Keywords

The Growth in Incarceration Rates

In 1971, there were about 200,000 prisoners in the daily US headcount. Since then, we have added approximately 1.3 million people to the daily population of those in prison, with 2016 year-end counts totaling 1,505,400. Counting all forms of incarceration, more than 2.1 million Americans are behind bars on any given day in the United States. These numbers are even more dramatic when one considers the entire correctional population. For example, at year-end 2016, about 2.6% of all US individuals aged 18 or older were under some form of correctional supervision (e.g., prison, jail, probation, and parole). Given these astonishingly high numbers, it is hard to imagine there was a time they were even higher. Yet, in 2007, counting probationers and parolees as people at risk of incarceration, there were over 7.3 million people for whom jail or prison time was a reality or a direct threat—an astonishing 2.4% of the adult population (Kaeble & Cowhig, 2018). As contrast, just over 4% of Americans will report a potentially lethal form of cancer each year (US Cancer Statistics Working Group, 2019).

After peaking in 2007, the US total correctional population has declined annually, with 6,613,500 individuals under the purview of the correctional system at the end of 2016. This represents an average annual decrease of 1.2% for the total correctional population, and an average annual decrease of 0.7% in the incarceration population (Kaeble & Cowhig, 2018). Despite these decreases in the past decade, the incarceration rate in the United States is globally unprecedented. Our current incarceration rate of 655 per 100,000 is the highest in the world, approached only by El Salvador (604), Turkmenistan (552), US Virgin Islands (542), and Thailand (526). Among Western democracies, we are an order of magnitude higher in the use of confinement: England (140), Australia (172), Canada (114), Germany (75), and France (100) all use prison at a rate that is a fraction of ours. Compared to third-world countries, we are at the top of a list that, among those making the most use of imprisonment, might not make us feel so progressive: compare our rate to the totalitarian states of Cuba (510) and China (118). Even with regard to some of the world’s more despotic governments, we still lead the pack.Footnote 1 Knowing nothing else about the United States than our rate of imprisonment, an unbiased observer would be more likely to think we are an economically underdeveloped dictatorship rather that the self-proclaimed “leading voice for freedom in the world.”

The United States has achieved its distinctive incarceration rate through a range of policies that have grown the penal system with little relationship to crime. The general picture is as follows: In the 1970s, the growing prison population closely mirrored increases in rates of felony crime; contrary to many predictions, the increase in imprisonment did not drive down the rate of crime. In the 1980s, property crime first fell precipitously, then rose at a roughly equivalent rate, ending the decade about where it started. Violent crime also fell during the early 1980s, but then rose steeply, ending the decade at a rate of 666.9/100,000 compared to 596.6 in 1980. Prison populations grew annually during these shifts in crime, largely as a consequence of a reduction in the use of probation as a sentence for felony crime. By the end of the decade, sentences to confinement outnumbered sentences to probation by a ratio of 2:1. In the early 1990s, after a peak in property crime in 1991, and a peak in violent crime in 1992, crime rates began to fall, and did so for more than a decade. Because of declining crime, the annual number of new felony commitments to prison also declined. Prison populations grew, nonetheless, because the amount of time served by those going to prison increased as much as 50%. [For a discussion of these three time periods, see Blumstein & Beck, (2005).]

At the turn of the century, apart from a rise in violent crime between 2004 and 2006 and again between 2014 and 2016, crime rates have continued to decrease. However, despite this downward trend in crime, the nation’s incarceration population only began decreasing in 2007. In fact, high incarceration rates have not been driven primarily by crime rates, but largely result from sentencing and correctional policies (Bushway, 2011; see also, Blumstein & Beck, 1999; Pfaff, 2008; Spelman, 2009). It is not surprising then that, in addition to decreasing crime rates, changes in sentencing and corrections laws have also led to the declining incarceration population we have seen in the past decade. These policy changes included drug reforms, pretrial reforms, and community supervision practices (Gelb & Stevenson, 2017). While these changes have provided welcomed reductions in the astronomically high incarceration rates, they are only a starting point. For example, despite drug reform policies, individuals incarcerated for drug offenses still dominate the federal system. Similarly, jail incarceration rates are predominately driven by pretrial detention, specifically bail practices (Sawyer & Wagner, 2019). Thus, even in the face of a decreasing incarceration population, today, we still have a prison population about seven times larger than it was in 1971; an incarceration rate seven times larger, yet a lower crime rate––marginally lower for violent crime and substantially lower for property crime (Bureau of Justice Statistics, 1982; Kaeble & Cowhig, 2018). The current rate of decline in imprisonment is sufficiently slow that unless it accelerates, it will take another 72 years to cut the peak incarceration by half (Ghandnoosh, 2019).

It is important to emphasize how unique the prison system is in the United States, and to accept how ingrained the prison system is in our sociopolitical psyche. Prisons grew in all areas of the country, under Democrat and Republican leadership, during good economic times and bad, while we were at war and during peacetime, before welfare reform and after, during the baby-boom years, and after they had ended. No other nation has this pattern of prison use. It is a peculiarly American idea to use the prison as the first-choice reaction to crime.

