Philippe Marcoul (University of Alberta) on 'Blood is thicker than water': Economic implications of food gifting within kinship networks

There has been a lot of discussion regarding the exact meaning of the word reciprocity in gift-giving. in a influential article, Marshall Sahlins developed a typology of reciprocity in gifting relationship based on whether there is or not a counter obligation of the receiver. In a relationship characterized by balanced reciprocity, the giver of the gift strongly expects the receiver to reciprocate in the (near) future with a gift whose value is close to her initial gift. In other words, any gift received gives rise to a clear debt-like obligation from the receiver. While there is some vague obligation to reciprocate in a generalized reciprocity setting, the strength of this obligation is weak and, unlike that in balanced reciprocity. In other words, reciprocation may or may not come and if it comes it may not match the initial gift value. In Sahlins' generalized reciprocity, giving is unconditional and failure to reciprocate will not put an end to gifting. In this work, following Platteau (1997), we hypothesize that in rural Tanzanian villages small food gifts take place in a context of generalized reciprocity. We develop a formal model of generalized reciprocity in a dyadic relationship where the giver and the receiver may have (or not) altruistic ties. In our model we assume that rural villagers produce their food by exerting effort in cultivation (e.g. weeding) but also by acquiring productive assets (e.g. seeder, plow). After harvest, some of the villagers may receive food gifts as a result of adverse shocks such as bad harvest or accident and food gifts act as insurance devices. Our formal model makes several basic predictions. First, when the altruistic strength of the tie between two villagers increases, gifts are more likely to happen. Second, those who do not make gift should perform higher levels of cultivation effort and invest more in productive assets. Finally, the relationship between altruism and those economic variable is non monotonic; for instance effort will first start to decrease when the altruism level is increased but will then increase again when this level become very strong. To empirically investigate these theoretical relationships, we have collected data from 550 households located in four different Tanzanian villages. For each household, we documented their effort in cultivation and their investment in productive assets. We also recorded each their giving and receiving instances and the exact tie between the receiver and the giver. Consistent with the theory, we find that when gifting is performed between villagers with closer family ties, effort in cultivation tends to be lower. Similarly, when family gifting ties are stronger, villagers will invest more in consumptive assets (e.g. TV, radio, couch) and less in productive assets (e.g. water tanks, seeders, fertilizers). We then discuss the development implications of our findings.