High rates of crime and incarceration impose tremendous costs on society, with lasting negative effects on individuals, families, and communities. Rates of crime in the United States have been falling steadily, but still constitute a serious economic and social challenge. At the same time, the incarceration rate in the United States is so high—more than 700 out of every 100,000 people are incarcerated—that both crime scholars and policymakers alike question whether, for nonviolent criminals in particular, the social costs of incarceration exceed the social benefits.

While there is significant policy focus on America’s incarceration policies, it is important to consider that crime continues to be a focus of concern for policymakers, particularly at the state and local levels. Public spending on fighting crime, including the costs of incarceration, policing, and judicial and legal services, is substantial; in addition, private spending by households and businesses is also substantial. There are also tremendous costs to the victims of crime, such as medical costs, lost earnings, and an overall loss in quality of life. Crime also stymies economic growth. For example, exposure to violence can inhibit effective schooling and other developmental outcomes (Burdick-Will 2013; Sharkey et al. 2012). Crime can induce citizens to migrate; economists estimate that each nonfatal violent crime reduces a city’s population by approximately one person, and each homicide reduces a city’s population by seventy persons (Cullen and Levitt 1999; Ludwig and Cook 2000, 104). To the extent that migration diminishes a locality’s tax and consumer base, departures threaten a city’s ability to effectively educate children, provide social services, and maintain a vibrant local economy.

The good news is that crime rates in the United States have been falling steadily since the 1990s, reversing an upward trend from the 1960s through the 1980s. There does not appear to be a consensus among scholars about how to account for the overall sharp decline, but contributing factors include increased policing, rising incarceration rates, and the waning of the crack epidemic that was prevalent in the 1980s and early 1990s.

Despite the ongoing decline in crime, incarceration rates in the United States remain at historically unprecedented levels. These high incarceration rates can have profound effects on disadvantaged communities in particular. Research has shown that incarceration may impede employment and marriage prospects among former inmates, increase poverty depth and behavioral problems among their children, and amplify the spread of communicable diseases among disproportionately impacted communities (Raphael 2007). In particular, these effects are especially harmful for particular demographic groups—often young minority males—who struggle to gain employment or lead productive lives post-incarceration. In addition, high rates of incarceration are expensive for both federal and state governments. On average, in 2012 it cost more than $29,000 each year to house an inmate in federal prison (U.S. Congressional Research Service 2013). In total, the United States spent over $80 billion on corrections expenditures in 2010, with more than 90 percent of these expenditures at the state and local levels (Kyckelhahn and Martin 2013).

A founding principle of The Hamilton Project’s economic strategy is that long-term prosperity is best achieved by fostering economic growth and broad participation in that growth. Elevated rates of crime and incarceration directly work against these principles, marginalizing individuals, devastating affected communities, and perpetuating inequality. In this spirit, we offer “Ten Economic Facts about Crime and Incarceration in the United States” to bring attention to recent trends in crime and incarceration, the characteristics of those who commit crimes and those who are incarcerated, and the social and economic costs of current policy.

Chapter 1 describes recent crime trends in the United States and the characteristics of criminal offenders and victims. Chapter 2 focuses on the growth of mass incarceration in America. Chapter 3 presents evidence on the economic and social costs of current crime and incarceration policy.

After a significant explosion in crime rates between the 1960s and the 1980s, the United States has experienced a steady decline in crime rates over the past twenty-five years. As illustrated in figure 1, crime rates fell nearly 30 percent between 1991 and 2001, and have subsequently fallen an additional 22 percent between 2001 and 2012. This measure, calculated by the FBI, incorporates both violent crimes (e.g., murder and aggravated assault) and property crimes (e.g., burglary and larceny-theft). Rates of property and violent crime have followed similar trends, falling 29 percent and 33 percent, respectively, between 1991 and 2001 (DOJ 2010b).

Social scientists have struggled to provide adequate explanations for the sharp and persistent decline in crime rates. Economists have focused on a few potential factors, including an increased number of police on the streets, rising rates of incarceration, and the waning of the crack epidemic, to explain the drop in crime (Levitt 2004). In the 1990s police officers per capita increased by approximately 14 percent. During this same decade, sentencing policies grew stricter and the U.S. prison population swelled, which had both deterrence (i.e., prevention of further crime by increasing the threat of punishment) and incapacitation (i.e., the inability to commit a crime because of being imprisoned) effects on criminals (Abrams 2011; Johnson and Raphael 2012; Levitt 2004). The waning of the crack epidemic reduced crime primarily through a decline in the homicide rates associated with crack markets in the late 1980s.

