Exclusion and Extraction: Criminal Justice Contact and the Reallocation of Labor

In the United States, almost seven million people are under correctional control. This includes 2.3 million held in the nation’s jails, prisons, detention centers, and involuntary commitment facilities. It also includes 4.5 million people in community corrections—3.7 million on probation and

t h e c r I m I n a l J u s t I c e s y s t e m a s a l a b o r m a r k e t I n s t I t u t I o n r s f : t h e r u s s e l l s a g e f o u n d a t i o n j o u r n a l o f t h e s o c i a l s c i e n c e s 2. The relatively high labor market stability that workers in Germany experience is attributable in part to the strong protections that German workers receive from employment protection laws and from extremely high rates of collective bargaining coverage-upward of 80 to 90 percent (Freeman 2002;DiPrete 2002). According to Thomas DiPrete (2002), German workers' pronounced labor market stability is also attributable to their tight coupling of education and job training, between school and work. Such strong protections, however, also discourage employers from hiring for fear that they might be stuck with bad employees or unable to dismiss workers during periods of economic shocks. Pettit and Lyons 2007;Sabol 2007). This is in part because for many individuals, employment typically amounts to day labor with no real prospects of further employment, let alone benefits (Sugie 2018). Thus the criminal justice system not only afects aggregate labor market participation, employment rates, and employment stability, but also erodes wages and earnings while driving up rates of poverty among the employed (Western and Beckett 1999;Western 2002;Western and Pettit 2005).
We draw from the economic and sociological bodies of research to define what we mean by labor market institutions and then explain how criminal justice policies serve as such, focusing on the role these policies have played to both exclude justice-involved individuals from labor market opportunities, but also to extract labor from the same population, often under oppressive conditions. We end our discussion by asking whether the United States is unique in its use of the criminal justice system as a key labor market institution? Is this yet another case of American exceptionalism? W hat are L abor M arke t Ins tIt u tIons?
Broadly defined, labor market institutions are organizations and policy interventions that have significant efects on short-and long-term wage and employment outcomes and on economic performance generally. According to Gordon Betcherman (2012, 2), "The laws, practices, policies, and conventions that fall under the umbrella of 'labor market institutions' determine inter alia [emphasis in the original] what kinds of employment contracts are permissible; set boundaries for wages and benefits, hours, and working conditions; define the rules for collective representation and bargaining; proscribe certain employment practices; and provide for social protection for workers" (see also Holmlund 2014).
Collectively, these institutional arrangements have a profound efect on individuals' well-being, shaping patterns of social and economic stability over the life course. They afect individuals' access to jobs and shape aggregate labor market participation, employment status, security, stability, and compensation, including wages and benefits. To the extent that labor market institutions afect training, the adoption of new technologies, the size and structure of firms, and the efficient reallocation of labor, they also afect productivity. And, insofar as labor market institutions contribute significantly to individuals' sense of economic security and stability, they also indirectly influence social cohesion, including individuals' sense of belonging to their community, their sense that equality of opportunities exists, and their civic engagement, among other things (Sampson 2011).
Employment protection laws, labor unions, minimum wage laws, and social insurance programs, most notably unemployment insurance, are the most often studied labor market institutions (Freeman 2007;DiPrete et al. 1997;DiPrete 2002;Gangl 2004aGangl , 2004bGangl , 2006Cigagna and Sulis 2013;Jaumotte and Buitron 2015). Employment protection laws regulate the hiring and firing of workers as well as the terms of temporary and fixed-term employment contracts. They can make it very difficult for employers to dismiss workers, thus providing the foundation for greater job security and greater labor market stability over the life course (DiPrete 2002). 2 Strong unions, which negotiate with employers for better wages, benefits, and working conditions, can also push to secure stronger employment protections for their members as well. As a result, union workers earn higher wages than their non-union counterparts. Furr s f : t h e r u s s e l l s a g e f o u n d a t i o n j o u r n a l o f t h e s o c i a l s c i e n c e s e x c l u s I o n a n d e x t r a c t I o n 3 ther, in industries and communities with higher union density, workers, no matter their union affiliation, earn higher wages, and levels of intergenerational economic mobility are much higher, even for children of non-union workers (Western and Rosenfeld 2011;Freeman et al. 2015). The minimum wage sets a floor under which most workers should not fall. Recent studies of local and state minimum wage increases reveal significant earnings increases for targeted and nontargeted workers, but a minimal efect on local unemployment rates (Allegretto et al. 2018); a growing body of evidence also indicates that higher minimum wage floors reduce racial gaps in earnings (Derenoncourt and Montialoux 2018). Collective social insurance programs can protect workers who experience triggering events, including job loss, from significant declines in living conditions (DiPrete 2002). For instance, through public pension systems for those retiring because of old age, unemployment insurance for those displaced through no fault of their own, and disability for those who fall ill, social insurance programs help manage the risks that families experiencing employment exits face by smoothing income streams during periods of income volatility (DiPrete et al. 1997, 323;Brady 2009;Gangl 2004bGangl , 2006. In so doing, they help decouple workers' living conditions and life chances from labor market fluctuations (Di-Prete 2002;Gangl 2004aGangl , 2004bGangl , 2006. Thus, labor market institutions shape economic performance and the flow of workers into and out of the labor market by altering incentives, as with minimum wage increases; by facilitating efficient bargaining, as with strong unions; and by increasing information and communication flows inside firms, which can facilitate better decision making by both managers and labor. The massive expansion and reach of the penal state-the set of institutions that have control over the power to punish (Garland 2013), including but not limited to those that adjudicate guilt and punish wrongdoers (Beckett and Murakawa 2012)-has altered the incentives of labor market participants, employers, and the state. In so doing, it has also profoundly and negatively afected employment and wages (Western and Beckett 1999), disproportionately so for low-income communities of color. Indeed, with its expansion, the penal state has come to play an outsized role in the reproduction of inequalities, increasing racial and class disparities across a number of important indicators (Western and Beckett 1999). It has also helped obscure the lack of progress the country has made toward racial equality (Western and Beckett 1999). Further, to the extent that penal expansion has removed individuals at high risk of unemployment, underemployment, or pure joblessness from the labor force, it has also helped conceal just how much the U.S. economy struggles to provide adequate employment for all who want it, in absolute terms and relative to other advanced capitalist economies (Western and Beckett 1999). Indeed, with few exceptions, mainstream economic analysis continues to ignore the impact that high rates of criminal justice involvement has, not only on rates of employment, but also on labor-force participation. Thus, the criminal justice system legitimizes a political economic system that increasingly fails to provide opportunities to achieve social and economic stability, security, and inclusion.
Just as the penal state shapes the flow of individuals into and out of the labor market, so the labor market shapes the flow of individuals into and out of the criminal justice systememployment laws and labor market agents surveilling, identifying, and penalizing individuals for perceived or actual criminal behavior (Levin 2018;Simon 2007). Because expanding incarceration also creates employment for some potential workers, often racially and geographically quite diferent than the communities from which the incarcerated are taken, this complex interaction has arguably become one of the major anchors for the enduring racial stratification in the labor market (Harcourt 2011). Further, the criminal justice system does not limit its moonlighting to its role in shaping labor market flow. In a society that relies heavily on the penal state to address rather than resolve its most pressing and seemingly intractable problems of poverty and race, it comes as little surprise that the penal system also functions as quasi-welfare, education, and healthcare institutions, providing housing, education t h e c r I m I n a l J u s t I c e s y s t e m a s a l a b o r m a r k e t I n s t I t u t I o n r s f : t h e r u s s e l l s a g e f o u n d a t i o n j o u r n a l o f t h e s o c i a l s c i e n c e s 3. This has been interpreted to mean that in anticipation of the social, political, and economic disorder that can emerge during periods of high unemployment, criminal justice authorities incarcerate individuals at risk for joblessness at higher rates than normal to head off potential troubles (D'Alessio and Stolzenberg 2002).
