Stronger Families, Safer Streets

Exploring Links Between Family Structure and Crime

By Rafael A. Mangual, Brad Wilcox, Seth Cannon, and Joseph E. Price December 2023
 TABLE OF CONTENTS

Executive Summary 3
Literature Review : The Psychological Effects of Family Instability on Crime 5
Father Absence, Young Adult Males, and Crime 8
The Current Study: Family Structure and Crime 11
Conclusion 19

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime
 Executive Summary

This Institute for Family Studies report finds that strong families are associated with less crime in cities across the
United States, as well in neighborhoods across Chicago. Specifically, our analyses indicate that the total crime rate in
cities with high levels of single parenthood are 48% higher than those with low levels of single parenthood. When it
comes to violent crime and homicide, cities with high levels of single parenthood have 118% higher rates of violence
and 255% higher rates of homicide. And in Chicago, our analysis of census tract data from the city shows that
tracts with high levels of single-parent-headed households face 137% higher total crime rates, 226% higher violent
crime rates, and 436% higher homicide rates, compared to tracts with low levels of single parenthood. We also find
that poverty, education, and race are linked to city and census-tract level trends in crime. In general, in cities across
America, and on the streets of Chicago, this report finds that public safety is greater in communities where the two-
parent family is the dominant norm.

The debate about how best to respond to urban crime—a debate that has become more important in light of recent
increases in violent crime and homicide in many cities across America—has tended to focus on two perspectives.
The first prioritizes tackling the “social structural factors” (unemployment, economic inequality, poverty, etc.) that are
thought to be the “root causes” of crime, and violent crime, in particular.1 A second perspective rejects this structural
approach in favor of a strategy that relies on traditional law-enforcement institutions (namely, police, prosecutors,
and jails/prisons), often citing the sharp violent crime declines of the 1990s and 2000s that occurred in the wake of
new policing and prosecutorial approaches—even in the face of structural realities said to be at the root of the urban
crime problem.2

But a third perspective seeks to understand how the fragile state of core social institutions—schools, churches, youth
sports leagues, and, above all, families—in too many of our cities may also have a hand in urban crime. Princeton
sociologist Patrick Sharkey, for instance, has argued that nonprofits “focused on reducing violence and building stronger
communities” are linked to lower rates of violent crime in cities across the country.3 In this Institute for Family Studies
report, we turn our attention to the core institution of family. Drawing on the work of scholars like Harvard sociologist
Robert Sampson—who found that “(f )amily structure is one of the strongest, if not the strongest, predictors of ... urban
violence across cities in the United States”4—we explore the relationship between family structure and urban crime in
the 21st century. Specifically, we address this question: How is family structure associated with crime, violent crime, and
homicide rates in American cities—and with these outcomes in Chicago neighborhoods?

1
April D. Fernandez and Robert D. Crutchfield. “Race, Crime, and Criminal Justice: Fifty Years Since the Challenge of Crime in a Free Society,” Criminology &
Public Policy 17.2 (2018): 397–417. p. 401.
2
RA Mangual, “Restoring Public Safety,” Manhattan Institute, December 2022.
3
Patrick Sharkey, Gerard Torrats-Espinosa, and Delaram Takyar, “Community and the Crime Decline: The Causal Effect of Local Nonprofits on Violent Crime,”
American Sociological Review 82, no. 6 (December 1, 2017): 1214–40.
4
Robert J. Sampson, “Unemployment and imbalanced sex ratios: Race-specific consequences for family structure and crime,” The Decline in Marriage Among African
Americans: Causes, Consequences and Policy Implications (Russell Sage Foundation, 1995): 249.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 3
 We find that cities are safer when two-parent families are dominant and more crime-ridden when family instability
is common. The same story applies to the neighborhoods of Chicago. More specifically, we find the total crime rate
is about 48% higher in cities that have above the median share of single-parent families, compared to cities that have
fewer single-parent families. That difference is even larger with respect to violent crime and homicide, specifically,
with cities above the median level of single parenthood experiencing 118% higher rates of violent crime and 255%
higher rates of homicide. In the Windy City, relying on an analysis of census tract level data, our research indicates
that neighborhoods above the median fraction of single-parent-headed households experienced 137% higher total
crime rates, 226% higher violent crime rates, and 436% higher homicide rates.

When controlling for additional factors such as racial composition, poverty rates, and educational attainment levels,
we find that the association between family structure and total crime rates, as well as violent crime rates, in cities
across the United States remains statistically significant. However, the association between family structure and
homicide in cities does not. In Chicago, the links between family structure and both violent crime and homicide
rates at the neighborhood level were significant, net of controls, but
not the total crime rate. In addition to the question of whether there
exists a statistical relationship between family structure and crime—a We find that cities are safer
question we generally answer in the affirmative—this study also
offers possible answers to the question of what might explain the
when two-parent families
relationships between family instability and crime. are dominant and more
crime-ridden when family
Drawing on an interdisciplinary body of social science research, we
theorize that this relationship is likely a byproduct of some mix of
instability is common. The
the heightened risk of family instability in the socialization of young same story applies to the
children, and the role that father absence plays in providing less neighborhoods of Chicago.
guidance and oversight for adolescent and young adult males.

Particularly in light of the preexisting literature on the role of family

structure in various life outcomes, these findings may have important implications for policymakers. They suggest the
need to encourage more young Americans—particularly those living in vulnerable neighborhoods with both high
rates of violence and out-of-wedlock childbearing—toward forming strong and stable families in marriage.5

5
Goesling, Brian, Hande Inanc, and Angela Rachidi, “Success Sequence: A Synthesis of the Literature, OPRE Report 2020-41 (Washington, DC: Office of
Planning, Research, and Evaluation, Administration for Children and Families, U.S. DHHS, 2020).

