In her Christmas 2007 campaign ad, Hillary Clinton was shown arranging presents labeled “Universal Health Care,” “Alternative Energy,” “Bring Troops Home,” and “Middle-Class Tax Breaks.” She then paused, looking somewhat puzzled, before delivering the punch line: “Where did I put universal pre-K?”

“Universal ­pre-­K” has become a politically popular campaign cause. Clinton is no longer a candidate, of course, but Barack Obama has promised an ambitious pre-kindergarten agenda; John McCain’s advisers have hinted that he will do the same. And why not? The rhetoric surrounding ­pre-­K programs is quite extraordinary: They close the achievement gap between low-income children and their more affluent peers; they prepare all children, including middle-income children, for school; and they provide financial relief to working mothers who have been paying for child ­care.

Yet as the Clinton TV spot unwittingly suggested, universal ­pre-­K programs do not have an obvious place in today’s crowded child-care world. Sometimes called “the new kindergarten,” pre-K is in most cases just what its name implies: a year of publicly funded ­half-­day school before ­kindergarten—­for all children, regardless of whether their mothers work and regardless of family income. Pre-K has hardly enjoyed a universal embrace. Twice in recent history, attempts to create similar national programs foundered on controversy and went down to defeat. In California, voters recently turned their backs on a statewide plan.

In a 2006 referendum, the Golden State’s voters rejected universal “free” preschool by a margin of three to two. Proposition 82, “Preschool for All,” was backed by the activist actor-director Rob Reiner and the California Teachers Association; it would have given all California four-year-olds “equal access to quality preschool programs” for three hours a day for about eight months a year—to be paid for by a 1.7 percentage-point increase in the tax rate for single individuals making more than $400,000 and couples making more than $800,000 (almost a 20 percent tax increase, by the way). Although attendance was theoretically voluntary, the proposition would have effectively withdrawn government subsidies from other forms of care, so that families needing or wanting a free or subsidized program would have had no choice but to use their local school’s pre-­K.

The referendum sparked a statewide debate that went beyond the typical mix of platitudes, generalizations, and exaggerations. Yes on 82, the prime sponsor of the referendum, repeated the oddly precise claim of RAND researchers that “every dollar California invests in a quality, universal preschool program will return $2.62 to society because of savings from reduced remedial education costs, lower high school dropout rates, and the economic benefits of a better-educated work force.”

Opponents pointed out, however, that more than 60 percent of California four-year-olds were already in a child-care center, a nursery school, or Head Start, and that the new program would have subsidized the ­middle-­class families now paying for child care while, in the words of a 

 editorial, establishing “a cumbersome bureaucracy . . . under the state Department of Education, which has done a disappointing job with K–12 schools.”

Strangely, the overwhelming rejection of universal ­pre-­K by the voters of our largest state has had no discernible impact on the national debate. It’s not that California just happened to have more preschool programs than the rest of the country. Nationwide, about 74 percent of ­four-­year-­olds now spend time in some form of organized child ­care.

To understand what is going on, a little history will help. Beginning in the 1950s, a steadily higher proportion of married women with children took jobs outside the home. Between 1950 and 1970, the proportion of 

 mothers in the work force doubled, rising from about 20 percent to about 40 percent. (Single mothers have always had little choice but to work, or go on welfare.) In 1971, spurred by this change, as well as the emerging women’s movement, a group of liberal Democrats led by Walter Mondale (D.-Minn.) in the Senate and John Brademas (D.-Ind.) in the House pushed the Child Development Act through Congress. It was an expansive measure, designed to create a federalized system of child development services. Children were to be enrolled regardless of whether their mothers worked and needed child care, on the ground that all children would benefit from a ­government-­supervised child development effort.

Initially, key senior officials in the Nixon administration supported the measure, seeing child care as an important component of their approach to welfare reform. But after some uncertainty, President Richard M. Nixon vetoed the bill, famously criticizing its “communal approaches to child rearing over [and] against the ­family-­centered approach.” His ­veto—­and the specter of “communal” child ­rearing—­not only killed the bill but took the political wind out of the child-care issue for a decade. Mondale himself became alarmed by the backlash even in his politically liberal home ­state.

