Evaluation of Swedish Youth Labor Market Programs

The new deal for young people: effect on the labour market status of young men

“…The key examples are Brodaty et al (2001), Gerfin and Lechner (2002) and Larsson (2003) for France, Switzerland and Sweden respectively. The variation in the results of these studies is to be expected given differences in the target groups, the details of the programmes and the outcomes considered.…”

“…The variation in the results of these studies is to be expected given differences in the target groups, the details of the programmes and the outcomes considered. 2 Gerfin and Lechner (2002) and Larsson (2003) estimate that EMP-type programmes increase by 6-17 percentage points the chances of being employed one year after programme start relative FTET-type programmes. Gerfin and Lechner (2002) also find that, relative to job creation programmes similar to VS and ETF, EMP increases the probability of employment by 15-18 percentage points.…”

“…2 For example,Brodaty et al (2001) andLarsson (2003) consider programmes targeting young people whileGerfin and Lechner (2002) consider all ages.…”

Differential effects of active labour market programs for the unemployed

“…Microeconometric studies looking at the relative effects of different programs includeFrölich et al (2004), Larsson (2003 andCarling and Richardson (2004) for Sweden, and e.g Gerfin and Lechner (2002). andGerfin et al (2005) for Switzerland;Brodaty et al (2001) andBonnal et al (1997) for France, and the earlier work byRidder (1986) for the Netherlands.2 This choice of comparison state is driven by a institutional set-up where individuals will sooner or later join a program provided they are still unemployed.…”

On the inefficiency of propensity score matching

“…Frölich (u) SIAW, Universität St. Gallen,Bodanstr. 8,9000 St. Gallen, Switzerland e-mail: markus.froelich@unisg.ch 1 See e.g., Black and Smith (2004), , Gerfin and Lechner (2002), Heckman et al (1998a), Frölich et al (2004), Larsson (2003), Lechner (1999Lechner ( , 2002, Sianesi (2004) or Smith and Todd (2005). 2 Hahn (1998) showed this for the case of experimental data, where the propensity score is constant and not affected by the X variables.…”