1

Policing: Connecting Evidence and PracticeSIPR Annual Lectures, 2007–2012

2

Contents Leadership from Bottom to Top: Community Policing in Chicago Professor Wesley G. Skogan

Evidence based policing: what we know, and how we know it Professor Larry W. Sherman

Discipline and Discretion: The role of discipline and discretion in delivering an improvement agenda Professor Nick Tilley

Observations from a decade on the inside: Policing cultures and evidence based policing Professor Betsy Stanko

The New Police Professionalism – Democracy, Evidence Based Practice and a 21st Century Profession Peter Neyroud CBE QPM

3

4

PrefaceIn 2007 the Scottish Institute for Policing Research (SIPR) was established as a strategic collaboration between Scotland’s universities1 and the Scottish police service. Its key aims are:

• Toundertakehighqualityindependentandrelevantresearch;• Tosupportknowledgeexchangebetweenresearchersandpractitionersin

ordertoimprovetheevidencebaseforpolicingpolicyandpractice;• ToexpandanddeveloptheresearchcapacityinScotland’suniversitiesandthepoliceservice;• Topromotethedevelopmentofnationalandinternationallinkswithresearch,

practitioner and policy communities.

SupportedbyinvestmentfromtheScottishpoliceservice,ScottishFundingCouncilanduniversities,SIPR has rapidly become established as a leading international centre for policing research and as anexemplarofcollaborativeworkingbetweenthepoliceandresearchcommunity2. As part of itsmissiontoadvanceevidence-basedpracticesinpolicing,theInstituteinitiatedtheSIPRAnnualLecture programme to create an opportunity for leading scholars and practitioners to share their knowledge about the role of evidence in understanding and improving police effectiveness and publicsafety.Drawinglargeaudiencesfromtheworldsofpolicing,policy,politicsandresearch,theselectureshavebeenahugesuccess,stimulatingdebateanddiscussionaboutthenatureanddevelopment of contemporary policing.

Against a background of a growing recognition of the value of evidence-based approaches to policing,thisbookletcontainsfiveSIPRAnnualLecturesdeliveredbetween2007and2012byWesleySkogan(NorthwesternUniversity),LawrenceSherman(UniversityofCambridge),NickTilley(UniversityCollege,London),BetsyStanko(MetropolitanPoliceService)andPeterNeyroud(UniversityofCambridge).Eachspeakerdrewontheirwealthofexperienceanddistinguishedrecord of academic and professional achievement to craft lectures which offer compelling arguments and illustrations of how research evidence can be used to enhance the effectiveness and legitimacy ofpolicing.Fromimprovingpolice-communityrelationsandpublicconfidencetotacklinghotspotsofcrimeanddisorder;fromunderstandingbettertheusesofpolicediscretiontotheimportanceofevidence-basedapproachesfordevelopingtheprofessionalstatusofpolicing,eachofthelecturesmobilizes a range of research to address key issues and innovations in contemporary policing as well as highlighting important challenges for the future. Collectively the lectures also show the richness of the ‘supply’ of research evidence across different areas of policing as well as the growing ‘demand’ from within police organizations to integrate evidence-based approaches into policy and practice.

1 Abertay Dundee, Dundee, Edinburgh, Edinburgh Napier, Glasgow, Glasgow Caledonian, Heriot-Watt, Robert Gordon, St Andrews, Stirling, Strathclyde and the West of Scotland Universities

2 Engel, R.S. and Henderson, S. (2013) identify SIPR as an example of international best practice for police-academic partnerships, arguing that ‘structured collaborations that span multiple universities and police agencies will be most effective at advancing evidence-based practices in policing agencies’ See Engel, R.S. and Henderson, S. (2013) ‘Beyond rhetoric: establishing police-academic partnerships that work’, in J. Brown (editor) The Future of Policing (London:

5

6

Thereis,ofcourse,moretobedonetoincreasethereceptivityofthepoliceservicetoresearchbut huge progress has been made over the last 20 years3. Institutions and organisations like theSocietyofEvidence-BasedPolicing,theCollegeofPolicingandSIPRareallindicativeofthemomentum that has developed around the evidence agenda and are playing key roles in working collaboratively with the police to advance and embed evidence-based practices.

BringingtogetherthisfirstsetofSIPRannuallecturesalsoprovidesmewithanopportunityto thank the many organisations and people who support SIPR and without whom our annual lectureprogrammewouldnotbepossible.InvestmentfromPoliceScotland,theScottishFundingCouncil and the consortium universities underpins all of SIPR’s activities and I am hugely grateful fortheircontinuingsupport.IwouldalsoliketothankApexScotlandwhowereourpartnersforthefirstSIPRannuallecturein2007andtoAlphaTranslating&InterpretingServicesLtdwhohavesponsoredthesubsequentlectures.Eachyearwehavebeenfortunateinbeingabletoholdour annual lecture in splendid venues and I would like to thank The Society of Writers to Her Majesty’sSignet(TheWSSociety),theScottishPoliceCollegeandtheUniversitiesofEdinburghandStrathclydeforhostingourannuallecturesoverthepastfiveyears.ThedetailedorganisationofeachlectureistheresponsibilityofSIPR’sKnowledgeExchangeManager,TimHeilbronn,andIamvery grateful to him for his hard work in ensuring the smooth running of these events and to Peter Wilson QPM for his skillful chairing of each of the lectures. Janine Hunter (University of Dundee) hasdoneanexcellentjobinconvertingaudiorecordingsofthelecturesintotext,andAngelaDunphy(UniversityofDundee,DesignPrintMarketingDepartment)hasusedherexpertskillsonthegraphicdesignofthispublication.Finally,thesuccessoftheSIPRAnnualLecturesultimatelyreflectsthequalityofthepresentationsmadebythespeakerssoIwouldliketothankthemfortheirexcellentcontributionsandtheirassistanceinpreparingtheirlecturesforthispublication.

Professor Nicholas FyfeDirector,Scottish Institute for Policing Research

3 See Sherman, L. (2013) ‘Targeting, Testing and Tracking Police Services: The Rise of Evidence-Based Policing, 1975-2025’, in M. Tonry (editor) Crime and Justice in America 1975-2025 (Chicago: University of Chicago Press).

7

Leadership from Bottom to Top: Community Policing in Chicago4

Wesley G. SkoganInstitute for Policy Research/Political ScienceNorthwestern UniversityChicago,USA

BiographyProfessorWesleySkoganhasbeenafacultymemberatNorthwesternUniversitysince1971,andhas held joint appointments with the Political Science Department and the University’s Institute forPolicyResearch.Hisresearchfocusesontheinterfacebetweenthepublicandthelegalsystem,incrimeprevention,victimservices,andcommunity-orientedpolicing.Since1993hehasdirectedanevaluationofChicago’sexperimentalcitywidecommunitypolicinginitiative.Hisbooksonthisproject include: Police and Community in Chicago(2006),On the Beat: Police and Community Problem Solving(1999),andCommunity Policing, Chicago Style (1997). In 2003 he edited a collection of originalessaysoninnovationinpolicing,Community Policing: Can It Work?His1990book,Disorder and Decline: Crime and the Spiral of Decay in American Cities won an award from the American SociologicalAssociation.HewasatechnicalconsultantfortheHomeOfficeResearchUnitduringthedevelopmentoftheBritishCrimeSurvey,andhaswrittentwoHomeOfficeResearchSeriesreports on contacts between the police and public in England and Wales.

ForewordMyremarksoncommunitypolicingweredeliveredatthe2007Apex/SIPRLecture,whichwasheld in the historic Signet Library in Edinburgh. This wonderful venue brought together leaders of Scotland’s criminal justice community and others. I was visiting Scotland at the invitation of theScottishInstituteforPolicingResearch,ajointresearchventureoftheAssociationofChiefPoliceOfficersinScotland,theScottishFundingCounciland12universities.AsIlearned,ApexandtheInstitutebotharededicatedtomonitoringandevaluationintheirareasofresponsibility,strengtheningtheevidencebaseonwhichpolicyandpracticearebased.ThisfittedwellwithmydiscussionofthelargestcommunitypolicingexperimentinAmerica,whichwasbasedonathirteenyearevaluationofitseffectiveness.Theaudiencewaslargeandparticipantsaskedexcellentquestions.Scotlandiswellservedbytheawarenessandmutualrespectthatcross-institutionalcollaborations of this sort engender.