Even though they do not “need” child care, about half of stay-at-home mothers place their children in a preschool or nursery school.

Most liberal commentators have seen only con­servative politics in the Nixon veto, but even many supporters of a federal child-care program thought the bill was deeply flawed, in ways that its congressional backers may not have understood. The legislation would have jumped past the states to fund hundreds if not thousands of “prime sponsors” (mostly local governments and nonprofit organizations)—all to be selected by officials of the U.S. Department of Health, Education, and Welfare. The prime sponsors were, in turn, supposed to establish local “child development councils” composed of parents, children’s services specialists, and community activists. These local entities would then fund as many as 40,000 individual ­providers.

If this web of federally administered, ­community-­based programs sounds like an echo of the War on Poverty, that’s because it sprang from the same social ­agenda—­and many of the same activists. They distrusted state and local governments and wanted “community groups” in control. The bill’s supporters boasted that this nationwide cadre of ­well-­funded organizations would be a strong political force for their favored causes. Maurien McKinley of the Black Child Development Institute explained: “It is to the advantage of the entire nation to view the provision of day care/child development services within the context of the need for a readjustment of societal power relationships. . . . As day care centers are utilized to catalyze development in black and other communities, the enhanced political and economic power that results can provide effective leverage for the improvement of the overall social and economic condition of the nation.”

In the next ­three-­plus decades, child-care advocates struggled to come up with a formula that would be more attractive to voters, but they repeatedly overestimated support for government-provided child care for middle-class children and underestimated the desire of parents for choice and flexibility.

In the years after Nixon’s veto, tens of millions of American mothers entered the labor force. By the 1990s, about 70 percent of married mothers had left ­full-­time child rearing for jobs outside the home, and child-care options had proliferated. According to the National Institute for Early Education Research (NIEER), about 74 percent of all four-year-olds are in “formal” child-care centers for at least part of the day, while the remainder are in “informal” arrangements, a category that includes care by anybody from their parents or relatives to the lady down the street.

Married mothers entered the labor force in waves. First came married women with older children, who were in school anyway and often could take care of themselves after school. Then came those with young children, who needed someone else to care for them. In 1975, only 34 percent of mothers with a child under age three worked outside the home; by 1990, 54 percent did. Moreover, new mothers are quick to return to work. About seven percent do so within one month of their child’s birth, and about 41 percent within three ­months.

Some think that American mothers are in the process of completely abandoning their traditional child-rearing role, but the picture is more mixed. The influx of married women with children into the labor force largely came to a halt in the 1990s. About 30 percent of all mothers today still do not work outside the home. Include those who work only part time—most often less than 20 hours a week—and you will find that almost 50 percent of all mothers, and almost 60 percent of those with a child under three, are not in the full-time labor force.

Although some of these women might take ­full-­time jobs if child care were free, most have decided to delay returning to the labor force until their children are older. In fact, even though they do not “need” child care, about half of stay-at-home mothers place their children in a preschool or nursery school (for at least a year) because they want them to be with other children in a structured learning environment. For these mothers, government-­funded ­pre-­K might be a welcome financial break, but it would have little or no educational ­effect.

Except among women on welfare, the great increase in working mothers had taken place by the late 1980s, when child-care advocates made their second major push for a universal program. In 1987, the Act for Better Child Care Services, or the “ABC bill,” as its supporters happily dubbed it, was introduced in Congress. Like the legislation Nixon had vetoed 15 years earlier, the ABC bill sought to create a nationwide system of child development ­services.

This time, however, there was no Great Society model; the states would administer the program, although they were to be guided by local advisory councils. Each year, the states would distribute $4.6 billion as grants to child-care centers or, in some circumstances, as vouchers to eligible families. Families would be eligible to receive assistance on a sliding scale if their income did “not exceed 115 percent of the State median income for a family of the same size.” In ­high-­income states such as Connecticut and New Jersey, that meant a family of four with an income of more than $100,000 would have been eligible. Nationally, the average income cutoff for eligibility for a family of four was about $79,000. (Unless otherwise indicated, all dollar amounts in this essay are in 2007 dollars.)

