The false promise of risk assessments: epistemic reform and the limits of fairness
Open accessThe DOI matches its Crossref record; 118 of the 120 references it cites were found in a registry or an index; 2 of the 97 sources its claims cite were checked for support.
Abstract
Risk assessments have proliferated in the United States criminal justice system. The theory of change motivating their adoption involves two key assumptions: first, that risk assessments will reduce human biases by making objective decisions, and second, that risk assessments will promote criminal justice reform. In this paper I interrogate both of these assumptions, concluding that risk assessments are an ill-advised tool for challenging the centrality and legitimacy of incarceration within the criminal justice system. First, risk assessments fail to provide objectivity, as their use creates numerous sites of discretion. Second, risk assessments provide no guarantee of reducing incarceration; instead, they risk legitimizing the criminal justice system's structural racism. I then consider, via an "epistemic reform," the path forward for criminal justice reform. I reinterpret recent results regarding the "impossibility of fairness" as not simply a tension between mathematical metrics but as evidence of a deeper tension between notions of equality. This expanded frame challenges the formalist, colorblind proceduralism at the heart of the criminal justice system and suggests a more structural approach to reform. Together, this analysis highlights how algorithmic fairness narrows the scope of judgments about justice and how "fair" algorithms can reinforce discrimination.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
2 Valid · 0 Not valid · 42 Not checked · 104 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“the impossibility of fairness mathematically proves that it is impossible to achieve substantive equality through mechanisms of formal equality.”
§ The False Promise of Risk Assessments: Epistemic Reform and the Limits of Fairness > 1 INTRODUCTION
What the sources say
- Partly supportedcites [23] matched to “Fair prediction with disparate impact: a study of bias in recidivism prediction instruments” (2017) sourceExperimental — accuracy not yet measured
Our copy does contain a genuine mathematical impossibility result, but a narrow one: in Section 2.1 (p. 2) Chouldechova shows that 'when the recidivism prevalence differs between two groups, a test-fair score S c cannot have equal false positive and negative rates across those groups,' and Section 3 (pp. 3-4) derives disparate impact under the MinMax penalty policy. The claim's leap to a general proof that 'substantive equality' cannot be achieved 'through mechanisms of formal equality' is a legal-philosophical framing that appears nowhere in the paper; indeed the Introduction (p. 1) expressly cautions that 'fairness itself—along with the notion of disparate impact—is a social and ethical concept, not a statistical one,' which cuts against casting the result as a mathematical proof about equality. So the statistical-trade-off half of the attribution is supported while the substantive/formal-equality conclusion is not. (LLM, conf=0.78)
- Not checked yetcites [91]
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
None of the sources cited for this claim have been checked yet.
In the paper
“objectivity is only partially aligned with these terms [37,129].”
§ The False Promise of Risk Assessments: Epistemic Reform and the Limits of Fairness > 2 OBJECTIVITY
What the sources say
- Not checked yetcites [37]
- Not checked yetcites [129] matched to “Trust in Numbers: The Pursuit of Objectivity in Science and Public Life” (1995) source
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“Risk assessments have proliferated in the United States criminal justice system.”
§ The False Promise of Risk Assessments: Epistemic Reform and the Limits of Fairness > Abstract
What the sources say
- Not checked yetcites [131]
- Not checked yetcites [132]
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Second, risk assessments provide no guarantee of reducing incarceration; instead, they risk legitimizing the criminal justice system's structural racism.”
§ The False Promise of Risk Assessments: Epistemic Reform and the Limits of Fairness > Abstract
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: unclear — It is unclear whether it could be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“are made ineffectual by their individualistic conceptions of risk and bias”
§ The False Promise of Risk Assessments: Epistemic Reform and the Limits of Fairness > 1 INTRODUCTION
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
85 more claims
Create a free account to see all 148 claims and to verify the 95 references still open.
