Child Support Enforcement

Soon after Ferguson, a police officer in North Charleston, South Carolina, shot Walter Scott in the back as he ran from a traffic stop. Scott feared the officer would see an arrest warrant for Scott’s nonpayment of child support (Robles and Dewan Reference Robles and Dewan2015). The state’s demand for child support also required that Scott get a job or go to jail, enforced by police action.

The work demands of child support enforcement reflect that system’s deep ties to welfare (Brito Reference Brito2012; Hatcher Reference Hatcher2007). In its efforts to “privatize dependency” (Fineman Reference Fineman1995), the welfare state requires claimants to exhaust resources both from their own wage income and from their kin (Hays Reference Hays2003, 19). A custodial parent’s access to benefits requires several forms of child support cooperation, including signing child support claims over to the state so that it can seize payments to offset its benefits outlay. The state thus acquires a financial stake in nominally “private” obligations between co-parents (Hatcher Reference Hatcher2007).

Front-line state and local child support enforcement operates under elaborate federal rules and infrastructure (Brito Reference Brito2012). This apparatus emphasizes noncustodial parents of children receiving means-tested public assistance, but it applies to all child support cases. About thirteen million noncustodial parents are included in this sprawling system (Office of Child Support Enforcement 2016, Tbl. P-101).

Child support cooperation requirements for custodial parents quickly yield work requirements for noncustodial parents. Most often, both custodial and noncustodial parents have quite limited employment prospects (Sorensen and Zibman Reference Sorensen and Zibman2001). In 2003–04, 70 percent of all child support arrears were owed by noncustodial parents with annual incomes below $10,000 (Sorensen, Sousa, and Schaner Reference Sorensen, Sousa and Schaner2007). Among low-income fathers in arrears, a plurality are African American (Sorensen and Zibman Reference Sorensen and Zibman2001). Contrary to a middle-class divorce story centered on the “deadbeat dad,” nonpayment (or underpayment) stems primarily from a lack of earnings to share, not a refusal to share the earnings in hand.

To make greater earnings available for payment, child support enforcement institutions directly demand work and evaluate work effort. For instance, the federal Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), the primary Clinton-era welfare reform statute, contains a little-studied provision requiring that state child support authorities be empowered to “issue an order that requires” noncustodial parents to participate in “work activities” (42 U.S.C. § 666(a)(15)); these are defined by cross-reference to the provision establishing PRWORA’s better-known work requirements for custodial parents receiving cash assistance.

This federal, welfare-related work enforcement infrastructure built upon older, less institutionally elaborate mandates. These include the standard enforcement tool of a “seek work order.” For instance, an Illinois statute provides that when a child support obligor is unemployed, “the court may order the person to seek employment and report periodically to the court with a diary, listing or other memorandum of his or her efforts in accordance with such order,” as well as to order participation in various workforce development programs (750 Ill. Comp. Stat. Ch. 5/505.1). This Illinois provision applies regardless of whether the supported child receives public assistance.

When a court directly orders an obligor to obtain work or participate in a “work activity,” noncompliance can be enforced with contempt sanctions, including incarceration. Unlike welfare, in child support enforcement there are no payments from the state to cut off as leverage; instead, the state seeks to extract money from the noncustodial parent.

Similarly, carceral sanctions ultimately enforce payment obligations themselves (Brito Reference Brito2012). Civil debt collection is ineffective against people lacking assets, income, and credit. Using incarceration, or its threat, to collect has given rise to what many critics characterize as a new form of debtors’ prison (Patterson Reference Patterson2008). Legally, incarceration for nonpayment is an available threat because traditional limitations on civil debt collection such as bankruptcy (Atkinson Reference Atkinson2017) and bans on imprisonment for debt (Hampson Reference Hampson2016) do not apply. Child support obligations arise not from private contracts but from family court orders enforceable through contempt proceedings that can authorize incarceration (Patterson Reference Patterson2008). Additionally, all states criminalize some child support nonpayment (National Council of State Legislatures 2015).

Even without a direct work order, work requirements are embedded in the legal principles governing what formally is incarceration for nonpayment. Constitutional doctrine forbids incarcerating someone for being too poor to pay (Colgan Reference Colgan2014; Birckhead Reference Birckhead2015). The legal touchstone is willful nonpayment: a choice not to pay despite the ability to pay.

Willful nonpayment includes willful failure to acquire the means to pay, mirroring the logic by which welfare means tests have long incorporated work and other behavioral considerations (Zatz Reference Zatz2012). New York courts routinely uphold incarceration for nonpayment, despite the obligor’s present lack of funds, because “the ability to pay support also includes the ability to find employment” (Commissioner of Social Services v. Rosen, 736 N.Y.S.2d. 42, 44 (App. Div. 2001)). For the poor, voluntary unemployment supplies part of the content of willful nonpayment. Carceral child support enforcement thus runs on the same analytical engine that has long driven the welfare state: distinguishing voluntary from involuntary unemployment, the undeserving from the deserving poor (Handler and Hasenfeld Reference Handler and Hasenfeld1991). Here, however, the baseline is lower: deservingness offers freedom from prison, not freedom from destitution.

The California Supreme Court illustrated these points in Moss v. Superior Court (950 P.2d 59 (Cal. 1998)), which reviewed a trial judge’s decision to imprison a child support obligor for contempt. The obligor had neither hoarded funds nor refused any job offer. Instead, the high court held that unemployment itself constituted contempt where someone “fails or refuses to seek and accept available employment for which the parent is suited by virtue of education, experience, and physical ability” (ibid., 76). The trial judge had reasoned that the obligor surely “could get a job flipping hamburgers at MacDonald’s [sic] …. He’s, in my mind, chosen not to” (ibid., 63).Footnote ⁴ Here the “criminalization of poverty” (Gustafson Reference Gustafson2011) entails specifically the criminalization of unemployment.

The factual setting underlying Moss was no outlier. Scattered reports suggest that incarceration for nonpayment can be substantial in some jurisdictions (Brito Reference Brito2012; Robles and Dewan Reference Robles and Dewan2015). Data from the Fragile Families and Child Wellbeing Study (FFCWS) support that conclusion at a national scale. The FFCWS is a longitudinal survey of mothers and fathers of a birth cohort of focal children, weighted to be representative of major cities in the United States (Reichman et al. Reference Reichman, Teitler, Garfinkel and McLanahan2001). Its detailed questions about child support obligations and enforcement reveal that, among noncustodial fathers subjected to any child support enforcement action, at least one quarter were incarcerated for noncompliance during the nine-year follow-up period. Moreover, such incarceration affected 5 percent of all fathers (custodial or noncustodial) and an astounding 15 percent of all Black fathers (Zatz et al. Reference Zatz2016; cf. Cozzolino Reference Cozzolino2018). Thus, obligors facing work demands have every reason to take seriously the threat of incarceration, and this threat touches a substantial proportion of the workforce, especially in low-income communities of color.

