Multi-agency case management: evidence and orthodoxy

Police Effectiveness in a Changing World 1 Multi-agency case management

Policing is increasingly concerned with the risksfaced by the vulnerable, the threats posed by thedangerous and reducing the harm caused to theformer by the latter. In response multi-agency casemanagement arrangements have proliferated andcome to represent a ‘new orthodoxy’ for the policeand their partners. However, evidence about theeffectiveness of such schemes is in short supply.This paper describes an attempt to reducerecurrent violent crime in Slough (England) throughthe application of multi-agency case managementtechniques, as part of the Police Foundation’sPolice Effectiveness in a Changing World project.Despite being modest in scale, the programme wassubject to a thorough impact and processevaluation, which showed that although it securedstrong local support, improved multi-agencyworking, increased information sharing andcatalysed ‘sensible’ activity; it did not achieve ameasurable reduction in violence. This paradoxraises a number of important questions: how canmulti-agency process improvements be translatedinto better outputs and outcomes? Can taskingwork better in a partnership setting? What canrealistically be expected from these schemes? Mostcritically it highlights the urgent need for moreevidence about what works in multi-agency casemanagement and in what circumstances multi-agency case management works.

Between 2011 and 2015 the Police Foundationresearch team worked with the police and theircommunity safety partners in two English towns– Luton and Slough – to identify persistent localcrime problems, improve the way in whichthese were understood, develop and implementappropriate interventions, and assess both theoutcomes of these and the challenges of doingso. In the process it was hoped that valuablelessons might be learned about the routes to,enablers of and dependencies for effectivepolicing under current conditions and in thecontext of change.

Later this year we will publish full reportssetting out the findings from each of the PoliceEffectiveness in a Changing World projectsites. In advance, the research team hasproduced a series of thematic papers bringingtogether key project learning from both townsand addressing the main issues that theseraise for policing in the second decade of thetwenty first century.

This is the third paper in this series dealingspecifically with the challenge of effectivepartnership working at a time when all agenciesare confronted with rapid social change, newpriorities and significant internal reform.

Police Effectiveness in a Changing World PAPER 3

Multi-agency case management:evidence and orthodoxyAndy Higgins, Gavin Hales and John Chapman – The Police Foundation, October 2016

Embargoed for release at 00:01 on Wednesday 14th September 2016


Policing in partnership: finallymainstream, new manifestations

It has become established wisdom that neither thepolice nor any other agency can effectively dealwith crime and public safety issues alone. This isbecause crime problems are often ‘wicked’ ones 1

that is, they have complex, social andinterconnected causes and thus require solutionsthat address these from different perspectives,using a range of tools, skills, powers and expertisein a ‘joined-up’ way. To this end, partnershipworking has become a progressively moreprominent feature of policing over the last halfcentury.2 Milestones, in the UK context, haveincluded Home Office Circulars in the 1960s, 70sand 80s, The Morgan Report on safer communities(1991), The Crime and Disorder Act (1998), whichestablished the statutory requirement for police andlocal authorities to work together in Crime andDisorder Reduction Partnerships (CDRPs) 3, and thenational roll-out of Neighbourhood Policing acrossEngland and Wales in 2008.

Although the challenge of assessing the contributionof such a broad and varied working approach issubstantial, enough evidence has amassed tosupport the conclusion that ‘on balance…theprinciple of applying partnership working as acomponent of initiatives to tackle crime and disorderproblems is effective’ 4. Working with other agenciesis also fundamental to delivering innovative and‘diverse’ 5 responses to crime including within aproblem-oriented approach 6, and collaboration is

therefore a key component in the concept ofinformed proactivity developed during this project todescribe the mode of working best suited todelivering effective policing 7.

The police have not always taken to partnershipworking with enthusiasm 8, however there isevidence of a sea-change in attitudes, with themainstream exposure of officers to multi-agencypractice, including through neighbourhood policing,leading to ‘pragmatic’ acceptance of its value andpalatability as a way of working 9. This is atransformation we recognise from our time in Lutonand Slough. Although in the early part of the projectwe found some evidence of partnership marginalisedto a specialism – a superintendent and small staff (ofthe kind described elsewhere as a ‘trumpets andpamphlets department’ 10 ) that ‘did partnership’ soothers could ‘get on with police business’ – this wasan anachronism rather than the norm.

Overall, and in particular during the implementation

phase of the project in 2014/15, we encountered

police officers in both towns who, along with their

colleagues in other agencies, genuinely embraced

collaboration and were engaged in numerous

examples of well-established multi-agency practice.

In small scale surveys of community safety

practitioners in each town, more than 80 per cent

thought local agencies worked (very or fairly) well

together to tackle crime, disorder and related social

problems, and more than half thought the way

agencies worked together had improved in recent

years – with most of the rest suggesting it had

stayed the same rather than deteriorated.

Based on the experience of this project it is clear thatmulti-agency working is now close to the mainstreamof local policing, however it is also clear that, in the

1 Rittel and Webber (1973).2 Crawford and Cunningham (2015), Berry et al. (2011).3 Which later became Community Safety Partnerships (CSPs).4 Berry et al. (2011).5 Based on a synthesis of the research literature, Weisburd and Eck (2004)concluded that there was little evidence to demonstrate the efficacy of‘standard model’ policing activities (general patrol, rapid response, reactiveinvestigation etc.). They argued that more promise can be found in innovationsthat extend from the traditional police activity-set along two axes; the diversity ofthe approach (for example working with new partners, as opposed to a relyingon narrow law enforcement tactics) and the degree of focus (for example onhotspots). Most promising of all were initiatives that combined diverseapproaches and were highly focused (for example problem-oriented policing).

6 Problem oriented policing developed from a critique of conventional policeactivities first made by Herman Goldstein (1979). Goldstein challenged theprevailing view of police work as a series of discrete incident responses andinstead advocated refocusing on the ‘problems’ that connected them. Theapproach emphasises collecting and analysing information from a range ofsources to improve understanding of problems and their causes, and enlistingthe support of other agencies and individuals to develop and deliver ‘upstream’solutions. The problem-orientated approach was field-tested by Eck andSpelman (1987) and codified into a four stage problem-solving process of

Scanning, Analysis, Response and Assessment (or ‘SARA’) which has beenshown to work at least ‘modestly’ well (Weisburd et al., 2008). The SARAmodel provided the template for the crime reduction work carried out in Lutonand Slough as part of this project.