4. Many institutionalized populations-the incarcerated and military personnel-are not counted in official statistics on labor market participation, and thus incarcerated individuals are not counted in conventional measures of employment either. Because the justice-involved would likely experience unemployment, underemployment, or joblessness at much higher rates, counting them in official statistics would not only produce higher rates of unemployment, but also significantly lower average wages and earnings. 5. For this reason, many point to justice-involved job seekers' pre-offense, individual-level attributes to explain poor employment outcomes post-contact, in that the characteristics that predict criminal behavior also help explain poor employment outcomes, post-offense (Grogger 1995;Useem and Piehl 2008).
6. Some restrictions are deployed unambiguously as an additional form of punishment; voting bans are an example. The appeal of others is that they protect society; for example, persons with convictions for theft, including identity theft, embezzlement, and fraud, cannot be hired for positions requiring fiduciary responsibility. and training, and treatment for many among the poor who struggle with the efects of poverty and social exclusion.
the CrIMIn a L Jus tICe sys teM as a L a bor M a rke t Ins tIt u tIon In a pivotal article, "How Unregulated Is the U.S. Labor Market? The Penal System as a Labor Market Institution," Bruce Western and Katherine Beckett (1999) note that by making significant investments in the expansion of the carceral system, the state through its criminal justice policy had a substantial, and dynamic efect on employment trends. In the short term, they argue, warehousing socially and economically marginalized, able-bodied men of working age artificially lowered conventional measures of unemployment by removing individuals at high risk of joblessness from the labor force. Indeed, independent of the actual volume of criminal activity, increases in unemployment rates tend to yield higher rates of incarceration (Yeager 1979;Chiricos and Delone 1992;D'Alessio and Stolzenberg 2002). 3 In this way, incarceration is for Western and Beckett a "hidden joblessness" of sorts. 4 They contend, however, that incarceration in the long term actually increases conventional unemployment rates by constructing serious barriers to employment for those returning home after time served. In the following section, we discuss the meaningful ways that policies related to criminal justice contact profoundly shape employment trends and economic performance. In some cases, exclusions and extractions appear coterminous, given that justice-involved individuals are directed away from promising employment opportunities while being coerced into taking some of the worst jobs at the low end of the labor market.

Labor Market Exclusion and Marginalization After Criminal Justice Contact
Even before penal contact, justice-involved people, who are disproportionately poor, uneducated, and of color (Western 2006;Wacquant 2009), generally struggle with higher rates of unemployment, and when employed garner lower hourly wages, work relatively few weeks per year, and have annual earnings that place them below the poverty line (Western and Beckett 1999;Western, Kling, and Weiman 2001;Western 2002Western , 2006. 5 After penal contact, however, their employment prospects dim further still (Western 2006;Visher and Kachnowski 2007).
On criminal convictions, courts impose punishment with sentences of varying degrees of severity, but sometimes even before their cases are fully adjudicated, the convicted face another set of sanctions-the formal and informal collateral consequences of criminal justice contact, which combine to erect significant barriers to employment. Formal collateral consequences, most of which are stipulated in criminal employment law, restrict or limit the rights that individuals with criminal records have to participate in social, political, and economic domains of life. This includes but is not limited to voting, parenting, holding public office, serving on a jury, and working. In the realm of employment, federal and state governments have not only restricted individuals' access to government employment, they have also enacted numerous provisions against extending licenses to justice-involved people for government-regulated, private occupations, with distressing results (Dale 1976;May 1995;Olivares, Burton, and Cullen 1996;Petersilia 2003;Bushway and Sweeten 2007;Mills 2008). Nationwide, roughly eight hundred occupations are formally of limits to ex-felons because of such statutes (Bushway and Sweeten 2007). And, as Cody Warner, Joshua Kaiser, and Jason Houle (2020) report in this volume, the accumulation of these penalties, what they call "hidden sentences," has serious consequences for employment and earnings. Analyzing the National Longitudinal Survey of Youth 97 (NLSY97) and the National Inventory of the Collateral Consequence of Conviction datasets, the authors find that state-level hidden sentences have a significant and negative efect on the employment prospects of justice-involved young adults. Relative to individuals residing in states with low levels of hidden sentences, those living in states with high levels experience greater unemployment and, when employed, lower earnings. This was especially so for black formerly incarcerated individuals for whom employment and earnings penalties were largest.
Unfortunately, individuals' eforts to engage the courts around the barriers created by hidden sentences also contribute to rupturing their connection to the formal labor market. In this volume, Michele Cadigan and Gabriela Kirk (2020) highlight the hurdles that individuals face when attempting to resolve issues relating to court-ordered debt. Drawing from hours of observations of court hearings in Illinois and Washington State, as well as in-depth interviews with residents of those states who had been sentenced to pay court-ordered monetary sanctions, Cadigan and Kirk deftly describe the bureaucratic procedural pressure points that individuals must negotiate to successfully navigate encounters with the courts and to keep their jobs. Not surprisingly, the barriers created by race, class, and gender disadvantaged many, and in the process severed all the more profoundly individuals' ties to the formal labor market.
Although studied less often, informal collateral consequences, including the loss of employment opportunities resulting from employers' eforts to avoid legal liabilities associated with criminal records, can also have pernicious efects, magnifying the social, political, psychological, and economic costs of criminal justice contact. As extensions of punishments meted out by the courts, both formal and informal collateral consequences raise serious concerns insofar as they indicate a lack of consideration made for proportionality, parsimony, citizenship, and social inclusion, ideals for how justice in any context should be meted out (Travis, Western, and Redburne 2014).

Civil Death: State and Federal Restrictions on Employment
Federal and state governments have not only restricted access to government employment for individuals with criminal records, but also enacted numerous provisions against extending licenses to justice-involved people for government-regulated, private occupations (Dale 1976;May 1995;Olivares, Burton, and Cullen 1996;Petersilia 2003;Bushway and Sweeten 2007;Mills 2008). For instance, in New York, ex-felons are restricted from owning barber shops, distributing commercial feed, and acting as emergency medical technicians (Uggen et al. 2006). In Florida, speech-language pathology and cosmetology are of limits. According to Kathleen Olivares, Velmer Burton Jr., and Francis Cullen (1996), six states have permanently denied access to public employment-Alabama, Delaware, Iowa, Mississippi, Rhode Island, and South Carolina. For the remaining states, restrictions in access vary in length and by given contingencies. For instance, in ten states, public sector employers have discretion to hire justice-involved individuals; in twelve states, hiring is contingent on whether the conviction is occupation-related; and in seventeen states, individuals are able to access public employment once they have completed their sentences.
Perhaps more profoundly damaging the employment prospects of the justice-involved are the numerous provisions against extending licenses to justice-involved people for government-regulated private occupations.
Trade or occupational licenses represent the formal permission that a government body gives to individuals, which allow them to engage in activities, including professional, skilled, or semi-and unskilled work, that would otherwise be of limits if not for the state's permission (Dale 1976;Rodriguez and Avery 2016). The number of occupations requiring licenses has grown significantly over the past half century, presumably for public protection. According to Morris Kleiner (2015), fewer than 5 percent of workers in the United States required state permission to work in their chosen profession seventy years ago, but today more than one in four workers must have a license to perform their occupational duties (see also Kleiner and Krueger 2010). Rates vary by state, however, the lowest share (12.4 percent) is in South Carolina and the highest (fully 33 percent) is in Iowa (Kleiner 2015).
Although the statutory requirements for obtaining occupational licenses vary widely by state and across occupations, all states have numerous provisions that efectively exclude the justice-involved from engaging in various forms of employment. Indeed, the American Bar Association's (ABA) inventory of penalties documents more than twenty-seven thousand state occupational licensing restrictions against individuals with a criminal record. These restrictions without question significantly reduce the quantity and quality of employment available to those with criminal records (Dale 1976;May 1995;Rodriguez and Avery 2016), impacts varying by race and ethnicity (Rodriguez and Avery 2016). According to one study, for instance, occupational restrictions are estimated to reduce by almost three million the number of people employed across the nation (Kleiner, Humphris, and Koumenta 2011;Kleiner 2015;Blair and Chung 2018). This would seem to have clear implications for the justice-involved.