Incarceration data alone cannot reveal whether and how state actors evaluated obligor work in the processes leading to incarceration. Given the prominence of work requirements within the legal and policy frameworks discussed above, however, they likely did in many cases.

Because these work obligations derive from payment obligations, work is not the exclusive means of compliance. Payment of the underlying obligation, sometimes as reduced to a “purge” amount deemed sufficient to escape contempt, will moot the work requirement. Those funds could be obtained not only through wages but also through charitable support, family contributions, or illicit activities (Haney Reference Haney2018). This is not unlike other work mandates. For instance, in early twentieth-century peonage cases, criminal statutes offered workers indebted to their employers a three-way choice—work, pay, or go to jail. The jail sanction could be avoided through payment from nonwage sources (Zatz Reference Zatz2016). And in the welfare context, a custodial parent’s access to alternate sources of income—including child support itself—will vitiate work obligations if that income substitutes for means-tested benefits payments.Footnote ⁵

Thus, while there are additional dynamics of importance at play, work mandates are one significant feature of child support enforcement. When sanctions are sought against someone who does not pay, whether they are subject to incarceration will turn in important part on their work behavior. In addition to the general “ability to pay” concepts noted above, for instance, an Ohio statute explicitly provides that compliance with a seek-work order may provide a defense against sanctions for nonpayment (Ohio Rev. Code 3119.06(A)).Footnote ⁶

These work requirements have until recently (Zatz Reference Zatz2016) largely been overlooked within the limited literature on child support enforcement and incarceration, primarily among legal scholars (Brito Reference Brito2012; Cammett Reference Cammett2011; Patterson Reference Patterson2008; but cf. Chung Reference Chung2011). This scholarship has focused principally on the actual imposition of incarceration (including procedural protections such as the right to counsel, which the Supreme Court addressed in Turner v. Rogers, 564 U.S. 431 (2011)) or on the child support consequences of prior incarceration; Haney’s sociological research recently united those perspectives (Reference Haney2018). Brito also notes how jailing obligors for contempt may disrupt future employment (Reference Brito2012), consistent with barriers analyses.

Conversely, the econometric literature on child support enforcement and employment has centered work but overlooked incarceration. It generally models child support enforcement exclusively as a tax on earnings (Miller and Mincy Reference Miller and Mincy2012; Cancian, Heinrich, and Chung Reference Cancian, Heinrich and Chung2013), without analyzing how nonwork is penalized. The dominant taxation model predicts employment disincentives, again cohering with a barriers analysis (Holzer, Offner, and Sorensen Reference Holzer, Offner and Sorensen2005).

The potential for child support enforcement to pressure obligors into work has, however, received attention in social policy circles. Lawrence Mead, the most influential conservative academic proponent of work-based welfare reform in the 1980s and 90s, has retooled his arguments to support analogous work mandates for noncustodial parents:

Notably, Mead’s analysis is not specific to the child support system and its internal goals. Rather, he sees child support as one institutional lever for addressing what he conceptualizes as a broader social problem of inadequate labor discipline. That analysis explicitly invokes what Mead characterizes as a racial occupational pathology, of which child support nonpayment is merely one symptom: “Young black men will often refuse to work for ‘chump change’ even if it means not working at all. Or they accept jobs but then find them unrewarding or abusive, so they leave in a huff or are fired” (ibid., 18).

With work effort front and center, familiar issues from welfare-to-work reemerge, including disagreements about the causes of un(der)employment and disputes over voluntary versus mandatory employment services (Sorensen Reference Sorensen2010; Mead Reference Mead2010). Race persistently influences these domains (Brito, Pate, and Wong Reference Brito, Pate and Wong2015), affecting the services offered, the work deemed “good enough,” the value of caretaking that competes for time with paid work, and whether noncompliance is excusable or merits sanction (Bonds Reference Bonds2006; Roberts Reference Roberts1994; Soss, Fording, and Schram Reference Soss, Fording and Schram2011).

Notably, the Obama Administration promoted, and often funded, child support work programs of the sort encouraged by the integration of custodial and noncustodial parent work requirements established by the 1996 PRWORA; this includes mandatory programs that expose obligors to contempt and incarceration for noncompliance (Turetsky Reference Turetsky2012; Office of Child Support Enforcement 2014).Footnote ⁷ Here is visible a potential bipartisan consensus around institutionalized work mandates broadly similar to, and building on, welfare reform; the differences are a pivot toward men and a shift from financial to carceral sanctions.

Criminal Legal Debt

Child support obligations have much in common with fines, fees, and restitution imposed by criminal legal institutions (“criminal legal debt” or “legal financial obligations” (LFOs)). The scholarship is fairly new (A. Harris, Evans, and Beckett Reference Harris, Evans and Beckett2010; Colgan Reference Colgan2014; A. Harris Reference Harris2016), but the issue gained prominence after Ferguson (Civil Rights Division 2015) and has moved to the forefront of racial justice, criminal justice reform, and anti-poverty advocacy (Bannon, Nagrecha, and Diller Reference Bannon, Nagrecha and Diller2010; Johnson et al. Reference Johnson, Herzing, Hooks and Stahly-Butts2016). Those with outstanding debt likely number in the tens of millions (A. Harris, Evans, and Beckett Reference Harris, Evans and Beckett2010), reflecting and exacerbating broader race and class disparities in policing, prosecution, and sentencing.

Systematic empirical LFO research has focused on felony defendants (A. Harris Reference Harris2016), but the Ferguson investigations and emerging research highlight LFOs accumulated through misdemeanors and lower-level traffic or public order violations (Natapoff Reference Natapoff2015). When unpaid, these, too, can lead to arrest warrants and incarceration (Stuart Reference Stuart2011), as well as intermediate sanctions like driver’s license suspensions. Both the underlying debt and subsequent arrests disproportionately burden low-income communities of color (Bingham et al. Reference Bingham, Castaldi, Della-Piana, Desautels, Dozier, Harootun and Herald2016). As with historical studies of convict leasing arising out of municipal courts (Haley Reference Haley2016), incorporating attention to nonfelony LFOs both enlarges the population caught in their net and increases the representation of women (cf. Bumiller Reference Bumiller, Carlton and Segrave2013).

Like child support obligations, LFOs typically are no ordinary civil debts. Instead, the payment obligation ultimately is enforced by the threat of incarceration (Beckett and Murakawa Reference Beckett and Murakawa2012; A. Harris Reference Harris2016), whether through contempt, freestanding prosecution for failure to pay, or other legal devices. Again, carceral sanctions are governed by a “willful nonpayment” standard under both federal constitutional law and state statutes that limit incarceration to those lacking an “ability to pay” (A. Harris Reference Harris2016; Zhen Reference Zhen2019).

