ABSTRACT
Food insecurity has been a persistent problem in the U.S., and yet over the past three decades, federally funded food programs have become more restrictive. Scholars and policymakers have inquired whether the nonprofit sector is increasing its food provision activities to address this unmet need. This study analyzes data from the U.S. Census and a nationally representative survey of religious congregations in the U.S. to examine trends in food insecurity and congregation-based food provision between 1998 and 2012. The objective of the study is to investigate the extent to which congregation-based food provision fluctuated with national food insecurity prevalence for the overall population, and for subgroups vulnerable to this condition. Results show an over-time correspondence between the prevalence of food-insecure households and the prevalence of congregations that provide food. Parallel patterns are observed between food insecurity in disproportionately affected subpopulations (e.g., African-Americans and immigrants) and food provision in the congregations likely to serve those households. These findings indicate that congregations are helping meet the needs of food-insecure households. However, research suggests that congregations and nonprofits are not an adequate substitute for federally funded programs. Policy recommendations include expanding access to federally funded programs like the Supplemental Nutrition Assistance Program (SNAP) to more immigrants and other groups vulnerable to food insecurity, as well as providing more systematic financial or federal support and quality control of congregation-based efforts.
Notes
1. In Waves II and III, if the respondent answered “no” to this question, the interviewer asked follow-up questions to probe into whether the congregation provided any type of social services. For consistency, in order to assess change over time related to this item, Wave II and Wave III responses generated from the follow-up questions were not included in the analysis.
2. In Wave III, congregations that offered four or fewer programs were asked to describe each program. Congregations that offered more than four programs were asked to describe their four most important programs. Among the congregations reporting some social service activity, 27% reported more than four programs. This methodological difference between Wave III and Waves I and II means that for comparison across waves, the percentage of congregations involved in food provision programs in Wave III may be underestimated if some of the congregations that offered more than four programs did not view their food provision program as one of their four most important programs and thus did not report it. However, because the percentage of congregations participating in food provision in Wave III exceeds the percentages in Waves I and II, an underestimation of this statistic in Wave III would not affect the upward trends observed nor the conclusions drawn.
3. the Current Population Survey (CPS) classifies a household based on the attributes of the household reference person. For example, if the household reference person is African-American, the CPS classifies the household as African-American.
4. For all of the analyses in this study, the NCS data are weighted using the following congregation level weight – wt_all3_cong_dup – which treats each congregation as one unit regardless of its size. This is the appropriate weight to use when assessing trends among congregations because a congregation’s likelihood of appearing in the NCS sample is proportional to its size and using the congregation level weight undoes the over-representation of larger congregations in the NCS (Chaves, Anderson, and Eagle Citation2014).