Policy Review Banner

Too Few Good Men

Thursday, December 1, 2005

Kathryn Edin and Maria Kefalas.
Promises I Can Keep: Why Poor Women Put Motherhood Before Marriage.
University of California Press. 312 pages. $24.95.

Jason DeParle.
American Dream: Three Women, Ten Kids, and a Nation’s Drive to End Welfare.
Viking. 432 pages. $25.95.

Everyone knows families are not what they used to be. Marriage has declined everywhere, out-of-wedlock births have increased across the board, and single parents are now commonplace. Although that’s the received wisdom, it’s only partly true. As Charles Murray noticed decades ago and demographers have known for some time, the structure of families has diverged drastically by social class. Among women with no more than a high school education, the out-of-wedlock birth rate has grown rapidly since the 1960s and is now approaching half of all births. In contrast, single motherhood is still rare among college graduates, representing less than five percent of births among this group overall. Almost all college graduates still marry eventually, but marriage rates are dropping steadily among those without a high school degree. Divorce has declined among the well-off since the 1980s but is climbing among the unskilled. Racial variations confound these differences. Among college-educated women, the rate of out-of-wedlock childbearing for blacks is ten times higher than for whites. Although white marriages have achieved greater stability over the past 30 years, black marriages at every level dissolve more frequently. As a result, many more black children in all income brackets grow up with one parent. As noted by Sara McLanahan in her recent presidential address to the Population Association of America, these trends tell an ominous story: The offspring of the well-off receive a growing share of parental time, attention, and investment and grow up in stable and orderly homes. The less privileged frequently endure a fractured and chaotic family life.

Understanding these trends presents a challenge for students of family life. Why have the well-off and well-educated continued to live fairly conventional domestic lives while the families of the lower classes have fallen apart? This puzzle broods heavily over Kathryn Edin and Maria Kefalas’s Promises I Can Keep: Why Poor Women Put Motherhood Before Marriage and Jason DeParle’s American Dream: Three Women, Ten Kids, and a Nation’s Drive to End Welfare. Both focus on families at the bottom of the income scale, exploring the lives of the women (and some men) who occupy the precarious juncture between the working and welfare classes, and both paint a revealing, candid, and sometimes lurid picture of their domestic, reproductive, and personal lives.


In American Dream, Jason DeParle, a journalist who covers poverty and welfare issues for the New York Times, follows three African-American women and their intertwined families in Milwaukee, Wisconsin, in the period surrounding welfare reform under the Personal Responsibility and Work Opportunity Act of 1996. His account is based on months of observation and discussion with the women and their family members, boyfriends, welfare caseworkers, and employers, as well as with administrators and program directors statewide. The result is a finely observed portrait of unskilled women confronting the demands of new welfare policies and the chaotic conditions of inner-city life, deftly interspersed with a legal, bureaucratic, and administrative account of poor relief in Wisconsin during the reform era. DeParle is particularly concerned with whether reform has achieved one of its chief declared goals: to reverse the decline in the nuclear family that coincided with welfare’s decades-long expansion and is sometimes blamed on its influence. On this point DeParle proceeds cautiously, advancing more questions than answers. In the end he concludes, reluctantly and provisionally, that work-based welfare reform has failed to produce the hoped-for benefits. The lives of his own subjects confirm that the changes wrought by the new rules have done little to shore up their fractured families. Although welfare reform succeeds, after fits and starts, in transforming these women from dependent welfare mothers into reasonably steady workers, domestic disorder continues to roil their lives. As parents they are loving but erratic and ineffectual. Their shifting contingent of consorts, housemates, and boyfriends is as feckless as ever. Respectable married life eludes them.

In Promises I Can Keep, Kathryn Edin and Maria Kefalas, urban ethnographers and social scientists, draw their account from interviews with 162 single mothers in eight Philadelphia-area low-income neighborhoods. The discussions probe the forces, motives, values, and life conditions that lead most of these women to have children while still young and unmarried. Although welfare reform and the vagaries of low-wage work form the story’s backdrop, the primary focus is on patterns of childbearing and the relations between the sexes. The authors’ goal is to understand why extramarital births have increased and marriage rates declined among the least educated and least skilled population of urban women.

