RESUMO
The study examines the effect of religious affiliation on livelihood activity choice amongst a sample of 1209 women in the Yilo and Lower Manya Krobo Districts of Ghana. We attempt to disentangle the effect religion would have through its effect on risk preferences, from the effect it would have due to creation of social capital or the effect of clustering due to social identity. We find evidence that there is a strong positive social capital or social identity effect of being Catholic, Pentecostal or Protestant when it comes to farming. This effect is also positive for Protestant and Catholic women for wage employment. This social capital and social identity effect is significantly negative for Catholic, Pentecostal and Protestant women in the case of petty retailing. Women who are willing to take risk are more likely to be in farming and less likely to be wage employed than risk averse women. However, the effect of risk aversion as a result of being Catholic, Protestant or Pentecostal only seems to have a negative effect in the case of farming. The risk aversion effect does not appear to act through religious affiliation when it comes to wage employment. It is recommended that any policy intervention aimed at enhancing livelihood activities of women in this area should take into consideration any relative advantage of having a specific religious affiliation. Such policies should take cognizance of both clustering and networking effects resulting from belonging to a religious group. Risk profiles of women are also important when implementing such policies. Other important considerations are age, household headship, education, dependency ratio, household farm ownership and household enterprise ownership.
