Article: Silver, Ian A., D’Amato, Christopher, and Wooldredge, John. 2024. Exacerbating Inequality over the Life-Course: Examining Race Differences in the Reciprocal Effects between Incarceration and Income. Social Forces, 102(3), 839–860. https://doi.org/10.1093/sf/soad113
1. Background (PDF and Article)
Racial disparities in income and incarceration are deeply entrenched in American society. Black individuals are more likely to be arrested, detained, and sentenced to longer prison terms than White individuals—even when controlling for offense severity. Simultaneously, Black workers earn substantially less than White workers across nearly every educational and occupational level. These disparities, however, have long been treated as distinct outcomes resulting from structural inequalities. In response to this longstanding tradition, Silver et al., 2024 assessed if income influences incarceration and incarceration influences income, and if this interrelationship was stronger for Black or White Americans
Existing research has generally examined the effects of incarceration on labor market outcomes, finding that incarceration reduces income, limits employment opportunities, and increases economic instability. However, these studies often rely on cross-sectional data or fail to account for the reciprocal influence of income on incarceration. For instance, individuals with lower incomes are more likely to be detained pretrial due to inability to afford bail and may face worse legal outcomes due to lack of quality legal representation—both factors increasing incarceration risk. Thus, low income can lead to incarceration, and incarceration can further suppress income, forming a vicious cycle.
Importantly, the magnitude and direction of these reciprocal effects may not be uniform across racial groups. Due to historical and ongoing discrimination, Black individuals face compounding disadvantages in both the labor market and criminal justice system. Despite this, few studies have utilized longitudinal data to examine how the relationship between incarceration and income unfolds over time—and fewer still have explicitly modeled race differences in these dynamics.
Silver, D’Amato, and Wooldredge address these gaps by employing a longitudinal panel of nationally representative youth followed from adolescence through early adulthood. Their central question: does incarceration suppress income more strongly for Black individuals than for White individuals, and does income, in turn, differentially influence future incarceration risk?
2. Summary of Findings
The results suggest that incarceration has a stronger and more persistent negative impact on income for Black individuals, while the effects are limited or absent for Whites. Four primary findings stand out.
First, Black respondents spent significantly more time incarcerated and earned significantly less than Whites between 2000 and 2016. The racial gap in income increased through young adulthood, with Blacks earning on average $15,000 less than Whites by 2016.
Second, a 1-month increase in incarceration resulted in a $300–$600 decrease in income the following year for Black respondents, while no consistent effect was found for Whites.
Third, changes in incarceration over time were associated with steeper reductions in income growth for Blacks. Blacks who experienced a one-month increase in incarceration from 2000–2014 had income growth that was $9,405 less than similarly incarcerated Whites and $16,905 less than non-incarcerated Whites by 2016.
Finally, while both groups demonstrated stability in incarceration and income from year to year, Blacks showed higher persistence in incarceration and lower year-to-year income growth.
3. Implications
These findings suggest that incarceration may act as a mechanism that contributes to economic inequality over the life-course. For Black individuals, incarceration is not just a temporary disruption—it has long-term consequences that reduce earnings potential and widen existing racial income gaps.
Policy interventions should aim to disrupt this cycle. Reducing incarceration through alternatives to detention, improving access to quality legal representation, and expanding diversion and reentry programs. Additionally, targeted investments in employment services, job training, and income support for formerly incarcerated individuals could help offset the economic harm associated with incarceration.
4. Data and Methods
This study uses data from the National Longitudinal Survey of Youth 1997 (NLSY97), a cohort study of 8,984 youth born between 1980 and 1984 living in the United States. The analytical sample included 5,918 respondents who self-identified as either Black (N = 1,925) or White (N = 3,993) and had valid data on key variables for at least 8 out of 14 observation periods between 2000 and 2016. Respondents with missing data on income or incarceration were imputed using multiple imputation with chained equations (predictive mean matching or logistic regression, depending on variable type).
Primary Variables:
Income: Annual pre-tax earnings (from wages, salary, or freelance work) were self-reported from 2000 to 2016. Income was top-coded and rescaled by multiplying by 0.00001 to facilitate model convergence without altering substantive interpretation.
Incarceration: Respondents reported start and end dates for all periods of confinement in jail or prison between 1992 and 2018. The number of months spent incarcerated was summed annually from 2000 to 2016.
Covariates: Eight time-invariant covariates were included to reduce confounding:
- Prior incarceration (1992–1999)
- Parental incarceration before age 16
- Childhood household net worth (1997)
- Early-life hardship (e.g., homelessness, living without utilities)
- Presence of both biological parents in 1997
- Age in 1997
- Biological sex
- Indicator for being in the cross-sectional vs. supplemental NLSY97 sample
Analytical Strategy:
To capture the reciprocal and time-varying effects between incarceration and income, the authors employed a Lagged Latent Curve Model with Structured Residuals (LCM-SR)—a technique that adjusts for both within- and between-individual effects. Unlike traditional fixed-effects or mixed-effects models, the LCM-SR can account for:
- The confounding influence of prior income on both future income and incarceration.
- The temporal stability of each construct (e.g., persistence of income or incarceration patterns over time).
- Latent growth trajectories that vary across individuals and racial subgroups.
A multi-group version of the LCM-SR was estimated to allow direct comparison between Black and White subsamples. Models were estimated using the Distributionally Weighted Least Squares (DLS) estimator in the lavaan package in R, selected for its robustness to violations of normality.
To benchmark findings, the authors also replicated several common statistical approaches used in the literature—including fixed effects and inverse probability weighted models—but found that only the LCM-SR fully captured the complex bidirectional relationship between income and incarceration while accounting for temporal proximity and unobserved heterogeneity.
5. Conclusion
This study reveals how incarceration influences racial inequality in income by creating feedback loops that disadvantage Black individuals across the life course. In particular, time spent incarcerated reduced subsequent income for Black individuals but not White individuals, suggesting that post-incarceration economic outcomes are not just a function of individual efforts. This suggests that structural barriers after incarceration could hinder Black individuals more than White individuals. Addressing this cycle and the cycle of reincarceration could potentially help reduce the gap in income between Black and White individuals in the United States.
Disclosure: This research brief was prepared by ChatGPT and reviewed/edited by Ian A. Silver.
ianasilver.com 