"Taken together, reciprocity, commonality and preference make for a sense of moral virtue that is rooted in privileged relationships between those who share some form of collective identity. Claims for assistance and support based on this kind of reciprocity are seen not only to have power but also to be legitimate, for patrimonialism is a moral force. As Lonsdale (1992b) has argued of the Kikuyu community in Kenya, leaders are expected to give back to the community, but not because there is any notion of a deserving poor. While 'wealth ineluctably incurred obligations' (1992b: 5), the moral purchase of claims on the wealthy come not from a shared belief in the virtue of redistributing wealth per se, but from the fact that the individuals concerned were embedded within a specific ethnic and moral community and were therefore subject to an established set of reciprocal expectations. At the same time, patrimonial virtue is internally contested. Personal ties, ethnicity and locality do not always pull in the same direction; claims or obligations may arise from shared school experience or involvement in a business organization or a club. These ties may enable electoral work: a candidate may raise campaign funds from former schoolmates (their OBs and OGs, as Ugandans call them) or fellow businessmen (Vokes 2016). But politicians, and voters, may have to weigh the relative significance of different kinds of tie: is it proper to trust, or help, other people from your ethnic group more than people who went to the same school? Your kin, rather than your neighbours? The sense that you must not let someone down - and its corollary, the belief that someone owes you something - is a powerful force, but patrimonial morality is chronically jealous: each patron therefore lives in constant fear of being replaced as a result of not having been attentive enough to their supporters' needs, and voters readily suspect theirMPof favouring immediate kin over the wider ethnic group, or - wo