However, the prolonged decline in total crime rates masks racial and socioeconomic differences. For example, in 2010 the homicide rates for African American men ages fifteen to twenty-four was eighteen times that for white men in that cohort (Heller et al. 2013). In general, racial and socioeconomic discrepancies have marked impacts on inequality in well-being.

The blue and green dotted lines in figure 1, corresponding to the right axis, show the cumulative share of families with income under various thresholds. Around 40 percent of families earn $40,000 or less each year, 54 percent of families earn $60,000 or less (demonstrated by the black dotted line), and 76 percent of working-age families earn $100,000 or less. For working-age families with children, earning over $100,000 is the exception, not the rule.

The vertical bars in figure 1, corresponding to the left axis, show the percent of families that fall within various income ranges. About 15 percent of working-age families (or approximately 5.6 million families) earn between $1 and $20,000 a year, while 19 percent of families (approximately 7.1 million families) earn between $20,001 and $40,000. On the opposite end of the distribution, fewer than 3 percent of families earn $260,001 or more.

Figure 1. Crime Rate in the United States, 1960–2012
After being particularly elevated during the 1970s and 1980s, crime rates have fallen nearly 45 percent between 1990 and 2012.

Source: DOJ 2010b; authors' calculations.
Note: The crime rate includes all violent crimes (i.e., aggravated assault, forcible rape, murder, and robbery) and property crimes (i.e., burglary, larceny-theft, and motor vehicle theft).

Across all types of personal crimes, victimization rates are significantly higher for individuals living in low-income households, as shown in figure 2. In 2008, the latest year for which data are available, the victimization rate for all personal crimes among individuals with family incomes of less than $15,000 was over three times the rate of those with family incomes of $75,000 or more (DOJ 2010a). The most prevalent crime for low-income victims was assault, followed closely by acts of attempted violence, at thirty-three victims and twenty-eight victims per 100,000 residents, respectively. For those in the higher-income bracket, these rates were significantly lower at only eleven victims and nine victims per 100,000 residents, respectively.

Because crime tends to concentrate in disadvantaged areas, low-income individuals are more likely than higher-income individuals to be victims. Neighborhood poverty, a lack of residential mobility, and unstable family structure can increase the likelihood of personal victimization. Notably, evidence from the Moving to Opportunity program—a multiyear federal research demonstration project that combined rental assistance with housing counseling to help families with very low incomes to move from areas with a high concentration of poverty—suggests that moving into a less-poor neighborhood significantly reduces child criminal victimization rates. In particular, children of families that moved as a result of receiving both a housing voucher to move to a new location and counseling assistance experienced personal crime victimization rates that were 13 percentage points lower than those who did not receive any voucher or assistance (Katz, Kling, and Liebman 2000).

Victims of personal crimes face tangible costs, including medical costs, lost earnings, and victim assistance programs, as well as intangible costs, such as pain, suffering, and lost quality of life (Miller, Cohen, and Wiersama 1996). There are also public health consequences to crime victimization. Since homicide rates are so high for young African American men, men in this demographic group lose more years of life before age sixty-five to homicide than they do to heart disease, which is the nation's overall leading killer (Heller et al. 2013).

Figure 2. Victimization Rates for Persons Age 12 or Older, by Type of Crime and Annual Family Income, 2008
In 2008 individuals with annual family incomes of less than $15,000 were at least three times more likely to be victims of personal crimes—such as rape and assault—than were individuals with annual family incomes of $75,000 or more.

Source: DOJ 2010a; authors' calculations. Note: The victimization rate is defined as the number of individuals who were victims of crime over a six-month period for every 100,000 U.S. residents.

