Locked Out of the Labor Market? State-Level Hidden Sentences and the Labor Market Outcomes of Recently Incarcerated Young Adults

INCARCERATION, RACE, AND LABOR MARKET OUTCOMES IN YOUNG ADULTHOOD

The labor market consequences of criminal justice contact have received a great deal of scrutiny. The bulk of this evidence shows that individuals struggle in the labor market following incarceration.¹ For example, audit studies show that callback rates for individuals with a criminal record are substantially lower than for individuals with no record (Pager 2003). Studies based on survey data have found an incarceration penalty on employment ranging from 10 to 30 percent, and a wage penalty of 10 to 40 percent (Apel and Sweeten 2010; Western 2002, 2006; Raphael 2007). Evidence also indicates that the labor market consequences of incarceration are not uniform, with several studies exploring heterogeneity by race-ethnicity. Devah Pager (2007) argues that a criminal record reinforces racial stereotypes, especially among black males, and intensifies the stigma of incarceration. For example, relative to formerly incarcerated whites, blacks have lower rates of labor market participation after prison and experience slower post-prison wage growth that extends well beyond the period of incarceration (Apel and Powell 2019; Lyons and Pettit 2011; Western and Sirois 2018).

Incarceration is conceptualized as a turning point in the life course that disrupts adult development, the effects of which reverberate and accumulate as people age (Pettit and Western 2004; Sampson and Laub 1992; Western 2002). After decades of growth, incarceration has become an increasingly common feature of the transition to adulthood among disadvantaged segments of the population (Pettit and Western 2004). It is important to examine the labor market consequences of criminal justice contact during the transition to adulthood because it is a period when risk of incarceration peaks and labor market inequalities start to emerge (Cheng 2014; Pettit and Western 2004). Incarceration during this critical stage can therefore contribute to a pattern of cumulative disadvantage in the worlds of education, work, and interpersonal relationships (Steinberg, Chung, and Little 2004; Western 2002).

Three primary mechanisms have been proposed to account for the labor market consequences of incarceration: selection, incapacitation, and stigma. According to the selection explanation, incarceration has no independent effect on labor market outcomes, but rather the criminal justice system disproportionately “selects” individuals who are not willing or able to hold down work (Pager 2007). Evidence is clear that formerly incarcerated individuals—even in the absence of incarceration—would struggle to be competitive in the labor market (Western 2006). But it is also the case that exposure to incarceration (that is, incapacitation) can fundamentally alter an individual’s employability through gaps in work history, loss of job skills, and the loss of the informal social networks often pivotal for landing employment. Finally, a criminal record works as a negative credential that prohibits labor market participation through social stigma and formal exclusion. Pager (2003) refers to a criminal record as a “status mark of dishonor,” impacting how employers view the reputation, reliability, and trustworthiness of those with a criminal record, black males in particular. In support of this, large shares of employers, when prompted, report that they would not knowingly hire someone with a criminal record (Holzer 1996). Thus, even though the criminal justice system disproportionally selects those who may struggle in the labor market, the evidence is clear that going to prison contributes to larger patterns of labor market disadvantage.

A criminal record is disqualifying not only because of incapacitation or informal exclusion through social stigma, but also for the formal, state-imposed sanctions that create barriers to full societal participation. This kind of exclusion, outlined in the following section, is often discussed but rarely directly measured, either as direct exclusion in the labor market or as a broader collection of policies that alter and limit participation in other social and economic spheres (Miller and Stuart 2017; Petersilia 2003; Travis 2002).

HIDDEN SENTENCES AS LABOR MARKET BARRIERS

As of 2015, more than thirty-five thousand laws across the United States impose more than forty thousand penalties on the justice-involved beyond visible forms of punishment like imprisonment and probation (Kaiser 2016). These policies restrict those with a criminal record from working certain types of jobs entirely or obtaining necessary certifications, from holding a driver’s license or voting, from receiving financial aid or government benefits, and from numerous other social and economic activities (Petersilia 2003; Travis 2002). Researchers refer to such policies using varying definitions and terms, including collateral consequences, collateral sanctions, civil disabilities, and invisible punishments (Chin 2011; Travis 2002; Uggen and Stewart 2014; Whittle 2018). Following Joshua Kaiser (2016), we adopt the terminology of hidden sentences to emphasize first, that these policies are state-imposed punishments based on a variety of criminal labels, such as conviction, indictment, or an arrest record (Chin 2011; Whittle 2018), and, second, that they are distinct from visible sentences largely because legal processes keep them obscured to different degrees throughout public and private spheres (Kaiser 2016; Travis 2002). Hidden sentences have the potential to affect individual labor market outcomes in three ways: direct restrictions on employment-related rights and privileges; indirect restrictions on employment through limits on other kinds of citizenship rights and social participation; and informal filters on the labor market due to overwhelming experiences of administrative burden, uncertainty, and stigmatization.

