New Versus Return Participation in the Supplemental Nutrition Assistance Program During the Great Recession Lloyd D. Grieger, Yale University Background and Methodology Participation in the Supplemental Nutrition Assistance Program (SNAP) climbed to historic levels during and immediately following the Great Recession. During the recession, about 1 in 8 persons received SNAP benefits and by 2011 annual Federal spending on the program amounted to $75.7 billion with over 44.7 million people in 21.1 million households receiving benefits in an average month that year. Researchers know a great deal about participation rates and the correlates of participation during the recession; however, it is unclear whether the recession-era increases in SNAP participation represented an influx of new participants or a mass return by former participants coinciding with the economic downturn. In addition, relatively basic facts about the demographic profiles of new and return SNAP participants are unknown. This analysis seeks to uncover: 1) the extent to which the Great Recession coaxed new versus return participation in SNAP among adults and 2) how the demographic profiles of new and return adult participants differ. In order to differentiate between new and return adult SNAP users, one must have access to information about individuals’ SNAP use over their entire adulthood. The Panel Study of Income Dynamics (SNAP), the longest-running nationally representative source of panel data in the United States, was employed for the analysis because it is the only data source with information that spans a period long enough to track SNAP participation over the entirety of individuals’ adulthoods. The PSID was first administered to about 18,000 individuals in 1968 and original panel members and their offspring were re-interviewed annually (biennially after 1997). As of the 2011 wave, the PSID included information from over 70,000 individuals spanning 42 years. SNAP participation is measured at the family level in each year of the panel. In each household, the head and wife report the amount of income received from food stamps in the previous year. After combining, if the value is non-zero all members of the family are considered to have been program participants in that year. Individuals were considered to be Great Recession SNAP participants if they lived in a family that received income from the program in 2008, 2009, or 2010. Individuals are included in the analytic sample if they are observed continuously (that is, in each survey wave) from the year they become age 25 until 2010 and have non-missing values on all other variables of interest. Continuous observation is required to ensure that SNAP participation for the entire period of adulthood, which is considered to begin at age 25, is observed. Thus, individuals may be observed for varying lengths of time – ranging from 43 years for individuals who turned 25 in the first wave of the PSID, to 1 year for those who turned 25 in the most recent wave. This yields an analytic sample of 7,680 individuals representing 98,143 person-years of information. Other important demographic measures were also harvested from the PSID, including: age, gender, completed education, race/ethnicity, household configuration (single or two-person household), household region (north central, northeast, west, south) and household location (urban, suburban, rural). The measure of household location was derived from a place-based spectrum/continuum very similar to the one employed by Office of Management and Budget and the USDA Economic Research Service, and slightly different from density-based measures employed by the U.S. Census. Findings Simple descriptive statistics about new and return SNAP participation revealed that over 2 in 5 Great Recession SNAP participants (42%) were first-time participants. This means that of the roughly 12% of adults in the sample who lived in households receiving SNAP benefits during 2008-2010, just under half had never before lived in a SNAP household as an adult. The average return recipient first participated in SNAP when they were 28.1 years old compared to 31.8 years old for new participants, meaning that first-time participation was more compressed toward younger ages for returners than for new participants. The fact that new SNAP usage among older people appeared to be more common during the Great Recession is notable because other research has shown that the likelihood of ever using SNAP conditional on no previous participation (i.e. first time participation) declines substantially with age. Multivariate statistical models were employed to explore the differences in the demographic profiles of new and return adult SNAP participation. The analyses demonstrated that new SNAP users during the Great Recession were very different from their return counterparts in ways that do not fit mainstream conceptions of the typical SNAP participant. After controlling for important observed characteristics, return participants were more likely than new participants to be suburban (than rural or urban), white, highly educated, and older. This means that the massive uptick in SNAP program participation during the Great Recession was not due solely to returners recycling back onto the program and cohort replacement but was also because of new participants from atypical demographic backgrounds. Future analyses will compare the Great Recession period to other recent economic downturns in order to determine if the patterns of new and return usage observed during the Great Recession are unique. Contact: Lloyd Grieger Department of Sociology Yale University P.O. Box 208265 New Haven, CT 06520-8265 Phone: 203-432-3313 E-mail: lloyd.grieger@yale.edu