What I’ll cover

1. Assessing the impact of a drug law enforcement operation against a heroin & crack cocaine marketDARC seminar – 7th February 2018

2. What I’ll cover • The evidence base relating to DLE activities • Recent developments in this area • Aims of the study • Methods and analysis • Results • Implications

3. UK evidence base around drug law enforcement is limited (2008) “The UK invests remarkably little in independent evaluation of the impact of drug policies, especially enforcement. This needs redressing if policy makers are to be able to identify and introduce effective measures in the future”. Source: http://www.ukdpc.org.uk/publication/tackling-drug-markets-distribution-networks-uk/

4. A situation acknowledged as being ‘unacceptable’ (2010) “Given the public money spent on the strategy and the cost to society, we find it unacceptable that the Department has not carried out sufficient evaluation of the programme of measures in the strategy and does not know if the strategy is directly reducing the overall cost of drug-related crimes.” Source: https://publications.parliament.uk/pa/cm200910/cmselect/cmpubacc/456/456.pdf

5. Local emphasis on generating evidence & priority setting (2012) “We encourage Police [and Crime] Commissioners to ensure they are fully informed about the relative effectiveness of different forms of drug-related policing...and to carefully consider the issue of how police time is best prioritised between different kinds of drug-related offences, whether simple possession, acquisitive crime, supply or trafficking.” Source: https://publications.parliament.uk/pa/cm201213/cmselect/cmhaff/184/184.pdf

6. Public concerns around drugs (1) Source: https://www.ipsos.com/ipsos-mori/en-uk/public-perceptions-policing-england-and-wales-2017

7. Public concerns around drugs (2) Source: https://www.london.gov.uk/sites/default/files/140624_mopac_challenge_substance_misuse_final.pdf

8. Evaluation of the 2010 drug strategy (2017) “Estimated spend on enforcement in 2014/15 was £1.6 billion covering a range of cross‑government activities...[but] there is a lack of high quality evidence to assess the impact of drug enforcement activities...Due to the lack of robust impact data it is not possible to adequately assess the cost benefit of enforcement activity.” Source: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/628100/Drug_Strategy_Evaluation.PDF

9. Study aims From discussions initiated by a DCI in 2008, a research study developed (over time) which sought to answer two key questions: • How can the routine impact analysis of policing activities be enhanced using existing data sources and systems? - Beyond conventional focus of arrests made, assets seized and convictions secured. • What do these data sources/systems tell us about the impact of specialist policing activities against OCGs (using a drug law enforcement operation as a case study – Operation Example)?

10. Background • ASB associated with an established drug market had long been a concern for local residents and businesses. • Perceived inaction considered to have fostered a sense of ambivalence, undermined community engagement and confidence, and led to an under-reporting of incidents to the police. • Growing interest from local councillors that ‘something should be done’ (Symbolic policing: Coomber, Moyle & Mahoney, 2017). “It felt very matter of fact that everyone knew [about the] dealing. You know, it was an issue, but people glossed over that, and it was the other issues that were around it that were really impacting on them…Like there’s a school next door but nobody really notices…there was open drug-taking and people drinking…I think it was just disrespect to the children really, and the families”.

11. Background – Op Example (1) • Prior to the ‘resolution phase’ of the operation, for evidential and intelligence purposes, 14 test purchases were made by plain-clothes officers from four separate sellers. • 1.7 grams of heroin (N=3 purchases) and 3.6 grams of crack cocaine (N=12 purchases) (one purchase was of heroin and crack in combination). • Average purity of these purchases was 14 per cent for heroin and 35 per cent for crack. • Purity rates were consistent with police seizures nationally during January to March 2011, which averaged 16 per cent for heroin and 31 per cent for crack cocaine (Coleman, 2011: 30).

12. Background – Op Example (2) • Targeted individuals operating from a commercial (betting shop) and neighbouring residential properties implicated in an established market for Class A drugs (heroin and crack cocaine), with warrants executed from mid-June 2011. • Resolution phase involved 90 officers & 13 warrants executed (£87.5k). • Resulted in arrests for conspiracy (4), supply (5) and being concerned in the supply of Class A drugs (4) offences. • Charges brought against nine arrestees (no further action taken against remaining four suspects), all but one entering a guilty plea. • Sentencing outcomes were available for six of the nine defendants. Three received custodial sentences totalling seven years. (Perceived leniency of sentences raised as a concern re: community confidence.)

13. Background – Op Example (3) “We didn’t yell to the community about it, because they all would have gone, ‘oh, that was a complete waste of time’, because that’s what I thought!”. “They [the public] think ‘why bother ringing the Police again’, or telling them what they’re doing”.

14. Methods The three main approaches involved: • analysis of administrative data covering a 24-month period (12 months pre and post ‘strike’) (CEDAR - Crime Evaluation Data Analysis and Recording; NDTMS - National Drug Treatment Monitoring System, N=932); • semi-structured interviews with key stakeholders, including ‘non-police’ and members of the local Neighbourhood Policing Team (NPT) (N=8); and • observations (of tasking and co-ordination meetings, pre and post-operation de-briefings, and observing warrants being executed).

