Commentary Magazine

What To Do About Welfare

In the 1992 campaign, Bill Clinton’s television ad promising to “end welfare as we know it” was one of his best vote-getters, so effective that it was the first choice for a heavy media buy in closely contested states at the end of the campaign. This should come as no surprise. No American social program has been so unpopular, so consistently, so long, as welfare. But why? What is wrong with welfare that evokes such a widespread urge to “do something about it”?

One obvious candidate is size and cost. Bill Clinton campaigned during a surging increase in the welfare rolls. By the end of his first year in office, more than fourteen million people would be enrolled in Aid to Families with Dependent Children (AFDC), representing more than 7 percent of American families and two million more recipients than had been on the rolls in 1989.

With so many working-aged people being supported by government, the amounts of money involved have mounted accordingly. But, as with so many other questions involving welfare, there is no uncontroversial answer as to exactly how much, because few can agree about where the definition of “welfare” begins and ends.

In 1990, before the most recent increase in the rolls had gotten well under way, figures cited by various parties in the welfare debate ranged from $21 billion to $210 billion. The lower figure, used by those who claim that welfare is really a piddling part of the budget, represents just AFDC. But no serious student of the issue denies that Medicaid, food stamps, and public housing are also part of welfare. That brings the total to $129 billion. But this number covers only part of the array of programs for low-income families. The upper-end figure of $210 billion is the bottom line for the Congressional Research Service’s report of state and federal expenditures on “cash and noncash benefits for persons with limited income” in 1990. Of that, $152 billion came from the federal government.

Two hundred and ten billion dollars works out to $6,270 for every man, woman, and child under the poverty line in 1990, only a few hundred dollars less than the official poverty threshold ($6,652 for a single unrelated individual in 1990). Statements such as “We could eliminate poverty tomorrow if we just gave the money we’re already spending directly to poor people” may be oversimplified, but they are not so far off the mark either.

One approach to the topic of “what to do about welfare” could thus reasonably involve ways to reduce expenditures. Yet, though complaints about wasting money on welfare loafers are commonly heard, and though the country truly does spend a lot of money on welfare, it is not obvious that money is really the problem. Suppose that for $210 billion we were buying peaceful neighborhoods and happy, healthy children in our low-income neighborhoods. Who would say that the nation could not afford it? Money may well become a decisive issue as the dependent population continues to grow, but it has not yet.

Instead, I will proceed from the assumption that the main source of the nationwide desire to do something about welfare is grounded in concerns about what welfare is doing to the health of the society. Judging from all that can be found in the press, on talk shows, and in the technical literature, an unusually broad consensus embracing just about everyone except the hard-core Left now accepts that something has gone drastically wrong with the family, that the breakdown is disproportionately found in poor neighborhoods, and that the welfare system is deeply implicated.

Different people put different emphases on just what has gone wrong. There are so many choices. In many welfare families, no one has ever held a regular job. This is bad for the taxpayer who supports such families, bad for the women who are trapped into poverty, and, most portentously in the long run, bad for children who need to be socialized to the world of work. In many welfare families, the mother works, but only sporadically and surreptitiously in the illegal economy. The welfare system becomes an instrument for teaching her children all the wrong lessons about how to get along in life.

In the vast majority of welfare homes, there is no biological father in the house. In many, there has never been a father. The male figure in the home is instead likely to consist of a series of boyfriends who do not act as fathers but as abusive interlopers.



These circumstances are damaging to children in so many ways that to list them individually would be to trivialize them. On this issue, the intellectual conventional wisdom has changed remarkably in just the last few years. The visible turning point was Barbara Dafoe Whitehead’s 1993 Atlantic article, “Dan Quayle Was Right,” but the groundwork had been laid in the technical journals in preceding years, as more and more scholars concluded that single parenthood was bad for children independently of poverty and other markers of socioeconomic disadvantage.

Statistically, measures of child well-being tend to order families by their structure: conditions are best for children in intact families, next best for children of divorce (it does not seem to help if the custodial parent remarries), and worst for children born out of wedlock (even if the woman later marries another man). This ordering applies to a wide variety of outcomes, from emotional development to school performance to delinquency to family formation in the next generation.

