Plan B (cont.)

Elsewhere, I described several strategies currently used by those scholars who, despite all the evidence, still seek to deny or minimize the effects of family structure on child well being. One of those strategies (I called it “Plan B”) is first to concede the worrisome correlations between one-parent homes and poor child outcomes, then perform some methodological hocus-pocus intended to make these correlations seem insignificant and misleading. Last time, I did my best to beat up on Robert W. Blum of the University of Minnesota, who deserved it. This time, the prize for the most brazen reliance on Plan B goes to Lynne C. Huffman and her colleagues at Stanford University, whose research on “Risk Factors for Academic and Behavioral Problems at the Beginning of School” is featured prominently in a recent report, Off to a Good Start, commissioned by the Child Mental Health Foundations and Agencies Network, a group of leading bigwigs from private foundations and U.S. government agencies concerned with families and children. The purpose of the report is to evaluate the “risk factors” that can undermine children’s social and emotional development and compromise their readiness for school.

Huffman and her colleagues conducted a comprehensive review of academic articles published on this general topic between 1986 and 1998. Based on this review, they identify 22 separate “risk factors” facing U.S. pre-schoolers, from low birth weight to low socioeconomic status. None of these risk factors will surprise you. And sure enough, one of them is “Family Composition.” Under this heading, we are briefly informed of several studies suggesting the harmful effects of divorce on children. The authors conclude: “Marital status and family composition, then, may be an important factor in school success or failure.”

Sounds reasonable so far, doesn’t it? But grab your wallet. Here comes Plan B. Huffman and her colleagues want us to understand that there are risk factors, and then there are risk factors. Some require immediate attention and urgent societal action. Others, however, hardly matter at all in any practical sense. These safe-to-ignore risk factors are “not a reasonable basis for structuring targeted interventions.”

How can this be? Watch this. There are actually three kinds of risk factors. Some are “fixed markers.” Their distinguishing trait is that they are “not amenable to intervention.” An example is race. Even if a child suffers (is “at risk”) for this reason, nothing can be done to change his or her skin color. Other risk factors, we learn, are “variable markers.” Even if changed, they are unlikely, in and of themselves, directly to affect outcomes for the child. They are interesting, but ultimately irrelevant. Finally, there are “causal risk factors.” Now we’re talking. These are the real deal. They are associated with poor outcomes, they are amenable to change, and, if changed, they are likely to improve child outcomes measurably. These let’s-get-busy risk factors for pre-schoolers include: learning problems, behavioral and emotional problems, conflicts with peer groups, poor relationships with teachers, age at school entry, poor parenting practices, and parents’ psychological problems.

“Family composition,” it turns out, is purely a non-starter. Granted, it “may be” important, but family structure is classified by these scholars as a “fixed marker,” just like skin color. Sorry, nothing can be done about it. Policy makers, philanthropists and social service agencies should therefore focus instead on factors that are actually “amenable to intervention.”

This is very silly stuff. First - as was also the case with Robert Blum - quite a few of the “causal risk factors” enumerated by these scholars (especially poor parenting practices, parents’ psychological problems, and children’s behavioral and emotional problems) are themselves causally linked to family structure. Yet according to these scholars, the problems disproportionately associated with one-parent homes are high priorities for intervention, whereas the demographic trend helping to generate the problems is a low priority for intervention. Go figure.

Second, notice the odd way in which many of the alleged causes of these problems (“causal risk factors”) become virtually indistinguishable from the problems themselves. We huff and we puff with our research, only to discover that having problems learning in school comes from . . . learning problems. Having trouble behaving in school come from . . . behavior problems. In practical terms, are we really making much progress here?

In fact, these almost comically tautological “findings” reveal an underlying conceptual problem that is widespread and anything but funny. For academics who study these issues, and for professionals who design and staff programs for children and families, there is constant pressure, when deciding where and how to intervene, to go causally as far downstream as possible, where a broad societal problem (such as family disintegration) can presumably be reconceptualized as a series of much narrower problems, each amenable to specialized treatment in clinical or quasi-clinical settings. Almost all current incentive structures in our society - financial, political, intellectual - encourage these professionals to adopt and defend this way of thinking. The model is essentially medical. The goal is always to make a narrowly scientific diagnosis, like a doctor identifying a disease, so that resources aimed at remediation can be carefully (here is a favorite word in this vocabulary) “targeted.”

This conceptual framework has obvious advantages when the issue is an actual disease, or when children suffer from problems that require drug prescriptions or specialized therapy. But this way of thinking is deeply flawed when the source of the problem - the “factor” that actually puts children at risk - is not little, but big; not narrow, but broad; not a medical disease, but a values-laden social trend. Rigidly attempting to specialize and medicalize the problem of why so many six-year-olds are not ready for school is an exercise that, under the aegis of specifying and “targeting” everything, takes us further and further away from social conditions which, a few miles upstream, are actually causing the problem.

As a result, good intentions notwithstanding, this whole approach ends up being a too-cheap solution, the easy way out. More and more children not ready for school? Well, let’s begin by making it clear that a 33 percent rate of unwed childbearing and the world’s highest divorce rate are subjects to which attention, in this regard, specifically need not be paid. Instead, let’s hire more specialists with advanced degrees, who can be in charge of a few more “targeted interventions.” Raise your hand if you think that this approach will solve the problem. 

Third and finally, we come to the curious idea of family structure as a “fixed marker,” an unalterable phenomenon. Now, obviously the majority of U.S. infants and toddlers in one-parent homes on any given day are also likely to be in one-parent homes the next day. Certainly very little of what help we currently offer these families - no prescription from the clinic, no parenting class for mom - is likely to return fathers to those homes from which they have been missing. So if we think only in terms of immediate casualties and current programs, and never in terms of prevention or longer-term attitudinal and behavioral change, then the “not amenable to intervention” label is partly defensible. I will also concede that, in our free society, government is well equipped to do some things, but directly shaping the pair-bonding and marrying behavior of people is not one of them.

At the same time, if I were one of the philanthropic or government leaders who commissioned this report, and based on its analysis I concluded that strengthening family structure was a legitimate and important national goal, I bet I could think of a dozen things to do. I might start by phoning someone in the state of Oklahoma, where a governor-led, public-private “Marriage Initiative” is now impressively mobilizing people and resources across the state to act on the idea that divorce and unwed childbearing rates in Oklahoma are “amenable to intervention.”

Bottom line? Forget the academic goobledy-gook about “targeted interventions” and “fixed markers.” It’s just smoke. If we as a society know that family disintegration harms children, then we ought at least to try to do something about it.

Sources: Lynne C. Huffman, Sara L. Mehlinger, and Amy S. Kerivan, “Risk Factors for Academic and Behavioral Problems at the Beginning of School,” in Huffman, et. al, Off to a Good Start (Chapel Hill, N.C.: FPG Child Development Center of the University of North Carolina, 2000).

First published Fall 2001.