Is illegitimacy a public-health hazard?

The need to protect our nation's children offers the most plausible case for interventionist social policy. And yet one form of behavior encouraged by our welfare policies

is among the leasing causes of infant death.

IN RECENT YEARS, the well-being of America's children has become a topic of intense public discussion. Many authoritative commentators have concluded that childhood has become a time of significant, and increasing, risk in America. Considerable evidence, in fact, has been adduced to make the case that the material condition of America's children has seriously worsened over the past decade and a half.

The outlines of the problem, as typically presented, are by now familiar. By the reckoning of the U.S. Census Bureau, real per-capita income in the United States rose by more than 25 per cent between 1973 and 1986; yet over those same years the poverty rate for Americans under 18 years of age increased by nearly two-fifths: from just over 14 per cent to almost 20 per cent. While America had four million fewer children in 1986 than in 1973, almost three million more children were estimated to be living in poverty. Between 1973 and 1986, both the childpoverty rate and the absolute number of children in poverty rose in all the ethnic categories the Census Bureau delineated ("White," "Black," "Spanish origin"); relatively and absolutely, the greatest increase in measured poverty was among "White" children.

Placed in broader perspective, these figures and trends seem even more disheartening. Although the United States continues to enjoy virtually the highest per-capita income in the world, a study just published in Science magazine (and co-authored by the chief of the Census Bureau's Center for International Research) concludes that the rate of child poverty is higher in the United States than in Australia, Canada, Sweden, the United Kingdom, or West Germany-higher, indeed, than in any of the other industrialized countries for which calculations were presented. The assessment by Senator Daniel Patrick Moynihan, long one of America's foremost experts on social-welfare policy, is more distressing still. In the printed version of his 1985 Godkin lectures at Harvard, Moynihan writes: "It is fair to assume that the United States has become the first society in history in which a person is more likely to be poor if young rather than old."

IT IS HARDLY SURPRISING that these data and judgments should have attracted widespread attention and evoked deep concern. More interesting, perhaps, have been the particulars of the responses. For some, the issue of child welfare seems to have sounded reveille for a new era of social-policy initiatives. And indeed, it is in the circumstances of children that the strongest possible case for statist welfare policy is to be made. Innocent, vulnerable, imperiled through no fault or action of their own, children in distress might seem an almost crystalline representation of the idea of the "deserving poor"-those persons who can (and indeed should) be helped through public assistance without fear that they will succumb to the "moral hazards" that necessarily accompany such charity.

Consider, for example, the New York Times's four-part pre-election editorial on child poverty, subtitled "An Issue for the No-Issue Campaign." The editorial warned that "the piecemeal proposals" the presidential "candidates have so far offered" would not be enough to "save the next generation of children at risk." "Tyrannized by deficits," it chided, "[Bush and Dukakis] talk only of token spending." The Times made more financially substantive recommendations: new health-care programs for children; national daycare services; additional support for young mothers (and fathers); "early childhood intervention" policies; and much more. Of these many proposals, there was scarcely one that did not require either the creation of new federal social programs, or additional funds for existing ones.

For all the importance of this issue, the Times editorial series, like so much of the public discussion of the wellbeing of America's children, seems to be proceeding under the influence of some consequential misperceptions, which may in large measure involve a simple misreading of data. Specifically, the danger posed to America's children by the explosive spread of illegitimacy through the fabric of American society seems to be widely ignored.

In part, this is because a critical piece of data has gone uncollected by our otherwise excellent national statistical system. Were such data to be regularly assembled, the numbers would show that out-of-wedlock birth is strongly associated with infant death-in fact, that it might rank as one of the leading syndromes associated with infant mortality in modern America. Such data would place contemporary discussions about the therapeutic potentialities of "early childhood intervention" through government social programs in a very different light.

The "poverty rate" is the statistic perhaps most commonly invoked in today's discussion about the condition of America's children. This is not the place for a detailed analysis of the technical and conceptual problems with that Census Bureau measure. Suffice it to say that the "poverty rate" is an ambiguous, and apparently increasingly unreliable, indicator of the actual well-being of American children. The poverty rate for children, for example, no longer tracks with the country's infant-mortality rate. (See table below.) Between 1973 and 1983, the incidence of child poverty in America is calculated to have risen by more than half. Over those same years, however, the infant-mortality rate fell by almost two-fifths. For "White," "Nonwhite," and "Black" children alike, the child-poverty rate and the infant-mortality rate moved in opposite directions over the course of the decade.