Professor Wesley SkoganMarch 2013

4 Held, in partnership with Apex Scotland, at the Signet Library, 11 September 2007

8

Today I will discuss some of the fundamentals of Chicago’s community policing programme and the thirteen year evaluation that we conducted to monitor its implementation and effectiveness. Chicago’sprogrammeofficiallybeganinAprilof1993,afterasurprisinglyshortdevelopmentperiodduringwhichseniordepartmentmanagers,outsideconsultantsandstafffromtheMayor’sofficecraftedaplanforthecity.Then,overthenexttwoyears,theyrefinedtheprogrammebytestingandreworkingitinseveraltestareas.Forpolicepurposes,Chicagoisdividedinto280smallpolicebeats,whicharegroupedinto25districts,andtheexperimentwascarriedoutinfiveofthem,overatwo-yearperiod.Theplannerstookaflexible,‘maybewecanmakeitworkandmaybewehavetoreformulateit’approach,gettingitoffthegroundrapidlyratherthantakingyearstoover-planit.Thisgave us evaluators a very valuable period during which police in most of the city were conducting theirbusinessasusual.Intheprototypedistricts,bycontrast,policehadsomeextraresourcesandagreatdealofmanagementattentionandextratraining.Theretheyreorganisedthemselvesinawaythateventuallytheentirecitywouldemulate.MyfirstbookonpolicinginChicagotookadvantageofdifferencesinwhathappenedintheseprototypedistrictsandelsewhere,todocumenttheimpact of the programme.

ThefirstcomponentofChicago’sprogrammeistodecentraliseanddevolveresponsibilitydownintheorganisation,closertowherepolicemeetthepublicandtheworkgetsdone.InChicago,thismeantbreathingnewlifeintothe280smallbeats.Thedepartmentcreatedwhatarecalled‘beatteams’;thesearegroupsofaboutnineofficerswho,inthemain,providestaffingforabeatcarthatisassignedtoanarea24/7.Ofcourse,otherofficersoccasionallygetdispatchedtohandleworkoverloads;butthegoalofthecomputeriseddispatchingsystemistokeepthebeatcarsbusyanswering calls from their area.

Asyoumightnote,thisindicatesthatChicagochosenottogothe‘specialunit’route.ManyAmericanandUKcitieshavetriedtostafftheircommunitypolicingprogrammeswithofficerswhoarereleasedfromtheroutinesof‘real’policework,butthatalmostinevitablyleadstotrouble.Oneliabilityofspecialunitsisthatofficersareconstantlybeingsiphonedoffforother,seeminglymorepressingduties;andChicagoknewfromitsinvestigationsthatthiswasadifficultytheywantedtoavoid.Anotherwasthatspecialunitscanleadtomoraleproblems.InFortWorth,Texas,forexample,specialcommunitypolicingofficerscametobeknownas‘emptyholsterguys,’andweredismissedbytherestoftheforce.Inothercities,whattheydoisknownas‘waveandsmilepolicing,’andNOP–NeighbourhoodOrientedPolicing–comestobeknownas‘NobodyonPatrol’.So,byorganisingteamsthatwouldspendtheirtimeansweringcallslikeeveryoneelse,onlyjustinoneneighbourhood,Chicagohopedtoavoidthemoraleandorganisationalproblemsthatplaguespecialunits.Thisstrategyalsoavoidedachargethathadcomeupinothercities,thattakingofficersawayfrompatrolwouldputthecommunityatrisk.InChicago,beatteamsdopoliceworkmostofthetime.The911emergencycallsystemmanagestheirworkload,soa)theyanswercallsthatcomefromtheirbeat;andb)sothey’renotfullyoccupied,andhaveotherthingsthattheycanalsododuringthecourseoftheirtour.Theteamsworkunderabeatsergeant,whoisresponsibleforquarterlymeetingsatwhichtheytalkaboutwhat’sup,andmakeplansforthenextquarter.

9

Atthesametime,Chicagochangedtofixedworkshifts,movingawayfromarotatingshiftpatternwhichusedtobeverycommonintheUnitedStates.Before,overthecourseofseveralmonths,officerswouldworktheirwayallaroundtheclock.Buttobuildstablerelationshipswiththepublicyou need stable shift assignments.

Thenextkeydevelopmentwastoprovidevehiclesforpublicinvolvementintheprogramme.InmanyAmericancities,publicinvolvementconsistsofacommitteeofthegreatandthegoodthatmeetsoccasionallywiththechiefofpolice.Itisabitexcitingforthem;theygetspecial,behindthescenestours,somecrimestatisticsandcrimemaps,andtheyfeellike‘insiders’.Buttheymeetonceaquarter,andnothingapparentlyhappens,exceptthattheyhaveaninterestingchat.

Instead,Chicagochosetodevelopasystemtofitthedevolutionofresponsibilityintheorganisationdowntobeats,andwhattheycameupwithisoneofthemostuniqueaspectsofthecity’sprogramme. Police and civic leaders come from all over the world to see beat meetings in action. Themeetingsaretobeheldeachmonth,andonaverage,theyactuallymeetabouttentimesayear.Eachhasaregularschedule–forexample,thefirstTuesdayofthemonth–andtheymeetinthe same location each time. They are held in venues like church basements and social halls and in parkbuildings;anditistheresponsibilityofthebeatsergeanttosecurealocationwheretheycanmeetregularly.Anaverageoffivepoliceofficersattend;mostaremembersofthebeatteam,andseveralofficerswhoareoffdutyatthetimeofthemeetingsarepaidovertimetobethere.Thebeatsergeantistypicallythere,andoftenspecialistsfromnichesinthedepartmentsuchasthegangunitaretherebecauseofspecialconcernsexpressedbylocalresidents.Butthekeyisthatmostoftheofficersworkinthebeat,drivingaroundandansweringcalls,andtheyaretheverypeoplewhoarelikelytoshowupatparticipants’doorstepsiftheycalltheemergencynumber.Theseofficershave direct responsibility for dealing with the concerns that come up during the meetings as well.

Theaveragemeetinglastsabout70minutes,andtheofficershavebeentrainedtofollowthesamegeneralagenda.First,thereisapresentationanddiscussionofcrimepatterns.Whenyouarriveatthe meeting there is always a ‘welcome table’ where you sign in and pick up information packets. Attendeestypicallygetacrimemap,andifthelastmeetingturnedouttofocusonaggravatedassault,there’llbeamapofalltheassaultsinthebeatinthelast30days.Thereisastandardanalyticreportcalleda‘TopTenList’,whichprovidesinformationaboutthetenmostfrequentcrimesinthisbeat during the last month. Often there will be summary reports about arrests in the area and a crimepreventionbrochureortwo.ThesematerialswillbeavailableinEnglish,SpanishandPolish,whichisthenextmostrequestedforeignlanguageinChicago.

Turnout at these meetings has been solid and very stable. It varies with the weather – we have serious weatherinChicago,asyoumayhaveheard–andSeptemberisourbestmonth;that’swhenyoushouldvisit,anditisalsothepeakmonthforparticipation.Initiallytherewasconcernthat,afterawhile,attendancewouldfade,whenthenoveltyoftrundlingdowntoyourlocalchurchhallandmeetingwiththepoliceworeoff.However,turnouthasbeenrunningatasteady65,000to67,000attendeesperyear.Between1993and2003,therewerejustover600,000participantsinthemeetings(seeFigure1).

10

Figure 1: Trends in Beat Meeting Attendance, 1995-2003.

The city-wide surveys we conduct show that community policing and beat meetings are very widelyknown.RecognitionishighestamongAfricanAmericans,andinfactit’sinAfricanAmericanneighbourhoodswheretheprogrammehashaditsmostsuccess.Aswetrackedit,recognitionwentupsteadilyovertime.Inourlastsurvey,89%ofAfricanAmericans(and80%oftheentirecity)knew about the programme. Much of this is doubtless due to the tremendous marketing campaign thatwentontogetthepublictocometothemeetings,andtokeeptheminformedaboutwhattheyneededtoknowtobeactiveparticipants.Theprogrammeisadvertisedusingmassmailings,flyersandsignspostedinthesubway.Churchesinsertlittleflyersaboutlocalmeetingsintheirweeklybulletins,andChicagoschoolchildrencomehomewithinformationattachedtotheirreportcards.Inonedistrict,thecommandersarrangedtohavepizzasdeliveredwithamapandabeatmeetingschedulestapledtothebox.Spanish-languageradioisusedveryextensivelytoreachoutto Chicago’s Hispanic community.