As with child support, one way carceral work mandates arise is through the work effort judgments incorporated into “willfulness”/“ability to pay” analysis (Colgan Reference Colgan2017; Zhen Reference Zhen2019). To date, this nexus has received little scrutiny in the burgeoning advocacy and research around LFOs. Legal challenges (American Civil Liberties Union 2018; Equal Justice Under Law 2018) have focused on widespread “auto-jail” practices (A. Harris Reference Harris2016) where nonpayment triggers incarceration without any individualized assessment of ability to pay. Tellingly, a recent LFO “ability to pay” lawsuit against the Los Angeles Superior Court sidestepped the issue by choosing plaintiffs receiving disability benefits premised on inability to work (Holland Reference Holland2016).

Today’s campaigns against the “new debtors’ prisons” keep the work question at bay by treating inability to pay as a fixed status: those incarcerated for nonpayment suffer this fate “simply for being poor” (Bannon, Nagrecha, and Diller Reference Bannon, Nagrecha and Diller2010, 21). Poverty status is separated from economic conduct, replicating the pattern of early welfare rights advocacy (Zatz Reference Zatz2012). This income/behavior distinction likely will come under pressure if “ability to pay”/“willful nonpayment” determinations become contested in practice. The concern will be “to ferret out which people … are not employed … because they don’t want to be employed or [instead] because they can’t find work” (A. Harris Reference Harris2016, 137). In this way, criminal legal debt offers a new stage for the “sturdy beggars” (Handler Reference Handler2004, 277) who have long haunted social welfare policy and whose invocation both reinscribes the work ethic and attributes poverty and criminal legal involvement to racialized personal pathology (Gans Reference Gans1994; Gustafson Reference Gustafson2013).

Several sources indicate how the work issue sits just below the surface, latent in the “ability to pay.” The relatively sparse LFO statutory provisions defining “ability to pay” sometimes refer to earnings capacity and future employment, not just present income (for example, Cal. Veh. Code § 42003), again echoing the better-established child support practice of setting orders based on imputed income. Although not focused on this point, Alexes Harris’s study of criminal legal debt found that corrections officers, prosecutors, and judges explicitly invoke judgments about work effort when determining whether nonpayment was willful (2016). One opined, in a statement mirroring the Moss trial court opinion, that “all nonpayment is willful because felons ‘can always go out and get a day job’” (A. Harris, Evans, and Beckett Reference Harris, Evans and Beckett2010, 1788 n. 28).

Indeed, such inquiries are invited by the touchstone of LFO “ability to pay” jurisprudence, the Supreme Court’s Bearden v. Georgia opinion (461 U.S. 660 (1983)). Bearden allowed incarceration only when a defendant “willfully refused to pay” or “failed to make sufficient bona fide efforts legally to acquire the resources to pay” (ibid., 672), including “efforts to obtain employment” (ibid., 674). Post-Bearden cases illustrate both this payment-work nexus and its intrinsic incorporation of work effort judgments. One Utah case, for instance, upheld incarceration for willful nonpayment of criminal restitution because the defendant, although employed full-time, did not take on a second job (State v. Brady, 300 P.3d 778 (Utah Ct. App. 2013)); the second job was necessary to enable payment because his existing earnings were largely devoted to paying child support.Footnote ⁸

Although the LFO context shares with child support enforcement these work mandates emergent from ability-to-pay jurisprudence, it currently differs in that the authorizing statutes rarely confer explicit authority to order defendants to seek and maintain employment. The same result, however, often is achieved when criminal legal debt payment obligations are supervised via probation or parole, which themselves incorporate direct work mandates, as discussed in the next section.

Nor have mandatory job search, placement, and training programs been widely promoted and institutionalized in the LFO context to the extent seen in child support enforcement during and after the work-based welfare reform efforts of the 1990s. Nonetheless, seeds have been planted. New Jersey, for instance, in the late 1980s operated a policy experiment called “MUST Earn Restitution” that mandated job search and training as part of an effort to increase LFO payments under threat of incarceration. (Weisburd, Einat, and Kowalski Reference Weisburd, Einat and Kowalski2008). In the 1970s, some states experimented with “restitution center” residential programs that incorporated mandatory private sector work in furtherance of payment (Temporary State Commission on Management and Productivity in the Public Sector 1977). These arose in the wake of court decisions that relied upon the Thirteenth and Fourteenth Amendments to strike down schemes that incarcerated defendants to force them to “work off” their LFO debts via prison labor (Anderson v. Ellington, 300 F. Supp. 789 (M.D. Tenn. 1969); Williams v. Illinois, 399 U.S. 235 (1970)). Mississippi currently operates four such restitution centers (Mississippi Department of Corrections 2019).

Although direct mandates of private sector employment are not currently widespread, another type of work mandate already receives substantial attention in LFO policy: unpaid community service. Rather than ordering defendants to “work off” LFOs in jail, some states shifted toward mandating service on public works projects or for nonprofit entities in lieu of cash payment.Footnote ⁹ Today’s liberal reformers often have embraced court-ordered community service as an “alternative to incarceration” for nonpayment (Bannon, Nagrecha, and Diller Reference Bannon, Nagrecha and Diller2010). If the LFO problem presents a dilemma as between debtors’ prison and paying ransom, then work seems to offer a third way. Debtors’ prison can be converted into debt peonage. As Gilmore warns, “debt robs. But debt also disciplines” (Reference Gilmore2011, 260). This conversion is generally absent from critical accounts of criminal legal debt that place financialization at the center of its distinctly neoliberal character (Page and Soss Reference Page, Soss, Schram and Pavlovskaya2017).

Even critiques of debtors’ prison that explicitly invoke peonage (Birckhead Reference Birckhead2015) often, though not always (Murch Reference Murch2016), miss the potential for racialized control and economic extraction mediated by labor rather than payment. Instead, the downside of community service is seen only as potential new “barriers to employment” (Birckhead Reference Birckhead2015; A. Harris Reference Harris2016), missing the labor involved in community service itself.

Promoting work mandates as an alternative to incarceration for nonpayment is a path already worn by child support policy (Turetsky Reference Turetsky2012). When promulgating new child support regulations, the Obama Administration distinguished sharply between orders to pay and orders that “require certain actions by the obligor, such as obtaining employment or participation in job search or other work activities.” Treating noncompliance with these orders as contempt punishable by incarceration is deemed suspect for nonpayment but appropriate for nonwork because there “the obligor has the present ability to do what is ordered of him or her” (Office of Child Support Enforcement 2014, 68557).