The ABC bill seemed headed for easy passage until controversy broke out among its liberal backers over a new provision barring the states from expending child-care money for “sectarian purposes or activities.” In other words, no money for child care by ­religiously ­oriented organizations—even though 28 percent of all ­center-­based programs in 1990 were operated by religious ­groups—­unless they removed all elements of religiosity from their ­premises.

That provision was a late addition to the bill, apparently at the urging of the National Education Association and the National Parent Teacher Association. These organizations were interested less in the theory of ­church-­state relations than in maximizing the money available for public schools and their employees. And they worried that by using vouchers (thus avoiding strictures against federal aid to religious institutions), the bill would create a precedent for vouchers in K–12 education. Many of the advocacy groups that originally supported the ABC ­bill—­especially those representing religiously based providers, such as the U.S. Catholic Conference and its allies—were incensed.

While the fight over aid to sectarian programs festered for almost two years, another, and ultimately more significant, rift developed among the Democrats who controlled Congress. Key leaders in the House, led by Thomas Downey (D.-N.Y.) and George Miller (D.-Calif.), decided that any new child-care bill should provide greater assistance to ­low-­income families rather than attempt to start a universal child development system, as the ABC bill would. It is unclear whether they opposed a universal federal program in ­principle—­as Marian Wright Edelman of the Children’s Defense Fund ­charged—­or were simply being pragmatic. Their own explanation was that a universal system was unlikely to be funded (at least in any meaningful way) and that, in the meantime, ­low-­income families needed ­help.

Meanwhile, Congress had passed legislation that encouraged mothers to leave welfare for work. Downey, Miller, and their allies wanted to “make work pay” for these ­mothers—­by providing government-funded child care and by supplementing low earnings through an expanded Earned Income Tax Credit (EITC).

In 1990, Congress and President George H. W. Bush finally agreed on a law, much different from the original 1987 ABC bill, that created a $1.3 billion annual program called the Child Care and Development Block Grant and a new half-billion–dollar entitlement for families ­“at ­risk” of becoming welfare recipients. It also doubled the EITC, from $11.9 billion in 1990 to $24.6 billion in ­1993.

It is difficult to judge what would have happened had the original ABC bill become law, but the narrower Downey-Miller approach was a boon to low-income families. The EITC is now a $45 billion-a-year program, providing financial assistance to more than 23 million families. And the administrative structure it ­created—­especially child-care ­vouchers—­became the basis of the massive expansion of child-care funding six years later under President Bill Clinton’s 1996 welfare reform law. That year, the Republican ­Congress—­pushed hard by the Clinton ­administration—­decided that if mothers were expected to work, the government should help pay for child ­care—­the same argument that had appealed to Republicans as far back as the Nixon administration. In only five years, from 1996 to 2000, federal and related state child-care spending almost doubled, rising from $7 billion to $13.6 billion. Add in funding for Head Start, and the total rose from $11.7 billion to $19.9 billion. Spending has remained relatively flat since ­then.

The result has been an unprecedented increase in the number of children in government-subsidized child care. But more needs to be done. Only half of all eligible four-year-olds with low-income working mothers (and only 18 percent of those under age two) receive child-care ­aid.

Both the Child Development Act of 1971 and the ABC bill of 1987 foundered, in part, on the seemingly wide political opposition to a universal child-care program that ignores the immediate needs of low-income families. But rather than learn from this lesson, advocates are pushing yet again for a universal program. This time, the selling point is “school readiness” rather than child development, and the focus is only on placing ­four-­year-­olds in public schools. But the result is the same: a ­middle-­class–oriented program that does not meet the needs of low-income ­families.