11

In1995and1996,about12,000neighbourhoodresidentswentthroughathree-weekendtrainingcycletolearnaboutneighbourhoodproblem-solvingfromtheirend,andtolearnabouthowtheycouldbecomeinvolvedincommunitypolicing.Ofcourse,policeofficersneedthemosttraining.Policingisahumanservicesoperation;youhire,train,andsuperviseyourpeopleinordertogetwhatyouwanttohappen.So,forexample,officersandsergeantsneededtolearnabouthowtorunameeting;thedepartmentmadeawonderfultrainingvideothatplayedouta‘beatmeetingfromhell’.Everythinggoeswrong;anobstreperoushecklerappears,asdoesaconfusedseniorcitizen.Theremainderofthesessionconsistsofofficersandtrainerstalkingabouthowtheywouldhave handled the situations that emerged.

Whoisitthatattendsbeatmeetings?Tofindout,weconductedroundsofobservationalstudiesin1995,1998and2002.Eachtime,weattendedhundredsofmeetings,toobservewhatwentonandsurvey the residents and police who attend. We found that people who come are not a particularly representativesliceofthepopulation;thisissomethingthateveryorganiserofacommunitypolicingprogrammehastoworryabout.Comparedwiththepopulationoftheirbeat,themeetingsover-representhomeowners,moreeducatedpeople,long-termresidents,seniorcitizens;peoplenolongerinwork,marriedhouseholdsandfamilieswithoutanychildrenlivingathome.Inshort,itisbetter-offandmoreestablishedmembersofthecommunity,withtimeontheirhands,wholearnabout,andtakeadvantageof,theseopportunitiestoinfluencepolicingintheirneighbourhood.

However,althoughtheylookdifferentfromtheirneighbours,itturnsoutthat,onmanydimensions,thosewhoattendadequatelyrepresenttheconcernsoftheirneighbours.Weassessedthisby comparing the priority problems reported by beat meeting participants with the same measuresinsurveysofthegeneralpublic,neighbourhoodbyneighbourhood.Itturnsoutthatbeatmeetings,byandlarge,doreflectthedistributionofconcernsinthecommunity.Participantsare more concerned than their immediate neighbours about crime and other problems – that isanimportantreasonwhytheyshowup.But,onmanymeasures,theydoaprettygoodjobofreflectingcommunitypriorities.Whyisthisthecase?Ithinkthatsmallisbeautiful;eventhoughparticipantsweremorelikelytoowntheirhome,forexample,beatsaresmall,andeveryoneisstillshoppingatthesamegrocerystore,walkingpastthesameschoolandcatchingabusatthesamestop(seeFigure2).Theyaresharingallthevisibleproblemsintheircommunityregardlessofdifferences in their backgrounds.

12

Figure 2: Demographic representation of those attending beat meetings, home ownership amongst attendees and beat population.

Importantly,thisturnsoutnottobetrueforsomethingsthatwehopecouldberepresentedatbeatmeetings.Oneoftheseisresidents’viewsofthequalityofpoliceserviceintheircommunity;ironically,itispeople’sviewsofpolicingthatareleastwellrepresentedatthemeetings.Whyisthat?TheprincipalreasonisthedemographicmismatchthatIdescribed;thatis,demographicmismatchespile up in ways that make the meetings more police-friendly. One culprit is older residents. There is a verystrongagegradientintherelationshipbetweenageandpeople’sperceptionsofthepolice;theolderyouare,themorepositiveyoutendtobe,byaverysharpmargin.

The second important factor that confounds the representativeness of the meetings is race. Whites who attend beat meetings in Chicago pretty much share the views of the white population in general – they both like the police. But it turns out that African Americans who come to the beat meetings are much more positive about the police than are their neighbours. These differential racialgapsinassessmentsofthequalityofpolicingensurethatbeatmeetingsdon’tclearlyrepresent the concerns of the public about the effectiveness with which they are doing their job.

13

One interesting factor we have tracked is the congruence between the views of participants andthoseofthepoliceofficerswhoattendthemeetingsandworkinthearea(seeFigure3).Policealsofilloutquestionnairesatthemeetings,andwehavefoundthattheirperceptionsofneighbourhood problems parallel those of residents to a surprising degree. Police are a little lessinterestedingraffitithanareresidents;residentsarealittlebitmoreconcernedaboutjunkandtrashinthestreetsandalleys.Butby-and-large,theprioritiesofthetwogroupsresembleeach other. This may be because of the repeated dialogue they engage in with participants at the meetings,butIthinkitisalsobecausetheturforientationadoptedbythedepartmenthasbroughtthemmuchclosertotheproblemsthatthey,too,seeastheypatrolafixedarea.Now,insteadofdrivingallovertowntowhereverthecomputersendsthemtonext,theystayon‘their’beat.

Priorities of Residents

burglarygraffiti

abandoned buildings

abandoned cars

car theft

public drinking

junk and trash

street crime

drugs

gangs

Priorities of Police

burglary

graffiti

abandoned buildings

abandoned cars

car theft

public drinking

junk and trash

street crime

drugs

gangs

Priorities of Residents

burglarygraffiti

abandoned buildings

abandoned cars

car theft

public drinking

junk and trash

street crime

drugs

gangs

Priorities of Police

burglary

graffiti

abandoned buildings

abandoned cars

car theft

public drinking

junk and trash

street crime

drugs

gangs

Figure 3: Officer and participant ratings of neighbourhood problems.

A third aspect of Chicago’s programme is interagency co-operation in problem solving. A key featureofcommunitypolicinginChicagoisthatitisnotthepolicedepartment’sprogramme;itisthecity’sprogramme.Infact,ifcommunitypolicingisjustthepolicedepartment’sprogramme,itisatriskoffailing.Withoutthesupportoftherestofthemunicipalserviceinfrastructure,it’snotgoing to be able to carry the freight.

Chicago’smodel,instead,requirestheactiveparticipationofmanyagencies:thepeoplewhotowabandonedcarsandpoisonratsinthealleys,andthecityworkerswhopaintoutgraffiti.Thereasonsforthisarethreefold.Firstly,politiciansandcivicleaderswhorunChicagohaveboughtahundred percent into what is known as the ‘broken windows’ theory of crime. They believe that youcantacklesomebigproblemsbytakingcareofsmallthings,asarecentpolicechiefputit.InChicago,tacklingabandonedbuildings,cars,graffitiandillegaldumpingisseenascrimeprevention.FixingthebrokenwindowsisoneofChicago’sapproachestocrime.

Secondly,thecityhasthismodelbecauseitfitsthemayor’s‘cleanandgreen’agenda.Chicago,likemanycities,istryingtocompeteintheglobalmarketplaceforcorporateheadquarters,tourists,conventionsandhigh-techstart-ups,andhowthingslookisverymuchafactorinthatmarketplace.Plus the mayor loves trees.

14

Finally,thosewhoplannedtheprogrammeknewthatwhenpeopleturnedupatbeatmeetingstovoicetheirconcernstheywerenotgoingtomakefinebureaucraticdistinctionsaboutwhoisresponsibleforwhat.Iftheirproblemisloosegarbageinthealleys,they’regoingtostandupandcomplainaboutit.Plannersknewthattheyhadtohaveanaffirmativeresponsewhenratsinthealleycameup.Iftheystoodthereandsaid,‘Ah,yeah,it’sterrible,butthat’snotpolicebusiness,’no-onewouldcomebacknextmonth.Beatmeetingswereinsteadstructuredtopayoff,withoutrespectforbureaucraticsilos.Evenbeforetheybegan,thecoordinationofabroadrangeofservices became very much part of Chicago’s programme.

Figure 4: CAPS Service Request Form.

Tomakethishappen,asystemwasdevelopedthatopenedaspecial‘window’thatofficers couldeasilygotoforservice.Atthemeetings,participants’complaintsgettranslatedtoservicerequestforms(seeFigure4).Everynight,thedistrictsfaxtheseformsdowntown,wheretheyareentered into a computer and allocated to the various departments. Each type of service has a requiredservicetime,andtheclockstartstotickassoonasitgoesinthecomputer.Themayoralsohasspecialauditorswhomakesurethatfallenstreetsignshaveactuallybeenputbackup, thatstreetlightsthatwereouthavefreshbulbs,andgraffitigetscleaneduporpaintedover.

