The focus of the present analysis is first to explain the different labour market performances of established OECD-countries and second to provide a theoretical and empirical contribution to the question how political and institutional variables can shape Public Policy-outcomes. To answer these questions, the paper analyzes unemployment and employment rates of OECD-21 from 1960-2007 by using pooled regression-analysis.
The broad focus is to explain a certain policy-outcome (labour market performance) by various political and institutional variables, like the partisan complexion of the government, veto players, and labour market institutions. However, it is unlikely to assume that these variables can affect labour market performance directly and by themselves: Governing parties, for example, apply certain policy instruments which can be linked to unemployment and employment. Concerning left-wing parties, one can argue that these instruments cover a Keynesian economic policy or higher spending on active labour market policies.
Therefore, two groups of independent variables will be analyzed: First, the “structural variables”, which consist of the political and institutional variables stated above (governing parties, veto players and labour market institutions). Second, the “policy variables” contain the concrete instruments or causal mechanisms, by which the structural variables affect the policy-outcome. The analyzed “policy variables” are mainly policy-output-variables like unemployment replacement rates, active labour market policies, employment protection and tax policy, but also some policy-outcome-variables like growth or interest rates.
In summary, the analysis acts from the assumption that structural and policy variables are able to affect labour market performance. However, they do not work independently from each other. Therefore, in the theoretical part, the causal links between these different variables will be described in detail to explore how political and institutional variables shape the analyzed policy-outcome. In the empirical analysis, models containing interaction-effects of the theoretically connected structural and policy variables are calculated.