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 4
 LITERATURE REVIEW

The Psychological Effects of Family Instability on Crime

Mental Disorder Prevalence Among Criminal Offenders

Though there is intense debate about the root causes of criminal violence, a large body of evidence establishes that
criminal offenders—young and old, male and female—are far more likely to have a mental disorder than the general
population. With respect to the criminal offending population, borderline and antisocial personality disorders (BPD
and ASPD), as well as substance use disorders (SUDs), are especially prevalent.

Among juvenile offenders in particular, conduct disorders (CDs)—a precursor to ASPD6—are quite common, with
a prevalence ranging between 33-70%, depending on the populations studied.7 And though prevailing estimates
suggest it only affects between 1-3% of the general population, estimates of the percentage of prisoners around
the world that can be diagnosed with ASPD range higher than 40 percent.8 With respect to substance use, a 2016
meta-analysis published by the Society for the Study of Addiction concluded that approximately 25% of newly
incarcerated male prisoners have an alcohol or drug-use disorder.9 Furthermore, as Rafael Mangual noted in a jointly
authored paper with Professors John Paul Wright and Matthew DeLisi:

A Bureau of Justice Statistics (BJS) report, using data from 2007–09, found that 58% of state prisoners and 63% of
sentenced jail inmates met the diagnostic criteria for drug abuse or dependence.10

Both measures are significantly higher than the prevalence measure for the general population of Americans aged 12
and older, which was recently reported to be 16.5% by the Department of Health and Human Services.11 And while
between 1-2% of the general population can be diagnosed with BPD, “rates among both male and female inmates
have been estimated at 12 percent to 30 percent,” according to a 2010 paper published in Corrections Compendium.12

Not only are these mental disorders more prevalent among criminal offending populations, their prevalence rivals or
surpasses that of poverty—often pointed to as a root cause of crime—among American prisoners.13 The questions,

6
Jason J. Washburn et al., “Development of Antisocial Personality Disorder in Detained Youths: The Predictive Value of Mental Disorders,” Journal of Consulting and
Clinical Psychology 75, no. 2 (2007): 221–31.
7
Bum-Sung Choi et al., “Comorbidities and Correlates of Conduct Disorder among Male Juvenile Detainees in South Korea,” Child and Adolescent Psychiatry and
Mental Health 11 (September 15, 2017): 44.
8
M.R. Rautiainen et al., “Genome-Wide Association Study of Antisocial Personality Disorder,” Translational Psychiatry 6, no. 9 (September 2016): e883.
9
Fazel, Isabel A. Yoon, and Adrian J. Hayes, “Substance Use Disorders in Prisoners: An Updated Systematic Review and Meta-Regression Analysis in Recently
Incarcerated Men and Women,” Addiction 112, no. 10 (October 2017): 1725–39.
10
Matt DeLisi, John Paul Wright, Rafael Mangual,“Psychology, Not Circumstances,” Manhattan Institute (blog), September 19, 2023.
Substance Abuse and Mental Health Services Administration (SAMHSA), “SAMHSA Announces National Survey on Drug Use and Health (NSDUH) Results
11

Detailing Mental Illness and Substance Use Levels in 2021,” HHS.gov, January 4, 2023.
12
Courtney Conn et al., “Borderline Personality Disorder Among Jail Inmates: How Common and How Distinct?” Corrections Compendium 35, no. 4 (2010): 6–13.
See: Prison Policy Initiative, “Prisons of Poverty: Uncovering the Pre-Incarceration Incomes of the Imprisoned,” accessed September 19, 2023, showing that 57% of
13

male prisoners earned less than $22,500 (in 2014 dollars) the year prior to their incarceration—a figure that is nearly double the 2014 federal poverty measure, which
was just $11,600 (see “2014 Poverty Guidelines,” ASPE, accessed September 19, 2023.)

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 5
 then, are whether there exists a basis upon which to suppose (1) that the developmental path of such disorders
begins in early childhood, and (2) that such disorders can be tied to experiences associated with family instability as
a child. The research on behavioral disorders and family suggests both questions can be answered in the affirmative.

Adverse Childhood Experiences, Behavioral Disorders, and Family Structure

When asked for his thoughts about incarceration in the United States, Academy Award-winning actor Denzel
Washington punctuated his answer with five seemingly simple words—words he has uttered on more than one
such occasion: “It starts in the home.”14 Arguing that by “the time the system comes into play, the damage is done,”
Washington noted to his interlocutor that “[police, prosecutors, and judges are] not locking up seven-year-olds.”
Though rarely heard in mainstream venues, Mr. Washington’s argument is not a new one. As suggested above, a
large literature—populated by scholars like Harvard’s Robert Sampson and the late Princeton sociologist Sara
McLanahan15—suggests that family instability is tied to crime.

Single parents tend to experience more stress and have fewer social supports than married parents. Research
consistently shows that single mothers experience more financial and other kinds of life stress, have fewer
opportunities to supervise their children, and tend to be more socially isolated.16 Accordingly, their children are more
likely to be exposed to Adverse Childhood Experiences (ACEs).17 As such, it should come as no surprise that a large
body of research shows that two biological parents are, for the purposes of child development, better than one.18

One caveat to that general finding is that this ceases to be the case when one parent is characterized by antisocial
behavior, in which case the effects of the anti-social parent’s presence on the family’s children are generally worse
than if that parent were absent.19 To be sure, none of this discounts the reality that in many cases, active, pro-social
step-parents, or adopted parents will successfully socialize children and put them on a path to success in later
life. However, the risks of maltreatment and other ACEs tend to be higher for children living with a step-parent,
compared to those being raised by their biological parents.20 These findings suggest the importance of an intact
family for child development insofar as stable, two-parent families reduce the likelihood of ACEs that are associated
with poorer outcomes in later life—including, and especially, criminal involvement.

