This article's purpose is to begin to develop an understanding of knowledge translation of early developmental crime prevention. Methods Involves a narrative review of experiments of early developmental prevention with measures of delinquency and criminal offending, ...
Purpose To begin to develop an understanding of knowledge translation of early developmental crime prevention.
Methods Involves a narrative review of experiments of early developmental prevention with measures of delinquency and criminal offending, and profiles two leading experiments, the Cambridge-Somerville Youth Study (CSYS) and the Montréal Longitudinal-Experimental Study.
Results While the roots of early developmental crime prevention can be traced to studies of human development, experiments of preventive interventions are at the heart of knowledge translation and policy influence. This can be seen in the form of replications, the process of scaling up effective interventions for wider dissemination, and inspiration for prevention scientists to launch new and innovative experiments—sometimes with the aim to improve upon past results. For example, far from curtailing policy interest in a developmental approach to delinquency prevention or dampening the need for prevention experiments, the harmful effects reported in the 30-year follow-up of the CSYS instead had an influence on some new longitudinal-experimental studies in developmental and life-course criminology.
Conclusions New experiments are needed to continue to advance early developmental crime prevention, and further research is needed to add to our understanding of knowledge translation in this area.
Keywords: developmental crime prevention; randomized controlled experiment; knowledge translation; public policy
“… an evil man is rather like a sturdy boy, or a man of childish mind, and evil is simply want of reason at an age when it normally accrues to men by nature governed by discipline and experience of harm …”
Thomas Hobbes (1998/1647)
Developmental crime prevention refers to “interventions designed to prevent the development of criminal potential in individuals, especially those targeting risk and protective factors discovered in studies of human development” (Tonry and Farrington, 1995, p. 2). It is largely premised on the idea that “criminal activity is determined by behavioral and attitudinal patterns that have been learned during an individual’s development” (Tremblay and Craig, 1995, p. 151). The central idea here is that earlier experience determines later behavior. In the case of crime, there is considerable evidence of behavioral continuity: childhood antisocial behavior is linked with juvenile crime, juvenile crime with offending in early adulthood, and offending in early adulthood with adult criminality (Farrington, 2013).
Developmental prevention is one of several crime prevention strategies that operate outside of the criminal justice system, representing an alternative to punishment and crime control tactics (Welsh and Farrington, 2012). One of developmental prevention’s distinguishing features (in contrast to criminal justice responses) is a commitment to prevention in the first instance—intervening before a delinquent act has been committed, before a child comes in conflict with the law. Relatedly, there is often a special focus on the early years of the life-course, with developmental theory and science postulating that the early years are most influential in shaping later experiences. As Duncan and Magnuson (2004, p. 101) note: “Principles of developmental science suggest that although beneficial changes are possible at any point in life, interventions early on may be more effective at promoting well-being and competencies compared with interventions undertaken later in life.” It is this view that guides our particular focus on early developmental crime prevention.
Another key feature of developmental crime prevention is its larger focus on improving the life chances of at-risk children and families, extending far beyond delinquent or criminal behavior. Here, there is a special interest in other important life-course outcomes, such as education, health, and employment. In many respects, developmental prevention aims to foster more productive, healthy, and socially responsible citizens.
Writing 25 years ago, Tonry and Farrington (1995, p. 10) characterized developmental crime prevention as the “new frontier of crime prevention efforts.” It seems safe to say that in the intervening years developmental prevention has more than lived up to this view. This can be seen in a dramatic growth in experimentation and scholarly research on early developmental prevention, the state of its effectiveness (Farrington et al., 2017), and its policy influence in many countries across the world (World Health Organization, 2014). Importantly, preceding this “new frontier” recognition is a long, rich history on the developmental origins of antisocial behavior, the beginnings of developmental theory (Tremblay, 2007), with prevention experiments starting almost a century ago (Farrington, 1983).
The central aim of this paper is to begin to develop an understanding of knowledge translation of early developmental crime prevention. Through a review of prevention experiments we are not offering a definitive guide, but rather a window into this process. We are particularly interested in the role knowledge translation has played at two levels. The first has to do with advancing understanding of the mechanisms of early developmental crime prevention interventions. The second has to do with contributing to the evidence base for improving the life chances of at-risk children and families. Recognizing that there are a number of different strategies of knowledge translation, we have adopted the definition that captures its broad reach: “synthesizing and disseminating results of research” (Wathen and MacMillan, 2018, p. 320). Experiments of preventive interventions are at the heart of this view of knowledge translation. This can be seen in the form of replications, the process of scaling up effective interventions for wider dissemination, and inspiration for prevention scientists to launch new and innovative experiments—sometimes with the aim to improve upon past results.
In discussing the role of experiments in advancing knowledge about mechanisms and contributing to policy influence, it is useful to briefly reflect on an ongoing debate on this issue in criminology. One view holds that experiments have great potential to advance understanding of causal mechanisms. One source of support for this view comes from the use of the longitudinal-experimental design (Farrington, 2006; 2013). This type of design combines a randomized controlled experiment with a prospective longitudinal survey of the development of offending.[1] As noted by Farrington (2013, p. 493), one of the unique features of this design is that “causal hypotheses could be generated in the longitudinal study from risk and protective factors and life events and then tested on the same individuals in the experimental study.”
Mediation analysis can also play a role in the ability of experimental evidence to advance knowledge on causal mechanisms. This allows for a test of the “degree to which a reduction in crime is directly related to a reduction in a risk factor and/or an increase in a protective factor” (Elliott and Fagan, 2017, p. 110). Several of the leading commercially packaged evidence-based interventions (e.g., Functional Family Therapy, Multisystemic Therapy) have conducted studies on mediating effects, offering further support for the causal mechanism that serves as the basis of the experimentally evaluated program. In fact, there is an increasing push for this type of analysis to be used as part of the scientific standard for certifying an intervention as evidence-based (Elliott and Fagan, 2017).