First, hidden sentences can limit access either to certain kinds of employment directly or to the occupational and business licenses necessary to them (Pager 2007; Petersilia 2003; Travis 2002). For example, felony convictions automatically exclude individuals from licenses in various medical occupations, barbering and beautician services, independent contracting, and other fields, and several kinds of convictions can create bans from public office, law enforcement, and segments of the civil service. In addition, many employers and certifiers, such as state bars or accounting boards, are empowered to exclude based on indictments, arrests, or other kinds of criminal justice contact. Second, hidden sentences that limit other citizenship rights and restrict societal participation can create barriers in the labor market. For instance, hidden sentences that curtail driver’s licenses can both prohibit the justice-involved from delivery services, commercial trucking, and other employment that requires driving, and restrict where and when individuals can work. Indeed, Michele Cadigan and Gabriela Kirk (2020) document how driver’s license suspensions, as a postconviction mechanism used to collect legal fines and fees, can affect labor market participation because of subsequent difficulties traveling to court hearings or places of employment. Hidden sentences that impose residence restrictions or prohibit even the ability to be present in certain locations (such as schools or day cares) also limit employment options (Beckett and Herbert 2010). Restrictions on firearms can prevent even nonviolent offenders from serving in private investigations and security, and disfranchisement can make the justice-involved unqualified to serve in some public employment even when other hidden sentences do not (Stavsky 2002; Uggen and Manza 2002).

Hidden sentences also enable employers and occupational gatekeepers (along with landlords, lenders, and other private parties) to run background checks and empower them to use arrest and conviction histories as factors in hiring, firing, and other employment decisions—thereby enabling the social stigma that puts the justice-involved at a disadvantage in the labor market (Pager 2003, 2007; Uggen et al. 2014). These more comprehensive employment barriers are often based on the broadest criminal labels. Such background checks and general criminal offense registries frequently include misdemeanors, drug and public order offenses, and arrests or indictments that never lead to convictions.

Third, state hidden sentences can affect how individuals approach and experience the labor market through administrative burden. As a policy regime increases in sheer size and complexity, so do the learning costs, compliance costs, and psychological costs of participating in the relevant programs or activities (Herd and Moynihan 2019). The result is a system of barriers that often cause avoidance, noncompliance, or participation in ways that are not predicted by a straightforward reading of the policies themselves. Research on the welfare state and economic regulation shows that the administrative burden of a particular policy regime can have negative impacts on participation in educational applications and financial aid (Hoxby and Turner 2015), immigration and nationalization (Heinrich 2018), voting and voter registration (Burden et al. 2014), medical insurance (Moynihan, Herd, and Ribgy 2016), and other social and economic activities.

Hidden sentence regimes can create similar experiences of administrative burden and therefore alter labor market participation in ways beyond formal restrictions themselves. Unlike the civil death laws of the past that excluded felons from society with a single law, hidden sentence regimes impact employment and citizenship status through a myriad of restrictions that vary by state and offense type, that apply under different circumstances and for different lengths of time, that encompass activities that are very narrow (for example, wool dealing or particular educational loans) or extremely broad (for example, all public employment in a state), and that depend on various employers’ and other decision-makers’ discretion (Ewald 2012; Kaiser 2016). As a result, these restrictions may “pile on” to create a hopelessly tangled policy of restrictions that take their toll one after another and can carry extreme compliance costs and offer very little benefit (Uggen and Stewart 2014). As a result, we expect the size and complexity of a state’s hidden sentence regime alone to amplify its impact on labor market outcomes.

Hidden sentences as a form of state surveillance may also create stigma and result in avoidance of formal institutions. Research shows that surveillance and coercion policies are so pervasive that individuals often avoid public institutions altogether, often in fear that interacting with institutions such as hospitals or banks can increase surveillance and lead to apprehension by authorities (Brayne 2014; Remster and Kramer 2018). Indeed, surveillance policies themselves are frequently associated with increasing levels of administrative burden (Waldo, Lin, and Millett 2007).

