(with Katrina Kosec and Valerie Mueller)
Migration is a known livelihood strategy for the rural poor. Research on mobility and its impact on household welfare is seldom investigated due to the omission of pertinent questions on standard household surveys. Furthermore, the framing of questions can lead to the computation of misleadingly, low migration rates and household well-being measures depending on the respondent, recall period, and inclusion of the absent migrant’s income. In 2013-4, 726 original households of the Pakistan Rural Household Panel Survey (1986-1991) were resurveyed. The 1991 rosters were used to ask the household head the whereabouts of all household members to track and include former members in the survey. Our paper exploits the unique survey to first understand the benefits of tracking migrants. We compare migration statistics calculated for the original households based on responses provided by the head and those verified by migrants assuming different definitions of migration. Regression analysis is used to determine which factors explain discrepancies between migration variables based on head and migrant responses. Our second objective is to examine the implications of each migration measure on the computation of well-being measures. Migration definitions bear consequences on who are considered household members and how household income and remittances are calculated. Recommendations for tailoring survey instruments to improve accuracy in migration and well-being measures will follow.