Financial Incentives and Retirement Decisions: Experimental Evidence

本研究在实验室实验中分析了财政激励对个人退休进入决策的影响。这一特定领域的经验证据是独一无二的，并提供了在受控实验室条件下对现收现付养老金制度的最佳设计的科学研究结果。该项目的主要研究问题可以概括如下。在第一步中，对退休时间方面的个人行为反应进行量化。在受控的实验室条件下，这些影响可以通过比较不同奖励计划的治疗组来衡量。第二，分析了现收现付制养老金制度中与财政激励有关的经济知识。理解并有能力进行定量计算，尤其是确定预期未来退休收入的现值，可能对退休时间产生重大影响。第三，分析了与风险规避有关的方面。在这种情况下，我们调查退休决策在多大程度上受到风险厌恶的影响，以及精算中性的养老金制度是否必须考虑到风险厌恶相应。如果我们观察到尽管精算中立，但在最早的允许退休年龄退休条目的聚集，这表明个人除了对消费和休闲的偏好之外，还害怕相对较早的死亡，因此领取养老金的时间相对较短。第四，分析了连续性与非经常性退休条目决定的差异。在德国公共养老金制度的现实中，个人通常可以自由地重新评估退休进入决定，直到他们真正发生，因此我们观察到一个顺序结构。因此，第四个研究问题的实质是方法论性质的。如果退休行为不受决策结构的影响，实验室实验是可行的，在低得多的费用，使用非经常性的退休进入decisions.By变化的单个参数，实验室实验能够隔离的影响，具体的政策对个人的决策。本研究填补了这一研究空白，为现收现付制养老金制度的优化设计提供了新的发现，并可能对退休年龄的相关政策设计产生启示。

This study analyses the impact of financial incentives on individual retirement entry decisions in a laboratory experiment. Empirical evidence in this specific field is unique and provides access to scientific findings on the optimal design of Pay-As-You-Go pension systems under controlled laboratory conditions. The central research questions of this project can be summarised as follows. In a first step, individual behavioural responses in the timing of retirement with respect to financial incentives are quantified. Under controlled laboratory conditions, those effects can be measured by comparisons between treatment groups that are subject to diverse incentive schemes. Second, economic literacy with respect to financial incentives in a Pay-As-You-Go pension system is analysed. The understanding and capability of conducting quantitative calculations and, most notably, determining the present value of expected future retirement income, may have a large impact on the timing of retirement. Third, aspects that are related to risk aversion are analysed. In this context we investigate to what extent retirement decisions are influenced by risk aversion and whether a actuarially neutral pension system must take into account risk aversion correspondingly. If we observe clustering of retirement entries at the earliest permissible retirement age despite actuarial neutrality, this indicates that individuals dread relatively early deceases and thus a relatively short period of pension benefit receipt in addition to preferences over consumption and leisure. Fourth, differences in sequential versus non-recurring retirement entry decisions are analysed. In the reality of the German public pension system, individuals are generally free to reevaluate retirement entry decisions until they really take place and therefore we observe a sequential structure. Therefore, the essence of the fourth research question is of methodological nature. If retirement behaviour is not influenced by the decision structure, laboratory experiments are feasible at much lower expenses, using non-recurring retirement entry decisions.By variation of single parameters, laboratory experiments are able to isolate the effects of specific policies on individual decisions making. This study fills this research gap by providing new findings for the optimal design of a Pay-As-You-Go pension system and it may as well generate implications for relevant policy design of the retirement age.