Edin and Kefalas are talented and sedulous ethnographers. Their field work is strong and careful and their portrait of poor women’s lives vivid and sympathetic. The book’s introduction offers promise of a fresh analysis: In the proper spirit of humility, the authors describe the fate of the working-class family as “perhaps the biggest demographic mystery of the last half of the twentieth century.” Unlike many academic sociologists, they adopt a strong pro-marriage tone and credit the evidence that stable, two-parent families benefit adults and their children. They likewise do not blink at the class split in family structure and marital behavior. They recite the bleak demographic facts and are skeptical of long-discredited explanations — such as a shortage of employable men — for the decline in marriage among the less educated.

Nonetheless, their book ultimately fails. Despite promising beginnings, the authors fall victim to tired social science dogmas. Their fealty to bad ideas hinders a full excavation of the rich lode of material they have so painstakingly assembled. They miss the message of their own fieldwork and the clear implications of broader social trends. The result is a lost opportunity to discover the true causes of family upheaval and to think constructively about the cures for its decline.

Why do the women in this study so rarely marry and so often end up as single mothers? Most express a strong desire to marry and view extra-marital childbearing as “second best.” Yet almost all remain single. The authors offer this explanation: Expectations for marriage have risen across the board. People now regard marriage as a luxury good rather than as a necessity. They refuse to tie the knot unless they have first achieved economic success. A house, a well-paying job, and enough money for a nice wedding are now needed before considering a trip to the altar. But few of the unskilled can make good on their aspirations because wages at the bottom have stagnated or declined. To their credit, the authors do not exaggerate the extent of these trends. Although they note (correctly) that unskilled men’s earnings have lost ground relative to college graduates’ and that some well-paying jobs have disappeared, they acknowledge that the overall economic prospects of men with a high school education or less are not significantly worse than in past decades when marriage rates were much higher. It’s not that most unskilled men are less able to support a family than they were decades ago; earnings for this group were always modest. Rather, the problem is that women — and men — expect far more.

In contrast, conclude Edin and Kefalas, having children carries no such inflated requirements. Babies need not await the achievement of an elevated position in life, because childbearing is a fundamental hallmark of female adulthood that is central to poor women’s dignity and identity. In the authors’ words, “women rely on their children to bring validation, purpose, companionship, and order to their often chaotic lives — things they find hard to come by in other ways.” In a perverse inversion of old values, these woman have come to regard lone motherhood as the ultimate heroic act, the proving ground of their responsible devotion to others.

At first blush, the authors’ theory about why marriage is unpopular among the less educated appears to explain demographic reality. Rising expectations generate a class divergence in marriage rates for the simple reason that the well-off are better able to fulfill those expectations than the poor and uneducated. Yet despite superficial appeal, the authors’ explanation just doesn’t fly. First and foremost, their conclusions are at odds with what their women subjects actually say. More broadly, the authors’ thesis cannot be reconciled with the full range of facts regarding racial and class differences in family structure. A growing body of social science evidence suggests that group mores and personal behavior, not insufficient resources, are the most important cause of marital decline.


This book is replete with evidence that men’s anti-social behavior, not unfulfilled economic expectations, is the main obstacle to matrimony among this group. The women do not complain of men’s failure to earn enough, but rather of their unwillingness to grasp opportunities, work steadily, and spend wisely. The objection is not to modest earning power, but to financial profligacy, defiant attitudes, and lack of work discipline. These women bear tales of their men mouthing off to bosses, alienating fellow workers, failing to get to work on time or at all, behaving erratically, quitting abruptly, or avoiding work altogether. What money the men manage to earn is seldom applied to family needs, but is dissipated on luxuries such as “alcohol, marijuana, new stereo components, computer accessories, expensive footwear, clothing and jewelry.” But poor work habits and financial irresponsibility are the least of it. The most vociferous complaints are reserved for men’s chronic criminal behavior, drug use, violence, and, above all, repeated and flagrant sexual infidelity. Most men made no effort to hide their frequent liaisons, which were often carried on simultaneously. More often than not, those relationships produced babies. Offspring by other partners loomed especially large as obstacles to stable and harmonious relationships. Women resented children fathered with other girlfriends as evidence of a man’s imperfect devotion and as a drain on his attention and resources. The presence of a woman’s children by previous boyfriends also produced conflict by undermining the man’s authority and engendering divided loyalties.