If the "poverty rate" is a poor predictor of the health of American infants over time, it seems not much better at indicating the survival chances of babies at any point in time. Data from the 1980 Census permit comparison of child-poverty rates for a variety of ethnic groups in America with the National Center for Health Statistics' breakdown on infant mortality. The result appears to be something less than strong correlation. For the group described as "Chinese," for example, child poverty was measured to be nearly a third higher than for the "White" group, yet its infant-mortality rate was less than half as high. Those classified as "Other Asian or Pacific Islanders" had a measured child-poverty rate about a fifth higher than the national average, yet their infant-mortality rate was almost 40 per cent lower than that for the nation as a whole.

DIFFERENCES in infant mortality among America's ethnic groups appear to be explained much more effectively by a different (though not entirely unrelated) factor: their illegitimacy ratios. There was a striking correspondence in 1980 between the proportion of births reported to unmarried mothers and babies' survival chances in the first year of life. "Chinese" babies may have been more likely than "White" babies to be born into families measured as poor, but they were far less likely to be born illegitimate.

The correlation between illegitimacy and infant mortality-arguably the most important indicator of the material well-being of the very young-should make us reflect upon some of the changes in the condition of childhood in America over the past generation.

Other things being equal, one would have expected the past generation to have witnessed a progressive improvement in the material circumstances of American children. Three factors would suggest such a judgment. First, the United States has enjoyed a substantial increase in both per-capita income and disposable percapita income since 1960. Second, the decline in fertility over the past generation has meant that a smaller fraction of children are born to fathers fifty or older-ages at which a decline or even total interruption of earnings becomes more likely. Third, the general increase in the health of the American population has dramatically reduced the chances that one or both parents will die during their offspring's childhood. In 1986, the percentage of children under 18 whose fathers had died was less than half of what it had been in 1930; the risk of being a full orphan was less than a twentieth of what it had been 56 years earlier.

Improvements in survival chances for parents, however, have not translated into improved survival chances for traditional family units. In 1960, more than 90 per cent of families with children were headed by a married couple; by 1986, that figure had dropped to 74 per cent. Whereas the number of families in America with one or more children under 18 increased by a quarter between 1960 and 1986, the number of femaleheaded families with one or more children under 18 tripled. In 1960, moreover, two-fifths of female heads of household with children were widows; by 1986, widows accounted for less than 7 per cent of such female heads of household. Divorcees, separated women, and unwed mothers accounted for the rest.

The rise of illegitimacy has been particularly rapid. In 1986, the number of births identified as being to unmarried mothers was nearly four times greater than in 1960, even though half a million more babies were born in 1960 than in 1986. By 1986, more than 23 per cent of all American births were identified as illegitimate-nearly four and a half times the fraction in 1960. With the spread of illegitimacy, the nature of the phenomenon has also changed. Illegitimate children, for example, are no longer predominantly black. To be sure, the fraction of children born illegitimate is much higher among American blacks than among American whites. But the rise in the number of illegitimate births has been much faster for whites than for nonwhites since 1960. In 1981, for the first time on record, more illegitimate children were born to whites than to blacks. By 1983, whites accounted for an absolute majority of the illegitimate births in America.

Illegitimacy, moreover, is no longer primarily associated with teenaged mothers. As recently as 1973, more than half of the illegitimate children in America were born to teenagers; by 1986, the fraction had dropped to less than a third. In 1986, in fact, almost as many illegitimate children were born to women 25 or older as to girls 19 and under. Illegitimacy has been increasing most rapidly among women in their twenties and thirties-women too old for us to ascribe such behavior to mistakes of youth or inexperience.

The rise in illegitimate births over the past generation has broadly coincided with an increase in the incidence of "welfare" recipiency. By 1984, more than a sixthof America's families were receiving one or more means-tested government benefits. In 1986, about one American child in nine was a recipient of Aid to Families with Dependent Children (AFDC). The AFDC program has been transformed by the spread of illegitimacy among its ranks, and diverted far from its original purposes. (See table this page.) In 1937 and 1938, shortly after the inception of the program, nearly half of the children on AFDC were paternal orphans; for nearly three-quarters of AFDC children, the father was either dead or incapacitated. By 1982, less than 5 per cent of all children on AFDC established their eligibility by such claims. By 1986, just under half of all AFDC children were identified as illegitimate (compare that fraction with the 3 per cent figure from the program's early years). In 1986, over 3.6 million illegitimate children were on AFDC; this would be more than three-fifths of all children living in households headed by a ncver-married mother. It seems fairly clear, then, that the AFDC program, as it currently operates, is to a great degree a vehicle for financing illegitimacy. One may debate causes and effects in this arrangement, and argue whether things would be worse for children if this program did not exist-but one cannot contest this fact.

