“…We selected these covariates used in the creation process based on theory and prior research describing the characteristics of recipients and the cost, type, or amount of child care used by, or available to, low-income families. For example, prior work has documented associations between families' child care preferences and parent race (Fuller, Holloway, & Liang, 1996), parent immigration status (Yoshikawa, 2011), family income (Early & Burchinal, 2002), parent marital status (Rose & Elicker, 2010), and mother's level of education (Hirshberg, Huang, & Fuller, 2005). There is considerably less research on how state-level characteristics are related to use of child care, but a 2008 review by Lippman and colleagues suggested that state-level aggregates of the above-cited characteristics, along with state policy and macroeconomic conditions, are key factors in explaining cross-state variation in the type, cost, and amount of child care that low-income families use.…”