Impact on Human and Social Capital in Poor Communities

The growth in imprisonment is not a random social phenomenon. Rather, it concentrates itself within society in four important ways: age, gender, race, and place.Footnote 2

Age

Confinement is disproportionately a young person’s experience. Americans aged 18–44 are about two-fifths of the US population, but they are more than three-quarters of the people behind bars. Young people end up in prison largely because crime is more prominent among the young. The peak age of arrest is the late teens. People rarely go to prison as a consequence of their first arrest, so it is a few years later when subsequent arrest leads to prison. For state prisoners released in 2016, the average time served was 2.6 years, while the median time served is 1.3 years (Kaeble, 2018). In 2006, Irwin and Austin calculated that the typical person who ends up behind bars went to prison for the first time at the age of 25, and, by the end of his sentence, will have spent a meaningful portion of his young adulthood behind bars.

Additionally, these same individuals are likely contributing to the aging prison population (Porter et al., 2016). Between 1993 and 2013, the number of state prisoners aged 55 or older increased by 400% (Carson & Sabol, 2016). While the increasing age of the prison population is often attributed to period effects (e.g., punitive sentencing reforms in the 1980s and 1990s that have led to a larger proportion of individuals serving longer sentences), research also suggests that cohort effects are causing the aging prison population (Carson & Sabol, 2016; Shen et al., 2020). In contrast to period effects that impact all birth cohorts during a specific period, cohort effects impact a specific birth cohort over their entire age (Shen et al., 2020). In their study, Porter et al. (2016) found greater support for cohort effects than period effects. Specifically, they found that individuals born in the 1960s, who came of age in the 1980s, are driving the increasing age of the prison population due to their increased probability of incarceration (Porter et al., 2016). Other research also confirms these findings (see King, 2019; Luallen & Kling, 2014; Shen et al., 2020). For example, Shen et al. (2020) found that birth cohorts who reached the prime age of crime during the 1980s and 1990s have a higher likelihood of incarceration throughout their life course than those who did not come of age during the crime wave.

Gender

Adult men are slightly less than half of the general population, but they account for approximately 93% of the prison population. In 2017, the imprisonment rates for men were 829 per 100,000 US male residents compared to 63 per 100,000 US female residents (Bronson & Carson, 2019). Similarly, in 2017, males were incarcerated in jail at a rate that was six times that females (394 vs. 69 per 100,000 male/female US residents, see Zeng, 2019). In fact, males are more prevalent in all aspects of the criminal justice system. In 2017, males represented 71% of all juvenile arrests (see Puzzanchera, 2019), and in 2018, males represented 73% of all adult arrests (see Federal Bureau of Investigation, 2018). Men end up in behind bars in part for the same reason young people do: they are more likely than others to be criminally active. At year-end 2017, 1.1% of adult males were serving a sentence of more than 1 year (Bronson & Carson, 2019). Although males are a defining feature of the incarceration population, the proportion of the prison population that is women has grown in recent years. For example, the female jail population grew by 20% between 2000 and 2017 (Zeng, 2019). Furthermore, even though the female prison population decreased between 2007 and 2017, it decreased at a lower rate than males, 13.2% versus 8% (Bronson & Carson, 2019).

Race/Ethnicity

The large racial and ethnic disparities characteristic of US prisons has been declining. Between 2007 and 2017, the black-white incarceration gap dropped by 20%. As noted above, while the total incarcerated population declined during this time, the decline was greater for black inmates (20%) than white inmates (13%). The Hispanic white incarceration gap also decreased during this time, with 100,000 more white than Hispanic inmates at year-end 2017 compared to 169,400 at year-end 2007 (Gramlich, 2019). Despite these decreases, the racial and ethnic composition of the prison population is still drastically different from the racial and ethnic composition of the United States. For example, in 2017, whites accounted for 64% of the US adult population, but only 30% of prisoners. On the other hand, blacks only represented 12% of the adult population, yet 33% of the prison population. Finally, Hispanics accounted for 16% of the adult population but 23% of the prison population (Gramlich, 2019). Thus, taken together, both blacks and Hispanics are disproportionality overrepresented in the prison population.

The imprisonment rate in the United States also highlights these racial and ethnic differences. For example, blacks are six times more likely to go to prison than whites (1,549 per 100,000 black US adult residents compared to 272 per 100,000 white adult US residents), and almost twice as likely as Hispanics (823 per 100,000 Hispanic adult US residents; Bronson & Carson, 2019). Unlike age and gender data, the prominence of blacks among those who break the law is less clear. For example, blacks in high school are slightly less likely to report illicit drug use than whites (see Beckett et al., 2006; Western, 2006). Nonetheless, black defendants are overrepresented at each stage in the justice system (Johnson, 2015; see also, Wooldredge et al., 2015; Spohn, 2015; Schlesinger, 2015), and almost one in three black men will be incarcerated in their lifetime (Wildeman & Wang, 2017).

