Jumping to Conclusions About Child Custody (Lawyers, Please Read)

The Monitor on Psychology, a publication for members of the American Psychological Association, has an “in brief” section where summaries of interesting research are presented. I have often had a beef or beeves with “in brief” summaries involving child care and parenting, which too frequently suggest a cause and effect relationship where none has been shown to exist. I am really disturbed by this in cases where people who assume that one thing causes a particular outcome for children may want to argue in court for particular orders about child custody or treatment, and may cite research as a support for their arguments even when the research does not actually provide the evidence they imply.

A recent case in point: in the “in brief” section, and under the headline “Shared Custody Matters” , the Monitor (December, 2017, p. 13) reports a Swedish study that compared psychological problems of preschool children in the joint custody of divorced parents with the problems of those who spent all or most of their time with one parent, and with those of preschoolers in intact families. (This study is summarized at DOI:10.1111/apa.140014. The article is by Bergstrom, Fransson, Fabian, Hjern, Sarkadi, & Salari – “Preschool children living in joint physical custody show less psychological symptoms than those living mostly or only with one parent”, Acta Paediatrica , 2017).  The researchers reported that children in intact families had the smallest number of psychological problems, followed by those in joint custody, and those in the care of only one parent had the most problems.

There are a number of Internet references to this study that suggest that shared custody has thus been shown to help children psychologically, and some of them use language like “children of divorce” that encourages readers to ignore the ages of the preschool children in the study and to generalize from this study’s results to conclusions about children of all ages.

So what’s my problem with this? Does the article not show that shared custody is better for children? No, actually it does not, and it can only be interpreted as doing so if the reader ignores the multiple factors that affect child psychological outcomes, as well as the bidirectional influences of parents on children and vice versa. Such an interpretation also avoids the fact that the groups of families studied may have had many important differences that directly affected their custody arrangements.

Given that many divorcing families choose joint custody (in the U.S. this is the default position of most judges, as well), why do some not follow this pattern? How are families with joint custody potentially different from those without it? The list of differences is substantial.

High-conflict parents may avoid shared custody, and exposure to conflict may cause psychological problems for children. Parents who have serious physical or mental illnesses may be less likely to be able to share custody, and children may be affected by the parents’ disorders. Children with existing physical or mental problems that make them challenging to care for may be avoided by one parent, who will be unwilling to share custody, whereas the same parents might be quite ready to care for healthier or “easier” children. Parents who have the assistance of their own parents or other family members may spend more time caring for their children than those who do not have such assistance, and depending on the family circumstances, this may or may not affect the children’s psychological health. Parents who have committed domestic violence may be regarded warily by co-parents, who may try to reduce the time the abusive person spends with the children – either the domestic violence of the past or the present fears of  one parent can affect the children psychologically.

Being able to name alternative causes is not the same thing as showing evidence that one or all of them are at work, but this exercise does underline the fact that a nonrandomized study of naturally-occurring events cannot be used to conclude that a single factor caused an outcome. I don’t say the original authors did this, but other interpreters of the research seem to have done so.

I’d like to touch on another point about the Bergstrom et al study. The children were preschoolers, ages 3-5 years, but some Internet references speak of them as “children of divorce” in a general way. There are real problems about generalizing from preschoolers to older children, or vice versa, as their developmental strengths and vulnerabilities may influence the ways they react to experiences. In addition, if all the children in a study are preschoolers, this means that the divorce took place no more than 5 years ago, and in many cases is likely to have been much more recent. These families are still working out new roles and functions in the wake of the divorce and may take two years or more to manage these tasks and arrive at a calmer period of their lives. “Children of divorce”, taken across the developmental range, are in many cases past the early chaos of post-divorce adjustment and have established relatively successful ways to function. For children in those later years, who are also more mature cognitively and emotionally than preschoolers are, psychological responses to care arrangements will not necessarily be the same as those of younger children.

I would hate to see the Bergstrom et al study dragged into the courtroom as an argument in child custody cases where there is a real concern about the children’s contact with one of the parents. Having plenty of contact with each of two “good” parents is beneficial for children, but plenty of contact with a “bad” parent is not necessarily desirable and may even be harmful. The Bergstrom study does not adequately support the claim that joint custody is superior across the board to single-parent custody, with respect to children’s psychological development. Let’s hope that lawyers and judges understand this and don’t jump to conclusions.