Another view questions a reliance on the randomized controlled experiment as the evidentiary base for public policy, and calls for greater attention to three concepts that connect causal effects and policy recommendations: mechanisms and causal pathways; effect heterogeneity; and contextualization (Sampson et al., 2013). Put another way, it is an attempt to add more precision to the complexities that undergird the scientific part of the “translational process” in moving from evidence to policy. Importantly, this view in no way negates the “move to greater rigor in demonstrations of causality” (Nagin and Sampson, 2019, p. 124), as advocated by the evidence-based movement in criminology (Weisburd et al., 2016).
We begin with some background on knowledge translation in the context of experimentation of early developmental crime prevention interventions. In the next two sections, we profile two leading developmental prevention experiments, the Cambridge-Somerville Youth Study and the Montréal Longitudinal-Experimental Study. This is meant to provide some historical context for prevention experiments, as well as to capture the intellectual and institutional connections that can exist between experiments. In the following section we highlight two other experiments that are representative of a new phase of developmental prevention experiments, and suggest a further lineage to earlier prevention experiments. The final section provides a discussion and some conclusions.
Tremblay and Craig (1995) carried out the first major review of developmental crime prevention interventions. The review included all the prevention experiments (and their latest follow-ups) at the time, including the two profiled in the next two sections. It also identified several factors related to effectiveness: (1) the interventions lasted for a sufficient duration—at least one year; (2) they were multimodal, meaning that multiple risk factors were targeted with different interventions; and (3) they were implemented before adolescence. More recent reviews of developmental crime prevention interventions have identified similar programmatic features related to effectiveness (see Farrington et al., 2017; Piquero et al., 2016). Moreover, these reviews demonstrate the effectiveness of different modalities of developmental prevention, including home visitation, cognitive-based preschool, behavioral parent management training, and social skills training.
These reviews are also important because they say much about the type of knowledge translation that we wish to explore here: incremental steps through experimentation for the purpose of advancing scientific knowledge and informing public policy. The process sometimes plays out as such: One experiment is initiated and follow-ups are carried out over time. Other experiments are initiated at varying time intervals, some for the purpose of replication, some to test new intervention modalities, and others because of direct inspiration from previous experiments. Here, there is some type of connective tissue among experiments carried out by different investigators. Of course, it needs to be acknowledged that developmental prevention experiments are also initiated outside of the influence of this process.
Two developments in particular over the last several decades have helped to shape this process of knowledge translation. These developments are prevention science and the evidence-based model. Both are best understood as frameworks that can be applied across a wide range of contexts, settings, and populations. At their core, each is concerned that the highest standards of science should be used to advance knowledge and inform public policy.
Prevention science has its roots in public health and begins with a commitment to prevention that is grounded in the developmental epidemiology of specific health or social problems. In the case of youth violence prevention, for example, “Prevention science has provided a bridge between an understanding of how chronic violence develops and how prevention programs can interrupt that development” (Dodge, 2001, p. 63). Prospective longitudinal studies are used to identify the most important risk factors for offending. Randomized experimental and rigorous quasi-experimental designs are used to evaluate the efficacy and effectiveness of preventive interventions designed to tackle these factors.
Prevention science is by no means restricted to developmental prevention. However, the idea of an alternative, non-punitive response to the prevention of delinquency and offending is a crucial element in prevention science’s mission. Its influence on developmental prevention can be found in its adherence to the highest scientific standards, its effort to base prevention on epidemiology, and its focus on the process of taking preventive interventions to scale to achieve population-level impacts (Gottfredson et al., 2015).
The evidence-based model embraces prevention science’s commitment to the use of the most scientifically valid studies to evaluate interventions. What it adds is the utilization of accumulated scientific research evidence on effectiveness. The evidence-based approach attempts to overcome a multitude of obstacles in policymaking by ensuring that the best available evidence is considered in decisions to implement a program or policy designed to prevent crime. Put another way, it is about making sure that this research is at center stage in political and policy decisions. As noted by Petrosino (2000, p. 636), “An evidence-based approach requires that the results of rigorous evaluation be rationally integrated into decisions about interventions by policymakers and practitioners alike.”
In the next section we discuss the first experiment of an early developmental crime prevention intervention, the Cambridge-Somerville Youth Study, which marks the starting point for our focus on knowledge translation. While the study’s connection to prevention science is far more evident, it has begun to take on greater relevance to an evidence-based approach as part of the need to understand the lessons from preventive interventions that cause harm (McCord, 2003; Welsh and Rocque, 2014).
Begun in June 1939, the Cambridge-Somerville Youth Study (CSYS) is a delinquency prevention experiment with a prospective longitudinal survey of the development of offending embedded in it. The study is recognized as the first randomized controlled experiment in criminology (Weisburd and Petrosino, 2004), as well as one of the earliest experiments of a social intervention (Forsetlund et al., 2007). It has also been referred to as the earliest example of a longitudinal-experimental study with criminological outcomes (Farrington, 2006).
Six hundred fifty underprivileged boys of average and difficult temperament (later reduced to 506), ages 5 to 13 years (median = 10.5), from Cambridge and Somerville, Massachusetts, were placed in matched pairs and one member of each pair was randomly assigned to the treatment group. Referred to as “directed friendship,” the preventive intervention involved individual counseling through a range of activities and home visits with the families. The intervention lasted a mean average of 5.5 years, and it ended in December 1945. Boys in the control condition received no special services.
The study was developed and directed by Richard Clarke Cabot (1868-1939), a renowned professor of clinical medicine and social ethics at Harvard University. The study encapsulated Cabot’s life work and his overarching desire to combine science with social ethics (O’Brien, 1985). Cabot was also influenced by the high recidivism rates of juvenile reformatories of the day, and he set out to intervene in the lives of at-risk boys—“predelinquents” as he called them—before contact with the formal justice system (Cabot, 1930). Other influences, most notably, the work of William Healy and Augusta Bronner and Sheldon and Eleanor Glueck, also played an important role in shaping Cabot’s vision of the study (Welsh et al., 2017; 2019a; see also Laub and Sampson, 1991).