Place

Poor people go to prison at rates much greater than the nonpoor. In his epic study of the role of prisons in inequality, Princeton sociologist Bruce Western (2006) shows how those who enter prison are predominantly those with low human capital: undereducated, underemployed, and underskilled. Due to racial and economic segregation, those who are incarcerated tend to come in concentrated numbers from impoverished neighborhoods. In fact, as noted by Western and Muller (2013), the growth of the US prison population has been concentrated among blacks and the poor (see also, Wacquant, 2009). For example, black men under 35, who dropped out of high school, have a higher chance of being incarcerated than finding a job (Western & Muller, 2013; Western & Pettit, 2010). Similarly, among unemployed 30-year-old men, approximately one-third has been—or is currently—incarcerated (Looney & Turner, 2018). Thus, some deeply poor neighborhoods in major cities have a significant percentage of their adult male residents behind bars on any given day. Of course, they cycle in and out at fairly high rates, so that over time, almost every family in some locations currently has or recently had a member in prison.

Poverty also has long-term consequences for the likelihood of incarceration. For example, a recent report by the Brookings Institute (2018) found that boys who grew up in households at bottom 10% of the US household income distribution were 20 times more likely to be in prison in their early 30s than boys who grew up in wealthy families, that is, those in the top 10% of the US household income distribution (Looney & Turner, 2018). Similarly, other research highlights the benefits of moving from high-poverty communities. For example, the Moving to Opportunities experiment found that relocating from high-poverty neighborhoods to low-poverty neighborhoods reduces arrest rates for men in the short term, and women in both the short and long term (Kling et al., 2005; Sciandra et al., 2013; see also, Looney & Turner, 2018).

The collective effect of these four types of concentration is that certain subgroups of Americans bear the brunt of US prison growth. Young black males who come from impoverished places and develop limited human capital are more likely to go to prison in their lifetimes (see Western, 2006; Western & Muller, 2013). This has substantial impact on social networks, social capital, and informal social control (Clear, 2008).

Social Networks

The array of personal relationships people find themselves involved in comprises their “social network.” It is upon one’s social network that one relies on for social support: when a problem arises, this is the set of relationships upon which a person can call for help; when opportunities are sought, the network is the cast of people whose real-world relationships are the foundation for those opportunities. Volumes have been written on social networks, much more than can be adequately reviewed here. But a few points are important to make concerning incarceration and social networks.

First, poor people—people who are likely to go to prison—have lower levels of social capital than nonpoor people (Warr, 2005). According to Field (2003), social capital is fostered through shared social norms, interactions, and networks that connect individuals (Warr, 2005). In general, individuals with diverse social networks have a stronger likelihood of developing social capital (Warr, 2005; see also, Erickson, 1996; Field, 2003; Granovetter, 1973). Unfortunately, poor people tend to have social networks that are diminished in several respects. Their networks are dominated by what is called “strong ties,” that is, ties that are reciprocal, to people whose networks include roughly the same array of relationships (Field, 2003; see also, Kunitz, 2001; McDonald, 1999; Portes, 1998). Family members, for example, are often strong ties. Furthermore, poor people tend to have fewer “weak ties,” or relationships with people whose networks include a far different cast of relationships. We develop weak ties with many of those at our workplaces. As the name suggests, weak ties are not as strong as strong ties; however, they are valuable in that they can provide individuals with access to a range of new relationships, opportunities, and resources (Warr, 2005; see also, Aguilera, 2002; Erickson, 1996; Granovetter, 1973). For example, through an important weak tie to someone in a person’s network, access is given to many people in the contact’s network but not in that person’s network. This exposure can connect individuals to other, additional resources and broaden their social experience.

Second, poor people’s social networks tend not to span outside of their local residential area, except for family ties. Thus, when a poor person is in need of some form of social support, poor people tend to be limited to relationships in physical proximity to where they live, which often stifles their access to anything more than impoverished resources and can lead to social isolation (Cohen et al., 2003; Tigges et al., 1998).

Third, young men play crucial relationship-building roles in social networks—they are referred to as “entrepreneurs.” They take jobs that generate employment-related social ties, they leave their local neighborhood to work and socialize, bringing outsiders into their networks, and, by contact to others, they generate valuable weak ties. Yet, when young, poor black men go to prison, those who are in their social networks, especially family members, are affected in important—but largely invisible—ways (Clear, 2007; Clear et al., 2001; Morenoff & Harding, 2014). First, many family members maintain contact with a young male who has gone to prison, especially when he goes for the first time and if the sentence is expected to be short. This means that the energy of those who remain behind, which could have been devoted to expanding and strengthening a social network, is spent maintaining the network with the person behind bars. Since these networks were thin to begin with, the effect of the imprisonment is to further weaken them. Studies of networks find that incarceration has a small destabilizing effect, reducing the size of already depleted networks (for a discussion, see Rengifo, 2007; Rengifo & Waring, 2005). The size of the network may be only part of the problem. Because young men should be contributing dynamically to the networks of those associated with them, men who are in prison constitute missing “entrepreneurs” whose absence invisibly diminishes the networks of others, not because of what they do but because of what missing men cannot do.

