Are Retirement Decisions Vulnerable to Framing Effects? Empirical Evidence from NL and the US

Abstract

This study investigates whether individual choices in the pension domain are vulnerable to the way alternatives are communicated to respondents. The analysis is based on a set of hypothetical questions posed in the DNB Household Survey as well as in the RAND American Life Panel on pension premium contributions and pension savings investment profiles. The design of the questions presented to the respondent in several alternative ways allows to test for the potential role of framing effects, as well as order and choice set effects. We find that framing has a significant and robust impact on individuals decisions. The effect is particularly strong for the alternative labeled as ‘standard’ option. In contrast, the answer categories order does not seem to be always significantly relevant. We also find that hypothetical preferences are consistent with the individual risk profile and actual portfolio allocations. The findings suggest that the presence of framing effects is strongly correlated with the complexity of decisions to be made and highlight the importance of communication with respect to retirement decisions.

Keywords: Framing effects, Individual decision making, Retirement behavior and pensions

JEL Classification: C5, C9, D12, G11