CITY SERVICE REQUESTCHICAGO POLICE DEPARTMENT

DATE

DISTRICT _______ BEAT_______ LOCATION___ ___ ___ ___ ___ ______________________ ____________________________________

SERVICE DESCRIPTION: LIGHT WATER LEAK HOLE CLEAN CAVE IN BULK CART HYDRANT SIGN HEAT TREE OTHER

STREET VACANT LOT CURB PARKWAY CHA MIDDLEOFSTREET CORNER SIDEWALK ALLEY VIADUCT HOUSE SEWER BLOCK

ALL ONE TRAFFIC OUT CUT TRIM HIGH NEW OPEN TRIM LOW PLANT STUMP DAMAGED REMOVE RESTORE MISSING REPLACEINSUFFICIENT OTHER ___________________________

GRAFFITI REMOVAL:

MATERIAL: BRICK PAINTED BRICKWAIVER SIGNED: YES NO

SIDING(SPECIFY) _____________________________________________________________________

OTHER ______________________________________________________________________________

OWNER(IFKNOWN) _________________________________________________________________

ABANDONED BUILDINGS: VACANT OPEN BRICK FRAME

GARAGE SECURED FLOORS___________________

ABANDONED AUTOMOBILES: MAKE _______________ YEAR ____________________________

LICENCE PLATES: YES NO # _________________ STATE ________ YEAR ________

CITY STICKER: YES NO # _________________

HAZARDOUS AUTO: YES NO FLOAT: FULL PARTIAL

COLOUR: ___________________________________________________________________________

VIN: ________________________________________________________________________________

OTHERINFORMATION: ______________________________________________________________

____________________________________________________________________________________

CRIMINAL HOUSING TASK FORCE: BUILDING LOCATION _________ NUMBEROFUNITS ___________ OWNER’S NAME COMPLAINT __________________________________________________________________________ ______________________________________________________________________________________

NUSANCE PAY PHONE: REMOVE

HUMAN SERVICES REFERRAL

LIQUOR CONTROL REFERRAL

COMPLAINT INFORMATION: NAME _______________________ ADDRESS ______________________________________________APT NO. __________ HOME PHONE ____________ WORK PHONE _________________________ORGANIZATION ______________________________________________________________________ ______________________________________________________________________________________

NEIGHBOURHOOD RELATIONS OFFICER

ONLY ________________________

PRIORITY (CHECK ONE)

EMERGENCY NON-EMERGENCY

NARRATIVE ___________________________________________________________________________ ______________________________________________________________________________________

REPORTINGOFFICER __________________________ STAR NO. ______________________________R.D NO. ________________________________ I&INO. _________________ ___________________

WHITE COPY - SEND TO MAYOR’S OFFICE OF ENQUIRY AND INFORMATION, CITY HALL, ROOM 100BLUE COPY - RETAIN IN NEIGHBOURHOOD RELATIONS OFFICE

(INITIALS)

(DIR)(NUMBERS) (STREET NAME)

15

InChicago,ifyouhaveaproblemandcometoabeatmeeting,youcangetitfixed.Onereasonfor high and stable participation in the meetings is that things happen as a result. We see this in oursurveys,inwhichover80percentofparticipantsreporttheyhaveseenchangestakeplaceintheir neighbourhood because of things that go on at the meetings. There is pay-off from linking the meetingstoservices.Andovertime,thisprocesshasactuallyremadeChicago’sservicedeliverysystem,enhancingitsresponsiveness,whichwasanotherofthemayor’sgoals.

Theresultscanbeseeninthedata.Aspartoftheevaluation,wetrackthedistributionofservicesin Chicago. We want to see what the impact of beat meetings has been on the service delivery process.Twoveryfrequentservicesthatwehavetrackedcloselyaregraffiticlean-upsandtowingabandoned cars. When the programme started there were reputedly (no one really knew the number)morethan10,000abandonedcarsonthestreetsofChicago,andgettingthemcleanedupwasoneofthebigfirstprioritiesoftheprogramme.

Iwillspareyouthecomplicatedstatistics,butouranalysisofbeatmeetingandservicedeliverydatafound a short list of factors that were important in determining which beats got more of what service.Oneimportantfactorwasresidentpriorities.Thiswasmeasuredbyourcity-widesurveys,aggregated to the beat level to make neighbourhood data. In areas where problems were of substantialconcern,theservicedeliveryratewashigher.Atthesametime,therewasanadditionaleffectoftheprioritiesofthepeoplewhocametobeatmeetings;wheretheywereconcerned,theservicedeliverywasevenhigher.Inaddition,inbeatswithahighturnoutrate,thereweremoreservices,evidencethatthe‘squeakywheelgetsgreased’.Finally,thisisChicago,sosupportingthemayorisimportant.Controllingforeverythingelse,themostimportantdeterminantofservicedelivery rates was the percentage of the vote that went for the incumbent mayor. This is absolutely normalinAmericanpolitics,andpeoplewouldbeastonishedifthatwasnottrue,especiallybecause this is Chicago.

Figure 5: Service Delivery Model: factors determining which beats got more of what service

resident priorities service

delivery rate

beat meeting

participant priorities

beat activism

support for incumbent mayor

land use patterns

16

Whatarethechallengesfacingtheprogrammeinthefuture?Thefirstquestionis,cancommunitypolicingsurviveCOMPSTAT?COMPSTATisahard-nosed,datadrivenmanagementaccountabilitythatbeganinNewYorkCity.Likemanycities,ChicagoadopteditsownversionsofCOMPSTATin 2001. The problem is that what matters in data-driven management is what’s measured. In COMPSTATmanagementmeetings,whichfeaturehugechartsthrownuponhigh-techLCDscreens,thetopbrassaskcommanders,‘Well,whatareyoudoingaboutthisdrugmarketandwhat are you doing about those shootings?’ The data in the department’s computers are about crime,arrests,responsetimesandcrashesonthewaytocalls.That’swhatismeasured,andthat’swhat’s important in this kind of accountability regime. This inevitably pushes resources away from community policing and in the direction of the traditional responsibilities of the organisation. Districtcommanderscomplainthatthingsthattheythinkareimportant,thattheyhavebeenfosteringwiththecommunity,simplydisappearfromviewwhentheyareinthehotseat.

Ourevaluationteamhasbeenputtingsomepressureonthedepartmentaboutthis,andithasincorporatedsomecommunitypolicing-typemeasuresintothereviewprocess.Districtcommanders,forexample,arecloselyquestionedaboutservicerequestrates,andbeatmeetingturnoutisanissuethat they can be called on the carpet to account for. But most of the data available to analysts at headquartersfocusesontraditionalactivities,andtheyinevitablybecomeparamount.

Thesecondquestionis,cancommunitypolicingsurvivetheendofgreatcrimedropoftheendofthetwentiethcentury?LikemanyAmericancities,crimeinChicagopeakedin1991,andthenbeganalongslidedown.Thisfreedupresources,tookshootingsandviolentcrimeoffthepoliticalagenda and gave departments’ breathing room in which to re-engage with the community. Between 1991and2006,anycrimewithaguninChicagodeclinedby67percent.Thiswasaverynoticeabledecline,tosaytheleast,andthebiggestdeclineshavebeeninAfricanAmericanneighbourhoodswhere problems were worst to begin with. But now there’s a hint of a turnaround in the United States.Nationwide,thecrimeratehasstoppeddropping,andinanumberofvisiblecitiesit’sbeguntocreepupagain.Chicago’scrimecounthassimplyflattened,notcreptup,butpressurefromthemediaandfromcommunitygroupsinareaswherethecrimeproblemlooksworse,furtherthreatens the resources devoted to community policing.

ThethirdquestionIdon’tknowtheanswertois,canChicago’sprogrammesurviveanewmayor?We’vebeenthroughanumberofpolicechiefs,andweareintheprocessofpickinganother,soIknowthatcommunitypolicingcansurvivethistransition.Chicago’sprogrammeisveryfirmlyrootedinitspoliticsandculture,anditistrulythecity’sprogramme.Peoplerunforpublicofficewiththefact that they are a community policing activist on their campaign résumé. This involvement is a big plusinneighbourhoodpolitics.Chicagoansknowtheirbeatnumber,theyknowwheretheirbeatmeetingsare,anditwouldbeverydifficulttodislodgetheprogrammepolitically.Itcouldbestarvedforresources,butIthinknopoliticiancouldfinditfeasibletoannouncethattheywerenolongergoingtobeahundredpercentbehindtheprogramme.Wewillfindoutsomedayifthisistrue,whenourmayor-for-lifefinallyretires,butinmanycities,turnoveramongchiefsofpoliceandmayorshasbeenarealtestingpointforhowfirmlycommunitypolicingisrootedinthecivicculture.

