Population Studies Centre, University of Waikato

2004-11-15

Room I1.09, University of Waikato

Meta-analysis constitutes a form of quantitative research synthesis originally developed in experimental medicine, later on extended to fields such as biomedicine and experimental behavioural sciences, specifically education and psychology. During the last decade, meta-analytic techniques have also been introduced to empirical economics, which is largely non-experimental. Growth in the number of applications has been particularly rapid in recent years and there are now more than 100 meta-analyses in economics published internationally. In this presentation I will focus on the issues that are particularly important in economic applications, such heterogeneity, quality variation across studies, statistical versus economic significance, 'within study' versus 'between study' variation, and publication bias. These issues are discussed by means of two examples from meta-analyses which I have conducted: one from microeconomics (the effect of the local unemployment rate on wages of individuals) and one from macroeconomics (the impact of government spending on economic growth).