Mass incarceration increases future crime by destroying social networks

The claim

At a certain scale, incarceration stops reducing crime and begins to produce it. The mechanism is structural: when a sufficiently large fraction of working-age men are removed from high-poverty urban neighborhoods — through a combination of aggressive policing, mandatory minimum sentences, and prosecutorial practices that are explicitly place-concentrated — the social institutions that informally regulate behavior collapse. The networks of mutual obligation, supervision, employment connection, and family stability that criminologists call “collective efficacy” depend on the presence of adults in communities. Mass incarceration, operating at American scale and with American geographic concentration, does not merely punish individuals — it amputates the connective tissue of specific communities, leaving behind elevated poverty, fractured families, reduced social trust, and ultimately more crime than would have occurred with a less aggressive carceral strategy.

This is not a claim that incarceration never reduces crime. Incarceration prevents crime by incapacitating people who would otherwise offend — the incapacitation effect is real and well-documented. The claim is about marginal returns at scale: that the United States, which incarcerates at roughly eight times the rate of Germany and ten times the rate of Norway, has long since exhausted the incapacitation dividend and is now operating in a regime where additional incarceration — and the community destruction it entails — is net-criminogenic.

The mechanism

Robert Sampson and colleagues’ research on collective efficacy provides the clearest theoretical framework. Collective efficacy is a neighborhood-level property: the combination of social cohesion among residents and their shared willingness to intervene in public-order problems. It is what allows communities to self-regulate — to confront disruptive behavior, maintain norms, and provide informal supervision over youth. Sampson, Raudenbush, and Earls (1997) showed in a landmark study of 343 Chicago neighborhoods that collective efficacy was the strongest predictor of violent crime rates, stronger than poverty, race composition, or residential instability taken individually. Its effects held after extensive controls.

Mass incarceration disrupts collective efficacy through several interlocking mechanisms. The direct removal of adults — predominantly men between 18 and 40 — reduces the density of stable relationships available to anchor community norms. Children lose fathers. Extended family networks lose members who provided employment connections, childcare, and economic stability. Social trust, which Sampson’s work shows takes years to accumulate and is damaged rapidly by chronic institutional contact with an adversarial criminal justice system, erodes.

Todd Clear’s “coercive mobility” thesis, developed in Imprisoning Communities (2007), extends this analysis. Clear argues that neighborhoods with high incarceration rates experience not just the loss of residents to prison but the disruptive churn of people entering and exiting incarceration — a pattern that prevents the stable relationships collective efficacy requires from forming at all. Communities that lose 1–2% of their adult male population to incarceration annually are perpetually reconstituting social networks, making durable community institutions structurally impossible to maintain.

The economic mechanism compounds the social one. Western’s Punishment and Inequality in America (2006) documents that incarceration dramatically reduces lifetime earnings — not just during the period of confinement but through employer stigma, network disruption, and occupational licensing restrictions that persist for decades. When concentrated in specific neighborhoods, this earnings penalty translates into neighborhood-level economic degradation: reduced household income, housing instability, and contraction of the local informal economy that employed people without formal credentials. Economic deprivation is itself a criminogenic condition, so mass incarceration reproduces the conditions it was purportedly designed to suppress.

The evidence

Clear’s natural experiment in Tallahassee

Todd Clear and colleagues conducted a systematic analysis of incarceration and crime rates in Tallahassee, Florida, tracking incarceration rates at the census-tract level and examining their relationship to subsequent crime. The finding was counterintuitive in a way that the aggregate national statistics obscure: neighborhoods where incarceration rates exceeded approximately 2% of the residential adult population showed subsequent increases in crime, not decreases. Below that threshold, incarceration had a modest crime-reducing effect consistent with incapacitation. Above it, the community disruption effect overwhelmed incapacitation, producing a net increase. The relationship was specifically non-linear and specifically place-concentrated — average incarceration rates in the city as a whole were not predictive; it was the concentration in specific communities that drove the criminogenic effect.

This dose-response relationship is the most compelling evidence that the mechanism operates as described. A claim that “incarceration generally increases crime” would be obviously false. The claim is that incarceration beyond a threshold, concentrated in already-disadvantaged communities, produces criminogenic community destruction. Clear’s data supports exactly that formulation.

