Abstract
This chapter provides a general introduction to forecasting criminal behavior in criminal justice settings. A common application is to predict at a parole hearing whether the inmate being considered for release is a significant risk to public safety. It may surprise some that criminal justice forecasts of risk have been used by decision-makers in the United States since at least the 1920s. Over time, statistical methods have replaced clinical methods, leading to improvements in forecasting accuracy. The gains were at best gradual until recently, when the increasing availability of very large datasets, powerful computers, and new statistical procedures began to produce dramatic improvements. But, criminal justice forecasts of risk are inextricably linked to criminal justice decision-making and to both the legitimate and illegitimate interests of various stakeholders. Sometimes, criticisms of risk assessment become convenient vehicles to raise broader issues around social inequality. There are, in short, always political considerations, ethical complexities, and judgement calls for which there can be no technical fix. The recent controversy about “racial bias” in risk instruments is a salient example.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
For readers will want to study more technical treatments, be sure to check what assumptions are being made. Too often, exciting results turn out on close inspection to depend on highly unrealistic requirements. Reading the abstracts alone will not suffice.
- 2.
Linear regression dates back to the late nineteenth century. Logistic regression was developed in the late 1950s. Both substantially predate machine learning, and for that matter, the discipline of computer science.
- 3.
Fit quality might be measured by mean squared error in the data being analyzed, and forecasting accuracy might be measured mean squared error in data used to evaluate the forecasts. A lot more will be said about both.
- 4.
In practice, procedures like stepwise regression and the lasso are interesting hybrids. The data analyst imposes some modeling structure but then allows the software find an acceptable model among many possibilities. As before, the model found is supposed to represent how the world works. Explanation is the primary goal. These are additional examples of conceptual boundaries that are quite permeable.
- 5.
Readers interested in such issues would benefit from studying Elements of Statistical Learning by Hastie et al. (2009).
- 6.
- 7.
In Philadelphia from January 1st to August 29th of 2018, at least 102 people 18 years of age and under were shot in the city, 18 of them fatally (Ubiñas 2018). That is more than the totals from all of the school mass shooting in the United States over that some time period (Amhed and Walker 2018) and probably over the past 2 years as well (Wilson 2018).
- 8.
This is a lot like much recent work in bioinformatics in which there is a lot of data but little theory to go with it. Microarrays and gene expression is a good example (Sánchez and Carme Ruíz de Villa 2018).
- 9.
This may seem inconsistent with the claim that in true forecasting situations the outcome is not known. To anticipate, test data require that the outcome be known because that is the only way forecasting accuracy can be determined. In real forecasting settings, the forecasting data do not include the outcome to be forecasted. The results from test data convey how accurate the forecasts can be.
References
Alexander, M. (2018) The newest Jm Crow: recent criminal justice reforms contain the seeds of a frightening system of “e-encarceration”. New York Times November 8th, https://www.nytimes.com/2018/11/08/opinion/sunday/criminal-justice-reforms-race-technology.htm
Amhed S. & Walker, C., (2018) There have been on the average 1 school shooting every week this year. CNN posted May 25th, 2018. https://www.cnn.com/2018/03/02/us/school-shootings-2018-list-trnd/index.html
Berk, R. A. (2008) Forecasting methods in crime and justice. In J. Hagan, K. L. Schepple, and T. R. Tyler (eds.) Annual Review of Law and Social Science 4 (173–192). Palo Alto: Annual Reviews.
Berk, R. A. (2012) Criminal Justice Forecasts of Risk: A Machine Learning Approach. New York: Springer.
Berk, R. A., Sorenson, S. B., & He, Y. (2005) Developing a practical forecasting screener for domestic violence incidents. Evaluation Review 29(4): 358–382.
Berk, R. A., Heirdari, H., Jabbari, S., Kearns, M., & Roth, A. (2018a) Fairness in criminal justice risk assessments: The State of the Art. Sociological Methods and Research, in press.
Borden, H. G. (1928) Factors predicting parole success.Journal of the American Institute of Criminal Law and Criminology19: 328–336.
Breiman, L. (2001b) Statistical modeling: two cultures (with discussion). Statistical Science 16: 199–231.
Burgess, E. M. (1928) Factors determining success or failure on parole. In A. A. Bruce, A. J. Harno, E. .W Burgess, and E. W., Landesco (eds.) The Working of the Indeterminate Sentence Law and the Parole System in Illinois (pp. 205–249). Springfield, Illinois, State Board of Parole.
Casey, P. M., Warren, R. K., & Elek, J. K. (2011) Using offender risk and needs assessment information at sentencing: guidance from a national working group. National Center for State Courts, www.ncsconline.org/.
Courtland, R, (2018) The bias detectives. Nature 558: 357–360.
Dean, C. W., & Dugan, T. J. (1968) Problems in parole prediction: a historical analysis. Social Problems 15: 450–459.
Elzayn, H., Jabbari, S., Jung, C., Kearns, M., Neel, S., Aaron Roth, A., & Schutzman, Z. (2018) Fair algorithms for learning in allocation problems. In Proceedings of ACM Conference on Fairness, Accountability, and Transparency (ACM FAT*?18).