The circumstances are no different in Jason DeParle’s picture of inner-city life in Milwaukee and in other recent accounts of urban working-class existence such as Adrian LeBlanc’s Random Family and Orlando Patterson’s Rituals of Blood. The men in these books also openly reject sexual fidelity and flout the most basic standards of male responsibility toward women and children. The women, for their part, haven’t figured out how to influence men for the better or have given up trying. They are resigned to the situation and actively complicit in it. They don’t hesitate to sleep with other women’s men and willingly bear children by more than one father. Their response is to take what they can get: temporary companionship, fleeting love and attention, a good time, and, above all, the children their men leave behind.

These stories of conflict, distrust, and disappointment do not point to rising expectations, economic or otherwise. These women do not hold their men to new and higher standards. They want what women have always wanted: men who are steady, faithful, considerate, and industrious. The virtues they seek in a husband — dependability, fidelity, honesty, frugality — are those that women have always sought. What has changed is men’s willingness to fulfill these requirements by living up to age-old standards. Although the women foster the men’s bad habits by having sex and babies with them despite the men’s irresponsibility, they still cling to the old expectations. Their dashed hopes transform marriage into an impossible and unattainable dream.

That dismal values and reckless choices are the chief obstacles to marriage in low-income communities is a truth so obvious that even Edin and Kefalas can’t help but succumb. Telling concessions are scattered throughout the book. The authors describe the one man who comes closest to displaying the traditional bourgeois virtues as his neighborhood’s “equivalent of a Rhodes Scholar.” They admit that although job market difficulties are a strain on relationships, unemployment is “seldom the relationship breaker.” Rather, the real problem is

the man’s unwillingness to “stay working” even when he can find a job. Or he may blow his earnings on partying or stereo equipment. But most women point to a larger problem than a lack of money, such as chronic womanizing.

Elsewhere the authors acknowledge that their women agree on the elements of responsible male adulthood: Sexual fidelity, steady work, abiding by the law, and assisting with childrearing are all on everyone’s list. Above all, their subjects want the fathers of their children “to become a responsible adult.” The authors more than once concede that the men in these neighborhoods behave “considerably worse” than the women, that they seldom fulfill expectations, and that this all but dooms prospects for stable family life. These admissions catch the authors in their own contradictions. Their notion that the decoupling of marriage and childbearing can be laid primarily at the feet of rising economic expectations seems to come out of nowhere. The idea becomes more fantastical and less persuasive with every page. In a last-ditch effort to shore up their thesis, Edin and Kefalas rely heavily on some women’s direct statements of their desire to achieve material well-being — to secure a steady job, accumulate money in the bank, and even own their own home — before getting married. But in taking these wishes at face value, Edin and Kefalas ignore their deeper roots. What comes through is these women’s failure to see marriage as a long-term cooperative venture. Instead of a mutual striving toward economic prosperity, marriage is a precarious gamble. Economic independence serves as a hedge against their men’s inconstancy. As one mother states: “I want to have everything ready in case something goes wrong.” Her partner of the moment provides little reason to think that things will go right.


At times, Edin and Kefalas seem to appreciate that the most plausible explanation for these women’s material aspirations is that their men cannot be relied on. They concede that failed relationships and distrust “make marriage seem risky” and that women “mitigate this risk by holding marriage to a high standard both in economic and relational terms.” But in the end they fail to draw the obvious lesson. They are blinded by their grand idea: The quest for prosperity before marriage manifests the broader cultural desire, peculiar to the zeitgeist, for “more” of everything.


Why do the authors hold out for this position in spite of all evidence to the contrary? One factor may be the policy agenda that follows — one that embraces activist government and economic redistribution. The view that nonmarriage and out-of-wedlock childbearing are all about economic opportunity fits with the conviction that public money, policy gimmicks, and political will — not basic good conduct — are the solutions to family disarray. The problem is not that people are behaving badly or that — heaven forbid — one class is more prudent than another, but that our policies are inadequate. Material conditions, not moral commitments, are the source of domestic chaos. To change behavior, we must give the poor more resources. Raise economic prospects for the least skilled men and women and the problem will fix itself.