- It may state two separate claims as one. 42
- Probably not a substantial claim. 27
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“by using arrests as a proxy for crimes committed, COMPAS can only exacerbate racist policing practices, rather than transcending them [7,13,23,37,39].”
From Measurement and Fairness (2021) · cites this paper as [23]
Experimental — accuracy not yet measuredThe check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“We note that many variables are explicitly deined as łsensitivež by speciic legal frameworks, see [25,129,130,191,192,209,268,280,306] and the references therein.”
From Fairness in machine learning: A survey (2024) · cites this paper as [129]
Experimental — accuracy not yet measured
Citations
Cites
Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment.
Brewer, R. M., & Collins, P. H. (1992) · Contemporary Sociology · doi:10.2307/2074808 · Cited by 19,445
The new Jim Crow: mass incarceration in the age of colorblindness
(2010) · Choice Reviews Online · doi:10.5860/choice.48-1766 · Cited by 4,407
Barocas, S., & Selbst, A. D. (2014) · doi:10.2139/ssrn.2477899 · Cited by 3,048 · cited 5×
“This tension runs throughout debates in areas ranging from equality of opportunity [49] to antidiscrimination [33] to big data [10].”
BackgroundInfluentialPager, D. (2003) · American Journal of Sociology · doi:10.1086/374403 · Cited by 2,840
Trust in numbers: the pursuit of objectivity in science and public life
(1995) · Choice Reviews Online · doi:10.5860/choice.33-1499 · Cited by 2,619
Designs on Nature: Science and Democracy in Europe and the United States
Jasanoff, S. (2005) · Princeton University Press eBooks · doi:10.1515/9781400837311 · Cited by 2,338
(2016) · Cited by 2,334
Great American City: Chicago and the Enduring Neighborhood Effect
Sampson, R. J., & Wilson, W. J. (2012) · doi:10.7208/chicago/9780226733883.001.0001 · Cited by 2,214
Fair prediction with disparate impact: a study of bias in recidivism prediction instruments
Chouldechova, A. (2017) · Big Data · doi:10.1089/big.2016.0047 · Cited by 2,159 · cited 3×
“To see this challenge of making fair predictions in an unequal society, consider the recent statistical results regarding the “impossibility of fairness” [23, 91].”
BackgroundThe Effect of Education on Crime: Evidence from Prison Inmates, Arrests, and Self-Reports
Lochner, L., & Moretti, E. (2004) · American Economic Review · doi:10.1257/000282804322970751 · Cited by 2,032 · cited 1×
“118, 150] and blacks have been subjected to myriad forms of oppression (including redlining and segregation [141], the war on drugs [3, 86], and severe underfunding of schools [46]) that contribute to increasing crime [93, 97, 139, 142].”
Background
Research works printed but not linked: 20
- Barabas Chelsea, 2018, Proceedings of the 1st Conference on Fairness, Accountability and Transparency, 81
- Butler Paul, 2017, Chokehold: Policing Black Men
- Ciccolini Julie, 2018, Rationing Justice: Risk Assessment Instruments in the American Criminal Justice System. EuropeNow
- The, 2018, Leadership Conference Education Fund.
- 2018, Every Second: The Impact of the Incarceration Crisis on America's Families.
- Ben Green. 2018. Data Science as Political Action: Grounding Data Science in a Politics of Justice. arXiv preprint arXiv:1811.03435 (2018). Ben Green. 2018. Data Science as Political Action: Grounding Data Science in a Politics of Justice. arXiv preprint arXiv:1811.03435 (2018). · arXiv
- Green Ben, 2018, Machine Learning: The Debates workshop at the 35th International Conference on Machine Learning.