There have been four follow-up assessments of the CSYS: 1948 (Powers and Witmer, 1951); 1956 (McCord and McCord, 1959a, b); 1975-79 (McCord, 1978; 1981); and 2016 to present (Welsh et al., 2019b). The latter two are the most noteworthy and will be briefly discussed here. It is important to note that the first two follow-ups indicated that the intervention had no measurable effect on official offending.
Between 1975 and 1976 (mean age = 45 years), McCord (1978) located records for 94.9% or 480 of the 506 participants (of whom 48 had died). Continued data collection up to 1979 (mean age = 47 years) eventually located records for 97.6% or 494 participants (McCord, 1981). This 30-year follow-up found iatrogenic effects. Compared to the control group men, significantly more men in the treatment group committed two or more crimes, were alcoholics, had high blood pressure or heart trouble, had difficulty finding employment, and died earlier.
McCord (1992; see also Dishion et al., 1999) proposed a “peer deviancy” hypothesis, observing that peer contagion among treatment group youth who had attended summer camps appeared to explain much of the iatrogenic effects. According to McCord (2003), these camps likely allowed for a great deal of unstructured socializing, representing an ideal environment for deviancy training to take place (see Gottfredson, 2010). Reanalyzing the 30-year follow-up data, McCord (2003; see also Dishion et al., 1999) found that for boys who attended summer camp only once (n=59), the odds ratio (OR) predicting undesirable outcomes was 1.33, which was marginally higher than the OR of 1.12 among treatment group boys who did not attend summer camp. For boys who attended camp more than once (n=66), the OR for undesirable outcomes was 10.0, meaning that participants were ten times as likely to experience undesirable outcomes as their matched mates.[2] McCord (2002, p. 235) concluded: “I strongly suspect that the boys from the Youth Study tended to bond together, encouraging one another’s deviant values.” This would come to have important implications for the delivery of group-based interventions.
The latest follow-up of the study is also being carried out in two waves. Between 2016 and 2018, records were located for 96.4% or 488 of the 506 participants, with a total of 446 participants confirmed deceased (88.1%) and 42 alive (8.3%). Mortality was the focus of this 72-year post-intervention follow-up (Welsh et al., 2019b). Matched-pairs analyses showed no significant differences for all outcomes of interest: mortality at latest follow-up; premature mortality (younger than 40 years); and cause of mortality (natural versus unnatural). A Cox proportional hazard regression analysis indicated no difference in time to death between the treatment and control group men. In not being able to detect iatrogenic effects on mortality, the main implication is that the iatrogenic effects on mortality experienced in middle age did not persist to old age (i.e., up to age 90).
One view of this change in intervention effects over time is that the observed iatrogenic effects on mortality in middle age “may have been a singular event, irrespective of their concordance with effects for a wide range of other outcomes” (Welsh et al., 2019b, p. 8). In commenting on the findings of this follow-up, Farrington and Hawkins (2019) suggest that the earlier iatrogenic effects may have been rather weak, as indicated by tests of statistical significance and the magnitude of the effect. The next wave of the follow-up, which is underway and focused on criminal offending, may provide further insight on an explanation for the study’s latest findings on mortality.
In the context of knowledge translation and policy influence, the CSYS helped to inform early developmental crime prevention in several key ways. One of these ways had to do with the nature of the study’s approach to preventing delinquency. This involved two important elements. First, by targeting pre-delinquent boys, it was driven by prevention in the first instance. Cabot was motivated by this approach out of deep concern for the high recidivism rates of juvenile reformatories and with the potential for labeling effects from youths being arrested, adjudicated, and incarcerated. Second, the study took a holistic view of delinquency prevention. Here, becoming a “good citizen,” “character development,” or “building constructive personalities” were seen as larger ambitions of the work being performed with the treatment group boys (Powers, 1950, p. 23). Later writings also emphasized this point: “Crime prevention involves socialization for good citizenship, and this involves teaching children not only how to act but also how to change and develop into productive adults” (McCord, 1996, p. 2).
Returning to the third follow-up of the CSYS, the identification of peer deviancy as a causal mechanism for the iatrogenic effects drew attention to the need for greater scrutiny of group-based preventive interventions (Dodge et al., 2005). McCord would partner with other researchers on a study reporting similar findings (Dishion et al., 1999; Poulin et al., 2001), and she would become part of a larger team of scholars that carried out new research on peer contagion and called for policy change in the delivery of group-based preventive interventions (Gifford-Smith et al., 2005). Avoidance of grouping delinquent youth of varying risk levels and unstructured socialization as well as greater adult involvement were some of the lasting contributions of this work.
Yet another key influence of the study is that the findings became important to the understanding that good intentions do not guarantee desirable effects (Dishion et al., 1999). Whether it is in the context of assessing the effectiveness of early developmental crime prevention or as part of the larger debate on prevention versus punishment, we must be mindful that all social interventions have the potential to cause harm (Welsh and Rocque, 2014). In the words of McCord (2003, p. 17), “Unless social programs are evaluated for potential harm as well as benefit, safety as well as efficacy, the choice of which social programs to use will remain a dangerous guess.”
Moreover, far from curtailing policy interest in a developmental approach to delinquency prevention or dampening the need for developmental prevention experiments, the harmful effects reported in the third follow-up instead helped to influence new longitudinal-experimental studies in developmental and life-course criminology (Tremblay et al., 2019). In the next section, we profile one of these studies, the Montréal Longitudinal-Experimental Study, which was “designed to prevent the iatrogenic effects of interventions she [Joan McCord] had identified” (Tremblay and Farrington, 2004, p. 6).