Advocates claim that ­pre-­K programs do not have to be in schools, and that they would be happy to see existing child-care centers improved with ­pre-­K funds (though that would leave out sectarian programs). But the “quality” requirements these programs impose, such as college degrees and specialized credentials for teachers, are, in the words of 

, “written in such a way to favor programs at public schools.”

In any event, given the strong political support for universal ­pre-­K from teachers’ unions and the allied educational establishment, it should not be surprising that most state ­pre-­K money has gone to new programs in public schools. In the 2003–04 school year, about 90 percent of children supported by ­pre-­K funds were enrolled in public ­schools.

Why add a new, ­school-­based program for ­four-­year-­olds when, as we have seen, about 70 percent of all three- and ­four-­year-­olds nationwide 

 spend at least some time in some form of ­center-­based child care or Head Start? Wasn’t this goal of universality the political and programmatic hurdle that brought down California’s Proposition 82? Would it not be sounder policy to expand the programs that already ­exist?

Perhaps the politicians supporting universal ­pre-­K do not know the extent of existing preschool services. (That seems to have been the case in California.) After all, like the rest of us, they are constantly exposed to a barrage of complaints about the inadequacy of child-care services. And some governors seem to have been persuaded that a ­pre-­K program would raise test scores, thus helping to prevent the financial penalties for failing to meet the standards of the No Child Left Behind ­Act.

The advocates of universal ­pre-­K, however, know exactly what they are doing. In public, they justify creating a new program by ­claiming—­often with some ­hyperbole—­that existing programs are of such poor quality that displacing them will be a net good. Thus, Nathan James, a spokesman for Rob Reiner, asserted that as few as 25 percent of the ­four-­year-­olds in day care were in quality programs. Care for the others “could be ­baby-­sitting or throwing a kid in front of a TV set,” he ­said.

That kind of exaggeration—with its remarkable suggestion that the majority of parents hand their children over to dreadful caregivers—distracts attention from the real question: Would it not make more sense to improve the existing programs than to start up a fresh group of efforts whose quality is far from guaranteed? For example, “Project Upgrade” (funded by the U.S. Department of Health and Human Services) used rigorous evaluation techniques to test a revised curriculum for child-care centers in Florida. It raised test scores on at least some elements of cognitive development as much as the best state pre-K programs—at a much lower cost. (Because pre-K pays teacher-level salaries, on an hourly basis it costs about 50 percent more than center-based care.)

In private, advocates give a more plausible explanation. They say that the phrases “universal preschool” and “universal ­pre-­K” are meant to suggest the extension of public education. The idea is to finesse the major reasons why past efforts to enact a universal child-care program failed. If ­pre-­K is just adding another year to schooling, then it is not taking over child rearing (a prerogative carefully guarded by American parents). And if it is an education program, it might attract the children of ­stay-­at-­home mothers and would certainly justify taxpayer spending on ­middle-­class and more affluent families. (After all, schools are free to all, regardless of income.)

Justifying free ­pre-­K is politically important because, contrary to what the news media imply, two-parent families in which the mother works are actually much wealthier than those with ­stay-­at-­home mothers. As 

 complained, universal ­pre-­K makes a ­“tax­­payer-­funded preschool available to ­middle-­class and rich families, which can easily afford it.” Al­though other factors are involved, consider that in 2006 the median income for households with two earners was $76,635, almost 40 percent more than that for married-couple households with only one earner ($55,372).

Only half of all eligible four-year-olds with low-income working mothers receive child-care aid.

The key to this “pre-K is ­just ­another ­year ­of ­school” argument is the claim that, unlike Head Start, pre-K programs provide educational benefits to all children, not just the disadvantaged. “All children make phenomenal gains” in pre-K, claims Libby Doggett, executive director of the advocacy group Pre-K Now. Rob Reiner told the National Governors’ Association that pre-K programs produce a “huger impact” on how all children do “in school and later on in life.”

At first glance, the idea that starting school a year earlier would boost the learning of ­middle-­class children might make sense. (Let’s pass on the worry that many experts have about the negative impact of starting formal education too soon.) We want our children to do the best they can in school, so, presumably, the earlier they start preparing for school, the ­better.

