activation policies. Crépon et al. (2005) evaluate the effects of intensive counseling schemes administered to French unemployed workers. The aim of these schemes is to improve the quality of assignment of workers to jobs. Some of these schemes significantly reduce unemployment duration, but they mostly improve the match quality in the sense that the time until unemployment recurrence is prolonged. Van den Berg and van der Klaauw (2006) use data from a social experiment to estimate the effect of counseling and monitoring on the transition rate to employment. If they find no evidence that counseling and monitoring affect the exit rate to work, they find that increased monitoring leads to a shift from informal to formal job search. Kjærsgaard et al. (2010) exploit random variation in the timing of planned meetings with case workers and find that job finding rates tend to increase during a period after the meeting has taken place.
Lalive et al. (2005) has looked at the effect of benefit sanctions in reducing the unemployment duration of unemployed Swiss workers. They focus on both ex ante (i.e. the effect of a warning) and ex post effects. They find that both warning and enforcement of benefit sanctions have large positive effects on the exit rate from unemployment. Svarer (2007) estimates ex post effects of sanctioning recipients of unemployment insurance, while Qureshi (2008) studies recipients of social assistance. Both find that benefits sanctions have very large positive effect on the exit rate from unemployment.
Summarising this selective literature review, there is a tendency towards finding positive effects of counseling and monitoring policies, although the evidence is not unanimous. With respect to threat effects and sanction, the evidence seems to be quite unidirectional; there are strong threat effects of monitoring and sanction policies as well as active labour market programmes, and administered sanctions have large effects. Hence, even if participation in activation programmes has negligible (or even negative) direct effects on participants, they may still be worthwhile due to their ex ante effects on job search behaviour. The discussion above implies that, even though direct participation effects are negative, there may still be positive overall effects on e.g. unemployment duration from having a system of active labour market policies.
Posted in :
