Association for Public Policy Analysis & Management

This paper addresses the problem of sampling a unique population: Americans living in extreme poverty. Recent studies have shown that about a fifth of all poor households with children, and about a fifth of all poor children, live below the extreme income threshold of only two dollars per person per day (Shaefer & Edin, 2013). Analyses of the Survey of Income and Program Participation suggest that the extreme poor may be a distinct group within the poor, and merit separate study. These data are further bolstered by two years of ethnographic work with 18 families living below the two dollars per person per day threshold in four locations across the United States: Chicago, Cleveland, Appalachia, and the Mississippi Delta (Shaefer & Edin, 2015; Shaefer, Edin, & Talbert, forthcoming).

To examine the unique population of the extreme poor, we use the metric of 25% of the federal poverty line, a measure slightly higher than the two dollars per person per day threshold. We seek to prove that sampling this distinct population of marginalized Americans is possible. We utilize two sampling strategies in both a rural and urban context proven successful in other geographically-based research projects.  In the first, we identify census block groups by mean income and majority race. We use the recruitment technique of knocking on randomly selected doors to assess the existence of extreme poverty among households. We then assess the necessary size of a random subsample of residential addresses within these block groups to attain an appropriately large sample of households living in extreme poverty. The second strategy is venue-based sampling, which has been successfully employed by other studies to sample hard-to-reach populations. This strategy addresses the obstacle of residential instability amongst individuals living in extreme poverty. We identify public sites—homeless shelters, food pantries, community centers, etc.—and engage in systematic time-space sampling of the people at these sites, a large proportion of whom we hypothesize live in extreme poverty.

We conclude that sampling families with children living in extreme poverty requires a nuanced understanding of the public and private spaces where these families are most likely to be. We suggest that, with appropriate pilot work to understand local geographies, this kind of sampling could be scaled up to a national project to better understand the dynamics and implications of extreme poverty in America. Such a national project would provide the necessary foundation for generating truly evidence-based policy interventions to support families living in extreme poverty. Systematic sampling of this population is an important step in generating tailored, evidence-based policy interventions.

Works Cited

Edin, K., & Shaefer, H. L. (September 2015). $2 a Day: The Art of Living on Virtually Nothing in America. Boston: Houghton-Mifflin Harcourt.

Shaefer, H. L., & Edin, K. (2013). Rising Extreme Poverty in the United States and the Response of Federal Means-Tested Transfer Programs. Social Service Review, 87(2), 250-268.

Shaefer, H. L., Edin, K., & Talbert, E. (Forthcoming). Understanding the Dynamics of $2-a-Day Poverty in the United States. RSF: A Journal of the Social Sciences.