Social Capital

As briefly mentioned above, social capital is the capacity a person has to obtain “goods” (support, resources, assistance, and materials) through relationships with others. Classic examples are the way parents can deal with the problems of sick children and college entrance applications. Adults with good social capital rely on friends to identify the best available medical care, or they use friendships to bring their children’s applications into a positive light. Thus, social networks are the foundation for social capital. That is why weak ties are so much more valuable than strong ties—weak ties add new layers of relationships to a network that can serve as potential sources of social capital when inevitable problems come along. One way of looking at the intersection of social networks and social capital is that when a person’s human capital —his or her own personal skills and abilities—and wealth are insufficient to deal with a problem in life, the resources available to one’s social networks are activated through the mechanism of interpersonal relationships, and the dormant capacity of those networks is a person’s social capital.

By definition, social capital is lacking in poor neighborhoods. People who live in poor places do not have many personal resources to call upon for social support, other than their immediate family and their personal capabilities. They tend to go to state-supported services for help when they are in need, making use of public welfare, free counseling, and drop-in health clinics. If these services are not adequate, they often do without help. And, as research suggests, the services in poor neighborhoods are often not adequate (Cohen et al., 2003).

The weakening of social networks that results from incarceration of young adults, especially men, has profound implications for social capital. This begins as social networks are affected in the way described above. This small deterioration in social networks adds up, when there is an entire community of people who are similarly affected. Each small diminishment in capacity is multiplied across family units and related networks. Most people whose social networks took a temporary “hit” would compensate by turning to others, but this is not possible in communities where networks were weak to begin with; on top of which, virtually every network is damaged by incarceration in much the same way. The impact of incarceration becomes multiplied when it becomes ubiquitous because the usual compensations are unavailable. The result is that state-sponsored and volunteer services grow in importance for places with limited social capital, simply as a result of the ever diminishing set of options (see Clear, 2007; Clear, 2008; Morenoff & Harding, 2014; Rose & Clear, 1998).

Informal Social Control

Hunter (1985) has defined three levels of social control : public, parochial, and private. Public controls are operated by the state: police, courts, and prisons on the one hand; schools, welfare, and social services on the other. Parochial social controls are community-level groups that stabilize a place’s community life; for example, the barbershop has historically played this role in black communities and religious institutions do the same. Private social control includes intimate interpersonal relations, most characteristically the family (for an expanded discussion, see Bursik & Grasmick [1993]). Two important points can be made from this classification. First, these levels of social control can operate independently, but they typically are in interplay with each other to provide public safety. Second, of the three, the public safety importance of public social control pales in comparison to private and parochial levels of social control.

If private social controls are effective, there is little pressure on parochial or public social controls. When private controls fail, parochial controls can be strained, and public controls attempt to enter the breach. Without the benefit of viable private and parochial controls, public social control is challenged to provide safety for communities.

The incarceration of large numbers of adults from particular communities affects both parochial and private controls (Garland, 2019; Morenoff & Harding, 2014; Sabol & Lynch, 2003). Parochial controls require enough adults to be around to participate in them. Participation is made more difficult when the adult population has limited long-term commitment to a residential area. For example, places with high levels of outward mobility do not sustain the long-term relationships that are the foundation for parochial control (Shaw & McKay, 1942). Instead, in these places a degree of isolation develops that keeps people from interacting with one another (Skogan, 1990), which in turn undermines the creation of social networks, community organizations, and thus, informal social control (Morenoff & Harding, 2014; Sabol & Lynch, 2003; Sampson et al., 1997, 1999).

Places with large numbers of adults going to prison are also places that have many single-parent families (Sabol & Lynch, 2003). The rate of adult incarceration is an indicator of the rate of deterioration in informal social control. (For a detailed discussion of American’s over reliance on public social control and its resulting impact on informal social control, see Garland, 2019.)

In this regard, the incarceration of women, although occurring in smaller numbers than men, may have impact exceeding the sheer question of numbers. Women play crucial roles in poor, especially black, communities. They provide the stability that makes parochial social control possible (in particular, churches and religious institutions) and they are the main providers of child socialization and childcare (Glaze & Maruschak, 2008). Their ability to perform functions of informal social control is impeded when men are absent because they have to focus their energies on filling in the missing functions men might have performed. Thus, the incarceration of men also produces consequences for women who are connected to them (see Braman, 2004; Comfort, 2008; Turney & Wildeman, 2018). Furthermore, research suggests that this is not a marginal number of women as approximately 50% of black women have a family member behind bars (Wildeman & Wang, 2017). Even worse, women are unable to provide any social control when they themselves are behind bars. (See below for a brief discussion of the consequences of paternal and maternal incarceration.)