The problem with this outlook is that it’s hopelessly outdated. Decades of experience belie the view that economic transfers can rescue disintegrating families or that government programs can substitute for good conduct. There is little evidence that poor relief and welfare policy, whether strict or lenient, can effectively promote marriage, reduce out-of-wedlock childbearing, or turn men into responsible husbands and fathers. As DeParle notes, recent developments are not to the contrary. Despite welfare reform’s resounding success in moving women from welfare to work, the two-parent family has shown few signs of recovery.

Given that public policy has never yet revived the nuclear family, it is not surprising that social science has yet to explicate the link between resources and family form. How lack of money dictates behaviors destructive of marriage remains an enduring mystery, especially since low-income families among some cultural groups remain exceptionally strong. Likewise, although more resources may ease family life, it’s not clear how they can cure the profligacy, violence, and sexual recklessness that destroy relationships. Despite decades of social science investigation, the black box of causation is firmly closed. The route from a higher hourly wage back to the bourgeois virtues remains obscure. The authors of these volumes can’t draw us a road map and they don’t even try.

If anything, the best evidence we have suggests that causation runs in the opposite direction. Moral commitment generates resources; wise behavior can secure economic well-being. As Christopher Jencks and others have argued, even people with little education and few skills can greatly reduce their risk of poverty and provide a decent upbringing for their offspring by following three simple rules — graduate from high school, work steadily, and marry before bearing children. Although the people depicted in these accounts rarely achieve this hat trick, the few who do confirm the point. Consider Jewell and Ken, a couple in American Dream. Despite a checkered history that includes drugs, out-of-wedlock childbearing, spells on welfare, a criminal conviction, and imprisonment, this couple manages to settle down to steady work and a sober and law-abiding life. Their jobs as a nursing aide and pizza delivery man, to which they apply themselves with devotion, bring in about $40,000 a year. They remain faithful to each other and have a child together. Although they can’t quite bring themselves to marry, they acquire a decent apartment, basic household appliances, and an impressive array of electronic gadgets, including cell phones, a personal computer, and a video camera. By dint of great effort, they achieve, in DeParle’s words, “a toehold on a lower-middle-class life.” The hope is that they will eventually move up from there.

Although Ken and Jewell prove it can be done, they are but one data point — what demographers contemptuously dismiss as “anecdata.” But Edin and Kefalas don’t really improve on this — their project is anecdata writ large. Although revealing in some respects, their book stands as a poster child for the dangers of academic ethnography. Their narrative method, which allows a focus on a narrow slice of society, makes it too easy to ignore inconvenient facts. Not only do their conclusions fly in the face of their own investigations, but they flout a growing consensus among sophisticated demographers that material and economic factors cannot fully explain the widening class divide in family structure.

Social scientists have long labored to explain the emerging disparities in family structure by pointing to social conditions and economic obstacles. Factors thought to impede marriage among the less skilled include stagnating male earnings, new women’s labor market opportunities, and lopsided sex ratios from male incarceration, early death, and male unemployment. But as Christopher Jencks and David Ellwood at Harvard have observed, the usual suspects do not account for more than a small portion of observed race and class shifts in marital and reproductive behavior.


Take the bromide that low marriage rates can be traced to male unemployment. That does nothing to explain the rapid decline in marriage among working-class men with jobs — a decline that accounts for a large portion of nonmarriage within the group. Nor are unskilled men’s low earnings a full explanation, as the mothers they fail to marry are increasingly likely to be in the workforce. If these working couples joined forces, their earnings would usually suffice to bring their families above the poverty line. Yet those couples remain apart. Racial patterns also challenge the received wisdom. Marriage rates among blacks are much lower than among other ethnic groups, but the reason usually cited for this disparity — too few marriageable men — explains only a small portion of observed patterns. The crux of the problem is that black men across the socioeconomic spectrum, from the lowest skilled to the best educated, marry at far lower rates than similarly qualified men from other groups. Incarceration and early death, although contributing to disparities, have a minor effect on the shortfall. To the extent that black male unemployment and low earnings reduce the number of desirable partners, the data suggest that these are as much an effect as a cause of low marriage rates. All this evidence suggests that nonmarriage among blacks is largely a matter of choice. The key is not the externals, but that blacks are responding differently from other groups to similar constraints.

