The Ethnosurvey in Theory and Practice

Abstract

This article describes a research approach designed to overcome the limitations of federal immigration statistics and to illuminate the social processes underlying aggregate patterns of migration. The principal weaknesses of existing data sources are that they underenumerate and imperfectly measure undocumented migration; they do not reflect the widespread circularity of modern international movements; they omit variables central to the immigration process; and their cross-sectional collection and tabulation precludes the analysis of immigration as a dynamic process. The ethnosurvey is a research design that ameliorates these problems through five specific design features: multimethod data collection, representative multisite sampling, multilevel data compilation, life history collection, and parallel sampling. These design features are described, justified, and tied to the broader methodological literature in social science. The ethnosurvey design is illustrated by its recent application to study Mexican migration to the United States, and empirical evidence is presented to show how it corrects the limitations of federal data on immigration.