Public Safety and Public Health Outcomes

There are good reasons, then, to believe that high rates of incarceration, concentrated in impoverished residential areas, will have negative impact on the health and safety of those areas. Studies of this question are mixed, but tend to bear this out.

Health outcomes include the direct effect that incarceration has on an incarcerated individual, as well as the indirect effect it has on the community, families of incarcerated individuals, and children of incarcerated parents.

Health Outcomes for the Incarcerated Individual

Several studies have investigated the relationship between incarceration and health. Compared to the general population, those incarcerated have higher rates of sexually transmitted infections (STIs), HIV, hepatitis C, hypertension, diabetes, and asthma, as well as substance use and mental health disorders (Wildeman & Wang, 2017; see also, Fazel & Danesh, 2002; Wilper et al., 2009). On the other hand, for some population groups, incarceration may have health benefits. For example, black male prisoners have lower rate of mortality than nonincarcerated black males. Although the mechanisms for this decreased mortality rate are debated, some research highlights that incarceration decreases the risk of violence, reduces a person’s access to substances, and provides increased access to health care. While the decreased risk of mortality does not extend beyond black males, many individuals behind bars are receiving their first access to preventive and chronic health care; yet, access to health care does not guarantee high-quality medical care (Wildeman & Wang, 2017; see also, Patterson, 2010; Rosen et al., 2011; Spaulding et al., 2011). Moreover, even for individuals whose health improved or remained stable while incarceration, there is often a disruption in their health care following release from prison or jail (Kulkarni et al., 2010). Thus, when viewed in the aggregate, incarceration worsens an individual’s physical and mental health (Wildeman, 2016; Wildeman & Wang, 2017).

Health Outcomes for the Community

In addition to the direct effect incarceration has on an incarcerated individual, high rates of incarceration in certain communities increases the likelihood of HIV and STIs, as well as teenage pregnancies (Dumont et al., 2013; see also, Johnson & Raphael, 2009; Thomas & Torrone, 2006; Thomas et al., 2007). Research has postulated that these increases are caused by the disruption social structures and social networks (e.g., gender ratios and stable relationships) in these communities (Dumont et al., 2013; Thomas et al., 2007; Thomas & Torrone, 2006). For example, given that males are incarcerated at higher rates than women, communities with high rates of incarceration have less men than women, which increases the likelihood of concurrent sexual partnerships (Johnson & Raphael, 2009; Thomas et al., 2007). And, as research suggests, concurrent partnerships facilitate the spread of STIs and the number of unwanted teenage pregnancies (Dumont et al., 2013; Thomas et al., 2007; Thomas & Torrone, 2006).

Moreover, it is well known that new––consensual and nonconsensual––partnerships are formed in prison. While the likelihood of HIV transmission behind bars is a result of the prevalence of HIV in the prison population, if HIV or another STI is acquired during a person’s incarceration, it is likely to spread within the community after release. For example, in their ethnographic interviews, Thomas et al. (2007) found that formerly incarcerated individuals expressed a strong desire for sex upon return to the community, yet a disinterest in monogamous relationships. These concurrent sexual partnerships––coupled with research suggesting that men released from prison are unlikely to disclose an infection, less likely to use a condom, and likely to have experienced a disruption in medical treatment––increase the likelihood of HIV and/or STI transmission (see Baillargeon et al., 2009; Johnson & Raphael, 2005; Stephenson et al., 2006; Thomas & Gaffield, 2003; Thomas & Sampson, 2005).

Health Outcomes for the Family

Research has also documented the deleterious outcomes associated with incarceration for families. In general, incarceration alters family structures by fracturing intact families and/or reducing the marital prospects for formerly incarcerated individuals or those in the communities to which they return, with research suggesting that women view formerly incarcerated men as less desirable (Apel, 2016; see also, Edin, 2000; Wakefield & Uggen, 2010). In fact, as found by Apel (2016), even short incarceration stints have detrimental effects on residential partnerships and long-term marriage formation. These findings confirm a handful of other research (see Massoglia et al., 2014; Western et al., 2004). For example, Western et al. (2004) found that men who have been to prison are less likely to marry after incarceration, than men of the same background who have never gone to prison. They cohabit, even in long-term relationships, but they tend not to marry (Western et al., 2004). It is no surprise, then, that places with high rates of incarceration have high rates of single-parent families (Sabol & Lynch, 2003). Moreover, given that research highlights that men who go to prison earn less money for the rest of their lives, and have declining employment prospectus (Apel & Sweeten, 2010; Western, 2006; Western & Pettit, 2005), they provide less economic support to the families with whom they cohabit. There is a pattern of mutually reinforcing problems here: places that produce lots of people going to prison are always places with very low economic resources; and the people who leave prison to return to these places will themselves face diminished earning capacity. The result is a cycle of diminished family health and well-being.