But the evidence accumulated so far tells only part of the story. Families that have been on welfare for long periods of time are overwhelmingly concentrated in communities where many other welfare families live. While it is unfortunate when a child must grow up in a family without a father, it is a disaster when a generation of children—especially male children—grows up in a neighborhood without fathers. The proof of this is before our eyes in the black inner city, where the young men reaching twenty in 1994 came of age in neighborhoods in which about half the children were born out of wedlock. Social science is only beginning to calibrate the extent and nature of the “neighborhood effects” that compound the problems associated with illegitimacy.

If these results were confined to the inner cities of our major cities, the effects on American society would still be grim enough. A look at the national mood about crime shows how a problem that is still localized (as the most severe crime rates still are, impressions notwithstanding) can nonetheless impinge on American life as a whole. But there is no reason to think that the effects will remain within the black inner city. The white illegitimacy ratio, which stood at 22 percent for all whites in 1991, is approaching the 50-percent mark in a number of working-class American cities. There is no good reason to assume that white communities with extremely high illegitimacy ratios will escape the effects of an unsocialized new generation.

These observations have led me to conclude that illegitimacy is the central social problem of our time, and that its spread threatens the underpinning of a free society. We cannot have a free society, by this reasoning, unless the great majority of young people come of age having internalized norms of self-restraint, self-reliance, and commitment to a civic order, and receive an upbringing that prepares them to transmit these same values to their children. We cannot achieve that kind of socialization without fathers playing a father’s role in the great majority of homes where children grow up.

For those who accept this pessimistic reasoning, extreme measures to change the welfare system are justified; for those who still consider illegitimacy to be one problem among many, more incremental reforms seem called for. Put broadly, four types of welfare reform are being considered in various combinations: workfare; the substitution of work for welfare; penalties for fathers; and the complete abolition of welfare.



Workfare refers to a variety of reforms that would make welfare beneficiaries show up at some sort of job, usually a public-service one, or lose their benefits. Softer versions of workfare call upon welfare recipients to attend job-training programs or risk losing their benefits. Offshoots include such things as “learnfare,” in which mothers lose part or all of their benefits if their children drop out of school.

The rationale for workfare that resonates with the voters is, roughly, “make them do something for the money we’re giving them.” Many also hope that the prospect of having to work for benefits will either deter young women from having babies in the first place or induce them to find real jobs on their own and leave welfare.

An additional intellectual rationale has been advanced by Lawrence Mead, a political scientist at New York University, who argues that what welfare recipients really lack is the ability to cope with the routines of ordinary life. Surveys show that they share the same aspirations as people in mainstream society, Mead says, but their lives are so chaotic and their discipline so ill-formed that the government must provide the framework that has been missing in their own lives.

Workfare is not an untried idea. Local attempts to force women to work for welfare have been made off and on in scattered jurisdictions for decades. The 1988 welfare-reform bill put the federal government’s imprimatur on such programs. The evaluation reports now add up to a fair-sized library, and they tell a consistent story. Participants in training and work programs usually have higher mean earnings than persons in the control groups. But these mean differences amount to hundreds of dollars per year, not thousands. The effects on long-term employment are small. The most successful programs tend to be located in small cities and rural areas rather than large cities.

A few exceptions to these generalizations are noteworthy. A program in Riverside, California, showed dramatic early results, apparently because of an energetic, decisive administrator who was given extraordinary freedom to define work rules, replace staff who did not perform, and enforce sanctions against welfare recipients who did not cooperate. If anyone can figure out how to duplicate these conditions nationwide, workfare might be able to produce much larger effects than shown in the typical evaluation.

As far as I know, no one has ever documented a deterrent effect for workfare. But evidence indicates that many welfare recipients, sometimes a significant portion of the total caseload, will drop out of a welfare program if a strictly enforced work requirement is installed.



In 1986, the social critic Mickey Kaus proposed an alternative to workfare that would scrap the existing welfare system and replace it with public-service jobs at the minimum wage. The government would provide medical care and child care for preschool children, but otherwise the woman would be on her own. If she showed up at one of the local job sites and worked, she would get a paycheck at the end of the week. If she chose not to work, that would be her business.

Kaus’s proposal, which he subsequently elaborated in his book, The End of Equality1 has much to recommend it. Workfare programs break down because of built-in contradictions. Welfare bureaucracies do not function well as employers. They have no incentives to reduce their caseloads and no incentives to make welfare recipients behave as real employees. Trying to enforce sanctions against uncooperative cases tends to become a long and tedious process. The Kaus system asks only that the government recreate a WPA-style agency for administering public-service jobs—something that the government did successfully in the 1930’s.

