Life-Cycle Responses to Pension Reform: The Role of Subjective Policy Beliefs
PDF: This study quantifies the impact of Statutory Retirement Age (SRA) reforms on individual behavior and welfare in the presence of subjective policy uncertainty and misinformation. We estimate a rich structural life-cycle model with forward-looking agents and dynamic policy belief updating. The model accounts for key determinants of retirement timing, precautionary savings, and old-age labor supply. In the model, agents have probabilistic expectations about the future evolution of the SRA and are misinformed about the penalty for early retirement (ERP). We derive subjective policy beliefs from new survey data from the German Socio-Economic Panel Innovation Sample (SOEP-IS). Our results show that uncertainty and misinformation reduce individual welfare and lead to agents overestimating the welfare cost of future reforms. However, these beliefs can support policy objectives. SRA reforms can lead to forward-looking reductions in savings and labor supply due to expected extensions of the working life. Uncertainty attenuates these anticipatory responses by 20-50 percent while maintaining effectiveness at the retirement margin. Eliminating misinformation would cause individuals to retire 1.3 years earlier and reduce lifetime labor supply by 2.4 percent, primarily among low-educated men.PDFPublic Appeals and Collective Crisis Mitigation
(Revise and resubmit at Journal of Economic Behavior and Organization)
PDF: Arrivals of crises often trigger public appeals from policy leaders, attempting to encourage crisis-mitigating behaviors. We ask whether the tone of an appeal changes its effectiveness, and to what extent policymakers know what tone to use. Using a controlled experiment in a large, general-population sample, we first study the impact of appeals and of their emotional tone on contributions to a well-defined crisis mitigation effort. Two equivalent appeals have either positive-tone or negative-tone wordings, and both increase contributions by about 20% compared to no appeal. Next, a sample of policymakers (n=88) is presented with our design and asked to predict the outcome. Although they correctly predict the impact of the positive appeal, they substantially underestimate the effectiveness of the negative one. PDF