We examine the duration of child care subsidies in Rhode Island using three entry cohorts: (1) a cohort of 2,814 families that began receiving child care subsidies between June and November 1996, (2) a cohort of 2,893 families that began receiving child care subsidies between June and November 1997 and (3) a cohort of 3,831 families that began receiving child care subsidies between June and November 2000.

          Using the 1996 entry cohort, we examine the dynamics of subsidy use over a seven-year period. We find considerable cycling on and off child care subsidies around the beginning and end of the school year.  This cycling may also be related to seasonal employment patterns. An examination of the probability that a family will exit child care in any given month shows clear patterns of exit related to the way in which the child care subsidy program is administered (i.e., re-determination requirements) and to the school year. Using Kaplan-Meier estimates, we find that the median duration of child care subsidies in Rhode Island during the period of our study is 9 months and the mean duration is 13 months. Fifty percent of families in the 1996 entry cohort have two or more spells of child care subsidies during our seven-year follow-up period.

          We find that the duration of child care subsidies differs significantly depending on (1) whether a family qualifies for subsidies because of low income or cash assistance receipt, (2) the level of family income, (3) the age/school status of the youngest child in the family, (4) the number of children receiving subsidies in the family and (5) the family’s place of residence. Specifically, families that are never observed on cash assistance receive child care subsidies for significantly longer periods than families that are current or former cash recipients (median duration of 12 months for those never on cash and 9 months for current and former cash recipients). Families with incomes greater than 125% of the Federal Poverty Level (FPL) receive child care subsidies for significantly longer periods than families with incomes less than or equal to 125% of FPL (median duration of 11 months for the higher-income families and 9 months for the lower-income families). Families with younger children receive child care subsidies for significantly longer periods than families with older children, particularly families that have only school-age children. The estimated median duration of child care subsidies for a family whose youngest child is less than age 3 is 10 months, while the median duration for a family whose youngest child is of school age is only 6 months. The estimated median duration of child care subsidies is significantly longer for families with more than one child in subsidized care than for families with only one subsidized child. The duration of child care subsidy spells is shortest in Central Falls and longest in Providence (median duration of 7 months in Central Falls and 10 months in Providence).

To estimate the impact of Rhode Island’s adoption of an entitlement to child care in May 1997, we compare child care continuation rates (technically known as “survival rates”) for the first spell of child care receipt for the 1996 entry cohort (this cohort entered before child care subsidies were an entitlement) and the 1997 entry cohort (this cohort entered one month to six months after child care subsidies became an entitlement), using a 72-month follow-up period. We find that families in the 1997 entry cohort received child care subsidies for significantly longer periods than families in the 1996 entry cohort. However, the increased duration appears to result from the welfare reform package as a whole, rather than only from the entitlement to child care subsidies.

To estimate the impact of Rhode Island’s increases in reimbursement rates and eligibility expansions that occurred between January 1998 and January 2000, we compare continuation rates[1]  for the first spell of child care receipt for the 1997 entry cohort (this cohort entered before these policy changes) and the 2000 entry cohort (this cohort entered 6 months to 11 months after the policy changes), using a 36-month follow-up period. We find that the duration of child care subsidies for the 2000 entry cohort does not differ significantly from the duration of child care subsidies for the 1997 entry cohort.  It is possible that Rhode Island’s increases in reimbursement rates and eligibility expansions would have increased the duration of child care subsidies if the economy of the 2000-2003 period (the follow-up period for the 2000 cohort) had been as buoyant as the economy of 1997-2000 (the follow-up period for the 1997 cohort).

          Finally, using the first spell of child care subsidies for the 1997 entry cohort and a 24-month follow-up, we compare the duration of child care subsidies in Rhode Island with the duration of child care subsidies in five other states. We find that Rhode Island has substantially longer periods of child care subsidy receipt than the other five states. For example, the median duration of child care subsidies for the 1997 Rhode Island entry cohort is 10 months, while Meyers, et al. (2002) report a median duration of child care subsidy receipt in Massachusetts for entrants between October 1996 and September 1998 of only 5 months. Rhode Island also appears to have a much larger number of families who continue to receive child care subsidies for an extended period of time. For example, while 25% of families in the 1997 Rhode Island entry cohort had periods of subsidy receipt which exceeded 20 months, Meyers, et al. (2002) report that 25% of Massachusetts families had durations of subsides that exceeded 11 months. 