Like the presumption that McDonald’s stands ready to hire all the world’s unemployed (Brito, Pate, and Wong Reference Brito, Pate and Wong2015), the creation of court-ordered community service “opportunities” enables all nonpayment to be characterized as voluntary insofar as the defendant has “chosen” not to work off his debt. Vice versa, debtors are told quite bluntly, go to work (for free) or go to jail (City of Wichita v. Lucero, 874 P.2d 1144 (Kan. 1994)). Even if the specific form of community service comes under criticism, the likely alternative will be variations on the basic formula of work requirements, albeit with a broader menu of specific “work activities” more akin to the welfare work regime. In 2017, for instance, a new Louisiana statute began allowing LFOs to be discharged, at the supervising authorities’ discretion, by “payment alternatives” including “substance abuse treatment, education, job training, or community service” (Louisiana C. Cr. P. § 875.1(D)(2)(c)).

Community Supervision

Carceral work mandates also arise from probation, parole, and other forms of noncustodial (“community”) criminal legal supervision. In any given year almost five million people, mostly people of color, are under post-sentence community supervision (Kaeble, Maruschak, and Bonczar Reference Kaeble, Maruschak and Bonczar2015), largely through “mass probation” (Phelps Reference Phelps, Ulmer and Bradley2018). Supervision also emerges from front-end “diversion” programs that defer prosecution or conditionally dismiss a conviction pre-sentence (Lynch Reference Lynch2012; Kohler-Hausmann Reference Kohler-Hausmann2018), as well as through pretrial alternatives to money bail (Steinberg and Feige Reference Steinberg and Feige2015). In each case, behavioral conditions are the price defendants pay to avoid incarceration. Violations of those conditions may not constitute a freestanding criminal offense, and they may be established without the procedural protections applicable to convictions (Beckett and Murakawa Reference Beckett and Murakawa2012).

Seeking and maintaining employment is a ubiquitous condition of supervision for both probation and parole. Indeed, it is among the most common (Doherty Reference Doherty2015; Petersilia Reference Petersilia2003; L. F. Travis and Stacey Reference Travis and Stacey2010), though recent research gives this little sustained attention (Petersilia Reference Petersilia2003; J. Travis Reference Travis2005; but cf. Simon Reference Simon1993). Thus, incarceration may result from an insufficiently vigorous job search, an inappropriate refusal of a job offer, or an unjustified quit (Gurusami Reference Gurusami2017).Footnote ¹⁰ Technically, that incarceration punishes the underlying offense of conviction. As a practical matter, though, the sanction is triggered by failure to work to the state’s satisfaction, much as we saw with child support enforcement formally characterized in terms of nonpayment but nonetheless explicitly contingent on work.

James Nolan’s research offers a vivid example:

One survey of hundreds of drug courts found employment or education requirements the second most common condition after drug treatment; community service mandates also were widespread (Rossman et al. Reference Rossman, Roman, Zweig, Rempel and Lindquist2011).

Again, relatively little is known about the nature or degree of street-level enforcement of these conditions. Research from the early 1990s reported minimal enforcement (Simon Reference Simon1993), but recent case studies cite employment pressure from parole and probation officers as a major feature of supervisory relationships (Gurusami Reference Gurusami2017; Augustine Reference Augustine2019). Moreover, data from the 2004 Survey of Inmates in State and Federal Correctional Facilities and the 2002 Survey of Inmates in Local Jails suggest that work-related carceral threats are carried out at nontrivial rates (Zatz et al. Reference Zatz2016). At any one time, about nine thousand individuals nationally report being incarcerated for violating a probation or parole employment condition, consistent with findings from other sources (Petersilia Reference Petersilia2003, 151). These work rule violations often coincided with those for failure to pay, while another eight thousand cases formally involve only nonpayment, which we have seen also implicates work. Among the twelve hundred cases based exclusively on work rule violations, two-thirds of respondents reported having worked full-time in the month preceding admission. Their work, however, appears to have been intermittent and/or very low-wage,Footnote ¹¹ suggesting that work enforcement was regulating the nature and intensity of employment, not bare labor market participation. Notably, already stark racial disparities in probation and parole incarceration generally (Phelps Reference Phelps, Ulmer and Bradley2018) were further intensified for work condition violations, with Black defendants representing 42 percent of the former and 67 percent of the latter.

Data on incarceration for violations does not capture those who comply with work demands but nonetheless are affected by them, including by accepting labor conditions they might otherwise resist (Hatton Reference Hatton2018a). One possible indicator of such effects is an anomaly in the literature on post-incarceration employment: despite the well-known barriers erected by a criminal record, recently released prisoners initially work at rates above their pre-incarceration levels (Pettit and Lyons Reference Pettit, Lyons, Bushway, Weiman and Stoll2007; Loeffler Reference Loeffler2013). Pettit and Lyons suggest in passing that this increase might reflect post-release supervision enhancing earnings capacity. Yet employment increases also could arise if work mandates created pressure, above and beyond financial incentives, to accept jobs that might previously have been forgone (Bumiller Reference Bumiller, Carlton and Segrave2013; Augustine Reference Augustine2019). The latter coheres better with Pettit and Lyons’s other finding that wage rates decreased immediately post-release.Footnote ¹² Other research found that the threat of probation revocation, not employment services, explained a program’s entire effects, though in that case only payments and not earnings were measured (Weisburd, Einat, and Kowalski Reference Weisburd, Einat and Kowalski2008).

Work mandates under community supervision may increase in importance alongside policy attention to “reentry” (Petersilia Reference Petersilia2003; J. Travis Reference Travis2005). Mead, for instance, touts probation and parole as, like child support enforcement, potentially powerful institutions with underutilized capacity to impose and inculcate labor discipline among low-income men of color (Reference Mead2011). He advocates incorporating mandatory work programs like those developed for welfare-to-work (cf. J. Travis Reference Travis2005; Simon Reference Simon1993, 263) and then child support enforcement. As with welfare reform, many liberals may agree, leaving policy disputes confined to the familiar balance between punitive “hassle” designed to lower expectations and supportive “help” designed to increase the returns to work (Mead Reference Mead2010; Sorensen Reference Sorensen2010). Underlying this agreement is the widespread view that employment is a foundation of personal and community flourishing (Wilson Reference Wilson1996), specifically including avoidance of future criminal legal system involvement (Uggen Reference Uggen2000). Sugie recently has noted, however, that the extremely precarious nature of some work available to recently incarcerated people “likely exacerbates day-to-day instability and strain” (Reference Sugie2018).

