Criminal Justice Contact Deepened Employment Income Losses During the COVID-19 Pandemic

A Criminal Record Became an Even Bigger Pay Cut During COVID-19

The economic shock of the COVID-19 pandemic did not hit all workers equally. For Americans who had previously been arrested or incarcerated, the disruption appears to have intensified an already persistent disadvantage, widening the gap between their employment income and that of people with no criminal justice contact. A longitudinal study of 7,502 participants found that the earnings penalties associated with arrest and incarceration became significantly larger in 2020 than they had been during the preceding decade. The results suggest that a criminal justice record is not simply a fixed obstacle encountered when someone applies for a job. Instead, it can become more damaging when the labor market contracts, employers have more applicants to choose from, and workers have less power to negotiate wages and hours. The study, published in the American Journal of Criminal Justice, provides one of the clearest comparisons yet of justice-related income losses before and during the pandemic.

The researchers, Hexuan Liu, Miranda Celeste and Francis Cullen, analyzed data from the National Longitudinal Survey of Youth 1997, a nationally representative panel that has followed people who were 12 to 16 years old at the end of 1996. The analysis focused on observations collected between 2011 and 2021, covering the years immediately before the pandemic and the period in which COVID-19 transformed the US labor market. After limiting the sample to respondents who were employed and had usable information on earnings and criminal justice histories, the final dataset contained 33,772 observations from 7,650 individuals. Participants reported annual income from wages, salaries, commissions and tips. The researchers converted all figures to constant 2020 dollars, allowing income comparisons across years without confusing inflation with genuine changes in earnings.

Criminal justice contact was measured in stages. Respondents were classified as having no recorded contact, having been arrested but not incarcerated, or having been incarcerated. The survey also recorded the cumulative number of arrests, the number of incarceration episodes and the total number of months spent incarcerated. This distinction mattered because contact with the justice system is not a single experience: an arrest may last hours or days, while incarceration can interrupt education, housing, employment and family life for much longer. To separate these effects from stable differences between people, the researchers used individual fixed-effects models. Statistically, this approach compares a person with themselves over time, removing the influence of characteristics that do not change, such as race at birth, birth year or other enduring traits. The models also adjusted for changing factors including education, occupation, health, marital status, number of children, region and urban or rural residence.

The descriptive pattern was striking even before the most detailed statistical tests. In 2010, participants who had been arrested but not incarcerated earned, on average, $5,391 less than those without criminal justice contact. By 2020, that gap had grown to $9,738. For people who had experienced incarceration, the difference increased from $18,046 in 2010 to $25,520 in 2020. Income rose for all groups over much of the study period, but growth slowed sharply between 2018 and 2020, and it slowed most severely among people with justice-system histories. By 2020, about 34 percent of participants had experienced an arrest and 9 percent had experienced incarceration, compared with approximately 31 percent and 6 percent, respectively, in 2010. The figures do not mean that every person with a record lost income because of that record, but they reveal a broad population-level pattern of widening economic separation.

The fixed-effects analysis estimated an average annual income reduction of $5,131 after arrest among respondents who had been arrested but never incarcerated. Following incarceration, annual employment income fell by an average of $12,140 relative to the same individuals’ pre-contact periods. When the researchers compared different years directly, the estimated penalties were largest in 2020. Income was approximately $7,228 lower after a first arrest without incarceration and $14,936 lower after a first incarceration in that year, after accounting for individual characteristics and time-varying conditions. Earlier estimates were smaller: the arrest-associated reductions ranged from $4,181 to $5,245 between 2010 and 2018, while incarceration-associated reductions ranged from $7,744 to $11,564. The interaction between justice contact and year indicates that the pandemic did more than preserve pre-existing inequality; it amplified the economic penalty linked to prior contact.

One possible explanation is that people with criminal justice histories were disproportionately concentrated in jobs that were vulnerable to shutdowns, reduced hours and layoffs. The study found that participants with justice-system contact were more likely to have lower educational attainment, work in service or manual occupations, report poorer health and be unmarried. Those jobs often provide fewer benefits, weaker employment protections and less control over scheduling. During the first months of the pandemic, businesses requiring close physical interaction were among the most disrupted, while workers in precarious positions had fewer opportunities to move into remote or secure employment. At the same time, employers facing an influx of applicants may have relied more heavily on quick screening signals, including criminal background checks. A record can therefore operate through several pathways at once: formal restrictions on particular occupations, employer stigma, interruptions to work experience and reduced bargaining power after someone is hired.

The findings also challenge the assumption that the severity of punishment alone determines the scale of economic harm. Among people who had been arrested, each additional arrest was associated with an estimated $1,969 reduction in employment income. By contrast, additional incarceration episodes were associated with lower income but did not produce a statistically significant relationship in the study’s models. Nor did additional months spent incarcerated predict further income reductions among those who had been imprisoned. These results suggest that the economic consequences may accumulate through repeated signals of justice-system contact rather than simply through the duration of confinement. The researchers caution that the measures of incarceration frequency and duration contained limited within-person variation, which makes those estimates less precise. Still, the overall pattern implies that the existence of a criminal justice history may itself function as a durable marker of vulnerability, regardless of whether the contact was brief or prolonged.

The burden was not distributed evenly across gender. Men generally earned more than women throughout the study period, and both arrest and incarceration were associated with income penalties that became more pronounced between 2018 and 2020. But among women, arrest without incarceration was linked to particularly substantial reductions in employment income. In some years, the losses associated with this relatively less severe form of contact approached those associated with incarceration. The researchers suggest that women’s greater exposure to fragmented employment, lower wages, caregiving responsibilities and occupational segregation may leave less room to absorb an additional disruption. Many women with justice-system contact may also be concentrated in service and care work, sectors heavily affected by school closures, workplace restrictions and the collapse of in-person demand. The statistically significant interaction between justice contact, year and gender indicates that the trajectory of the penalty changed differently for women and men during the study period.

Race and ethnicity revealed another distinct pattern. Justice-related income penalties increased especially sharply during the pandemic among White and Hispanic participants. Among Black participants, however, the penalty associated with incarceration remained consistently large across the entire period rather than showing a uniquely dramatic jump in 2020. That difference does not indicate less harm for Black workers. Instead, it suggests that the economic disadvantage associated with incarceration was already deeply entrenched, reflecting longstanding racial disparities in policing, employment, wages, occupational access and reentry conditions. Earlier research has documented a “double jeopardy” in hiring, in which Black applicants with criminal records face both racial discrimination and record-based stigma. The new analysis indicates that large-scale economic disruptions can interact with these existing structures in different ways: for some groups, the penalty may spike during a crisis, while for others it remains high in both stable and unstable labor markets.

The study has important limits. Because the analysis included employed respondents who reported employment income, it does not measure the full effect of criminal justice contact on labor-force participation. People who left the labor market entirely, could not find work or lacked reportable earnings were not captured in the main outcome. The income measure also excluded unemployment benefits, stimulus payments, informal support and other resources that may have buffered financial losses during the pandemic. The observational design cannot prove that an arrest or incarceration directly caused every measured reduction, even though fixed-effects models and robustness tests strengthened the analysis. The researchers found that the results were not explained by age differences or by an unusual increase in first-time justice contacts between 2018 and 2020. Taken together, the findings point to a broader lesson: criminal justice contact can act as a form of cumulative disadvantage, with effects that are shaped by the surrounding economy. When jobs disappear and employers become more selective, an old record can become a new economic crisis.