Farrington, D. P. & Tarling, R. (1985) Prediction in Criminology. Albany: SUNY Press.
Gigi, A. (1990) Nonlinear Multivariate Analysis New York: Wiley.
Glaser, D. (1955) Testing correctional decisions. The Journal of Criminal Law, Criminology and Police Science 45: 679–684.
Goel, S., Rao, J.H., & Shroff, R. (2016) Precinct or prejudice? Understanding racial disparities in New York City’s stop-and-frisk policy. The Annals of Applied Statistics 10(1) 365–394.
Gottfredson, S. D., & Moriarty, L. J. (2006) Statistical risk assessment: old problems and new applications. Crime & Delinquency 52(1): 178–200.
Hastie, R., & Dawes, R. M. (2001) Rational Choice in an Uncertain World. Thousand Oaks: Sage Publications.
Hastie, T. & Tibshirani, R. (1993) Generalized Additive Models. New York: Chapman & Hall/CRC.
Hastie, T., Tibshirani, R., & Friedman, J. (2009) The Elements of Statistical Learning. Second Edition. New York: Springer.
Hyatt, J.M., Chanenson, L. & Bergstrom, M.H. (2011) Reform in motion: the promise and profiles of incorporating risk assessments and cost-benefit analysis into Pennsylvania Sentencing. Duquesne Law Review 49(4): 707–749.
Hvistendahl, M. (2016) Crime forecasters Science 353(6307): 1484–1487.
Knuth, D., (1968) The Art of Computer Programming: Fundamental Algorithms New York: Addison Wesley.
McCaffrey, D.F., Ridgeway, G., & Morral, A. (2004). Propensity score estimation with boosted regression for evaluating adolescent substance abuse treatment. Psychological Methods 9(4): 403–425.
Mease, D., Wyner, A.J., & Buja, A. (2007) Boosted classification trees and class probability/quantile estimation. Journal of Machine Learning Research 8: 409–439.
Monahan, J. (1981) Predicting Violent Behavior: An Assessment of Clinical Techniques. Newbury Park: Sage Publications.
Ohlin, L. E., & Duncan, O. D. (1949) The efficiency of prediction in criminology. American Journal of Sociology 54: 441–452.
Ohlin, L. E., & Lawrence, R. A. (1952) A comparison of alternative methods of parole prediction. American Sociological Review 17: 268–274.
Perry, W.L.. McInnis, B., Price, C.C., & Hollywood, J.S. (2013) Predictive Policing: The Role of Crime Forecasting in Law Enforcement Operations Santa Monica, CA: Rand Corporation.
Pew Center of the States, Public Safety Performance Project (2011) Risk/needs assessment 101: science reveals new tools to manage offenders. The Pew Center of the States. www.pewcenteronthestates.org/publicsafety.
Reiss, A. J. (1951) The accuracy, efficiency, and validity of a prediction instrument. American Journal of Sociology 17: 268–274.
Robinson, D. & Scognigs, C., (2018) The detection of criminal groups in real-world fused data: using the graph-mining algorithm “GraphExtract.” Security Informatics, published online, https://doi.org/10.1186/s13388-018-0031-9.
Roehl, J., O’Sullivan, C., Webster, D., & Campbell, J. (2005) Intimate partner violence risk assessment validation study, final report. National Institute of Justice, U.S. Department of Justice.
Skeem, J. .L., & Monahan, J. (2011) Current directions in violence risk assessment. Current Directions in Psychological Science 21(1): 38–42.
Sánchez, A., & Carme Ruíz de Villa, M. (2018) A tutorial review of microarray data analysis. Working paper, Department of Statistics, University of Barcelona. http://www.ub.edu/stat/docencia/bioinformatica/microarrays/ADM/slides/A_Tutorial_Review_of_Microarray_data_Analysis_17-06-08.pdf.
Ubiñas, H. (2018) In Philly, we’re burying our children, not our weapons. Philadelphia Daily News at Philly.com, posted August 24th, 2018 http://www2.philly.com/philly/columnists/helen_ubinas/helen-ubinas-philadelphia-violence-erase-the-rate-philadelphia-police-20180824.html.
Wager, S. & Athey, S. (2017) Estimation and inference of heterogeneous treatment effects using random forests. arXiv:1510.04342v4 [stat.ME].
Wilkins, L. T. (1980) Problems with existing prediction studies and future research needs. The Journal of Criminal Law and Criminology 71: 98–101.
Wilson, C. (2018) This chart shows the number of school shooting victims since Sandy Hook. Time Magazine, posted February 22nd, 2018. http://time.com/5168272/how-many-school-shootings/.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Berk, R. (2019). Getting Started. In: Machine Learning Risk Assessments in Criminal Justice Settings. Springer, Cham. https://doi.org/10.1007/978-3-030-02272-3_1
Download citation
DOI: https://doi.org/10.1007/978-3-030-02272-3_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-02271-6
Online ISBN: 978-3-030-02272-3
eBook Packages: Computer ScienceComputer Science (R0)