Finally, hidden sentence policies have the potential to play a role in larger patterns of racial inequality. Government policies can either narrow or widen gaps in broad outcomes, including the labor market. Affirmative action programs, for example, contributed to a narrowing of wage inequality between whites and blacks (see Lyons and Pettit 2011, 258). On the other hand, the expansion of the criminal justice system is often conceptualized as a broader policy approach that has widened racial inequality. Given how common criminal justice contact is in the lives of minority citizens, and thus the accompanying hidden sentences, it is thus plausible that any exclusion created by hidden sentences could contribute to racial inequalities.

In sum, in this article we contribute to the large and growing literature on the labor market consequences of incarceration during young adulthood and consider whether and how state-level hidden sentence regimes are implicated in this relationship. Specifically, we ask how and to what extent state-level hidden sentences are responsible for producing inequalities in labor market outcomes between those with and without a history of incarceration. We ask three key research questions. First, are state-level hidden sentences associated with labor market outcomes among justice-involved young adults, net of other individual and state-level characteristics? Second, are disparities in labor market outcomes by incarceration status larger in states with more punitive hidden sentences? And, third, is there evidence that state-level hidden sentences are more consequential for justice-involved young adults of color than white young adults?

DATA AND APPROACH

To answer these questions, we draw data from several sources. Individual-level data are drawn from the 1997 cohort of the National Longitudinal Survey of Youth (NLSY97), a nationally representative longitudinal dataset that has regularly surveyed a cohort of American young adults born between 1980 and 1984. Respondents have been interviewed a maximum of seventeen times since the first interview in 1997, and were age thirty to thirty-six at the most recent round of data collection. The data have been used to examine the impact of incarceration and other forms of criminal justice contact on a variety of outcomes, including earnings and labor market participation (Apel and Powell 2019; Apel and Sweeten 2010). To capture state-level hidden sentences, we use data from the National Inventory of the Collateral Consequence of Conviction (NICCC), a collaborative effort of the American Bar Association and the National Institute of Justice that identifies all postconviction hidden sentences in all U.S. jurisdictions. The database covers a range of hidden sentences, including those related to employment, occupational licensing, housing, voting, and education (for more information on the NICCC, see Kaiser 2016, 129).² Finally, we capture other potentially relevant state-level characteristics (see below) by drawing on data from the U.S. Census, American Community Survey, Bureau of Justice Statistics, and the University of Kentucky Center for Poverty Research. These state-level data are linked to a restricted version of the NLSY97 that includes state of residence identifiers for respondents.

For all individual-level dependent and independent variables, we standardize the data by age to focus on a specific period of young adulthood, from age twenty-five to age thirty. We focus on this range for two reasons: first, because most young adults in this age range have completed their education, begun their transition to adulthood, entered the workforce, and attempted to begin their adult careers (Danziger and Ratner 2010); and, second, to be consistent with research on labor market outcomes in the transition to adulthood, which also focuses on young adults within this age range (Danziger and Ratner 2010; Silva 2012; Sironi 2018; Swartz, McLaughlin, and Mortimer 2017). By 2015, the last round of data collection, all NLSY97 respondents are at least thirty years old; research has established that the period of young adulthood is pivotal is establishing earnings trajectories and inequalities (Rindfuss 1991). Employment information is available for 7,599 respondents at or around the age thirty interview, and earnings are available for 7,505 respondents. We use listwise deletion on dependent and independent variables to omit a total of 1,818 respondents from the analyses, leaving us with a sample of 7,166.

RESULTS

We start by noting that that hidden sentences are ubiquitous across U.S. states, averaging more than eight hundred hidden sentences that may apply following a criminal conviction or other criminal status (mean = 868.27, SD = 312.73). In addition, states have an average of more than four hundred hidden sentences that are categorized as mandatory (mean = 441.29, SD = 173.94). We show state variation in hidden sentences, based on the levels of mandatory hidden sentences described, in figure 1. This map indicates that states with high levels of mandatory hidden sentences tend to be concentrated in the southern United States, although both high- and low-level states are observed in nearly every region.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Mandatory Hidden Sentences in the United States

Source: Authors’ compilation of NICCC (American Bar Association 2013).

Table 1 shows descriptive statistics for the full sample and by incarceration status. Between the ages of twenty-five and thirty, 5.6 percent (n = 405) of the NLSY97 respondents experienced a spell of incarceration and these individuals were disadvantaged in the labor market. Relative to their never-incarcerated counterparts, recently incarcerated individuals are less likely to report that they worked (either at all or full-time) leading up to the age thirty interview, and reported approximately $20,000 lower annual earnings.

