Association for Public Policy Analysis & Management

The existing literature examines many factors that play a role in the decisions families make with regard to non-parental care for their children, including whether to use non-parental care, what type and how much. Type of care decisions have been associated with child age, mother's education, family income, race and ethnicity, and family structure as well as parents’ reported preferences. Yet few studies of child care selection are able to take into account the characteristics of all the providers nearby the family. In particular, the role of distance in child-care decision-making is under-studied. While descriptive information suggests parents use child care providers close to home, no studies directly measure the trade-off between price and distance in choosing child care.

This study uses unique data on low-income children in Minnesota combined with data on child care providers to analyze the role of distance in families’ child care decisions. We apply standard methods of analyzing consumer demand for differentiated products accounting for variation in the local child care market—what is available geographically nearby to families—and examine the role of distance and quality in families’ decisions. We have data on child care use of low-income families who receive subsidies to help pay for care, whose access to and use of high-quality care are of interest to policymakers.

This study uses administrative data on all children receiving child care assistance in Minnesota in 2016, including information on the specific child care providers that are paid by the Child Care Assistance Program (CCAP) for each child. We link the CCAP data with data on the local child care market near where the family lives, including all licensed child care centers and family child care providers. Provider information included licensed capacity, price, location, accreditation and quality rating of the provider from the Parent Aware Quality Rating System.

We estimate a multinomial logit model to describe the demand of consumers for child care, given the characteristics of the local child care market. There are two key innovations. The first is to view the distance between a family and a provider as an interaction between household and provider characteristics, that is, a factor specific to each household-provider combination. Any two households will (usually) be a different distance from a given provider. We measure distance between each household and each provider (using straight-line and driving distance measures). Interactions of household and provider characteristics allow us to test how child care decisions differ across various dimensions such as household income categories and quality rating of the provider (while controlling for distance). The second is to model the full choice set, rather than aggregating providers into categories and modeling the choice of category. This captures the fact that different households face different choices depending on the structure of their local child care market and accounts for distance. This study builds on prior literature on child care decisions by accounting more fully for the amount and characteristics of the supply near the family and how that influences their decisions.