Policing the future, disrupting urban policy today. Predictive policing, smart city, and urban policy in Memphis (TN)

ABSTRACT

Significant resources and efforts have been devoted, especially in the USA, to develop predictive policing programs. Predictive policing is, at the same time, one of the drivers of the birth, and the ultimate material enactment of, the anticipatory logics that are central to the smart city discourse. Quite surprisingly, however, critical analyses of the smart city have remained divorced from critical criminology and police studies. To fill this gap, this article sets out the first critical, in-depth empirical discussion of Blue CRUSH, a predictive policing program developed in Memphis (TN, USA), where its implementation intersects long-term austerity for urban policy. The article, first, shows that there is no evidence of Blue CRUSH’s capacity to prevent crime, thus adding empirical material to skepticism over the role of predictive policing as a policy solution in the first place. And, second, it argues that, rather than making crime a matter of technological solutions, predictive policing shifts the politics therein – in short, it contributes to the expansion of policing into the field of urban policy at the same time as it disrupts present police work. These takeaways allow to further the critique of the salvific promises implicit in the smart city discourse.

KEYWORDS:

• Urban security
• critical urban studies
• crime prevention
• anticipatory logics
• preemption
• algorithmic governance

Acknowledgments

Simone Tulumello acknowledges the financial support by the Fundação para a Ciência e Tecnologia (DL57/2016/CP1441/CT0007) and a Fulbright Visiting Scholar grant (US – Italy Commission) sponsored by the US Department of State and the Italian Ministry of Foreign Affairs. A preliminary version of this article was presented at the II International Conference Urban E-Planning (Lisbon, April 2017). We are grateful to the editors and two anonymous reviewers for their constructive commentaries.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Available at www.ucrdatatool.gov/.

2. Five interviews and three work meetings with: academicians (University of Memphis, Criminology and Criminal Justice); municipal civil servants (MPD; Parks and Neighborhoods); a retired criminologist and former consultant of MPD; an activist from the Mid–South Peace and Justice Center; and a lawyer, chair of Memphis Crime Commission.

3. Thanks to a partnership between the University of Memphis, Department of City and Regional Planning (activities coordinated by Laura Saija and Antonio Raciti), and the local Community Development Corporation (CDC). Fieldwork consisted in the participation to monthly meetings of the CDC, and further events and activities; and the collaboration in the production of a participatory plan for community development.

4. Hot-spot policing is grounded on the evidence of concentration of crimes, in US cities, in “very small places”: “the appeal of focusing limited resources on a small number of high-activity crime places is straightforward. If we can prevent crime at these hot spots, then we might be able to reduce total crime” (Braga et al., 2014). See Hope (2018) for a compelling critique of the very roots of this understanding.

5. Theories such as routine activity (Brunet, 2002) or opportunity (Felson & Clarke, 1998), which explain crime as the result of the encounter of a rationally motivated criminal with a potential victim in a favorable time/space context. If truth be said, Brantingham (2013; see also Maguire, 2018) has theorized a more complex rationale of car thieves, who would follow sub-optimal choices guided by evolutionary patterns. But, first, other crimes, for instance, burglary, have long been discussed by positivist criminology as following rational choice patterns; and, second, the idea of police patrols as preventative means nonetheless builds on a paradigm of situational prevention (see also the “Crime Prediction and Prevention” page on IBM’s website: www.ibm.com/industries/government/public-safety/crime-prediction-prevention; accessed 15 November 2019).

6. Our emphases. Modesto Police Chief Galen Carroll and LAPD Foothill Division Captain Sean Malinowski,quoted in www.predpol.com/testimonials/(accessed 15 November 2019). See also the case for predictive policing made by Los Angeles chief of detectives in three arguments (Beck, 2009): first, predictive policing improves efficiency by reducing costs; second, evidence-based policing makes a neutral and technical issue of crime; and, third, by emulating forecasts used in distribution and retail operations, it allows public policy to be managed like a business.