Two types of requirements in occupational licensing statutes create the greatest barriers to obtaining licenses for the justice-involved (Dale 1976;May 1995;Rodriguez and Avery 2016). The first are blanket bans, which automatically disqualify individuals with certain records. According to the ABA Inventory, individuals found guilty of any felony face more than twelve thousand restrictions in occupational, professional, and business licenses categories; those with any type of misdemeanor face more than six thousand such restrictions. Further, permanent, or lifetime, disqualifications number more than nineteen thousand; mandatory disqualifications, for which licensing agencies have no choice but to deny a license, total more than eleven thousand (Rodriguez and Avery 2016). Statutes requiring good moral character and the like also erect significant barriers to justice-involved individuals' ability to get occupational licenses and find employment. Embedded in many licensing laws is a character component (May 1995). As highlighted in the literature, however, requiring good moral character raises at least two issues. The term is itself vague and subjective, leaving decisions about who qualifies as having such character to the discretion of licensing agencies. Also, having been convicted of a crime can easily be marshaled as evidence that one lacks good moral character, an argument with which both the courts and licensing agencies have generally agreed (May 1995). Statutes that require a "good reputation," "reputable character," or "honest and trustworthy character" are less demanding insofar as they allow for the possibility of rehabilitation and redemption. The plethora of vague statutory terms (such as good moral character) and widely varying requirements and procedures, both across states and within occupations, creates a lack of transparency and predictability in licensing systems, magnifying the burdens that the justiceinvolved face as they undergo the process to become licensed and to find work (Rodriguez and Avery 2016). In terms of restricting licensing boards' ability to take criminal records into consideration, a comprehensive evaluation of states' licensing laws revealed generally poor standards across the board (Rodriguez and Avery 2016).
Although it remains unclear what impact occupational licensing restrictions, and the failure of many states to check them, have had on employment, evidence indicates that the effects are significant. In Florida, for instance, 40 percent of jobs have been subject to statemandated criminal background checks or emr s f : t h e r u s s e l l s a g e f o u n d a t i o n j o u r n a l o f t h e s o c i a l s c i e n c e s e x c l u s I o n a n d e x t r a c t I o n 7 ployment restrictions based on having a criminal record (Mills 2008). Even for defendants, legal stigma forecloses access to some jobs while their cases remain open, and thus they are also vulnerable to sanction even before their cases are adjudicated and even if their cases are eventually dismissed (Kohler-Hausmann 2013).

The Impact of Legal Liabilities on Employers' Hiring and Retention Practices
Although most attention on the collateral consequences of criminal justice involvement focus on its formal aspects, the result of statutory restrictions enforced by state or public actors, Benjamin Levin (2018) makes a compelling argument for researchers' focused attention on the informal, where nonstatutory legal frameworks shape the actions and decision making of private actors, namely, private employers. Specifically, Levin argues that private employers have also become key players in extending the punishments meted out to the justiceinvolved. The reason is simple. Because some of the liability they face is tied to their employees' criminal justice involvement, the legal system incentivizes employers to both surveil and sanction their employees' behavior, both on and of the job. In efect, they become agents of criminal justice.
The primary routes through which the penal system does so are tort doctrines of negligent hiring and retention (Glynn 1998;Holzer 1996;Holzer et al. 2001;Levin 2018). These doctrines link employers' liability to employees' identities and histories of criminal justice involvement, making employers responsible for employees' misconduct wherever this misconduct occurs. As examples, employers can be held liable if, without investigating potential red flags, they hire someone who then goes on to cause harm in the workplace. Employers can be held responsible for negligent retention if they retain employees with reputations for misconduct that could put others in harm's way, even if patterns of misconduct occur outside the em-ployment context and regardless of whether the employer has acted to prevent employees' misconduct on the job.
Such liabilities create strong incentives for employers to discover and weed out applicants with criminal records. According to Ben Goldberg (2016), 72 percent of companies perform background checks on every new employee; among these, 82 percent screen potential employees for their criminal histories. They do so in good part to avoid claims of negligent hiring, but also to avoid other threats to the workplace. 7 But these liabilities also incentivize employers to monitor and sanction, where necessary, the behaviors of current employees, both on and of the job, for signs of misconduct. Levin points to employee conduct policies that ban criminal conduct that occurs of duty and away from the workplace, which, along with or instead of state sanction, privately punishes individuals. Both sets of actions are attempts to safeguard against being held legally responsible for their employees' actions. To do otherwise could be quite costly. Levin points to a 2001 study revealing that the overwhelming majority of employers lose negligent hiring cases that go to trial, settlements averaging more than $1.5 million. Although it is unclear just how much such doctrines indirectly depress the employment opportunities available to the justiceinvolved by disincentivizing employers to hire and retain justice-involved individuals, one doubts that their efects are negligible.

Employers' (and Job Contacts') Aversion to Justice-Involved Individuals
Employers do not want to hire individuals with criminal records. In absolute terms, employer surveys indicate that more than 40 percent would not knowingly hire them, and fewer than 12 percent report that they would definitely hire them (Holzer 1996;Holzer et al. 2007). These figures are likely overestimates of employers' willingness, however. Devah Pager and Lincoln Quillian (2005) report that many employers who say that they are willing to hire individuals 7. The use of criminal background checks has skyrocketed, despite the significant flaws associated with the system, including inaccurate criminal records data that would almost certainly disqualify applicants who might otherwise be great prospects (Duane, Reimal, and Lynch 2017). 8. Many, however, is not all. For instance, in industries-such as manufacturing-where workers have little contact with customers, employers express a greater willingness to hire them. During tight labor markets, too, employers' willingness to hire increases modestly, as evidenced by the greater demand during such times for these job seekers (Holzer et al. 2001(Holzer et al. , 2007. Indeed, recent reports indicate that, in the expanding markets and low unemployment of the current labor market, and given public discussion about the need to reduce the barriers to employment that individuals with criminal records have (Garsd 2019), employer views about hiring justice-involved individuals might be changing, to the benefit of those with records. 9. Here again, exceptions exist because some employers prefer to hire recently released former prisoners because they perceive them as more motivated, hard working, and willing to perform dirty tasks (Pager 2003).
with criminal records are no more likely to do so than those who report being unwilling. 8 In relative terms, employers much prefer applicants without criminal records to those who have had criminal justice contact (Schwartz and Skolnick 1964;Boshier and Johnson 1974;Pager 2003). Two studies are worth highlighting. One, by Richard Schwartz and Jerome Skolnick in 1964, revealed that any type of criminal justice contact would reduce the odds of employer interest in hiring lesser-skilled workers. The authors designed a field experiment in which one hundred employers were randomly assigned to one of four treatments: an application from someone with no criminal record, an arrest for assault with acquittal and letter from judge, an arrest for assault and acquittal without a letter, and an arrest and conviction for assault. All else about applicants' details were the same. Rates of positive responses by employers varied substantially by level of criminal justice contact. Thirty-six percent of employers who received applications indicating no criminal record responded positively, relative to 24 percent of employers who received applications indicating acquittal with a letter, 12 percent who received applications indicating acquittal without a letter, and 4 percent who received applications from those convicted. Thus, even when criminal justice contact did not result in conviction, Schwartz and Skolnick's research indicates that it would significantly diminish individuals' employment prospects.
Devah Pager's (2003) field study has advanced our understanding of the efect of the criminal record on employment outcomes by highlighting how felony conviction status and race afected employer callback rates. She discovered that whites without a criminal record were twice as likely to receive a callback than equally qualified whites with a felony convic-tion. Among blacks, the ratio was worse, at 3:1. Further, blacks without a felony conviction appeared no more likely than whites with a conviction to get a positive response. Regardless of race, then, having a criminal record has a substantial and negative efect on making it through the hiring process, but those odds are far worse for blacks than for their white justiceinvolved counterparts (see also Pager, Western, and Sugie 2009;Pager, Western, and Bonikowski 2009). 9 As discussed earlier, some of employers' reluctance to hire can be attributed to fears of being found liable for negligent hiring, but employers are also clearly driven by a general distrust of a pool of applicants who have essentially been certified untrustworthy by the penal system (Schwartz and Skolnick 1964;Boshier and Johnson 1974;Holzer et al. 2007; Ispa-Landa and Loeffler 2016; but see Atkin and Armstrong 2013), a negative efect amplified by the race of justice-involved individuals (Pager 2003(Pager , 2007. Most employers generally perceive the justice-involved to be too risky to trust with business operation and assets (Schwartz and Skolnick 1964;Boshier and Johnson 1974;Pager 2003;Pager and Quillian 2005;Holzer et al. 2007).