14
“Denzel: the System Is Rigged but It Starts in the Home,” 2020, via YouTube: https://www.youtube.com/watch?v=rojskHocO7w.
15
Cynthia Harper and Sara McLanahan, “Father absence and youth incarceration,” Journal of Research on Adolescence 14.3 (2004): 369-397.
Marsha Weinraub and Barbara M. Wolf, “Effects of Stress and Social Supports on Mother-Child Interactions in Single- and Two-Parent Families,” Child
16

Development 54, no. 5 (1983): 1297–1311; Ga Eun Kim and Eui-Jung Kim, “Factors Affecting the Quality of Life of Single Mothers Compared to Married
Mothers,” BMC Psychiatry 20, no. 1 (April 15, 2020): 169.
17
Roy Wade Jr et al., “Adverse childhood experiences of low-income urban youth,” Pediatrics 134.1 (2014): e13–e20.
18
W. Bradford Wilcox et al., “Less Poverty, Less Prison, More College: What Two Parents Mean for Black and White Children.” Institute for Family Studies, June
17, 2021, accessed November 6, 2023; David Popenoe, Families without Fathers: Fatherhood, Marriage and Children in American Society, 1st edition (New Brunswick,
N.J: Routledge, 2009); Paul Amato and Alan Booth, A Generation At Risk: Growing Up in an Era of Family Upheaval. (Harvard University Press, 1997).
19
Rafael A. Mangual, Criminal (In)Justice (Center Street, 2022).
See e.g., Marinus H. van IJzendoorn et al., “Elevated Risk of Child Maltreatment in Families With Stepparents but Not With Adoptive Parents,” Child
20

Maltreatment 14, no. 4 (November 1, 2009): 369–75.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 6
 Indeed, the CDC states that ACEs “can have a tremendous impact on future violence victimization and
perpetration”21 (emphasis added). The CDC defines ACEs as “potentially traumatic events that occur in childhood,”
as well as “aspects of the child’s environment that can undermine their sense of safety, stability, and bonding.”22
Examples of ACEs given by the CDC include “experiencing violence, abuse, or neglect,” “witnessing violence in the
home,” and “growing up in a household with… substance use problems.”23

Signs of the more serious personality and substance use disorders prevalent among criminal offenders often become
apparent in the first few years of life.24 The first warning signs of risk for a serious personality disorder can begin
with symptoms of early childhood behavioral disorders such as Oppositional Defiant Disorder (ODD), a condition
whose core symptoms include the refusal to comply with behavioral requests from parents and caregivers, and
defiant reactions to assertions of authority, and Conduct Disorder (CD), a term described in the DSM-5 as a
deviation from behavioral norms characterized by outward aggression and delinquency.

In the Canadian Journal of Psychiatry, scholars from the University of Cambridge identified several risk factors for
the development of CDs. Among them are:

• poor parental supervision

• punitive or erratic parental discipline
• a cold parental attitude
• physical abuse
• parental conflict
• disrupted families
• antisocial parents
• large family size and
• low family income.25

A New Zealand longitudinal birth cohort study published in the Journal of the American Academy of Child &
Adolescent Psychiatry lists “exposure to socioeconomic adversity, parental maladaptive behavior, [and] childhood
exposure to abuse and interparental violence” as factors predictive of both CDs and ODDs.26

21
CDC, “Adverse Childhood Experiences (ACEs),” September 5, 2023.
22
Id.
23
Id.
Terrie E. Moffitt, “Life-Course-Persistent and Adolescence-Limited Antisocial Behavior: A 10-Year Research Review and a Research Agenda,” in Causes of Conduct
24

Disorder and Juvenile Delinquency (New York: The Guilford Press, 2003), 49–75.
25
Joseph Murray and David P. Farrington. “Risk factors for conduct disorder and delinquency: Key findings from longitudinal studies.” The Canadian Journal of
Psychiatry 55.10 (2010): 633–642.
Joseph M. Boden, David M. Fergusson, and L. John Horwood, “Risk Factors for Conduct Disorder and Oppositional/Defiant Disorder: Evidence from a New
26

Zealand Birth Cohort,” Journal of the American Academy of Child & Adolescent Psychiatry 49, no. 11 (November 1, 2010): 1125–33.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 7
 Studies have also shown that, compared to those in intact (especially married) families, children in single-parent
families experience:

• lower parental supervision levels27

• higher parental conflict levels28
• higher risks of low family income and poverty29 and
• higher risks of abuse and neglect.30

In other words, one of the potential mechanisms driving the relationship between family structure and crime is
the fact that children raised by single parents or unstable families are more likely to be exposed to many of the risk
factors that can metastasize into more serious mental disorders. Such disorders, in turn, are strongly associated with
adolescent and adult criminal behavior.

Father Absence, Young Adult Males, and Crime

The previous section of this paper discussed one mechanism that may explain the positive correlation between
single-parent family structure and crime. The psychological and developmental effects of being raised as a young
child in a single-parent or unstable household appear to increase the kinds of mental health challenges that can lead
to criminal activity. This section discusses a second possible explanation for the protective effect of intact families
upon crime: the presence of fathers in the lives of young adult males.

Fathers play a significant role in preventing juvenile delinquency by contributing to the healthy development of
their children, especially sons, as they transition into adulthood. The research on family structure and delinquency
among boys indicates “the effect of intact versus ‘broken’ families is a consistent and real pattern of association [with
the] prevalence of delinquency.”31 In particular, scholarship indicates that both family structure and family conflict
are tied to worse outcomes, insofar as “research on antisocial behavior [among children] consistently illustrates that
adolescents in mother-only households and in conflict-ridden families are more prone to commit delinquent acts.”32