7. Our emphasis. as cited in Lum and Isaac (2016, p. 18) and originally found on PredPol website (www.predpol.com/about/), but not available at the time of writing.

8. Aradau and Blanke suggest that, “through the featurization of time and space, PredPol has, for example, pre-emptively dis-activated accusations of discrimination” (Aradau & Blanke, 2017, p. 386). This argument is theoretically intriguing, but is nonetheless problematized by the fact that, as shown by the literature here reviewed, discrimination has been proven virtually every time predictive policing has been empirically scrutinized.

9. If truth be said, one study authored by the developers of PredPol found no evidence of racial bias in arrests made through predictive policing in Los Angeles (Brantingham et al., 2018).

10. The report is available at www.muckrock.com/foi/elgin-7770/foia-elgin-police-dept-predpol-documents-51858/#file-190432 (accessed 15 November 2019) and was made public by Lucy Parsons Labs, a Chicago-based project that focuses on the intersection between digital rights and “on-the-streets issues” (see https://lucyparsonslabs.com/).

11. See www.memphis.edu/fedex/SmartCities/(accessed 15 November 2019).

12. Blue CRUSH was heavily reliant on overtimes, as we heard from two interviewees (former MPD consultant and activist of Mid–South Peace and Justice Center) and a document made available by a local councilman confirmed (www.memphistn.gov/Portals/0/pdf_forms/bluecrushanalysis1.pdf; available through Internet Archive, https://archive.org/).

13. More recent UCR crime data are provided in different formats and may be not comparable.

14. Monthly number of violent crimes, violent crimes solved and sworn officers, January 2012 to June 2016. MIMEO (we can share the files upon request). We are grateful to MPD’s Office of Public Relations.

15. The following claims are based on our elaboration of UCR data, available since 1985, and, again, should be considered with a pinch of salt, like all reported crime data (for some series of data, see Tulumello, 2018b).

16. An organization with the mission to “engage, organize and mobilize communities to realize social justice through nonviolent action” since 1982. See https://midsouthpeace.org/about-us/history/(accessed 15 November 2019).

17. We do not use the term “community policing” because in Memphis, and in the U.S.A. more generally, the ideal type of the latter – which is characterized by co-decisional practices – has never been achieved, and the label has being used to refer to proximity practices such as aggressive order maintenance and nuisance abatement (Goetz & Mitchell, 2003; Tulumello, 2018b).

18. Mayor’s Weekly Digest Bulletin, 17 June 2016.

19. The Commercial Appeal did not follow up on the topic and our requests of information were never answered by the newspaper.

20. Ken Shackleford is listed in the page “About SkyCop” of the company website at the time of writing: www.skycopinc.com/index.php?p=about. In a document (without data) available on SkyCop website, Shackleford is quoted as MPD Lieutenant, explaining the decision to buy SkyCop equipment: www.skycopinc.com/assets/sitemedia/PDFs/SkyCopMemphis.pdf (accessed 15 November 2019). Our requests of interview were not answered by SkyCop.

21. In 2008, a firm of lawyers, representing the City of Memphis and MPD, sent a takedown notice to an online shop, for the selling of t-shirts and bumper stickers with Blue CRUSH logo. See www.citizen.org/documents/MemphisTakedownNoticetoZazzle.pdf (accessed 15 November 2019).

22. See www-03.ibm.com/press/us/en/pressrelease/32,169.wss#release (accessed 15 November 2019).

23. This is a topic that deserves further specific empirical investigation. For an early example, see Sanhu and Fussey’s ethnography (2020) in police agencies that make use of predictive policing in the UK, which has emphasized a tendency by police officers to be reluctant in following software’s instructions.

24. Quite interestingly, some of the most poignant critiques of predictive technologies have been made by former data scientists become novelists (cf. Lepore, 2020).

25. When not available online, articles from The Commercial Appeal have been downloaded from NewsBank online resource (subscription required).

Additional information

Funding

This work was supported by the Fundação para a Ciência e a Tecnologia [DL57/2016/CP1441/CT0007]; US.-Italy Fulbright Commission [Research Scholar Grant].