Referrals from trusted individuals, such as current employees, can go far in convincing employers to hire the justice-involved (Fahey et al. 2006), but it is unclear under what circumstances current employees would act in this capacity for job seekers tainted by a criminal record. After all, the stigma that informs employers' perceptions of former prisoners likely also shapes the way potential job contacts perceive them. Indeed, public opinion about exofenders tends to be quite negative. A number of older studies indicate that because most contacts would be uncomfortable having formerly incarcerated individuals as coworkers, fearing 10. Alfred Blumstein and Kiminori Nakamura (2009) estimate a "point of redemption" and find that for individuals with nonviolent offenses, it takes roughly 4.8 years without a new arrest to achieve a risk of re-arrest that is comparable to, or lower than that for, an individual of the same age in the general population; for individuals with violent offenses, eight years clean achieves a risk similar to that of the generation population (see also Bushway, Nieuwbeerta, and Blokland 2011). Few employers know about these points of redemption, however, and many laypeople would struggle to provide job-matching assistance to the justice-involved without evidence that an individual's risk of reoffending had declined significantly (Smith 2018). their own safety and that of their coworkers, they would also be unwilling to help them during the hiring process (Conklin 1975;Kutchinsky 1968;Simmons 1965). This is because diverse populations see ofenders as outsiders, low-class, unattractive and prone to violence (Reed and Reed 1973;Roberts 1992Roberts , 1997Saladin et al. 1988;Shoemaker and South 1978;Simmons 1965). They also assume that those who have been convicted of crime have several priors and are quite likely to commit new crimes in the future (Roberts 1997). In the minds of most, past is prologue.
To varying degrees, then, concerns about former prisoners' trustworthiness and fears about their risk of reofending would likely animate the thoughts of many potential job contacts and afect their decisions to act as personal intermediaries. After all, individuals' perceptions of the justice-involved also afect their behaviors toward them. In general, although flexible and modifiable (Cullen et al. 1990;Flanagan and Caufield 1984;Sandys and McGarrell 1995), the American public's attitudes about crime and punishment tend toward punitiveness. Despite some support for rehabilitative measures, including counseling and educational and vocational training programs for individuals with some types of offenses, the last four decades have been characterized by a general tenor of harshness and a clear focus on retribution, constant surveillance, and incapacitation (Garland 2001;Mc-Corkle 1993;Simon 2014). Still, levels of support for such punitive criminal justice policies and practices vary, and variations have been linked to cultural frames about punishment, deterrence, rehabilitation, and racial injustice (Bobo and Johnson 2004;Bobo and Thompson 2006;Cullen et al. 1990;McCorkle 1993).
But cultural frames-the meanings that we attribute to events or situations-do more than shape support for criminal justice policies.
They also shape whether or to what extent potential job contacts help formerly incarcerated job seekers to find work. Drawing from indepth interviews with 126 racially and ethnically diverse jobholders at one large, public sector employer, Sandra Smith (2018) shows that jobholders' willingness to assist depended in good part on whether they embraced one of two cultural frames-the second chance frame and the signaling change frame. Jobholders who argued that all individuals were capable of change and entitled to more chances to prove themselves were strongly inclined to help the formerly incarcerated to find work. Jobholders who tended to be noncommittal either referenced the nature of ofenses for which job seekers were punished-a proxy for their ability to change-or evidence that job seekers had changed-a proxy for former prisoners' commitment to do better. Without strong evidence that job seekers had been rehabilitated and become oriented more toward the job market, these jobholders were too uncomfortable to offer much in the form of job-matching assistance. Racial and ethnic background also informed the extent, nature, and quality of jobholders' experiences with the formerly incarcerated and shaped which frame or set of frames jobholders deployed when considering whether they might provide help with job search.
The findings reported in Smith (2018) suggest that, as with employers, formerly incarcerated individuals' friends, relatives, and acquaintances also want assurances that they have changed, that they are committed to work in the formal wage economy, and that they are no more likely than those without criminal records to cause harm to the physical, financial, or reputational well-being of the workplace. 10 Without these assurances, potential job contacts would be far less open to ofer help, even when they are able to do so. Such assurances, however, are difficult to come by. Potential job contacts do not often know enough about job seekers to make informed decisions about who is work-ready and who, for whatever reasons, might not yet be there, increasing the odds of facilitating poor employment matches.
Community corrections could ofer a remedy for the asymmetric information problem faced by the friends, relatives, and acquaintances. Potential job contacts need strong and efective signals to diferentiate between those who are and are not work-ready; additionally, through transitional employment-based reentry programs, members of probation and parole departments can be trained to provide the institutional supports needed for diferentiation to happen. The point of a signaling program is not to efect desistance but to identify those who are likely to be good bets for desistance and employment stability. As Shawn Bushway and Robert Apel (2012) contend, even if such programs do not have a causal impact on employment and recidivism, because those who choose to participate and complete such programs are likely distinct from those who do not, the programs can play an important role in the reentry process by signaling who is workready and who is not. In the process, they provide employers and potential job contacts with the information they need to make hiring and job-matching assistance decisions, respectively.
To be an efective signal of change and workreadiness, however, signaling programs would have to have at least three key features. First, they would have to be voluntary, because interested parties would have difficulty assessing change and work-readiness among those coerced into participation. Second, because work program requirements are likely more costly to individuals who are not yet ready to give up criminal ofending or to earn income legitimately, opportunity costs must be high enough to weed out those who are not yet ready but not so high as to discourage participation among those who are truly committed. Third, participants must actively participate and successfully complete the program. According to Bushway and Apel (2012, 33), these "explicit mechanisms . . . will allow the 20% to 30% of returning prisoners who have desisted from crime to self-identify to employers or other decision makers[potential job contacts] as soon as they step beyond the prison walls, or very shortly after release from prison." After voluntary enrollment, active participation, and successful completion, program graduates can receive a certificate of completion-perhaps ofered at a formal graduation that family members and friends can attend-that they can then share not only with employers but also with potential job contacts from their personal networks. Further, not only should employers be encouraged to hire those certified as work-ready, as is already done in many probation and parole departments across the country, community members should be encouraged to help as well, providing whatever assistance they can, including word of mouth, to help graduates to find work.

Consequences for Justice-Involved Individuals' Employment and Wages
In part because of the legal and social barriers formerly incarcerated individuals face, their probability of being employed is reduced significantly (Freeman 1991a(Freeman , 1991bWaldfogel 1994;Western 2006), and they work many fewer weeks per year (Freeman 1991a(Freeman , 1991b. Indeed, almost two-thirds remain without work one year after release (Travis 2002;Finlay 2008). It is not only the formerly incarcerated, however, who have diminishing employment prospects. Several studies also link arrest, conviction, and probation to employment problems as well. For instance, using data from a random sample of Philadelphia residents, Terrence Thornberry and R.L. Christenson (1984) sought to test the hypothesis that crime and the factors presumed to cause it (that is, unemployment) are actually instead mutually reinforcing. The evidence supported their hypothesis. Not only did unemployment appear to increase the probability of getting arrested, but having been arrested also increased the probability of being unemployed. Grogger's 1992 study had two objectives-to determine the efect that arrest had on employment and to ascertain the extent to which black-white gaps in employment could be explained by racial diferences in their rates of arrest. Analyzing two datasets-merged arrest and employment records of men arrested in California and the 1980 wave of the National Longitudinal Survey Youth Cohort-Grogger finds that arrest did reduce employment prospects. Further, racial diferences in arrest rates also helped explain why blacks had lower rates of employment relative to whites. In a more recent study, Bushway (1998) analyzes the National Youth Survey data and found that firsttime arrest had a strong and negative efect on employment stability, reducing by almost eleven weeks the amount of time an individual spent at their main job.
