Preventing Violence: Understanding and Addressing Determinants of Youth Violence in the United States

This brief summarizes a review of research and evaluation studies, as well as promising and proven interventions, to identify programs, policies, and practices that can contribute to reducing high levels of violence in the United States. Reducing violence is not a topic of controversy – virtually everyone would like to see reductions in injury, harm, and mortality due to violence. The question is, how can violence be reduced?

Download Publication

Additional Downloads


Preventing ImageRates of all types of violence have declined substantially in the United States over recent decades. Nevertheless, rates of violence and the numbers of children and youth affected by violence remain high compared with other countries. Moreover, data indicate great variation across states and communities. The fact that there is so much variation across states and countries suggests that there is substantial opportunity to reduce high rates of violence.

Violence comes, of course, in many forms. In this brief, we use the following definition of violence: “The intentional use of physical force or power, threatened or actual, against oneself, another person, or against a group or community that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment, or deprivation.

While we focus on the child in this review, violence is often intergenerational; hence, adults are frequently critical actors. Our review includes varied forms of violence, including child maltreatment, crime/delinquency, gang violence, intimate partner violence, suicide, self-harm, and general physical aggression.

Key Findings and Implications

  • Violence appears in many forms, but there are common risk and protective factors found across types of violence. A child or family that experiences multiple risk factors and few protective factors faces a particularly high risk of experiencing violence, either as a victim, as a perpetrator, or both.
  • While the United States has high rates of violence compared with other countries, many programs and approaches have been identified that could reduce violence, if scaled up with quality.
  • Prevention of violence is preferable to treatment, but emerging evidence from neuroscientists indicates significant plasticity (in other words, malleability) of the human brain, including in individuals experiencing trauma, supporting the perspective that treatment can make a difference.
  • Social and economic disparities are strongly correlated with violence and are malleable; however, we have not focused on these because other interventions seem more immediately realistic.
  • Interventions are available at the level of individuals, the family, schools, and communities.
    • For individuals, problems with self-regulation, sleep, hostile attributions about other people’s intentions, and abuse of substances are risk factors. While mental health problems are not generally a cause of violence, the combination of substance use and mental health issues does elevate the risk of violence. Individuals with mental health issues and/or disabilities are more likely to be victims of violence.
    • Family factors represent an important determinant of violence. Potential interventions include the prevention of unintended pregnancy, programs to prevent and treat intimate partner violence, and parenting education.
    • Schools are another important locus for intervention, and efforts to improve school climate include a focus on improving engagement, safety, and environment by developing social and emotional skills, reducing of bullying and other physical and emotional safety issues, and creating consistent and fair disciplinary policies.
  • High levels of violence across the United States compared with other countries suggest that there are beliefs, values, and policies underlying our national culture that, if better understood and thoughtfully discussed, could reduce violence.
  • Many of the interventions that might be pursued to reduce violence are useful in their own right (for example, reducing substance abuse); the fact that these interventions can also reduce violence should give them added importance and urgency.

Identifying the Determinants of Violence

This brief summarizes a review of research and evaluation studies, as well as promising and proven interventions, to identify programs, policies, and practices that can contribute to reducing high levels of violence in the United States. Reducing violence is not a topic of controversy – virtually everyone would like to see reductions in injury, harm, and mortality due to violence. The question is, how can violence be reduced?
We have drawn on available research to identify a broad range of factors that predict a similarly broad range of types of violence. These are depicted in the chart below, which arrays varied types of violence across the top and identifies potential causes or determinants of violence along the left side. Each cell summarizes our sense of the strength of the research evidence linking each determinant of violence with each type of violence. A bold X indicates strong evidence of an association, while an unbolded, orange X indicates more moderate evidence, and a small x indicates weak evidence. Weak evidence can reflect a lack of research or a small association, or it may reflect an uneven research literature, such that some determinants have been heavily researched while others have not been as widely explored. In addition, some factors have been explored in rigorous studies that control for confounding influences, while others are based on weaker research methods. Alternatively, it may be that some predictors have effects that are more universal, while others do not. One longitudinal study (following children over time) that examines a broad range of types of violence and a broad array of risk and protective factors would help resolve this question.

