Willingness to wait covaries with endogenous variation in cortisol

Abstract

It has been argued that one dimension of the cycle of poverty is that poverty is a state of chronic stress and that chronic stress impairs decision-making. These poor decisions, made under chronic stress, might include carrying high-interest loans, failure to buy health insurance, gambling or drug use. As such, these decisions can contribute to the cycle of poverty. More specifically, a few studies suggest that increased stress may lead to more risk-aversion and steeper delay-discounting. While the deleterious effects of chronic stress on brain function are well established, much less is known about how chronic stress influences financial decision making specifically. Here, in a longitudinal design within six weeks period we aimed to incorporate biological mechanisms to improve our understanding of how stress influences economic decisions. We used a combination of decision-making tasks, questionnaires, saliva and hair samples within-subject (N=41). We assessed time and risk preferences using hierarchical Bayesian techniques to both pool data and allow heterogeneity in decision making and compared those to cortisol levels and self-reported stress. We found only weak links between endogenous variation in stress and model-based estimates of risk and time preferences. In particular, we found that fluctuations in the stress level measured via hair sample were not only positively correlated with time preferences in the short delay task and risk preferences, but also the decision noise in the risk task. However, relationships for the risk task disappeared when an outlier was removed. Also, we found model-free task measures in the short delay task to be moderately related to both hair cortisol as well as the stressful life events questionnaire measure. For example, we observed that endogenous stress fluctuations and the life change units were negatively correlated with the proportion of later choices. Finally, we established that for the reaction times the curvilinear relationship was preferred to the linear one for those with increase in biological stress level compared to the baseline: when cortisol level increased slightly, participants decided slower, but when stress increased to higher levels, they decided quicker.