7 Higgins and Hales (2016).8 Bullock et al. (2006).9 O’Neill and McCarthy (2014).10 Crawford and Cunningham (2015).

Collaboration is a key componentof informed proactivity.



context of public sector austerity, and the localorganisational restructures that have followed, it hasbecome more difficult to sustain – as the followingquotations from practitioners in both towns explain:

“When I first started all the agencies had reallyrobust working relationships, but now it’sdifficult; it’s much harder… Funding’s cut,staffing’s cut, there are less staff doing thesame amount of work. A lot of agencies can’tget to the meetings…so you don’t have thatnetworking part as well”.

“It is getting harder now because there aren’tso many resources around, and I suppose thedifferent areas of the partnership are focusedon what they need to focus on and their corebusiness”.

“I think the biggest problem is there’s too manychanges in other agencies all the time, there’snew faces all the time, so you’re constantlygetting to know new people”.

On the whole however (as reflected in the nextquotation), we encountered agencies andpractitioners committed to facing these challengestogether.

“I think in austerity, partnership is probably of fargreater importance, because rather than oneorganisation trying to solve everything on theirown and throwing everything at it, eachorganisation can do their little bit – that willhopefully make for a more efficient [response]in the first place”.

Partnership was also a defining characteristic of theproject methodology, which involved PoliceFoundation researchers accompanying localpractitioners through a long-term revolution of theproblem oriented policing ‘SARA’ cycle. Thescanning processes, which selected the crimeproblems for attention in each town, were carriedout following extensive local consultation; analysiswas conducted on datasets from a range ofagencies, culminating in a set of workshops thatbrought practitioners together to develop thefindings into options for responses, some of whichwere then developed and delivered by multi-agencyworking groups 11.

In Luton this resulted in Police Community SupportOfficers (PCSOs) working with colleagues from theFire and Rescue Service and local authorityenvironmental health teams to conduct ‘street-survey’ visits to burglary hotspots, identify vulnerableproperties and engage with residents. It also led tothe local Home Improvement Agency and localauthority housing department agreeing to takepolice referrals for home security improvement work,and to a council-employed community developmentofficer working with police officers to start aneighbourhood improvement group in a transientneighbourhood with high burglary rates.

In Slough, building on analysis that characterisedviolence in the town as diffuse and varied (ratherthan concentrated and easily framed as a coherent‘problem’), but which also showed that someindividuals were involved in violent incidents time andagain, the Violence Multi-Agency Panel (henceforthVMAP) was formed. VMAP was a ‘virtual team’comprising practitioners from local neighbourhoodpolicing teams, social services (and other localauthority safeguarding functions), housing, youngpeople’s services, probation, drugs and alcoholtreatment services, mental health services, domesticabuse support and several other local third-sectorsupport organisations. The group came together ona fortnightly basis to take a systematic approach todevising and delivering preventative interventions inthe cases of individuals recurrently coming to policeattention in relation to violence.

This paper focuses on the second of theseinitiatives as an example of a form of multi-agencyworking that is becoming increasingly important to


11 Summarised in Higgins and Hales (2016).

Multi-agency working is now closeto the mainstream of local policing,

however it is clear that it hasbecome more difficult to sustain.

Multi-agency case managementis becoming increasingly importantto contemporary policing and has

begun to consume a greaterproportion of police resources.


contemporary policing and, as issues of harm andvulnerability increasingly move centre-stage, hasbegun to consume a greater proportion of policeresources 12. In addition to close collaborationbetween agencies, this mode of working ischaracterised by a case-based approach whichaddresses crime, public protection and/orsafeguarding issues at the level of priorityindividuals (or sometimes couples or families)with personal case-histories and ‘risky’ futures.It stands apart from approaches which take eitherincidents or ‘problems’ as the principle unit of focus– and is one which represents relatively newterritory for the police compared to their colleaguesin probation or social services, for whomcase-work is core business.

Multi-agency case management

Multi-agency case management has become adistinct mode of working, but it represents theconvergence of two streams of activity. The first isprincipally concerned with managing offenders andreducing reoffending and incudes practices such asMulti Agency Public Protection Arrangements(MAPPA), through which criminal justice agenciesdischarge their statutory duty to manage violent andsex offenders. It also includes schemes such as theProlific and Priority Offender (PPO) initiative –introduced in 2004 to coordinate enforcement andrehabilitation for the most prolific (usually acquisitive)offenders – which was later subsumed within abroader approach to Integrated OffenderManagement (IOM) 13.

The second stream also involves the police andothers in coordinated case-work, but focuseson keeping vulnerable individuals safe, includingthrough Multi Agency Risk AssessmentConferences (MARACs), which coordinate theresponse to the highest risk victims of domesticviolence, and Multi Agency Safeguarding Hubs(MASH), which deal with cases of children andother vulnerable people at risk of harm and abuse.

The caseloads of many of these schemes haveexpanded in recent years 14 but in addition, the

palate of practices and techniques which theyshare (risk and needs assessment, case-conferencing, information sharing, key-workingetc.), has increasingly becoming familiar and ‘best’practice for the police and others, and has foundnew applications in national schemes such as thegovernment’s Troubled Families programme andin myriad localiterations (of whichVMAP was justone example).