Bruce Western, for instance, sharply criticizes how mass incarceration generates racial and labor market inequality (Reference Western2006) and proposes massive annual spending on transitional jobs and other supports for recently released people (Reference Western2008). Nonetheless, like Mead, he would incorporate work mandates into parole. Despite opposing full revocation for technical violations, Western endorses “short jail stays up to several weeks” (Reference Western2008, 19) to sanction nonwork (cf. J. Travis Reference Travis2005). These would enforce work discipline and instill “rudimentary life skills of reliability, motivation, and sociability with supervisors and coworkers” (Western Reference Western2008, 14), echoing Mead’s focus on a “disorderly lifestyle” (Reference Mead2011, 31) and “breakdown in work discipline” (Reference Mead2011, 16). Likewise, the widely celebrated Center for Employment Opportunities (CEO) model provides transitional employment to instill “the habits and competencies that make for a good employee emphasized through the routine of reporting for work each day, cooperating with colleagues, and following supervisors’ directions” (Broadus et al. Reference Broadus, Muller-Ravett, Sherman and Redcross2016, iii). Participants enter through parole or probation officer referrals and sometimes are specifically mandated to participate (Redcross et al. Reference Redcross, Millenky, Rudd and Levshin2012).

Work requirements are only one component of community supervision and may operate in synergy with conditions mandating payment of child support (Haney Reference Haney2018) or criminal legal debts, community service, or other mandatory programs. Forrest Stuart has shown how diversion programs in Los Angeles’ Skid Row leverage quality-of-life arrests into mandatory participation by unhoused people in residential social services they previously had avoided (Reference Stuart2011). Although Issa Kohler-Hausmann suggests that the content of such programs is unimportant, with community service functioning as a “neutral currency of hassle and performance capacity” (Reference Kohler-Hausmann2018, 230), Stuart’s analysis points toward the broader political economy in which they are situated: in these programs “arrestees also constitute the manual labor—maintenance, cleaning, and food preparation—necessary to keep the facility’s operating costs to a minimum” (Reference Stuart2011, 208). Similarly, recent investigative journalism has revealed a more extreme practice of Oklahoma courts mandating, as conditions of probation or diversion, participation in what are denoted residential drug treatment but in practice consist primarily of forced labor for commercial poultry processing plants (A. J. Harris and Walter Reference Harris and Walter2017). The result was both a captive labor force for the plants and a revenue source for the programs, obviating any need for either public funding or participant payment.

Depressing Labor Standards by Defining Voluntary Unemployment

Conventional employment law sets labor standards by establishing a floor beneath which employers may not go. The purpose and, where enforced, effect is to raise working conditions at least to that floor. Work requirements also set labor standards and shape conditions but in a different way. Most simply, they penalize nonwork. For some people who would not work absent the threat of penalty, the threat pushes them to work instead. Schematically, the threat lowers reservation wages. Likewise, bargaining power is lost among workers who absent the carceral danger could credibly threaten to quit if the employer cut (or refused to raise) wages. For both reasons, wages decline. This is essentially the old logic of “less eligibility” (Rusche and Kirchheimer [1939] Reference Rusche and Kirchheimer2003; Handler and Hasenfeld Reference Handler and Hasenfeld1991), which calibrates the conditions of nonwork (whether set by public benefits or prison conditions) to sit below those deemed appropriate for work.

This simple model leaves out the process of defining the “nonwork” that triggers the penalty and the “work” that avoids it. That process directs the administration of work requirements toward the characteristics of jobs. This connection flows through the voluntary/involuntary unemployment distinction. For someone deemed unable to work, nonwork triggers no penalty. That is why work requirements—in the classic public benefits context and in the carceral ones discussed here (State v. Ingram, 142 P.3d 338 (Kan. Ct. App. 2016); People v. Likine, 823 N.W.2d 50 (Mich. 2012)) —always are restricted to people variably described as “employable,” “able-bodied,” “not disabled,” and the like, complicated and normatively charged as those categories turn out to be (Diller Reference Diller1996).Footnote ¹³

The most difficult challenges arise from assessing the voluntariness of unemployment for people who are deemed “employable” in the abstract. If there are no jobs available, unemployment is “involuntary.”Footnote ¹⁴ But which jobs count? For instance, is a job “available” if it requires a long commute or even a cross-country move? For Social Security disability purposes, yes; for most unemployment insurance purposes, no. What about a job entailing illegal working conditions: subminimum wages, subjection to sexual harassment, or safety hazards? Generally speaking, a worker may choose to refuse such jobs and yet still be considered involuntarily unemployed. That much is clear in the public benefits context (Zatz Reference Zatz2006), and it would be surprising to find something different with carceral work requirements (but cf. Schmeets v. Turner, 706 S.W.2d 504 (Miss. Ct. App. 1986)). The distinction between voluntary and involuntary employment is a thoroughly normative one (Williams Reference Williams, Handler and White1999).Footnote ¹⁵

The classic criticisms of the unemployed are that they are being too choosy by rejecting some jobs and too lazy by settling for inadequate jobs or no job at all. Recall Mead’s complaints about refusals to work for “chump change” (Reference Mead2011, 18). Thus, actual work requirements do not simply demand work in the abstract. They enforce expectations that the un(der)employed worker search harder, work longer, and accept less (Lafer Reference Lafer2004). One child support obligor, for instance, was found in contempt, despite working forty hours per week, because he declined an opportunity to earn more by working “ten to fifteen hours per day, sometimes seven days per week,” in a different job that required relocating to another state (McDaniel v. McDaniel, 878 So. 2d 686, 690 (La. Ct. App. 2004)); Haney reports one child support judge held obligors to the standard set by a father who worked four jobs and mowed lawns to boot (Reference Haney2018). Gurusami found probation officers pressuring people to abandon part-time or temporary work to pursue full-time work even while condemning as “lazy” attempts to enhance employability through education (Reference Gurusami2017, 445). Thus, not only does some nonwork avoid sanction, but some work may be deemed noncompliant.

The same basic points apply to unemployment through job loss. A worker who quits may be deemed voluntarily or involuntarily unemployed depending on the reasons for quitting. Recall Mead’s complaint about “leav[ing] in a huff” merely because the job is “unrewarding” (Reference Mead2011, 18). So, too, for terminations, depending on the worker’s responsibility: one obligor went to jail because he got fired for fighting with a coworker (Batton v. Com. ex rel. Noble, 369 S.W.3d 722 (Ky. Ct. App. 2012)).

If carceral work mandates have any effect, they must exert downward pressure on labor standards. In principle, all actually existing nonwork could be deemed “involuntary,” but that obviously is not the point. Rather, the rationale is that too many workers are shirking their responsibilities and need the firm hand of state power to whip them into shape (Mead Reference Mead2011).

This combination of increased employment at lower wages has been found in research on welfare (Cancian et al. Reference Cancian, Haveman, Meyer and Wolfe2002), parole (Pettit and Lyons Reference Pettit, Lyons, Bushway, Weiman and Stoll2007; cf. Haley Reference Haley2016, 185), and, more tentatively, child support work requirements (Schroeder and Doughty Reference Schroeder and Doughty2009). Sugie’s research on recently incarcerated job-seekers also suggests how this downward ratchet could operate. More extensive employment was a function not of better opportunities but rather greater “willing[ness] to take on poorer quality work”; those who worked more reported worse hours, pay, and overall conditions than those who worked less (Reference Sugie2018, 1478). Other research also finds sectors where employers will hire formerly incarcerated workers, so long as they accept abusive conditions (Bumiller Reference Bumiller2015; Peck and Theodore Reference Peck and Theodore2008). One function of carceral work mandates is to manufacture that willingness to accept less.