Juveniles make up a significant portion of offenders each year. More than one quarter (27 percent) of known offenders—defined as individuals involved in a crime incident that had at least one identifiable characteristic, whether or not an arrest has been made—were individuals ages eleven to twenty, and an additional 34 percent were ages twenty-one to thirty; all other individuals composed fewer than 40 percent of offenders. As seen in figure 3, this trend holds for all types of crimes. More specifically, 55 percent of offenders committing crimes against persons, such as assault and sex offenses, were ages eleven to thirty. For crimes against property (such as larceny-theft and vandalism offenses) and crimes against society (including drug offenses and weapon law violations), 63 percent and 66 percent of offenders, respectively, were individuals in the eleven-to-thirty age group.

A stark difference in the number of offenders by gender is also evident. Most crimes—whether against persons, property, or society—are committed by men; of criminal offenders with known gender, 72 percent were male. This trend for gender follows for crimes against persons (73 percent), crimes against property (70 percent), and crimes against society (77 percent) (DOJ 2012). Combined, these facts indicate that most offenders in the United States are young men.

Some social scientists explain this age profile of crime by appealing to a biological perspective on criminal behavior, focusing on the impaired decision-making capabilities of the adolescent brain in particular. There are also numerous social theories, including those that emphasize youth susceptibility to societal pressures, or their concern with identity formation, establishing their independence, and peer reactions (O'Donoghue and Rabin 2001).

Figure 3. Number of Offenders in the United States, by Age and Offense Category, 2012
More than 60 percent of known criminal offenders are under the age of thirty, with individuals eleven to twenty years old constituting roughly 27 percent of this group, and individuals twenty-one to thirty years old making up an additional 34 percent.

Source: DOJ 2012; authors' calculations.
Note: The FBI defines crimes against persons as crimes whose victims are always individuals. Crimes against property are those with the goal of obtaining money, property, or some other benefit. Crimes against society are those that represent society's prohibition against engaging in certain types of activity, and are typically victimless crimes (DOJ 2011). Offender data include characteristics of each offender involved in a crime incident whether or not an arrest has been made; offenders with unknown ages are excluded from the analysis. Additionally, incidents with unknown offenders—1,741,162 incidents in 2012—are excluded. For more details, see the technical appendix.

Youths from low-income families (income at or below 200 percent of the federal poverty level) are equally likely to commit drug-related offenses than are their higher-income counterparts. As seen in figure 4, low-income youths are more likely to use marijuana by age sixteen, and to use other drugs or sell drugs by age eighteen. In contrast, low-income youths are more likely to engage in violent and property crimes than are youths from middle- and high-income families. In particular, low-income youths are significantly more likely to attack someone or get into a fight, join a gang, and steal something worth more than $50. In other words, youths from low-income families are more likely to engage in crimes that involve or affect other people than are youths from higher-income families.

A standard economics explanation for this socioeconomic profile of property crime is that for poor youths the attractiveness of alternatives to crime is low: if employment opportunities are limited for teens living in poor neighborhoods, then property crime becomes relatively more attractive. The heightened likelihood of violent crime among poor youths raises the issue of reactive behaviors—in other words, youths responding to perceived threats—which has become the focus of recent research in this field. However, the similar rates of drug use across teens from different income groups is consistent with a more general model of risky teenage activity associated with the so-called impaired decision-making capabilities of the adolescent brain.

Some intriguing recent academic work has proposed that adverse youth outcomes are often the result of quick errors in judgment and decision-making. In particular, hostile attribution bias—hypervigilance to threat cues and the tendency to overattribute malevolent intent to others—appears to be more common among disadvantaged youths, partly because these youths grow up with a heightened risk of having experienced abuse (Dodge, Bates, and Pettit 1990; Heller et al. 2013). Some experts have consequently begun promoting cognitive behavioral therapy for these youths to help them recognize and rewire the automatic behaviors and biased beliefs that often result in judgment and decision-making errors. Promising results from several experiments in Chicago—in particular, improved schooling outcomes and fewer arrests for violent crimes—suggest that it is possible to change the outcomes of disadvantaged youths simply by helping them recognize when automatic judgments and decision-making may trigger the wrong outcomes (Heller et al. 2013).

Figure 4. Adolescent Risk Behaviors by Family Income Level
Although youths from low-income families are as likely to use or sell drugs as are their higher-income counterparts, the former are significantly more likely to engage in criminal activities that involve or affect other people.

