C:\cygwininstall\home\root\archive\docs\$ASQ2005-02-18_isa_proceeding_58866.docLast printed 3/4/2005 4:07 PMLast saved by
armingeon
9
passive policies before investing in policies that may work or may not work and that
frequently show effects only in the medium to long run.
Comparative research on the political determinants of ALMP found systematic
evidence for political variables too. ALMP originally has been a social democratic strategy in
particular in the Scandinavic nations. It comes as no surprise that left governments spend more
on activation (Janoski 1994). In a recent contribution Martin and Swank (2004) argued that
employer organization is a major determinant both for size and efficiency of ALMP spending.
According to Martin and Swank the more employers are coordinated the more likely is their
participation in and support of ALMP policies.
3 Research Design
The dependent variables of this analysis are ALMP effort and efficiency. Frequently
ALMP effort is measured as spending on active measures in % of GDP. But this does not take
into account that generally spending for labor market policies increases with the level of
unemployment and it does not take into account the relative size of passive in comparison to
active policies. Therefore in this analysis effort is measured as active spending per person
unemployed relative to GDP. However, countries that score high on this measure often have
large spending on passive measures too. Therefore as a second indicator I use the share of
ALMP spending (% GDP) in total labor market spending (% GDP) (cf. OECD Employment
Outlook 2003: 193/194.) Both indicators are correlated, but the correlation is far from perfect.
Figure 1 shows averages for both operationalizations by countries, 1985-2002. Sweden and
Norway are the most active countries while the Anglo-Saxon nations score low on these
measures.
Convention