Such downward pressure, however, is not unlimited. And that limit could be set somewhere above the floor established by conventional employment protections. For instance, employers generally are permitted to impose mandatory overtime, so long as it is properly compensated; nonetheless, welfare work requirements typically prohibit agencies from mandating more than forty-hour work weeks (Zatz Reference Zatz2006). So a worker could be fired legally, based on her choice to refuse overtime, and yet be deemed involuntarily unemployed and thus compliant with work requirements. Longstanding points of legal conflict involve whether it is excusable to refuse a transfer that would require a long-distance move (required by McDaniel, supra), to miss work or restrict availability in order to care for an ill family member or a young child, and so on (Williams Reference Williams, Handler and White1999).

Similarly, workers may be able to limit their search to their chosen occupation, one that they have trained for, or for wages commensurate with their prior experience. Such restrictions traditionally are incorporated into unemployment insurance’s “suitable work” concept; those protections often weaken as time passes and unemployed workers are expected to lower their standards and accept downward mobility (Nicholson and Needels Reference Nicholson and Needels2006). Similar concepts, without much calibration, are echoed in the California Supreme Court’s Moss opinion limiting child support obligors’ work expectations to “employment for which the parent is suited by virtue of education, experience, and physical ability” (950 P.2d 59, 76 (Cal. 1998)).

The bounds of “involuntary unemployment” can vary significantly with legal context. While unemployment insurance allows workers temporarily to hold out for jobs comparable to their previous work, “Work First” or “Labor Force Attachment” programs in the welfare-to-work context enforce the notion that any (legal) job is better than no job. In a Portland welfare-to-work program often touted as an alternative to “Work First,” a key feature was allowing participants to hold out for jobs with supra-minimum wages and some prospect of advancement (Berlin Reference Berlin2002). Notably, the Obama Administration’s child support enforcement regulations initially rejected funding “services to promote access to better jobs and careers” and instead prioritized “rapid labor force attachment” programs that take labor market prospects as-is (Office of Child Support Enforcement 2014, 68558). The Michigan Supreme Court recently justified its especially stringent “impossibility” standard for involuntary unemployment by referencing the child support context’s specific “concern for paying the debt one owes to one’s child, which arises from the individual’s responsibility as a parent” (People v. Likine, 823 N.W.2d 50, 70 (Mich. 2012)).

Just as some conception of what jobs are good enough is intrinsic to “involuntary unemployment,” so too is some conception of what life is good enough for the worker: what standard of living is sufficient, how much time should be devoted to work, what level of stability or uncertainty is acceptable, with what consequences for family members, and so on. Such conceptions of appropriate work and life invite interpretations shaped deeply by the race, class, and gender positions of the worker. Although the particularities have varied, racial labor hierarchy long has been justified by attributing lesser material needs to workers of color, greater need for strict discipline, and unsuitability to higher status work (Glenn Reference Glenn2002; Lamont Reference Lamont2000; Davis Reference Davis and James2000); it also is sustained by “colorblind” disregard of racialized barriers to employment (Brito, Pate, and Wong Reference Brito, Pate and Wong2015). At a micro level, decision makers must “apply a logic of appropriateness to the events that unfold at their desks (asking what is permissible, reasonable, and right to do in a case such as this), they must assign meaning to a client’s action—treating it as a fluke, an avoidable mishap, or an intentional act of noncompliance and deceit” (Soss, Fording, and Schram Reference Soss, Fording and Schram2011, 300). The racial content of such decision making is amplified by the institutional context in which such judgments are made (Soss, Fording, and Schram Reference Soss, Fording and Schram2011). We are considering, after all, workers already defined in various combinations by debt, by nonmarital childrearing, by criminal conviction.

A striking feature of carceral work mandates is the minimal formal elaboration of these suitable work concepts. Child support work requirements sometimes do come with limitations, such as “employment commensurate with the obligor’s education, training, professional or occupational qualifications, job skills, work history, and the availability of employment opportunities in the community” (56 Okla. St. Ann. § 240.10(B)(5)). Even these have far less detail than typical of public benefits contexts, and they sometimes impose such broad obligations as obtaining “any employment that the obligor is capable of performing” (Mont. St. Ann. § 40-5-291(1), (5)). In the criminal legal debt and community supervision contexts, however, there is virtually no formal guidance in statute or case law.

The lack of authoritative standards does not mean that lines are not being drawn. It simply affects which institutional actors are drawing them. The implication is that labor standard setting is currently submerged in the discretion of traffic court judges, child support enforcement agencies, probation officers, and the like (Gurusami Reference Gurusami2017; Haney Reference Haney2018; A. Harris Reference Harris2016). The discretion, and resulting interpersonal power, of street-level bureaucrats (Lipsky Reference Lipsky1980) is itself an old theme in the public benefits literature on work requirements (Handler Reference Handler2004; Brodkin Reference Brodkin1997).

Empowering Employers to Trigger Carceral Enforcement

Carceral work mandates also confer on employers power over workers at moments of potential exit. This shapes the ongoing practice of workplace governance and discipline. What happens when employers and workers conflict over working conditions, job performance, and so on? The employer’s day-to-day economic power normally comes from its ability to fire workers and cut off their wages. That power expands in both kind and degree when the employer also can threaten to get the worker arrested, an instance of what Hatton terms “social coercion” more generally (Reference Hatton2018a).

The analogous point has been developed regarding employers’ ability to trigger immigration enforcement, which gives them a potent means of retaliating against immigrant workers for workplace complaints, union organizing, or the like (R. Smith, Avendaño, and Martínez Ortega Reference Smith, Avendaño and Ortega2009). The analogy may seem curious because, for unauthorized workers, the state forbids work rather than mandating it. But the crucial insight concerns an employer’s control over bringing down the power of the law, not the formal content of the law’s command. Because the state relies on employers as delegated immigration screeners (Lee Reference Lee2009), employers can hire unauthorized workers and then hold over them the threat of triggering detention and deportation with a simple call to Immigration and Customs Enforcement (ICE) (Kim Reference Kim2015).

Employers’ ability to leverage state violence is even clearer with employer-specific guestwork visas. Because any rupture in the employment relationship can trigger deportation, guestwork programs have long been criticized for creating a relationship akin to involuntary servitude (Ontiveros Reference Ontiveros2006; Gordon Reference Gordon2006). Indeed, the Supreme Court’s peonage jurisprudence equates a regime that “authorize[s] the employing company to seize the debtor and hold him to the service” with one that enlists “the constabulary to prevent the servant from escaping, and to force him to work out his debt” (United States v. Bailey, 219 U.S. 219, 244 (1911)). Either way, the day-to-day power relations at work operate in the shadow of what workers can expect from defiance or departure (Pope Reference Pope2010).

