Social networks and labour migration in South-Asia

We study motives and explanations for network use in labour transactions in low income countries and the implications for employers, and in particular for the segment of poor unskilled workers. One explanation favoured among economists is that networks are vehicles for effective information transmission about vacancies. Alternatively, the high incidence of network use may reflect problems of asymmetric information and incomplete contracts. By engaging in labour transactions, employers and workers expose themselves to behavioural risk, and in connection with internal or international migration, these risks intensify. If operating on his own, a migrant may worry that he will not be paid as agreed, or that he may lose his job once an economic downturn sets in. An employer may worry that the migrant is less diligent than he appears, or that he may soon leave. In contexts where labour laws are rarely enforced and worker opportunism is common, finding jobs or recruiting through networks becomes more attractive. The downside is that entry into lower end labour markets becomes highly selective. Those lacking networks may face powerful mechanisms of exclusion. This project will develop theoretical models to study the incentive problems confronting workers and employers and their variation across the skill-level of jobs, production technologies, industries etc. In a factory with piece rate pay for collective effort, workers may put pressure on their employer to recruit through their own networks. In a fixed wage system, it is the employer that has an incentive to use employee referral to control moral hazard. Using primary data from detailed village and destination studies in India and Nepal, we will test alternative theoretical explanations for network use. The ultimate aim is to provide knowledge on mechanisms that may exclude the poor from these networks to be able to give advice on policy interventions that may improve the access of the poor to external labour markets.