Table 1 also shows that that most NLSY97 respondents live in states classified as either mid- or high-hidden sentences (low = 9.2 percent, mid = 45.3 percent, high = 45.5 percent). There are no differences in exposure to hidden sentences based on a history of recent incarceration. That is, the table shows that the levels of hidden sentences in the states where respondents with or without a recent incarceration live are not significantly different.

Hidden sentences do not appear to play an independent role in young adult employment outcomes, given that the results in table 2 show that a recent incarceration decreases the probability of employment at age thirty, and this relationship is not explained by exposure to hidden sentences.

After accounting for a host of individual-level potential sources of spuriousness and pre-incarceration employment, and consistent with existing work (see Apel and Ramakers 2019), we find in model 1 of table 2 that young adults with a recent incarceration are 16.6 percentage points less likely to report having been employed in any month in the prior year. This association is robust to the addition of the hidden sentences in model 2, and hidden sentences themselves are not directly associated with employment in the full sample. The results change little with the inclusion of other state-level variables in model 3. However, the results in model 4 indicate that hidden sentences exacerbate the consequences of a recent incarceration for young-adult employment. Recently incarcerated respondents living in states with low-levels of hidden sentences are not significantly less likely than their nonincarcerated peers to have worked at some point in the year leading up to their age thirty interview. Instead, the decreased probability of employment in young adulthood is concentrated among respondents living in states with middle and high levels of hidden sentences.

If hidden sentences work in combination with a recent incarceration to restrict any involvement in the labor market, they appear to be less consequential for full-time employment. The final model of table 2 (labeled also as model 4 to reflect that the only change is the outcome measure) shows that, among those who reported any employment in the year leading up to the age thirty interview, the association between incarceration and full-time employment is not significantly moderated by state-level hidden sentences.

In table 3, we predict any (panel A) and full-time employment (panel B) separately for white, black, and Hispanic young adults, and find evidence that hidden sentences are more strongly associated with adverse labor market outcomes for blacks than whites.

Looking first at model 1, which accounts for all covariates, incarceration is negatively associated with employment for all groups, but the association is stronger for blacks and Hispanics than it is for whites. When we include the interaction between hidden sentences and incarceration in model 2, we find that the employment penalty associated with incarceration is strongest in states that have more hidden sentences among blacks, but not among whites and Hispanics.⁴ This is consistent with the notion that hidden sentences may have more deleterious consequences for formerly incarcerated black young adults than for white young adults.

We find less evidence that hidden sentences moderate the association between incarceration and full-time employment by race (panel B), similar to our results for the full sample. In panel B, recently incarcerated blacks residing in states with middle and high levels of hidden sentences are still less likely than their nonincarcerated peers to be employed full time, but these coefficients do not reach conventional levels of statistical significance.

We estimate the association between incarceration, hidden sentences, and logged earnings in table 4. Across all models, we find evidence that the negative association between incarceration and earnings is stronger in states that have more hidden sentences. After adjusting for state- and individual-level controls in model 3, we replicate research and find that young adults with a recent incarceration report significantly lower earnings than their counterparts. In model 4, we find that this association is stronger in states with moderate and high levels of hidden sentences. In states with low hidden sentences, formerly incarcerated young adults report 38 percent lower wages than those who were not incarcerated, but this difference is not statistically significant. Both interaction terms, however, are negative and significant. This suggests that the incarceration wage disparity is substantially larger and statistically significant in states with more hidden sentences. And that, in our sample, the incarceration wage penalty is primarily driven by those who live in states with a more robust hidden sentence policy regime.

However, one question is whether this association is driven by non-earners, or whether we also observe disparities among wage earners. When we restrict our analyses to earners only (final column of table 4), we find less evidence that state hidden sentences moderate the association between incarceration and wages. When we restrict the dependent variable to those respondents who report some wages, the coefficients for the interaction terms—though consistent in direction with those in model 4—mostly fail to reach conventional levels of statistical significance. This implies that to the extent that state hidden sentence policies exacerbate disparities in wages by incarceration status, they do so by preventing access to employment, not by creating disparities among wage earners.

Finally, in table 5 we ask whether incarceration wage disparities across states vary by race-ethnicity. We find some evidence that incarceration wage disparities are largest in states with moderate and high levels of hidden sentences, and that these disparities are larger for blacks than they are for whites and Hispanics.