27
Jeffrey Cookston, “Parental Supervision and Family Structure,” Journal of Divorce & Remarriage 32 (September 1, 1999): 107; Susan Harkness, Paul Gregg, and
Mariña Fernández-Salgado, “The Rise in Single-Mother Families and Children’s Cognitive Development: Evidence From Three British Birth Cohorts,” Child
Development 91, no. 5 (2020): 1766; Indiran Rathinabalan and Sridevi Naaraayan, “Effect of Family Factors on Juvenile Delinquency,” International Journal of
Contemporary Pediatrics 4 (October 24, 2017): 2082; Paul R. Amato, “The Impact of Family Formation Change on the Cognitive, Social, and Emotional Well-Being
of the next Generation,” The Future of Children 15, no. 2 (2005): 75–96.
28
Op. Cit., Amato, 83.
29
Op. Cit., Harkness and Fernández-Salgado, 1765.
30
Richard J. Gelles, “Child Abuse and Violence in Single-Parent Families: Parent Absence and Economic Deprivation,” American Journal of Orthopsychiatry 59, no. 4
(1989): 1; Margo I. Wilson, Martin Daly, and Suzanne J. Weghorst, “Household Composition and the Risk of Child Abuse and Neglect,” Journal of Biosocial Science
12, no. 3 ( July 1980): 333; Peter Sidebotham and Jon Heron, “Child Maltreatment in the ‘Children of the Nineties’: A Cohort Study of Risk Factors,” Child Abuse
& Neglect 30, no. 5 (May 1, 2006): 497.
31
L. Edward Wells and Joseph H. Rankin, “Families and Delinquency: A Meta-Analysis of the Impact of Broken Homes,” Social Problems 38, no. 1 (1991): 87.
32
David H. Demo and Alan C. Acock, “The Impact of Divorce on Children,” Journal of Marriage and Family 50, no. 3 (1988): 639.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 8
 Fathers play a significant role in
preventing juvenile delinquency by
contributing to the healthy development
of their children, especially sons, as they
transition into adulthood.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 9
 And it is not just delinquency, but crime itself on the part of young adults, especially young males, that is tied to
family instability. As criminologists Michael Gottfredson and Travis Hirschi have noted,

In most (but not all) studies that directly compare children living with both biological parents with children living in
‘broken’ or reconstituted homes, the children from intact homes have lower rates of crime. The presence of a father in a
stable, married home protects teenage boys and young men from succumbing to the lure of delinquency and crime.33

But how and why are fathers so important for the conduct of their children, especially young boys? The first reason
has to do with role modeling and character development. Fathers are essential role models for their children,
particularly their sons: it is primarily through their fathers, as sociologist David Popenoe has observed, that young
boys see and learn about “male responsibility and achievement, how to be suitably assertive and independent, and
how to relate acceptably to the opposite sex.”34 Moreover, an active and affectionate style of paternal engagement is
linked to greater compassion, self-control, and empathy in sons, all of which deter criminal behavior.35 Without the
care of their father, boys are less likely to develop the virtues that protect them from engaging in delinquency and
crime as young adults.

Second, fathers tend to be more effective at setting limits: they are more likely to get “quick action” from their
children, have a greater tendency to assume the role of “disciplinarian, and are more likely to stress principles like
justice and duty over care and sympathy,” as Popenoe notes.36 Psychologists Kyle and Marsha Kline Pruett take a
similar view, writing that

Fathers tend to be more willing than mothers to confront their children and enforce discipline, leaving their children
with the impression that they in fact have more authority.37

This parenting style is helpful in establishing a climate of order in the family, and minimizing the odds that young
males end up engaging in criminal activity.

Third, the involvement of fathers is linked to positive outcomes in academic and intellectual development. Several
studies have found that father-involvement predicts children’s academic achievement, especially in mathematics
and verbal skills.38 This finding has been established for both sons and daughters but, unsurprisingly, it is especially
pronounced among boys. The presence of married fathers is also protective against school suspensions and expulsions,

33
Michael R. Gottfredson and Travis Hirschi, A General Theory of Crime (Stanford: Stanford University Press, 1990): 103.
34
David Popenoe, Families without Fathers: Fatherhood, Marriage and Children in American Society, 1st edition (New Brunswick, N.J: Routledge, 2009): 142.
35
David Popenoe, Families Without Fathers. p. 149; W. Bradford Wilcox and Kathleen Kovner Kline, eds, Gender and Parenthood: Biological and Social Scientific
Perspectives (Columbia University Press, 2013): 130–132.
36
Ibid. Popenoe, pg. 146.
37
Brad Wilcox, “The Distinct, Positive Impact of a Good Dad,” The Atlantic, June 14, 2013.
38
Op. Cit., Popenoe, 146.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 10
 as well as the risk of dropping out of high school. This is important because young men who do poorly in school or are
suspended and expelled are more likely to end up engaging in criminal behavior and being incarcerated.39

In summary, young males who are raised in an intact family with their father are more likely to steer clear of
delinquency, crime, and incarceration. They are more likely to have the benefit of a father’s attention and affection,
which is associated with better educational performance and behavior, as well as less delinquent activities. All
these factors, in turn, reduce the odds that young males engage in criminal conduct. By contrast, young males
raised in fatherless homes are significantly more likely to engage in criminal activity. For these reasons, we also
expect that communities with large numbers of boys and men raised in single-parent homes experience more
crime and are less safe.

The Current Study: Family Structure and Crime

This report analyzes the relationship between family structure and violent crime across U.S. cities. We perform
our analysis separately at two different levels: by city and by Chicago census tract. For the city level, our crime
data comes from the FBI’s Crimes Known to Law Enforcement tables (published by the FBI’s Uniform Crime
Reporting system), which report the number of crimes by state and city. The variable “Total” is constructed by adding
the table-reported values “Violent Crime” (total number of violent crimes) and “Property Crime” (total number of
property crimes). Crime rates are then calculated by dividing the number of total crimes by the reported population
for each city.40 We do this relying on data from 2015-2019 for our analysis of American cities.

We gather data on family structure, race, poverty, education, age distribution, and size of population for the cities
from the Census Bureau’s American Community Survey (ACS). The ACS is a long-form survey sent out each year
by the Census Bureau to gather more detailed information on communities than can be obtained by the general
census. The Integrated Public Use Microdata Series (IPUMS) from the University of Minnesota gathers this data,
and it is from their collections that we pull our statistics. Annual data is only available for cities with populations
above 65,000, so we restrict our city sample to the 613 cities that meet this criterion.