Several studies have also found a causal link between conviction and employment problems. Researchers have found that having a conviction can reduce the probability of employment by between 2 and 6 percentage points (Freeman 1991b;Waldfogel 1994). It can also negatively afect employment stability. For instance, Daniel Nagin and Joel Waldfogel (1995) examine the efect of conviction on the number of weeks and found that conviction added two additional weeks of unemployment.
Research also indicates that probation also negatively afects employment, reducing the probability of employment by up to 10 percentage points (Freeman 1991a;Waldfogel 1994). In this volume, Josh Seim and David Harding (2020) examine the efect of parole, not probation, on employment. Specifically, they investigate whether parole supervision increases the odds of employment and, importantly, whether employment while on parole reduces the odds of recidivism. Analyzing administrative data from the Michigan Department of Corrections matched with data on employment and earnings from Michigan's Unemployment Insurance Agency, Seim and Harding find that though parole supervision did increase the odds of employment, parole employment did not reduce the likelihood that individuals would recidivate, as has been hypothesized. That supervised parolees did not desist calls into question arguments that parole-related employment acts as a protective factor against future criminal involvement.
Perhaps, instead, the issue with parolerelated employment is one of job quality. Even when employed, the quality of jobs that are available to the justice-involved is generally poor (for a general discussion of bad jobs, see Kalleberg, Reskin, and Hudson 2011). Despite their general desire to work in jobs with regular hours, a guaranteed income, and benefits (Ispa-Landa and Loeffler 2016, 399; Sugie 2018; Kalleberg, Reskin, and Hudson 2011), the justiceinvolved are often relegated to informal, temporary, and part-time work (Ispa-Landa and Loeffler 2016). According to Christy Visher and Vera Kachnowski (2007), of the minority of former prisoners from Illinois employed within eight months of release, the overwhelming majority worked in construction, manual labor, and maintenance-among the few industries still generally open to hiring them-and earned between $750 and $900 per month; only among childless, single adults making the upper range would these earnings break the poverty threshold (see also Western 2002Western , 2006Western and Pettit 2005).
In this volume, Joe LaBriola (2020) explores the issue of post-prison employment quality and its efects on recidivism, using data on the employment outcomes of all prisoners paroled in Michigan in 2003. LaBriola finds that job quality matters. Parolees who found work in high-quality industries-those with above average earnings, job tenure, firm-level earnings, and state-level union coverage-were six times less likely to recidivate than parolees who found work in low-quality industries-industries below average on these four indicators. Further, those in low-quality industries were no more likely to desist than unemployed parolees. Insofar as low-income people of color are more likely to reside in and return to neighborhoods and communities where relatively few job opportunities exist, much less "good jobs," their probability of employment, regardless of penal disposition, will be further compromised and the likelihood of recidivism will be greater (Sabol 2007;Wang, Mears, and Bales 2010;Bellair and Kowalski 2011).

Reform Efforts
Reform eforts are under way, and they take many forms. According to the Institute for Justice, since 2015, eighteen states have reformed their occupational licensing laws to make it easier for ex-ofenders to find work in statelicensed fields. Reforms include new rules preventing licensing boards from blanket bans  Holzer, Steven Raphael, and Michael Stoll (2001) report that, ironically, when employers did conduct criminal background checks, black applicants had a higher likelihood of being offered employment because many fewer had criminal records than employers assumed. When employers did not check, they would assume that black applicants likely had criminal records and so would be less likely to offer them employment. Employers' willingness to hire ex-offenders affected the size of the gap. Among employers willing to hire ex-offenders, those who conducted criminal background checks were only 4.8 percentage points more likely than employers who did not to have recently hired a black applicant. Among employers who are unwilling to hire ex-offenders, the gap is larger. Those who run background checks were 10.7 percentage points more likely to have recently hired a black applicant than employers who did not check. In general, Holzer and colleagues find this pattern of decision making for stigmatized workers-welfare recipients, high school noncompleters, workers with spotty work histories, and the long-term unemployed.

Harry
(ten states), instituting new reporting requirements (five states), prohibiting the use of vague language related to moral character (four states), and allowing individuals to petition licensing boards at any time to determine whether their record will disqualify them from being licensed (eight states). 11 Further, to remove some of the barriers to employment that the justice-involved face, more than 150 cities and counties, thirty-four states, and the District of Columbia have also adopted ban-the-box (BtB) policies, a part of a larger fair chance employment initiative. These policies, which began appearing in the mid-2000s, are intended to improve employment outcomes of the justice-involved by preventing employers from taking criminal records into consideration until the latter stages of the hiring process (Doleac and Hansen 2016; Rodriguez and Avery 2016; Vuolo et al. 2017). For instance, California's new statewide policy, efective January 2018, prohibits employers from inquiring about arrest or conviction records until after a conditional job ofer is made. Thus, employers, public and private, are disallowed asking whether applicants have a criminal record on applications or during interviews; nor are they allowed to run a criminal background check until after a conditional ofer is made. With criminal records in hand, employers cannot discriminate at all about arrests that did not lead to convictions, participation in diversion or deferral of judgment programs, or convictions older than seven years (a few exceptions aside). If, after background checks, em-ployers decide against completing the hiring process, they must provide the applicant with a copy of the background check, explain how the conviction informed the hiring decision, and provide the applicant with an opportunity to respond.
To date, research on the efectiveness of such policies is limited. What does exist suggests that such policies do afect employers' behavior, increasing the callback rates and hiring for people with criminal records (Atkinson and Lockwood 2014;Agan and Starr 2016;Berracasa et al. 2016;Shoag and Veuger 2016). However, whether BtB works appears contingent on the race of the applicants and the sector of employment in question. For instance, in the private sector, BtB policies actually reduce the likelihood that justice-involved applicants who are black and Latino receive callbacks or are offered jobs (Agan and Starr 2016;Doleac and Hansen 2016). Consistent with prior research (Holzer, Raphael, and Stoll 2001), this finding suggests that when a person's criminal history is unavailable, employers in the private sector engage in statistical discrimination. 12 In the public sector, however, BtB policies increase the likelihood that formerly incarcerated individuals will receive callbacks and job ofers (Craigie 2017). Still, the efects of BtB policies are relatively small, and it is therefore not clear how much such policies can improve the labor market outcomes of justice-involved job seekers. Significant limitations in the scope and reach of such policies, the general lack of awareness that the justice-involved have about such policies, and the lack of accountability systems to ensure employers' compliance, will almost certainly limit their efectiveness.
The reform of laws that allow individuals to seal or expunge their criminal records (expungement) is another pathway by which some have sought to block the devastating efects of formal and informal collateral consequences. Expungement limits public access to arrest and conviction records, thus allowing the justiceinvolved to pursue opportunities, including employment and housing, that might otherwise be of limits. Today, the overwhelming majority of states have procedures in place to permit sealing or expunging criminal records. Typically, one must petition the court, often after a waiting period, and judges are empowered to decide, with determinations based on the type, severity, or number of convictions on an individuals' record. Expungements, however, can also be automatic or, on request, mandatory.
In an era of mass criminalization and easy public access to those records, various eforts are afoot to deploy expungement as a weapon against the collateral consequences of criminal records (Roberts 2015). Perhaps pushed by advocacy organizations and think tanks, numerous states have considered either adopting new expungement laws or expanding those they already have. In 2018 alone, twenty states reformed their expungement laws to further limit public access to criminal records. Pennsylvania was one, becoming the first state to automatically expunge criminal convictions of minor, nonviolent misdemeanors after ten years of desistance (Prescott and Starr 2019). Multiple efforts are under way, however, to expand the number of expunged cases. As Jenny Roberts (2015) explains, public defenders, civil legal aid offices, and reentry clinics are increasingly aiding clients through the expungement process, and law school training is more likely now to include modules on expungement.
Increasingly, too, researchers are studying how sealing and expungement afects individuals' outcomes. Early evidence indicates that expungement makes a significant diference, offering the justice-involved a second (or first) chance in a social, political, and economic landscape that is otherwise extremely difficult to navigate. In a compelling new study, for instance, J. J. Prescott and Sonja Starr (2019) study the efects of expungement laws on rates of recidivism and employment outcomes. They combined criminal records data from the Michigan State Police-which included individuals whose criminal records had been set aside because of expungement as well as the criminal history records for similarly situated individuals whose convictions were not set aside-with detailed wage and employment data from the state's unemployment insurance program for the same individuals. Although few who were eligible applied for expungement ( just 6.5 percent), among those who did, set-aside crime rates were comparable to that of the general population. Their rates of recidivism were so low that they could not be considered a public safety threat. In addition, set-asides were more likely to be employed and had better jobs than their counterparts whose records had not been expunged (Prescott and Starr 2019).
