Global Poverty Research Group

How does social capital form?


The work now being undertaken by Abigail Barr is building on earlier project funded by DFID entitled ‘From Strangers to Neighbours’ which came to an end in 2002. The primary objective of this project was to further our understanding of how social capital emerges and investigate whether anything can be done to promote its generation. The research focus was a sample of Zimbabwean small-holder farmers who resettled into new villages made up of stranger-households in the early 1980s. The strategy was to investigate how these farmers had gone about building new stocks of within-village social capital and thereby improving their ability to solve shared problems and take advantage of collective opportunities, to identify the characteristics that distinguished the new villages that have been successful in this regard from the ones that have been unsuccessful, and to establish whether and how policy interventions might support their endeavours.

     This project is important also because it incorporates a significant departure from the typical survey-based approach to micro developmental issues. We treated this project as an opportunity to explore the potential value of field experimentation as a data generation technique.

     In field experiments people are involved in the playing of games under controlled conditions. The games are designed to capture certain aspects of decisions or types of decision that those people take in everyday life. If we are interested in how they make decisions about risky courses of action, the game will involve choosing between different gambles, i.e., different uncertain outcomes, or choosing whether to gamble or not. Alternatively, if we are interested in whether they are kind to each other because they expect their kindness to be reciprocated, a game in which they have a chance to be kind to others or to be kind back is appropriate. Whatever the design, the games always involve real incentives, i.e., they are played for money or food or something else of value. The decisions that each person makes combined with the role of a die or toss of a coin in the case of gambling games or with the decisions made by others in the case of interactive games determine how much money they take home. As long as the amounts of money at stake in the game are high enough, this causes them to take their decisions seriously and thereby reveal their preferences and underlying motivations through the way they play.

     Traditionally, such experiments are conducted in university laboratories with students as subjects. This affords the experimenters a great deal of control, but has caused many to question the relevance of experiments to our understanding of how people in general and especially in developing countries make decisions. Now, an increasing number of economists and anthropologists (although still probably not more than 30) have started running field experiments. These experiments have involved farmers, urban workers, truck drivers, school children, hunters, gatherers, whalers, and many other types of people. They have been conducted on doorsteps, in sitting rooms, in village meeting places, under large trees, in worker common rooms, on beaches, in truckers’ cafes, in club houses, and in school classrooms, but in every case they have still involved playing games for real money or food or something of value.

In Zimbabwe, we conducted a series of economic experiments in 28 Zimbabwean villages in order to generate data relating to trust, trustworthiness, cooperation, and social norm enforcement. We combined the resulting information with data generated through surveys of households and individuals, group interviews and participatory research exercises and, thereby, built a very rich and detailed picture of how and to what extent the resettled households had transformed the strangers they had found themselves surrounded by in the early 1980s into neighbours.

The experimental methodology turned out to be not only an excellent data generation technique but also a remarkable facilitation tool in participatory research endeavours. The games involving trust, reciprocity and cooperation had a profound impact on people’s willingness to talk frankly about how well their villages functioned as communities. Prior to the games, it proved very difficult to get people to diverge from an account of how people ought to live within a community. We were told about how cooperative people were, how everyone trusted everyone else, and so on. Attempts to get people to talk about what happened when cooperation or trust broke down either failed completely or led to one-on-one discussions about particular incidents. The latter were interesting but did not facilitate the development of a picture of the cooperativeness of each village as a whole. The villagers instantly understood that through the games all, or at least much, was being revealed about how trusting and cooperative they really were. Following the games, the villagers would enter into new discussions about how ‘trustingly’ or ‘cooperatively’ they had played and what this said about their communities. Many of these post-play discussions were initiated not by the researchers but by the villagers themselves. They were fascinated by the metaphor that the game represented and wished to explore its relevance to their everyday life. They were also keen to find out how trustingly or cooperatively they had played relative to other villages and what insights the researchers had as to why such differences existed. Thus, our primarily non-participatory research activities became participatory at the instigation of the research subjects.

This innovatory methodology generated many interesting findings. When faced with low stocks of ‘traditional’ kin-based social capital resettled villagers encourage offspring to intermarry and form civil associations. Intermarriage facilitates informal insurance, while the emergent civil society provides an environment within which trust can develop. Many saw resettlement among strangers as a chance for a fresh start and this fresh start was associated with a persistent increase in people’s willingness to cooperate. Older resettlers play an important role in the accumulation of social capital; they are more cooperative and socially active.  Women are also more cooperative and socially active and they also play an important role in enforcing pro-social behavioural norms. A willingness to comment on the appropriateness or inappropriateness of others’ behaviour, facilitates greater cooperation and social conformity. However, an aversion to the conflict associated with social sanctioning can cause individuals to withdraw from cooperative endeavours.

We found no evidence of sub-optimal investments in civil society upon which to base a recommendation for interventions designed to increase such investments. Social capital generating interventions would better be focused on mutual trust building. The experimental methods we used could be adapted to assist in the building of mutual trust among resettled and displaced people both in Zimbabwe and elsewhere. Resettling entire villages together would allow the resettled villagers to take with them their traditional kin-based social capital. However, this could mitigate the fresh start effect which caused the resettlers to achieve new, more cooperative approaches to social dilemmas. Planners should not focus solely on the physical capacities of potential new community members as in so doing they may underestimate the potential contribution of the older generation as community-builders.

Recent publications

Barr, A., ‘Social dilemmas and shame-based sanctions: Experimental results from rural Zimbabwe,’ CSAE working paper WPS/2001.11, 2001.

Barr, A. and B. Kinsey, ‘Do men really have no shame?’ CSAE working paper WPS/2002.05, 2002.

Barr, A., ‘Forging effective new communities: The evolution of civil society in Zimbabwean resettled villages,’ in World Development as part of a special issue on land reform in Zimbabwe, forthcoming 2003.

Bourdillon, M. and M. Shambare,  ‘Gossip and social control revisited,’ Anthropology Southern Africa, forthcoming 2003.

Barr, A., ‘Trust and expected trustworthiness: Experimental evidence from Zimbabwean villages,’ The Economic Journal, forthcoming 2003.

Reseachers to contact for this project

Abigail Barr