The Montréal Longitudinal-Experimental Study (Boisjoli et al., 2007; Castellanos-Ryan et al., 2013; Tremblay et al., 1992; Vitaro et al., 2001) was initially conceived as a prospective longitudinal study of the early origins of juvenile delinquency. One of the main influences in planning the study was West and Farrington’s 1973 book, Who Becomes Delinquent? The major innovation was to start four years earlier (at age 6) than the Cambridge Study in Delinquent Development of 411 boys in London (at age 10). A sample of 1,031 kindergarten boys from 53 schools in low socioeconomic neighborhoods of Montréal was selected. The boys were assessed annually until age 15 years and at different time points up to early adulthood. While conducting the first assessment in the Spring of 1984, the research team realized that they had an opportunity to integrate a prevention experiment within the longitudinal study. Having read McCord’s article on the long-term iatrogenic effects of the CSYS (McCord, 1978), Tremblay contacted McCord to ask her to help plan the Montréal experiment.
The global objective of the experiment was to test to what extent an intensive preventive intervention between ages 7 and 9 years would prevent delinquency during adolescence for the most aggressive, hyperactive, and oppositional kindergarten boys. A total of 250 boys were randomly assigned to one of three groups (prevention [n=69], attention-control [n=123], or control [n=58]). The attention-control group was involved in numerous contacts over the years to study their behavior with parents and peers, but there were no efforts to change their behavior. The control group was assessed only once a year. The 250 families were offered to participate in the experiment with 172 families (69%) agreeing to participate. However, all the at-risk boys (n=250) were assessed yearly and were included in intent-to-treat analyses.
The preventive intervention was based on previous developmental prevention experiments. However, the intensity of the intervention was substantially increased by simultaneously targeting three important sources of influence on children’s social development: peers, parents, and teachers. This was achieved by integrating three types of preventive interventions that had been tested separately: (1) social skills training for the disruptive boys (Michelson et al., 1983), aiming to foster prosocial behavior towards peers; (2) parent training in effective child rearing, based on the Oregon Social Learning Center model (Patterson et al., 1975); and (3) support for teachers concerning the at-risk boys, based on the same model that was used for parents, to insure that the parents and the teachers would use similar child rearing practices.
The iatrogenic effects observed in the CSYS played an important role in designing the social skills training part of the Montréal intervention. As part of her work with the Montréal team, Joan McCord hypothesized that participation in summer camps, where deviant children spent much time together, could be part of the mechanisms that produced the iatrogenic effects in the CSYS.[3] This led to an important innovation. Rather than group the at-risk boys to work on their social skills, groups were created that had the following ratio: one disruptive boy to three highly prosocial boys. This strategy was an attempt to use peers as positive role models. There is evidence from the follow-up assessments that the treatment boys had friends who were less deviant than the control boys (Tremblay et al., 1992). Grouping one or two deviant boys with a majority of highly prosocial boys appears to have created lasting prosocial friendships. Indeed, alongside the behavioral parent training, this would prove to be an important causal mechanism in the effectiveness of the preventive intervention.
The intervention lasted two school years, from September 1985 to June 1987. Boys were 7 years old when the intervention started and 9 years old when it ended. Over 85% of the boys who participated in the intervention attended a minimum of two-thirds of the social skills training sessions. Parents were given as many sessions as needed to master the parenting skills, and 75% of the parents covered at least two-thirds of the content and objectives of the planned training. Unfortunately, teachers demonstrated low interest and limited availability; they were generally not able to spend much time discussing teaching strategies for a given child. Therefore, meetings with teachers were much fewer than planned. Interventions with parents, children, and teachers were carried out by three full-time university-trained professionals (psychologist, social worker, and psycho-educator).
The first follow-up, two years after the end of the intervention, indicated desirable effects according to self-reports. Compared to the control group, the treatment group reported significantly less fighting and theft (Tremblay et al., 1991). Follow-ups from 11 to 17 years of age indicated that the treatment boys, compared to the controls, reported less physical aggression, vandalism, and theft (Lacourse et al., 2002). Also, the treatment boys, compared to the controls, reported less substance use between ages 14 and 17 years (Castellanos-Ryan et al., 2013). The latter analysis indicated that the reduction in substance use was partly explained by a decrease in impulsivity, antisocial behavior, and affiliation with deviant peers during pre-adolescence. The latest follow-up, at 24 years of age, showed that the treatment group, compared to the control group, had a higher rate of high school graduation and lower rates of criminal convictions (Boisjoli et al., 2007).
It is important to note that these positive results are still far from what could be considered the solution for preventing most cases of juvenile delinquency or adult criminality. Longitudinal studies beginning during pregnancy or at birth have now shown that behavior problems do not start in kindergarten or in elementary or secondary school. The most serious cases of behavior problems start during early childhood. It has been clearly shown that the frequency of physical aggression peaks between 2 and 4 years of age. Most children learn to control these natural tendencies (e.g., temper tantrums, physical aggression toward others), but those who do not learn before school entry tend to live in families with numerous problems and are at higher risk of behavior problems from early childhood to adolescence (Tremblay, Vitaro, and Côté, 2018). From this perspective, it appears reasonable to offer preventive interventions that begin during pregnancy. This is the focus of the next section.
The idea of initiating preventive interventions very early in the life-course, although not as early as pregnancy, likely did not originate with the Cambridge-Somerville Youth Study, but it did become a central feature of the intervention. The study’s founder and first director, Richard Clarke Cabot, wanted to intervene in the lives of at-risk boys as early as possible, initially seeking “to get them as young as we can” (Cabot, 1935)—no older than 7 years of age (Powers, 1949). However, the clearest suggestion to start interventions before birth was formulated in the first World Health Organization report, published three years after the organization’s creation in 1948: Psychiatric Aspects of Delinquency (Bovet, 1951). After reviewing the research on juvenile delinquency in the previous 50 years, Bovet (1951, p. 56) concluded: “Treatment of the non-delinquent little girl or teen-age girl might perhaps have been the most efficacious prophylaxis for the delinquency which a few years later will break out in her sons.” Longitudinal research from infancy to adulthood in the past 40 years has indeed shown that the best predictors of children’s behavior problems, especially physical aggression, are maternal characteristics, which include antisocial behavior and mental health problems (Tremblay et al., 2018). In this section, we discuss two leading prevention experiments that were initiated in pregnancy and provided continued interventions to children and parents through the child’s formative years. They represent a new phase of experiments of early developmental crime prevention.
