Behavioral Evidence
In an aging society, the extent to which aging affects financial decision making becomes increasingly important to understand. We (Carsten Diener, Chrsitian Fiebach, Christiane Schwieren, Gil Sharvit and I) investigated decision making under uncertainty and how it is affected by age in two separate studies. Uncertain decisions are defined as choices with some known (risk) or unknown (ambiguity) probability of yielding a positive outcome. Risky decisions can be subdivided into conditions with a priori probabilities (where probabilities are explicitly given) and with statistical probabilities (which have to be learned through exploration).
In the first study, we review the literature on age differences in decision making under uncertainty and extract three methodological factors which may explain the inconsistent results found in the literature: (1) uncertainty conditions, (2) feedback on decisions and (3) learning. We ran two experiments with young adults (exp. 1: N = 103; exp. 2: N = 50) and older adults (exp. 1: N = 97; exp. 2: N = 50) systematically varying these factors. For uncertainty conditions, the results revealed no age differences in choice behaviour under risk with a priori probabilities. In contrast, age differences were found in risky choices with statistical probabilities, an effect that can be explained with age differences in sensitivity to prior choices. If feedback is provided, we found that older adults are less ambiguity averse than young adults. Moreover, the presence of feedback had a positive effect on the propensity to gamble in uncertain conditions by influencing subjective probabilities for both age groups. In a second step, we compared the results from our experiments with survey data, demonstrating that education is positively related to risk taking, despite the negative relation of risk taking and age. We conclude that differences in uncertainty-processing exist between young and older adults, leading to the second study where we investigate possible explanations of these differences.
Neuroeconomic Evidence
In the second study, a group of young adults (< 30 years, N = 25) and a group of older adults (> 58 years, N = 21) performed a task investigating decision making under uncertainty while brain activity was measured using fMRI. We confirmed the age-related differences in behavior and found age differences in the neural activity of decision-making related brain regions. An interaction effect between age group (young vs old adults) and types of uncertainty (certain vs. risk vs. ambiguity) was observed in clusters of the anterior insula, the amygdala, and the prefrontal cortices. Region of interest analysis further confirmed the localization of this effect and in addition revealed a main-effect of age in the right orbitofrontal cortex. Finally, the observed interaction modulated the connectivity between the left anterior insula and the ventromedial prefrontal cortex. We explore the relationship between age differences in behavior and the effects observed in the brain. We address different widespread processes (e.g choice-valuation, affective uncertainty processing) that may be involved in the observed age-related differences both in behavior and in brain-activity.