Source: Kent 2009.
Note: Original data are derived from the 1997 National Longitudinal Survey of Youth. Adolescent risk behaviors are measured up to age eighteen, except for marijuana usage, which is measured up to age sixteen. Low-income families are those whose incomes are at or below 200 percent of the FPL. Middle-income families have incomes between 201 and 400 percent of the FPL. High-income families have incomes at or above 401 percent of the FPL.

Incarceration rates—defined as the number of inmates in local jails, state prisons, federal prisons, and privately operated facilities per every 100,000 U.S. residents—increased during the past three decades, from 220 in 1980 to 756 in 2008, before retreating slightly to 710 in 2012 (as seen in figure 5). Incarceration rates are driven by three factors: crime rates, the number of prison sentences per number of crimes committed, and expected time served in prison among those sentenced (Raphael 2011). Academic evidence suggests that increases in crime cannot explain the growth in incarceration rates since the 1980s (Raphael and Stoll 2013). However, the likelihood that an arrested offender will be sent to prison, as well as the time prisoners can expect to serve, has increased for all types of crime (Raphael and Stoll 2009 2013). Given that both the likelihood of going to prison and sentence lengths are heavily influenced by adjudication outcomes and the types of punishment levied, most of the growth in incarceration rates can be attributed to changes in policy (Raphael and Stoll 2013).

Policymakers at the federal and state levels have created a stricter criminal justice system in the past three decades. For example, state and federal laws—such as the Sentencing Reform Act of 1984—established greater structure in sentencing through specified guidelines for each offense. Between 1975 and 2002 all fifty states adopted some form of mandatory-sentencing law specifying minimum prison sentences for specific offenses. In fact, nearly three quarters of states and the federal government—through laws like the Anti-Drug Abuse Act of 1986—enacted mandatory-sentencing laws for possession or trafficking of illegal drugs. Many states also adopted repeat offender laws, known as "three strikes" laws, that enhance the sentences of those with prior felony convictions. These policies, among others, are believed to have made the United States tougher on those who commit crime, raising incarceration rates through increased admissions and longer sentences (Raphael and Stoll 2013).

The continued growth in the federal prison population contrasts with trends in state prison populations. Between 2008 and 2012 the number of inmates in state correctional facilities decreased by approximately 4 percent (from roughly 1.41 million to 1.35 million), while the number of inmates in federal prisons increased by more than 8 percent (from approximately 201,000 to nearly 218,000) (Carson and Golinelli 2013). This increase in federal imprisonment rates has been driven by increases in immigration-related charges. Between 2003 and 2011 admissions to federal prisons for immigration-related offenses increased by 83 percent, rising from 13,100 to 23,939 (DOJ n.d.).

Figure 5. Incarceration Rate in the United States, 1960–2012
Federal policies, such as the Sentencing Reform Act, and state policies, such as the "three strikes" legislation, were major contributing factors to the 222 percent increase in the incarceration rate between 1980 and 2012.

Source: Austin et al. 2000; Cahalan 1986; Carson forthcoming; Census Bureau 2001; Glaze 2010, 2011; Glaze and Herberman 2013; Raphael and Stoll 2013; Sabol, Couture, and Harrison 2007; Sabol, West, and Cooper 2010; authors' calculations. Note: Incarceration rate refers to the total number of inmates in custody of local jails, state or federal prisons, or privately operated facilities within the year per 100,000 U.S. residents. The three events highlighted in this figure are examples of the many policy changes that are believed to have influenced incarceration rates since the 1980s. For more details, see the technical appendix.

The United States is an international outlier when it comes to incarceration rates. In 2012 the incarceration rate—which includes inmates in the custody of local jails, state or federal prisons, or privately operated facilities—in the United States was 710 per every 100,000 U.S. residents (Glaze and Herberman 2013). This puts the U.S. incarceration rate at more than five times the typical global rate of 130, and more than twice the incarceration rate of 90 percent of the world's countries (Walmsley 2013).

The U.S. incarceration rate in 2012 was significantly higher than those of its neighbors: Canada's and Mexico's incarceration rates were 118 and 210, respectively. Moreover, the United States had a higher incarceration rate than Russia and other Western European nations. The U.S. incarceration rate is more than six times higher than the typical rate of 115 for a nation in the Organisation for Economic Co-Operation and Development (OECD) nation (Walmsley 2013). As seen in figure 6, in recent years incarceration rates in OECD nations have ranged from 47 to 266; these rates are relatively comparable to the rates seen in the United States prior to the 1980s. Indeed, mass incarceration appears to be a relatively unique and recent American phenomenon.

