Altruism and the role of affect: an investigation into the causal effects of positive emotions and its rationality on altruism

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
There is now well established evidence that shows that growing up in poor neighbourhoods can keep people locked in poverty. One response to neighbourhood distress – i.e., situations in which neighbourhoods serves as poverty traps – is to move people out of such situations. Experiments conducted in the United States of America (for example, the Moving to Opportunity experiment) provides a compelling case for this approach. However, when the scale of neighbourhood distress reaches a majority of the population, as is does in South Africa, where the legacy of Apartheid spatial inequality persists, alternative approaches are necessary. One such alternative is to seek to better understand the behavioural mechanisms at work that tilt communities toward the low level equilibria characterising a poverty trap and to target those behaviours that can aggregate to produce neighbourhood distress. Behaviour modification programmes attempt to do this, for example, by targeting the underlying drivers of crime, violence, risky sexual behaviour and fatalism that characterises so many marginalised communities in South Africa. This dissertation uses data from a rare lab-in-the-field experiment to investigate one important mechanism of behaviour modification: the impact of emotions on pro-social preferences. The results show that positive emotions causes an increase in individual altruism and do not imply irrationality according to standard axiomatic frameworks in economics. This contrasts with the view of earlier investigators that emotions are disruptive in decision making. This dissertation is divided into five chapters. Chapter one focuses on a general introduction of endogenous preferences, the link between emotions, pro-social preferences and decision making. In chapter two I discuss the novel behaviour modification experiment that this dissertation is based on: the Activate! Change Driver's programme (hereafter, simply referred to as Activate!), an empowerment programme aimed at youth that live in distressed neighbourhoods in South Africa. Activate! consists of a nexus of interventions that aims to promote prosocial preferences (i.e., preferences for altruism, trust, and commitment to the public good), as well as interventions that are aimed at changing mindsets and perceptions that foster destructive risk-taking, myopia and civic apathy. The programme runs as a series of three modularised workshops covering self-belief, goal-orientation, creative thinking, problem solving, resilience, communication skills, trust building, project management, and political engagement. These interventions are theorised to lead to an actualisation of greater pro-social preferences as well as better outcomes in terms of risk taking, tolerance for delayed gratification, civic engagement, and economic opportunity. In chapter three, I investigate the causal effect of positive affect on altruism. I employed two estimation strategies where in the first strategy, I used an instrumental variable approach to estimate the effects of affect on altruism. In the second strategy, I relaxed the assumption that affect does not directly impact altruism by using a mood inducement experiment to vary affect within the subjects. Both identification strategies support the same conclusion: positive affect is shown to be a significant and positive cause of altruism. Subjects in the positive affect treatment send significantly more tokens to anonymous others in a dictator game as compared to subjects in the mild neutral affect treatment. This result is robust in all specifications considered. Chapter four focuses on characterising whether emotions disrupt rational choices in altruism. In psychology, emotions are theorised to influence decision making since they influence a person's decision to exert effort and be productive at work. There is also evidence indicating that emotions reduce time preferences over money, that they can increase cognitive flexibility, reduce spending and willingness to pay as well as increase reciprocity in gift exchange. I made use of data from a modified dictator game experiment and employed the axioms of revealed preferences to test for economic rationality. I found that subjects treated to positive affect elicited altruistic preferences that are indifferent to subjects treated to neutral affect and that their preferences fit the standard economic definition of rationality. Thus, by this account even though emotions are seen to be disruptive, they do not imply irrationality. In summary, the dissertation presents a number of implications. First, behaviour modification programmes especially in the context of marginalised communities may result in increasing pro-social outcomes and therefore have the potential to reverse fatalistic preferences among young people. Second, behaviour modification and ultimately belief updating towards pro-social preferences has the potential to spark civic mindedness among young people. Third, positive emotions mediate pro-sociality and do not blur individual rationality in decision making.