It is important tonote, however,that despite therapid expansion ofmulti-agency casemanagement,there is limited evidence of its effectiveness as anapproach. Reoffending analyses have indicatedthat the introduction of MAPPA coincided with afour percentage point reduction in reoffending byeligible offenders 15 and that the PPO cohortcommitted fewer offences after the introduction ofthe scheme 16, however neither of these studieswas able to generate a sound counterfactual orstrong evidence of a causal link. IOM has beenrecognised as ‘promising’ 17 yet quantitativeevaluations are scarce and show either negativeor highly tentative outcomes 18. Evaluations ofmulti-agency case management in thesafeguarding sphere typically report very positivequalitative appraisals from those involved, butinclude little outcome data – and none (that weare aware of) make use of control or comparisongroups 19. Early findings from the Troubled Familiesprogramme appeared promising 20 but weregreeted with some scepticism, 21 and it has been

12 College of Policing (2015).13 Home Office and Ministry of Justice (2010).14 College of Policing (2015).


Despite the rapidexpansion of

multi-agency casemanagement,there is limitedevidence of

its effectivenessas an approach.

15 Peck (2011), Bryant et al. (2015).16 Dawson and Cuppleditch (2007).17 HM Inspectorate of Probation and HM Inspectorate of Constabulary (2014).18 Dawson et al. (2011), Williams and Barak (2012).19 Steel et al. (2011), Crockett et al. (2013).20 DCLG (2014).21 Crossley (2015).

Evaluations of multi-agency casemanagement report very positivequalitative appraisals but include

little outcome data.


reported that a recent (as yet unpublished)evaluation shows that it has failed to live up toambitious expectations 22.

More broadly, of the most robust studies onmulti-agency practice assembled for a RapidEvidence Assessment, none that reported asignificant positive impact could be described asa case-management approach 23 and the ‘matrix’of crime reduction studies (compiled by Lum andcolleagues 24) indicates that approaches targetedat individuals tend to be less effective than thosefocused on small places. It is in the context ofthis shallow and equivocal evidence-base thatthe findings of the VMAP evaluation should givepause for reflection.

The Slough Violence Multi-Agency Partnership (VMAP)

The Slough VMAP programme was an attempt tocut violent crime by focusing attention on thoseindividuals who were involved in violence time andagain. In the priority wards on which it focused,analysis showed that one in four victims of violence,and one in three violent offenders, had previouslybeen involved in violent crime within the precedingyear. It also showed that the same individualsfrequently came to notice in different roles (as bothvictims and offenders) and in differentcircumstances (both domestic abuse and otherviolent incidents), and that recurrent violence wasstrongly predictive of further involvement in violence.With few situational options available to tackleviolent crime, and acknowledging that – other thanfor the minority who qualified for processes such asMARAC or MAPPA – recurrence went largelyunrecognised, bespoke case-level interventions toaddress the risk of future offending and/orvictimisation held the promise of sustainedreductions in crime, harm and service demand.

VMAP attempted to systematise multi-agencyattention by using police recorded crime data. On a

fortnightly basis, an analysis of recent crime reportsidentified those individuals who had come to notice(as violence victims and offenders) within four pilotwards, for at least the second time in the last year.The resulting list was shared among participatingagencies (under the terms of a data sharingagreement) and subjected to research across multi-agency databases. Once every two weeks theVMAP panel convened25, shared case informationand expertise, and carried out a case-basedproblem solving process that sought to formhypotheses about the factors contributing to on-going risk (of recurrence) in individual cases – andthen task appropriate intervention actions, based onthe theories generated.

During the pilot year, the panel met on 26occasions, considered nearly 300 cases (althoughsome of these were filtered out through a ‘triage’process) and generated more than 650 actions.Throughout the process the Police Foundationteam provided programme support, acted asobservers and conducted process and impactevaluations – feeding back emerging findings toshape delivery, as part of an on-going actionresearch approach.

Process evaluation:the joy of partnership

In terms of multi-agency working, VMAP can beconsidered a resounding implementation success.From the outset, it benefited from strong practitionerbuy-in, reflected in sustained attendance figuresand endorsement from a broad set of participatingagencies. Although not without its teethingproblems, including a caseload that ballooned tooquickly and initial difficulties in getting relevant caseinformation ‘in the room’ for meetings, the hard workof the coordinating team and fine-tuning throughaction research, resulted in a process that maturedrapidly, as reflected in improved survey ratings fromthe practitioner group between the third and eighthmonth of the pilot (see Figure 1, overleaf).

22 Cook (2016).23 Berry et al. (2011) – in one study (Scott et al., 2002) a positive impact on

recidivism of violent offenders was observed, however the approach wasprogrammatic (all individuals were the subject of a standard treatment) ratherthan using a tailored case management approach; other positive impactsrelated to initiatives that were not targeted at individual cases.

24 Lum et al. (2010).

25 VMAP panel meetings were attended by an average of 14 representativesfrom relevant police and local authority functions (community safety, socialservices, safeguarding, young people’s services, housing etc.), probation,community mental health services, drugs and alcohol treatment services,domestic abuse support and local third sector organisations. Meetings wereco-chaired by a police chief inspector and a local authority community safetymanager with analytic and coordination support provided by the police.



As previously noted, VMAP was implemented into alargely hospitable environment; many of thoseinvolved had experience of partnership working andwere convinced of its benefits (and a minority whowere not quickly ‘self-selected’ out withoutdetriment) – however it was also clear that theinitiative became valued for the additional depth andbreadth of multi-agency collaboration it catalysed,as the quotations below illustrate.

“I always ‘big up’ VMAP…if you come to VMAPyou’ll have these people basically on thedoorstep, you can ask them anything aboutanything…you build that relationship with themthrough VMAP...they’re not only there to dealwith VMAP cases, because they can help youwith other cases as well.”

“Networking; everyone has got to really knoweach other…and that really helps as well forthe support we are giving as a whole.”

“It has done nothing but strengthen our linkswith key partners.”

Practitioners also reported benefits for their own

work in terms of bringing new cases to attention,

avoiding duplication, improving case knowledge,

personalising referral processes and, perhaps most

importantly, for presenting a consistent and unified

response to service users.

“They [service users] are not happy to tell their

story to everyone, time and time again, but if it

can be approached as one. That is what a lot

of our clients will say, that we don’t want to

keep telling everyone what’s happening. So,

we should be coming together.”

The extent to which VMAP acted as an enabling

mechanism for neighbourhood policing was

particularly encouraging, with the contrast between

the ‘usual’ and new approaches summed up neatly

by one neighbourhood officer.