A variety of factors can explain the divergence in incarceration rates. One important factor is higher crime rates, especially rates of violent crimes: the homicide rate in the United States is approximately four times the typical rate among the nations in figure 6 (United Nations Office on Drugs and Crime [UNODC] 2014). Additionally, drug control policies in the United States—which have largely not been replicated in other Western countries—have prominently contributed to the rising incarcerated population over the past several decades (Donahue, Ewing, and Peloquin 2011). Another important factor is sentencing policy; in particular, the United States imposes much longer prison sentences for drug-related offenses than do many economically similar nations. For example, the average expected time served for drug offenses is twenty-three months in the United States in contrast to twelve months in England and Wales and seven months in France (Lynch and Pridemore 2011).

Figure 6. Incarceration Rates in OECD Countries With an incarceration rate of 710 inmates per 100,000 residents, the United States stands in stark contrast to all other OECD nations, with a typical incarceration rate of 115.

Source: Glaze and Herberman 2013; Walmsley 2013; authors' calculations. Note: All incarceration rates are from 2013, with the exception of the rates for Canada, Greece, Israel, the Netherlands, Sweden, Switzerland, and the United States; of these countries, all rates are from 2012, with the exception of Canada, whose rate is from 2011–12. The incarceration rate for the United Kingdom is a weighted average of England and Wales, Northern Ireland, and Scotland. For more details, see the technical appendix.\

For certain demographic groups, incarceration has become a fact of life. Figure 7 illustrates the cumulative risk of imprisonment for men by race, education, and birth cohort. As described by Pettit and Western (2004), the cumulative risk of imprisonment is the projected lifetime likelihood of serving time for a person born in a specific year. Specifically, each point reflects the percent chance that a man born within a given range of years will have spent time in prison by age thirty to thirty-four. Notably, most men who are ever incarcerated enter prison for the first time before age thirty-five, and so these cumulative risks by age thirty to thirty-four are generally reflective of lifetime risks.

Men in the first birth cohort, 1945–49, reached their mid-thirties by 1980 just as incarceration rates began a steady incline. For all education levels within this age group, only an 8-percentage point differential separated white and African American men in terms of imprisonment risk (depicted by the difference between the two solid lines on the far left of figure 7). As incarceration rates rose, however, discrepancies between races became more apparent. Men born in the latest birth cohort, 1975–79, reached their mid-thirties around 2010; for this cohort, the difference in cumulative risk of imprisonment between white and African American men more than doubled (as seen on the far right of figure 7).

These racial disparities become particularly striking when considering men with low educational attainment. Over 53 percentage points distance male high school dropouts in the latest birth cohort (depicted by the difference between the two dotted lines on the far right of figure 7), with African American high school dropouts facing a nearly 70 percent cumulative risk of imprisonment. For African American male high school dropouts, this high risk of imprisonment translates into a higher chance of spending time in prison than of being employed; for African American men in general, it translates into a higher chance of spending time in prison than of graduating with a four-year college degree (Pettit 2012; Pettit and Western 2004).

Figure 7. Cumulative Risk of Imprisonment by Age 30–34 for Men Born 1945–49 and 1975–79, by Race and Education
Among men born between 1975 and 1979, an African American high school dropout has nearly a 70 percent chance of being imprisoned by his mid-thirties.

Source: Western and Wildeman 2009.
Note: Cumulative risk of imprisonment is the projected lifetime likelihood of imprisonment for a person born in a specific range of years. For more details, see the technical appendix.

In the United States the system of taxes and transfers plays an important role in determining the amount of income a family ultimately has at its disposal. Taxes (such as federal and state income taxes, payroll taxes, and property taxes) typically reduce family income, but the federal tax system also provides credits (such as the Earned Income Tax Credit [EITC] and the Child Tax Credit [CTC]) that can increase the cash income for qualifying families.

In 2010 the United States spent more than $80 billion on corrections expenditures at the federal, state, and local levels. Corrections expenditures fund the supervision, confinement, and rehabilitation of adults and juveniles convicted of offenses against t