There are few more controversial or politically charged topics than parenting. Advice columns and radio/television shows abound providing tips on the best way to raise children. Parenting, especially in this day and age of social media, when every move is constantly scrutinized, has become a touchy subject.
In the arena of juvenile justice, however, parenting has long been considered an important focus of intervention. The relationship between parenting styles and behavior is one of the most researched in all of criminology and crime prevention.
Yet in recent years, the causal relationship between parenting and behavior has come into question, particularly by scholars who argue that biology plays a larger role than the environment in producing outcomes in children. For example, in her book “The Nurture Assumption,” Judith Rich Harris argues that the long-standing belief that parenting styles affect the way children turn out is wrong — genes and peers are more influential. Others have followed in Harris’ footsteps.
Specifically referring to criminal conduct, Brian Boutwell has written several pieces arguing that parenting does not have much to do with children’s behavior. He begins his quintessential essay provocatively: “I want you to consider the possibility that your parents did not shape you as a person.” He then goes on to review all the reasons why parenting may not matter and that designs that cannot account for genetics have been misleading.
Why does this matter for juvenile justice? Because if it is true that parenting does not matter, then juvenile delinquency programs and crime prevention approaches should look elsewhere. If parents do not “shape” children, then programs and interventions that seek to improve parents’ supervision or relationships with youth won’t have much impact on delinquency.
The origins of this skepticism of the effects of parents revolves around the use of research methods that can account for genetic effects. In other words, traditional social science methods that correlate parenting practices with their children’s behavior is insufficient because such results may be confounded by a third variable (e.g., genes). In other words, the long-standing findings that parenting matters may be spurious, some argue.
This work is to be commended in demonstrating the deficits of observational research and the power of biology. It is very likely true that parenting effects have been overstated. But does that mean parenting does not make a difference at all in juvenile delinquency or other outcomes?
On this score, the record is clear. Parenting does matter, and it matters a lot. Evidence — not just observational social science, but randomized, experimental studies (the only kind that can uncover unambiguous causality) show this. Further, the kind of genetic, twin studies that critics of parenting effects use to dismiss them do not actually measure parenting styles or any other kind of interaction.
Experimental studies are essential to understand for juvenile justice practitioners. When the criticism is levied that “correlation does not equal causation” in observational studies, this means that the relationship between X and Y may be actually driven by an unmeasured variable. Thus X does not in fact cause Y.
In experimental studies, groups of individuals who meet some set of inclusion criteria are randomly assigned to a treatment and control group. By virtue of this randomization, before the treatment (say, a cognitive-behavioral program), the two groups should not vary with respect to any variable, measured or unmeasured. Thus the genetic variations in group A should match the variations in group B.
Then the program is instituted, time passes, and outcomes are evaluated. Because the groups were similar pretreatment, any differences that arise after treatment can be confidently attributed to the program. This is why experiments are considered the “gold standard” in evaluation research.
So what does the experimental evidence show with respect to parenting and juvenile justice?
First, programs that focus on parenting skills early in the child’s life have shown long-lasting effects. For example, David Olds’ Nurse Family Partnership (NFP), which actually began before the children were born, helped at-risk mothers with parenting behaviors such as discipline techniques and focused on the importance of healthy development (for both themselves and their children).
The results of the first randomized study testing this program were astounding. When the children were 15 years old, they were doing better than their control group counterparts on a host of outcomes, including education and criminal behavior.
Projections of the overall impact of the NFP, which has been rolled out to other communities, indicate that by the year 2031, it will prevent 90,000 violent crimes by youth, 594,000 property and public order crimes (e.g., vandalism, loitering) by youth, 36,000 youth arrests and 41,000 person-years of youth substance abuse. These are large and meaningful effects.
The studies that show little to no parenting effects are called “behavior genetic” studies — they do not actually measure genes but are able to control for them. What about studies that measure genes directly? These are called “molecular genetic” studies.
One interesting program, the Strong African American Families program, focuses on parenting styles to reduce risk behavior of juveniles. The results have shown in a randomized experiment that it works to improve parenting and thus behavior. Genes? The researchers, led by Gene Brody, found that the program reduced the effect of a genetic polymorphism (variant) that has been linked to risk behavior. Thus parenting matters — perhaps particularly for those with genetic risk.
On the juvenile justice front, programs for at-risk and delinquency-involved youth have also shown — again, experimentally — that parenting matters. For example, Blueprints for Healthy Youth Development is an initiative that helps collate effective prevention and rehabilitation approaches for juveniles. Their highest standard of evidence requires a program to have been shown effective with a randomized trial in more than one location — such approaches are given the designation of “model program.”
One example of a model program, used in my home state of Maine, is Functional Family Therapy (FFT). FFT focuses on parenting behaviors and parent-child interactions. The results of randomized experiments of FFT, in multiple states, show it reduces juvenile delinquency. A meta-analysis (a statistical compilation of studies) indicated that the program has a moderate effect on juvenile behavior.
Overall, parenting programs have been shown to positively affect both parents and children. Often, only short-term effects have been demonstrated, which, rather than indicating parenting does not matter, suggests that programs must do more to have a lasting impact on parenting behaviors.
Thus, despite considerable confusion regarding parenting — which is the best approach, does it actually matter? — the best available evidence suggests that parents and families remain a fertile site for intervention if we wish to improve juvenile outcomes and increase public safety. Juvenile justice practitioners should continue to explore programs, both for prevention and for juvenile delinquency, that target parents and families.
Michael Rocque is an assistant professor of sociology at Bates College. His research interests include life-course criminology, race and justice, and corrections.