“Well, it [violent crime] has been addressed in

the past by reacting to what’s gone on and I’ll

The VMAP initiative became valuedfor the depth and breadth of multi-agency collaboration it catalysed.

Figure 1: VMAP participant practitioners’ views on process and working arrangements(surveys conducted October 2014 (n=20) and March 2015 (n=17))


Discussing cases at VMAP leadsto sensible actions and activity thatwould otherwise not have happened

In VMAP meetings we usually haveenough information to make good

decisions about cases

VMAP meetings usuallyrun smoothly

October 2014 March 2015 October 2014 March 2015 October 2014 March 2015

35% 56%

10%

25%

35% 55%

0%

6%

35% 38%

10% 44%

■ Strongly Agree ■ Agree

Practitioners reported benefits fortheir own work – most importantlypresenting a consistent and unified

response to service users.


go to it one day, she [a colleague] will go to itthe next, he will go to it the next, and there’sabsolutely no continuity in how to deal with thatparticular person…now we’re talking abouthow all of us around that table can work todeal with that person.”

Local officers felt they benefited from improvedflows of information, both out of police systems, andfrom colleagues in other agencies to ‘the front line’;a benefit that was clearly reciprocal.

“I don’t think we’d [previously] have got all theresearch which is what underpins everything. Icertainly wouldn’t have had anybody to be ableto put it into some sort of format and, youknow, to look at … actually, this is the issue.So there’s been a lot of benefits to it.”(Neighbourhood police officer)

“[Other VMAP practitioners] will give me thepersonalities of the people we’re talking aboutbecause, they’ll know a little bit about them thatI don’t know, and I think that’s really important.So that young lad…[who] assaulted somebody,one of the youth workers was telling me whathe’s like and what he enjoys and stuff. If I goand visit him I know that, don’t I?”(Neighbourhood police officer)

“What I find quite useful, is information thatwe’re given from the neighbourhood policeofficers, on that day to day basis how thatperson’s behaving in the community, becausethat service user’s not necessarily going to giveus that information and that’s something wecan [use to] challenge that individual.”(Non-police offender manager)

As these comments illustrate, at its best VMAPpromoted the kind of informed proactivity which theevidence base tells us is most likely to be effective26.

It is also pertinent to note that VMAP appeared toperform well against the checklist of enablingmechanisms associated with effective partnershipworking 27. It benefited from strong leadership,both operationally and from the Local Policing Area(LPA) commander and local authority chief executivewho offered reputational sponsorship and the

occasional injection of ‘hard power’. The force chiefconstable also took a personal stake in the projectby funding a part-time coordinator post. VMAPmanaged to surmount traditional barriers in relationto data-sharing. Through shared training and withthe right agreements in place, participants clearlybecame mindful of the dangers and drawbacks ofnot sharing information.

“Everyone’s got so much information here, andwhen we look at Serious Case Reviews thatare going on all around the country, it’s like,‘oh, yes, they knew that, they knew that’… andwe sit together, and it’s like we’ve got a wholepicture of this person now... we didn’t knowthat you were involved, and you didn’t know wewere involved, and, actually, that’s not great, sowe need to work better at that.”

“That’s where it [VMAP] sort of comes in,because sometimes people can be working insilos, the whole idea is you open it up, soeveryone is aware of the situation, and if theycan add additional benefit and support to thoseindividuals, then that’s the place to do it.”

Although co-location was not a feasible option,partners clearly recognised improvements in inter-agency communication – when surveyed towardsthe end of the pilot 80 per cent of participants saidthat they were now more likely to get in touch withcolleagues in other agencies and that VMAP hadgiven them a better understanding of what otherorganisations do 28. Programme structures weresuitably flexible (for example equivalent roles werefilled by a police inspector in one area and aconstable in another, as these were felt to be themost appropriate individuals for the role), the PoliceFoundation team provided a research and feedbackmechanism and, although there were some initialgaps in terms of agency coverage, these wereaddressed, leading to comprehensive representationand a genuine team ethos. VMAP also benefited

VMAP promoted the kind of informedproactivity which the evidence basetells us is most likely to be effective.

26 Higgins and Hales (2016).27 Berry et al. (2011), Crawford and Cunningham (2015). 28 Surveyed in July 2015, n=18.



from extensive experience, including from a highlyrespected police chief inspector who chaired mostmeetings as well as established and experiencedpersonnel from a number of agencies.

As a result, VMAP came to be highly valued bymany of those involved in its delivery; it was seen asexemplifying the way agencies should work togethertowards common, long-term goals, and ofsubstantial benefit in addressing the sharedchallenges presented by complex and concerningindividual cases within the town. There was strongconsensus that these efforts should continuebeyond the end of the pilot.

“I think it’s a fantastic tool that can be used togreat benefit in the long run.”

“Surely if we look at all the money that is beingsaved, if we are being preventative. You know,we’re trying to prevent all the police call-outs,and getting to serious case reviews, and all ofthat, which we’re not wanting. It will save afortune, so it is much better to be preventative.”

Impact assessment: the realityof evidence-based evaluation

Set against this positive experience, the evaluationfinding that VMAP had no identifiable impact on

recurrent violent crime was a difficult one forpractitioners to reconcile.

Careful analysis of police-recorded crime datashowed that the cohort of 298 individuals who metthe criteria for VMAP were involved in 132 furtherviolent crimes within the town during the pilotperiod after VMAP had considered their cases andset in motion intervention activities; that is they wereinvolved in at least one ‘subsequent occurrence’ 29.Comparisons against equivalent groups who wouldhave been referred into VMAP had it beenoperational in previous years, or in the parts of

29 The subsequent occurrence rate was calculated as follows: 1) A count wasmade of all violent offences, during the pilot year, that involved one or moreindividuals whose involvement in earlier violent offences would have qualifiedthem for referral into the VMAP process (ie these are offences that involvedindividuals who had been involved in at least two previous violent incidents,one within the pilot year and one within the year prior to that). As the referralcriteria is ‘objectively’ derived from crime data, the count can also be appliedand to alternative time periods and geographic areas to provide comparisons.2) The count was then divided by the sum of the total number of daysbetween all individuals in each cohort first meeting the referral criteria for VMAPand the end of the pilot year. This approach took into account both differencesin the number of eligible cases between periods/locations and the distributionof these throughout the year. In the chart the rate of subsequent occurrence isshown per 1,000 VMAP ‘protected’ days. For a full methodology seeChapman, Higgins and Hales (forthcoming).