Still, in an era when criminal records are readily available and at low cost online, expungement laws will likely be of limited utility, unless a number of other steps are taken. According to Roberts (2015), to maximize the effectiveness of expungement laws, we must also guarantee the accuracy of criminal records, restrict how decision makers can use criminal records, and efectively regulate companies that profit from providing access to such records.
Policy interventions that seek to improve the employment prospects of the justice-involved should also shape how well they can support their families. In this volume, Allison Dwyer Emory and her colleagues (2020) draw from both the Fragile Families Study data and a dataset of state employment protection laws to examine the efect of incarceration on noncustodial fathers' ability to contribute formally and informally to their children's household economy. As might be expected, the authors find that incarceration reduces these odds, but they also report that state policies implemented to protect the justice-involved from employment discrimination moderated these efects in noteworthy but complicated ways.
14. It is unclear, however, whether or to what extent this pattern of job search-high levels of search intensity at the early stages of search followed by a significant curtailment of search activity-is unique to formerly incarcerated individuals. Examining the job-search strategies of unemployed, white-collar job seekers, Ofer Sharone reach and more teeth will be needed to fundamentally improve the employment prospects of the justice-involved. The Legal Action Center advocates taking four steps to fully embrace fair chance hiring. These include eliminating unreasonable legal restrictions on hiring, improving background checking systems and processes, providing strong antidiscrimination protections for workers with criminal records, and limiting negligent hiring liability for employers who follow fair hiring practices. 13 Anything short of these combined steps will not likely make a huge diference in the employment chances of the justice-involved. Further, other labor market institutions might militate against incarceration for individuals otherwise at high risk of criminal justice contact. In this volume, Bryan Sykes and Amy Bailey (2020) examine how military employment afected the risk of confinement for black and white men. They show that among black, male, high school dropouts, veterans' status was a protective factor against contact with the criminal justice system. Black veterans without a high school diploma were significantly less likely than their nonveteran counterparts to experience incarceration. The authors point to evidence of "institutional castling," where shifts in the prominence of competing institutionsthe military and the criminal justice systemafected the underlying risk of military enlistment and penal confinement for different demographic groups. This important work highlights the role that various institutions have played, independently and in concert, in profoundly shaping the social and economic fortunes of society's most vulnerable.

Implications for Job Search
Given the number and nature of the demandside constraints described, job search costs are unquestionably higher for justice-involved job seekers (McCall 1970). They would have to expend more efort to find work than those who have not had criminal justice contact, and much more efort for "good" jobs. They know this. The majority seem well aware that penal contact substantially diminishes the quantity and quality of employment opportunities, and so they expect that job-finding will be very difficult (Sullivan 1989;Harding 2003;Visher and Kachnowski 2007;Gofman 2009). In general, such perceptions among job seekers are associated with reduced search efort and withdrawal from search entirely (Kanfer, Wanberg, and Kantrowitz 2001); in this way, the justiceinvolved are no diferent.
A growing body of research suggests that because of discouragement borne from the anticipation of stigma, and because of frustration borne from early job search failures, many do not put in the amount of efort required to find a job (Apel and Sweeten 2010;Sugie 2018). Using the NLSY97, for instance, Robert Apel and Gary Sweeten (2010) investigated the factors that lay behind incarceration's apparent efect on employment outcomes, contrasting the experiences of convicted young men who had been incarcerated with convicted young men who had not. They showed that formerly incarcerated young men were less likely to be employed in good part because they were less likely than their nonincarcerated counterparts to search for work. For Apel and Sweeten, this detachment from the labor market contributed significantly to the lower wages that formerly incarcerated individuals earned when employed. Time without employment further eroded the skills, education, and training they brought to the labor market, which negatively afected wage outcomes as well. More recently, Naomi Sugie (2018) reports that immediately after release from prison, the formerly incarcerated in her sample overwhelmingly searched for work, but within one month their search eforts plummeted, likely also the result of frustration and discouragement (see also Visher and O'Connell 2012). 14 Finally, to investigate whether and how criminal justice contact-arrest, conviction and incarceration-altered search patterns and, through search, afected search success, Sandra Smith and Nora Broege (2019) analyze the 2001-2011 panels of the 1997 cohort of the National Longitudinal Survey of Youth (NLSY97). Focusing solely on men, who still make up the vast majority of those who have had penal contact, they examine whether and how young, justice-involved blacks, Latinos, and whites searched for work. In the process, the authors implicate both nonsearch engagement and the use of inefective search methods in job seekers' relative lack of success in finding jobs. After penal contact, individuals were less likely to search for work; for whatever reasons, they appeared to detach from labor-force participation. Those who did search tended to use fewer methods of job search and abandoned search methods that were more efective and efficient at producing jobs, such as direct application. 15 This resulted in less successful job search episodes and more frequent unemployment. Smith and Broege also show that whether and how individuals searched mattered not only for former prisoners, but also for arrestees and nonincarcerated convicts. For the full sample, all three penal dispositions showed patterns of search that difered from the search eforts observed before contact with the criminal justice system. Further, these changes in job search patterns contributed significantly to justice-involved individuals' lower odds of search success, especially for blacks. Thus, although we continue to study the proportion of justice-involved individuals who continue to search for work, more research needs to be done to better understand the process by which some individuals opt out of labor-force participation altogether or alter their search patterns to the point of inefectiveness. e x tr aC tIng a nd e x pLoItIng L a bor a f ter CrIMIn a L Jus tICe ContaC t Over the past three decades, the bulk of research at the intersection of labor market out-comes and criminal justice contact has focused on how such contact contributes to individuals' exclusion and marginalization from labor market opportunities, as well as the short-and long-term consequences borne from such exclusions. Historically speaking, however, labor extraction and exploitation after contact have been the rule, servicing not only the needs of political and ideological forces (Garland 1990), but also of economic production (Rusche and Kirschheimer 2009), producing not only goods but also potentially more tractable, docile, and useful workers (Foucault 1977;Melossi and Pavarini 2018). Early criminologists and religious promoters of improving the habits of the poor perceived idleness as the major cause of criminality in the poor and advocated hard physical labor and tight regulation on the leisure lives of the poor to stem the problem (Melossi and Pavarini 2018, 121). From its origins in the nineteenth century (McLennan 2008), U.S. prisons across the country sought to defray a significant portion of the total costs of imprisonment through forced extraction of labor from prisoners. In some instances, the prisoners were contracted out to private employers; in others, prison officials supervised the labor and sold the output. Outside the South, the era of profitable labor began to decline in the late nineteenth century as unions and private competitors won legislation limiting the market for prison-made goods. In the South, however, the end of slavery led to the widespread adoption of the convict lease system in which prisoners, essentially only black prisoners, were leased out to private contractors for ruthless exploitation (Lichtenstein 1996;Oshinsky 1996). Even after the convict lease system ended at the turn to the twentieth century, many southern and western states retained an extensive focus on agricultural labor. It was not until the federal courts began to intervene against the harsh conditions in the 1970s that this changed in meaningful ways (Feeley and Rubin 1998). As mass imprisonment took of at (2013) finds a similar pattern in the United States (but not Israel, his other site of study) and linked this to the messages job seekers received in the labor market institutions in which they were embedded.
15. Post-contact, job seekers are generally less inclined to apply directly to employers, to go-it-alone, despite greater odds of search success when using this method. In prior work, David Harding (2003) suggests that the end of the twentieth century, labor in most prisons was limited to service functions for the prison itself and deemphasized as a way of either paying costs or "training" prisoners. However, as Eric Hatton (2018) cogently argues, most prisoners in most states are still working and the labor they perform is essentially forced labor.
