Myopia, pension payments and retirement: an experimental approach

The behavioral economics literature on time discounting has suggested that individuals may systematically undersave when planning for retirement. Hence, pension systems have developed to enable, or indeed force, individuals to save more for retirement. Of course, the saving aspect and the timing of retirement are connected, in the sense that the expected length of retirement determines what is meant by adequate post-retirement resources, and vice versa. Despite this, the timing aspect rarely enters into policy discussions, although the same behavioural phenomena that lead to undersaving – in this paper, myopia and present bias – may also have implications for the retirement decision. Moreover, the form of pension payments may also affect the timing decision when individuals do not have time consistent preferences. This paper presents a model of saving and retirement timing where saving rates are mandated, and pension payment may come in either a lump-sum or an annuity. It tests the model using data collected through a new experiment. The experiment presented has a particular novel feature which made it uniquely suited for testing the theoretical model. Specifically, participants in the experiment came back to the laboratory on a weekly basis over a two month period. This decision to return to the laboratory (or, to leave the experiment and collect a pension) became in itself the main variable of interest. The experiment therefore exploited the effort it takes for participants to come to the laboratory to capture preferences over time-use and leisure. The results shown that plans over leaving the experiment tend not to reflect preferences, whilst actual leaving times were lower for more impulsive individuals and those who gave up more time to participate. This suggests a tradeoff between increasing saving through pension systems and earlier retirement. Payment group had no effect on retirement timing, most likely because the small rewards meant participants were indifferent between the two forms of payment. The results suggest individuals may have time-inconsistent preferences over leisure choices, leading to the incidences of unplanned early retirement.