Health Outcomes for Children of Incarcerated Parents

The incarceration of a parent is associated with a range of problems, such as truancy, academic underperformance, depression, anxiety, and violent acting out (Wakefield & Montagnet, 2019; see also, Andersen, 2016; Christian, 2009; Comfort et al., 2011; Foster & Hagan, 2007; Hairston, 2007; Murray et al., 2014; Poehlmann, 2005; Turney & Haskins, 2014; Wakefield & Wildeman, 2011, 2013). And, given the vast number of individuals incarcerated in the United States, these negative outcomes impact a handful of minor children. Although the exact number of children with a parent incarcerated is unknown (Wakefield & Montagnet, 2019), a recent survey estimated that 7% of all children will have experienced parental incarceration at least once during their childhood (Murphey & Cooper, 2015; Wakefield & Apel, 2017). Furthermore, the incarceration of a parent not only leads to the fracturing of a parent and child bond but it also can cause reduced social capital, economic deprivation, dysfunctional parenting practices, and an increased risk for intergenerational crime and incarceration (Andersen, 2016; Wakefield & Apel, 2017). Regarding intergenerational crime, research has postulated that children learn criminality through interaction with their parents (Wakefield & Apel, 2017; see also, Akers, 1973; Sutherland & Cressey, 1978). Overall, although the quality of studies on parental incarceration varies, as do the findings (for more detail on data problems, see Kirk & Wakefield, 2018), the fact remains that parental incarceration impacts a child’s well-being in a heterogeneous manner.

Incarcerated women are more likely than men to have lived with their child prior to incarceration (Arditti, 2015; Turney & Wildeman, 2015, 2018). While this would appear to suggest that children are impacted by the discontinuity of care that results from their mother’s incarceration, more so than their father’s incarceration, this is not always the case (Arditti, 2015). Instead, research suggests that maternal incarceration is associated with extreme disadvantage, more so than paternal incarceration; thus, children whose mother is incarcerated likely experienced heightened disadvantage long before their mother’s incarceration. In fact, research suggests that mothers in the criminal justice system are some of the most vulnerable women in the United States (Arditti, 2015). This could explain the inconsistent findings for the effects of maternal incarceration on children. For example, as noted by (Wildeman and colleagues 2013), on average, it is unclear whether maternal incarceration improves, diminishes, or has no effect on child well-being (see also Cho, 2009a, 2009b, 2011). The complexity of the impact of a mother’s incarceration contrasts with research on paternal incarceration, with the bulk of the evidence suggesting a negative or nonsignificant impact on child well-being (Wildeman, 2009; Wildeman et al., 2013).

Safety Outcomes

The original argument that very high rates of imprisonment would lead to higher rates of crime was first made by Rose and Clear (1998). They argued that there would likely be a “tipping point” after which the deterrence and incapacitation benefits of imprisonment would be outweighed by the way it destabilized the neighborhood’s capacity for informal (parochial and private) social control. In a later paper (Clear et al., 2003), they found strong evidence in support of the tipping point thesis, analyzing crime and incarceration data in Tallahassee, Florida. They show that neighborhoods with low levels of incarceration in 1 year experience less crime in the following year; however, neighborhoods with the highest levels of incarceration experience increases in crime rather than decreases in the following year. Similar effects have been found in Portland, Oregon, Columbus, Ohio, and Chicago [see Chap. 7 of Clear, 2007].

The “tipping point” thesis and the evidence in support of it are controversial because the exact relationship is difficult to model. Crime and incarceration are reciprocally related; that is, incarceration is undoubtedly a result of crime because with only the rarest of exceptions, people do not get locked up unless crimes have been committed. Yet the model also posits a complicated relationship in the opposite direction: up to a point, incarceration will reduce crime; after a point, it will increase it. The available data and math to statistically model this sort of effect are not very satisfactory.

For example, Lynch and Sabol (2004) analyzing data in Baltimore and Cleveland found a similar kind of pattern to incarceration and crime as Clear and his colleagues when they replicated that modeling method. However, when they used a different statistical technique (instrumental variables), not only did the nature of the impact change but also it reversed itself, suggesting that higher rates of incarceration decrease crime (though it also increased fear of crime and had other problematic effects). They conclude that the tipping point models of crime and incarceration are inconclusive, at best, and potentially wrong.

Taylor et al. (2006) entered this debate by investigating the impact of adult incarceration on serious juvenile crime. They argue that using juvenile crime is a solution to this problem because it breaks the reciprocity of the crime-incarceration relationship: it is irrational to argue that juvenile crime “causes” adult incarceration. When they investigate incarceration and juvenile crime (arrest) rates in Philadelphia neighborhoods, they find support for the tipping point thesis: high adult incarceration rates in 1 year lead to increasing juvenile crime rates in the years that follow.

In a review of this and other work on the topic, Clear (2007) argues that the weight of the evidence supports the tipping point thesis. The argument is simple. To argue in favor of the tipping point thesis is to argue that incarceration, at high rates of concentration, clearly destabilizes families, weakens bonds between parents and children, decreases economic well-being, diminishes the capacity of social networks, reduces long-term job market viability, increases serious juvenile delinquency, and increases crime. The opposite argument is that concentrated incarceration has all of these same negative effects, each of which might be expected to increase crime, yet it does not increase crime.