Insofar as punishment, or the threat of it, has compelled individuals to work specifically in service of economic production, the criminal justice system acts as a labor market institution, facilitating the reallocation of labor and doing so on terms quite favorable to economic and political elites. In the following section, we discuss eforts to compel or coerce labor from the justice-involved, in and outside the prison context, where work, often uncompensated or poorly compensated, is used as punishment or as the threat of punishment.

Prison Labor
Georg Rusche and Otto Kirschheimer (2009) were perhaps the first to put forward a clearly articulated thesis, neo-Marxist in nature, that the criminal justice system operated as a labor market institution. In Punishment and Social Structure, first published in 1939, they drew on historical material that spanned the Middle Ages to the turn of the twentieth century to explain why specific forms of punishment were deployed or cast aside in specific historical moments. Embedded in their question was the assumption that, although all systems of punishment were at least in part a response to crime, they were also always designed in response to some other social, political, and, perhaps most important, economic project, and in particular projects aimed at controlling the behavior of the poor.
It is through this theoretical lens that we are to understand the birth of the modern prison. The house of correction, which Rusche and Kirschheimer (2009) describe as combination poorhouse, workhouse, and penal institution, emerged in response to problems of early capitalism. Toward the end of the mercantilist period, markets were expanding rapidly at home and abroad, but population stagnation meant that labor reserves were too low to meet demand. There were simply not enough workers, at the wages and work conditions that labor could insist on, empowered in the moment by tight labor markets. With the upper hand in labor negotiations, workers would either garner the high wages and improved working conditions they desired, or they could choose to withhold their labor power. For capitalists, the situation was untenable.
According to Rusche and Kirschheimer, to resolve their labor shortage crisis, capitalists sought assistance from the state. The state responded in numerous ways, including the deployment of relatively new forms of punishment, which essentially made withholding one's labor power impermissible. In addition to galley slavery and transportation, Rusche and Kirschheimer ofered the house of correction, whose purpose was to force unwilling, able-bodied people to work. By exploiting labor and training new labor reserves, early prisons helped solve problems inherent to early capitalism, and, in the process, to varying degrees served the needs of the penal institution itself, of capital production, and of the national economy.
justice-involved men who successfully searched alone adopted different impression management strategies, including failing to disclose their criminal justice status, in an effort to either completely eliminate the negative consequences for employment of having a criminal record, or to blunt its negative effect, and he linked impression management strategies to individuals' employment outcomes. Not everyone can so easily manipulate employers' impressions for their own benefit, however; race matters in how employers respond when justiceinvolved men disclose (Pager 2007). Increasingly, too, the justice-involved have much less control over the impressions they make. As employers gain ever greater access to applicants' personal records (SEARCH Group 2005; Bushway and Sweeten 2007)-criminal records and credit histories-it would seem that attempts to manage impressions by not disclosing one's contact with the penal system will fail to achieve its intended goal, since most employers are both loathe to hire the justice-involved and strongly inclined to verify their status and creditworthiness. For this reason, despite the potential benefits associated with using this method of job search, job seekers may perceive that with less control over the impressions they make, they are less likely to get the results they are hoping for. During the Industrial Revolution, efficient systems of mass production and efectively organized labor movements made precarious the role that houses of corrections played in helping to produce new labor reserves easily exploitable for industry, and so prison work was transformed from a productive, profitable form of training into a rational system of deterrence. Indeed, Rusche and Kirschheimer (2009) predicted that insofar as the prison had become obsolete in solving important problems in the economic realm, it might also soon be cast aside as a useful form of punishment. It did not.
Building on Rusche andKirchheimer (2009), historian Michel Foucault (1977) suggests that even where the prison abandoned the goal of economic productivity, the core elements of the prison-its cellular structure and its emphasis on precise controls over the body-reproduced the disciplinary form of power that was itself essential to the emergence of capitalist production and mass democracies. Dario Melossi and Massimo Pavarini (2018) point as well to the merging of punishment and a capitalist vision of the laboring body. The prison was less a factory to produce goods than a factory to produce obedient and pliable labor.
Despite some challenges, prisons have avoided the extinction that Rusche and Kirchheimer had predicted. Indeed, prison industries have been thriving. As evidence, from 1980 to 1994, although the number of federal and state prisoners increased by 221 percent, both the number of inmates employed in prison industries and the amount of prison industry sales grew by more than 350 percent, the latter growing to $1.3 billion (Erlich 1995). Whether operating in prisons or in collaboration with prison industries, corporations are clearly and increasingly profiting from punishment.
Still, not everyone in prison engages in productive labor. One of the distinctive features of mass incarceration is the frequency with which idleness rather than overwork now constitutes one of the pains of imprisonment (Travis, Western, and Redburne 2014). The overwhelming majority of those who do work have "regular" jobs that do not generate profit for industry. Instead, they provide institutional support in the form of maintenance, food preparation, laundry services, and the like, work that aids in the operation of the prison (Sawyer 2017). For these nonindustry jobs, prisoner-workers in 2001 were paid between $0.86 and $3.45 per hour, a decline because a growing number of states, all southern, chose not to pay inmates anything for performing in these positions (Sawyer 2017). 16 Work release programs, work camps, and community work centers ofer additional options for low-risk inmates and those preparing to be released, and in these situations it is not unusual for employers from the community to pay inmates' wages.
Roughly 6 percent of prison inmates work in state-owned businesses, which provide goods and services that are typically sold to government agencies (Sawyer 2017). UNICOR, the trade name for the Federal Prison Industries program, is perhaps the oldest such organization. Established in 1934, it currently has more than seventy factory locations across the United States with a diverse portfolio of products and services. 17 Through UNICOR, roughly eighteen thousand prisoners annually produce apparel and accessories, electronics, eyewear, food service products, license plates, office furniture, print products, awards and plaques, mattresses and bedding, office seating, and signage. The services they ofer include computer-aided design; data services; distribution, warehousing, and logistics; electronic recycling; call center solutions; and printing and bindery services. 18 In 2017, from more than $450 million in sales these operations combined to yield almost $17 million in net income, an increase of roughly $13 million in net income from the prior fiscal year. Meanwhile, during the same period, stateowned businesses like UNICOR paid their prisoner-workers on average between $0.33 cents and $1.41 per hour. 19 Although these amounts are approximately twice what the same prisoner-workers would make in regular prison jobs or privately run facilities, the significant profits they help generate would seem to warrant a greater piece of the prison industries pie. Unlike in Europe, where human rights conventions require that prisoners receive minimum wage, the United States has no laws protecting prisoners against forced labor, and, notoriously, the Thirteenth Amendment of the U.S. Constitution, which bars "involuntary servitude" or slavery, explicitly excludes those convicted of crimes.
Finally, a small number of incarcerated people work for businesses that contract with correctional agencies through the Prisoner Industry Enhancement Certification Program (Bureau of Justice Assistance 2018), a program that allows private companies to run aspects of their business operations inside prison walls using prison labor. At least thirty-seven states have legalized this practice, and a number of prominent companies are taking advantage. 20 Although these companies are mandated to pay prisoner-workers the "prevailing wages" for these jobs in the local area (Pelaez 2018), in the end, workers might see only a small fraction of the gross. Up to 80 percent of their earnings will likely be deducted mandatorily for various fees, including deductions for court-assessed fees, fines, and surcharges; victim restitution; room and board; and family support (Sawyer 2017).
In such contexts, how do prisoners gain a sense of dignity, especially as it relates to work? In this volume, Michael Gibson-Light (2020) reports on the discursive strategies that inmates in one medium-security men's prison deployed to claim self-worth in prison. Drawing from ethnographic fieldwork data and eighty-two in-depth interviews with prisoners and stafers, Gibson-Light ofers the "sandpile of dignity" metaphor to describe a context in which work roles are constantly shifting and prisoners attempt to make claims to legitimacy in part by erecting symbolic boundaries between themselves and the prisoners in lower status jobs, those whose dedication, motivation, and work ethic could be challenged by virtue of their lower position. In the process, they are able to justify their standing in the ever-shifting prison-work hierarchy.