Unfortunately, no scientifically rigorous evidence supports the claims of ­pre-­K’s impact on middle-class children. James Heckman, a University of Chicago Nobel laureate in economics, is one of the strongest voices in favor of early education for ­low-­income children, but here is what he says about applying the model to the middle class: “Advocates and supporters of universal preschool often use existing research for purely political purposes. But the solid evidence for the effectiveness of early interventions is limited to those conducted on disadvantaged populations.” As Bruce Fuller, an education professor at the University of California, Berkeley, and author of 

(2007), explains, “For ­middle-­class kids the quality of preschool centers would have to approach a ­nirvana-­like condition to present radically richer environments than the majority of ­middle-­class homes, or ­home-­based caregivers.”

It’s not that knowledgeable ­pre-­K backers don’t know this. Fuller reports on a conversation he had with one of the key foundation funders of the ­pre-­K movement: “When I asked [universal pre-K] benefactor Sue Urahn of the Pew Charitable Trusts why government should subsidize preschools for all families, rich or poor, she acknowledged that ‘you probably won’t get the degree of benefit for ­middle-­class children that you would for poor kids.’ But, she added, universality may bolster the political will to widen children’s access to, and to improve the quality of, preschool.”

So that’s the strategy: promise the middle class a free ­lunch. Thus far, it seems to be working. Each year sees an increase in the number of children in pre-K programs. In the 2006–07 school year, the NIEER reports, 14 states had 25 percent or more of all four-year-olds in pre-K, and three states had reached 50 percent.

In most places, ­pre-­K programs are simply being added to the mix of preschool programs, with little or no attempt to coordinate them with existing child-care programs or Head Start. The eventual goal, apparently, is to have universal ­pre-­K programs substitute for all programs that now serve ­four-­year-­olds.

But is it the right strategy? What about the nearly 500,000 ­four-­year-­olds in Head Start? And what about the almost 1.6 million ­four-­year-­old children of ­full-­time working women—children who need more than ­part-­time care while their mothers are on the ­job?

Pre-­K is already eating into Head Start enrollments. Last year, Congress responded to what was called “underenrollment” by allowing Head Start grantees to enroll more infants and toddlers, and to raise income eligibility ceilings. This is, at best, a temporary fix to a ­long-­term ­problem.

Nonprofit and ­for-­profit child-care centers face a subtler threat. ­Full-­time working mothers who use pre-K (whether because of its presumed quality or because it is free) no longer need their services. And because ­pre-­K fills only a few hours of each day, these mothers tend to patch together some combination of before– and ­after–­pre-­K activities for their children. Because they generally cannot use child-care ­centers for this purpose, children are more likely to wind up in informal care, provided by neighbors, relatives, and ­others—­the very care that pre-K advocates criticize most.

When researchers studying New York State’s universal ­pre-­K program raised the possibility that ­pre-­K programs “could negatively impact the enrollment of four-year-olds at nonpublic child-care centers and preschools,” a ­pre-­K advocate asked, “Is this necessarily an ­all-­negative outcome?”

Or perhaps advocates would prefer the Oklahoma solution. Using mostly federal funds, the state simply pays child-care centers for a full day for each child, even if the child is only present for four hours. (This practice is documented in government reports, but the folks in Washington either don’t know or don’t care about it.)

Another troubling aspect of the ­pre-­K movement is that it is a retreat from parental choice in early childhood arrangements, an approach that has been nurtured since the passage of the block grant bill in 1990. Since then, more than $100 billion in child-care subsidies has been distributed through ­vouchers—­with nary a ­problem—­while ­low-­income parents have had the freedom to choose the providers they want, largely without government constraints. (Even unlicensed providers can be used in most states.) But parents in neighborhoods served by pre-K have only one choice: send their children to the public program or dig into their pockets to send them to one of their own ­choosing.

