Social networks and economic behavior: Impacts of gifting in Tanzania

Informal institutions such as social networks are often drawn upon in order to protect households from market failures. Social networks often involve gift giving both within and outside the family network to cope with production uncertainties and food insecurity. But some community members avoid giving out gifts to either family members or outsiders. Following literature reviewed, there are three key reasons or motivations for gifting – altruism, reciprocity, and social norms. Although all three motivations may be for any gifting between households, each motivation is more associated with different types of households than others. In this thesis, I assume gifting between households of the same family network is more likely motivated by altruism, or social 30 norms that create obligations, whereas gifting between households of different families is more likely motivated by reciprocity. However, these gifting behaviors may involve behavior such as free-riding or investment in social capital.As such, gift transfers within or outside family networks can affect the productivity activities of those who give and receive gifts. Therefore, this paper seeks to examine how gifting behavior of farmers, among friends and family networks in the Kongwa and Mvomero districts in Tanzania, affects their productive activities.To do so,using primary data obtained from 552 households from 4 villages in each of two districts in Tanzania, a probit model is run to examine the determinants of a household decision to give gifts or not. Also, I estimate a tobit model with household weeding effort and an OLSmodelwithagriculturalcrop yields as a function of different family types while holding constant other factorsthat affect agricultural productivity. My results support the idea that the formal economy (i.e. access to formal institutions, or services such as savings and credit institutions) is a substitute to social networks in these rural areas. I also provide estimates that suggest that productive efforts of farmer households differ depending on whether they are engaged in gifting with members of another household unit but within the same family network, or outside their family network. To be specific, the empirical evidence suggests that, relative to households not engaged in gifting, nonfamily gifting households have higher yields whereas family gifting households invest lover productivity efforts by weeding less