The results in panel A show that recently incarcerated whites and blacks who live in states with higher levels of hidden sentences earn significantly less than their nonincarcerated counterparts. Although the interaction terms are larger for formerly incarcerated black young adults, the difference between the interaction terms for whites and blacks is not statistically significant. When the outcome is restricted to earners, in panel B, some evidence indicates that formerly incarcerated blacks and Hispanics earn less if they live in states with higher levels of hidden sentences, but the coefficients mostly fail to reach conventional levels of statistical significance.

DISCUSSION

According to recent estimates, between seventy and one hundred million Americans have a criminal record (Vallas and Dietrich 2014), 7.3 million adults have been incarcerated (Shannon et al. 2017), and more than six hundred thousand are released from prison every year (Carson 2018). Incarceration—and criminal justice contact more broadly—has therefore become an increasingly common turning point in the life course for millions of people, especially young men of color. Research shows that these young people experience long and lasting consequences in the labor market that accumulate over time, one of the many ways that the criminal justice system creates and reinforces existing social inequality (Kirk and Wakefield 2017; Wakefield and Uggen 2010). The informal social stigma attached to a criminal label is often implicated in the labor market struggles of formerly incarcerated individuals (Pager 2003). In this study, we advance research by examining the role that formal stigma, operating through state-level laws and policies that limit social participation (Travis 2002), plays in the relationship between incarceration and labor market outcomes.

Our primary conclusion is that state-level hidden sentences appear to exacerbate the consequences of multiple kinds of criminal justice contact for labor market outcomes. Young adults with a recent incarceration, in particular, are less likely to find employment if they live in states that have higher levels of hidden sentences, and the earnings penalty of incarceration is larger in states with more hidden sentence laws and policies. In other words, the association between incarceration and labor market outcomes appears to depend in part on the state in which one lives and the overall burden that results from high levels of hidden sentences that accompany criminal justice processing. Individuals with recent incarceration spells in states with low levels of hidden sentences pay a smaller penalty than their peers in other states. Moreover, our supplemental analyses shows that the association between recent arrests or convictions and employment is larger in states with higher levels of hidden sentences. These findings align with those of Cadigan and Kirk (2020), who show how the management of monetary sanctions by courts can shape labor market experiences. Through the scheduling of compliance hearings or imposition of additional sanctions (such as suspension of a driver’s license, bench warrants, short period of incarcerations), courts create pressure points that make it difficult to find and maintain employment. Combined, these studies show that consequences of incarceration and other forms of criminal justice contact emerge as a result of postconviction policies, laws, and mechanisms of compliance.

We find that the moderating role of hidden sentences is a stronger predictor of entry into employment than it is of disparities among those who are employed (including wages among earners and full-time employment among the employed). This suggests that hidden sentences may limit initial access to the labor market, but if justice-involved individuals are able to get a job, these policies may play a smaller role in labor market inequalities. Future research should consider whether hidden sentences amplify firm-level discrimination that prevents the formerly incarcerated from gaining access to employment (Pager 2007). That is, the question remains whether hidden sentences limit access to the labor market, broadly, or only to certain sectors of the labor market that the justice-involved are likely to enter. Additional research should therefore examine how hidden sentences are linked to entry into specific occupations. In particular, it would be useful to better understand how individuals interact with the labor market in the face of hidden sentences, including segments of the labor market that are entirely blocked off, and whether shifts are discernable in occupational sectors of the justice-involved before and after incarceration via hidden sentence policies.

Our findings also raise concerns that hidden sentences may exacerbate racial disparities in socioeconomic outcomes, at least among the justice-involved. We find that hidden sentences are more strongly associated with access to employment for black formerly incarcerated respondents than for whites or Hispanics. This finding is in line with, and supports, recent research that shows evidence for a “racialized reentry” (Western and Sirois 2018), where the labor market penalties of incarceration are stronger for blacks than for whites (Apel and Powell 2019; Lyons and Pettit 2011). One interpretation of these findings is that hidden sentences institutionalize discrimination and make it easier for employers to discriminate against people of color. Pager (2003) argues that a criminal record works in combination with minority status to intensify stigma for blacks relative to whites; our findings show that this may work formally through state-level policies in combination with the informal stigma accompanying a conviction. Furthermore, even in the absence of discrimination or race-specific effects, hidden sentences may exacerbate racial disparities in employment outcomes given that blacks are far more likely to be subject to the pernicious consequences of hidden sentences than whites are (see Sugie and Turney 2017).