Figure 2: Subsequent occurrence rate for VMAP and ‘rest of Slough’ groups – 2006/07 to 2014/15

The evaluation finding that VMAP hadno identifiable impact on recurrentviolent crime was a difficult one for

practitioners to reconcile.

2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/152006/07

Year (mid-July to mid-July)

4

3

2

1

0

VMAPintervention

VMAP area

VMAP area – mean

VMAP area – threshold for statistically significant reduction (at 95% confidence level)

Rest of Slough


Slough not covered by the scheme (who had agenerally lower tendency for recurrence), gave noreason to suggest that this rate would have beenany different had VMAP not been in place (seeFigure 2 on previous page).

Further analysis for various sub-groups of thecaseload (including separating out those referred inas victims and offenders, and those who hadreceived most attention), as well as of the‘seriousness’ of subsequent incidents and of awider indicator of demand on police resources, allsupported the conclusion of no programme effect.

The paradox of evidenceand orthodoxy

How can we explain this failure to achieve effectiveimpact in the face of ‘text-book’ partnershipcollaboration and resounding local endorsement?One possibility of course is that, with its focus ontackling ‘root causes’, the benefits of the VMAPapproach did not have time to filter through intocrime (and demand) outcomes. This was asentiment frequently offered by practitioners:

“We have had a huge success in terms ofbringing VMAP together, bringing partnerstogether, sharing information and establishing Isuppose what some people would see is quitea powerful partners’ meeting. But in terms ofreducing violence, I think probably the projectshould’ve had another 12 months to try toachieve that.”

“VMAP will only impact the community if it’sgiven the longevity to be able to impact thecommunity. Today it won’t change anything, itjust won’t because it hasn’t been working longenough or running long enough to be able todo that. My only fear is that all this hard workgets thrown out the window one day withoutactually really understanding why and thatwould be it. Because VMAP is a process thatyou’ll know in five years’ time what a fantasticthing it was to start.”

There are certainly challenges for evidence-basedevaluation in assessing the maturity of initiatives,however in this instance we would draw attentioninstead to the level of active intervention generated

through the VMAP process. Without seeking todismiss the value of the more ‘intangible’ benefitsalready suggested (stronger inter-agencyrelationships, improved information flows etc.)analysis of VMAP’s formal outputs – the 657 actionsset during meetings – gives some cause to questionits potential for potency. Almost two thirds of theseactions can be categorised as ‘information actions’that is, they were either tasks for VMAP members tomonitor and report back on the outcome of on-goingprocesses (investigations, court cases, bail hearingsetc. – 38 per cent), or to conduct further research(23 per cent), while a further eight per cent ofactions were ‘administrative’.

Only 18 per cent of the actions set during VMAPrequired a direct attempt by practitioners to engagewith an individual victim or offender (this equates to97 actions relating to 57 members of the caseload– about 20 per cent). A further five per cent ofactions related to engagement with (or other actionpertaining to) linked parties (often children), whileother less frequent activities included initiating moredetailed case conferences outside of the VMAPforum, formal referral to other agencies, and‘flagging’ on various systems (eg for a particularcourse of action to be enacted in event of arrest). Inthe light of this it seems reasonable to hypothesisethat VMAP was not effective quite simply becausenot enough of substance happened as a result.

It is worth stating here that although some concernsabout outputs were raised (gently) by the researchteam as part of the action research process, thoseinvolved never expressed reservations atopportunities missed or under-performance. Insurveys most thought VMAP stood a good chanceof reducing violence, more than 70 per cent thoughtVMAP would improve service in individual cases and,(after initial teething problems) 80 per cent agreedthat VMAP generated sensible activity that wouldotherwise not have happened. There was a clearconsensus among participants that VMAP was doingeverything it could, and that that was worth doing.

Four observations are pertinent to understandingwhy both outputs and outcomes were not moreconspicuous.

First, it proved extremely difficult to achievemeaningful ‘problem-solving’ in the cases



identified through the VMAP process. Therewere certainly ‘success stories’; in one case, afamily with feuding teenage sons were helped tofind more suitable accommodation. In another, acouple with a long history of domestic abuse(forced back into co-habitation by deterioratinghealth) were supported to resume independentliving. In a third, a homeless drug and alcoholdependent man, making progress with histreatment, was provided with help intoaccommodation and a more stable lifestyle. It is fairto say that these outcomes would not havehappened without VMAP. In many other cases,however, practicable intervention avenues wereeither not apparent and/or subjects refused toengage with the services on offer. More was notdone because it was not obvious what more couldbe done and, as problematic and seeminglyintractable cases began to re-present, practitioners’frustrations began to show:

“That’s the thing that probably galls a lot ofpeople. We’re sitting here, we’re giving up ourvaluable time, and we’re having a ten minuteconversation about someone that’s going roundand round and round the system, and we, kindof, know that they will continue to do that.”

One notable feature of VMAP was the extent towhich examining cases of recurrent violenceopened the door to a host of complex long-termneeds (including drug and alcohol dependency,homelessness and mental health), often in thecontext of chaotic lifestyles and destructive but co-dependent relationships. Our secondobservation is that the level of additionalservice resource required to meaningfullyattempt to ‘problem-solve’ these issues, wasjust not available in Slough (and, we suspect,would also not be available in many other places).As one practitioner explained, even getting to thepoint of understanding the drivers of harmful andrisky behaviour in complex cases could requireintensive investment of resources.

“Unless somebody is willing to talk about whythey need to drink a litre of vodka to getthemselves through the day [we will not makeprogress]… it needs [support from] somebodythat can find the key…but I guess it’s finding theright person. And of course, as services arebeing withdrawn or are being cut, that wholeability for people to spend the time and the effortand the energy…is starting to be reduced.”