Employers gain similar power when carceral work mandates merge the employer’s supervisory interests with those of the mandating institution (cf. Levin Reference Levin2018). In the drug court episode recounted above, Judge McKinney offered to “add another weapon to [the employer’s] arsenal” by promising to “put him in jail, on your say so” (Nolan Reference Nolan2002). Similarly, in the early twentieth century, “[p]aroled black women who performed domestic service for white employers were always under the threat of being sent back to the chain gang or state farm if they broke a rule or failed to work up to their employers’ standard” (Haley Reference Haley2016, 176), and employers stayed in close touch with criminal legal authorities (cf. Simon Reference Simon1993). Today, one reentry program in Louisiana touts itself to employers as providing “Oversight: Probation Officers and Case Managers are your HR Department.”Footnote ¹⁶

In principle, a worker’s vulnerability would be limited by the “involuntary unemployment” concepts discussed above. But disputes over working conditions often are murky. As an early peonage case explained, state coercion survives legal protection for quitting with “sufficient excuse” because that protection “could never be known to be available except at the risk of, and at the end of, a criminal prosecution” (Toney v. State, 37 So. 332, 334 (Ala. 1904)). This is particularly true where workers under carceral threat already have been rendered suspect by the intersection of formal legal status and race (Soss, Fording, and Schram Reference Soss, Fording and Schram2011, 80). Authorities may find credible employers’ accounts of insufficient willingness to work on their terms and may crack down based on employers’ “say so.”

Creating Legally Unprotected Forms of Work

The previous two sections assumed that the carceral threat could drive labor standards down toward the legal floor but not below it, underenforcement aside. Those floors, however, generally apply only to work legally recognized as employment. Carceral work mandates also may operate to remove that floor by enabling the legal classification of work as outside employment. The result is a new tier of subordinated labor subject simultaneously to enhanced coercion and diminished protection. This comports with longstanding patterns by which the labor citizenship associated with white, male breadwinners has been dissociated from work deemed to occur outside the labor market and, instead, to be embedded legally in separate institutional spheres of family, of immigration, of welfare, and/or of criminal law (Dawson Reference Dawson2016; Gilmore Reference Gilmore2007, 185; Hatton Reference Hatton2015; Zatz and Boris Reference Zatz and Boris2014).

Prison labor illustrates this dynamic. US courts place it outside protected employment relationships. The rationales shift with context but ultimately treat labor’s embeddedness (Granovetter Reference Granovetter1985) within the institutional practice of punishment as incompatible with the employment status associated with market exchange (Zatz Reference Zatz2008). This idea of a “separate world of the prison” (Vanskike v. Peters, 974 F.2d 810, 814 (7th Cir. 1992)) does not apply directly to those not incarcerated but instead at work “in the community.” Indeed, the doctrine routinely attributes employment law protections to people who, while serving a carceral sentence, nonetheless are temporarily free on “work release,” where they are treated as “free laborers in transition to their expected discharge from the prison” (Bennett v. Frank, 395 F.3d 409, 410 (7th Cir. 2005)). Little effort has been devoted, however, to mapping that boundary, and there is reason to doubt that it is firm (Zatz Reference Zatz2008).

In several contexts, labor protections already have been stripped from work outside incarceration but under carceral threat. The first involves the Thirteenth Amendment’s bedrock prohibition of involuntary servitude. Because “involuntariness” generally applies to labor under threat of physical violence, whether private or public, work pursuant to carceral threat seemingly constitutes involuntary servitude under the constitutional analysis applied to criminal laws creating Jim Crow peonage, all the more so where debt creates the trilemma of pay, work, or jail (Zatz Reference Zatz2016). To be sure, complications arise from the Thirteenth Amendment’s criminal punishment exception (Raghunath Reference Raghunath2009), which some see as swallowing the rule (Childs Reference Childs2015; but cf. Pope forthcoming). Child support enforcement, however, seemingly provides a clean case because the duty to work arises without any prior criminal conviction.

Nonetheless, courts have upheld incarceration of unemployed child support obligors against constitutional challenge. The limited case law relies on two lines of reasoning. The first finds no “servitude” so long as the work obligation is not specific to any one employer. In that sense, the worker is forced into the labor market where she remains free to choose (Zatz Reference Zatz2016; Pope Reference Pope2010). The second argument, more relevant here, takes the opposite tack. In United States v. Ballek (170 F.3d 871 (9th Cir. 1999)), the Ninth Circuit explained that “not all forced employment is constitutionally prohibited” because obligations “traditionally … enforced by means of imprisonment” are outside the constitutional prohibition (874). Judge Kozinski extended a set of late nineteenth- and early twentieth-century precedents that had upheld compulsory work by private sailors and conscription into military service or public road construction. Forced labor to generate child support vindicated a public duty “of vital importance to the community,” not least because it offset the cost of public assistance. Thus, Ballek posits some purely “economic” sphere within which the Amendment ensures private economic freedom, but that freedom does not extend to work grounded in the principles organizing some other sphere. This simply generalizes the logic that renders penal servitude compatible with—because separate from—a “free labor system” in the market (Pope forthcoming).

With regard to private sector employment, these rulings appear limited to Thirteenth Amendment protections. No court has stripped private employment of statutory protections by virtue of its mandatory character. However, both criminal legal debt and community supervision have generated specialized work programs, backed by carceral threat, that suggest how unprotected work could take root outside both the prison and the conventional labor market. As such, they suggest how “neoliberal paternalism” (Soss, Fording, and Schram Reference Soss, Fording and Schram2011) or, relatedly, the “roll-out of market-making policies” (Grewal and Purdy Reference Grewal and Purdy2015, 5) may not capture the full scope of work mandates; they do not invariably point toward marketized forms of work.

This dynamic is illustrated by court-ordered community service, which arises not only to “work off” criminal legal debt but also as an important noncarceral sanction generally (Tonry Reference Tonry and Petersilia1997). New York City judges extensively grant an “adjournment in contemplation of dismissal” (ACD) (Kohler-Hausmann Reference Kohler-Hausmann2018) conditioned on a misdemeanor defendant’s performance of “services for a public or not-for-profit corporation, association, institution or agency” (N.Y. Crim. Proc. Law § 170.55(6)). Although not technically linked to any financial obligation, one rationale for such work is that “[d]efendants who do not have money to make restitution should, when practical, pay for their offense through community service” (Doyle v. City of New York, 91 F. Supp.3d 480 (S.D.N.Y. 2015)).