Developed by David Olds, the Nurse-Family Partnership (NFP) is perhaps the most well-known and influential prevention experiment—actually group of experiments—that began in pregnancy and has a direct measure of delinquency or criminal offending. NFP was first tested in Elmira, New York, in the early 1980s. One of its targets was the prevention of child abuse, which has often been linked to delinquency (Spatz Widom and Brzustowicz, 2006). Four hundred first-time and disadvantaged mothers were randomly assigned to receive home visits from nurses during pregnancy, or to receive visits both during pregnancy and during the child’s first two years of life, or to a control group that received no visits. Visits lasted about one hour and the mothers were visited on average every two weeks. The nurses gave advice about prenatal and postnatal care of the child, about infant development, and about the importance of proper nutrition and avoiding smoking and drinking during pregnancy.
Findings of this experiment showed that prenatal plus postnatal home visits caused a significant decrease in recorded child physical abuse and neglect during the first two years of life, especially by poor, unmarried, teenage mothers; 4% of visited versus 19% of non-visited mothers were guilty of child abuse and neglect (Olds et al., 1986). In a 15-year follow-up, which included 330 mothers and 315 children, significantly fewer treatment compared to control group mothers were identified as perpetrators of child abuse and neglect (29% vs. 54%) and, for the higher risk sample only, significantly fewer treatment mothers, in contrast to the controls, had alcohol or substance abuse problems or were arrested. At the age of 15, children of the treatment mothers had committed significantly fewer violent and other major criminal acts than their control counterparts (Olds et al., 1998). In the latest follow-up at age 19, compared to the controls, girls of the full sample of mothers had incurred significantly fewer arrests and convictions and girls of the higher risk mothers had significantly fewer children of their own; few effects were observed for the boys (Eckenrode et al., 2010). These results suggest that, from an intergenerational perspective, the causal mechanism of intensive, nurse-delivered support for at-risk young pregnant women will have long-term positive impacts on their daughters who hopefully will select better adapted mates and have better adapted sons (Tremblay et al., 2018). This mechanism was also key to the betterment of the mothers’ immediate circumstances.
Large-scale replication experiments of NFP in Memphis, Tennessee, and Denver, Colorado, have also shown desirable effects on a wide range of outcomes for both nurse-visited mothers and their children (Kitzman et al., 2010; Olds et al., 2014). Moreover, the success of NFP has resulted in its use in almost 600 counties in 42 states across the United States, serving more than 34,000 families each year (Nurse-Family Partnership, 2019). It is also being replicated in many other countries, including Australia, Canada, Germany, Norway, and the United Kingdom.
A more recent example of a preventive intervention that started during pregnancy is the Preparing for Life (PFL) study in Ireland (Côté et al., 2018). Beginning in 2008, more than 200 pregnant women from a disadvantaged community in Dublin were randomly assigned to a treatment group (n=115) or a control group (n=118). The treatment group received regular home visits by trained professionals (known as PFL mentors), beginning in pregnancy to the child’s fifth birthday, as well as baby massage and a parenting intervention delivered in small groups. The control group received some services (e.g., annual stipend for child developmental toys and books, assistance with access to enhanced preschool, public health workshops, local services). The mentors came from a variety of professional backgrounds, including education, social care, youth studies, psychology, and early childcare and education. The children’s behavioral problems (externalizing, internalizing), cognitive skills (general, vocabulary), and health service use (number of health clinic visits) were regularly assessed from 6 months to 4 years of age.
An early assessment found that the preventive intervention improved children’s cognitive development from 18 months of age onwards. Compared to their control counterparts, children who received the high treatment supports had better general cognitive functioning and more of them scored above average from 24 months onwards. By the age of 4 years, results showed that children in the treatment group, compared to the controls, were more likely to follow high-level cognitive development and vocabulary skills trajectories. There was also a positive impact on the development of behavior problems (Côté et al., 2018). At the latest assessment, by school entry, the PFL had a significant and large impact on children’s cognitive development. Children who received the high treatment supports, compared to the controls, had better general cognitive functioning, spatial abilities, non-verbal reasoning skills, and basic numeracy skills. This means that they were better at understanding information, seeing patterns, solving problems, and working with numbers (PFL Evaluation Team, 2016).
This paper set out to begin to develop an understanding of knowledge translation of early developmental crime prevention, with a focus on the role of experiments in identifying causal mechanisms and contributing to the evidence base for improving the life chances of at-risk children and families. Our starting point was the Cambridge-Somerville Youth Study and its inspiration for another pioneering experiment, the Montréal Longitudinal-Experimental Study, followed by a new phase of developmental prevention experiments. It is through this lineage of prevention experiments, albeit by no means direct, that we can begin to develop an understanding of knowledge translation at work.
Crucial to this work are intergenerational mechanisms involved in the development of antisocial behavior. While not the central focus of this paper, we view it as important to raise here as a way to draw connections with some ongoing research and new directions for research on prevention experiments. Almost a century ago, the work of the Gluecks (Glueck and Glueck, 1930) and later the work of Joan McCord (McCord and McCord, 1959a, b) helped to establish the idea that there are strong intergenerational mechanisms involved in the development of antisocial behavior. It has now become clear that imitation of parent’s behavior and poor parenting are not the only intergenerational mechanisms involved. Intergenerational studies of mental health problems with national population samples (Nordsletten et al., 2016) have shown that assortative mating (having children with a partner who has similar problems) is frequent and increases the probability of negative outcomes, both at the environmental and genetic levels (Tremblay et al., 2018). Studies with twins have also shown strong genetic effects on the ability to control physical aggression as early as two years of age (Lacourse et al., 2014), while epigenetic studies (control of gene expression) are showing that the genes transmitted from parents to children can be affected by the parent’s behavior before, during, and after pregnancy (Tremblay et al., 2018). A recent analysis of the impact of the NFP intervention showed that child maltreatment impacts the expression of genes and it had an impact on DNA methylation related to abuse and neglect 25 years after the end of the intervention (O’Donnell et al., 2018).