We construct our measure of family structure based on the fraction of households within a city that are headed
by a single parent. This measure ranges from less than 5% in cities like South Jordan, Utah and Weston, Florida,
to over 75% in cities like Youngstown, Ohio in 2016 and Gary, Indiana in 2017. These cities also differ in other
dimensions, so we also control for the racial mix of the city, poverty rates, education rates, the age distribution,
and the overall population.

Nicholas Zill and W. Bradford Wilcox, “The Black-White Divide in Suspensions: What Is the Role of Family?” Institute for Family Studies, November 19, 2019;
39

Cynthia Harper and Sara McLanahan, “Father absence and youth incarceration.” Journal of Research on Adolescence 14.3 (2004): 369-397.
40
For more information on the dataset, see https://www.fbi.gov/how-we-can-help-you/more-fbi-services-and-information/ucr/publications.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 11
 To understand potential unobserved factors at the within-city level, we also conduct a separate analysis that zooms
in on individual census tracts within Chicago. We examine the relationship between the fraction of single-parent
households and violent crimes occurring in census tracts. Again, our key measure of family structure is the fraction
of households headed by a single parent within the census tract. We include similar controls for the characteristics of
the census tract (race, poverty, education, and age).

For this second analysis, our crime data comes from the Chicago Police Department. The department records the
date, approximate location, and description of the crime and makes this data publicly available a week later.41 Here,
we expand our analysis back to 2005, examining all crimes from 2005-2019, totaling 5,141,260 crimes occurring
over 796 populated Chicago census tracts. Using the Illinois Uniform Crime Reporting (IUCR) code, we determine
if a crime was classified as violent. We then map each crime from the given longitude and latitude into a Chicago
census tract, after which we once again use the ACS to gather the same family structure and control variables for the
Chicago census tracts. However, results by census tract are only available in 5-year aggregations, so we use 2005-
2009, 2010-2014, and 2015-2019 variable levels for the Chicago analysis.

In Table 1, we provide some descriptive information about our two datasets. In each case, we split the observations
based on whether they are in a location that is above or below the sample median fraction of households headed
by a single parent. When we compare total crime rates of cities above or below the median fraction of single-
parent households, we find that the total crime rates in the cities above the median are about 48% higher than for
cities below the median. When we focus specifically on violent crimes or homicides, we find the difference is even
larger, with a 118% increase in the rate of violent crimes and a 255% increase in the rate of homicides in cities that
have higher levels of single parenthood. When we just look at the census tracts within Chicago, we see that these
differences are even larger still, with a 137% increase in the total crime rate, a 226% increase in the violent crime rate,
and a 436% increase in the homicide rate in communities with above the median rates of single parenthood.

The results in Table 1 are all raw differences in crime rates, so they don’t consider any of the other differences in
covariates between these cities and census tracts. The other statistics reported in Table 1 indicate that cities and census
tracts with a higher fraction of single-parent households also tend to have a higher fraction of African American
individuals, greater rates of poverty, and lower levels of education. In the regression analysis that follows, we control for
these additional covariates to focus in on the independent effect of family structure (net of these other controls).

41
See: https://data.cityofchicago.org/Public-Safety/Crimes-2001-to-Present/ijzp-q8t2

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 12
 Table 1. Descriptive Statistics comparing above/below median cities and census
tracts in terms of the fraction of households headed by single parents.
CITY LOW CITY HIGH CHICAGO LOW CHICAGO HIGH

TOTAL CRIMES (/1000 POP) 28.46 42.05 85.45 202.55

(13.28) (15.03) (118.80) (135.40)

VIOLENT CRIMES (/1000 POP) 3.14 6.83 25.79 84.07

(2.08) (3.64) (26.76) (52.20)

HOMICIDES (/1000 POP) 0.03 0.10 1.56 8.37

(0.03) (0.10) (1.93) (6.14)

AREA OF CENSUS TRACT (ACRES) 189.70 182.02

(244.50) (267.00)

BLACK/AFRICAN AMERICAN 0.07 0.24 0.082 0.66

(0.07) (0.18) (0.134) (0.38)

SINGLE PARENTS 0.18 0.38 0.23 0.61

(0.05) (0.10) (0.09) (0.16)

POVERTY 0.11 0.20 0.14 0.31

(0.05) (0.07) (0.09) (0.13)

HIGH SCHOOL GRADUATE 0.94 0.91 0.86 0.77

(0.03) (0.03) (0.13) (0.12)

COLLEGE DEGREE 0.52 0.24

(0.25) (0.14)

YOUNG MEN AGE 15-29 0.14 0.15 0.12 0.12

(0.04) (0.03) (0.05) (0.04)

UNMARRIED YOUNG MEN AGE 15-29 0.12 0.13 0.13 0.14

(0.03) (0.03) (0.06) (0.049)

POPULATION (THOUSANDS) 175.23 259.60 3.82 3.10

(279.80) (640.10) (1.89) (1.67)

OBSERVATIONS 1365 1364 5970 5970

Note: mean coefficients displayed; s.d. in parentheses. The variables Total Crimes, Violent Crimes, and Homicides
are all the number of crimes per 1,000 people. “Low” represents having a single-parent proportion below the 50th
percentile. “High” represents having a single-parent proportion above the 50th percentile. Institute for Family Studies

Table 2 provides regression-based estimates of the effect of the fraction of households headed by single parents on
our three measures of violent crime for U.S. Cities. We standardize our outcome measure of crime per capita to
have a mean of zero and a standard deviation of one for all three measures of violent crime. Therefore, the regression
coefficients estimate the change in units of standard deviations that would result from a 100% increase in the
proportion of single parents within a city/census tract.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 13
 Table 2. Regression-based estimates of the relationship between
single-parent households and crime in U.S. Cities.
TOTAL CRIME VIOLENT CRIME HOMICIDES