17

OurfinalprobleminChicagoiswhattodoaboutournewimmigrants.Communitypolicinghasfailed to engage with Chicago’s burgeoning Latino community. Chicago is divided into three great communities:weareasimpleplace,populatedbywhites,AfricanAmericansandHispanics,thelatteroverwhelminglyfromMexico(seeFigure6).TheLatinofractionofthepopulationistheonlypartthat’sgrowing.Thewhitepopulationdroppedby13percentbetween1990and2000,andtheblackpopulationisstable.Latinoneighbourhoodsareoverflowing,schoolsthereareovercrowded,and that’s where the city’s future lies. Community policing has been unable to successfully penetrate thelargeandgrowingbarriosthathaveemergedinChicago;placeswhereyoucanliveyourentirelifespeakingonlySpanish.Twothirdsofthecity’sLatinosliveinmajority-Latinobeats,andthatproportionhasbeengrowing.Theyhavebecomemoreconcentratedovertime,astheirnumbershavegrownthroughimmigrationandnaturalgrowth.Findingwaystorespondtothis,andtoengagewiththiscommunity,isperhapsthecity’slargestchallengeinthisnewcentury.

Figure 6: Chicago is grouped into 25 districts, divided into 280 small police beats, populated by whites, African Americans and Hispanics.

18

ReferencesSkogan WG. 2006. Police and Community in Chicago: A Tale of Three Cities.NewYork:OxfordUniversity Press.

Skogan WG. 2004. Community Policing: Can It Work? Cengage Learning.

TheChicagoCommunityPolicingEvaluationConsortium:WesleyG.Skogan,LynnSteiner.2004.CAPS at Ten: Community Policing in Chicago. An Evaluation of Chicago’s Alternative Policing Strategy. http://www.northwestern.edu/ipr/publications/policing.html

SkoganWG,SteinerL,DuBoisJ,GudellEJ,FaganA.2002.TakingStock:CommunityPolicinginChicago.Chicago:InstituteforPolicyResearch,NorthwesternUniversity.http://www.ncjrs.gov/pdffiles1/nij/189909.pdf

SkoganWG,HartnettSM,DuBoisJ,ComeyJT,KaiserM,LovigJH.2000.ProblemSolvinginPractice:ImplementingCommunityPolicinginChicago.Chicago:InstituteforPolicyResearch,NorthwesternUniversity.http://www.ncjrs.gov/pdffiles1/nij/179556.pdf

SkoganWG,HartnettSM,DuBoisJ,ComeyJT,KaiserM,LovigJH.1999.On the Beat: Police and Community Problem Solving.Boulder,Colorado:WestviewPress.http://skogan.org/files/On_The_Beat_Police_and_Community_Problem_Solving.pdf

SkoganWG,HartnettSM.1997.Community Policing, Chicago Style.Oxford:OxfordUniversityPress.

Skogan WG. 1990. Disorder and Decline: Crime and the Spiral of Decay in American Cities Berkeley: University of California Press.

19

ScottishPoliceCollege,TulliallanCastle,1October2009

Evidence based policing: what we know, and how we know it

Lawrence W. ShermanDirectoroftheInstituteofCriminology,UniversityofCambridge,UK

BiographyProfessorShermanisDirectoroftheInstituteofCriminologyattheUniversityofCambridge,where he has been Wolfson Professor of Criminology since 2007. He also serves as a non-executiveDirectoroftheCollegeofPolicing(UK),aswellasDirectoroftheCambridgePoliceExecutiveprogramme.His1998PoliceFoundation(US)Lectureon“Evidence-BasedPolicing”ledtothedevelopmentofthecurrentgraduatepolicingcurriculumatCambridgeUniversity,andisreflectedinthemissionofthenewCollegetoidentifyandpromotewhatevidenceofwhatworksinpolicing.TheformerPresidentoftheAmericanandInternationalSocietiesofCriminology,hewaselected Honorary President of the Society of Evidence-Based Policing in 2010.

ForewordSinceIdeliveredtheSIPRLectureattheScottishPoliceCollegein2009,manythingshavechangedinevidence-basedpolicing.Myownthinkinghasexpandedtheframeworktodescribethe“Triple-T”ofEBP,presentedatgreatdetailinmycontributiontoVolume42ofCrimeandJustice,editedbyMichaelTonry.Muchmoreresearchevidenceon“targeting,testingandtracking”theuseofpoliceresourcesisnowavailable.Someoftheplannedstudiesdescribedinthelectureneverhappened,whilemanyothersnotdescribedhavebeencompleted.ThegrowthofrandomizedexperimentsinUKpolicinghasbeenastonishing,andwehavemanyyoungpoliceleaderstothank.Manyofthemcame to Scotland in early 2013 for a SIPR-sponsored meeting of the Society for Evidence-Based Policing.Thenew,potentiallyRoyal,CollegeofPolicinginEnglandandWaleshasbeenmandatedtodevelopandpromoteevidenceofwhatworks.Allofthismakesthelecturebelowincomplete,butnot out of date. The basis ideas are as enduring as the Scottish Enlightenment from which they are drawn. It was an honour to bring them back to Scotland in this form.

Lawrence Sherman,CambridgeApril,2013

20

Evidence based policing: what we know and how we know it (October, 2009)Whenwelookbackatterribleevents,therearetwokeyanalyticalquestionsthathavetoberaised.Oneistheextenttowhichthetragedyispredictable,foreseeableinadvance;andthesecondis,evenifitwaspredictable,wasitpreventable?Doweknowhowtopreventatragedyofthatkind?Thesequestionshitparticularlyclosetomyfamilywhenour1819houseinCambridgeshirecaughtfirelastweekafewhoursafterweleftforwork.TheyhavealsobeencentraltotheUKpolicediscussionaboutwhetherthetragicdeathsofLouisePilkingtonandherdaughter,aftershecalledLeicestershire Police on repeated occasions about anti-social behaviour.

Intermsofprediction,Iknowinourownsituation,wedidn’thaveanyoftheriskfactorsthatareassociatedwithfires;otherthanthefactthatitwasanoldhouseandwasthereforehadsomeoldwiringinit.Wealsoknowthattheriskofthatfirewasextremelylow,andwethereforehavetoacceptthattherewasnowayofsaying‘youaregoingtohaveafireinthathouse’.Butevenwithalowrisk,thereisstillaveryhighuncertaintyaboutwhetherornotsomethingbadisgoingtohappen.IfyoulookatitfromtheperspectiveofLeicestershireConstabulary,atthepointofthefirstphonecallbyMsPilkingtontothepoliceandthethirtiethcallsevenyearslater,youmayfindthatyouareinthedenominatorpoolthatissobigthattheoddsofatragedyoccurringmayhavebeenalmostasunlikelyasthefireinourhouse.Thequestionofhowpredictableorcomprehensible something is when looking backwards from an evidentiary standpoint is completely irrelevant,becauseyoudon’thaveabasisforsayingthatofcourse,allthosecallsoverthattimeperiod could result in the kind of tragedy we saw in Leicestershire.

Whetherthefireinourhouse,ortheLeicestershiretragedy,hadbeenpreventableraisesaseparatesetofquestions.Whatisthethresholdofindicatorsorriskfactorsthatwouldneedtobemetbeforeyou would kick into place a policy for action that’s designed to prevent the tragedy? If I look at Anti-SocialBehaviouralOrders(ASBOs)orfamilyordersasanevidencebasedcriminologist,Iwouldsaythatno-onehasanyideaoftheeffectivenessofthosestrategies,eveniftheyhadbeenused.

