Faculty of Economics and Business Administration Publications Database

Estimating the macroeconomic effects of active labour market policies using spatial econometric methods

Authors:
Rodrigues, Paulo M.M.
Wolf, Katja
Source:
Volume: 30
Number: 7
Pages: 648 - 671
ISSN-Print: 0143-7720
Link External Source: Online Version
Year: 2009
Keywords: Data analysis; Economic policy; Germany; Labour market
Abstract:

Purpose – The paper aims to present an analysis of the indirect and direct effects of active labour market policy measures at the regional level for Western Germany. Design/methodology/approach – Most evaluation studies of active labour market policy focus on the microeconometric treatment effect using individual data and do not account for possible indirect effects like deadweight and substitution effects. The present study uses a dynamic specification of the augmented matching function at the regional level. A dynamic panel data model is estimated using monthly and regional variation of different labour market programmes as explanatory variables. Furthermore, spatial interactions are taken into account by adding a spatially correlated error term. Findings – Almost no significant negative effects are found of the stock of participants in programmes of labour market policy on the number of outflows from unemployment into regular jobs. Thus, contrary to findings at the individual level, no lock-in effect is found. The number of programme participants does not reduce the number of outflows from unemployment. On the other hand when looking not at the stocks but on the outflows from programmes, no positive effects on outflows from unemployment at the regional level are found. Research limitations/implications – Because of data limitations only a period up to six months after completing a programme is used. Originality/value – The authors distinguish between the effects of the stock of programme participants and of the outflows from programmes. Furthermore, the authors account for spatially correlated error terms by using a GM estimator proposed by Mutl in 2006.

back