With one out of every two American mothers returning to paid employment before her baby's first birthday the call for high-quality affordable child care is being heard with increasing frequency and urgency. Many employed parents complain that licensed child care is too expensive and that they must make do with informal arrangements of uncertain quality and dependability. For some mothers, the high cost of child care discourages them from seeking employment; other women remain childless or limit the size of their families for the same reason.
In a recent report, Who Cares for America's Children?, the National Research Council (NRC) set out an agenda for child care policy in the 1990s, identifying three major policy goals: to achieve uniformly high quality in child care; to improve accessibility; and to make child care more affordable.
To achieve higher quality in child care, the report recommends a low child/staff ratio and more education for child care workers. Although the report's policy goals are all equally admirable, they may not, in fact, all be achievable. Indeed, the NRC's recommendations for child care workers potentially conflict with its other two goals of accessible and affordable care.
Discussions of child care policy must, at the outset, address two conflicting economic issues: the high cost of child care and the low pay of child care workers, whose hourly earnings are among the lowest of any major occupation. Child care workers earn 30 percent to 60 percent less per hour than women who serve children as kindergarten or elementary school teachers. Even more startling, the hourly earnings of child care workers are 40 percent less than those of other women workers who have the same level of education.
Why do child care workers earn so little? Their wages, like everyone else's, are influenced by the interplay of numerous factors, many of which can be analyzed under the familiar categories of supply and demand: Do child care workers have the qualifications that would enable them to earn more in other occupations? Are there considerations other than wages that lead women to accept the low pay of child care rather than seek higher paying jobs? How much are those who buy the services of child care workers willing and able to pay? What value do they place on higher qualifications for child care workers? What is the relation between child/staff ratios and wages? The answers to these questions can help to frame the policy debate over child care, so that we may determine how much Americans are willing and able to pay for child care, and how that money should be spent.
Why Earnings Are So Low
In part, wages are low because of the characteristics of child care workers (though that is only part of the story, since greater demand for child care might raise wages and attract higher-skilled workers). To understand the "supply-side" part of the story, we looked at sample data from the 1980 Census on the hourly earnings of child care workers, teachers, and women in other occupations. (Data from the Current Population Surveys for March 1980 and March 1987 indicate that conclusions based on 1980 data should hold for the late 1980s as well.)
After various exclusions, our final sample consisted of 103,210 employed women, of whom 859 were child care workers. Of this group, 264 were employees in the private sector, 380 were public employees, and 215 were self-employed as providers of family day care. Because the level of earnings and the occupational differences in earnings depend on whether the workers are private or public employees or are self-employed, this three-way distinction is maintained throughout the analysis.
Table 1 shows the hourly earnings for each occupation-class of worker group as a percentage of the earnings of private employees in "all other occupations." We see that private child care workers earn only 59 percent as much as the base group. Public child care workers earn 75 percent of the base group, and 67 percent of what public employees in "all other occupations" earn (75/112). The self-employed child care workers have the lowest earnings of all. The average wage of the base group was $5.20 per hour in 1980; by 1990 this would have grown to approximately $9 per hour. The relative wages shown in Table 1 probably understate the compensation gap between child care workers and other employed women because they do not take into account fringe benefits such as health insurance and pensions. These benefits are probably smaller, relative to wages, for child care workers than for teachers or other employed women.
What kind of women work at child care? They differ from teachers and women in other occupations in several important ways, as is shown in Table 2. These differences can help us understand the large wage gaps revealed in Table 1. Years of schooling, for instance, often explain much of the difference in occupational wages, and we see that child care workers have much less education than teachers. The average schooling for "all other occupations," however, is about the same as for child care workers. Women who work in child care are much less likely to be childless and more likely than other women workers to have three or more children of their own. Five years earlier, only 39 percent of the child care workers had a full-time paid job; and another 39 percent did not have any paid work. In short, teachers and other women workers show a much firmer commitment to the labor force.