Given that prison labor typically lacks employment protections, does not qualify for minimum wage, and is ineligible for overtime pay, and that prisoner-workers are prohibited from organizing and collectively bargaining for improved wages and working conditions, they are stuck as one of the most vulnerable and exploitable sources of cheap labor in the Global North and their conditions appear comparable to some in the Global South. Indeed, many contend that with the loss of economic citizenship, prison labor in the United States is socially, politically, and economically equivalent to working in free enterprise zones in Africa, Asia, and Latin America, 21 a reality made all the worse by the fact that the vast majority of labor performed in prisons fails to translate into marketable or transferable skills on the outside. Thus, most will be released with almost no financial resources, despite hours of labor, and skills no more developed than when they arrived, making the prospect of finding work after release that much more difficult.

Coercing Labor from Nonincarcerated Workers
When considering the collateral consequences of criminal justice contact, we tend to think specifically of those consequences, formal and informal, that create barriers to employment, such as occupational licensing restrictions or employer liabilities for negligent hiring. As a small but growing body of research indicates, however, collateral consequences of criminal justice contact might also have the opposite effect, forcing work from the justice-involved under the threat of further punishment, including the loss of freedom (Zatz et al. 2016). To the extent that penal contact threatens individuals' economic citizenship rights, it also makes the justice-involved vulnerable to labor exploitation and extraction, and in the process, deepens the precariousness that many of these individuals already face.
In important new work, Noah Zatz and his colleagues tackle this very issue, identifying the routes of legal authority through which the justice-involved become vulnerable, and then highlighting the negative consequences for them and non-justice-involved coworkers, including deteriorating conditions of labor for the working class and the working poor generally. Probation and parole are one route. The vast majority of probationers and parolees are required to abide by multiple conditions to remain free from further sanction. Indeed, virtually no ofenders are released on probation without stipulation, and almost all (99 percent) had one or more conditions to their probation sentence (Travis and Petersilia 2001;Rainville and Reaves 2003;Siegel and Senna 2007). Of those released conditionally, approximately 84 percent were required to pay some sort of fee or fine; 35 percent were required to find employment or enroll in some type of educational or job training program-41 percent of those convicted of felonies and 27 percent of those convicted of misdemeanors (Bonczar 1997). The courts are the other route. As with probation and parole, the courts have the power to mandate that the justice-involved pay their criminal justice and child support payments or face the consequences. Zatz and his colleagues (2016) explain that to please the court or probation officers and to avoid sanction for noncompliance, including possible jail time, justiceinvolved individuals are required to pay or to find a job, a better job, or two or more jobs. For those who cannot pay their fines or fees, it is fairly typical for the courts to mandate community service to work of the debt or jail time.
These are not idle threats, and, unsurpris-ingly, they disproportionately afect low-wage workers of color. According to a 2016 report (Zatz et al. 2016), approximately nine thousand individuals nationwide were jailed for violating probation or parole requirements that they hold a job; roughly two-thirds of those sanctioned were black or Latino. Five percent of all fathers in major cities were jailed for failure to pay child support; the figure for black fathers was 15 percent. Further, African American fathers make up nearly 80 percent of those incarcerated by the child support enforcement system and are incarcerated at a rate ten times higher than other fathers. In Los Angeles, the authors report that between fifty thousand and one hundred thousand residents were courtordered to engage in community service to work of their legal debt. In other words, they must work to remain free. Most of these were from Los Angeles County, where residents are disproportionately of color and low-income. The problem, of course, is that many who face these mandates lack the resources to do better. Take, as an example, payment of legal financial obligations and child support. Amounts vary somewhat by state, but analysis of data from Washington State revealed court assessments ranging from a minimum of $500 (mandatory for all felony convictions) to a maximum of $256,257; the median amount assessed per felon was $5,254; the mean $11,471 (Harris, Evans, and Beckett 2010). Many justice-involved fathers are weighed down by child support arrears as well. One in five incarcerated parents has a child support obligation, and these, too, can be daunting. In Boston, for instance, parents entering prison with a child support case owed roughly $10,000 in arrears; upon release, their average debt had doubled (Thoennes 2002).
These figures are not inconsequential. The vast majority of justice-involved men and women are poor or near poor (Western 2006). According to Alexes Harris, Heather Evans, andKatherine Beckett (2010, 1776), white, Hispanic, and black male felons have median legal debt loads that are roughly 60 percent, 36 percent, and 50 percent of their annual incomes, respectively; mean debt loads are roughly 10 percent, 69 percent, and 222 percent of their average annual earnings, respectively (see also Evans t h e c r I m I n a l J u s t I c e s y s t e m a s a l a b o r m a r k e t I n s t I t u t I o n r s f : t h e r u s s e l l s a g e f o u n d a t i o n j o u r n a l o f t h e s o c i a l s c i e n c e s 22. These large debt liabilities can discourage work, even with the threat of jail. Harris, Evans, and Beckett (2010) report that 80 percent of their respondents found their legal debt obligations to be "unduly burdensome," and their heightened financial stress actually had the unintended consequence of reducing commitment to work and related search effort. Despite the possibility of jail time for nonpayment, some of their respondents chose not to work, relying on crime, or, where this was allowable for legal debtors, cash assistance. High child support arrears balances and aggressive enforcement policies have also been found to have similar effects on fathers' work effort (Pate 2002;Bartfeld 2003;Meyer, Ha, and Hu 2008;Cancian, Heinrich, and Chung 2009). Thus, in the form of heavy monetary sanctions, penal interventions can worsen before-search options to the point of eroding commitments to work and thus engaged job search. 2014). Thus, most simply could not aford to fulfill their obligations in the short term, and given the accumulation of interest on courtimposed sanctions, fulfillment over the long term is unlikely. Even small monthly payments could reduce take-home pay substantially, but most states have the authority to garnish much more-as much as 35 percent of wages for legal fines and fees and as much as 65 percent for back child support debt (Sandberg 2010). Such deductions from already-meager earnings would make it near impossible to meet other needs and obligations, such as buying groceries and paying rent. In this situation, even if already employed, poor or near poor parents would either have to accept community service that forced them to work for free, or they would have to serve time in jail, encouraging system avoidance (Brayne 2014).
In addition, the mandate to pay, work, or to go to jail puts pressure on those already in precarious economic circumstances to accept employment arrangements that they might otherwise forsake, including jobs ofering low pay, no benefits, few employment protections, and poor working conditions. 22 By deferring to employers and allowing them to report on whether the justice-involved workers are in compliance with work requirements, the state essentially bestows on employers the power not only to hire and fire, but to jail as well. Thus, workers might feel pressure to silence their dissent and to accept employers' demands, no matter how unreasonable or illegal, deepening the sense of precarity that many caught up in the criminal justice system already face. Recent empirical evidence suggests that this is the case (see Hatton 2018). In this volume, Noah Zatz and Michael Stoll (2020) explore the question, examining the extent to which child support enforcement techniques, especially punitive ones such as incarceration, afect noncustodial fathers' employment outcomes. Analyzing from the Fragile Families Study data, they find that in cities that rely heavily on incarceration sanctions, noncustodial fathers who are most vulnerable to incarceration work more hours and at lower wages than those who are not so vulnerable. Thus, their study is one of the first to provide empirical support for the hypothesis that the threat of liberty lost will push vulnerable individuals to take on jobs with significantly poorer work arrangements and, likely, poor work conditions as well. Through the criminal justice system, it seems, we have created multiple mechanisms through which to trap justiceinvolved individuals in poverty.
ConCLusIon: Is the unIted s tat es u nIqu e?
As is true of many policies, organizations, and institutions, the criminal justice system has a hand in shaping the flow of individuals, especially those in the lower classes, into and out of the labor market as well as afecting economic performance generally. The United States is not unique in this regard. Perhaps what is unique is the scale at which the penal state is deployed at least in part in service of a winner-take-all economy, where worker protections, social insurance programs, and active labor market systems are weak. In such a context, the twin actions of exclusion and extraction, though seemingly at odds, make sense. What is also unique is the degree to which, in an otherwise largely unregulated labor market, where workers receive few protections against unemployment and antidiscrimination laws are systematically underenforced, the criminal justice system stands out as an aggressive form of state intervention in the labor market. With criminal justice contact, individuals are automatically excluded from a large swath of employment op-