 

---

In this paper, we use a wide array of administrative data that covers the welfare reform period (1996-2001) to assess how the quality of child care changed as a result of welfare reform and concurrent social, political and economic changes. We compare the group care of children receiving child care subsidies, children living in poverty neighborhoods and children that neither received subsidies nor lived in poverty neighborhoods. Our study area is Miami-Dade County, Florida.

We find many differences between providers that participate in the CCDF child care subsidy program (CCDF providers) and those that do not participate. CCDF providers more frequently violate minimum-standards regulations than other providers. Further, their minimum-standards violations tend to be more serious (including numerous instances of child-staff ratios in excess of minimum-standards requirements) than the violations of other providers. CCDF providers also have substantially more complaints filed against them than other providers, including other providers in poverty neighborhoods. CCDF providers have a larger proportion of unfilled (vacant) child care slots than other providers. However, CCDF providers report a smaller proportion of their staff with low levels of education (high school or less) than other providers, and CCDF providers are more likely to report use of a curriculum than other providers. CCDF providers are both more likely to be accredited and more likely to be profit-seeking firms than other providers. During the course of our study, we find a large increase in the proportion of faith-based providers. This trend is particularly strong for CCDF providers. We believe that this reflects Florida Governor Bush’s faith-base initiatives.   

A composite measure of quality helps to address many policy, administrative and research questions.  We develop such a composite measure using principal components analysis (PCA) and multiple quality variables. The results of the PCA analysis are reasonable (e.g., they identify quality factors that are associated with various important aspects of child care quality).  Specifically, the single composite indicator of quality that we develop using PCA reflects 80% of the total variation in our array of quality measures. This composite indicator could be used to develop a quality rating system that is based on multiple quality measures rather than on a single quality measure, such as accreditation, as is commonly done. The composite quality indicator also could be used to identify a group of very low-quality providers that administrators might want to target for quality-enhancing interventions. Finally, both the cross-sectional and time series variation in the composite indicator could be used to evaluate the impact of quality interventions. For example, using our composite indicator, we find an increase in the median overall quality of CCDF providers relative to providers that do not participate in the CCDF program after responsibility for the CCDF program was shifted to a local not-for-profit agency with strong leadership, quality-enhancement initiatives and a school-readiness focus.   This finding provides evidence that CCDF providers are responsive to quality initiatives.

In this report, we study the care choices of 19,386 Rhode Island families that received child care subsidies between July 1998 and June 2002.We use these RI data to answer four major questions. Question 1: What were the impacts of RI’s extensive child care reforms on child care choices made by subsidized families?  Controlling for all of the factors listed in Table 1 of the full report, we find strong evidence that RI’s child care reforms significantly increased the odds that both current and former cash recipients and families that never received cash would choose formal care for their subsidized children. The reforms with the greatest impacts were: (1) RI substantial increases in reimbursement rates for formal care, (2) RI more stringent enforcement of requirements for informal providers(3) RI separation of eligibility determination and provider enrollment and (4)RI introduction of portable child care vouchers.   Question 2: How do household characteristics affect the child care choices of families with child care subsidies?  We find that the age of the head of household, the age of the children receiving child care subsidies, the percent of children of various ages, race/ethnicity, citizenship status and education all have significant effects on the child care choices of households with child care subsidies.  Question 3: What impact do community characteristics, particularly the availability of formal care, have on child care choices?   We find that increased availability of center care for preschoolers, Head Start and kindergarten significantly increases the probability that households with child care subsidies will purchase center care.  Question 4: Do households with more than one child in subsidized care behave differently than households with only one child receiving subsidized care?  We find that families with more than one subsidized child behave significantly differently than families with one subsidized child. These results indicate that it is important to analyze care choices for all the children in the family rather than just for one child (typically the youngest) as it is generally done.