To our knowledge, this is the first attempt at quantifying the link between hidden sentences and labor market outcomes of justice-involved young adults, and we show that hidden sentences compound a history of incarceration, especially in states where such policies are more pervasive. However, our study is not without limitations. First, we are unable to directly measure the mechanisms linking hidden sentences with labor market outcomes. Although we argue that hidden sentences may limit access to the labor market directly through occupational restrictions and licensing, indirectly through social exclusion and limiting full citizenship, and more broadly through administrative burden, we are unable to directly measure these mechanisms. We do, however, find indirect support for them. In supplementary models, our findings were strongest for mandatory (relative to discretionary) hidden sentences, suggesting that hidden sentences are more consequential when they are more likely to be enforced or implemented. Our results were also similar for both employment-related hidden sentences and hidden sentences that are not directly related to employment (such as those that affect drivers’ licenses, welfare access, or residence locations). This constitutes suggestive evidence that hidden sentences—even those that do not restrict specific occupations—may play a role in labor market outcomes. That said, a careful examination of exactly how hidden sentences contribute to labor market struggles after incarceration is warranted.

Second, looming over any of the mechanisms we suggest is the broader issue of enforcement. Some hidden sentences are automatically triggered, such as those that automatically disseminate criminal records to various state and nonstate employers, impose mandatory civil fines and forfeitures, or make the justice-involved completely ineligible for public employment in many states. Others require discretionary action to be activated, such as when statutes grant medical, barbering, and other professional boards the power to consider arrests and convictions in the granting or renewal of occupational licenses. We focus here on those hidden sentences that are categorized as automatic, but it is clearly difficult to know exactly how or when hidden sentences become activated. In Wisconsin, for instance, barbering and cosmetology boards are required to deny, restrict, or suspend the licenses of those who are convicted of a felony, but it is unknown how often the boards actually comply with such a mandate. Such questions are more relevant with “mandatory” laws that involve more interpretation. In Delaware, for instance, anyone convicted of a crime “substantially related” to the practice of barbering or beautician services is ineligible for a license, but no legal guidelines list such crimes. One suggested avenue for future research, then, would be to follow a cohort of individuals as they progress though the system, examining what hidden sentences become activated at various stages and what actors drive those decisions. It would also be useful to examine whether hidden sentences are context dependent, given that larger views on crime and crime control across jurisdictions could affect the activation of hidden sentences.

Third, further research is needed to more clearly examine how a state’s labor force is affected by hidden sentences. The current framework of the NICCC database prohibits connecting specific hidden sentences to specific occupational domains, but this issue clearly warrants additional attention. Fourth, although we are interested in the effect of these policies on labor market outcomes for the justice-involved, our study is only correlational. Although we control for an array of characteristics that may confound our association of interest, we cannot speak to causality with this study design, and it would be useful to test our conclusions using methodologies that can more fully account for observed and unobserved sources of spuriousness. Finally, our measure of hidden sentences is time stable, and thus we cannot determine whether changes in hidden sentences change labor market outcomes. This is an important area for future research, especially in light of recent attempts to remove postconviction barriers and ease reintegration (Love and Schlussel 2019).

Additional research should also consider whether hidden sentences affect other outcomes associated with incarceration. Restrictions on access to financial aid, for example, could play an important role in the link between incarceration and educational attainment. More broadly, hidden sentences may drive larger patterns of socioeconomic inequality through differential access to credit and opportunities to accumulate wealth (Sykes and Maroto 2016). Given that states have increasingly been charged with and granted power to implement social and economic policies, future research might also consider how additional aspects of the state social policy regime are associated with outcomes among the justice-involved. For example, state-level policies regarding child support could be useful in understanding the relationship between parental incarceration and child outcomes. This is especially important given that Noah Zatz and Michael Stoll (2020) show that the threat of incarceration to enforce court-mandated work requirements can push noncustodial fathers into low-wage employment. Thus, more research is needed to determine how hidden sentences affect individual, family, and community outcomes following arrests, convictions, and periods of incarceration.

Taken together, our findings provide important insights on the relationship between criminal justice contact and socioeconomic inequality through labor market participation. In response to recent calls to push beyond average effects of incarceration (Apel and Ramakers 2019; Kirk and Wakefield 2017), our findings show how state-level policies of social exclusion interact with criminal justice contact to put justice-involved individuals at a distinct disadvantage in the labor market. This provides important insights on how the overall consequences of criminalization play out in jurisdictions across the country.