Vouchers are controversial for K–12 education, but they have been widely accepted in the child-care ­world—­because the context is so different. Remember, the children involved are three-year-olds and four-year-olds. Even some strong critics of vouchers for the schools, such as John Witte, a political scientist at the University of Wisconsin, Madison, have concluded that 

 a “voucher system seems to be the best choice to maximize opportunity and equity and educational efficiency.”

Besides encouraging responsive programming and service improvement, vouchers provide a high degree of flexibility needed to accommodate the disparate needs of families. Some parents want, or need, only ­half-­day care; some need evening or ­after-­hours care; others need ­full-­day care, perhaps with extended hours. Some parents want their children cared for by other family members; some want to use neighbors; others want a nursery school; still others prefer a care center, perhaps in a church. Some parents may want all their children of different ages in one place; others may not care. Some parents will want their children close to home; others will want them close to work. The variations are almost infinite. Accommodating such variation is all but impossible in a ­top-­down, ­pre-­K ­regime.

Perhaps most troubling, universal ­pre-­K does little, if anything, to solve the most vexing educational problem facing America: the achievement gap that puts low-income, mostly minority children so far behind more fortunate children. On a host of important developmental measures, low-income children suffer large and troubling social and cognitive deficits compared with others. This translates into a ­life­long achievement gap that curtails the educational attainment, employment opportunities, and earnings potential of large numbers of ­children—­especially among African Americans, Latinos, and other disadvantaged ­minorities.

The achievement gap has many causes, from the poverty stemming from a history of discrimination and restricted opportunity to the ­child-­rearing styles of many disadvantaged families. Cause and effect are intermingled in multiple and controversial ways. Early childhood education is a potentially important remedy to some of these problems, but the plain fact is that the family is the primary teacher of young ­children—­and compensatory programs face a much larger challenge than pre-K advocates’ rhetoric commonly suggests. What parents do (and do not do) counts much more than any early education ­program.

Debate rages about how best to close the achievement gap, but all specialists agree that to be successful, programs must be focused on the children’s deep needs and be intense enough to make a difference. That means multiple years of educational and support services for the parents as well as the ­children—­and that simply is not something ­pre-­K and its three or four hours of ­school-­based services will ­provide.

Some observers think that, if ­pre-­K programs really worked for the middle class, they would widen the achievement gap. Bruce Fuller points out, “The well-orchestrated universal preschool campaign at once says their ­silver ­bullet will help all kids and close early achievement gaps. That’s pretty difficult to pull off. It means that children from ­middle-­class and wealthy families will accelerate in their development, and then poor kids will accelerate even more.”

Perhaps sometime in the future all American children will be in free child care, at least by the time they are four years old. But we seem far from that goal. One research group estimates that a universal pre-K system would cost roughly $55 billion a year, more than six times the roughly $9 billion the federal and state governments now spend on four-year-olds. If past estimates for the costs of other social programs are any guide, it would not be unreasonable to double that forecast.

Universal pre-K might be a boon to the middle class—depending on whether, in the end, it is their tax dollars that pay for it—but it would still leave unmet the much more serious needs of low-income children. Half of all eligible low-income working mothers still do not receive child-care subsidies. Would it not be wiser policy to help them purchase better child care than to channel more funding into pre-K programs that serve higher-income children whose parents do not necessarily work?

Twice before, efforts to create a universal program stalled in Washington. But this round’s education-based strategy may work. Although it failed with the voters of California, special interests hold much greater sway in the nation’s capital. So, to answer Hillary Clinton’s question: Universal pre-K is caught in the midst of middle-class and interest-group politics. As usual, the most disadvantaged children may lose out.

 is the Joseph J. and Violet Jacobs Scholar at the American Enterprise Institute for Public Policy Research and a professor at the University of Maryland School of Public Policy. He was the first director of the U.S. National Center on Child Abuse and Neglect, and is the author of

 Recognizing Child Abuse: A Guide for the Concerned and other books. 

 is a research associate at the University of Maryland School of Public Policy.

8078450265_dd33b0cab2_o.jpg

The case for universal pre-kindergarten isn’t as strong as it seems.

Douglas J. Besharov