Prison-Bound Boys

Which eight-year-old boys will be in serious trouble by the age of eighteen?  According to a study recently published in the Journal of the American Academy of Child and Adolescent Psychiatry, boys from broken homes are particularly prone to punctuate their late adolescence with illegal acts. 

Conducted by researchers affiliated with half a dozen American and Finnish universities, the authors of the study carefully examine data for 2,713 Finnish boys born in 1981.  Those data include information on the boys’ family backgrounds when they were eight years old and a collation of reports on the boys’ “problem behavior” at that age, collected from parents, teachers, and the boys themselves.  Also included in the data are all reports in the national police register of crimes later committed between 1998 and 2001, when the boys in the study were between ages 16 and 20. Statistical analysis of these data permits the researchers to identify which characteristics of eight-year-olds predict later criminality.  Conduct problems or hyperactivity when the boys were eight years old predicted later criminality.  Survey profiles indicating that at age eight boys were “living in a broken home” or were living in a home characterized by “low parental education level” also were statistical predictors of later criminality. 

The researchers identify family structure as a particularly important statistical predictor of later criminality. The data indicate that family structure predicts all five types of crime examined in this study: drug offenses, violence, property offenses, traffic offenses, and drunk driving.  Indeed, by using a final statistical model that takes into account all other significant explanatory variables, the researchers highlight a child’s living in a broken home as the only independent predictor of all these types of criminality.   

When looking at the statistical linkage between family structure and criminality, the researchers also discern a pattern not evident for any other predictor of criminality.  The researchers underscore a “predictive association between family structure and the number of crimes [committed that] was linear: it was much stronger among those who had committed more than five crimes [Odds Ratio of 4.5] than among those who had committed one or two crimes [Odds Ratio of 1.6] or three to five crimes [Odds Ratio of 2.1].”

Given the high levels of divorce and out-of-wedlock childbearing in the United States in recent decades, American readers may well regard this new study as a sobering prediction of future social disorder. 

(Source: Andre Sourander et al., “Childhood Predictors of Male Criminality: A Prospective Population-Based Follow-up Study From Age 8 to Late Adolescence,” Journal of the American Academy of Child and Adolescent Psychiatry 45 [2006]: 578-586.) 

Forget the Neighborhood, Look at the Family  

Does living in a bad neighborhood expose children to a distinctively high risk of injury?  This quite sensible question recently prompted sociologists from the Universities of Leicester and Nottingham to start parsing data from Accident and Emergency (A & E) departments, looking for meaningful social patterns.  In publishing their final results in Social Medicine & Medicine, the British researchers acknowledge, however, that they “did not find a neighbourhood effect for A & E attendances that could not be explained by family-level characteristics.”

Carefully scrutinizing data collected for 1,717 Nottingham households, the British research team finds that neighborhood characteristics—such as access to a nearby hospital, number of parks and play areas, and presence of school-crossing patrols—fail as statistical predictors of children’s visits to an A & E department.  In contrast, the researchers find that family characteristics—such as the presence of a teenage mother, residence in rented accommodations, or single-parent family structure—do indeed serve as strong statistical predictors that children will end up in an A & E department.  In looking particularly at the effects of family structure, the British scholars calculate that children living in a single-parent family are 21% more likely to visit an A & E department than are children from two-parent families and are a remarkable 82% more likely to be actually admitted to a hospital. 

In the view of the authors of this new study, “The greater variation in injury rates between families than between neighbourhoods suggests reducing inequalities in injury rates may be achieved more effectively by focussing prevention at families rather than neighbourhoods.”

Might “focussing prevention at families” mean reversing the retreat from wedlock evident in the United Kingdom—and the United States—in recent decades?

(Source: Denise Kendrick et al., “Relationships between child, family and neighbourhood characteristics and childhood study,” Social Medicine & Medicine 61 [2005]: 1905-1915.)

Is Preschool Really Necessary?  

Even as the enrollment of children in kindergarten remains optional in most states, the daycare lobby and “early learning” advocates would like to make preschool universal or mandatory on the presumption that pre-K programs promote the “school readiness” of children. Yet a study by Lisa N. Hickman at the Ohio State University challenges their agenda, finding that children who attend daycare or preschool the year prior to kindergarten do not gain greater social or cognitive skills and in some measures end up lagging behind their peers who enjoy the attention of their parents exclusively.

Hickman looked at data from the Early Childhood Longitudinal Study, Kindergarten Cohort, which began tracking more than 21,000 children who started kindergarten in the fall of 1998. Improving upon the methodology of existing early childhood studies, she conducted both cross-sectional as well as longitudinal tests, the latter of which more accurately isolate effects of various preschool experiences over time.

Her cross-sectional tests confirm some existing research that finds that children who are enrolled in daycare or preschool start kindergarten with significantly higher cognitive skills, although that advantage is cut in half in tests that control for family background characteristics. At the same time, her cross-sectional analysis also confirms that children who follow the more traditional pattern of parental care start kindergarten with significantly better social skills in three of four different measures in tests with and without controls.