Whether the government could do as much again is open to question. The typical WPA male worker in the 1930’s came to the program with a set of motivations much different from those of the typical AFDC mother in the 1990’s. Yet it seems plausible to me that the Kaus system would not only achieve substantial effects on work behavior among AFDC mothers but also have a substantial deterrent effect.

The program’s cost, which Kaus himself set at $43 billion to $59 billion for national implementation, might not be as large as expected. Since we know that large proportions of the caseload have taken themselves off the rolls when a strict work requirement was imposed, we could expect a similarly large drop if the Kaus plan were implemented. And while it is difficult to imagine the federal government adopting a scrap-welfare-for-work proposal with the pristine purity necessary to make it succeed, it is possible to imagine a state doing so, if states were given the option of folding all the money currently spent on AFDC, food stamps, and public housing into a public-service jobs program.



Enforcement of child support among unmarried fathers is one of the most popular reforms under consideration, not least because it gives people a chance to say the right things about the responsibilities of the male. Like workfare, enforcement of child support is an old idea. Toughly worded laws are already on the books requiring child support, and the federal government is spending about $2 billion a year on the Child Support Enforcement program originated in 1975.

Despite these efforts, paternity is not established for about two-thirds of illegitimate births. The failure rate is so high partly because of poor enforcement, but mainly because the law asks so little of the unwed mother. The government has leverage only when she wants to qualify for AFDC benefits. For this, she is required merely to cooperate in identifying the father, a condition that can be satisfied by giving the name of a man whose whereabouts are unknown or even by her earnest statement that she does not know who the father is.

The proposed reform with the most teeth is to withhold all AFDC benefits unless the father is actually identified and located. Would such a threat help control the behavior of males? Perhaps—if the father had a job in the aboveground economy, if the state had in place methods of garnisheeing his wages, and if the state were able summarily to jail fathers who failed to meet their obligations.

Yet to list these conditions is to expose the reasons not to expect much from reforms of child support. Many unwed fathers have no visible means of support, and an even higher proportion will flee into that category, or disappear entirely, if child-support enforcement is tightened.

Would such measures nonetheless “send the right signals” about the responsibilities of men for their children? Many think so; I am a holdout. The alternative “right signal” is to tell young women from the outset—from childhood—that they had better choose the father of their babies very carefully, because it is next to impossible for anyone, including the state, to force a man to take on the responsibilities of fatherhood.



This brings us to the fourth option, scrapping welfare altogether, a proposal with which I have been associated for some years. I am under no illusions that Congress is about to pass such a plan nationally. But, as with the Kaus plan, a state can do what the federal government cannot. And it is conceivable that Congress will pass reforms permitting the states wide discretion in restructuring the way they spend their welfare budgets.

The main reason for scrapping welfare is to reduce the number of babies born to single women. The secondary reason is to maximize the chance that children born to single women are raised by mature adults who are able and willing to provide a loving, stable, nurturing environment—a result that will ensue because more children will be given up for adoption at birth, and because single mothers who choose to keep their babies in a no-welfare society will be self-selected and thus their number will be limited to those who have the most resources for caring for children.

These goals presume that ending welfare will have a drastic effect on behavior. One must ask whether there is good reason to believe that it will.

One way of approaching the question is to ask whether welfare causes illegitimacy in the first place. I have written two reviews of this debate in the past two years—one long and technical, the other shorter and nontechnical2—and will not try to cover all of the ground here. These are the highlights plus a few new points:

Academics have focused almost exclusively on comparisons of illegitimacy based on the differences in welfare payments across states. It is now generally if reluctantly acknowledged by these scholars that the generosity of welfare benefits has a relationship to extramarital fertility among whites. More recent work is showing that a relationship exists among blacks as well. The size of the effect for whites seems to be in the region of a 5-percent change in extramarital fertility for a 10-percent change in benefits, with some of the estimates substantially larger than that.

This effect is called small by those unhappy to admit that welfare has any relationship at all to extramarital fertility. I treat the fact that any effect has been found as I would treat favorable testimony from a hostile witness—the analyses have generally consisted of regression equations with a multitude of independent variables, making it as hard as possible to show an independent effect for AFDC.

