Sampling and Surveying Hard-to-Reach Populations for Demographic Research

Abstract

Abstract Because household-based survey designs are notoriously ineffective in studying hard-to-reach groups such as migrants, these groups, however numerically large they may be, are rarely represented in demographic analyses. In this article, we report on the application of a workplace-based stratified probability sampling design (a variant of the time-location sampling approach), response rate, and item-specific refusals in a study of female migrants from Kyrgyzstan, Tajikistan, and Uzbekistan working in bazaars, eateries, and small retail outlets in Moscow, Russia. We argue that the workplace-based survey approach, while not flawless, provides a uniquely feasible and cost-effective tool for studying migrants and similar hard-to-reach populations in metropolitan settings. (ProQuest: ... denotes non-US-ASCII text omitted.) 1. Introduction Systematic sampling and data collection among hard-to-reach and hidden populations have always been a major challenge for social scientists. Marginalized groups such as homeless people, drug users, sex workers, or illegal immigrants are notoriously difficult to access in a systematic way (Faugier and Sargeant 1997; Marpsat and Firdion 1999; Valenzuela 2002). In most demographic studies, where representative household survey is the golden standard for data collection, such marginalized population segments are lost by definition or, at best, are grossly underrepresented. An additional set of problems concerns the willingness of individuals from marginalized, hard-to-reach groups to provide information to survey interviewers, especially about the type of sensitive matters that define their marginality: it can be argued that the same social characteristics and constraints that hinder access to these individuals may also impair their willingness or ability to answer survey questions. This article reports on how these two types of challenges have been addressed in a survey of irregular, international, female labor migrants conducted in Moscow, Russia, in the autumn of 2010. We define these migrants as irregular as most of them are in violation of at least some of arcane laws and regulations that govern foreigners' entry, duration of residency, and work possibilities in the Russian Federation. The main aims of the survey were to assess the feasibility of a stratified probability sampling approach tailored to the occupational distribution of this hard-to-reach population. In addition, the survey sought to test a battery of sensitive questions such as those dealing with sexual partnerships and HIV/STD risks as well as those concerning legal status, income, and living arrangements. Large-scale labor migration from Central Asia to Russia is a historically recent phenomenon. In the early years after the collapse of the U.S.S.R., the nations that emerged from its ruins saw massive emigration of ethnic Russians and other ethnic groups of European origin primarily to Russia. At the same time, political instability, especially in Tajikistan spurred sizeable refugee migrations. However, since the second half of the 1990s, migration from Central Asia has been increasingly dominated by labor migrants from native ethnic groups, which in the Soviet era were characterized by very low mobility (Ergeshbayev 2006). The Russian Federation has always been the destination-country of choice for the overwhelming majority of Central Asian labor migrants. Among primary Russia- bound migration streams, Tajik migration is the oldest, but economic hardship and endemic unemployment in Kyrgyzstan and Uzbekistan have fueled increased migration from those nations as well (Zayonchkovskaya 2009). Russia's Central Federal Region is a major magnet for labor migrants from Central Asia; especially the city of Moscow, Europe's largest metropolis with a population of more than 10 million, and the surrounding region of Moscow Oblast (Dolotkeldieva 2010; Zayonchkovskaya 2009). …