Logo ipums


Publications, working papers, and other research using data resources from IPUMS.

Balistreri, Kelly 2012. Family Structure, Work Patterns and Time Allocations: Potential Mechanisms of Food Insecurity among Children.

Project Goals. Over 469 thousand households in the U.S. experienced very low food security among children, a severe condition characterized by reductions in food intake due to an inability to afford enough food. But food insecurity is not simply about economic resources. There exists a paradox in which some poor households with children are food secure while some non-poor households with children are food insecure. This study moves beyond a singular focus on income and considers how the family context may protect or generate risk of food insecurity for children. The goals of the proposed project were to: 1) to provide a detailed profile of an understudied group, households with children experiencing very low food security; 2) to consider the food security paradoxhouseholds that are poor but food secure, and households that are food insecure but non-poor, and 3) to examine how family context (structure and parental time allocations) is related to food security among households with children.Data. The study uses multiple rounds of the CPS Food Security Supplement, taking advantage of new cohabitation and parent pointers to explore more refined measures of family structure from the perspective of the child. For the main analysis, data from 2007 through 2010 are pooled, excluding any households surveyed twice due to the 4-8-4 sampling structure of the CPS. The analytic sample (N= 64,860) is composed of children ages 0 to 17 with household-level child food security information and household composition from the child perspective attached. The final research question is addressed by linking multiple years of the American Time Use Survey (ATUS) to the FSS. The American Time Use Survey (ATUS), a nationally representative sample of U.S. household that collects detailed information on family and household composition as well as time diary information on how individuals allocate their time.Methods. The analytic strategy begins with a descriptive portrait of households (using the pooled FSS) across childrens food security status levels 1) high or marginal, 2) low, and 3) very low. Several sets of multinomial logistic regression models are estimated. First, the likelihood of being in a particular food security group was estimated with the full FSS sample in order to assess the independent effects of factors such as family structure and parental work patterns while controlling for other characteristics of the child and household. Further analyses limit the FSS sample to only low income households, to uncover characteristics associated with resiliency or risk, paying particular attention to parental work patterns. The final portion of the project explores the relationship between childrens food insecurity (broadly measured) and parental time allocations using linked FSS and American Time Use Survey data.Results and Implications. The results presented here indicate that children at the most risk of very low food security are more often children being raised in immigrant families. While under a quarter of all children in the United States is the child of an immigrant, a disproportionate amount (40%) comprise the population of children living under the most severe conditions of food insecurity. Results from multivariate models suggest that family structure is a key predictive factor among low-income families. Net of economic and household characteristics, children living with an unpartnered parent or living in a more complex family are at an increased risk of low or very low food security compared with children living in either a 100% biological family or a stepfamily. Notably, mothers work patterns among low-income families are much stronger predictors of childrens food insecurity among stepfamilies than in 100% biological families. Other results suggest that disability among adults living with children greatly increases the likelihood of the more extreme form of child food insecurity. Net of individual and household characteristics, children living with a disabled adult have almost three times the odds of living under conditions of very low food security than children living in a household without the presence of a disabled adult.The exploratory analysis yielded limited results, possibly due to the small sample size of the combined FSS and ATUS sample. Restricting the sample to the shortest time frame between food security interview and the ATUS interview, as well as restricting the sample to include only respondents interviewed on weekdays greatly reduced the explanatory power of the models. However, several findings are worth note. It is reasonable that parents in households in which the children are completely food secure would allocate more time to work; more time spent on work among parents often yields more economic resources. Yet unadjusted differences in the time spent on food preparation and cleanup are higher among parents living with children experiencing any food insecurity regardless of family structure. Once household resources and number of children were controlled for, the relationship between time spent in food preparation and childrens food insecurity became marginal at best, and only among two-parent households. Restricting the analysis to employed parents finds that time spent in the care of non-household members may be associated with a higher likelihood childrens food insecurity. While these results do not imply that parental time spent in food preparation or care of non-household family members causes childhood food insecurity, it may suggest that parents in food insecure environments use their time differently than do parents in food secure households. Future research will address these issues using more inclusive measures household-level food insecurity.