These complex intergenerational transmission phenomena require new approaches to the study of early preventive interventions. One new strategy is the use of identical twins. Because they share the same genes, one of the twins can serve as a control for the other twin in an experimental design. The United States’ space agency, NASA, recently used this strategy to study the effects of long missions into space. A pair of identical twins (monozygotic) was trained as astronauts. One of the two was sent into space for a long period and the other stayed on earth. Medical and psychological assessments of both twins before and after the space flight led to a detailed study published in Science (Garrett-Bakelman et al., 2019). Similar experiments could be done from pregnancy onwards with monozygotic twins born in high risk families.
As important as this process of knowledge translation has been to advancing understanding of the mechanisms of early developmental prevention, it has also played a role in contributing to the evidence base for more effective early developmental crime prevention interventions. The Montréal Longitudinal-Experimental Study and the Nurse-Family Partnership are two notable examples in the field, as reported in respective systematic reviews (Piquero et al., 2016; Duffee et al., 2017). Past prevention experiments have given rise to novel ideas for new preventive interventions, inspiration for prevention scientists to improve upon past results, and replications. And it is here that further research is needed to add to our understanding of knowledge translation in early developmental crime prevention. Learning from the connections between the past and the present can be informative, as well as offer guidance in the years to come.
We are especially grateful to the journal editors and the anonymous reviewers for insightful comments.
Boisjoli, R., Vitaro, F., Lacourse, É., Barker, E. D., & Tremblay, R. E. (2007). Impact and clinical significance of a preventive intervention for disruptive boys: 15 year follow-up. British Journal of Psychiatry, 191, 415-419.
Bovet, L. (1951). Psychiatric aspects of delinquency. Geneva, Switzerland: World Health Organization.
Cabot, R. C. (1930). Foreword. In S. Glueck & E. T. Glueck, 500 criminal careers (pp. vii-xiii). New York: Knopf.
Cabot, R. C. (1935). Letter to Dr. H. A. Murray, Jr., October 19, 1935. HUG 4255: Box 97. Richard Clarke Cabot Papers, Pusey Library, Harvard University Archives.
Castellanos-Ryan, N., Séguin, J. R., Vitaro, F., Parent, S., & Tremblay, R. E. (2013). Impact of a 2-year multimodal intervention for disruptive 6-year-olds on substance use in adolescence: A randomised controlled trial. British Journal of Psychiatry, 203, 188-195.
Côté, S. M., Orri, M., Tremblay, R. E., & Doyle, O. (2018). A multicomponent early intervention program and trajectories of behavior, cognition, and health. Pediatrics, 141(5), 1-14 (e20173174).
Dishion, T. J., McCord, J., & Poulin, F. (1999). When interventions harm: Peer groups and problem behavior. American Psychologist, 54, 755-764.
Dodge, K. A. (2001). The science of youth violence prevention: Progressing from developmental epidemiology to efficacy to effectiveness to public policy. American Journal of Preventive Medicine, 20, 63-70.
Dodge, K. A., Dishion, T. J., & Lansford, J. E. (Eds.). (2005). Deviant peer influences in programs for youth: Problems and solutions. New York: Guilford Press.
Duffee, J. H., Mendelsohn, A. L., Kuo, A. A., Legano, L. A., & Earls, M. F. (2017). Early childhood home visiting. Pediatrics, 140(3), 1-14 (e20172150).
Duncan, G. J., & Magnuson, K. (2004). Individual and parent-based intervention strategies for promoting human capital and positive behavior. In P. L. Chase-Lansdale, K. Kiernan, & R. J. Friedman (Eds.), Human development across lives and generations: The potential for change (pp. 93-135). New York: Cambridge University Press.
Eckenrode, J., Campa, M., Luckey, D. W., Henderson, C. R., Cole, R., Kitzman, H., Anson, E., Sidora-Arcoleo, K., Powers, J., & Olds, D. L. (2010). Long-term effects of prenatal and infancy nurse home visitation on the life course of youths: 19-year follow-up a randomized trial. Archives of Pediatrics and Adolescent Medicine, 164, 9-15.
Elliott, D. S., & Fagan, A. A. (2017). The prevention of crime. New York: Wiley-Blackwell.
Farrington, D. P. (1983). Randomized experiments on crime and justice. In M. Tonry & N. Morris (Eds.), Crime and justice: An annual review of research, vol. 4 (pp. 257-308). Chicago: University of Chicago Press.
Farrington, D. P. (2006). Key longitudinal-experimental studies in criminology. Journal of Experimental Criminology, 2, 121-141.
Farrington, D. P. (2013). Longitudinal and experimental research in criminology. In M. Tonry (Ed.), Crime and justice 1975-2025 (pp. 453-527). Chicago: University of Chicago Press.
Farrington, D. P., Gaffney, H., Lösel, F., & Ttofi, M. M. (2017). Systematic reviews of the effectiveness of developmental prevention programs in reducing delinquency, aggression, and bullying. Aggression and Violent Behavior, 33, 91-106.
Farrington, D. P., & Hawkins, J. D. (2019). The need for long-term follow-ups of delinquency prevention experiments. JAMA Network Open, 2(3), 1-3 (e190782).
Forsetlund, L., Chalmers, I., & Bjørndal, A. (2007). When was random allocation first used to generate comparison groups in experiments to assess the effect of social interventions? Economics of Innovation and New Technology, 16, 371-384.
Garrett-Bakelman, F. E., Darshi, M., Green, S. J., Gur, R. C., Ling, L., et al. (2019). The NASA twins study: A multidimensional analysis of a year-long human spaceflight. Science, 364(6436), 1-20 (eaau8650).