15. Analysis – CEDAR data Three levels of analyses, using a comparative, pre/post design: (i) within a pre-defined radius of 300 metres of the betting shop which served as the focus for the ‘strike’ (the target area); (ii) the Local Policing Area (LPA) (comparison 1); and (iii) across the wider Police Force Area (PFA) (comparison 2). Impact on four outcomes considered: (i) all CFS incidents (N=1,032,361); (ii) all recorded crimes (N=311,555); (iii) all recorded ‘acquisitive’ offences (N=182,905); and (iv) all CFS incidents relating to anti-social behaviour (ASB) (N=123,728).

16. Analysis (2) • Descriptive statistics. • Odds ratios (OR) calculated to measure proportional change observed in the LPA or PFA, relative to target area (Welsh & Farrington, 2009). • An OR greater than 1.00 indicates a desirable effect from an intervention, and an OR less than 1.00 indicates an undesirable effect. • Analysis replicated and supplemented using a weighted displacement quotient (WDQ) (Bowers & Johnson, 2003). • WDQ can detect any change due to displacement from the target area to the wider LPA. • All interview data were recorded, fully transcribed and manually coded to identify key themes and issues.

17. Results (1) There was no evidence to indicate that: • Op Example reduced CFS incidents (both general and ASB-related) or recorded crimes within the target area over a 12-month follow-up period, relative to the scale of reduction observed across the wider LPA and PFA. • Most persuasive evidence of a positive benefit related to falls in levels of acquisitive crime (down 25% vs. 9% & 15%). • Reductions unlikely to be accounted for by an increase in numbers presenting for treatment at local drug services (which fell 25 per cent in the 12 months after the operation). • Anecdotally, some evidence of ASB displacement elsewhere within the wider LPA.

18. Results (2)

19. Results (3)

20. Results (4) – Other activity Response to police and council requests to local residents to report such incidents in order to gather evidence in support of an application for a dispersal order: “We did a big campaign in the Spring of last year telling people to start reporting things to the Police, because we were trying to get the evidence for the dispersal order, and obviously unbeknown to us they were working on the other thing [Operation Example]. So building up that kind of awareness, ‘here’s the number, here’s the number, here’s the number’, and then when the big operation happened I think that gave people the confidence that something was being done”.

21. Results (5)

22. Results (6) – What went well? Aspects of the operation that were considered to have gone particularly well related to: • its planning (e.g. dedicated team at an off-site location); • the involvement of key partner agencies (early and regular engagement with the Crown Prosecution Service); and • the degree of community reassurance that followed: • leafleting the local community about the operation; • deploying targeted/high visibility patrols; and • introducing a local business watch scheme.

23. Results (7) – Lessons learned The two most significant problems identified related to the: • planning process (other operational demands in parallel reduced the planning time available for Op Example); and • involvement of partnership agencies (delays with the local council installing CCTV cameras post-operation). • Anecdotally, no adverse impact on the availability of Class A drugs, or their quality. • Reports of some unintended consequences: • previously peripheral players within the market began to occupy more prominent roles; • separation of two local markets for Class A (heroin and cocaine) and B (cannabis) drugs, which had previously existed, was thought to be less obvious following enforcement activity.

24. Results (8) – Lessons learned “They had control over that. Since we’ve taken them, the bigger players, out we’ve now got all the smaller peripheral people in who ship the drugs out. I think that’s the users who have been brought in to do that, which is why we’re getting more people dealing throughout the area now unfortunately”. “The only difference is that now both heroin, crack cocaine, and cannabis, so Class A and B, are all freely available for people all round the community. Whereas in the past it was more, it seemed to be more a case of if you wanted Class A they knew they’d go to [the betting shop], if you wanted cannabis then you’d go to [neighbouring market location]”.

25. Implications (1) • Scope to enhance impact analysis using existing data sources and systems – but takes time i.e. 12 months+ • Review needed of how different sources could be better integrated and made more accessible: • i.e. drug samples submitted for forensic analysis following test purchases or routine seizures to monitor changes in drug prices and/or purity • Value of integrating qualitative insights, including those from individuals/groups outside the organisation-where possible. • Focus on unintended as well as intended consequences (‘coal face’ NPT input seemed especially useful for this). • Requires analysts to be more proactive however; not only engaging police personnel across a range of departments, but also other stakeholders in external agencies.

26. Implications (2) • Existing best practice knowledge resides with a small number of officers who have direct experience of running these types of operation. • No mechanism exists whereby learning can be disseminated Force-wide, or appropriately deposited in order to inform subsequent enforcement activities. • A need for this PFA to enhance organisational memory of best practice around specialist policing activities. • Consider establishing a function tasked with generating a body of local ‘what works’ knowledge, drawing upon a range of internal and external sources (e.g. knowledge gleaned from structured de-briefs; work of CoP etc).

27. Any questions? McSweeney, T. and Gyateng, T. (2013) Improving the impact analysis of specialist policing activities: A drug law enforcement case study. London: ICPR. http://www.icpr.org.uk/media/35873/Op_Example_summary.pdf