Using repeated cross-sectional data to examine changes in early care and education arrangements over time: results from the US National Survey on Early Care and Education 2012 and 2019

The US National Survey of Early Care and Education (NSECE) was described as “the first national portrait of the availability of early care and education” when it was first collected in 2012. The second nationally representative wave of the NSECE recently became available for 2019. This study uses these repeated cross-sectional surveys to report and compare primary childcare arrangements for infants, toddlers and preschool-aged children across two time periods. Our analyses include all sampled low-income households with children aged zero to five with an additional focus on Hispanic households—a population over-represented in the NSECE but considered greatly under-represented in federally subsidized early care and education programs. Reauthorization of the federal Child Care and Development Block Grant Act in 2014 led a number of states to make changes to promote access by low-income families to federal childcare subsidies. This study attempts to exploit the differential changes in policy adoption over time across states to examine whether there is a correspondence between three major policy changes (increases in the redetermination period, reductions in required household copayments, and efforts to improve access to subsidies for households speaking languages other than English) with changes in primary care arrangements over time. While the lack of pre-trend data prevents the identification of causal policy impacts, there is evidence that policy changes associated with the 2014 reauthorization were associated in some cases with reductions in the use of relative care and increases in the use of center-based care, especially for infants in Hispanic households residing in states that lowered the copayment between 2012 and 2019. Moreover, improvements in accessibility for non-English speaking households corresponded with reductions in the use of relative care for Hispanic toddlers during this period. Identifying causal impacts of policies rather than associations would require more than two repeated waves of data, something that might be possible once the NSECE 2024 becomes available.