The pension-''frame in thought'' and its impact on willingness to reform. Evidence from a German survey.

Recent and unpopular reforms of state pension systems in Europe are often justified by referring to a twin demographic crisis that will unfold over the next decades. Politicians and experts alike claim that an increase in life expectancy will be accompanied by a decrease in the number of contributors, so that structural reforms are deemed necessary to keep the system running. However, little is known about whether the European publics find this argument convincing. Relying on the concept of framing, we argue that voters who mentally connect demographic developments with pension issues are more likely to support reforms than other members of the public. Usually, framing effects are tested by studying small and homogeneous samples within experimental research designs. While this research strategy is commendable and widely accepted, all too often the ensuing problem of external validity is ignored. Put differently, it is not clear whether effects that can be demonstrated in a laboratory setting to have any impact on the way in which ordinary citizens think about political problems in real-world situations. Therefore, our paper investigates an alternative approach to framing effects in the context of demographic change and pensions reforms by employing a quasi-experimental design to survey data. Our findings suggest that methods developed in the field of econometric evaluation research provide a useful complement to genuine experiments, and may substantiate the inferences from classic framing experiments. Based on a survey conducted in the German State of Hesse in 2008, the paper demonstrates that propensity score matching estimation techniques can provide strong and robust evidence for a positive effect of the pension-''frame in thought'' on voters'' likelihood to support pension reforms. Moreover (and subject to cross validation by pure experiments), our findings suggest that policy makers should emphasise the link between demographic developments and their consequences on state pension systems in order to increase support for unpopular reforms. Keywords: framing effects, propensity score matching, pension reforms