SINGLE PARENTS 0.05** 0.85** 0.31

(0.02) (0.26) (0.21)

BLACK/AFRICAN AMERICAN 0.03* 0.86*** 1.69***

(0.01) (0.19) (0.22)

POVERTY 0.26*** 2.76*** 1.91***

(0.05) (0.67) (0.53)

HIGH SCHOOL GRADUATE 0.13** -1.22* -0.94*

(0.04) (0.52) (0.41)

YOUNG MEN AGE 15-29 0.61* 1.21 0.90

(0.25) (1.80) (1.08)

SINGLE YOUNG MEN AGE 15-29 -0.81** -2.64 -2.95*

(0.27) (2.00) (1.29)

POPULATION (THOUSANDS) 0.00 0.00 0.00

(0.00) (0.00) (0.00)

N 2230 2298 2326

Note: Standard errors in parentheses. Variables Single Parents, Black/African American, High School Graduate,
and Poverty all are the proportion of the city that falls into that category. The variables Total Crime, Violent
Crime, and Homicides are all standardized values with a value of 0 representing a city with the average number of
crimes per capita. Standard errors are clustered at the city level.
* p < 0.05, ** p < 0.01, *** p < 0.001
Institute for Family Studies

In Table 2, we find that a 100% increase in the proportion of single parents predicts that the total crime rate
increases by 0.05 standard deviations. More intuitively, a 10% increase in the proportion of single parents results in
a 0.005 standard deviation increase in total crime, and a 0.085 standard deviation increase in violent crime. These
results are statistically significant with a p-value <0.01 or a confidence level of 99 percent. The estimated effect of
our family structure measure on homicides is not statistically significant. The single-parent proportion of U.S. Cities
from 2015 to 2019 ranges from 2% to 80 percent. If we look at the predicted effect of moving from the lowest
proportion to the highest in cities across the United States, we expect to see a 0.039 standard deviation increase in
total crime and a 0.663 standard deviation increase in violent crime.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 14
 Table 3 provides regression-based estimates of the effect of the fraction of households headed by single parents
on our three measures of violent crime for Chicago Census Tracts. We standardize these measures following the
methods used in Table 2. We find that a 10% increase in the proportion of single parents results in a 0.046 standard
deviation increase in violent crime, and a 0.043 standard deviation increase in homicides. There is no statistically
significant increase in total crime for Chicago Census tracts associated with single parenthood; however, the effect
on violent crime and homicides is significant at the 99% confidence level. The single-parent proportion of Chicago
Census tracts from 2005-2019 ranges from 0 to 1. Moving from the tract with the lowest proportion of single
parents to the tract with the highest would, therefore, predict an increase in violent crime of 0.46 standard deviations
and an increase in homicides by 0.43 standard deviations.

We keep the U.S. Cities and Chicago Census tract regressions comparable by including similar controls in each.
Both regressions control for the percent of the population that is Black/African American, the poverty and high
school graduation rates within a unit, the proportion of young men and single young men ages 15-29, and the total

Table 3. Regression-based estimates of the relationship between single-parent

households and crime within census tracts in Chicago.
TOTAL CRIME VIOLENT CRIME HOMICIDES

SINGLE PARENTS 0.20 0.46*** 0.43***

(0.13) (0.09) (0.08)

BLACK/AFRICAN AMERICAN 1.03*** 1.19*** 1.35***

(0.05) (0.04) (0.03)

HIGH SCHOOL GRADUATE -0.75*** -0.71*** -0.77***

(0.18) (0.11) (0.09)

COLLEGE DEGREE 0.36** 0.06 -0.16**

(0.12) (0.07) (0.05)

POVERTY 1.10*** 1.26*** 0.64***

(0.11) (0.09) (0.09)

YOUNG MEN AGE 15-29 2.79*** 1.28* 0.46

(0.64) (0.52) (0.46)

SINGLE YOUNG MEN AGE 15-29 -2.48*** -1.44** -0.56

(0.48) (0.47) (0.44)

POPULATION (THOUSANDS) -0.08*** -0.08*** -0.08***

(0.01) (0.00) (0.00)

POPULATION DENSITY -2.32*** -1.73*** -1.11***

(0.40) (0.29) (0.20)

N 11940 11940 11940

Note: Standard errors in parentheses. Variables Single Parents, Black/African American, High School Graduate,
College Degree, Poverty, and Unmarried Young Men all are the proportion of the census tract that falls into that
category. The variables Total Crime, Violent Crime, and Homicides are all standardized values with a value of 0
representing a census tract with the average number of crimes per capita. Population Density is the number of
thousands of residents per acre in the census tract.* p < 0.05, ** p < 0.01, *** p < 0.001 Institute for Family Studies

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 15
 population in thousands. The Chicago Census tract regression contains two extra controls: the proportion of the
population within the tract with college degrees and the population density within the tract. Both sets of regressions
(city- and tract-based) standardize all three violent crime outcome measures as previously mentioned and define the
single-parent proportion as the number of single mothers divided by the total number of mothers.

