This project will use longitudinal data from the Three-City Study to examine the determinants of food hardship among low-income children in Baltimore, Chicago, and San Antonio, the three cities covered by the survey.  The Three-City study asked questions regarding children’s food hardships at three separate points in time (1999, 2001, and 2005), and it collected detailed data at all three surveys on income, labor force activity, participation in all major food assistance programs and emergency food assistance, as well as labor force and participation histories.  In addition, administrative data on monthly SNAP/FSP and TANF participation and benefits have been added to the data set, allowing detailed and accurate analyses of program histories and permitting a study of survey response error. Lastly, administrative data on quarterly earnings are available for one city, allowing for analyses of earnings histories and volatility.  The data will permit the study of several policy-related aspects of children’s food hardship determinants that have not been possible with other data sets.

Our project can be expected to yield new information that could lead directly to policy questions of interest to the USDA, particularly to the goal of reducing food hardships among children in the U.S.  Our new data and estimation methods to be applied to food hardships for a single child could yield more precise and potentially superior information on the determinants of hardship among children.   The rich set of covariates in the TCS will allow us to determine whether there are additional demographic or economic groups of children who are at substantial risk of food hardship, and this could allow policy-makers to address current program features and new program initiatives focused more on such groups.  Our use of administrative data will permit us to make more precise determinations of food hardship among true FSP program participants, which will improve our understanding of the puzzling phenomenon of significant rates of hardship among recipients of such benefits.   Our data allowing us to study food hardship among FSP entrants and those who have recently left the FSP will add to our understanding of the importance of that program to alleviating child food hardship.  Our study of the NSLP, the SBP, and WIC will provide new information on the relationship between participation in those programs and child food hardship, an understudied subject.  Our data on the use of community food banks and other food sources, as well as our questions on coping strategies, will generate new information which could lead policy-makers to better support efforts to alleviate the type of short-term need that leads to child food hardship.