Sampson, concentrated disadvantage, and collective efficacy

Sampson’s decades of research on Chicago neighborhoods, synthesized in Great American City (2012), provides the structural context. Concentrated disadvantage — the geographic clustering of poverty, unemployment, family disruption, and residential instability — predicts violent crime far more powerfully than individual-level risk factors. Mass incarceration actively produces concentrated disadvantage by stripping communities of economic actors, creating stigmatized returning citizens who face formal and informal exclusion, and imposing monitoring conditions (probation, parole, GPS supervision) that restrict mobility and employment.

Sampson and Loeffler (2010) extended this analysis to show explicitly that incarceration rates are spatially clustered in precisely the neighborhoods already characterized by high collective disadvantage — and that this concentration means incarceration’s costs and disruptions are geographically amplified in communities least equipped to absorb them. The counterfactual comparison is not evenly distributed incarceration across the whole population; it is the specific geographic targeting that characterizes American carceral practice.

Wakefield and Uggen: the stratification pathway

Wakefield and Uggen’s (2010) review in the Annual Review of Sociology documents the multigenerational stratification consequences of mass incarceration. Children of incarcerated parents show elevated rates of behavioral problems, educational disruption, housing instability, and food insecurity. These are not trivial effect sizes: paternal incarceration increases child behavior problems by approximately 3 percentage points after controlling for pre-incarceration family characteristics. Given that approximately 1.7 million children had an incarcerated parent on any given day during the peak of American mass incarceration, the aggregate developmental harm is substantial.

This pathway matters for the long-run crime question because childhood behavioral problems and educational disruption are among the strongest predictors of subsequent criminal involvement. Mass incarceration’s effects on children — transmitted through family economic instability, residential disruption, caregiver stress, and the direct trauma of parental separation — reproduce criminogenic conditions in the next generation. The strategy that claimed to reduce crime by incarcerating the parents is producing the conditions for elevated crime among their children.

The NAS 2014 report and marginal deterrence

The National Academy of Sciences 2014 comprehensive review The Growth of Incarceration in the United States — a panel of prominent criminologists, economists, and sociologists convened specifically to assess the evidence — reached a set of conclusions that are important for this claim. The panel found that the quadrupling of US incarceration between 1972 and 2007 had uncertain and likely modest effects on crime rates, with most crime decline attributable to other factors (demographic change, policing strategy, labor market conditions). The marginal deterrent effect of incarceration — the crime-reduction benefit of adding additional incarcerated people to an already large prison population — was found to be near zero or potentially negative at current American scale. The NAS review did not make the stronger claim that mass incarceration increases crime, but its finding that enormous additional incarceration produced no detectable crime reduction is consistent with Clear’s threshold hypothesis: above a certain scale, incapacitation benefits and criminogenic community costs approximately cancel, producing null aggregate effects.

Cross-national evidence

Germany incarcerates at 69 per 100,000, Norway at 54, Finland at 51, the Netherlands at 59, Canada at 104. The United States incarcerates at 531 per 100,000 — roughly eight to ten times these peer-nation rates. If incarceration were the primary mechanism of crime suppression, these countries would face severe crime problems that the United States does not. They do not. Germany’s homicide rate is approximately 0.8 per 100,000; Norway’s is 0.5; Finland’s is 1.6; the Netherlands’ is 0.9; Canada’s is 1.8. The US homicide rate is approximately 6.3 per 100,000 — three to twelve times higher than peer nations that incarcerate at a fraction of the American rate.

This comparison does not establish that American incarceration causes American violence. The confounding factors are substantial: gun availability, inequality, historical racial violence, urban geography. But the cross-national evidence is decisive against the premise of the opposing claim — that mass incarceration is what protects American communities from even higher crime. Peer nations have demonstrated that crime at American levels is not the structural consequence of low incarceration. The American pattern of high incarceration and high violence is exceptional, not representative.

State-level decarceration evidence

Pew Charitable Trusts tracked states that reduced their prison populations between 2007 and 2012 against states that increased them, controlling for economic conditions. The decarceration states saw crime rates fall as fast as the incarceration-increasing states — no crime penalty was detectable from reducing the prison population at the margins. New York dramatically reduced its prison population from the mid-1990s onward while experiencing the sharpest urban crime decline of any major American city. New Jersey reduced its prison population by 26% between 1999 and 2012 while violent and property crime continued to fall. Connecticut cut its prison population by 17% between 2008 and 2012 without an increase in crime. These are not anomalies — they are the systematic result across the states that have reduced incarceration at modest scale, and they are inconsistent with a model in which current incarceration levels are necessary for crime suppression.