But,thatdoesn’thelpwhenthereisnecessity–forpoliticalorculturalreasons–tofindsomebodytoblame.Ifthatnecessityarises,thereisalwayssomebodywhocanbeblamed,nomatterhowfairlyorunfairly.Thereis,forexample,theSimmonsfamily;nowgettingdeaththreatsandabuse[thelocalauthorityhadpreviouslytakenoutaninjunctionagainstStevenandSuzanneSimmons,whosechildrenwereallegedtobeinvolvedintheabuse;theinjunctionwasnotenforced].Questionscouldberaisedifthat’sanappropriatefocus,or,asthegovernmenthaschosentodo,tofocuson the Leicestershire Constabulary. The Northern Ireland Policing Board has demanded and receivedanapologyfromtheirnewChiefConstable,MattBaggott,whowasChiefConstableofLeicestershire Constabulary at the time of the case. Theproblemwiththeblamecultureofcourse,isthatblameassumesboththepredictabilityandthepreventabilityoftheevent.Italsonarrowstheresearchquestiontoasinglecaseoutofahugedenominatorofcases,wherein99%ofcasesthesamedenominatoroccurredbutnothingterriblehappened. Asking ‘who’s to blame?’ rather than ‘why did it happen?’ doesn’t allow us to ask how to prevent such harm in the future and learn how to change systems.

21

But,ifwelookforward,inthewaythatotherfieldshavedone,Ithinkwecanbeencouragedbysuccessstoriesfromevidencebasedimprovement.Onemajorexampleispracticesintheairlineindustry,whereoverthelast30yearstherehasbeenahugereductioninthedeathsfromcommercialjetlinercrashesperpassengermilesflown.Ifyoulookatautoaccidentdeaths,atleastinEnglandandWales,theyarenowdownattheleveltheywerein1924,despitehugeincreasesinpopulationdriving,anddrivingmiles.FiredeathsaredownallovertheUSandEurope,asfarasIhavebeenabletodetermine,andformegoingforward,Ithinkwe’vesimplydecidedthatwe’regoing to rewire the house.

So what kind of evidence are we talking about? What we are interested in is systematic and helpful evidence about how we can get better results if systems are changed in certain ways. That’s why in 1998IpresentedtothePoliceFoundationinWashingtonDCalectureonevidencebasedpolicing–andfailedtocopyrightthetermortrademarkit,sothateverytimesomebodysays‘what’sthe evidence base for policing?’ we could keep track of it (www.policefoundation.org/content/evidence-based-policing).

We have many different conceptions of what evidence based policing (EBP) is about. Even I keep changingmymind,soIthankyouforthisinvitation,whichhelpedmeclarifywhatIthinkitreallyaddsupto.EBPisadecision-makingprocessthatusesreliable,unbiasedquantitativeevidenceonpredictionandpreventionasaprimarycriterionforsettinggoals,choosingpriorities,makingpolicies,makingdecisions,managingcompliance,assessingresults,andimprovingpolicies.It’sthinkingaboutevidenceasadashboardforsteeringpolicyandguidingdecisions,onaturnbyturnbasis;butitallowsforU-turns,becausethere’snothingmoreusefulthanaU-turnonceyouhavetheinformation if you’re going in the wrong direction. I’m appalled every time the press goes after a governmentforchangingitsmindonpolicy,ifit’sadecisionbasedongoodevidencethatthepolicyisn’tworking.Ithinktheyshouldgetmedalswhentheydothat,ratherthanknee-jerkcriticism.

AsIoftendescribeittoourstudents,EBPisa‘DRIVER’process,thatis:

• Diagnosis—localevidenceontheproblem• Response—reviewsofpublishedevidence• Implementation—localevidence• Value-added—actualversuspredicted• Evaluation—isthisthebestwecando?• Revision—startalloveragain Itbeginswiththediagnosis;atlocallevels,frompatternsandtrendsinwhat’sgoingonwithinapoliceagency,andthenrelatingthosefindingstotheinitiativesthathavebeenrigorouslyevaluatedandpublishedintheliterature,orevenunpublishedliteraturesuchassystematicreviews.Implementationofthepolicyagainreliesonlocalevidence,whichcouldthenhelptounderstandwhetherthepolicyisaddingvalue;recalling,asthepresssooftenforgets,thatthefundamentalcauseofacrimeordisorderproblemisnotinadequatepolicing,butsocialconditions.Allofthiscanthenbeevaluatedastowhetherwe’reimplementingthepolicywellenough;ifwehavetheright

22

policymatchforthelocaldiagnosis,andisthereanywayinwhichweoughttorevisethepolicy,gathermoreevidence,andstarttheprocessagain?It’sinspiredbywhatDrW.EdwardsDeming(whoestablishedtheconceptofqualitycontrol)calledthe‘continuousqualityimprovementcircle’(‘Plan,Do,Study,andAct’).

Figure 1: Evidence based policing (Sherman, 1998).

ThisisthecirclewhichIlaidoutin1998,inwhichyoutakeyourin house evidence on what’s going onintheagency;andthenyougototheliterature,what’sknownabouttheproblemandhowbesttodealwithit;thishelpstodevelopinternalornationalpolicyguidelines;howyouactandtheoutcomes to be accomplished in certain conditions.

ShouldcertainthingshavebeendoneinLeicestershirethatweren’t?AsfarasIcantell,nonationalpolicywasviolated,nolocalornationalguidancewasignored;thepoliceweredoingthingsthatwerewithintherangeofacceptableconduct.Theycouldhavedonemoreperhaps,ifyoulookatitunderamicroscope,butthatcouldbesaidaboutmanythingsinhindsight.Thelargerquestionofoutcomesthen,relatesnottoanyparticularcasebuttoanoverallpattern.Ifyouaregettingbetterresultsusingthepolicythanifyouweretotrysomethingelse,howwouldyouknowunlessyou—or someone else before you-- had actually tested your hypothesis?

Allofthatbringsustoevidence-basedmedicine(EBM),whichwasthemostdirectmodelforEvidence-basedpolicing,onlysixyearsafterEBMasamovementbeganinCanadaandEnglandin1992,especiallyundertheleadershipoftheenterprisingScotobstetrician,SirIainChalmers.HewastheguidingforceforcreatingtheCochraneCollaboration,whichisdevotedtotakingthemassofmedicalresearchandsystematicallyintegratingitforpractitioners.Inmedicinetoday,practitionerscanincreasinglygotoacomputer,putinadiagnosis,andgetbackthebestevidenceonhowtodealwiththatproblem.EBMhasrevolutionisedmedicalpractice,notjustintheUK,butaroundtheworld,wheretherearenowsome70countriesparticipatingintheintegrationofmedicalpracticeevidence.Thisevidenceismadeavailabletopatients,notonlymedicalpractitioners,sothateverybody concerned with achieving the best possible outcome of a particular case has the best evidence on how to make it happen.

Literature

In-House

Guidelines

Outputs

Best Evidence

Outcomes

23

EBMisvitallyimportant,evenifsomedoctorsdon’twanttodoit.Manydoctorswouldstillratherpractice‘experiencedbased’medicine,relyingonclinicalwisdomandanecdotalexperienceratherthan systematic analysis. That is how such dangerous medical practices as blood-letting were practicedforcenturies.WhenthatpracticewasfirsttestedinParisinthe1830s,manydoctorswerefuriousaboutit.Theysaidthatitwouldtakeawaythe‘art’ofmedicine,whichwasthekeytoitssuccess,bystressingmeretechnicalities.Ofcourse,thefindingsshowedthatpeopleweremorelikelytodiefromhavingtheirbloodremovedandlesslikelytodiewithoutit,butitstilltookdecades to put a stop to that practice.

SirIainChalmershassaid:‘inmedicine,theproblemwithnothavinganevidencebaseisthatwhenweinterveneinthelivesofotherpeople,wenotinfrequentlyendupdoingmoreharmthangood’.IfweweretothinkintermsofaHallofShame,andnon-evidence-basedmedicine,wehavetogototheAmericanDr.BenjaminSpock,thepaediatricianwhomayhaveindirectly“killed”alotofbabies. He used untested theories in giving advice to parents on the critical issue of how you put yourbabytosleep:onitsback,oronitsstomach,oronitsside?Hewasveryclearaboutputtingababytosleeponitsstomach,forpurelytheoreticalreasons.Thiswasfromhisunsystematicexperience,hispersonalbiasedsample,ratherthanasystematictestofhypothesis.Hedidnotdo what was needed in order to interpret cause and effect about what he thought he knew. He ignored the empirical revolution promoted by the Scottish Enlightenment philosopher David Hume (1711–1776),whoadvisedustobemorescepticalaboutwhatwethinkweknow,andhowwethinkweknowit,anddon’tevenbegintothinkwehavetheinformationnecessarytoinfercausationfrompersonalexperiencealone.

