Migrant Network Analysis

Cumulative causation is the self-reinforcing process by which each act of migration changes the origin community in ways that make further migration more likely. As people leave, more residents acquire ties to migrants, which spreads the social capital that lowers the cost and risk of moving, so the next round of migration is larger and easier still. Massey argued this feedback also reshapes local economies, expectations, and the prestige of migration, embedding it in the community's social structure. The result is that a migration stream can acquire a momentum of its own, growing and persisting even after the wage gaps or shocks that first triggered it have disappeared, which is why network-driven flows are so hard to reverse.

Social capital here means the resources a person can access through ties to others who have migration experience — information about jobs and routes, help with documents, a place to stay, and financial backstops. Empirically it is captured at the individual level by whether a person has kin or friends who are current or former migrants, and at the community level by migration prevalence, the share of the community connected to migration. Massey treats prevalence as the key summary because it indexes how widely the enabling ties have diffused. The measures are usually built from detailed retrospective and longitudinal community surveys that record each respondent's relationships and the migration histories of their relatives and associates.

Because once networks are dense, migration no longer depends on the conditions policy can most easily change. Deterrence aimed at wages, borders, or initial push factors leaves the accumulated social capital intact, and that social capital keeps lowering the cost and risk of moving for the many people now connected to established migrants. The cumulative-causation feedback means each move replenishes and expands the network, so restrictive measures may raise costs at the margin without dismantling the underlying social infrastructure. This is the central policy lesson of network analysis: mature, self-perpetuating migration systems respond far less to interventions targeting their original causes than to anything affecting the networks themselves.