Role of uncertainty, stress and failure in dietary and economic decision making
Item statusRestricted Access
Embargo end date31/07/2022
This thesis studies how economic circumstances and psychological factors in the decision environment affect individuals’ behaviours and choices, particularly in the context of dietary and economic decisions. Using both theoretical models, observational data and behavioural experiments, the role played by economic uncertainty, stress and failure is studied. Understanding what drives individuals’ dietary choices and economic decisions is an important basis for developing policies aimed at tackling obesity and poverty. The first chapter examines the relevance of economic uncertainty as a pathway in the causal link from education to dietary choices and body weight, both theoretically and empirically. Previous research argues that instincts to hedge starvation risks in historic times of food insecurity might cause individuals today to non-optimally gain body weight in the face of uncertainty. In my theoretical analysis, I show that rational economic decisions can also lead to adjustments in diet when facing uncertainty. If eating healthy is an investment in future health, rational utility-maximizing individuals will reduce their consumption of healthy foods when facing economic uncertainty. In my empirical analysis, data from the British Cohort Study 1970 (BCS70) and the Quarterly Labour Force Survey (QLFS) is used to examine the relevance of financial, wage and job security as mediating factors in the link between education and body weight. My results suggest job and wage security among men and financial security among both genders to be significant mediating factors. The second chapter examines whether stress leads to unhealthy dietary choices using a lab experiment, conducted jointly with Michèle Belot, Jonathan James and Martina Vecchi. We propose a novel stress protocol that aims to mimic everyday stressors experienced by low socioeconomic status individuals, specifically mothers of young children. The protocol consists of computerized tasks under time and financial pressure. We evaluate the impact of (mild) stress on immediate and planned food choices, comparing a group exposed to our stress protocol relative to a control group. Immediate consumption is measured with in-laboratory consumption of low calorie and high calorie snacks; planned consumption is measured with an incentivized food shopping task. The stressfulness of the stress protocol is evaluated using subjective assessments, as well as physiological measurements (heart rate and salivary cortisol levels). We find no statistically significant effect of stress on the nutritional content of immediate or planned food consumption, thus no support for the hypothesis that everyday stressors are a likely explanation for unhealthy food choices. The third chapter examines the individual and combined influences of short-term stress, failure and success on economic decision making using an online experiment with a novel experimental protocol, conducted jointly with Martina Vecchi. Participants completed two online sessions, during each of which they carried out a variety of economic decision making tasks: decisions on altruistic behaviours, choices under risk, intertemporal substitution choices, and effort choices. To introduce experimental variation in participants’ stress levels, they were assigned to complete an incentivized cognitive task aimed at inducing mild stress, either at the beginning of the first or the second session. The incentive structure for the stress task involved a deduction from participants’ pay-off if they performed below an unknown threshold. With the second experimental treatment we introduced variation in participants’ perceived failure or success by randomizing feedback about their performance relative to the threshold and by varying the threshold. Independent of the timing of the stress task, participants were assigned to receive either no feedback, feedback with a low threshold level (success condition) or feedback with a high threshold level (failure condition) prior to the economic decision making tasks in the second session. The stress protocol was perceived as significantly more stressful than a control task, and it induced a sizeable and significant rise in state anxiety. The provision of negative feedback (“failure”) significantly lowered participants’ assessment of their performance, induced feelings of failure and raised state anxiety. Despite successfully inducing psychological short-term stress and failure, we find no statistically significant evidence of an impact of stress, failure or success on economic decision making.