Economic factors likewise fail to account for reproductive and sexual practices of the educated elite. Almost all white women with at least a college degree marry before bearing children. Economic prospects would appear to be the first-line explanation. After all, well-educated husbands are more desirable because they earn more. But the ability to snag a prosperous husband ultimately fails to explain upper-class women’s propensity for marriage because it looks to only half the equation. Privileged women are far better equipped, economically and otherwise, to go the single-motherhood route than their less educated counterparts. Yet unlike their less well-off sisters, they still insist upon marriage before children.


In the most sophisticated economic model developed so far, George Akerlof and his colleagues attribute the recent secular decline in marriage and increase in out-of-wedlock childbearing to the technological “shock” of the birth control pill. Their theory is that, by increasing the availability of sex outside marriage, the pill both subverted men’s willingness to submit to shotgun marriages and caused more pregnancies. The result was more babies born to unmarried women. But this model fails to account for the emergence of social class differences. Although women in every class abandoned the norm of premarital chastity, privileged women continue to avoid premarital childbearing and to insist upon marriage before children. There is no evidence that differences in access or skill in using birth control explain this pattern. Rather, poor women somehow became more willing to have babies without getting married and less likely to marry.

What we know of why marriages endure suggests that male behavior may be crucial to these class divisions. Would college-educated white women long stand by men who were as habitually and notoriously unfaithful as the men depicted in these books? Even assuming that educated men are better providers and more effectively socialized to the world of work, it strains credulity that their female counterparts would put up with such antics just for the money. One possibility is that better-off men more often honor monogamy and strive for sexual fidelity. Which doesn’t mean they never cheat. But how they cheat, and how often, may make all the difference. As Jonathan Rauch has noted, discretion and hypocrisy are the hallmarks of middle-class adultery. In its effects on family stability, the occasional or hidden lapse is worlds apart from infidelity as a way of life. Likewise, “multiple partner fertility,” which is a potent relationship killer, would seem to be unusual among educated men. At least the numbers show that unmarried women of their class are unwilling to bear “love children.” And these men’s open acknowledgment of such children would appear to be relatively rare. But the very discretion and restraint that make sexual adventurism less destructive of better-off families also make it harder to investigate. Edin and Kefalas interview no comparison group, so one cannot know from their study whether higher income women would make comparable claims. There is remarkably little hard data on class differences in extra-marital sex and little evidence of attempts to document them. Most likely that is because few social scientists are eager to posit such differences or to stress their potential importance.

In spite of these uncertainties, this much is clear: Economic disparities can’t begin to account for observed patterns. This suggests that marital and sexual behavior are more a matter of values than of money. Cultural commitments, social norms, and individual choices, rather than access to material resources, best account for class and ethnic differences in family structure.


Seeing culture as primary is at odds with the position, which Edin and Kefalas so clearly favor, that government programs and economic redistribution are the way to solve social problems. As already noted, if economics is paramount and culture unimportant, family disintegration must be ascribed to outside factors such as lack of opportunity — problems that only the government can solve. But if mores are the key to family structure and some choices better promote well-being under similar constraints, it follows that moral commitments loom larger than external conditions. Material circumstances do not dictate behavior, and manipulating resources won’t automatically cure what ails fragile families. A stress on cultural norms points to individual and community reform, not government action, as the solution to family decline. Yet that stance is thought to “blame the victim.”

Another source of hostility to cultural explanations is a misguided egalitarianism that insists that the poor are no different from you and me. A fundamental tenet is that we all share the same values and aspirations. It follows that everyone would respond to similar hardships the same way. Indeed, Edin and Kefalas claim that marriage is so valorized and respected among their study population that couples hesitate to take the plunge if there is even a chance of failure. To the extent culture enters the picture at all, their subjects participate in a “trend affecting the culture as a whole.” Just like everyone else, they regard marriage as a luxury good rather than as a necessity. But the effects of that trend “look somewhat different for the middle class than for the poor.” What it all boils down to, of course, is money. If resources and opportunities were only better, the disadvantaged would have the same families as everyone else. On this view, there are no behavioral subcultures by class or race and no distinct group norms that critically shape behavior.

