Association for Public Policy Analysis & Management

Current research and policy analysis suggests two- generation programs that simultaneously provide educational services to mothers and children may improve outcomes for low-income families (Chase-Lansdale & Brooks-Gunn, 2014). However, mothers have to balance educational participation with managing a household, working, and providing child care, often with few supports (McLanahan, 2004). Compounding these challenges, low-income families often live in disadvantaged neighborhoods with little access to educational services (Brooks-Gunn, Duncan, & Aber, 1997). Access to child care can help address some of these challenges because securing adequate child care is a primary barrier to mothers’ participation in education (Gelbach, 2002). However, only some low-income mothers whose children are in care participate in education, and the reasons for this are not well understood. The limited research that has explored barriers to educational participation has typically focused on individual characteristics. Yet, research on barriers to low-income mothers’ employment identifies contextual predictors of participation in employment, including access to transportation and local labor market conditions. Such contextual factors could similarly affect participation in education.

This exploratory research uses data from the Head Start Impact Study (U.S. Department of Health and Human Services, 2010) to examine: 1) individual, child care center, and neighborhood factors that promote mothers’ participation in education; and, 2) whether the effects of these factors are moderated by baseline maternal education. Individual predictors include demographics, maternal depression, and maternal sense of control; center predictors include total hours open, quality, and services (the center provides information about educational options); neighborhood predictors include norms (the percent of individuals with less than a high school education and services), the density of education, employment, and health services.

The analytic sample consists of 1870 families randomly assigned to 336 Head Start centers in different neighborhoods. Twenty-four percent of these mothers participated in education in the year following random assignment.

Multi-level logistic regression analyses illustrated that many individual characteristics, including high sense of control, having at least some college education, and being younger, were predictive of mothers’ participation in education. Contrary to expectations, none of the child care center characteristics were predictive of educational participation. Neighborhood characteristics were related to participation in education in interesting ways that depended on baseline level of maternal education: 1) greater density of neighborhood services increased participation in education more for mothers who had no college education at baseline, perhaps because mothers with some college were already in school; 2) neighborhood educational attainment increased participation in education more for mothers with some college education. Average neighborhood educational attainment may be salient to educational participation because it sets norms where education is expected, and neighborhood social networks can provide information about educational opportunities. However, these norms may only be relevant for mothers who have some college because norms are not a strong enough influence to remove barriers to education for mothers who are not already motivated towards education. These results suggest that individual and neighborhood characteristics predict educational participation for certain groups of mothers. Potential implications for increasing participation in two-generation programs will be discussed.