To test whether these patterns persist as the children move into higher grades, Hickman’s longitudinal tests control for fall test scores in kindergarten and first grade. During kindergarten, whatever advantages daycare or preschool children enjoy in math and reading become statistically insignificant in tests with and without background controls. During the first grade, the daycare/ preschool children have significantly lower math scores (p<.05). In both grades, these children scored significantly lower in the “approaches to learning” measure, which measured teacher perception of student attentiveness and persistence, a reversal of what was found in the cross-sectional test.

The longitudinal model also reveals even more so than the cross-sectional analysis that daycare/ preschool children exhibit poorer social skills throughout kindergarten. Such children have worse self-control, have worse interpersonal skills, and externalize problems more than their peers under parental care (p<.001 for each coefficient in tests with and without background controls). The only social measure (internalizing problem behaviors) where these children outperformed their parental-care peers in the first model is now insignificant.

While these findings will not endear Hickman to the “early learning” crowd, they nonetheless suggest that something other than the welfare of children may be driving the current pre-K craze.

(Source: Lisa N. Hickman, “Who Should Care for Our Children? The Effects of Home Versus Center Care on Child Cognition and Social Adjustment,” Journal of Family Issues 27 [May 2006]: 652-684.)

Another Fault of No-Fault  

The enactment of no-fault divorce legislation has been documented to have significantly increased the divorce rate in the United States. Now comes a study by two Canadian economists and their Taiwanese colleague that claims the long arm of the law of no-fault in the United States extends to the recent rise in the age of first marriage. That may not sound tragic, but the economists claim the trend represents a “small loss” in the welfare of the average man or woman and a greater loss to those who value marriage highly as they, now being forced to search harder and longer for life-partners to ensure against the risk of an easy divorce, remain single for longer periods of time.

The research team analyzed six million marriage records from the states and held by the National Center for Health Statistics from 1970 to 1995, comparing distributional statistics by divorce law, state, and year. In regressions that isolated the marginal effects of no-fault statutes, the mean log age at first marriage among women is statistically significant and positive, in some cases amounting to seven months. Men in no-fault states were also found to marry from one month to five months later than men in fault states. The mean increase in age for both men and women was small, roughly one or two months.

But they also discovered that the “spread” of the marriage age distribution changed more dramatically after the introduction of no-fault. Those who would have married earlier under traditional fault statutes now marry later, and those who would have married later before no-fault (presumably those who value the institution less) now marry earlier as they now have an easier exit strategy if the relationship turns sour. In tests that controlled for state-specific effects on the age of first marriage and for national trends over time, they found that no-fault divorce laws correlate robustly with a one to five percent decrease in the standard deviation of the age log at first marriage for both men and women.

While the increase in the mean age of marriage triggered by no-fault (one or two months) may seem trivial, the economists claim that the larger compression in the marriage-age distribution suggests that “some people had to search much longer [for a life-mate] whereas others had to search much less.” And even as an additional five months for some men or eight months for some women may still not seem alarming, their findings provide additional evidence that no-fault has come with a cost, especially for Americans who highly value marriage.

(Source: Douglas A. Allen, Krishna Pendakur, and Wing Suen, “No-Fault Divorce and the Compression of Marriage Ages,” Economic Inquiry 44 [July 2006]: 547-558.)

Family Structure Is Key  

Parents of teenagers are barraged with advice, especially when it comes to helping their children negotiate so-called “responsible sexual decisions.” The standard line is that if parents share dinnertime, participate in shared activities, maintain good relationships, and communicate about sex with their teens, they have little to fear. While this appeal contains a kernel of truth, a study by sociologists at the University of Texas and Penn State suggests not only that this therapeutic strategy has been oversold, but also that maintaining an intact family anchored in marriage is the more critical factor in keeping teen children on the straight and narrow.

Looking at data from the 1995 and 1996 waves of National Longitudinal Study of Adolescent Health, the researchers zeroed in on 8,663 teens, all of which had reported being virgins in the first wave. They discovered that for most teens, living in households headed by something other than both their biological parents is significantly associated with losing one’s virginity by the second wave. In the total sample, and among the boys when broken down by gender, the correlations were statistically significant for every one of the six alternative living arrangements.

More poignant, the researchers found that the four parental involvement indicators mentioned above—some of which also correlated with teen virginity—in most cases did not explain the robust correlations between family structure and teen sex, which remained statistically significant in models that introduced the four parental involvement variables. “The overall results suggest that parenting (or those aspects of parenting studied here) is not an important mechanism for the effects of family structure on sexual decision making.”

Among the parental involvement correlations, only among girls and among whites were all four variables significantly associated with teen sexual behavior. Having family dinners, participating in shared activities, and maintaining good relationships were each independently associated with these categories of teens remaining chaste. But having more conversations with parents about sex—what the experts often berate parents for not pursuing—was significantly associated with a teen losing his or her virginity between the two waves. In the general sample, teens whose parents discussed sex with them more often were 16 percent more likely to have lost their virginity relative to their peers with fewer discussions.