The evidence that social expectations and normative commitments, not economic circumstances, best account for class differences in family structure undermines this view. People are not equal in their ability to handle newfound sexual freedom. The sexual revolution, with its laxer standards of conduct in favor of self-directed sexual innovation, has hit the less privileged harder than the affluent. Foresight, restraint, capacity for self-governance, and prudence, which make for occupational and economic success, are also qualities that make for orderly families. If these attributes are more common or more valued among the economically successful, then that group’s adoption of a distinct sexual subculture in the wake of the sexual revolution would be no surprise and could fully account for observed differences in family structure.

Social scientists’ tendency to discount culture and stress material circumstances is not a matter of political ideology alone. The quest for “science” also plays a role. Seeing group norms as unimportant and external conditions as primary fits with a view of human nature that is amenable to systematic explication and with a science of human decision-making that aspires to the rigor of established fields. The rational actor model is a darling of this view. If individuals are rational decisionmakers motivated primarily by a personal cost-benefit calculus, then human choice becomes rule-like, quantifiable, and predictable.


This model, if taken to extremes, distorts the realities of social life. Nonetheless, its influence in social science is pervasive. Edin and Kefalas do not escape: The model’s conceptual commitments dominate their work. A central tenet is that everyone is equally rational: All persons can be expected to take similar steps to maximize their individual well-being. Since identical circumstances will elicit the same behavior, it follows that disparate conditions — not different behavioral choices — are the ultimate cause of divergent results.

There is little room in this paradigm for distinct cultural values as a principal source of success or failure. Nor can this worldview even acknowledge the possibility of personal or group dysfunction. On the rational actor model, everyone is doing the best he can under the circumstances. Either persons cannot do otherwise or we cannot reasonably expect them to. Ambient conditions both explain and dictate human action.

If taken to its logical limits, this paradigm has disconcerting implications. Not only is it hard to square with a robust conception of moral choice, but it also sits uneasily with normative judgments of human conduct. The desire to create an “objective,” value-free social science fuels the resistance to assessing some ways of life as superior to others or to attributing bad outcomes to poor decisions or deficient values. But the reluctance to label any behavior as self-destructive is a signal defect in any approach to social life. Some people and groups clearly accomplish more and achieve greater success under similar constraints. Some groups are better at promoting beneficial behavior and fostering a higher morality. As Thomas Sowell has argued repeatedly, any account of human life that fails to acknowledge these realities ultimately proves unsatisfactory.

Finally, the reluctance to see culture as primary may reflect social science’s methodological limitations. Culture resists precise dissection and quantification. The methods now available to analyze social life are inadequate to the task. There is no fully satisfying exposition of the relationship of culture to behavior and no comprehensive theory of personal and group dysfunction. We do not know why some individuals harm their own interests or why some groups are more successful in developing and cultivating virtue and success. David Brooks’s recent call, in a New York Times op-ed column, for heightened attention to “cultural geography” has not yet been heeded. No academic stampede in that direction is likely, if only because negative comparison between groups is now so politically incorrect.

Social science is far from achieving a full understanding of how cultural values affect social life, but it is not wholly devoid of ideas for approaching these complexities. Although work on group norms is in its infancy, concepts such as “contagion” or “tipping” are beginning to enter the lexicon and influence understanding. Game theory also holds promise. Game theoretic models of group interactions show how customs that are harmful to individuals and groups can become entrenched. As economist Robert Frank has also noted, practices can arise that benefit individuals (at least in the short term) but harm the group — that are good for one and bad for all. Game theory also sheds light on the vital function of morality: Frank and others have observed that moral rules coordinate cooperative social strategies and foster group success.

These approaches have much to teach scholars of the family. Their implications are not wholly unsympathetic to the dilemmas individuals face. Group norms, once entrenched, are tenacious. Individuals who buck the crowd pay a price or risk futility. Ostracizing felons becomes harder if most men have been to jail. Single motherhood loses its stigma if every mother lacks a husband. A faithful husband is a chump if everyone else is playing the field. The challenge comes down to a problem of collective action — of reforming failed group norms from within. That task requires first owning up to internal failure. In depicting single motherhood as the expected outgrowth of external conditions and broader cultural trends, Edin and Kefalas undermine that project. They do their subjects no service.