The regression-based estimates that we have used so far in the report are all based on the assumption that there is a
linear relationship between the fraction of single-parent households and crime. In the next set of figures, we explore
this assumption by plotting the non-linear relationship between single-parent households and crime rates.
We create these figures by generating a two-way scatter plot of crime category and single-parent proportion and
using the two-way fractional-polynomial prediction plots command in Stata to overlay a line of best fit based on
a fractional-polynomial prediction plot of predicted y on x using regression. The figure entitled “Family Structure
on Total Crime for Chicago Census Tracts” uses a quadratic model for the line of best fit, and others incorporate

Family Structure on Total Crime Family Structure on Total Crime

For Chicago Census Tracts For U.S. Cities in 2015-2019
1000 150
Total Crimes Per 1,000 Residents

800
Total Crimes Per 1,000

100
600

400
50
200

0 0
0 20 40 60 80 100 0 20 40 60 80
Single-Parent Households (%) Single-Parent Households (%)

Family Structure on Violent Crime Family Structure on Violent Crime

For Chicago Census Tracts For U.S. Cities in 2015-2019

400 25
Violent Crimes Per 1,000 Residents

Violent Crimes Per 1,000

20
300
15
200
10

100 5

0 0
0 20 40 60 80 100 0 20 40 60 80

Single-Parent Households (%) Single-Parent Households (%)

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 16
 Family Structure on Homicides Family Structure on Homicides
For Chicago Census Tracts For U.S. Cities in 2015-2019
40 1
Homicides Per 1,000 Residents

30 .8

Homicides Per 1,000

.6
20
.4
10
.2

0 0
0 20 40 60 80 100 0 20 40 60 80

Single-Parent Households (%) Single-Parent Households (%)

higher-level polynomials to best fit the data. The figures depicting data on violent crime and homicides show that
the greatest increase in crime is associated with moving across Chicago census tracts with 40-80% of single parents
or across U.S. cities with 20-60% of single parents.

Finally, the estimates in this report have centered on the cross-sectional relationship between family structure and
crime using data from 2005-2019. While we have included some of the most important covariates in our analysis
(race, poverty, education, age), there are likely other factors that we are not including. Though the within-city analysis
that we do for Chicago is helpful in controlling for any factors that might vary at the city-level, there are likely still
unobserved variables at the census tract level that we would control for if possible. We also acknowledge that efforts
to reduce crime may be disproportionally targeting neighborhoods where both crime and family instability are high.

Ideally, we would identify a natural experiment or instrumental variable that influences family structure while having
no direct impact on crime. Other scholars have been able to identify other factors for which there is quasi-random
variation to examine other factors that impact crime rates, such as lotteries for housing vouchers or unexpected
changes in local police policies. There are several reasons why such quasi-random variation is difficult to find in
terms of family structure.

One strategy that has been used is to estimate a fixed effects model in which we examine how changes in family
structure influence changes in crime. We include city or census-tract fixed effects to control for any factors that are
fixed across time for the same location.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 17
 Tables 4 and 5 show our regression coefficients when we include city or census-tract fixed effects. Our regression
coefficients for Chicago Census tracts and U.S. cities become statistically insignificant when these fixed effects are
included in the regression. This may indicate that differences in family composition and crime are not drastic enough
across cities/census tracts within any given year to produce a significant effect. It is also worth noting that increases
in family instability plateaued across much of the nation in the 1990s. This means there may not have been enough
variation in family structure in the early part of the 21st century to drive additional increases in crime in urban
America. One of the challenges of using fixed effects in this type of analysis is that it requires that strong assumptions
be made about the timing of when family structure influences crime rates. The way that we include fixed effects in this
model would only capture effects in which changes in family structure in the year of analysis affect crime rates in the
same year. But it is likely that the influence of family structure accumulates over the lifetime of a youth or young adult.
This could explain why our fixed effects models did not find significant effects for family structure.

Table 4. Fixed Effects Regression-based estimates of the relationship between

single-parent households and crime within census tracts in Chicago (2005-2019).
TOTAL CRIME VIOLENT CRIME HOMICIDES

SINGLE PARENTS -0.11 -0.14 -0.01

(0.18) (0.13) (0.12)

BLACK/AFRICAN AMERICAN 0.60 0.66* 0.24

(0.46) (0.30) (0.21)

HIGH SCHOOL GRADUATE -0.65 -0.66* -0.26

(0.52) (0.31) (0.24)

COLLEGE DEGREE -0.48 -0.08 -0.31

(0.75) (0.42) (0.18)

POVERTY -0.47 -0.31 -0.44*

(0.30) (0.23) (0.20)

YOUNG MEN AGE 15-29 -3.11 -1.05 0.67

(1.97) (1.22) (0.79)

SINGLE YOUNG MEN AGE 15-29 1.41 -0.27 -1.50*

(1.49) (1.02) (0.73)

POPULATION DENSITY -5.10 -4.14 -4.42

(5.33) (4.21) (3.73)

N 11940 11940 11940

Note: Standard errors in parentheses. Variables Single Parents, Black/African American, High School Graduate,
College Degree, and Poverty all are the proportion of the census tract that falls into that category. The variables
Total Crime, Violent Crime, and Homicides are all standardized values with a value of 0 representing a census
tract with the average number of crimes per capita. Population Density is the number of thousands of residents per
acre in the census tract. Standard errors are clustered at the census tract level.
* p < 0.05, ** p < 0.01, *** p < 0.001 Institute for Family Studies

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 18
 Table 5. Fixed Effects Regression-based estimates of the relationship between
single-parent households and crime in U.S. Cities (2015-2019).
TOTAL CRIME VIOLENT CRIME HOMICIDES

SINGLE PARENT 0.004 0.045 -0.156

(0.004) (0.065) (0.089)

BLACK/AFRICAN AMERICAN -0.017 0.148 -0.434

(0.014) (0.203) (0.299)

POVERTY 0.000 -0.000 0.000

(0.000) (0.000) (.)

HIGH SCHOOL GRADUATE -0.004 0.169 0.096

(0.018) (0.242) (0.371)

YOUNG MEN AGE 15-29 0.000 -0.000 0.000

(0.000) (0.000) (.)

SINGLE YOUNG MEN AGE 15-29 -0.000 0.000 -0.000

(0.000) (0.000) (.)