Our review identifies a number of common predictors or determinants of violence. These are factors that are consistently found associated with higher levels of violence across varied types of violence. That is, whether violence takes the form of delinquency, suicide, or domestic violence, there are many common predictors. These determinants represent many of the forms of trauma experienced by children and youth incorporated as “adverse childhood experiences” or ACEs, but the set of determinants goes beyond these factors.

The critical take-away from this chart is that many of the predictors of violence affect many or even most of the types of violence. Child maltreatment, for example, predicts every type of violence; that is, every cell is filled with an X. This suggests that reducing child abuse and addressing related trauma would have a number of positive effects on varied types of violence and suggests another reason (beyond the inherent importance of preventing harm to children) to prevent these adverse experiences.

Other common determinants include domestic violence, gun availability, harsh and dysfunctional parenting, low self-control, and a lack of school connectedness. Domestic violence/ intimate partner violence predicts every type of violence. Other predictors appear to be related to just some types of violence, for example, attribution of hostile intent to others, poor quality sleep or not enough sleep, a sense of efficacy in acting collectively, and unintended pregnancy, which has been found to be associated with about half of the varied types of violence.

Chart on Detriments of Violence

Preventing Image02

Examining Misperceptions

Some misperceptions were also identified regarding the causes of violence. For example, despite the media emphasis on mental health issues as a major cause of violence, research indicates that mental health problems only modestly increase the probability of violence, though whether certain mental health conditions create an elevated risk is a topic for additional research. Substance abuse is a far more substantial determinant of violence; and the combination of substance abuse and mental health problems is also a source of violence. Individuals with mental health issues are, though, more likely to be victims of violence. Moreover, parent mental health can represent a risk factor for children, as well as parents being unable to build positive relationships with their children and provide consistent positive parenting.

Focusing on approaches to reduce these common determinants of violence represents an important direction for prevention and treatment. Accordingly, in the course of our review, we examined in depth a number of factors that, if addressed, could reduce multiple types of violence.

In addition, to inform strategies to address these common determinants of violence, we have identified rigorously evaluated programs that have impacts on these factors. We have also sought to identify new approaches, where possible, to expand the range of opportunities to address the high and costly levels of violence in the United States. We have highlighted varied policies and initiatives that go beyond programmatic approaches, though we find a dearth of rigorous research on these apparently important factors. The same is true for cultural factors. There is little understanding of the cultural beliefs or values that underlie the high rates of violence found in the U.S.

How the U.S. Compares to Rest of the World

Rates of Violence: United States and Worldwide
Intentional homicide, 2012: 437,000 deaths worldwide
-36 percent in the Americas
-31 percent in Africa
-28 percent in Asia
-5 percent in Europe
-0.3 percent in Oceania
Under 15 – 8 percent of all victims
Ages 15-29 – 43 percent of victims
Overall homicide rate, 2012:
United States – 4.7 per 100,000 population
France – 1.0
Germany – 0.8
Canada – 1.6
Firearm death rate: The overall U.S. firearm homicide rate is 20 times higher than the combined rates of 22 countries that are our peers in wealth and population, and American children die by guns 11 times as often as children in other high-income countries.
Non-fatal youth violence: Globally, for every youth homicide there are around 20 to 40 victims of non-fatal youth violence receiving hospital treatment. The rates of non-fatal violent injuries tend to increase dramatically during mid-adolescence and young adulthood.
UNODC, 2014. Global study on homicide 2013: trends, contexts, data.
Garbarino, Bradshaw, & Vorrasi, 2002. Mitigating the Effects of Gun Violence on Children and Youth.
Richardson & Hemenway, 2010. Homicide, Suicide, and Unintentional Firearm Fatality: Comparing the United
States With Other High-Income Countries, 2003.
Krug, Dahlberg, Mercy, Zwi, & Lozano 2002. World report on violence and health.