In recent years, additional studies have investigated the relationship between incarceration and crime. For example, in their 2010 study of Chicago, Sampson and Loeffler explored community-level incarceration rates between 1990 and 2006. While they mainly focused on investigating the impact of incarceration across different neighborhoods, their study also explored the relationship between incarceration and crime. Sampson and Loeffler (2010) argued that if incarceration deterred crime then decreases in crime rates should follow incarceration; yet, neither increases nor decreases in crime were found in their study. However, as discussed at the beginning of this chapter, the crime drop of the early 1990s was associated with a steep rise in incarceration. In other words, as crime decreased, incarceration increased––signifying a negative relationship. Despite these findings, Sampson and Loeffler (2010) echo the sentiments of other studies, arguing that the relationship between incarceration and crime is too complex to make causal claims. Overall, although the assertion that incarceration at the highest levels increases crimes has not been definitively proven, the evidence to support that conclusion is strong.

Incarceration, Public Health, and Public Safety

If the arguments of this chapter are correct, then what is to be done?

It goes without saying that to stop locking people up from certain communities, just because that neighborhood has reached a “tipping point,” is untenable. There are all kinds of people convicted of crimes in these places, some of whom would seem dangerous to any eye. Likewise, it seems implausible to segment a certain stratum of crime, say drug sales, and treat residents of this location differently than others when convicted of this crime merely because of where they live. There is no obvious way to treat residents here differently, in terms of punishment, and not raise basic objections of equity and justice.

There are at least two possible routes, however: sentencing reform and justice reinvestment.

Sentencing Reform

The prison population is fully determined by two numbers: how many people go in and how long they stay. To reduce the prison population, either number (or both) must be changed.

It is possible to reduce the prison population proportionately by reducing (average) length of stay. It is difficult to see how public safety will be affected much, if at all, by releasing people a few months earlier than we do now. Almost everyone who goes into prison comes out. The length of stay in prison has little or no effect on recidivism (if there is an effect, it is that longer stays lead to higher rates of failure). If the general population experienced a 3-month, across-the-board reduction in stay, it would make people’s return to the streets more rapid in the marginal sense, and would have potentially significant impact on the number of people behind bars. One group of criminologists estimated that a 3-month reduction in length of stay for felony offenders, and a 6-month reduction for arrested parole violators, would result in a reduction in the average daily prison population by over 200,000 prisoners (Irwin & Austin, 2006).

In general, additional research has supported this argument. For example, Mears, Cochran, Bales, and Bhati (2016) found that the effects of incarceration differed by the amount of time served. Specifically, Mears et al. (2016) identified three differences: (1) for sentences of 1 year or less, incarceration increases recidivism; (2) for sentences between 13 and 24 months, time served decreases recidivism; and (3) for sentences between 25 and 60 months, time served did not impact recidivism. Taken together, Mears et al. (2016) argue that there is not one single effect of incarceration on recidivism. However, most importantly from a policy perspective, the results of their study indicate that lengthier prison sentences do not reduce recidivism; instead, lengthy prison sentences may increase recidivism. For example, sentencing an individual to a year of incarceration, rather than a month, may result in a 10% increase in recidivism. Mears et al. (2016) note that their study is not an argument against lengthy prison sentences; instead, their findings should spark policy considerations for the use of shorter sentences, as doing so does not appear to reduce public safety through increased recidivism. Similarly, a recent study conducted by Rhodes, Gaes, Kling, and Cutler (2018) echoes these findings. Specifically, Rhodes et al. (2018) found that a 7.5-month reduction in an individual’s prison sentence does not impact recidivism rates; thus, arguing that prison sentences can be reduced with a negligible effect on recidivism rates and public safety.

Justice Reinvestment

There are two “good news” items in this story. First, the number of areas in our major cities that are negatively affected by high incarceration rates is not large; usually, it is only a handful of places––less than three or four neighborhoods in most large cities. It means that the target of change is not jurisdiction wide but much more targeted. The small number of affected places opens the possibility for targeted strategies that focus their efforts in those places.

Second, there is already a great deal of money being spent on the public safety problem. In 2016, state and local governments spent about $187 billion on police and corrections (Urban-Brookings Tax Policy Center Data Query System, 2019). This is a large amount of money, especially at a time of declining budgets, which has caused policy makers to question whether extensive reliance on incarceration is an effective approach (Harvell, Welsh-Loveman, Love, & La Vigne, 2016; see also, Ho et al., 2013). In fact, at the turn of the century, scholars have called attention to the potential benefits of community justice models that are focused on heavily affected communities and divert existing resources in a strategy called “justice reinvestment” (Tucker & Cadora, 2003).