Beyond a part time coordinator post funded by thechief constable, some funded training and theresearch and project development capacityprovided by the Police Foundation, VMAP receivedno additional resource. It was delivered within theexisting capacity of partner agencies and had no‘outreach function’ or dedicated key-workers. Itrequired busy practitioners, from stretched services,to find time in their schedules to attend meetings,carry out database checks and complete whateveradditional intervention activity they could. In themajority of cases, finding capacity to ‘own’ casesand attempt concerted engagement, provedunrealistic. It is also pertinent to note that severalparticipating commissioned and third-sectorservices declared themselves ‘at capacity’ duringthe period, while other services lamented tightenedservice frameworks that constrained the contributionthey were able to offer. More was not taskedbecause it was generally understood that only alittle more could be done.

The third observation relates to the style oftasking and accountability monitoringadopted within the VMAP process. A keyfactor in generating the positive momentum andstrong practitioner buy-in that amassed behindVMAP was the highly consensual and inclusivespirit in which it was managed; this owed much tothe tone set by the chair and appeared to be verymuch ‘the way things are done’ in Slough. Whilethis approach generated willing volunteering attimes, it is also possible that it served (incombination with the factors previously mentioned)to hamper decisive tasking and accountability (the


In many cases practicable interventionavenues were not apparent

and/or subjects refused to engagewith the services on offer.

In the majority of cases, finding capacityto ‘own’ cases and attempt concerted

engagement, proved unrealistic.


second quotation below, though intended as apositive reflection, is perhaps revealing):

“People are going ‘put this down as an actionfor me, I’ll do this’, and it’s just a reallyrefreshing way of looking at things.”

“No-one gets put on the spot here, nobodygets put on the spot. It’s all transparent...youknow what you’re coming into.”

It is also relevant to note that 40 per cent of allactions set (largely by the police chair) were assignedto other police personnel. In the chair’s words;

“…my own staff, my own teams I know very well.I know what I can and can’t say to them. And it’seasier to put the pressure on if I need to.”

This is a fundamental dilemma when deliveringactivity in partnership: how to harness the benefitsof lateral collaboration, without diminishing thepower of vertical command and control structures?It was, incidentally, also a challenge encountered indifferent circumstances, during our parallel project inLuton, and one which is likely to become morepressing as multi-agency arrangements continue towiden and deepen.

Fourth and finally, we draw attention to therole of orthodoxy; the way activity wasframed and influenced, albeit subtly andsubconsciously, by habitual ways of doingthings learned in other similar partnershiparrangements. Research draws attention to thepitfalls of a strategy of multiple aims 30 in whichaccommodating the varied interests and goals of allcollaborating partners, can lead to a lack ofprogramme focus. We raise the question of

whether this plurality of purpose has, to somedegree, become internalised within agencies – notleast the police – and the extent to which this mighthave undermined the focus on this programme’sexplicit goal (of crime reduction). VMAP benefitedfrom a consensus view that violence was‘everybody’s business’, however the specific aim ofreducing violent crime incidents within the case-load perhaps got lost in a more general sharedeffort to ‘make a difference’. One reflection onVMAP case discussions is that they often seemedto default to risk management, for exampleensuring social services were aware that childrenwhere present in a house where violence occurred.While this is sensible ‘joined up’ working, itscontribution specifically to reducing recorded violentcrime, (especially in the short term) is probablytenuous at best.

30 Crawford and Cunningham (2015).

A fundamental dilemma when deliveringactivity in partnership is how to harnesslateral collaboration without diminishing

the power of vertical command.

More was not tasked because itwas generally understood that only

a little more could be done.


31 Quoted in Wood (2016).

VMAP benefited from a consensusview that violence was ‘everybody’sbusiness’, however the specific aimof reducing violent crime incidentswithin the case-load may have gotlost in the more general sharedeffort to ‘make a difference’.

A related concern – albeit one that was never explicit– is that participants liked VMAP because it (perhapssubconsciously) provided reputational cover foragencies dealing with risk, and as such participantswere satisfied simply to be discussing cases theyknew about in a forum with multiple partners. AsBaroness Jay has cautioned however “An issuewhich belongs to everybody round the…tableeffectively belongs to nobody” 31. The dangerousallure of diffused responsibility within multi-agencyworking must be recognised and guarded against.

Implications: confrontingthe paradox

The emergence of risk, harm and vulnerability asmatters of prime concern for the police is a centralfeature of the way their world is changing. Due toincreased awareness of complexity, the need forefficiency and the growing imperative to


demonstrate due-process, multi-agency casemanagement has become the dominant paradigmfor addressing these new priorities. There areperhaps worryingly few other big ideas.

In this context, the findings of the Police Effectivenessin a Changing World project in Slough may raisesome unsettling questions for policy makers andpractitioners. There is potential for genuine existentialdiscomfort here in response to an example of serviceprovision, enthusiastically delivered by dedicated andexperienced professionals, optimised andcoordinated by multi-agency collaboration, that still‘did not work’. To compound matters, where anevidence-base might provide some reassurance anddirection, there is worryingly little to reach for.Uncomfortable conclusions however must beconfronted and their implications considered. Wefinish by summarising the most striking of these andoffer five recommendations for progress.

VMAP was a small programme – it operated in fourwards of one town, for a one year period. Due toits modest scale, the evaluation techniquesavailable were of limited rigour, however theevidence generated about its (in this case lack of)impact, surpasses that in existence for a number ofmuch larger, more widespread and more heavilyresourced schemes that share similar principles.Of course this does not mean that those schemesdo not work; what it shows is that adoptingorthodox ‘best’ practice – forming caseloads,sharing information, engaging in multi-agency casediscussions and setting sensible actions – doesnot guarantee impact, even if it feels to thoseinvolved that it should.