Inthe1980s,thedirectoppositeofSpock’stheorywastested:thatbabieswouldbebetteroffsleepingontheirback.Usingnotarandomisedtrialbutquasi-experimentalevaluationoverafifteenyearperiodofadvisingparentstoputbabiesontheir“backtosleep,”thedeathratefromcotdeathor sudden infant death syndrome dropped substantially. That evidence strongly suggested that Dr. Spock was wrong. But it was not the kind of evidence that is valued the most.

The problem with looking at trends and inferring causation is the logical fallacy of inferring that ‘after this’ therefore means ‘because of this’. We have seen it in our own research into restorative justiceinCanberrafrom1995to2000,wherethereisaveryimpressive29%dropincrimeforthosecasesthatparticipatedinrestorativejusticeforcrimesofviolence.Butofcourse,ifthecasesofsimilarcharacteristicsthatweredealtwithwithoutrestorativejusticehaddroppedby50%,theconclusion would have been wring despite a declining trend. Simply looking at a trend without a comparison leaves you still in the dark about what effect the intervention has had. In fact it looks likethis(seeFigure2),withthedottedlinebeingtherandomlyassignedcontrolgroup,gettingevenworse over the two years after the random assignment. What we’re seeing is actually a bigger effect than we thought just by looking at the trend in relation to the implemented programme.

24

Figure 2: Restorative justice for a sample of violent offenders in Canberra from 1995 to 2000; comparison of crime levels amongst of those participating in restorative justice (solid line) and the randomly assigned control group (dotted line).

The problem is that we can’t distinguish a trend after programme implementation from what would havehappenedanyway--thenaturaltrendorhistoricalpattern.That’soneexampleofwhatismeantbytheterm“spurious”explanations:plausibletheoriesthatcanbeexplainedawaybyotherfactors.

Thepointaboutusingcontrols,orbestofall,randomisedcontrolledtrials(RCTs),isthatyoucanruleoutexplanationsyouhaven’teventhoughtof.Nomatterwhatwasgoingonatthesametimeasthetrendintheexperimentalgroupchanged,itwasalsogoingonwiththecontrolgroupaswell.Thatiswhytheimpactisnotfoundinthetrendofonegroup,butinthedifference between the trends of two groups. That is what made the pioneering development of RCTs in the UK – a Britishinventionpioneeredin1948bySirAustinBradfordHill–thegoldstandardforwhatworks,with over a million trials in medicine being completed since Hill’s. It is widely agreed in medicine thatotherresearchdesignsaremorepronetobiasandspuriousexplanations.

IntheEBPliterature,Ihavetriedtomakeastrongcasetoignorenon-randomisedtestingtothedegreepossibleandtopreferRCTs.Forpredictions,RCTsarenotrelevant,becausepredictionscanbereliablewithoutunderstandingcausation.Even,then,however,standardsmatter.Whilethein-housedataisnecessarilydescriptiveandaboutpatternsandtrends,itcanbeusedinaveryprecisewayintermsofprediction;whereareeventsmostlikelytooccur,involvingwhom,whattimeofday,etc.

LetmeofferthreeexamplesofthepossibleapplicationofEBP.Firstly,inthecaseofmurder,whereithasn’tyetbeenusedinpolicing,buthasbeenusedinprobation.Secondly,thenow20-yearhistoryofconcentratedpolicepatrolsinhotspotshasspreadwidelyacrosstheUS.Thirdly,EBPhas been half-successful in the complicated history of the police responses to common assault in a domestic environment.

101%

121%

71% 71%

28%

49%

100%

20%

40%

60%

80%

100%

120%

140%

Yr (-2) Yr (-1) Yr (+1) Yr (+2)

CourtOffenders (n=59)

DCOffenders (n =62)

25

Example 1: MurderIntermsofmurder,evidence-basedpredictionscangreatlyimproveonexperience-basedpredictions. I don’t know if Scotland has the Multi-Agency Prevention Partnership Agreements (MAPPAs)whicharewidelyusedinEnglandandWales.Iwouldsuspectthatifyoudo,itwouldbebasedonclinicalprediction;thatis,subjectiveassessmentofwhoisdangerousbylookingattheircriminalrecords,asopposedtostatisticalforecasting,usingtensofthousandsofcasesusingdataminingandnon-linearanalysisofthepatternsofdevelopmentwhichareprecursorsforextremelyrare events. These are the needles in the haystack that statistical forecasting can identify in ways thatnoindividualscan,becausenoindividualisevergoingtohave30,000casestodealwithinthecourse of their career. No one person can absorb the full information that can be gleaned in an actuarial analysis using supercomputers that have only become available in this century.

The other problem with clinical prediction as it is currently used in multi agency prevention partnerships is that it doesn’t take false negatives systematically into account. The failure to identify that somebody is dangerous before they kill somebody is rarely used to hold clinical prediction accountableforitserrors.Peopleleftoffthe“danger”listarenotplacedundersurveillance,norunder review.

Amorecost-effectivewaytoapproachthepreventionofmurderorotherextremeharmislaidoutinanarticlebyRichardBerkandIandourcolleaguesin2009:‘Forecastingmurderwithinapopulationofprobationersandparolees:ahighstakesapplicationofstatisticallearning’,publishedinthe Journal of the Royal Statistical Society.(http://onlinelibrary.wiley.com/doi/10.1111/j.1467-985X.2008.00556.x/abstract).

WelookedatPhiladelphiaprobationcasesoveranextendedperiodoftimeinacitywhichhas300–400murdersperyearinapopulationof1.5millionpeople.Sowhatwecouldsaytoanation,suchasScotland,withsixmillionpeopleandaround110murdersayear?Wehavethedatatodothingsthatwouldtakeyoualotlongertodo;andperhapstherearesomeinsightsfromhavingamuch higher murder rate.

One insight is the incredible importance of age as the most powerful predictor affecting the risk ofcommittingmurder,andtheenormousfallfromtheageof18to40.However,ahiddenfactoristheageatfirstadultdisposition;iftheageoffirstprosecutionasanadultwas12,thenthatindividualwillnotageoutoftherisk,soevenasa40or50yearoldpersontheyhavethehighestriskofcommittingmurderofanyage.Becausetheywere12yearsoldwhenfirstchargedwitharmedrobbery.Andthatisnotsomethingthatisbeinglookedatclinically;norisittheentirestory.

If you look at the article’s analysis of some 17 risk factors in terms of how much error would be added to the forecasting by false negatives and false positives if it was not known if they were a factor,youcanseethatageisalmosttwiceasimportantasthenextfactor,ageoffirstcontact;havinga‘gunprior’isthethirdmostimportantpredictoroferror,andfourthisbeingmale,whichislow compared to what we would normally think in terms of its power as predictor.

26

Figure 2: False negatives and false positives.

Themostimportantpointisthatyoucannotreducethisforecastingevidencetoachecklist;thepersonalhistoryofeachoffenderhastobeineffect“droppeddown”thestatisticalmodeleverytimeaparticularcriminalhistorypatterngoesintoit,soyoucomeupwithaforecastthatwouldtakeintoaccountboththelikelihoodofafalsepositive,whereyouarepredictingthatsomebodyislikelytocommitamurder,andthentheydon’t;aswellasafalsenegativewhereyoupredictthattheywon’tcommitamurder,buttheydo(SeeFigure2).

Figure 3: Philadelphia probation population (40,000) over a two year period: 2% at high risk of committing a murder; 60% at low risk, and 38% neither at high risk or low risk.

That is the basis for what turned out to be a very interesting pyramid prepared by Dr. Geoffrey Barnesamongtheprobationpopulationofsome40,000peopleinPhiladelphia,whichisthat2%ofthem,lookingforwardoveratwoyearperiod,wereinfactathighriskofcommittingamurder;60%wereatlowriskofcommittingamurderandtherestwereinbetween.Intermsofforecastingmurder,themodelidentifiedpeoplewhowerechargedwithmurderorattemptedmurder75timesmoreoftenthanthelowestriskgroup.Now,that’sstillbasedonahigherrorrate;it’snotperfect,it’snotcertain,it’snotguaranteedandit’snotsomethingyou’dnecessarilywanttoputsomeoneinprisonfor,whichiswhatalotofpeoplesay,whentheylookatthis.ThejudiciaryinPhiladelphiaisquitekeentogetthiskindofreportbeforeeverycase.ThenewDistrictAttorneywantstousethisasabasisfordecidingwhotodivertfromprosecutionaltogether,becauseheiscommittedtoloweringtheincarcerationrateandthisgiveshimsomebasisonwhichtodothat,withoutputting

Actually Happens Actually does NOT Happen

Predicted to Happen 1. True Positive 2.FalsePositive

Predicted NOT to Happen 3.FalseNegative 4. True Negative

High Risk (2%)

Neither High nor Low Risk (38%)

Low Risk (60%)

27

thepublicatgreaterrisk.Thisraisesethicalissues,butIwouldarguethatthoseissuesarealreadypresent even without these data. Judges and others are making the same kind of assessments on a casebycasebasis,byusingclinical,biasedandindeedraciallyprejudicedtoolsratherthanrationaltoolsandstatisticalanalysis.Why,then,woulditbelessethicaltouseunbiasedforecasts?

