Final Report Summary - COOPERATION (Putting Strong Reciprocity into Context: The Role of Incentives, Social Norms, and Culture for Voluntary Cooperation)
The main goal of COOPERATION is to understand the human tendency to cooperate with unrelated others. Here we are particularly interested to shed light on conditions under which people work for the common good ("cooperation") rather than pursuing selfish interests. Our specific ambitions are to investigate how cooperation is shaped by individual differences and social context as provided by institutions (Work Package WP1); the behaviour of other people and social norms (Work Package WP2); the societal and cultural environment (Work Package WP3); and personality and gender (Work Package WP4). This is an empirical project with a strong interdisciplinary perspective. The main tool is economic experiments, in particular using the “gift exchange game” and the "public goods game with and without punishment" which I helped to establish as important behavioural models to understand cooperation.
In WP1 we investigate how material incentives influence voluntary cooperation. One of the questions we ask is how the social and personal norm of reciprocity is affected when people are exposed to material incentives to induce them to cooperate. In one project, we found that incentives influence people's behaviour strongly and as predicted by economic theory. Reciprocity was completely "crowded out". We also found, however, that the extent to which such "crowding out" happens depends on people's experience with trust and reciprocity. Utilising a novel "multi-game" environment, where people at the same time make decisions in two different situations involving different people, we found that the exposure to incentives in one situation negatively affects voluntary cooperation in another unrelated domain. The significance of this finding is that incentives may not only "crowd out" cooperation directly in the domain where incentives are applied, but may spill over to other domains. For example, the ongoing incentivisation of the work place may have negative effects on civic-mindedness as an important form of social cooperation.
In WP2 we investigate the role of peer effects on reciprocity and generosity. Here we investigate to what extent the strengths of reciprocity and generosity are affected by "peers", that is, relevant comparison others. We found strong peer effects in reciprocity: people who observe strong reciprocal reactions in others are more generous themselves, than people who observe others to be more selfish. Similar conclusions hold for generosity. The significance of this finding is that reciprocity and generosity are not "fixed" properties of a person's personal norms, but are shaped by the people around them. To understand our findings theoretically, we performed an extensive analysis of all major theories of social preferences and found that almost all of them, except a theory of inequality aversion, are inconsistent with our empirical findings, whereas theories of social norms can explain our results. We therefore conducted a "horse race" between the two theories and found surprisingly strong support for the theory of inequality aversion over the theory of social norms.
WP3 investigates the role of cultural and societal background for cooperation and people's social preference. One important character trait that is crucial for social cooperation and the smooth functioning of societies is honesty. To investigate the extent to which societal conditions affect honesty we ran experiments in 23 countries which differ strongly in terms of society-wide practices of dishonesty, such as political fraud, tax evasion, and corruption. To measure honesty, we used a very simple die-rolling task where people get paid by the number they report, and reports are unverifiable. The design ensures that dishonesty is not detectable at the individual level but only at the population level. We found that people in societies with wide-spread practices of dishonesty are also more dishonest in our experiment than people from societies with limited dishonesty. We have also collected data on cooperation games in 43 countries around the globe and found that people are much more likely to behave prosocially when they live in a society governed by the rule of law that when the rule of law is weak. The significance of these findings is that prosociality is strongly shaped by cultural and societal background.
Our main goal in WP4 is to understand the role of personality and gender differences in cooperation. To make progress, our approach was to assemble a large data set involving many individual studies and comprising data from thousands of individuals. Having a large data set, rather than just small single data sets, will allow understanding gender and personality differences systematically and with unprecedented statistical power. Our results show strong gender differences in strategic cooperation.
In WP1 we investigate how material incentives influence voluntary cooperation. One of the questions we ask is how the social and personal norm of reciprocity is affected when people are exposed to material incentives to induce them to cooperate. In one project, we found that incentives influence people's behaviour strongly and as predicted by economic theory. Reciprocity was completely "crowded out". We also found, however, that the extent to which such "crowding out" happens depends on people's experience with trust and reciprocity. Utilising a novel "multi-game" environment, where people at the same time make decisions in two different situations involving different people, we found that the exposure to incentives in one situation negatively affects voluntary cooperation in another unrelated domain. The significance of this finding is that incentives may not only "crowd out" cooperation directly in the domain where incentives are applied, but may spill over to other domains. For example, the ongoing incentivisation of the work place may have negative effects on civic-mindedness as an important form of social cooperation.
In WP2 we investigate the role of peer effects on reciprocity and generosity. Here we investigate to what extent the strengths of reciprocity and generosity are affected by "peers", that is, relevant comparison others. We found strong peer effects in reciprocity: people who observe strong reciprocal reactions in others are more generous themselves, than people who observe others to be more selfish. Similar conclusions hold for generosity. The significance of this finding is that reciprocity and generosity are not "fixed" properties of a person's personal norms, but are shaped by the people around them. To understand our findings theoretically, we performed an extensive analysis of all major theories of social preferences and found that almost all of them, except a theory of inequality aversion, are inconsistent with our empirical findings, whereas theories of social norms can explain our results. We therefore conducted a "horse race" between the two theories and found surprisingly strong support for the theory of inequality aversion over the theory of social norms.
WP3 investigates the role of cultural and societal background for cooperation and people's social preference. One important character trait that is crucial for social cooperation and the smooth functioning of societies is honesty. To investigate the extent to which societal conditions affect honesty we ran experiments in 23 countries which differ strongly in terms of society-wide practices of dishonesty, such as political fraud, tax evasion, and corruption. To measure honesty, we used a very simple die-rolling task where people get paid by the number they report, and reports are unverifiable. The design ensures that dishonesty is not detectable at the individual level but only at the population level. We found that people in societies with wide-spread practices of dishonesty are also more dishonest in our experiment than people from societies with limited dishonesty. We have also collected data on cooperation games in 43 countries around the globe and found that people are much more likely to behave prosocially when they live in a society governed by the rule of law that when the rule of law is weak. The significance of these findings is that prosociality is strongly shaped by cultural and societal background.
Our main goal in WP4 is to understand the role of personality and gender differences in cooperation. To make progress, our approach was to assemble a large data set involving many individual studies and comprising data from thousands of individuals. Having a large data set, rather than just small single data sets, will allow understanding gender and personality differences systematically and with unprecedented statistical power. Our results show strong gender differences in strategic cooperation.