It is a familiar plea, but there is an urgent needfor more evidence, both on what works inmulti-agency case management and in whatcircumstances multi-agency case managementworks. The VMAP experience should act as anantidote to complacency and a call for bravery.Recommendation 1. Those operatingsuch schemes should seek out robustevaluation (and those designing newschemes should build in evaluation fromthe start), both for the sake of their owneffectiveness and efficiency and to advancethe paradigm as a whole.

One cautionary lesson from VMAP is that outputsas well as outcomes need to be closelyscrutinised and objectively appraised. In ourmonitoring role we should admit our failure to seebeyond process improvements until it was too late.Recommendation 2. We would imploreothers to learn from this and (placing‘intangible’, internal benefits to one side)dispassionately weigh the additionality ofthe quantity and quality of interventionactivity generated, against the time andresource expended on process coordination.

We are also not blind to the fact that by taking acase based approach to addressing violence inSlough we returned somewhat ironically, to the kindof individuated response model that problemorientated policing approaches attempt to movebeyond. Managing priority cases will always beimportant, however in doing so there is a danger offailing to identify the common features that link themtogether. In this instance, analysing crime gave uslittle except priority cases to work with; however wespeculate that a secondary process of analysingthose priority cases (a process we start in our full sitereport) might provide insights into issues or areas ofservice provision in which strategic change might beconsidered. It is another matter however, given thepressures and constraints currently facing services,whether these could realistically be delivered.Recommendation 3. Agencies should usemulti-agency data-sets to systematicallyanalyse the characteristics and needs of thecases referred into multi-agency casemanagement schemes, and seek tounderstand local drivers and gaps incombined service provision.

Tasking and accountability within a multi-agencyframework appears to present a dilemma to whichfurther attention might usefully be given. Themanagement style required to create buy-in,positivity, willingness and occasional outbreaks of


Adopting orthodox ‘best’ practice –forming caseloads, engaging inmulti-agency case discussionsand setting sensible actions –does not guarantee impact.

AcknowledgementsThe Police Foundation is very grateful to the Dawes Trust for funding the Police Effectiveness in aChanging World project. We would also like to thank the leadership and staff of BedfordshirePolice, Thames Valley Police and Luton and Slough Borough Councils, along with their localpartners, for all their efforts in facilitating this project and participating in the research.

The authors would like to acknowledge the contribution of current and former colleagues at thePolice Foundation and to express their gratitude to the project’s National Advisory Group for theirwisdom and guidance throughout the project.

About the Police FoundationThe Police Foundation is the only independent think tank focused entirely on developing knowledgeand understanding of policing and crime reduction, while challenging the police service and thegovernment to improve policing for the benefit of the public. The Police Foundation acts as a bridgebetween the public, the police and the government, while being owned by none of them.


volunteering, appears at odds with the directivecommand and robust challenge sometimes requiredfor task-oriented delivery. In this regard, the idea thatlocal partner agencies might become increasinglyintegrated within public-protection/harm-reductionunits, operating with pooled budgets and singlecommand structures32 may offer a route to progress.Recommendation 4. In considering the casefor closer and deeper service integration,agency leaders should recognise theinherent weaknesses in multi-agency taskingarrangements and the potential benefits of amore unified command structure.

Any such integrated function, however, must haveclarity about its purpose. Despite their ostensiblecrime-cutting mission the police have learned – inpart through working with others – to also try to berisk-managers, problem-solvers, harm-reducers,needs-addressors, duty-doers, demand-cutters,reputation-managers, confidence-generators andreassurance providers. A shared orthodoxy hasemerged that can often suggest ‘sensible’ caseresponses in line with this broad ‘difference making’project, but which may do little to address theparticular purposes set for the specific programmeof which that case forms a part. It is important,therefore, to consider what it is realistic to expectfrom multi-agency case management arrangements.

With hindsight, given the level of complex needencountered, and the level of resource available,genuine problem-solving and crime-reduction wereperhaps never realistic outcomes for the VMAPprogramme. Harm-reduction, risk-management, oreven procedural compliance might be more realisticgoals for multi-agency case management, butachieving clarity and consensus, and findingsuitable metrics on which to assess performanceand measure success, are likely to prove formidablechallenges requiring leadership from the top. Theformal discourse lags behind multi-agency partners’intuitive appreciation of the synergies andinterdependencies between their work.Recommendation 5. Politicians and serviceleaders should clarify and better articulatethe shared purpose of the partnershipproject, and the roles that particular servicesshould play within it.

What we expect of our police service is changing.The new mission makes their work increasinglyintertwined with that of other agencies, and thereappears to be no alternative but to work ever-moreclosely together. The VMAP experiment reminds ushowever that closer cooperation is not in itselfsufficient to bring about transformative change torisky lives. It also highlights the urgent need to builda better understanding of what those agencies thatmake risk and vulnerability their business can dotogether, to reduce harm and keep people safe.


32 National Debate Advisory Group (2015).


ReferencesBerry, G., Briggs, P., Erol, R. and van Staden, L. (2011) The effectiveness of partnership working in a crimeand disorder context: A rapid evidence assessment. London: Home Office.https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/116549/horr52-report.pdf

Bryant, S., Peck, M. and Lovbakke, J. (2015) Reoffending Analysis of MAPPA Eligible Offenders. London:Ministry of Justice. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/407139/reoffending-analysis-of-mappa-eligible-offenders.pdf

Bullock, K., Erol, R. and Tilley, N. (2006) Problem-Oriented Policing and Partnerships. Cullompton: Willan.

Chapman, J., Higgins, A. and Hales G. (forthcoming) Police Effectiveness in a Changing World: Slough sitereport. London: The Police Foundation.

