Can behavioural economics make us healthier and more ethical?
This thesis discusses potential motivations behind unhealthy or unethical behaviour. With the experimental method and drawing on insights from behavioural economics, it identifies features of the decision making process which enhance the appearance of unethical or unhealthy behaviour. The aim is that of informing policy makers or future interventions on the ways to improve these behaviours. The first chapter "Groups and Socially Responsible Production: An Experiment with Farmers", is about the main drivers of entrepreneurs' social responsibility. With a lab in the field experiment with farmers in Italy, I study how and whether group decision making affects the social responsibility and the ethicality of production choices. Does corporate social responsibility decrease when corporate decisions are taken by several people (a corporate board) instead of an individual entrepreneur? And if so, why? I ask to 126 farmers in Tuscany to choose between an ecological and non-ecological but more profitable product to use in their farm. To study the effect of collective decision making, I introduce two experimental variations in a 2x2 design: (i) the number of people responsible for the decision (one vs three) and (ii) the number of people receiving a payoff from the decision (one vs three). I find that collective payoff leads to less socially responsible decisions, possibly because it provides them with the moral wiggle room to be less pro-social. On the other hand, sharing the responsibility of the decision with others does not change behaviour in this setting, meaning that there is no diffusion of responsibility, in contrast to what has been found in laboratory experiments. To shed light on the external validity of my results, I find that my experimental measure of social responsibility correlates with measures of social responsibility outside the lab. The second chapter "Rewarding with a food makes that food more appealing", studies with a field experiment the impact of using a food as reward. The chapter is joint work with Jan Michael Bauer, Michele Belot and Marina Schroder. Parents often use sweet and calorie dense foods to reward their children. We hypothesize that such practices may contribute to the formation of unhealthy food preferences. To test this hypothesis, we conducted a randomized field experiment with 214 children in 3 schools in Germany. In the treatment classes, children were asked to complete a cognitive task in 6 visits over 3 weeks, and received dried apples as a reward. The task consisted of counting a number of random dots in several different pictures. In the control group classes, children received the dried apple unconditionally. Receiving the food for solving the tasks might provoke a positive association of the food with the positive feeling of achievement. It could also be that having to provide effort to obtain a food enhances the value of that food. We split the treatment groups into two, varying the number of pictures between the two treatment groups. This allows us to identify the role of effort in driving a change in preference for dried apple. Our results show that rewarding children with food does increase their liking for the food reward. Also, increasing the effort required to obtain the reward does not impact the liking. These findings suggest that parents and carers should avoid using unhealthy food as rewards and may even use this mechanism to increase the liking of healthy food by using such foods as rewards. The third chapter, "Stress and Food Preferences: A Lab Experiment with Low Income Mothers", discusses the influence of stress on food preferences. Jointly with Michele Belot, Jonathan James and Nicolai Vitt, we conduct a lab experiment with 196 low-income mothers in the UK to study the impact of acute stress on immediate and planned food choices. We propose two channels through which stress might affect food choices: (i) by affecting individuals' preferences and (ii) by affecting their ability to make sound decisions. With a novel incentivised stress task designed to mimic stressors often experienced by mothers from a low socio-economic status, we experimentally induce acute stress on a group of participants. We take measurements of participants' salivary cortisol and heart rate over the course of the experimental sessions to assess the stressfulness of the stress task. Afterwards, we ask them to purchase food items in a "virtual supermarket" and also we offer them high-and low-calorie snacks. We use the nutritional content of the chosen food-shopping basket and the quantity of snacks consumed to determine the impact of acute stress on planned and immediate food consumption choices, respectively. Contrary to previous findings in the literature, we find no evidence of an effect of acute stress on immediate or planned food choices.