A broader observation about these studies is that trying to analyze the relationship of welfare to illegitimacy by examining cross-state variation in AFDC benefits has a number of serious methodological problems that are bound to limit the magnitude of the effect that AFDC is permitted to show. I have been pointing to such problems in print for many years. So far as I know, none of the analyses using cross-state benefits has even acknowledged the existence of these technical problems, much less tried to deal with them.

Last summer, 76 social scientists signed a statement saying that the relationship of welfare to illegitimacy was small. When I replied that the very studies they had in mind were consistent with something in the neighborhood of a 50-percent drop in white illegitimacy if welfare were eliminated, there were cries of outrage—but not because my statement was technically inaccurate. It was a straightforward extrapolation of the 5-percent (or more) change in white fertility per 10-percent change in welfare benefits that has been found in recent research.

I should add that I do not place much faith in such linear extrapolations in this case. Indeed, I argue from other evidence that the effects would most likely steepen as the reductions in welfare approached 100 percent. But this is speculative—no one has any empirical way to estimate how the curve might be shaped.



Meanwhile, two characteristics of illegitimate births imply a stronger relationship to welfare than that indicated by the cross-state analyses.

The first of these characteristics is that the illegitimate birth rate has been increasing while the legitimate birth rate has been decreasing. The rate in this case refers to the production of babies per unit of population, in contrast to the more commonly used statistic, the illegitimacy ratio, representing the proportion of live births that are extramarital.

The logic goes like this: birth rates are driven by broad historic forces that are so powerful and so consistent that they have applied everywhere in the West. Put simply, birth rates fall wherever women have an option to do something besides have babies. The options are brought about by better medical care (so more babies survive to adulthood), increased wealth and educational opportunities, and the opening of careers to women. Improved technology for birth control and access to abortion facilitate the effects of these forces.

Thanks to all this, among both blacks and whites in America, the number of legitimate babies per unit of population has been falling steeply. But during this same period, concentrated in the post-1960’s, the number of illegitimate babies per unit of population has been rising. In other words, something is increasing the production of one kind of baby (that born to single women) at the same time that the production of the other kind of baby (that born to married women) is dropping.

The scholars who say that welfare cannot be an important cause of the breakdown of marriage and the encouragement of illegitimacy have yet to offer an explanation of what this mysterious something might be. The existence of a welfare system that pays single women to have babies meets the test of parsimony.

Perhaps, however, the “mysterious something” is the lack of these new options for disadvantaged women. But why specify single disadvantaged women? That brings us to one of the most provocative features of illegitimacy, its relationship to poverty—not poverty after the baby is born, but before. It is one of the stronger reasons for believing that the welfare system is implicated in the production of illegitimate babies.

Begin with young single women from affluent families or women in high-paying jobs. For them, the welfare system is obviously irrelevant. They are restrained from having babies out of wedlock by moral considerations, by fear of the social penalties (both of which still exist, though weakened, in middle-class circles), by a concern that the child have a father around the house, and because having a baby would interfere with their plans for the future.

In most of the poorest communities, having a baby out of wedlock is no longer subject to social stigma, nor do moral considerations still appear to carry much weight. But the welfare system is very much part of the picture. For a poor young woman, the welfare system is highly relevant to her future if she has a child, easing the short-term economic penalties that might ordinarily restrain her childbearing. The poorer she is, the more attractive the welfare package, and the more likely that she will think herself enabled by it to have a baby.

The implication of this logic is that illegitimate births will be concentrated among poor young women—and they are. This may be inferred from the information about family income from the Bureau of the Census data, showing that in 1992, women with incomes of less than $20,000 contributed 73 percent of all illegitimate babies, while women with incomes above $75,000 contributed just 2 percent.

But these data are imprecise, because income may have fallen after the baby was born (and the woman had to quit work, for example). The logic linking welfare to illegitimacy specifically refers to women who are poor before the baby is born. For data on this point, I turn to one of the best available bases, the National Longitudinal Survey of Youth (NLSY),3 and ask: of women of all races who were below the poverty line in the year prior to giving birth, how many of their children were born out of wedlock? The answer is 56 percent. Among women who were anywhere above the poverty line, only 11 percent of babies were born out of wedlock.



