Home Fires

Beyond Research

Why the latest day-care studies–and the oldest ones–shouldn’t matter.

No one can say what the welfare-reform bill will bring, but one unlikely outcome already shows signs of taking place. Child-care policy may finally receive practical, rather than polemical, attention. For more than three decades, social scientists have been studying the costs and benefits of child care, and their work has yielded only one consistent result: Their research is seized upon, waved about, politicized, and in the end, nothing much gets done. But the polarized stalemate can’t continue much longer, for two reasons. Child care has become an inescapable public issue, and social scientists themselves have been expressing a new humility about child-care research. Politicians now have to stop hiding behind the experts and start improvising policy.

Does the welfare-reform bill provide enough money to take care of the kids whose single parents are now forced to work? The lines that get drawn on that central question are fairly predictable. Liberal child advocates, the Children’s Defense Fund prominent among them, regard the funding–$20 billion, available over six years–as almost surely insufficient to offer the poor a true choice of quality day care. “What we know from child-care research,” as a press release from Child Trends Inc. puts the standard position, is that “the quality of child care matters for children’s development, and it matters the most for low-income children.”

Conservatives are content to assume that a majority of women who leave the rolls will rely on free (or cheap) unregulated child care by relatives and neighbors, as the poor have generally done in the past. Ron Haskins, an influential Republican staff director on the House Ways and Means Committee, has been outspoken in invoking social science to defend this position. The latest research, he argues, has been unable to demonstrate with any certainty that the quality of day care has much of an impact on a child’s future. Why invest more money if it won’t make a difference in the long run?

The interesting twist here is that, as far as the science is concerned, conservatives are right, and liberals lag a wishful step behind when they cite sure proof that early day care decisively influences a child’s development. To make sense of the shakier expertise both sides now confront, it helps to trace the history of child-care research through three unsettling waves during the past quarter century.

The first wave crested in the 1970s, when researchers faced the question: What happens when mothers work? The mood was apprehensive, the phrasing negative–what harm was being done to babies whose mothers leave them in the care of others? Psychologists looked at maternal-infant bonding as measured by the so-called “Strange Situation” test, in which they observed the partings and reunions of babies and mothers and graded the baby’s behavior from secure to “avoidant.” All in all, the evidence seemed to suggest that infants who spent more than 20 hours a week in nonmaternal care risked being less attached to their mothers, and risked displaying uncooperative behavior later in life.

When these findings left the lab, they were promptly construed as political propaganda. They were part of a “backlash against the women’s movement,” declared Sandra Scarr, a professor of psychology at the University of Virginia and prominent day-care expert (now the CEO of KinderCare), in the strident mood of the moment. “The advice for women has always been to get out of the work force. This is just another way of saying the same thing.”

The second wave of studies followed in the 1980s. They were more textured and hopeful, the question behind them less dour: What kind of nonmaternal care can help infants and children? Longitudinal studies sought to measure children’s “development”–their language skills, school achievement, cognitive gains, social and emotional adjustment, etc.–given different kinds and quality of care in their early years. When the widely varying studies detected short-term benefits from better-quality care, liberals and child advocates trumpeted them. Meanwhile, conservatives seized on findings that such benefits faded in later years.

Day-care research since the early 1990s has been characterized by what the experts call a more “ecological” approach. Studies have lately aimed to take into account children’s family characteristics–parental living arrangements and attitudes, income, experience with welfare, etc. But as Sandra Scarr confessed in a recent paper surveying the child-care landscape, the results have pulled the rug out from under two decades of work. Taken together, these studies seem to show that, given basically safe child-care settings, the quality of care has next to no measurable, independent effect on children’s long-term fates. Poor day care doesn’t seem permanently to harm kids whose lives are otherwise in OK shape. Nor can good day care be solidly proven to give a dramatic or lasting boost to kids whose home lives are a mess. Sometimes it can help, but a child’s family appears to be what really counts.

In retrospect, it’s clear that earlier studies made too much of small effects on development that were derived from very small samples. The complex interaction between family and day-care situations–to say nothing of individual temperaments–defies ready measurement. But where does that leave us? Liberals and child-care advocates, no longer armed with evidence to bolster their demands for big investments in child-care quality, have two choices. They can either rely on outmoded studies to argue with conservative politicians, who, for the moment, have the more authoritative pretext for scrimping on day-care cost and quality. Or they can follow the example set by Scarr. Acknowledging the new research (and doing some of it herself), Scarr still champions quality, but now on more immediate, less quantitative grounds. Her approach points the way for liberals to gain credibility by shedding the onus of being “social engineers,” and take their turn at playing the populist anti-expert card.

Liberals and child advocates can now explain, as conservatives have before them, that social policy isn’t about enforcing officially approved “choices.” They can say that, regardless of what the experts tell us, it’s perfectly obvious that cheap care by relatives and neighbors shouldn’t become the only feasible alternative for struggling mothers. Parents can make the best decisions for their kids, and if they are looking for quality day care, it isn’t likely to be because someone in a lab coat tells them it will mean an IQ 4.6 points higher at age 15. It will be because they hope it may mean a happier, more secure week for their kid and a less anxious one for themselves. What’s more, the chance to take an active choice in the matter as a parent may itself be a step toward the more successful, better monitored arrangements that poor working parents–and more middle-class parents, too–say they need and want.

A similarly pragmatic spirit beckons on the conservative side. Republican Gov. Tommy Thompson of Wisconsin and Mayor Rudolph Giuliani of New York, the conservative vanguard on the issue, show no inclination to exploit research that says, in effect, Why care about day-care quality? Instead they’re busy scrounging for funds to spread good care more widely. Thompson recently came up with $25 million more for “affordable quality child care” as welfare mothers head to work–$20 million to ensure that the working poor aren’t edged out of the subsidized care they count on, and $5 million to increase the supply and quality of overall care. Of course, this decision raises plenty of questions of its own. Is $25 million more going to be enough in a state like Wisconsin? How should that money be used–on more informal arrangements or on more day-care centers, on stiffer regulations or on other financial incentives to improve quality? On some of everything? The political debate over those questions will be more practical and less ideological, paradoxically enough, if scientific claims about a long-term payoff, or the lack of one, are left out of it.