Justice reinvestment, which relies on data-driven evidence, has two broad goals. First, it aims to reduce the size of the prison population and, thus, reduce criminal justice expenditures by reforming sentencing and revocation policies. Second, the money saved is reinvested in communities that are negatively affected by criminal justice (Clear, 2011; Harvell et al., 2016). The intended result of justice reinvestment is win-win: the community is improved through targeted efforts, and the effects of incarceration are ameliorated because the projects both replace incarceration and target the negative consequences of incarceration, such as substandard housing, school failure, and economic decay.

Although the Council of State Governments Justice Center experimented with the idea of justice reinvestment in the early 2000s, the Justice Reinvestment Initiative (JRI) formally began in 2010 through a partnership between The Pew Charitable Trusts and the Bureau of Justice Assistance. Through the work of various technical assistance providers, JRI was implemented in various states around the country. While the approaches vary across states, policies generally attempted to reduce the number of people incarcerated, limit the time spent incarcerated, and strengthen community programming. States have evaluated their progress on reducing prison population, strengthening community programs, cost savings, and the reinvestment of these savings (Harvell et al., 2016). In their evaluation of the first 6 years of JRI, the Urban Institute found that the initiative generated a greater reliance on data-driven, evidence-based corrections practices in the 28 participating states (Harvell et al., 2016). In terms of savings and reinvestment, for the most part, states have reduced prison costs and managed to reinvest this money into evidence-based public safety strategies within communities hit the hardest by incarceration. For example, in 2016, JRI participating states saved more than $1.1 billion and reinvested more than $450 million into various community reform strategies (Harvell et al., 2016).

Despite these promising findings, other independent reviews of JRI are both limited and critical (Sabol & Baumann, 2020; see also, Austin et al., 2016; Brown et al., 2016; Fox et al., 2013). One of the major criticisms of justice reinvestment was its inability to specify how large reductions in prison populations could be reached without more radical sentencing approaches. For example, some scholars have argued that to achieve substantial reductions in prison populations, sentencing reforms need to reclassify some crimes as misdemeanors, reduce arrests for drug crimes, shorten length of stays, expand early release policies, and eliminate mandatory minimums and life sentences for certain crimes (Sabol & Baumann, 2020; see also, Austin, 2011; Austin et al., 2013). An additional criticism of justice investment was its assumption that money from prison population reductions could—and would—be redirected into the communities, rather than used to tackle other needs (Sabol & Baumann, 2020; see also, Tonry, 2011). Finally, even if the savings were reinvested into the community, the cost savings were overstated (Sabol & Baumann, 2020, see also Pfaff, 2016; Tonry, 2011). For example, as noted above, although cost savings have been achieved through JRI, these savings are insignificant to the insurmountable amount of money spent on corrections.

Taken together, although JRI accelerated sentencing reforms and the adoption of evidence-based practices, JRI has not yet brought about the promised substantial change (Sabol & Baumann, 2020). For example, evidence on its ability to reduce prison populations, achieve cost savings, and redirect these savings into impacted communities has been limited. It is clear that JRI, as the policy mechanism to carry our justice reinvestment, has been disappointing. Yet, the promise of the strategy—to deal with criminal victimization more effectively by spending funds differently—remains a compelling idea in the policy arena that has yet to be fully tested.

Conclusion

This chapter has linked incarceration to the health and safety problems of certain subgroups of the US population, and it has shown how concentrated incarceration within those subgroups has had deleterious effects on the neighborhoods within which they live. Research demonstrating the negative impact of high rates of incarceration on public health and public safety has been summarized, with attention to both settled matters of empirical fact and controversial matters currently under debate. The chapter concludes with some suggested ways of addressing this problem.

There is a moral imperative that has not yet been stressed. The penal apparatus in the United States is meant to be an instrument of public safety and institutional justice. To the degree that people who break the law should not be allowed to do so without consequences that symbolize the wrongfulness of that conduct, the criminal justice system is an essential instrument of social order. Yet in the United States, the system has grown to the point that it no longer can claim the high moral ground. In some places, the punitive apparatus of the penal system is now one of the proximate causes of declining public safety. For the children who grow up in those areas, the criminal justice system is now a source of social injustice that robs them of life chances and places a low ceiling on their lifelong prospects.

Although recent policy changes have resulted in reductions in the incarcerated population, this is only a starting point. For example, as argued by Shen et al. (2020), policy changes such as reductions in mandatory minimums can only do so much; instead, policy efforts should focus on how a person’s prior criminal history affects their sentence, as this is one of the reasons for the aging prison population. Yet, as Shen et al. (2020) also note, given that these individuals are aging out of the criminal justice system, the incarcerated population may continue to drop even without policy changes. However, our prison population has grown too large––and is associated with far too many social costs––to merely wait for individuals to age out of the system. We need to change our overreliance on imprisonment; for, as the research in this chapter suggests, it is an ineffective public safety strategy with far-reaching collateral consequences. It is also a strategy that, much evidence suggests, has outgrown its usefulness.