Why should illegitimate births be so much more likely to occur among women who are already poor? The common argument that young women with few prospects “want something to love” may be true, but it has no answer to the obvious rejoinder, that single poor young women in the years before the welfare system began probably wanted something to love as well, and yet the vast majority of them nonetheless made sure they were married before bearing a child. Other things being equal, poor single young women face the most daunting prospects if they have a baby without a man to help take care of it, and that reality used to govern the behavior of such young women. Of course the sexual revolution has changed the behavior of young women at all levels of society, but why has it produced babies predominantly in just one economic class?

Once again, an answer based on a welfare system that offers incentives only to poor women meets the test of parsimony. Once again, the scholarly literature has yet to offer an alternative explanation, or even to acknowledge that an alternative explanation is called for.

There is one additional characteristic of women who are at most risk of giving birth to children out of wedlock: they generally have low intelligence. This point is new to the welfare debate. Richard Herrnstein and I discuss it at length in The Bell Curve, again using the National Longitudinal Survey of Youth, which administered a high-quality cognitive test to its subjects when the study began. The chances that a poor young woman’s baby would be born out of wedlock were 68 percent if she had an IQ of 85, but only 26 percent if she had an IQ of 115.

Lest it be thought that this result is conflated with racial complications, it should be noted that the relationship held among whites as powerfully as among the population as a whole. Lest it be thought that the result is conflated with the opportunity that smart women have to go to college, it should also be noted that the relationship holds as powerfully among women who never got beyond high school as it does for the population as a whole. Lest it be thought that this is a reflection of socioeconomic background, the independent importance of IQ is still great after holding socioeconomic status constant. Conversely, the independent importance of socioeconomic background after holding the effects of IQ constant is severely attenuated.

Summarizing the overall picture: women in the NLSY (in their mid-twenties to early thirties when this observation applies) who remained childless or had babies within marriage had a mean IQ of 102. Those who had an illegitimate baby but never went on welfare had a mean IQ of 93. Those who went on welfare but did not become chronically dependent on welfare had a mean IQ of 89. Those who became chronic welfare recipients had a mean IQ of 85.



Now back to the first and most crucial goal of welfare reform, that it drastically reduce the number of children conceived by unmarried women. In trying to develop methods for accomplishing this goal, we know from the outset that both sex and the cuddliness of babies are going to continue to exert their powerful attractions. We know that decisions about whether to have sex and whether to use birth control are not usually made in moments of calm reflection.

Therefore, any reform must somehow generate a situation in which a young woman, despite not being calm and reflective, and often despite not being very bright, is so scared at the prospect of getting pregnant that she will not have intercourse, or will take care not to get pregnant if she does.

This means that the welfare reform will have accomplished one of two things. Either the change has been so big, so immediate, and so punishing that even a young, poor, and not very smart girl has been affected by it; or else the change has directly motivated people around that young woman to take an active role in urging her not to have the baby.

Bill Clinton’s program, based on the threat of “two years and out, if you’ve had a reasonable chance at job training and a reasonable chance to find a job,” is not calculated to meet this criterion. Two years is an eternity to a young girl. The neighborhood is filled with single women who have been on welfare for ages and have not gotten thrown off. Is a sixteen-year-old going to believe that she will really be cut off welfare two years down the road, or will she believe the daily evidence around her?

Other commonly urged recommendations—sex education, counseling, and the like—are going to be just as futile. A major change in the behavior of young women and the adults in their lives will occur only when the prospect of having a child out of wedlock is once again so immediately, tangibly punishing that it overrides everything else—the importuning of the male, the desire for sex, the thoughtlessness of the moment, the anticipated cuddliness of the baby. Such a change will take place only when young people have had it drummed into their heads from their earliest memories that having a baby without a husband entails awful consequences. Subtle moral reasoning is not the response that works. “My father would kill me” is the kind of response that works.

From time immemorial, fathers and mothers raised the vast majority of their daughters, bright ones and dull ones, to understand these lessons. Somehow, in the last half-century, they began to lose their capacity to do so—curiously, just as social-welfare benefits for single women expanded. I want to press the argument that the overriding threat, short-term and tangible, which once sustained low illegitimacy ratios was the economic burden that the single woman presented to her parents and to the community. I do not mean to deny the many ways in which noneconomic social stigma played a role or to minimize the importance of religious belief, but I would argue that much of their force was underwritten by economics.

