ACF OPRE: National Study of Child Care of Low-Income Families

Chapter Six: Summary of Findings and Their Implications

The state and community substudy relied on state information on child care expenditures data from 1997 to 2001; state child care subsidy use data from 1997 to 2002; and key informant interviews over the period from 1999 through 2002. Together they provide a comprehensive picture of the late 1990s and early 2000s. These years represent a unique period for Federal and state child care subsidy policy. The period began with unprecedented growth in state subsidy programs: after many years of level funding, states had the opportunity to expand their programs rapidly between 1997 and 1999, using funds made available by PRWORA. New flexibility that the law provided, coupled with expanded funding, gave states the opportunity to extend the subsidy program to many more families, as well as to increase the value of the subsidy. The beginning of the period was characterized by new pressures on administrative systems, as states needed to spend the influx of new resources quickly and as they responded to the increased visibility of child care subsidies that came with the additional dollars.

Starting in 2000, states faced a very different set of concerns. Confronted with state deficits and level Federal spending, most of the states in the study focused on consolidating the gains that had been made in previous years, and achieving costs savings in ways that did not jeopardize the integrity of the systems they had put into place when funding was more plentiful. During that period growth slowed or stopped, and, in a few of the states, the number of children who received subsidies decreased slightly from previous years.

This chapter summarizes the information from prior chapters and discusses some of the implications of those findings:

Despite the contraction in funding that characterized the second half of the study period, the average spending by states in the study more than doubled between 1997 and 2001. Although at the end of the study period some states had made substantial cuts, on the whole funding for subsidies was much more generous in 2001 than it was in 1997.
Throughout the study period, most of the states made liberal use of the Federal TANF Block grant as a source of child care funds. At its peak in 2000, TANF funding accounted for 20 percent or more of child care expenditures in 15 of the 17 states, and 40 percent or more in six states.
States served many, many more children with subsidies in 2002 than they did in 1997. In most of the states in the study, the number of children who received subsidies increased by 75 percent or more; in five of the 14 states that could provide information for both 1997 and 2001, the subsidy caseload more than doubled. In 10 of those 14 states, at least 25 percent of families estimated to be eligible under state rules were supported by subsidies. The great infusion of subsidy funds in the study communities during this time period meant that a substantial portion of the regulated supply was paid for through the subsidy system, even though in some states subsidies also supported considerable amounts of unregulated care.
In addition to increases in the numbers of children served by subsidies, there were significant increases in efforts to improve the quality of child care purchased, as evidenced by the average amount of subsidies spent per child served, as well as the proportion of the CCDF set aside for quality-improvement activities. In 10 of the 13 states that provided sufficient data, the amount of subsidy per child served increased between 1997 and 2001. In that same period, the adjusted amount of quality spending doubled or more than doubled in 13 of the 16 states reporting sufficient data.
In the majority of states in the study, most of the arrangements supported by subsidies were in centers. Whether in centers or in family child care homes, most of the subsidized slots in the majority of states were also regulated. These two general statements, however, mask a great deal of variation among the states. The distribution of types of care purchased by subsidies differed greatly across the study states. Individual state patterns were fairly consistent across the study period, despite concerns at the outset of the study that subsidies would increasingly be used to support unregulated providers.
Although there were some important trends in spending and in the choices states made, there were always many exceptions to every statement made in this report. There were substantial differences across the states in virtually every program dimension. Although the study was able to document this great variation, it was not possible to develop a method to cluster state child care policy decisions in order to classify the states in any meaningful way.

Period of Growth in Funding, Followed by Period of Some Targeted Cuts

The data collection for the state and community substudy took place during a unique period in Federal and state child care subsidy policy. After adjusting the calculation to address state differences in the cost of child care, the study team found that growth in child care expenditures more than doubled between 1997 and 2001. In 1997, after several years of nearly level Federal funding, states had the opportunity to expand subsidies rapidly, both through the funds from the Child Care and Development Fund and by using surplus TANF funds. As a result of this expansion in funding, states were confronted with broader interest in and political pressure related to subsidies. They also had to make a variety of decisions, including: which groups to serve and with what system of priorities, how to streamline delivery systems and make them more efficient; how to expand services rapidly despite state and local administrative constraints; and which parts of the system should be improved first. The first three years of data collection for the study documented these issues and how they were resolved.

The year 2000 was a turning point in many states. Although subsidy spending held steady or grew by slight amounts in the majority of states, in two states spending decreased slightly between 1999 and 2000, and three other states reported decreases between 2000 and 2001.

