Marriage Matters.

The Case for Marriage: Why Married People Are Happier, Healthier, and Better Off Financially, by Linda J. Waite and Maggie Gallagher (Doubleday, 260 pp., $24.95)

The Unexpected Legacy of Divorce, by Judith S. Wallerstein, Julia M. Lewis, and Sandra Blakeslee (Hyperion, 347 pp. $24.95)

Neither of these books could have been published forty years ago because then we did not have the facts on which they are based. But neither would have been published because nobody needed to be told that marriage is a good idea and divorce a bad one. Facts, and the times, have changed. And in our present culture, it is not yet clear that the message of either book will prove persuasive. As Linda Waite and Maggie Gallagher write on the first page of their book, over the last few decades we have managed to transform marriage, "the most basic and universal of human institutions, into something controversial." And as any academic expert reading Judith Wallerstein's book on divorce will know, a lot of scholars do not think divorce is such a bad idea.

Reviewing these books cannot be separated from reviewing a culture that has made what my parents thought was common sense into grounds for contention. The argument, of course, is not on the surface of things. Almost every American thinks having a happy marriage is a good idea and only a tiny minority believe that remaining single is an important goal. But what is true in theory is not true in practice. A lot of children are born to unwed mothers, many women getting married have to apologize for apparently giving up their "freedom," and divorce ends roughly half of all marriages. As sociologist Norval Glenn has shown, the proportion of people who are happy in their marriages has gone down since the early 1970s.

Waite and Gallagher set out to show that marriage is in fact very good for people in countless practical ways. Waite is a professor of sociology at the University of Chicago and former president of the Population Association; Gallagher directs the Marriage Project at the Institute for American Values in New York City. Both are well equipped to display the evidence, and it is extraordinarily persuasive.

Married people live longer than unmarried ones, and while they are alive have better sex lives than their unmarried counterparts. Being married helps them accumulate more wealth than single people and to live more cheaply than two individuals. Two cannot live more cheaply than one, but they can live more cheaply than two people who date each other but live apart. Married men are much less likely than bachelors to be criminals, alcoholics, or drug addicts, and this helps explain why they live longer. A wife tells them to settle down and stay away from nights at the barroom or the street corner, and most men respond. As a result their mortality rate is lower.

An obvious rejoinder to these facts is self-selection: Men who get money, avoid drugs, and stay away from crime are also the men who are likely to marry (and to find a woman who will marry them). Being married, therefore, explains nothing; it simply identifies people who know how to survive.

That argument cannot be entirely rebutted without a controlled experiment in which men are randomly assigned to marriage or bachelorhood; happily, social scientists can't do that. But the evidence strongly suggests that self-selection is not the key factor. No matter how men become unmarried-by being bachelors, becoming separated or divorced, or becoming widowers-they lose ground in terms of health.

Though marriage helps men's health and earnings, it is not true, as some suppose, that women are net losers. Married women acquire more wealth, their mental health improves, and their happiness increases. Raising children can, of course, put a lot of stress on women, but those who raise them single are psychologically worse off than those who are married. And for most women, having children is vastly more beneficial than whatever it costs in midnight feedings and endless chores.

Since marriage has so many benefits, especially for men, it is astonishing that it should be in so much trouble. If men on average are going to live longer, be happier, and earn more money when they marry, why do so many men reject it? A few have no interest in marriage because they are gay or confirmed bachelors, but for many decades an increasing number have also rejected marriage to women whom they have impregnated. It is as if something good were offered to people and, preoccupied with what is in their immediate interest, they reject it in favor of the pleasures of unmarried sex, brief cohabitations, unmarried parenthood, and frequent divorce. A bachelor may notice other men who are happily married, but he does not link their happiness to marriage. Instead, he repeats the tired jokes about not being "tied down" to the "old lady" when he could be "partying" with "babes." What babes? Sex for single men is less frequent and less enjoyable than for married ones. Men who choose to remain single spend most of their evenings at home, eating Cheez Dip and watching wrestling on television.

But a puzzle remains. If marriage is so beneficial, why has the divorce rate shot up, and what consequences has it had for children? Divorces are more common because they are more respectable. No-fault divorce laws passed in the 1960s and 1970s have encouraged divorce, but these new laws were not controversial when they were put on the statute books. The culture has changed, and the law changed to catch up with it. A culture that once made divorce next to impossible became one that made it easy.

The culture made it easy, in my view, because it has fully adopted the view that everyone has equal rights, that all agreements are private contracts that can be renegotiated at will, and that no higher authority or external public has the authority to ask people to act contrary to their preferences so long as they put no one else in danger. But Waite and Gallagher show that the erosion of marriage does put other people at risk, and Judith Wallerstein and her coauthors suggest how grave a risk it is.