At this point, we reach a question that cannot be answered by more social-science research but only by experience: if welfare were to be abolished in the late 20th century, would a revival of the economic threat be enough to drive down illegitimacy? Or do we need a contemporaneous revival of the moral sanctions against illegitimacy to make the economic penalties work? The good news is that the two forces can be counted on to work together, because of a built-in safety mechanism of American democracy. Welfare will not be abolished until the moral sanctions against illegitimacy have also gained great strength. There will not be enough votes until that mood is broad and deep.

It is only because of the sea change in the conventional wisdom about the deficiencies of single-parent families that proposals to end welfare are now being taken seriously. So far, that change has been couched in utilitarian terms. The next step, already well under way, is for language to change. Now, the elites are willing to say, “Having a baby if you are young and single is ill-advised.” It seems to me that the truer way to put the issue is this: bringing a new life into the world is one of the most profoundly important moral acts of a person’s life. To bring a child into the world knowing that you are not intellectually, emotionally, or materially ready to care for that child is wrong.

When the elites are broadly willing to accept that formulation, and not before, welfare will be ended. And at that stage, we can also be confident that the financial penalties of single parenthood that ending welfare would reimpose are going to be reinforced by moral suasion.

Different parts of the country will reach this state of affairs sooner than others. In Utah, for example, with its low illegitimacy ratio plus the moral force of the Church of Latter Day Saints and that church’s elaborate system of social welfare, one may be confident that if the entire federal welfare system disappeared tomorrow, the result would be overwhelmingly positive, with only the most minor new problems. But if the same legislation were to apply to Harlem, where more than 80 percent of children have been born out of wedlock for a decade and nongovernmental social-welfare institutions are scattered and in disarray, one may be equally confident that the short-term result would be chaos on a massive scale.



Drawing these strands together, here are the characteristics of legislation that might have a chance of passing in the next several years:

The centerpiece of the legislation should be freedom for the states to experiment. Congress knows beyond doubt that the welfare system it currently mandates for the entire country is a failure. The next thing for Congress to learn is that it does not have a one-size-fits-all answer to amend that failure. The solution is to permit the states to adopt a wide variety of plans.

Thus, Congress should develop a simple formula whereby states can take the money that would otherwise flow into them in the form of AFDC payments, food stamps, and housing benefits (and as many other means-tested programs as possible) and use it for other ways of dealing with the needs of children currently supported by the welfare system.

One example of a simple formula is to base the amount of the allocation on the budgets for those programs in the last year before the federal legislation is passed. States should also be permitted to end those programs altogether and forgo federal funds completely, though it is doubtful whether any state would choose to go that route.

Initially, most states would probably opt for modest reforms along the lines Congress is contemplating—more workfare, more job training, perhaps soft time limits. But a few brave states are likely to try something more ambitious. Probably one or two will adopt much more aggressive workfare or time limits than the ones in the Clinton plan. Perhaps a state somewhere will choose to adopt a version of the Kaus plan, funding public-service jobs in lieu of welfare benefits. My hope is that some state will also end welfare. If a state should consider doing so, here are some guidelines that I would recommend:

Grandfather everyone now on the system, letting them retain their existing package of benefits under the existing rules. The reasons for grandfathering are both ethical and pragmatic. For many women, welfare has turned out to be a Faustian bargain in which the government plays the role of the devil. Having made this bargain, many of the women on welfare are so mired in the habits of dependency and so bereft of job skills that it is unethical for the government now to demand that they pull themselves together. Pragmatically, grandfathering is probably a prerequisite for getting any such plan through a state legislature.

I should add that some grace period is also necessary between the passage of the legislation and the time it takes effect. Nine months and one day is the symbolically correct period. Practically, a year seems about right: long enough to allow the word to spread, abrupt enough to preserve the shock value that is an essential part of changing behavior.

Limit the reform to unmarried women. This step is primarily to facilitate building a political coalition that can get the legislation passed, but it also can be taken without jeopardizing the desired result. Divorced and abandoned women are not at the heart of the welfare problem. On the contrary, most of them treat welfare as it was originally intended: as a temporary bridge. When you read statistics such as “half of all women get off welfare within two years,” it is divorced women who have brought down the average.

It may be objected that to limit the reform to unmarried women provides an incentive for pregnant girls to enter into a marriage of convenience. This may well be true, but it is a good result, not a bad one. Men who sign a marriage certificate are much more easily held to account for support of the child than men who do not.