A federal court recently rejected the first known employment law claim by ACD community service workers. They sought minimum wages under the Fair Labor Standards Act for collecting garbage and maintaining city parks and bridges. Relying on prison labor precedents, the court reasoned that these were not employment relationships because the work functioned not “to earn a living” but instead “to avoid further prosecution and the risk of conviction” and, in doing so, to discharge “a legal obligation to serve the public good” (Doyle, 488–89); note the echoes of Ballek on child support.

The Doyle workers received no direct economic benefit from their efforts, similar to “volunteers” generally excluded from employment law (Tsuruda Reference Tsuruda, Collins, Lester and Mantouvalou2018). This feature, however, may not prove limiting. Consider community service that “works off” criminal legal debt. Such workers never see a penny for their efforts, but they are credited with debt reductions in amounts based on a nominal hourly wage (Cal. Penal Code § 1209.5). In Los Angeles, debtors must sign a participation agreement declaring themselves to be “volunteers” “performing unpaid community service,” “not an employee of the [agency],” and “not entitled to recover any workers’ compensation benefits” in the event of injury (Zatz et al. Reference Zatz2016). Welfare agencies have similarly denied the employee status of workfare workers, though courts have sometimes disagreed (United States v. City of New York, 359 F.3d 83 (2d Cir. 2004); Goldberg Reference Goldberg2007).

A further half-step beyond nominally unpaid “community service” lie the “transitional jobs programs” (TJPs) that currently are in vogue in reentry policy. These include both the CEO model discussed above and social entrepreneurship approaches like the well-known Homeboy Industries. The particular jobs, or the entire business, are created specifically to provide employment for disadvantaged workers and often incorporate supportive services. This facilitates integration of such programs into the job placement efforts of child support enforcement or community supervision.

In the welfare-to-work context, TJPs were deliberately structured as wage-paying, legally protected employment in contrast to “workfare.” Yet TJPs themselves are susceptible to efforts to structure and legally classify them in ways akin to classic “prison industries.” Those often self-consciously mimic conventional businesses in their pay, supervision, and performance of “real work” making products for sale. Despite these features, courts have placed prison industries outside legal “employment” because the prison funnels workers into them and does so for “penological” or “rehabilitative” purposes (Zatz Reference Zatz2008). Thus, while the Doyle court repeatedly emphasized that ACD community service did not involve “work for the benefit of a private enterprise or … produc[ing] goods or services that were sold in commerce” (ibid., 488) and implied that this feature was important to the result, such distinctions have not always held in prison labor jurisprudence.

Some recent developments suggest how work programs outside of prison could follow the prison industries model and allow Doyle to extend even to paid work yielding products for sale. For instance, Los Angeles’ much-heralded $15 local minimum wage included a partial carve-out for transitional employers such as Homeboy (Reyes Reference Reyes2015). Similarly, in the Oklahoma poultry processing scheme mentioned above (A. J. Harris and Walter Reference Harris and Walter2017), “drug court” defendants were forced to work the poultry processing line of major commercial firms and perform work functionally indistinguishable from regular employees. These work assignments, however, were structured through an intermediary, Christian Alcoholics & Addicts in Recovery (CAAIR), which purportedly provided residential drug rehabilitation services. Like a temp agency, CAAIR was paid by the poultry processor, but CAAIR pocketed the entire wage to fund its operation. Technically, the court ordered the defendants into rehab on CAAIR’s terms. CAAIR required them to sign waivers declaring themselves to be “clients,” not “employees,” much as the Los Angeles community service intermediaries do.

These examples suggest how Doyle and Ballek might be generalized. By segregating work from “regular” employees, employing alternative currencies of debt relief (Zelizer Reference Zelizer1997), and assigning and monitoring participation through specialized institutions, such work structures might do sufficient “relational work” (Zelizer Reference Zelizer2005; Zatz Reference Zatz2008) for courts or legislatures to place them outside conventional employment law. In that position, the downward pressures identified in the previous subsections could operate without any restraining floor and could create a legally second-class tier of work available to public agencies and private firms alike.

Disciplining and Displacing Other Workers

Even workers not personally subject to carceral work mandates may feel their effects. As Angela Davis recounts, “DuBois made the astute observation that so-called ‘free’ black labor was, in a very concrete sense, chained to black convict labor” (Reference Davis, Lawson and Kirkland1999, 356). The availability of one vulnerable labor force tends to discipline the demands of others. The latter fears displacement by the former. Harris & Walter’s reporting on CAAIR (Reference Harris and Walter2017), for instance, notes how a chicken processor was laying off paid employees while becoming increasingly reliant on workers subject to carceral work mandates.

Such discipline and displacement is a familiar issue in public benefits, immigrant labor, and prison labor contexts. Consider the “prevailing conditions” rule in unemployment insurance (Kilbane Reference Kilbane1998). If a job offers terms below what prevails for that occupation in the local labor market, an unemployed person may refuse it without jeopardizing their benefits, even if the job is otherwise suitable. This limits downward pressure on labor standards from a pool of unemployed job seekers. Similar principles underlie wage regulation in agricultural guestwork and commercial prison labor programs (J. J. Lee Reference Lee2017; Zatz Reference Zatz2008). So, too, with the standard “competitive fairness” rationale (S. D. Harris Reference Harris2000) for applying employment laws to stigmatized workers such as unauthorized immigrants (Motomura Reference Motomura2010), welfare recipients, and prison inmates (Zatz Reference Zatz2008): directly protecting one group of workers also indirectly protects others (and the firms hiring them) with whom they compete. Similar concerns apply to how employers could use carceral threats to disrupt worker collective action. Employer-triggered immigration enforcement, for instance, not only provides retaliatory leverage against individual workers but also may undermine organizing that crosses immigration status lines and could benefit authorized workers.

Efforts to equalize labor standards across work status can limit employer incentives to utilize vulnerable workforces to lower costs or expand control. A more direct approach bars employing such workforces unless “regular” workers are unavailable and thus not at risk of displacement. Such anti-displacement provisions are common in welfare work programs (Dietrich, Emsellem, and Yau Reference Dietrich, Emsellem and Yau1996–1997) and agricultural guestwork (J. J. Lee Reference Lee2017), and similar principles have steered prison labor expansion toward industries that otherwise would move production abroad (Weiss Reference Weiss2001).

Rather than limiting, or upgrading, precarious work, however, the competitive fairness rationale can simply concentrate unprotected or vulnerable workers in jobs that seem not to threaten “regular” workers. One crude example is rules limiting prison labor or welfare work assignments to governmental or non-profit enterprises (McLennan Reference McLennan2008) or to enterprises that do not produce goods or services for sale. Recall how a similar principle animated Doyle’s denial of employee status to New York’s ACD community service workers. In this fashion, anti-displacement concerns may actually amplify the dynamics that create separate, second-class workforces.