The United States prison system is a potent example of American exceptionalism. No other country incarcerates nearly as many people. To give some perspective, if all 2.2 million prisoners constituted their own city, they would be the fourth largest in the country, just short of Chicago. If they comprised their own state, they would be more populous than fifteen states. In addition, the rate of incarceration, 707 out of every 100,000 residents, far exceeds the average incarceration rate of European countries (a paltry 140 out of 100,000 residents). Such a widespread social phenomenon would seem to have important consequences for economic, political, and social institutions. This is precisely what Amy Lerman investigates in her excellent new book, The Modern Prison Paradox: Politics, Punishment, and Social Community.

In theory, prisons can generate net benefits for society in several ways. Prisons can be a deterrent against committing crimes. They incapacitate people, thus preventing victimization while the person is incarcerated. They may also rehabilitate inmates so that upon release, a person does not reoffend. It may be, however, that going to prison makes a person more likely to reoffend. As Lerman explains, “the modern American prison presents us with a troubling paradox…the crime control politics of the past half-century have given rise to institutions that (re-)create the conditions that arguably gave rise to criminality in the first place” (p. 12). If so, the rise of mass incarceration since the early 1980s would actually undermine the crime control ends that these policies apparently sought.

Lerman’s book takes a rigorous approach to empirically measuring the influence of relatively harsher prison settings in California on people’s attitudes and behaviors. The book provides a refreshing look at how stricter confinements affect both inmates and correctional officers working in prison. In doing so, Lerman is able to assess a small, but crucial, part of the broader question about how prison affects society.

Prisons can affect inmates in several ways. Prison may be a “holding pattern” that has no notable influence on inmates. Prison may rehabilitate an inmate, teaching him more socially acceptable attitudes and behaviors. Finally, prison may have criminogenic effects, “hardening” the inmate. Lerman tests these claims with the help of a creative quasi-experiment. She takes advantage of the security classification scoring scheme that officials use to determine the prison where an inmate will be held. The score is determined by a relatively large number of risk factors. The total score determines at which security level an inmate is housed. In California, prisons rate from Level I (low security) to Level IV (high security). She compares inmates who score just below (26 and 27) and above (28 and 29) the threshold that determines whether they will go to a Level II or Level III prison. These inmates are not significantly different on most observable measures, so comparing their attitudes and behaviors provides a way to identify the effect of harshness of prison conditions.

Based on inmate surveys, she finds no overall difference for social isolation and feelings of belonging. However, a difference exists between those inmates who have had relatively less contact with the criminal justice system. Assignment to a higher security prison is associated with less social isolation for those inmates with relatively little criminal history. In particular, these inmates report relatively greater contact with friends who provide support and companionship. Lerman argues that this reflects the strengthening of social bonds. In particular, part of this social bonding is with friends and acquaintances who are gang members. Inmates just above the cutoff threshold, with low criminal history, also report differences in norms of aggression. They reported greater agreement with statements like, “Some people just don’t deserve respect and should be treated like animals” (p. 111). Finally, she finds that harsher prisons have a significant effect on the likelihood of recidivism (p. 117). In sum, for inmates with relatively less contact with the criminal justice system the harsher prison environment shapes the social networks and attitudes that they develop and influences their post-release behavior.

What affect does the harshness of prisons have on correctional officers’ attitudes and behaviors? Past work has had difficulty overcoming the selection effects of both who becomes a correctional officer and where they choose to work. Officers with a stronger belief in rehabilitative and liberal ideals, for example, may choose to work in low security facilities, while someone with more punitive beliefs may prefer working in high-security prisons. The empirical challenge is to isolate the effect of the harsher prison on attitudes and behaviors.

Lerman exploits a bureaucratic procedure for allocating new officers to a particular prison to identify this effect. In California, new cadet officers must do a 24 month apprenticeship at one of the state’s prisons. Assignment to prisons takes place in two ways. A cadet can select ten prisons (out of the state’s 33 prisons) at which he or she would like to work. Or, if the cadet prefers, he or she can express a willingness to go to any prison in the state. The latter group overcomes the problem of self-selection bias. Lerman also argues convincingly that, for a variety of bureaucratic, economic, and geographic reasons, even those officers who list their preferred facilities still have relatively little choice over where they are assigned. Because apprentice officers have relatively little choice over where they work, the institution and security level where he or she works are likely to be uncorrelated with individual characteristics. She compares the attitudes and behaviors of first-year apprentice officers working at Level III and Level IV facilities.

Among officers assigned to higher security prisons, officers are less supportive of specific rehabilitation programs, including vocational programs. However, there is no difference between a general, ideological support for rehabilitation between officers at the higher and lower-security prisons. Officers in high-security prisons rate their interactions with inmates slightly worse. Officers at high-security prisons are more likely than those at low-security prisons to believe that the opportunity for inmates to work in the prison increases violence. The behavior of these groups of officers differ as well. Apprentice officers at high-security prisons are less likely to contact a supervisor if they need assistance and are more likely to contact the California Correctional Peace Officers Association. As with inmates, harsher prisons have real, unintended consequences for prison staff.

The Modern Prison Paradox is a well-executed empirical investigation into a major government institution. It provides compelling evidence that socialization occurs within prison; socialization is often negative in nature; and harsher prisons affect how inmates and staff think and behave. Far from the idea that inmates are isolated from society or that officials “lock them up and throw away the key,” nearly 95 % of inmates will eventually be released. It matters therefore whether these people will be reformed or impaired. Amy Lerman’s informative analysis and persuasive account is a major advance in our knowledge of what kinds of people prisons produce.