Is not limiting the reform to unmarried women discriminatory? Yes, that is one of the main points of doing anything about welfare. I am not enthusiastic about using government policy positively to reward marriage, but it is another thing to end government policy that undermines marriage—as welfare for single women undeniably does.

“Ending welfare” should mean at a minimum cutting off all payments which are contingent on or augmented by having a baby. The core benefit to be ended altogether is AFDC. Medicaid benefits for the child should be left in place, because the existence of Medicaid has gutted the alternative ways in which medical care could be made reliably available to poor children (whereas there remain many alternative ways of providing children with food, shelter, and nurturing).

What about housing and food stamps? I doubt if it is possible to end them altogether. If a woman is poor enough to qualify for housing benefits and food stamps without a child, it seems unlikely that the courts would allow those benefits to be cut off because a child has been born.

Instead, a state that adopts the “end-welfare” option should simply become neutral with regard to births out of wedlock. In principle, the best way for the state to become neutral is the approach advocated by Milton Friedman: dismantle the entire social-welfare structure with its multiplicity of benefits and bureaucracies, replace it with a cash floor using the mechanism of the negative-income tax, and make that cash floor invariant regardless of the number of children. But I cannot imagine Congress giving states the option of converting all federal-subsidy programs into a negative-income tax (though it is certainly an intriguing idea). Some steps short of that need to be worked out.

One attractive possibility is to return to the original intention of the 1935 act that created welfare. AFDC would continue to be available for widows with young children and for divorced or abandoned women with young children, with a higher cash payment to compensate for the cuts in housing and food benefits. Unemployment benefits would also remain available for men and women alike, with or without children. I favor broadening and strengthening the unemployment-insurance system as part of this approach.

Limit the initial legislation to teenagers. It is widely assumed that if welfare is ended, some other mechanisms will be required to replace it. Most of these options (I will describe some presently) involve extensive interventions in loco parentis. Limiting the initial legislation to teenagers has two merits. First, it is much better to let the government act in loco parentis for minors than for adults. Second, a political consensus already exists about single teenage girls having babies that has not yet consolidated about single adult women having babies.



If a state ends welfare in the ways I have just described, a large behavioral impact may be expected—somewhere in the region of a 50-percent reduction in illegitimate births among whites (and probably among blacks as well) if the cross-state analyses are taken seriously.

Other effects are hard to predict. Some people assume that large numbers of pregnant women will move across the border to the next state, others predict a surge in abortions. The type and size of the effects will also depend on the nature of a state’s caseload. The effects in a mostly rural plains or mountain state are likely to play out much differently than the effects in states with large cities. In any case, a substantial number of single women will continue to get pregnant. What happens to them and their children? These measures should be considered:

Actively support adoption at birth. Today, the welfare system and its satellite social-work agencies typically discourage adoption. The pregnant single woman who wants to give up her child for adoption is more likely to be encouraged to keep the baby than to be praised. This is perverse. In America, the pool of mature, caring adoptive parents is deep, not just for perfect white babies but for children of all races and for children with physical and mental handicaps, if—the proviso is crucial—the child can be adopted at birth. Any comparison of what is known about child abuse and neglect, emotional development, or educational success suggests that the child of a never-married teenager has a better chance in an adoptive home.

If welfare has been ended, many more pregnant women will be looking at adoption, and the state can do much to help. Changes in laws can encourage a larger pool of adoptive parents by reinforcing the rights of the adoptive parents and by strictly limiting the rights of the biological parents. Adoption agencies can facilitate the adoption of black children by ending restrictions on transracial adoption.

Offer group living for pregnant women. For a pregnant young woman from a functioning family and a functioning community, the best support network consists of friends and relatives. One of the chief reasons for ending welfare is to revitalize those networks. But one of the saddest aspects of today’s burgeoning illegitimacy is that many pregnant young women have no friends and relatives who are competent to provide advice and nurturing during the pregnancy, let alone to help think through what will happen after the baby is born. This will continue to be true when welfare is ended.

States that end welfare should therefore look carefully at the experience of the homes for pregnant single women that dotted the country earlier in the century, most notably the Florence Crittendon homes. In a modern version of such a home, the young woman would receive the kind of prenatal care and diet—meaning, among other things, no drugs, alcohol, and tobacco—that would help children of unwed mothers get off to a better physiological start. Group homes of this sort can also be excellent places to help young women come to grips with their problems and prepare for their futures.



