A behavioral economic approach to multiple job holdings with leisure

Financial constraints or economic needs, career development, psychological satisfaction as well as demographic and situational factors cause workers to seek more than one job while enjoying leisure time. In this paper we examine how a worker with prospect theory type of preferences allocates her time between leisure, a safe job and a risky job. Optimal time allocation for a sufficient loss averse worker depends on the reference level which in turn determines whether the worker is willing to experience relative losses or not. When the reference level is relatively low then the sufficiently loss averse worker will allocate some of her time to leisure and will hold both jobs in order to diversify risk and reduce income loss arising from the risky job. However, if the probability of a good state of nature is very high and the reference level is very low, the worker spends time only on leisure and the risky job while avoids the safe job. Loss aversion does not affect the optimal time allocation to the three activities as the time allocation results in avoiding relative losses for any state of nature. When the reference level is relative high, but not too high, the worker
will allocate her time between both safe and risky jobs as well as to the leisure. Worker with very high reference level will avoid the safe job and will divide her time between the risky job and the leisure. In both cases the worker is willing to accept relative losses in the bad state of nature provided it is compensated with relative gains in the good state of nature. Here the allocation of time to the three activities depends on the degree of loss aversion. When the reference level is relatively low, but not too low, an increase in the reference level will reduce leisure time, reduce time in the risky job and increase time in the safe job. At very low reference levels, an increase in the reference level will result in the worker re-allocating her time from leisure to the risky job assuming the probability of a good state of nature is higher than a threshold. When the reference level is high the opposite effects are observed. We also examine other comparative statics including the effect of changes in the wage rate.