Travel time to work is important because jobs in residential areas tend, on average, to pay less than those that must recruit workers from longer distances. Child care workers, on average, have the shortest commute. The last two characteristics listed in Table 2 are based on all the workers (women and men) in the occupations. Many researchers have found that wages for both women and men tend to be lower in occupations dominated by women. The last line of Table 2 shows that approximately half of the child care workers work less than a full week, whereas the proportion of part-timers among elementary and secondary school teachers and "all other occupations" is only about one in five. This difference is important because occupations that include large numbers of part-time employees tend to pay lower wages not only to the part-timers but to full-timers as well.
How different would the wages in these occupations be if the women in them shared the same characteristics (at least those listed in tables 2 and 3)? To answer that question, we used the statistical technique of multiple regression. Even with this statistical adjustment, child care workers are still far below the norm. In the private sector their adjusted wages are only 74 percent as large as in "all other occupations"; in the public sector the adjusted ratio is 83 percent. When child care workers are compared with elementary school teachers, the adjusted wage ratio for private employees is 80 percent; for public employees it is only 69 percent.
The large wage gaps shown in Table 3 are probably attributable to several different factors for which data are not available. Unionization, for instance, undoubtedly accounts for a portion of the high adjusted wage enjoyed by teachers in the public schools. Numerous studies of the effects of unions on wages, however, suggest that even if teachers were completely unionized, their wages would be only about 15 percent higher than an occupation with zero unionization. Data covering quality of schooling, subjects studied in school, and other worker characteristics valued in the labor market are also missing and might help explain the gap in wages. Finally, it is probable that many child care workers enjoy working with young children and accept the low wages even though they could earn more in other occupations.
Our multiple regression analysis provides a basis for estimating the contribution of the various characteristics to the explanation of the unadjusted wage gap shown in Table 1. This contribution depends on two things: the size of the effect of the characteristic on wages in general, and the difference in the characteristic between the occupations under comparison. If a characteristic does not have much effect on wages in general, or if there is not much difference between the occupations with respect to that characteristic, it cannot be a significant explanation for the wage differences between occupations.
When comparing child care workers with "all other occupations," the biggest effect comes from the opportunities for part-time work afforded by child care. This one feature of child care jobs would explain about an 8 percent difference in wages. Other characteristics that help explain the wage differential are work status five years earlier and travel time to work. For the comparison of child care workers and elementary school teachers, by far the biggest effect is attributable to education. On average, child care workers have about four years less schooling than teachers; this difference alone would explain about a 28 percent difference in wages. Part-time hours is second in importance, and work status five years earlier is third. Travel time to work is important among private employees, but there is little difference in this characteristic between child care workers and elementary school teachers in the public sector.
In short, child care workers earn less partly because their individual qualifications would command less in other jobs as well, and partly because the positive features of child care attract some women even though they could earn more doing other work.
The low wages are also partly explained by the fact that the parents and others who buy child care are apparently unwilling, or unable, to pay for more highly educated workers or for those with more work experience. Education and experience are rewarded much less in child care than in teaching. Through separate multiple regressions for each class of worker-occupation group, we estimate that for child care workers, an additional year of schooling is associated on average with only 3 to 4 percent higher wages. By contrast, for teachers an additional year is associated with 8 to 14 percent higher wages. (This is more than the typical salary increment within a single school district for one more year of schooling. Some of the increase probably comes from higher-paying districts attracting teachers with more schooling.)
For child care workers, previous work experience seems also to have little effect on wages. There was only a 4 percent difference between those who had a full-time job five years earlier and those who had no paid work at all; among teachers the differential ranged from 18 to 29 percent. These differences are all highly significant; that is, there is a very small chance that they are the result of sampling errors.
How Can Earnings Be Raised?
If child care workers earn so little, why is good child care so expensive? The major explanation is the low child/worker ratio. The wages that a child care center can offer its workers depend entirely on three factors: (1) the gross revenue per child; (2) the proportion of revenue taken by overhead; and (3) the number of children served by each worker. To raise wages for American child care workers, it will be necessary to increase the revenue per child, decrease the overhead percentage, or increase the number of children cared for by each worker. These economic realities need to be moved to the center of the current debate over child care policy, since they constitute the crux of the dilemma.