Gifford-Smith, M., Dodge, K. A., Dishion, T. J., & McCord, J. (2005). Peer influence in children and adolescents: Crossing the bridge from developmental to intervention science. Journal of Abnormal Child Psychology, 33, 255-265.
Glueck, S., & Glueck, E. T. (1930). 500 criminal careers. New York: Knopf.
Gottfredson, D. C. (2010). Deviancy training: Understanding how preventive interventions harm. Journal of Experimental Criminology, 6, 229-243.
Gottfredson, D. C., Cook, T. D., Gardner, F. E., Gorman-Smith, D., Howe, G. W., Sandler, I. N., & Zafft, K. M. (2015). Standards of evidence for efficacy, effectiveness, and scale-up research in prevention science: Next generation. Prevention Science, 16, 893-926.
Hobbes, T. (1998/1647). De cive; or The citizen. Cambridge, UK: Cambridge University Press.
Kitzman, H. J., Olds, D. L., Cole, R. E., Hanks, C. E., Anson, E. A., Arcoleo, K. J., et al. (2010). Enduring effects of prenatal and infancy home visiting by nurses on children: Follow-up of a randomized trial among children at age 12 years. Archives of Pediatrics and Adolescent Medicine, 164, 412-418.
Lacourse, É., Côté, S., Nagin, D. S., Vitaro, F., Brendgen, M., & Tremblay, R. E. (2002). A longitudinal-experimental approach to testing theories of antisocial behavior development. Development and Psychopathology, 14, 909-924.
Lacourse, É., Boivin, M., Brendgen, M., Petitclerc, A., Girard, A., Vitaro, F., Paquin, S., Ouellet-Morin, I., Dionne, G., & Tremblay, R. E. (2014). A longitudinal twin study of physical aggression during early childhood: Evidence for a developmentally dynamic genome. Psychological Medicine, 44, 2617-2627.
Laub, J. H., & Sampson, R. J. (1991). The Sutherland-Glueck debate: On the sociology of criminological knowledge. American Journal of Sociology, 96, 1402-1440.
McCord, J. (1978). A thirty-year follow-up of treatment effects. American Psychologist, 33, 284-289.
McCord, J. (1981). Consideration of some effects of a counseling program. In S. E. Martin, L. B. Sechrest, & R. Redner (Eds.), New directions in the rehabilitation of criminal offenders (pp. 394-405). Washington, DC: National Academy Press.
McCord, J. (1992). The Cambridge-Somerville Study: A pioneering longitudinal-experimental study of delinquency prevention. In J. McCord & R. E. Tremblay (Eds.), Preventing antisocial behavior: Interventions from birth through adolescence (pp. 196-206). New York: Guilford Press.
McCord, J. (1996). Prevention and the grounds for action. Paper presented at the Seminar on Early Prevention, Stockholm, Sweden, December 2-3, 1996.
McCord, J. (2002). Counterproductive juvenile justice. Australian and New Zealand Journal of Criminology, 35, 230-237.
McCord, J. (2003). Cures that harm: Unanticipated outcomes of crime prevention programs. Annals of the American Academy of Political and Social Science, 587, 16-30.
McCord, J., & McCord, W. (1959a). A follow-up report on the Cambridge-Somerville Youth Study. Annals of the American Academy of Political and Social Science, 322, 89-96.
McCord, W., & McCord, J. (1959b). Origins of crime: A new evaluation of the Cambridge-Somerville Youth Study. New York: Columbia University Press.
Michelson, L., Sugai, D. P., Wood, R. P., & Kazdin, A. E. (1983). Social skills assessment and training with children. New York: Plenum Press.
Nagin, D. S., & Sampson, R. J. (2019). The real gold standard: Measuring counterfactual worlds that matter most to social science and policy. Annual Review of Criminology, 2, 123-145.
Nordsletten, E., Larsson, H., Crowley, J. J., Almqvist, C., Lichtenstein, P., Mataix-Cols, D. (2016). Patterns of nonrandom mating within and across 11 major psychiatric disorders. JAMA Psychiatry, 73, 354-366.
Nurse-Family Partnership (2019). Nurse-Family Partnership: National snapshot. Available at: www.nursefamilypartnership.org. Accessed May 17, 2019.
O’Brien, L. (1985). ‘A bold plunge into the sea of values’: The career of Dr. Richard Cabot. New England Quarterly, 58, 533-553.
O’Donnell, K. J., Chen, L., MacIsaac, J. L., McEwen, L. M., Nguyen, T., Beckmann, K., Zhu, Y., Chen, L. M., Brooks-Gunn, J., Goldman, D., et al. (2018). DNA methylome variation in a prenatal nurse-visitation program that reduces child maltreatment: A 27-year follow-up. Translational Psychiatry, 8(15).
Olds, D. L., Henderson, C. R., Chamberlin, R., & Tatelbaum, R. (1986). Preventing child abuse and neglect: A randomized trial of nurse home visitation. Pediatrics, 78, 65-78.
Olds, D. L., Henderson, C. R., Cole, R., Eckenrode, J., Kitzman, H., Luckey, D., Pettitt, L. M., Sidora, K., Morris, P., & Powers, J. (1998). Long-term effects of nurse home visitation on children's criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial. Journal of the American Medical Association, 280, 1238-1244.
Olds, D. L., Holmberg, J. R., Donelan-McCall, N., Luckey, D. W., Knudtson, M. D., & Robinson, J. (2014). Effects of home visits by paraprofessionals and by nurses on children: Follow-up of a randomized trial at ages 6 and 9 years. JAMA Pediatrics, 168, 114-121.
Patterson, G. R., Reid, J. B., Jones, R. R., & Conger, R. D. (1975). A social learning approach to family intervention, vol. 1. Families with aggressive children. Eugene, OR: Castalia.
Petrosino, A. (2000). How can we respond effectively to juvenile crime? Pediatrics, 105, 635-637.
PFL Evaluation Team (2016). Promoting child development by supporting parents: Learning from the evaluation of the Preparing for Life home visiting programme. Dublin, Ireland: Geary Institute, University College Dublin.