IF ILLEGITIMATE CHILDREN could expect to be as healthy as children born to married couples, the correspondence between public assistance and illegitimacy might be seen merely as a matter of financial or moral interest. But there is considerable reason to believe that illegitimate children cannot expect to be as healthy as children born into wedlock. In 1981, according to the National Center for Health Statistics, illegitimate black babies were more than 40 per cent more likely to have "low birth weight" (roughly, less than five and a half pounds) than legitimate black babies; for whites, the risk of low birth weight was more than 50 per cent higher for illegitimate babies than for legitimate ones. The difference is consequential. A major study of 1980 data by the Centers for Disease Control estimated that the infant-mortality rate for low-birth-weight babies was about twenty times higher than for children whose birth weight was five and a half pounds or more. Other figures from the National Center for Health Statistics reinforce the impression that illegitimacy is a risk to the health of infants.

For both black and white American babies, the risk of low birth weight is higher the later pre-natal medical care begins. (Pre-natal care itself does not wholly account for this difference; the decision to seek early pre-natal care is indicative of other parental attitudes and actions.) Black babies who receive no pre-natal care are two and a half times as likely to be born low-birth-weight as those whose pre-natal care begins in the first or second month of pregnancy; for white babies, the risk increases by a factor of almost three. According to the latest data available (1985), unmarried black mothers were two and a half times as likely as married black mothers to go into delivery without any pre-natal care; white babies were over five times as likely to have received no pre-natal care if the mother was unmarried.

The easiest way to assess the health threat posed to the newborn by illegitimacy, of course, would be to compare infant-mortality rates for legitimate and illegitimate children in America. Unfortunately, this cannot easily be done. The United States collects wide-ranging and reliable statistics on all manner of demographic phenomena, but infant mortality by mother's marital status is one that we have, until now, neglected to compile regularly. Other Western countries, however, do make it a habit to collect such data. (See table above.) Perinatal mortality is considerably higher for illegitimate children in all of them.

There is, at this writing, only one detailed study of nationwide infant-mortality rates for legimate and illegitimate children in America; it is for the year 1960. For that year, infant mortality was more than a third higher for illegitimate "nonwhite" babies than for legitimate ones, and more than 50 per cent higher among "whites." Differences in the age of mothers did not account for these discrepancies: for "whites" and "nonwhites" alike, infant mortality was consistently higher for illegitimate babies regardless of the mother's age.

The National Center for Health Statistics is currently in the process of trying to link up the country's birth and death records for children under the age of one. Preliminary, unpublished data for 1983 are now available. These show that infant-mortality rates were 35 per cent higher for illegitimate black babies than for legitimate ones, and more than 60 per cent higher for illegitimate white babies than for legitimate ones.

On the basis of some very rough calculations (death rates for babies by age of mother, for example, are not available in these preliminary data) it appears that illegitimacy was associated in 1983 with about three thousand to four thousand "surplus" infant deaths. That is to say: if infant mortality by race had been as low for illegitimate children as for legitimate ones, about 3,800 fewer babies would have died in 1983. Alternatively, 3,500 fewer babies would have died in America in 1983 if the proportions of illegitimate and legitimate children by race had stayed the same as they were in 1960.

By such computations, illegitimacy would emerge as one of the major killers of American children. Illegitimacy, of course, is a social condition associated with early death, rather than a medical cause of death in and of itself. Nevertheless, a comparison of this rough estimate with actual cause-of-death figures for American babies in 1983 is instructive. In that year, for example, just over one thousand infants died from accidents and homicide (including child abuse). About eight hundred died from pneumonia and influenza. About nine hundred died from infections specific to the perinatal period. Illegitimacy would, by such calculations, be associated with more infant deaths in America than all these causes combined. Indeed, if it were a medical condition rather than a social syndrome, illegitimacy would have been ranked the third or fourth leading cause of death for infants in America.

Health may be the best synecdoche for overall social well-being; for children, low infant-mortality rates may be the best synecdoche for health, By that reckoning, if we are truly concerned with the well-being of our children, we can only view the continuing spread of illegitimacy with the deepest dismay.

The data and computations given above cast a troubling light on the social programs that have been constructed to address the needs of the youngest and the most vulnerable among the deserving poor. On the one hand, it is apparent that social-welfare programs may provide children with food, medical services, and other benefits that they might not otherwise be able to obtain. On the other hand, it now seems to be the case that these same programs subsidize or finance practices on the part of parents that are distinctly perilous to children. (From this perspective, it is immaterial whether such social programs have themselves created pernicious behavior or merely accommodated it.) It is to be hoped that recognition of the connections between social-welfare programs, illegitimacy, and child health will lead to public policies that are more genuinely protective of the well-being of our children. One might also wish that, after a generation and more of trying to avoid the issue, the realization that illegitimacy is a direct hazard to the well-being of the young will help direct public discussion of the child-welfare problem back where it should have been all along: toward the attitudes, practices, and responsibilities of our nation's parents.