The study collected expenditure data from 1999 to 2001, information on child care use through 2002, and information on child care policy changes through the spring of 2003. The fiscal picture drawn by this report therefore lags behind the other information gathered by the study team. In 2002 and 2003, many of the states in the study made cuts in some of aspects of their subsidy programs while trying to maintain the core program in terms of basic rates to providers, eligibility guidelines, or parents’ co-payments. These measures often included changing rules related to presumptive eligibility or handling of child absences, or eliminating higher rates for special circumstances such as care offered during non-traditional hours. While states may not have changed the basic rates, these other changes often resulted in decreased payment amounts to providers.

Some states, however, resorted to more stringent cuts, particularly in 2002 and 2003. For instance, by the summer of 2003, New Mexico had decreased its eligibility ceiling from 200 percent of the Federal poverty level (FPL) to 100 percent, Washington had cut its eligibility ceiling from 225 percent of FPL to 200 percent, and Ohio had reduced its eligibility ceiling from 185 to 150 percent of FPL. In fact, eligibility ceilings in Illinois, Indiana, Louisiana, and Michigan went down 20 percent or more between 1999 and 2003. Also, over the course of the study period, North Carolina, New Jersey, Minnesota, and Illinois increased co-payment requirements for families.

These cuts in the generosity of subsidies, while painful to many state administrators, were made in a context markedly different than in pre-PWRORA times. The tough decisions about how to deal with reduced funding were made from a base of a much larger amount of Federal funding for child care than had been available in the mid-1990s and before. Cuts were not made across the board, nor did states necessarily reduce the generosity of the program. For instance, during the time period 1999 (a peak year) to 2002, eligibility ceilings went down by 5 percent or more in eight of the states in the study, but they also went up by 5 percent or more in four of the states. And while co-payments for families at 33% of SMI went up by 10 percent or more in four states, co-payment burdens for these families were reduced by 10 percent or more in six states.

Perhaps part of the reason that these cuts felt so painful to many state administrators is the perception that, even at its peak, subsidy funding was insufficient to meet demand. Consistently throughout the study period, the majority of the states in the study were unable to serve all the eligible families that requested subsidies. In 1999, 12 of the states had waiting lists; at the end of the study period, the number grew to 13. In addition to trying to provide subsidies to all eligible families that requested them, many state policymakers described consistent pressures on them throughout the study period, even when there was rapid expansion in funding. These included pressures to make the subsidy more valuable—by making co-payments less burdensome and payments to providers more generous.

If, at its peak, funding was perceived to be insufficient in many of the study states, how much subsidy funding was enough? The study did not attempt to answer this question, nor to estimate how much of the demand for child care was either met or remained unmet in the study states. These are complex questions, which are related to numerous factors. One major factor is the nature of the child care subsidy itself. In each state, the subsidies for parents were different in terms of the maximum price of care for which they would pay, the maximum amounts that parents could earn, the amount that parents at varying income levels were required to contribute, and the rules that providers had to agree to in order to receive subsidies. It stands to reason that a relatively more valuable subsidy, for which relatively more families were eligible, would be more in demand than a relatively less valuable one with more restrictive eligibility. It would therefore be difficult to sum up the answer to the question across the study states, even if the amount of unmet demand (i.e., the number of eligible families who applied for subsidies but were turned away) could be accurately measured.

Sources of Funding for Child Care Subsidies

The TANF block grant was an important part of the story of the unprecedented increase in the amount of subsidies available to low-income families during the study period. The CCDF has not been the only major Federal source of child care support. Starting in 1999, the TANF block grant was a major source of funding for child care in the study states. The use of TANF funding peaked in 2000, when, in 15 of the 17 states in the study, more than 20 percent of their child care expenditures came from their Federal TANF Block grant; in six of these states, 40 percent of child care expenditures came from the TANF block grant.

While the CCDF and TANF block grants were responsible for the large majority of spending, other sources of child care funding were important for some states. Several of the states in the study— California, Illinois, and Massachusetts, for instance—have a long tradition of spending state dollars on child care subsidies. Throughout the study period, states with a history of spending their own dollars for child care subsidies (above and beyond the amount needed to draw down Federal funds) continued to do so consistently, while those that in the past had not spent more than was necessary to draw down their Federal allotments tended to continue to do that as well.

How much states would spend of their own dollars on child care was an open question when the PWRORA was passed in 1997. There had been some concerns that the act would change states’ incentives to invest their own dollars. With more TANF funding available in states in the late 1999s, due to falling TANF caseloads, some of these concerns were amplified because some policymakers and advocates saw TANF as a ready source of funding to replace state dollars spent on child care assistance. However, key informants from many states considered child care a fundamental part of their support to low-income families and, in fact, an extension of welfare reform efforts, in that by providing child care assistance to low-income families they were stabilizing their employment and reducing families’ need for direct assistance.