Wallerstein has been following a group of middle-class, mostly white children of divorced parents for 25 years. In her previous books, she has argued that these children were harmed by divorce; in this one, she shows that the harms extend into their adult lives. Wallerstein and Lewis, both psychologists, together with Blakeslee, a science writer for the New York Times, have written what is essentially a biographical account of their subjects-Karen, Gary, Larry, Carol, Paula, Billy, and Lisa-with chapter-length stories about each. The authors' conclusions are that divorce not only hits many children hard when it occurs, but that the hurt lasts and sometimes deepens. At first they are frightened and angry, terrified of being abandoned. After that initial shock their home becomes a lonely place, one that often moves as their custody parent relocates or they shuttle back and forth between two parents. When they become adults, the suffering is greatest, as they find they have no model for love and commitment. "No one taught me," they complain. Some products of divorce learn to manage this; others do not, and many go on to create the same problems for their own children.

There is a problem in this analysis that has been pointed out over the years by several of Wallerstein's critics. The harms that befall children can result from tension and conflict in a married family as well as from divorce, and it is difficult to know whether pre-divorce conflict or divorce itself is the culprit. Moreover, studying divorced families without a control group of similar intact ones makes it hard to draw strong conclusions. Frank Furstenberg at the University of Pennsylvania and Andrew J. Cherlin at Johns Hopkins, both important family scholars, have suggested that divorce helps children who would otherwise be trapped in violent families or has a harmful effect that is short-lived compared to the long-term effect of incessant parental quarreling. Wallerstein and her coauthors do not reply in any clear fashion to these criticisms other than to say that now, many years after their research began, they have assembled and studied a comparison group. But that group was formed long after their study group, the two are not statistically identical, and its members were suggested to the authors by the children of divorced parents from among their one-time friends.

This book, then, like Wallerstein's previous books, is not so much a work of social science as a collection of biographies linked with therapeutic suggestions. As a social scientist, I am worried when people draw large conclusions from a study that was originally designed to do little more than suggest hypotheses for research. Most unsettling is the fact that while some of Wallerstein's critics have come around to her way of thinking, she does not refer to these developments or make any effort to show that other data can support her findings. Cherlin, for example, who has been quite critical of her work, published in 1998 an important study based on a close analysis of all the children born in Great Britain in the first week of March 1958. These people, their parents, and their teachers have been studied repeatedly over the last forty years. Cherlin and his colleagues showed that some of the emotional problems the subjects reported at age 33 could be explained by characteristics revealed when they were only seven years old, in most cases long before a divorce had occurred. These problems clearly intensified after divorce, suggesting that "divorce and its aftermath may have effects that persist into adulthood."

This statement is a very important admission in the long-running academic debate between those who minimize and those who maximize the effects of divorce, and it is based on a very sophisticated statistical analysis of long-term information. It would have been better if Wallerstein had told her readers more about what we actually know, as well as what she herself has suggested, not only to confirm some of her findings, but to suggest that others remain subject to revision or even refutation.

But if, as I think, her fundamental argument that divorce hurts children is true, what should be done about it? If we wish to have stronger marriages, and most Americans want this very much, how can this be brought about? Neither book, in my view, gives us much helpful advice. Both Waite and Gallagher and Wallerstein, Lewis, and Blakeslee say what anyone will say when we confront a massive cultural change: Get the facts out, rely on education, enlist the churches, "strengthen marriage." However, if we indeed confront, as I think we do, a profound cultural change, then education and the churches and all right-thinking people have caused precisely what these authors would now like them to change. Waite and Gallagher want to change tax and marriage laws to be more supportive of marriage, but not, I think, with much confidence that ending the IRS's marriage penalty will help keep families together. If a few hundred dollars a year can make a difference in marriage rates, then this country is in even more trouble than I fear. Wallerstein and her coauthors want paid family leave and more flextime, but they adduce no evidence that these changes would make any difference.

There may well be things we can do to ease somewhat the rate of out-of-wedlock births and slow down a bit the divorce rate. But to expect that we can make major changes is to deny the most obvious fact: Virtually every Western nation faces exactly the same problems as we do. Improving the prospects for marriage, therefore, cannot mean simply fixing some mistakes Americans have made; it must mean changing how a large portion of the most prosperous and best-educated part of the world has evolved. Not only can Americans not end a phenomenon that extends throughout the Western world, it is unlikely that anyone can change it by the conventional means of public policy.