CoChat

Incarceration, Recidivism, and Employment

Open access
Bhuller, Manudeep, Dahl, Gordon B., Løken, Katrine V., Mogstad, Magne · 2020 · Journal of Political Economy

The DOI matches its Crossref record; 57 of the 57 references it cites were found in a registry or an index; 1 of the 21 sources its claims cite were checked for support.

Abstract

Using a random judge design and panel data from Norway, we estimate that imprisonment discourages further criminal behavior, with reoffense probabilities falling by 29 percentage points and criminal charges dropping by 11 over a 5-year period. Ordinary least squares mistakenly reaches the opposite conclusion. The decline is driven by individuals not working prior to incarceration; these individuals increase participation in employment programs and raise their future employment and earnings. Previously employed individuals experience lasting negative employment effects. These findings demonstrate that time spent in prison with a focus on rehabilitation can be preventive for a large segment of the criminal population.

Overview

CoChat Summary

TL;DR

Using random assignment of Norwegian criminal cases to judges as an instrument, this paper finds that incarceration causally reduces subsequent criminal behavior, contrary to OLS estimates showing a positive association. The preventive effect is concentrated among defendants who were not employed before incarceration, for whom prison increases job-training participation and employment while cutting recidivism.

Key findings

  • IV estimates show incarceration lowers the probability of reoffending within 5 years by 27 percentage points and reduces the number of criminal charges over the same period by 10.
  • OLS estimates are positive: with all controls, incarceration is associated with a 6.6 percentage point higher probability of ever being charged over months 1–60 and 4.128 more charges.
  • The reduction in crime is driven by the previously non-employed: their likelihood of reoffending within 5 years falls from 96% to 50% (a 46 percentage point reduction), and they commit 22 fewer crimes; their employment rate rises by 40 percentage points within 5 years, and job-training participation rises by 34 percentage points.
  • For previously employed defendants, incarceration has no significant effect on reoffending or number of charges, but causes an immediate 25 percentage point drop in employment that persists to 5 years.

Methods

  • The paper links Norwegian administrative court, crime, prison, and labor-market registers covering the entire population, using a baseline sample of 33,509 non-confession criminal cases processed 2005–2009, with 23,345 unique defendants and 500 judges.
  • It exploits the random assignment of cases to judges who differ in stringency, measured as the leave-out mean judge incarceration rate in other cases, as an instrumental variable for incarceration in 2SLS models estimating local average treatment effects.
  • The paper reports first-stage, reduced-form, IV, and OLS estimates, plus specification checks for random assignment, exclusion, monotonicity, and heterogeneity by prior labor-market attachment.

Limitations

  • The paper states that it is a proof-of-concept and that evidence from other settings or populations would be useful to assess generalizability.
  • The paper notes that it observes charges, not actual crimes, so it is theoretically possible that ex-convicts learn to avoid being caught rather than reduce criminal activity; it argues this is unlikely because incarceration increases formal-sector employment.
  • The text makes evident that the findings are specific to Norway’s rehabilitative prison system, short average prison time (175 days actual time served), and institutional context, and that IV estimates apply to compliers whose incarceration status could change with judge assignment.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1086/705330

Cites

57 of 57 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

1 Valid · 2 Not valid · 14 Not checked · 228 Own results and interpretations

119 more claims

Create a free account to see all 245 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 46
  • It may state two separate claims as one. 70
  • The quote only partly states this claim. 22

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

3 of 3

Citations

Cites

Research works printed but not linked: 28

  1. Aebi M., 2015, Strasbourg: Council of Europe.
  2. Autor D., 2019, A.E.R., 109, 2613
  3. Buonanno P., 2013, A.E.R., 103, 2437
  4. Bureau of Justice Statistics. 2014., 2013, Prisoners
  5. Bureau of Justice Statistics. 2014., 2014, Prisoners
  6. Carneiro P., 2011, A.E.R., 101, 2754
  7. Dobbie W., 2018, A.E.R., 108, 201
  8. Dobbie W., 2015, A.E.R. (3), 1272
  9. Doyle J. J., 2007, A.E.R., 97, 1583
  10. Growing Inmate Crowding Negatively Affects Inmates, Staff, and Infrastructure
  11. Gottfredson D. M., Sentencing Decisions on Criminal Careers.
  12. Harrendorf S., 2010, Helsinki: European Institute for Crime Prevention and Control, affiliated with the United Nations (HEUNI).
  13. Henrichson C., 2012, New York: Vera Institute of Justice.
  14. Kling J., Discussion Paper no. 208
  15. Kling J., 2006, A.E.R., 96, 863
  16. Kostøl A. R., 2014, A.E.R., 104, 624
  17. Kristoffersen R., Finland, Iceland, Norway and Sweden 2009–2013.
  18. Lipton D., 1975, Albany: New York Office of Crime Control Planning.
  19. Maestas N., 2013, A.E.R., 103, 1797
  20. Martinson R., 1974, Public Interest, 35, 22
  21. Dømmes av Likemenn [Judged by Peers]
  22. Independent Budget Office, NYC Independent Budget Office Annual Data
  23. Pew Center, State of Recidivism: The Revolving Door of America’s Prisons
  24. How Effective Is Correctional Education, and Where Do We Go from Here? Santa Monica
  25. Raphael S., 2013, New York: Russell Sage Foundation.
  26. Redcross C., 2012, OPRE Report, 2011
  27. Stevenson M. T., 2018, J. Law, Econ., and Organization, 34, 511
  28. Walmsley R., World Prison Population List

Cited by

See all 335

Metadata

Authors
Bhuller, M., Dahl, G. B., Løken, K. V., & Mogstad, M.
Affiliations
  • Statistics Norway (NO)
  • National Bureau of Economic Research (US)
  • Center for Economic and Policy Research (US)
Venue
Journal of Political Economy
Published
2020-04
Volume
128
Issue
4
Pages
1269-1324
ISSN
  • 0022-3808 (ISSN-L)
  • 1537-534X
OpenAlex
W2516386981
Cited by
335
References
42
Citations by year
2026: 35 · 2025: 51 · 2024: 66 · 2023: 49 · 2022: 53 · 2021: 24 · 2020: 22
FWCI
57.26
Type
journal article
Language
en
Open access
Licence
cc-by-nc-sa
Locations
  • Journal of Political Economy (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • NORA - Norwegian Open Research Archives (submittedVersion, cc-by-nc-sa)
  • NORA - Norwegian Open Research Archives (submittedVersion, other-oa)
Funders
Norges Forskningsråd
Topics
  • Social Sciences › Sociology and Political Science › Criminal Justice and Corrections Analysis
  • Social Sciences › Law › Criminal Justice and Penology
  • Social Sciences › Sociology and Political Science › Crime Patterns and Interventions
Keywords
incarceration, recidivism, employment, criminal behavior, employment programs, earnings, rehabilitation, criminal justice policy

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-09-30, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-09-30 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.