Crime and Disorder Act 1998, ch.1. Available at http://www.legislation.gov.uk/ukpga/1998/37/contents

College of Policing (2015) College of Policing analysis: Estimating demand on the police service. Ryton: Collegeof Policing. http://www.college.police.uk/News/College-news/Pages/First-analysis-of-national-demand.aspx

Cook, C. (2016) Troubled Families report ‘suppressed’. BBC News [online]. 8 August.http://www.bbc.co.uk/news/uk-politics-37010486

Crawford, A. and Cunningham, M. (2015) Working in Partnership: The Challenges of Working acrossOrganizational Boundaries, Cultures and Practices. Ch. 4. In Fleming, J. (ed.) Police Leadership –Rising to the Top, pp. 71-94, Oxford: Oxford University Press. https://www.researchgate.net/publication/288488191_Working_in_Partnership_The_challenges_of_working_across_organisational_boundaries_cultures_and_practices

Crockett, R., Gilchrist, G., Davies, J., Henshall, A., Hoggart, L., Chandler, V., Sims, D. and Webb, J. (2013)Assessing the Early Impact of Multi Agency Safeguarding Hubs (MASH) in London. London: London SafeguardingChildren Board. http://www.londonscb.gov.uk/wp-content/uploads/2016/04/mash_report_final-1.pdf

Crossley, S. (2015) The Troubled Families Programme: the perfect social policy? London: Centre forCrime and Justice Studies. https://www.crimeandjustice.org.uk/sites/crimeandjustice.org.uk/files/The%20Troubled%20Families%20Programme,%20Nov%202015.pdf

Dawson, P. and Cuppleditch, L. (2007) The Impact of the Prolific and other Priority Offender programme.Home Office Online Report 08/07. London: Home Office. http://webarchive.nationalarchives.gov.uk/20100413151441/http:/www.crimereduction.homeoffice.gov.uk/ppo/rdsolr0807.pdf

Dawson, P., Stanko, E., Higgins, A. and Rehman, U. (2011) An evaluation of the Diamond Initiative: year twofindings. London: London Criminal Justice Partnership. https://www.london.gov.uk/sites/default/files/slp_reducing_reoffending_board_-_may_2011_-_info_item_-_diamond_year2_final_120411.pdf

Department for Communities and Local Government (DCLG) (2014) National Evaluation of the TroubledFamilies Programme: Interim Report Family Monitoring Data. London: DCLG. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/335261/National_Evaluation_of_report_web_copy-ecorys.pdf

Eck, J. and Spelman, W. (1987) Problem solving: Problem-oriented policing in Newport News.Washington DC: Police Executive Research Forum.

Goldstein, H. (1979) Improving Policing: A Problem-Oriented Approach. Crime and Delinquency, 25, pp.236-258.

HM Inspectorate of Probation and HM Inspectorate of Constabulary (2014) A Joint Inspection of the IntegratedOffender management Approach. London: HMIP and HMIC. http://www.justiceinspectorates.gov.uk/crown-prosecution-service/wp-content/uploads/sites/3/2014/04/Integrated-Offender-Management-report.pdf



Higgins, A. and Hales, G. (2016) Cutting crime in the 21st century: Informed proactivity in the midst of socialand organisational change. London: The Police Foundation. http://www.police-foundation.org.uk/uploads/holding/projects/changing_world_paper_1.pdf

Home Office and Ministry of Justice (2010) Integrated Offender Management: key principles.London: Home Office and Ministry of Justice. http://library.college.police.uk/docs/APPref/Integrated-Offender-Management-Key-Principles-2010.pdf

Lum, C., Koper, C. and Telep, C. (2010) The Evidence-Based Policing Matrix. Journal of ExperimentalCriminology, 7, pp.3-26.

Morgan, J. (1991) Safer Communities: The local delivery of crime prevention through the partnership approach.London: Home Office.

National Debate Advisory Group (2015) Reshaping policing for the public. London: HMIC.https://www.justiceinspectorates.gov.uk/hmic/wp-content/uploads/reshaping-policing-for-the-public.pdf

O’Neill, M. and McCarthy, D. (2014) (Re)negotiating police culture through partnership working: Trust,compromise and the ‘new’ pragmatism. Criminology and Criminal Justice 14(2), pp. 143-159

Peck, M. (2011) Patterns of Reconviction among Offenders Eligible for MAPPA. London: Ministry of Justice.https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/217373/patterns-reconviction-mappa.pdf

Rittel, H. and Webber, M. (1973) Dilemmas in a General Theory of Planning.Policy Sciences 4 (1973), pp.155-169. Amsterdam: Elsevier.http://www.uctc.net/mwebber/Rittel+Webber+Dilemmas+General_Theory_of_Planning.pdf

Scott, K., Tepas, J., Frykberg, E., Taylor, P. and Plotkin, A. (2002) Re-thinking Violence – Evaluation ofProgram Efficacy in reducing adolescent Violent crime recidivism. The Journal of Trauma: Injury,Infection and Critical Care 53(1), pp.21–27.

Steel, N., Blakeborough, L. and Nicholas, S. (2011) Supporting high-risk victims of domestic violence: a reviewof Multi-Agency Risk Assessment Conferences (MARACs). Home Office Research Report 55. London: HomeOffice. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/116537/horr55-report.pdf

Weisburd, D. and Eck, J.E. (2004) What Can Police Do to Reduce Crime, Disorder, and Fear? Annals of theAmerican Academy of Political and Social Science, 593, pp.42-65.

Weisburd, D., Telep, C.W., Hinkle, J.C. and Eck, J.E. (2008) The Effects of Problem-Oriented Policing onCrime and Disorder. Campbell Systematic Reviews 2008:14. DOI: 10.4073/csr.2008.14.http://campbellcollaboration.org/lib/download/228/

Williams, A. and Barak, A. (2012) The Bristol Integrated Offender Management Scheme: A Pseudo-Experimental Test of Desistance Theory. Policing: A Journal of Policy and Practice 7 (2) pp.123-134

Wood, A. (2016) Wood Report: Review of the role and functions of Local Safeguarding Children Boards.London: Department of Education. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/526329/Alan_Wood_review.pdf


© 2016: The Police Foundation

All rights reserved. No part of this publication may be reproduced,stored in a retrieval system or transmitted in any form or by anymeans, without the prior permission of The Police Foundation.

Enquiries concerning reproduction should be sent to The PoliceFoundation at the address below.

The Police FoundationThe Foundry17 Oval WayKenningtonLondon SE11 5RR

020 3752 5630

www.police-foundation.org.uk

Charity Registration Number: 278257


Documents

Multi-agency case management: evidence and orthodoxy