Rethinking recidivism through a causal lens ...
Predictive modeling of criminal recidivism, or whether people will re-offend in the future, has a long and contentious history. Modern causal inference methods allow us to move beyond prediction and target the "treatment effect" of a specific intervention on an outcome in an observational...
| Main Authors: | , |
|---|---|
| Format: | Article in Journal/Newspaper |
| Language: | unknown |
| Published: |
arXiv
2020
|
| Subjects: | |
| Online Access: | https://dx.doi.org/10.48550/arxiv.2011.11483 https://arxiv.org/abs/2011.11483 |
| id |
ftdatacite:10.48550/arxiv.2011.11483 |
|---|---|
| record_format |
openpolar |
| spelling |
ftdatacite:10.48550/arxiv.2011.11483 2024-09-09T19:38:12+00:00 Rethinking recidivism through a causal lens ... Shirvaikar, Vik Lakshminarayan, Choudur 2020 https://dx.doi.org/10.48550/arxiv.2011.11483 https://arxiv.org/abs/2011.11483 unknown arXiv arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Machine Learning cs.LG Computers and Society cs.CY Applications stat.AP FOS Computer and information sciences Article article Preprint CreativeWork 2020 ftdatacite https://doi.org/10.48550/arxiv.2011.11483 2024-06-17T09:18:46Z Predictive modeling of criminal recidivism, or whether people will re-offend in the future, has a long and contentious history. Modern causal inference methods allow us to move beyond prediction and target the "treatment effect" of a specific intervention on an outcome in an observational dataset. In this paper, we look specifically at the effect of incarceration (prison time) on recidivism, using a well-known dataset from North Carolina. Two popular causal methods for addressing confounding bias are explained and demonstrated: directed acyclic graph (DAG) adjustment and double machine learning (DML), including a sensitivity analysis for unobserved confounders. We find that incarceration has a detrimental effect on recidivism, i.e., longer prison sentences make it more likely that individuals will re-offend after release, although this conclusion should not be generalized beyond the scope of our data. We hope that this case study can inform future applications of causal inference to criminal justice analysis. ... : 16 main pages, 1 appendix page, 3 figures, 8 tables ... Article in Journal/Newspaper DML DataCite |
| institution |
Open Polar |
| collection |
DataCite |
| op_collection_id |
ftdatacite |
| language |
unknown |
| topic |
Machine Learning cs.LG Computers and Society cs.CY Applications stat.AP FOS Computer and information sciences |
| spellingShingle |
Machine Learning cs.LG Computers and Society cs.CY Applications stat.AP FOS Computer and information sciences Shirvaikar, Vik Lakshminarayan, Choudur Rethinking recidivism through a causal lens ... |
| topic_facet |
Machine Learning cs.LG Computers and Society cs.CY Applications stat.AP FOS Computer and information sciences |
| description |
Predictive modeling of criminal recidivism, or whether people will re-offend in the future, has a long and contentious history. Modern causal inference methods allow us to move beyond prediction and target the "treatment effect" of a specific intervention on an outcome in an observational dataset. In this paper, we look specifically at the effect of incarceration (prison time) on recidivism, using a well-known dataset from North Carolina. Two popular causal methods for addressing confounding bias are explained and demonstrated: directed acyclic graph (DAG) adjustment and double machine learning (DML), including a sensitivity analysis for unobserved confounders. We find that incarceration has a detrimental effect on recidivism, i.e., longer prison sentences make it more likely that individuals will re-offend after release, although this conclusion should not be generalized beyond the scope of our data. We hope that this case study can inform future applications of causal inference to criminal justice analysis. ... : 16 main pages, 1 appendix page, 3 figures, 8 tables ... |
| format |
Article in Journal/Newspaper |
| author |
Shirvaikar, Vik Lakshminarayan, Choudur |
| author_facet |
Shirvaikar, Vik Lakshminarayan, Choudur |
| author_sort |
Shirvaikar, Vik |
| title |
Rethinking recidivism through a causal lens ... |
| title_short |
Rethinking recidivism through a causal lens ... |
| title_full |
Rethinking recidivism through a causal lens ... |
| title_fullStr |
Rethinking recidivism through a causal lens ... |
| title_full_unstemmed |
Rethinking recidivism through a causal lens ... |
| title_sort |
rethinking recidivism through a causal lens ... |
| publisher |
arXiv |
| publishDate |
2020 |
| url |
https://dx.doi.org/10.48550/arxiv.2011.11483 https://arxiv.org/abs/2011.11483 |
| genre |
DML |
| genre_facet |
DML |
| op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
| op_doi |
https://doi.org/10.48550/arxiv.2011.11483 |
| _version_ |
1809907178566320128 |