Piquero, A. R., Jennings, W. G., Diamond, B., Farrington, D. P., Tremblay, R. E., Welsh, B. C., & Gonzalez, J. (2016). A meta-analysis update on the effects of early family/parent training programs on antisocial behavior and delinquency. Journal of Experimental Criminology, 12, 229-248.
Poulin, F., Dishion, T. J., & Burraston, B. (2001). 3-year iatrogenic effects associated with aggressive high-risk adolescents in cognitive-behavioral preventive interventions. Applied Developmental Science, 5, 214-224.
Powers, E. (1949). An experiment in prevention of delinquency. Annals of the American Academy of Political and Social Science, 261, 77-88.
Powers, E. (1950). Some reflections on juvenile delinquency. Federal Probation, 14, 21-26.
Powers, E., & Witmer, H. L. (1951). An experiment in the prevention of delinquency: The Cambridge-Somerville Youth Study. New York: Columbia University Press.
Sampson, R. J., Winship, C., & Knight, C. (2013). Translating causal claims: Principles and strategies for policy-relevant criminology. Criminology & Public Policy, 12,587-616.
Spatz Widom, C., & Brzustowicz, L. M. (2006). MAOA and the “cycle of violence”: Childhood abuse and neglect, MAOA genotype, and risk for violent and antisocial behavior. Biological Psychiatry, 60, 684-689.
Tonry, M., & Farrington, D. P. (1995). Strategic approaches to crime prevention. In M. Tonry & D. P. Farrington (Eds.), Building a safer society: Strategic approaches to crime prevention (pp. 1-20). Chicago: University of Chicago Press.
Tremblay, R. E. (2007). The development of youth violence: An old story with new data. European Journal on Criminal Policy and Research, 13, 161-170.
Tremblay, R. E., & Craig, W. M. (1995). Developmental crime prevention. In M. Tonry & D. P. Farrington (Eds.), Building a safer society: Strategic approaches to crime prevention (pp. 151-236). Chicago: University of Chicago Press.
Tremblay, R. E., & Farrington, D. P. (2004). In memoriam: Joan McCord, 1930-2004. The Criminologist, 29(2), 6-7.
Tremblay, R. E., McCord, J., Boileau, H., Charlebois, P., Gagnon, C., Leblanc, M., & Larivée, S. (1991). Can disruptive boys be helped to become competent? Psychiatry: Journal for the Study of Interpersonal Processes, 54, 148-161.
Tremblay, R. E., Vitaro, F., Bertrand, L., LeBlanc, M., Beauchesne, H., Boileau, H., & David, L. (1992). Parent and child training to prevent early onset of delinquency: The Montréal Longitudinal-Experimental Study. In J. McCord & R. E. Tremblay (Eds.), Preventing antisocial behavior: Interventions from birth through adolescence (pp. 117-138). New York: Guilford Press.
Tremblay, R. E., Vitaro, F., & Cȏté, S. (2018). Developmental origins of chronic physical aggression: A bio-psycho-social model for the next generation of preventive interventions. Annual Review of Psychology, 69, 383-407.
Tremblay, R. E., Welsh, B. C., & Sayre-McCord, G. (2019). Crime and life-course, prevention, experiments, and truth seeking: Joan McCord’s pioneering contributions to criminology. Annual Review of Criminology, 2, 1-20.
Vitaro, F., Brendgen, M., & Tremblay, R. E. (2001). Preventive intervention: Assessing its effects on the trajectories of delinquency and testing for mediational processes. Applied Developmental Science, 5, 201-213.
Wathen, C. N., & MacMillan, H. L. (2018). The role of integrated knowledge translation in intervention research. Prevention Science, 19, 319-327.
Weisburd, D., Farrington, D. P., & Gill, C. E. (Eds.) (2016). What works in crime prevention and rehabilitation: Lessons from systematic reviews. New York: Springer.
Weisburd, D., & Petrosino, A. (2004). Experiments, criminology. In K. Kempf-Leonard (Ed.), Encyclopedia of social measurement (pp. 877-884). San Diego: Academic Press.
Welsh, B. C., & Farrington, D. P. (2012). Science, politics, and crime prevention: Toward a new crime policy. Journal of Criminal Justice, 40, 128-133.
Welsh, B. C., & Rocque, M. (2014). When crime prevention harms: A review of systematic reviews. Journal of Experimental Criminology, 10, 245-266.
Welsh, B. C., Zane, S. N., & Rocque, M. (2017). Delinquency prevention for individual change: Richard Clarke Cabot and the making of the Cambridge-Somerville Youth Study. Journal of Criminal Justice, 52, 79-89.
Welsh, B. C., Dill, N. E., & Zane, S. N. (2019a). The first delinquency prevention experiment: A socio-historical review of the origins of the Cambridge-Somerville Youth Study’s research design. Journal of Experimental Criminology, 15, 441-451.
Welsh, B. C., Zane, S. N., Zimmerman, G. M., & Yohros, A. (2019b). Association of a crime prevention program for boys with mortality 72 years after the intervention: Follow-up of a randomized clinical trial. JAMA Network Open, 2(3), 1-11 (e190782).
West, D. J., & Farrington, D. P. (1973). Who becomes delinquent? London: Heinemann.
World Health Organization (2014). Global status report on violence prevention 2014. Geneva, Switzerland: World Health Organization.
[1] All of the studies reviewed in this paper use longitudinal-experimental designs.
[2] As noted by Dishion et al. (1999, p. 761), these “comparisons are based on random assignment within pairs matched prior to the treatment.” This means that the original design of the experiment—matching plus random allocation to treatment and control conditions—was not compromised in testing the peer deviancy explanation for the program’s iatrogenic effects.
[3] It is important to note that McCord’s hypothesis about peer deviancy was developed after her 1978 article (McCord, 1978), and was shared with Tremblay in advance of its first published account in 1992 (McCord, 1992).
Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC-BY-NC-ND 4.0)