Opportunities to Reduce Violence

Overall, violence in the U.S. has been declining since the mid-to late-1990s, although rates for some kinds of violence have remained flat or increased somewhat recently.

The review identified numerous opportunities for further reducing violence, including some overlooked opportunities. For example, a lack of school connectedness and, to a lesser extent, poor school performance are both linked to greater violence. Clearly there are many reasons to foster academic achievement and connectedness. Preventing violence represents an additional and very important reason.

Family planning programs represent another overlooked opportunity. We find that unplanned pregnancy is a predictor of many forms of violence directed at the mother, such as domestic violence, and the child, such as child maltreatment. Unplanned childbearing is also a correlate, as the child grows up, of an increased risk for delinquency, crime, and gang violence. Again, while there are many reasons to assist couples in avoiding unplanned pregnancy, helping to reduce violence represents another, relatively ignored, reason.

In general, the importance of socioemotional learning needs to be elevated in the discussion. Risk factors such as poor self-regulation provide malleable points of intervention that could have a number of positive outcomes, including a reduction in violence.

Recent advances in technology (such as computerized screeners in waiting rooms) make it easier to screen youth for violence and associated risk factors, and technology is increasing the reach of some proven programs. (For example, some home visiting programs send text messages, and some parenting programs deliver some content via videos that can be accessed from any computer with an internet connection.) Widespread use of texting and smartphone applications can potentially increase the reach of already-proven programs to a larger audience, and open the door to innovative new approaches, such as using video games to teach and reinforce skills in a medium that is embraced by youth.

Electronic technologies are also being used to help train professionals in the field, interactively and with a more flexible schedule. Training can be done when individuals have time, rather than their having to attend a webinar or conference. Virtual trainings that include the use of avatars to help teachers and health professionals hone important skills related to violence prevention can also help to broadly disseminate evidence-based practices.

Prevention interventions can also take advantage of emerging computer and communication technologies. Finally, there are video games that teach and reinforce positive skills such as problem solving and self-regulation in a medium that is embraced by youth.

Positive media represents another approach that seems to fly under the radar screen. Characters that provide role models for positive behaviors, including positive approaches to conflict resolution, relationships, and interaction with peers and family, can help children, and even youth, learn better social and emotional skills.

Exploring the Role of Culture and Social Factors

Unfortunately, some issues, such as the role of American culture, have been difficult to explore.

It is clear that the United States has higher levels of violence than most developed nations, but it is not clear which cultural values or beliefs drive or permit such high levels of violence. Changing the public’s understanding of violence seems like an important avenue for efforts to reduce violence, but it may be necessary to conduct research on the values that citizens hold and how they are framed in order to understand how cultural values may contribute to ongoing high levels of violence.

It is important to recognize that the antecedents of violence include well-documented disparities, particularly poverty, parent education, race/ethnicity, neighborhood quality, and family structure. While socioeconomic differences are theoretically malleable, we have not focused on these in this paper because other routes to reducing violence appear to be more pragmatic. Despite this, it is critical to note that these disparities underlie and magnify the importance of other risk factors. Accordingly, achieving reductions in social and economic disadvantages needs to be on any list of strategies to reduce violence.

Parenting behaviors have proven difficult to change, but harsh and dysfunctional parenting represents an important risk factor for children’s development, and we perceive considerable support for empowering parents to be the best parents for their child that they can be. Helping to prevent child abuse and neglect represent particularly critical paths, and approaches to identify trauma and treat children and parents are being developed.

The Role of the Education, Health, Justice, and Community Sectors

The education sector. A focus on academic achievement has expanded to encompass the importance of non-academic—or socioemotional—skills in enhancing school success and supporting student development. Initiatives to improve school climate and build student connectedness include efforts to reduce bullying, develop student self-regulation, and reduce the frequency of attributing hostile intentions to the behavior of others. Like many of the interventions to reduce violence, it is likely that these interventions will improve school outcomes, such as attendance and academic performance, as well as the predictors of violence.

