Contrary to the way social cohesion is often framed in policy agendas in the aid sector, it is not one single, clear 'thing'. Different dimensions of social relations are affected in different ways by various factors, and relations within and between sub-groups can vary greatly.
Policy implication: Talking about social cohesion generically is very unlikely to be useful. It is necessary to know which aspect of social relations are important to displaced people and those they live among, and for what reason, and then to speak about these specifically and explicitly.
People manage their own social relations. They know what is important to themselves about their social relations and they know better than outsiders how best to manage them.
Policy implication: Those who want to ensure good social relations for displaced people should focus on giving them more agency to manage their own relations, rather than on trying to socially engineer better social relationships for, and on behalf of, others.
Managing social relationships involves trade-offs. The price paid for acceptance could be financial or conforming to social norms.
Policy implication: Discussions about improving social cohesion must explicitly include discussion of these trade-offs. People must decide for themselves which trade-offs they are willing to make.
Aid flows in Pakistan were small. Their main impact on social relations was that unclear targeting and diversion by elites caused tensions within the refugee or Pakistani communities, and not between them. Resentments expressed about aid flows to refugees were really a convenient expression of underlying resentments that had other causes.
Policy implication: Aid agencies should show caution in believing that redirecting aid away from refugees is likely to be an effective appeasement strategy. Those concerned with social cohesion should worry more about transparent and fair targeting of aid within any population group.