Abstract

This study examines the circumstances under which rural households can use outmigration to cope with negative shocks. In theory, when financial markets are imperfect and when migration involves a fixed cost, the impact of economic shocks on migration can depend on the extent to which shocks are common across households. When shocks are idiosyncratic, shocks are likely to raise migration. But aggregate shocks may make it more difficult to pay fixed migration costs, and so can actually lead to less migration. In a rural household panel from El Salvador, idiosyncratic and aggregate shocks do have opposite effects on migration. Households become differentially more likely to have a migrant relative in the year following an idiosyncratic shock . By contrast, aggregate shocksCproximity to the massive 2001 earthquakesClead to large differential declines in migration. Increased difficulty in financing migration=s fixed cost is likely to help explain declines in migration in areas closest to the quakes: households in quake-affected areas also experience differential declines in granted credit. Empirical evidence does not support alternative explanations for the quake-induced migration decline, such as a desire for family unity, increases in outside aid, a need for family labor in reconstruction, or improved local income opportunities