Prevalence and Correlates of Food Insecurity and Homelessness Among University Students

Recently, there has been a steep increase in attention to insecurity among college students for basic needs (e.g., Miles, McBeath, Brockett, & Sorenson, 2017; Morris, Smith, Davis, & Null, 2016); however, published research on student food insecurity and housing insecurity remains sparse. It is critical to understand the prevalence of these challenges because they are associated with mental health functioning, academic success, and graduation rates (see Goldrick-Rab, Richardson, Schneider, Hernandez, & Cady, 2018). Prior studies (Miles et al., 2017; Tsui et al., 2011) point to high co-occurrence of food insecurity and homelessness, so the intersection of these challenges should be explored. To prevent basic needs insecurity and offer appropriate support for those who are affected, we must understand correlates of college food insecurity and student homelessness. Thus, the purpose of this investigation was to answer questions of prevalence, co-occurrence, and correlates of food insecurity and homelessness. A novel feature of this study was our examination of the degree to which cumulative marginalized/ minoritized characteristics predicted food insecurity and homelessness. Cumulative risk models, in which risks that tend to cluster together are equally weighted, are widely used in developmental psychology and have advantages over studies of single predictors of outcomes (Ashworth & Humphrey, 2019; Evans, Li, & Whipple, 2013); to date, such models have not been applied to studies of student food insecurity or student homelessness, even though predictors tend to be correlated (Crutchfield & McGuire, 2019).

METHOD

The research office at our large public university in the Southeastern US randomly selected 7,000 students; 1,923 completed our online survey (27.5% response, much higher than most prior studies). They represented the student body in race (4.7% African American; 5.3% biracial or multiracial; 14.9% Asian; 71.3% White; 3.8% other), ethnicity (6.0% Hispanic), and degree sought (71.0% undergraduate, 27.8% graduate, 1.1% associate's). Mean age was 21.74 years (SD = 4.4; 17—61 years). The vast majority, 94.7%, were enrolled full time. The sample slightly overrepresented women, 51.3% compared to the university population with 45.6%; 39.6% were men; and the remainder used a different term when asked about their gender identity (e.g., genderfluid, transMale). Most, 78.8%, considered themselves to be heterosexual, 9.0% did not reply to the item, 5.7% were bisexual, 2.7% were gay/lesbian, and the remainder were asexual, questioning, or used a different term. An e-mail invitation to voluntarily complete the online survey was sent to the 7,000 students in Fall 2017. The survey was a modification of the one developed by the California State University system (Crutchfield & Maguire, 2017) and the Hope Center (Goldrick-Rab, Richardson, & Kinsley, 2017). Our IRB approved the study.

To measure food insecurity, the USDA Household/Individual Food Security Survey Module (FSSM) 10-item version was used with a 30-day time frame (Bickel, Nord, Price, Hamilton, & Cook, 2000). Two of the 10 questions asked participants to report the number of days various food situations occurred in the past 30 days. The response format might have been confusing, because almost no participant answered those 2 questions; therefore, we adjusted the coding for the 10-item survey using a conservative approach to defining very low food security. Those who provided affirmative responses to none (0) of the FSSM items were food secure, participants with scores of 1—2 were marginally food secure, those with scores of 3—5 were low food secure, and those with scores of 6—8 were very low food secure. We generated a food insecurity variable by combining students who were low food secure or very low food secure.

The U.S. Department of Education (2016) definition of homelessness includes individuals who lack a fixed, regular, and adequate nighttime residence. To assess homelessness using this definition, we asked students to indicate all the places they had stayed in the past 12 months, with 9 items indicative of homelessness (e.g., at a shelter, temporarily staying with friends, outdoor location). The items were recommended by Crutchfield...