At the peak of the spending, in 2000, key informants described concerns about relying heavily on TANF funds to maintain relatively higher levels of funding for child care subsidies. Even a mild recession, they feared, could result in higher TANF caseloads. This event would be likely to reduce the surplus. States would be faced with the choice of replacing Federal dollars with state dollars to maintain the current high spending levels, or of cutting subsidies for many low-income families.

However, in many states, during the recession that began in 2000, TANF caseloads did not, in fact, climb back to their 1997 levels. And while the expenditure of TANF funds, either directly or transferred into the CCDF, may have decreased substantially in later years, there was no strong indication that this was happening consistently in 2001, the last year for which the study collected financial data. In 2000, 20 percent or more of subsidy expenditures came from TANF in 15 of the 17 of the states; this was still the case in 13 states in 2001. TANF allocations, as a percentage of spending, went down in 10 of the states between 2000 and 2001 but also increased in seven of the states.

Growth in Numbers of Children Served and Amount of Regulated Supply Supported

By every measure, the increased expenditures described above were reflected in the growth of numbers of children served in the study states. In 10 of the 14 states that could provide data for both 1997 and 2002, the number of children receiving subsidies grew by 67 percent or more.

Measurement of numbers served was an important and thorny issue for the research team. Given that both the eligibility rules and the value of the subsidy are different in each state, the team faced the challenge of identifying ways to draw useful comparisons among the states. One could look at the percentage of state-eligible children served in each of the study states, but the denominator in each state is different. Using this measure, it is possible to determine the degree to which each state was meeting its own goals, but a direct comparison of the proportion of similar children the different states served is not possible. In order to do the latter, in the first interim report the team used as a comparison group all families with employed parents, children under age 13, and incomes below 85% of SMI; in this report we used 62% of SMI to develop a denominator to compare states’ performance. Both approaches (i.e., comparing numbers served to a number of “state-eligible,” and using a comparison group of a number below SMI) have drawbacks if they are to be used to determine the degree to which states’ served potentially “eligible” families.

The first drawback is that families’ eligibility stems from their need for child care for specific times in the day that single or married parents are working or, in some cases, attending job preparation, education, or training programs. Some parents have arranged their work schedules to coincide with their children’s school hours or, in the case of two-parent families, to maximize the amount of time that a parent can be home to care for the child. Therefore, families’ earnings levels and information about whether or not parents are employed are insufficient to determine eligibility. Second, if subsidies are valuable enough, it is possible that they could change incentives so that non-working parents could choose to work. Therefore, a more valuable subsidy may influence unemployed parents to become employed, which would move them from the “ineligible” to “eligible” group. The final drawback is the fact that there is still public debate about which families should be eligible for subsidies, and setting the maximum income at any particular level for comparison might indicate a policy viewpoint that this is the accepted level of “need” and the proportion served is the amount of “need” that has been “met.”

The use of 62% of SMI to establish a comparison group in this report, however, is meant to enable the reader to compare the experiences of one state with those of another by developing a per capita number served, so the experiences of a state with a relatively small population could be understood when compared with that of a state with a relatively large population. It was chosen because this was the average eligibility ceiling among the states in 2002. And it did serve to illustrate the wide differences among the states. For instance, in 2002, Washington provided subsidies to nearly two and a half times more families under 62% of SMI than did Virginia.

Whichever measure one ultimately chooses to use, it is still clear that the gains it reflects are considerable. In 2002, in 11 of the 14 states, at least 25 percent of the children estimated to be eligible by virtue of income and earnings under state rules received subsidies. In Illinois, 54 percent and, in Michigan, 44 percent of state-eligible children were served. In 2002, in Washington, 32 percent of the families with incomes under 62% of SMI received subsidies compared with 20 percent in 1997; in New Mexico, 24 percent of these families were served in 2002 compared with eight percent in 1997.

What were the effects of subsidies on families’ child care and employment decisions? This substudy does not provide any answers to these questions; however, the Patterns of Child Care Use Report of this study, which describes the results from a household survey within the 25 study communities, addresses questions related to subsidies and the type of care that families select.

A second set of questions relates to the effects of subsidies on the child care market. The state and community substudy collected data about the regulated supply from CCR&Rs and other organizations in 25 counties in a peak year of 2000. By using the number of the regulated slots, and comparing it with the number of arrangements paid for by subsidies, the research team was able to estimate the proportion of supply supported by subsidies. In eight of the 20 counties that were able to report data on regulated supply and on subsidized arrangements, nearly one-fourth or more of the regulated county-level supply was at least partially supported by subsidies. (Many subsidized families also made co-payments to these providers for the slots.) We would expect that, in low-income neighborhoods within these counties, this proportion would be much higher. To the degree that these 25 counties represent those with moderate or high poverty rates in the country,1 it appears that the government has become a major purchaser of regulated care in such areas. Although the study documented the growth in subsidies but not the effects of that growth, it seems reasonable to conjecture that subsidies must be responsible for influencing at least some aspects of child care supply in terms of its price, its quantity, and/or aspects of its quality.

