Purpose: To develop a theoretical framework for understanding household resilience in the context of food insecurity in the United States. Examining household resilience and its relationship with risk for food insecurity is an emerging concept. Most of the formative work has been conducted outside of the United States – largely in developing nations.

Methods: Semi-structured interviews have been completed with a demographically diverse sample of adults (n=45) across five states (MD, TN, AR, NE, and CA). Interviewees were recruited from sites that serve populations at risk for food insecurity (e.g., food pantries, low-cost clinics, and low-income community programs). The qualitative approach of this study is best characterized as phenomenological research with a goal of understanding the commonalities of the lived experiences of food insecurity and its perceived relationship with household resilience. The study will include a thematic analysis with Creswell’s “lean coding” technique. Themes will be largely inductive, but will be explored within a priori household resilience concepts identified in the literature. These concepts include absorptive capacity (household’s ability to minimize exposure to financial shocks); adaptive capacity (household’s ability to make informed choices about alternative livelihood strategies); transformative capacity (conditions largely outside of the household that are necessary to foster long-term resilience). Interviews have just completed, and analysis has begun (below are preliminary findings, with final analysis to be completed prior to the conference).

Results: Preliminary emerging themes within absorptive capacity include income adequacy, buffers against financial shocks, instrumental social support, and stability of necessities. Preliminary emerging themes within adaptive capacity include awareness and social connectivity, self-efficacy and stress coping, human capital, adaptive barriers, and assistance utilization. Preliminary emerging themes within transformative capacity include access to opportunities, governance, civic engagement/power, community social factors, and future oriented thinking.

Conclusions: This foundational work is being completed as part of a larger measurement development project. Findings will directly inform survey module development and testing. The resulting validated survey can be used for monitoring food insecurity risk related to household resilience and for informing the tailoring of community intervention approaches aimed at addressing root causes of food insecurity.

Purpose: Research on college students in the US has produced high rates of food insecurity, with prevalence rates three to five times the national average. Given these high rates of food insecurity, researchers have questioned whether the USDA food security survey module is adequate for assessing food hardship among college students. Emerging research on this topic indicates that college students are responding to food-insecurity questions differently than other populations, which may contribute to higher-than-expected rates of food insecurity among college students. In this study, we quantify the extent to which college students are misclassified as food insecure based on existing food insecurity measures at two large, diverse universities.

Methods: Cross-sectional data on food insecurity provided by college students (N = 2,651; 51% non-white; 72% female) from two large public universities collected between 2015 and 2021 were pooled and standardized. A student’s food insecurity was assessed with either the six- or ten-item food-insecurity measure. Rasch models were used to calibrate food-insecurity scales and assign each student a continuous estimate of their food insecurity. The misclassification methodology developed by Rabbitt and Engelhard (2021) for food insecurity research was used to estimate false positive and negative rates of food insecurity, examining differences by length of the survey.  

Results: Nearly 6% of college students in our sample were classified as food insecure when they were truly food secure based on existing food-insecurity measures; however, about three percent of students were classified as food secure when they were truly food insecure. If we were to account for this measurement error, our estimate of food insecurity among college students would decrease from 37.5 to 35.2 percent. Moreover, misclassification rates are the highest when shorter food-insecurity measures are used. College students administered the six-item food-insecurity measure (14% misclassified) were twice as likely to have their food insecurity status misclassified as those administered the 10-item food-insecurity measure (7% misclassified).

Conclusions: Refinement of food-insecurity measures for the college student population are needed in the future and where possible researchers should consider longer survey modules to reduce measurement error for ongoing assessments.

Purpose: Experience-based food insecurity indicators can be used to rank households or individuals across the continuum of levels of severity of food insecurity but often times studies only report HFI as a dichotomous variable. This study examined the policy relevance of reporting different levels of severity of Household Food Insecurity (HFI), e.g. food secure, mild- , moderate-, severe- HFI.

Methods: Nationally representative surveys were examined to describe the distribution of different levels of HFI by key socioeconomic characteristics from the national to the municipal level. A literature search was conducted to identify studies (N=16) across world regions examining the relationship between HFI severity level with: a) mental health, b) physical health, c) weight status, d) the double burden of malnutrition, d) chronic diseases including type 2 diabetes and hypertension among adults, e) maternal anemia, f) child stunting and b) early childhood development.

Results: The great majority of studies measured HFI with versions of the U.S. Household Food Security Survey Module, the Food Insecurity Experience Scale (FIES), the Latin American and Caribbean Food Security Scale (ELCSA), or The Brazilian Food Insecurity Scale (EBIA). There was strong variability in the distribution of different HFI levels across regions, states and municipalities and by socio-economic indicators. In countries like Brazil there were HFI reductions between 2004 and 2014 in response to more equitable social and economic policies followed by a rebound in severe HFI as a result of major social, economic and political instability. The great majority of studies detected dose-response or curvilinear relationships between FI levels and diverse physical and mental health outcomes, and early childhood nutrition and development indicators. The dose-response relationships were strongly consistent across world regions for overall wellbeing and mental health outcomes among adults and stunting among children.

Conclusions: Reporting on different HFI levels is key for designing, targeting and evaluating policies and programs. HFI can only be properly understood and effectively addressed by assessing and reporting from local to global settings on all its levels of severity.