- Harris Kamala, 2017, 115th Congress
- Jasanoff Sheila, 2004, The idiom of co-production. In States of Knowledge: The Co-Production of Science and the Social Order, Sheila Jasanoff (Ed.). Routledge, Chapter, 1, 1
- Jasanoff Sheila, 2004, Ordering knowledge, ordering society. In States of Knowledge: The Co-Production of Science and the Social Order, Sheila Jasanoff (Ed.). Routledge, Chapter, 2, 13
- Karakatsanis Alec, 2015, Policing, Mass Imprisonment, and the Failure of American Lawyers, Harvard Law Review Forum, 128, 253
- Karakatsanis Alec, 2019, The Punishment Bureaucracy: How to Think About "Criminal Justice Reform, The Yale Law Journal Forum, 128, 848
- Mayson Sandra G., 2018, Dangerous Defendants, Yale Law Journal, 127, 490
- McLeod Allegra M., 2013, Confronting Criminal Law's Violence, The Possibilities of Unfinished Alternatives. Unbound: Harvard Journal of the Legal Left, 8, 109
- Martha Minow. 1991. Making All the Difference: Inclusion Exclusion and American Law. Cornell University Press. Martha Minow. 1991. Making All the Difference: Inclusion Exclusion and American Law. Cornell University Press.
- Review Harvard Law, 2019, Introduction, Harvard Law Review, 132, 1568
- Stevenson Megan T., 2017, Risk Assessment: The Devil'sin the Details. The Crime Report
- Sylvester Jared, 2018, Machine Learning: The Debates Workshop at the 35th International Conference on Machine Learning.
- Visvanathan Shiv, Science and Citizens: Globalization and the Challenge of Engagement.
- Yang Crystal S., 2017, Toward an Optimal Bail System, New York University Law Review, 92, 1399
Cited by
Baker, R. S., & Hawn, A. (2022) · International Journal of Artificial Intelligence in Education · doi:10.1007/s40593-021-00285-9 · Cited by 993
Fairness in machine learning: A survey
Caton, S., & Haas, C. (2024) · ACM Computing Surveys · doi:10.1145/3616865 · Cited by 646
Jacobs, A. Z., & Wallach, H. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445901 · Cited by 362
“Indeed, by using arrests as a proxy for crimes committed, COMPAS can only exacerbate racist policing practices, rather than transcending them [7, 13, 23, 37, 39].”
The flaws of policies requiring human oversight of government algorithms
Green, B. (2022) · Computer Law & Security Review · doi:10.1016/j.clsr.2022.105681 · Cited by 263
“15 Policymakers should also consider additional factors at this stage, such as whether the algorithm in question actually represents a desirable approach to reform (Green, 2020).”
BackgroundSelf-citationThe Flaws of Policies Requiring Human Oversight of Government Algorithms
Green, B. (2021) · doi:10.2139/ssrn.3921216 · Cited by 259
Self-citation
Metadata
- Authors
- Green, B.
- Affiliations
- Harvard University (US)
- Corresponding author
- Green, Ben
- ORCID
- Green, Ben
- Venue
- Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency
- Publisher
- ACM
- Published
- 2020-01-27
- Pages
- 594-606
- OpenAlex
- W3004155201
- Semantic Scholar
- 2c4cc62a1fe667e41ac7f577f173d57cbf26dec5
- Cited by
- 80
- References
- 120
- Citations by year
- 2026: 5 · 2025: 9 · 2024: 18 · 2023: 18 · 2022: 14 · 2021: 12 · 2020: 4
- FWCI
- 12.45
- Type
- conference paper
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- ACM Conference on Fairness, Accountability, and Transparency (FAccT) (publishedVersion)
- Funders
- National Science Foundation (DGE1745303)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Sociology and Political Science › Criminal Justice and Corrections Analysis
- Social Sciences › Law › Criminal Justice and Penology
- Keywords
- risk assessment, criminal justice reform, algorithmic fairness, algorithmic discrimination, structural racism, incarceration, equality, algorithmic decision-making
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 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash).
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: The false promise of risk assessments: epistemic reform and the limits of fairness. https://app.cochat.ai/papers/10.1145/3351095.3372869
- The paper
- Green, B. (2020). The false promise of risk assessments: epistemic reform and the limits of fairness. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. https://doi.org/10.1145/3351095.3372869
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.