Consider the problem of increasing gross revenue. The U.S. General Accounting Office (GAO) has gathered data from "high quality" full-day, full-year, preschool and early childhood education programs accredited by the National Association for the Education of Young Children. These centers, which represent the very top of the field, reported gross revenues per child of $4,200 (or $4,800, if the value of in-kind donations is included) in fiscal year 1988. In that same year American elementary and secondary schools received only $4,000 per child. Can child care centers realistically expect to receive much more per child than the schools do? An annual charge of $6,000 per child would put child care beyond the reach of most families or government agencies. Tax-supported subsidies could supplement private payments, but public aid is likely to be channeled into improved services for poor children, most of whom do not currently receive even the average standard of care, let alone the high standard prevailing in the centers included in the GAO study.
We could, as a change in public policy, transfer some of the $75 billion per year that now goes into higher education into subsidies for child care centers. Even a massive shift of public funds, however, is less likely to increase revenue per child than to create new opportunities for more children in such centers. At present only one in five preschool-age children is enrolled in a day-care center or nursery school; that number could double if more subsidies were available. The complaint of social critics that "there aren't enough places available" completely misses the point. The supply of child care places is very responsive to effective demand; the problem is not one of supply but of the willingness and ability of parents and others to pay for care.
The overhead expenses in the centers studied by the GAO amounted to 52 percent of gross revenues. After deductions for nonteaching salaries, rent, equipment and supplies, insurance, payroll taxes, and the like, 48 percent was available for the wages of the teachers and teachers' aides. A reduction in overhead would make more funds available for workers' pay, but many of these expenses are fixed by law and regulation. Furthermore, there is more pressure on centers to increase overhead than to reduce it. For example, improvements in safety and better supervision and training of child care workers would all raise the overhead percentage. There is little prospect for wage increases coming from this source.
The GAO report did not specifically describe the child/worker ratio, but it can be inferred from the annual salaries of teachers and teachers' aides, which the report states averaged approximately $12,000. Since approximately $2,100 per child was available for salaries (48 percent of $4,200), the child/worker ratio must have been approximately 6:1. This ratio is in line with the standards prescribed in the NRC report, which recommends a ratio of 4:1 or less for children under age two, from 3:1 to 6:1 at age two, and from 5:1 to 10:1 for children ages three through five.
An increase in the child/worker ratio seems to offer the greatest opportunity for higher wages. This is precisely the trade-off that is made in France. The French system of care for preschoolers is often hailed as a model for the United States, partly because French child care workers are as well educated and as well paid as their elementary school teacher counterparts. The French child/worker ratio, however, is double that of American nursery schools and daycare centers.
Failing a change in the child/worker ratio, child-care wages are likely to stay at their extremely low levels. If they do, the education and work experience of the child care workers will remain low, and the turnover rates in this occupation will continue to be high.
The Census and GAO data indicate that the parents who buy child care are apparently unwilling or unable to pay more for workers with additional qualifications. Or, seen from another perspective, these parents value a low child/worker ratio more than they do better-educated child-care workers who would also command higher salaries.
The contrast between child care centers and elementary schools puts the issue in sharp perspective. Revenue per child and overhead expenses are similar in the two settings. The only reason why elementary schools can pay wages that are more than twice as large as in child care centers is that the child/staff ratio is usually more than twice as great.
Should the United States follow the French model, giving up small groups in exchange for more highly trained, better-paid child care givers? The answer to that question depends upon the goals of child care and knowledge of the effects of such a change on children. We cannot say for sure what the effects would be, and, in any case, the choice rests, at least in part, with parents. Whether or not we embrace the French model, however, we cannot continue to evade the fundamental realities of child care economics. There is simply no way simultaneously to hold down the cost of care, reduce the child/worker ratio, and increase the pay and qualifications of child care workers.