Confirming the role of family structure, the sociologists also found that teens that experienced a change in family structure such as a parental divorce or remarriage between the two waves were more likely to have lost their virginity than those whose household composition remained the same.

While the women researchers were surprised that parental involvement indicators did not explain the links between family structure and teen sex, most parents intuitively know what matters most to teens is being able to come home to a married mom and dad.

(Source: Jennifer Pearson, Chandra Muller, and Michelle L. Frisco, “Parental Involvement, Family Structure, and Adolescent Sexual Decision Making,” Sociological Perspectives 49 [2006]: 67-90.)

Single Moms, Sick Children  

Epidemiologists interested in children’s health often look carefully at those children’s parents.  Among the parental characteristics scholars examine when predicting children’s health, ethnic identity and educational level often figure prominently.  But new research suggests that parents’ marital status actually predicts children’s health better than does their ethnic identity or educational attainment. 

To be sure, parents’ ethnic identity and educational attainment do receive attention in a new study in which researchers from Albert Einstein College of Medicine examine the ways  “social disadvantage” affects children’s health.  Parsing nationally representative data collected in 1994 and 1995 for 57,553 children, the Einstein scholars examine the effect on children’s health of four different measures of social disadvantage: poverty, minority ethnic status, low parental education, and family structure. 

But not all of these measures of social disadvantage consistently predict health problems for children in the researchers’ statistical tests.  “Minority race/ethnicity,” the Einstein scholars report, “was not consistently related to poor child health outcomes.”  The researchers likewise conclude that “low [parental] education was an inconsistent risk factor.  Sometimes it added significantly to [children’s health] risk, and sometimes it did not.”

Poverty and family structure, however, predict children’s health problems in all of the researchers’ statistical tests.  Not surprisingly, poverty statistically predicts an elevated incidence of children’s chronic health problems, an elevated likelihood of a health-related activity limitation, and a depressed likelihood of very good or excellent overall health. But family structure likewise predicts all three of these health outcomes for children.  “Those [children] not living in two-parent families,” the Einstein team acknowledges, “were in poorer health than those in two-parent families.”  Even in statistical models that account for other background characteristics, family structure—like poverty—predicts chronic health problems, activity restrictions, and impaired overall health status. 

In their concluding remarks, the researchers suggest that activists pushing for universal health insurance are not seeing the complete picture: “Although universal access [to health care] will deliver much-needed medical services to the nation’s children,” the researchers point out, “ . . . it will do little to reduce the effect of social disadvantage on child health.”  And since family structure is one of the key determinants of social disadvantage, perhaps it is time for activists to worry less about insurance and more about enduring matrimony. 

(Source: Laurie J. Bauman, Ellen J. Silver, and Ruth E. K. Stein, “Cumulative Social Disadvantage and Child Health,” Pediatrics 117 [2006]: 1321-1327.)

The Dark Shadow of Abortion  

For more than 30 years, feminist activists have been congratulating themselves on having won for American women a nearly unfettered right to abortion. That these activists ignore the unborn lives snuffed out has deeply troubled many Americans.  But new research is raising troubling questions even for Americans willing to focus exclusively on the well-being of the women who seek abortions.

Recently published in the Journal of Child Psychology and Psychiatry, a study conducted by researchers at Christchurch Health and Development Study indicates that when young women abort their unborn children, they often subsequently suffer from a range of serious psychological problems.  Examining data collected from 520 New Zealand young women for the first 25 years of their lives, the Christchurch scholars adduce evidence that “young women reporting abortions had elevated rates of mental health problems when compared with those becoming pregnant without abortion.”  The litany of psychological problems associated with abortion includes “depression, anxiety, suicidal behaviours, and substance use disorders.”

The association between abortion and psychological vulnerability persists in statistical models that take into account differences in psychological history and disparities in economic and social circumstances.  The researchers interpret the persistence of this linkage as strong evidence of “a possible causal linkage.”    Further statistical analysis clarifies the direction of causation, with the data indicating that “abortion leads to increased risk of mental health problems.”  The data do not support the view that “mental health problems lead to increased risk of abortion.”

The Christchurch scholars acknowledge that their findings are “inconsistent” with a statement published by the American Psychological Association (APA) assuring Americans that “risk of psychological harm is low” for women who undergo abortions.  However, the Christchurch scholars complain that the APA’s “relatively strong conclusion about the absence of harm from abortion was based on a relatively small number of studies which had one or more . . . [methodological] limitations.”  Underscoring their skepticism with regard to the APA’s assurances, the Christchurch scholars find it “difficult to disregard the real possibility that abortion amongst young women is associated with increased risk of mental health problems.” 

(Source: David M. Fergusson, L. John Horwood, and Elizabeth M. Ridder, “Abortion in young women and subsequent mental health,” Journal of Child Psychology and Psychiatry 47 [2006]: 16-24.)