Offer group living for teenage single mothers. Another intriguing suggestion is to extend the Florence Crittendon concept to the period after birth. The mother who keeps her baby is no longer given welfare services, but she is given the option to live in a group home. She and her child receive food and shelter; the mother receives training in parenting and job skills; and the child is in an environment where at least some of the adults understand the needs of infants and small children.

Maintain a clear bright line short of coercion. Adoption services or group homes must be purely optional; no young woman should be required to use these services. This bears on a broader point. Having a baby you are not prepared to care for is wrong, but this does not mean that the state has the right to prevent you from doing it—a nice distinction between immoral acts and the state’s power to regulate them that could easily be ignored once the Left decides that illegitimacy is a bad thing.

An idea gaining favor—requiring welfare recipients to use Norplant—illustrates the danger. From a legal standpoint, I find nothing objectionable about the idea. Welfare is not a right but largesse, and the state may legitimately place conditions upon dispensing largesse. But once the government requires any use of birth control, a barrier has been broken that has frightening possibilities.

For the same reason, the government must be passive regarding the encouragement of abortion. If enough people think that low-income women should have easier access to abortions, let the subsidies come from the philanthropies that private citizens choose to support. The process of ending welfare must unambiguously represent a withdrawal of the state from personal decisions, not new intrusions.

Enforce the existing laws on child neglect. One of the most common questions about ending welfare is, “What happens to the woman who keeps her baby anyway?” The answer is that some women will indeed choose to keep their babies. As I have already suggested, the self-selection process imposed by the end of welfare also means that such women are likely to be those who have the greatest commitment to their children. They are likely to be the ones who have done the best job of lining up support from relatives and friends, or the ones who have well-paying jobs.

But the main point is that single women who keep their babies will be in exactly the same situation as every other parent who takes a baby home from the hospital: that child is now the parent’s responsibility. There is no need to keep a special watch on how a single mother does; rather, she falls under the same laws regarding child neglect and abuse as everyone else, to be enforced in the same way.



And that, finally, should be the overriding theme of what we do about welfare: treating the human drama of “having a child” as the deeply solemn, responsibility-laden act that it is, and treating all parents the same in their obligation to be good parents. The government does not have the right to prescribe how people shall live or to prevent women from having babies. It should not have the right even to encourage certain women to have babies through the granting of favors. But for 60 years the government has been granting those favors, and thereby intervening in a process that human communities know how to regulate much better than governments do. Welfare for single mothers has been destructive beyond measure, and should stop forthwith.

The present article is the fifth in a series that began with James Q. Wilson on crime (September), followed by Gertrude Himmelfarb on the universities and Chester E. Finn, Jr. on the schools (October), and Eliot A. Cohen on national defense (November).


1 See the review by Michael Horowitz in the December 1992 COMMENTARY.—Ed.

2 “Welfare and the Family: The American Experience,” Journal of Labor Economics (January 1993), and “Does Welfare Bring More Babies?,” Public Interest (Spring 1994).

3 The NLSY is a very large (originally 12,686 persons), nationally representative sample of American youths who were aged 14 to 22 in 1979, when the study began, and have been followed ever since.

About the Author

Charles Murray is the W.H. Brady scholar at the American Enterprise Institute and the author most recently of In Our Hands: A Plan to Replace the Welfare State (2006). This article has been adapted from a presentation at the annual Herzliya Conference in Israel in January.

Pin It on Pinterest

Welcome to Commentary Magazine.
We hope you enjoy your visit.
As a visitor to our site, you are allowed 8 free articles this month.
This is your first of 8 free articles.

If you are already a digital subscriber, log in here »

Print subscriber? For free access to the website and iPad, register here »

To subscribe, click here to see our subscription offers »

Please note this is an advertisement skip this ad
Clearly, you have a passion for ideas.
Subscribe today for unlimited digital access to the publication that shapes the minds of the people who shape our world.
Get for just
Welcome to Commentary Magazine.
We hope you enjoy your visit.
As a visitor, you are allowed 8 free articles.
This is your first article.
You have read of 8 free articles this month.
for full access to
Digital subscriber?
Print subscriber? Get free access »
Call to subscribe: 1-800-829-6270
You can also subscribe
on your computer at
Don't have a log in?
Enter you email address and password below. A confirmation email will be sent to the email address that you provide.