The health sector. Health insurance can play a valuable role in addressing substance use, mental health issues, and treatment of injury. The availability of health insurance coverage for screening is not consistent. Recognizing that prevention is cheaper in every sense of the word than treatment, ways to support preventive approaches merit consideration. The health system also provides screenings and services for parents and can therefore address varied determinants of violence, including parental depression, harsh discipline and dysfunctional parenting, domestic violence, and intimate partner violence. In addition, as unintended pregnancy is another determinant of violence, the health sector can help address high rates of unintended pregnancy.

The justice sector. Developing better approaches to addressing child welfare and juvenile justice represents a critical challenge. Again, stronger prevention and treatment programs and policies are needed. For example, treatment of behavior problems rather than incarceration represents one valuable direction for many youth. Similarly, alternate approaches to incarceration for parents convicted of non-violent offenses is another strategy to consider, if families can be strengthened and supported rather than further disrupted.

The community sector. Media campaigns have been used to good effect to address many issues, such as smoking and sudden infant death syndrome, and thus represent an approach worth considering. More direct cross-sector approaches to building neighborhood and community collective efficacy have been explored; they are difficult to evaluate but, importantly, they recognize that high rates of violence are concentrated in particular communities and thus that this sector is also relevant to reducing violence. Initiatives include Defending Childhood, the National Forum on Youth Violence Prevention, My Brother’s Keeper, and Community-Based Violence Prevention. Another strategy, Safe and Sound, is being implemented in several sites and focuses on treatment with evidence-based approaches to reduce costly programs such as foster care; the savings are then invested in evidence-based prevention programs.

Evidence-Based Programs for Reducing Violence

Our review identified a number of programs that have been rigorously evaluated and found to have significant impacts on reducing varied forms of violence. These are depicted in Figure B, ordered according to the ages when the programs are appropriate. These programs are described in detail in Child Trends’ LINKS database of experimentally evaluated social programs for children and youth (

However, the extent to which these programs are offered in the United States and the proportion of all children and youth receiving any of these interventions are not known, nor is the extent to which they are reaching at-risk populations. In addition, evaluations frequently do not assess the long-term impacts of even these fairly well-known effective programs. Moreover, we find that many programs have only been evaluated from a narrow perspective. That is, many programs have only been evaluated for a particular, specific outcome, though it appears likely that the program affects multiple outcomes or a constellation of related outcomes. For example, Botvin’s Life Skills Training program was developed to address substance use but was subsequently found to also affect delinquency. While we do not endorse fishing for impacts, it may be appropriate for program evaluators to identify several theory-based confirmatory outcomes as well as a broader set of exploratory outcomes. Also, the aggregate impact of implementing multiple complementary programs needs to be assessed.

Proven Programs by Target Age

Preventing Image03




Most of all, it is critical to focus on prevention. Once a violent act has occurred — be it bullying, child abuse, suicide, or murder — the consequences cannot be undone. Advocates often say that we know what to do; we just need to do it. Researchers, however, often say that more research is needed before action is taken. In this case, while further research and evaluation would be beneficial, enough is known to warrant action. Understanding how to build the private and public will to support the implementation of evidence-based programs, practices, and policies may represent the most urgent research need.

For more detailed information on this study, read the full final report at


© Child Trends 2015. May be reprinted with citation.
Child Trends appreciates the contribution and partnership of Futures Without Violence in conducting this review. This report was developed with support from Futures Without Violence based on a grant from the Robert Wood Johnson Foundation.


Kristin Anderson Moore
Brandon Stratford
Selma Caal
Carl Hanson
Shelby Hickman
Deborah Temkin
Hannah Schmitz
Joy Thompson
Susannah Horton
Alyssa Shaw