POPULATION (THOUSANDS) -0.000* -0.001 -0.001

(0.000) (0.000) (0.000)

N 2230 2298 2326

Note: Standard errors in parentheses. Variables Single Parents, Black/African American, High School Graduate,
and Poverty all are the proportion of the city that falls into that category. The variables Total Crime, Violent
Crime, and Homicides are all standardized values with a value of 0 representing a city with the average number of
crimes per capita. Standard errors are clustered at the city level.
* p < 0.05, ** p < 0.01, *** p < 0.001 Institute for Family Studies

Conclusion

This report finds that the streets are safer and violent crime is lower in cities across America where the family is
stronger. Our analysis, in keeping with earlier research, finds that violent crime rates are generally lower in cities with
more two-parent families; the same basic story applies to Chicago neighborhoods. Our analysis specifically shows
that total crime and violent crime rates are higher in cities with fewer two-parent
families, even net of socioeconomic controls. And in Chicago, violent crime and
homicide rates are higher in neighborhoods with fewer two-parent families, even The streets are safer
net of controls. Unsurprisingly, the health of our most basic institution—the
family—is intimately related to the safety of our cities.
and violent crime is
lower in cities across
But, in keeping with previous research, our analysis also finds that single
America where the
parenthood is not always, nor the only, predictor of crime rates. Factors like
poverty and race also are linked to violent crime rates in cities across the nation family is stronger.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 19
 and in ways consistent with the “structural perspective” on crime. Likewise, recent increases in crime, which have
occurred without marked changes in family life, suggest that changes in law-enforcement and the prosecution of
criminals have also had a hand in the recent uptick in violent crime in American cities. These changes are consistent
with classic theories that stress the value of smart policing and prosecution when it comes to reducing crime. So,
family appears to be one—but not the only—factor associated with violent crime. Efforts to reduce crime will have
to strengthen neighborhood institutions, improve schools, and bolster the effectiveness of local law enforcement.

However, it is undeniable that family instability is significantly predictive of violent crime, even in the presence of
control variables on both the local and national level. Given our findings, we think that shifts from the 1960s to the
1990s away from stable families have left some cities, and especially some neighborhoods, vulnerable to higher rates
of crime, especially violent crime. Serious attempts to make our streets safer must also grapple with the challenge of
strengthening and stabilizing urban family life—and reducing the odds that young males grow up without a father
in the home, and in neighborhoods in which prosocial fathers are less prevalent. We need to realign material and
cultural incentives in our cities to favor marriage and stable families, not undercut them, especially in vulnerable
neighborhoods where crime is common and stable families scarce.

A comprehensive agenda to strengthen family life is beyond the scope of this IFS report. Nevertheless, we suggest
the following targeted policy recommendations:

• First, advertise and advocate the “success sequence”—the idea that a high school education, a full-time
job, and marriage should precede parenthood—in schools and social media across America’s cities.42
• Second, eliminate marriage penalties in federal means-tested programs, like Medicaid, that discourage
marriage among lower-income families.43
• Third, steer more young adults, especially young men, who are not on the college track toward
high-quality, vocational and apprenticeship programs that boost their employment, income, and
marriageability.44

Measures like these would help to strengthen and stabilize marriage and family life in city neighborhoods where too
few young adults are on track to forge stable, married families. This, in turn, would decrease the odds that young men
in our nation’s cities experience the kind of family instability that can make them more vulnerable to negative peer
influences, delinquency, and violent crime. Thus, all of us who are dedicated to making our city streets safer must also
focus on making our families stronger.

42
Ron Haskins and Isabel V. Sawhill, “Creating an Opportunity Society” (Washington, DC: Brookings Institution, 2009); Wendy Wang and W. Bradford Wilcox,
“The Millennial Success Sequence: Marriage, Kids, and the ‘Success Sequence’ among Young Adults,” American Enterprise Institute, June 14, 2017; Nat Malkus,
“Uncommonly Popular: Public Support for Teaching the Success Sequence in School,” American Enterprise Institute, December 2021.
43
W. Bradford Wilcox, Joseph P. Price, and Angela Rachidi, “Marriage, Penalized,” American Enterprise Institute, July 26, 2016.
44
James J. Kemple, “Career Academies: Long- Term Impacts on Work, Education, and Transitions to Adulthood,” MDRC, June 2008.

Stronger Families, Safer Streets: Exploring Links Between Family Structure and Crime 20
About the Authors
Rafael Mangual is the Nick Ohnell fellow at the Manhattan Institute for Policy Research, a member of
the Council on Criminal Justice, and author of Criminal (In)Justice: What the Push For Decarceration And
Depolicing Gets Wrong And Who It Hurts Most. Rafael has, on multiple occasions, testified before committees
of the U.S. House Of Representatives, U.S. Senate, and U.S. Commission on Civil Rights. His work has
appeared in The New York Times, Wall Street Journal, Washington Post, and City Journal, where he is also a
Contributing Editor.

Brad Wilcox is the Future of Freedom fellow at the Institute for Family Studies, visiting scholar at the
American Enterprise Institute, and the director of the National Marriage Project at the University of
Virginia. Wilcox has published widely on marriage, cohabitation, fatherhood, and the welfare of children.
He is the author of the forthcoming book, Get Married: Why Americans Should Defy the Elites, Forge Strong
Families, and Save Civilization.

Seth Cannon is an undergraduate student studying economics at Brigham Young University and a
Wheatley Scholar.

Joseph Price is a Professor of Economics at Brigham Young University. He is a Faculty Research Fellow at
the National Bureau of Economic Research, a Research Fellow at the Institute of Labor Economics, a Senior
Fellow at the Austin Institute, and the Director of the BYU Record Linking Lab. He has published over 50
articles including articles in the Quarterly Journal of Economics, Journal of Labor Economics, Journal of Human
Resources, Journal of Health Economics, Demography, and Management Science.

Acknowledgements
We thank Wendy Wang for substantive advice and suggestions, Alysse ElHage for editing this report, Michael
Toscano for administrative direction, and Brandon Wooten for design and graphics. This IFS report was
produced with the cooperation of the Manhattan Institute.

© Copyright 2023 Institute for Family Studies. All rights reserved.