Who benefits

CoreCivic and GEO Group, the two largest private prison companies in the United States, have revenues that depend directly on occupied beds. Their combined 2022 revenues exceeded $4 billion. Both companies have explicitly listed criminal justice reform — sentence reduction, decarceration, alternatives to incarceration — as material risk factors in SEC filings. CoreCivic spent approximately $1.4 million on federal lobbying in 2022; GEO Group spent approximately $1.2 million. Their political donations historically concentrate on candidates and ballot initiatives that oppose sentencing reform.

Rural politicians and the communities whose economies depend on correctional facilities have structural incentives to oppose decarceration. In states like Texas, Louisiana, and Mississippi, correctional facilities provide public-sector employment in regions with few alternatives, and reducing prison populations threatens both employment and the Census-based political representation that counts incarcerated people in facility locations rather than home communities. This produces a lobby for continued high incarceration that is distinct from crime-reduction arguments and operates through appropriations committees and state legislatures rather than public advocacy.

Law enforcement associations — the National Sheriffs’ Association and state prosecutors’ organizations — generally oppose sentence reduction and diversion programs that would reduce caseloads and departmental resources. Fear-of-crime rhetoric, which mass incarceration claims to address, is a resource mobilization tool for these organizations, making high crime perception functionally useful even when crime is declining.

The counter

The strongest challenge to this claim is the incapacitation evidence. Loftin and McDowall (1982) and subsequent researchers have documented that incarceration does reduce crime through incapacitation: a person who is confined cannot commit crimes in the community. William Spelman’s influential work estimated that incarceration accounted for perhaps 25% of the crime decline of the 1990s — a substantial share. Steven Levitt’s (2004) widely cited analysis found significant crime-reduction effects of incarceration. These are not fabricated results; the incapacitation mechanism is real and the evidence for it is not negligible.

The individual-level heterogeneity argument also has empirical support. A small number of high-frequency offenders account for a disproportionate share of violent crime, and targeted incarceration of this group could achieve meaningful crime reduction without the community-destruction effects of mass incarceration. The problem is that American incarceration practice has never approximated surgical targeting of high-frequency offenders — it is heavily weighted toward low-level drug offenses and non-violent property crimes that do not involve the high-rate violent offenders whose incapacitation would provide the largest crime-reduction dividend.

The timing argument presents a genuine complication. American crime rates did fall sharply from the early 1990s through the 2010s, and this period overlapped with peak incarceration. The magnitude of the incarceration contribution remains contested: Sharkey, Torrats-Espinosa, and Takyar (2017) attribute substantial crime reduction to expanded non-profit social service organizations; Blumstein and Rosenfeld have emphasized demographic factors; Chalfin and McCrary emphasize policing. The honest position is that incarceration contributed something to 1990s crime decline, that its marginal contribution became negligible or negative as it reached current scale, and that the communities bearing the costs of concentrated incarceration are not the communities experiencing the concentrated benefits of incapacitation.

References

Clear, T. R. (2007). Imprisoning communities: How mass incarceration makes disadvantaged neighborhoods worse. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195305791.001.0001

Loftin, C., & McDowall, D. (1982). The police, crime, and economic theory: An assessment. American Sociological Review, 47(3), 393–401. https://doi.org/10.2307/2095074

National Academy of Sciences. (2014). The growth of incarceration in the United States: Exploring causes and consequences. National Academies Press. https://doi.org/10.17226/18613

Sampson, R. J. (2012). Great American city: Chicago and the enduring neighborhood effect. University of Chicago Press.

Sampson, R. J., & Loeffler, C. (2010). Punishment’s place: The local concentration of mass incarceration. Daedalus, 139(3), 20–31. https://doi.org/10.1162/DAED_a_00022

Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277(5328), 918–924. https://doi.org/10.1126/science.277.5328.918

Sharkey, P., Torrats-Espinosa, G., & Takyar, D. (2017). Community and the crime decline: The causal effect of local nonprofits on violent crime. American Sociological Review, 82(6), 1214–1240. https://doi.org/10.1177/0003122417736289

Wakefield, S., & Uggen, C. (2010). Incarceration and stratification. Annual Review of Sociology, 36, 387–406. https://doi.org/10.1146/annurev.soc.012809.102551

Western, B. (2006). Punishment and inequality in America. Russell Sage Foundation.

Western, B., & Muller, C. (2013). Mass incarceration, macrosociology, and the poor. The Annals of the American Academy of Political and Social Science, 647(1), 166–189. https://doi.org/10.1177/0002716213475421