Growth in Efforts to Boost the Value of the Subsidy and Improve Quality

At the outset of the study, there were concerns that pressures on states to increase the numbers served by subsidies would cause state policymakers to shortchange concerns about the quality of the services supported. These concerns, however, were not borne out by the experiences of many states.

The CCDF and other subsidy spending could potentially boost quality in two major ways. The first is that policymakers could improve the value of the subsidy in terms of the maximum payment rate to providers and by making other rules more generous, such as paying for child absences, so that more providers would be willing to accept the subsidy. They could also reduce co-payment amounts, making the subsidies more desirable for families, and the prospect of accepting them less risky for providers who are worried about collecting large co-payments from parents. An increased subsidy value would increase the purchasing power of parents as well as increasing the desirability of accepting the subsidy—giving parents the potential to select higher quality options available in their neighborhoods than they would without subsidies, or with subsidies that were less valuable.

The second potential way to improve quality is by providing support directly to providers in ways that are separate from the subsidy payment system, or by providing information to parents to help them become more informed child care consumers. These activities have been well-documented by many reports, including the Substudy’s First Interim Report (2000).

During the study period, many of the states in the study did both of these things. The adjusted value of the subsidy per child served increased between 1997 and 2001, in 10 of the 13 states that could report data sufficient data for both time periods; in 8 of these states it went up by 24 percent or more. Funding for quality-improvement and supply-expansion activities also tripled in 11 of the 16 study states reporting data for 1997 and 2001; it increased by nearly fivefold or more in seven of these states.

Increased Access to Regulated Care

At the outset of the study, there was concern that a combination of events—tight labor markets, new pressures on regulated child care caused by families moving from welfare to work, parents preferring to keep subsidy dollars “in the family” by using relatives, and relatively few center-based providers willing to accept subsidies—would mean that subsidies would go to license-exempt providers in unprecedented amounts. However, in nine of the 16 states that reported sufficient data in 2002, over half the subsidized arrangements were in centers, and, in 13 of the 16 states, at least 50 percent of the arrangements were considered regulated by the states. In 11 of these states, 70 percent or more of the subsidized arrangements were considered regulated.

Even though the majority of funding went to regulated sources, did subsidies go to higher proportions of relatives and unregulated non-relatives in 2002 than they did in 1997? Again, the answer is that they did not, in 10 of the 13 states that could report data for both years. The three exceptions were New York, Illinois, and Michigan. Michigan and Illinois experienced a substantial decrease in center-based care and a corresponding increase in relative care over the course of the study period.

These summary statements mask a good deal of variation. For example, in Michigan, 16 percent of subsidized arrangements were in centers, compared with 84 percent of arrangements in North Carolina. In 7 of the 16 states, 1 percent or less of subsidized arrangements occurred in the child’s home with a non-relative, while such care accounted for 9 percent or more of the arrangements in four states. It appears that subsidy policy, coupled with local labor and child care markets, all affect the patterns of subsidized care. The interactions of these three factors, however, could not be explored by this substudy.

Variations in Subsidy Policy

A challenge of the report was to describe the changes that occurred in state’s subsidy systems over the study time period using summary statements, while not masking the underlying, great variation in nearly every aspect of the subsidy systems that we examined in this study. Each summary statement needed to be accompanied by a multitude of caveats and qualifications. In the process of analysis, our hope was that we could simplify the process of describing states by discerning different patterns and configurations of states’ policy decisions that could characterize the states of the study in a few, relatively uncomplicated clusters. However, this goal proved to be impossible given the extent of variation in subsidy take-up rates, subsidy use for different types of care, reimbursement rates, and regulatory and other policies, and the relatively small number of states in our sample. Rather, the study highlights the fact that each state’s child care subsidy policies are unique and interact with the state’s child care regulatory environment, its other social policies, and local child care and labor markets.

In another report of this study, which describes results from the Community Survey of 2,500 low-income families that use non-parental, out of home care, we were able to identify ways in which the receipt or absence of a subsidy may influence parents’ decision-making. However, additional and future research is needed to untangle the effects that subsidy policies have on parents, children, caregivers, and child care and labor markets as a whole.

1 The study used the sampling frame provided by the National Child Care Study of 1990, but excluded counties with child poverty rates below 14 percent in 1993. (back)

Table of Contents | Previous | Next