Inprobation,offendersarealreadysentencedandundercommunitysupervision.Thequestionswhich we answered in this article were about managing sentenced offenders. If you wanted to directmoreresourcestopeoplewhowereextremelydangerous,howwouldyoupickthem?Andthen,onceyouknowthat,howcouldyourestructureyourprobationagentsorofficers?Usingouranswers,PhiladelphiaProbationofficialshaveessentiallyputmostofitsoffendersonlowintensitysupervisions–twovisitsayeartotheprobationofficeforthelowriskgroup.ThisstrategywastestedinanRCTshowingthatitmakesnodifferencecomparedtosixor12visitsayear,whichweren’t being delivered anyway. So it turns out that compared to the best they can deliver in frequencyofcontact,withacaseloadofwhatwasabout175offendersperprobationofficer,therewas no harm in lowering the intensity of visits in the lowest group. We’re about to launch cognitive behaviouraltherapyandotherinterventionsfortheverysmallnumberofhighriskcases,toseeifitwill reduce serious crime for people who have already been sentenced to community supervision. Butthetrade-offisadegreeofuncertaintyandimperfection;wehadsomebody,forexample,whowasclassifiedaslowriskwhowaschargedwithamurderthispastsummer.Hissupervisionwouldn’thavebeensubstantiallydifferenthadhebeenhigh-risk,butitfedtheprobationofficerunionoppositiontorisk-basedclassification.

28

Example 2: Hot spotsUsedbypolicearoundtheUS,hotspotswereoriginallyunderstoodtobeverysmallmicrolocations,aboutthesizeofthislecturehall,withhighconcentrationsofcrimeanddisorder.Theseconcentrationsprovidedameansofpredictionthatcouldbeaguidetoefficientprevention.Justtoreturntothisthemeofpredictionandprevention,wehaveseenthatpredictioniskeytopreventionnotonlyintermsofsettinggoals,butsettingprioritieswithefficiency;focusingtheresources on high risk rather than on low risk targets and the peaks and valleys of crime probability.

Figure 4: Peaks and Valleys of Crime (Distribution of Violent Offenses in Tokyo)

The Japanese National Police Agency has brilliantly portrayed this in a density map of time and spacewhichshowstheprobabilityofviolentcrimepersquarefootinTokyo.Itshows“volcanoes”ofcrimes,mostofwhichareactuallylocationsofpedestrianconcentration.Thatisanexplanation,butnotarefutation,inthesensethatifthat’swherethepeopleare,ifthat’swheretheminorcrimesare,thenarguablythat’swherethepoliceshouldbe.AnAmericanpoliceleadernamedOrlandoWWilson,inhis1950bookPolice Administration,saidthatpoliceshouldpatrolthroughouttheentirejurisdictioninordertocreatetheappearanceof“omnipresence”asaneffectivedeterrent.Butthatwas merely theory-based policing. What we know from random assignment from over a hundred hotspotsinMinneapolisisthatwhenhalfofthehotspotsgotanaverageoftwohoursextrapolicing a day the result was a long term trend of crime reduction in the overall crime rate in the city,almostentirelyduetoreductioninthehotspots.

29

What David Weisburd and his colleagues found in Seattle is that crime rates change primarily in the small portion of a city where hot spots are located. About half of all crime in Seattle occurred inthe15%ofblockswhicharethemostactivecrimelocationsinthecity.Astheirarticle‘CrimeTrajectories at Places: A Longitudinal Study of Street Segments in the City of Seattle’ (Criminology 2004:42:283–322),almostallofthedecreaseincrimeover15yearsoccurredinthoselocations.Ifpoliceconcentrateonthesevolcanoes,thedisplacementargumentisthatifyousquashitdownhereit’sgoingtopopupoverhere.Yetalmosteverytimewe’veactuallytestedthatargument,inProfessorDavidWeisburd’s(DistinguishedProfessorofCriminology,LawandSocietyandDirectoroftheCenterforEvidenceBasedCrimePolicyatGeorgeMasonUniversity,Virginia,USA)workandinotherpeople’swork,wefailtofindsubstantialevidence.Therearesomescatteredindications,buttheevidenceisclearlyagainstdisplacementinanearbytimeandplace.Whatweseehereistheencouragingideaoffocusingonhotspots,whichinthiscasetheSeattlepolicedidnotspecificallydo,butoverthattimeperiodtherewasarevolutionininformationsystems,withimproved crime analysis directing police patrols.

Butit’sstillnotgoodenough,andthat’swhytheGreaterManchesterPolicehavenowjoinedwithCambridgeUniversityintheTacticalExperimentsandStrategicTesting(TEST)programme,whichwillextendtheUShotspotspatrolexperimentsintotheUKinidentifiedManchesterhotspots.TheMinneapolis data was based on observing – with a stop watch – the number of minutes the police werestandingorpatrollingineachhotspot;bothinthecontrolgroupandintheexperimentalgroup.Theyfoundthattheoptimalcontroltimeinahotspotisabout10to15minutes,measuredbythenumberofminutesafterthepoliceleavetothetimethefirstcrimeoccurs;or,howlongdoesittakebeforethefirstcriminalstickshisheadupovertheparapetandsays‘Ithinkthecopsaregone,bang,givemeyourpurse!’orsomethingtothateffect.

Figure 5: Effects on Crime and Disorder Maximized by 14-15 Minute Stops: the Koper curve (Koper C, ‘Just enough police presence: reducing crime and disorderly behavior by optimizing patrol time in crime hotspots’,

Justice Quarterly 1995: 12: 649-672).

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

30

ThisiswhattheKopercurvelookslike;it’sloggedinrelationtosome5000observationsofpolicepresenceinhotspots;whenthepolicearrive,whentheyleave,andtheoccurrenceofcrimeisrecordedbytheobservers.(Thatwasanexpensiveapproach,butsuchmeasurementcannowbedonecheaplybyGPSlocatorsinpolicecarsandradios.)InMinneapolis,inacontextofgenerallyrisingcrime,therewasasubstantiallylowerrateofincreaseintheextrapatrolhotspots,usingcallstothepoliceastheindicator.Forsoftcrime,therewasnotasgreataneffectasforhardcrime,includingrobberies,assaultsandotherpredatorycrimes,whichwerereducedbyextrapatroldespitetheoverallriseinthecrimerate.Inthecontrolareas,thehardcrimesremainedconstant.

The research designed in Greater Manchester Police (GMP) would take advantage of the size of thatagency,wherewecanidentify200hotspotsthataresmallenoughsothatonepolicecarcanbe seen from anywhere in that hot spot. The hot spot must be far is enough from others to be independent so they are not confused in the causal analysis. It is also good to add a displacement cushionaroundthemeasuredhotspots,with100ofthemreceivingtwohours’extraconstables’patroldaily;theother100receivingnoextrapolicing;andthencomparethecrimetrendsbeforeandaftertheexperimenttookplace.

Studies can measure police presence all over the city of Manchester using the automatic radio locator system which is just coming online and should be fully operational within a year. Analysts canalsolookatarrestsmadeateachhotspotandthenthequalityofobservations.We’recertainlyveryconcerned,atatimewhenpolicehavemadecomplaintshavereachedrecordlevelsinEnglandandWales,thathotspotpolicingbenonprovocative,andnotgeneratemorecomplaintsthanthecontrolgroup.Thekeyquestionisthedistinctionbetweenaproactivepresencewithoutaproactiveintervention,whichiscertainlysomethingthathascomeupinrelation to the G20 protests in London.

ThefinalpointhereisthateveryotheragencyintheUK–andindeedintheworld–isinvitedtojoinin.Iwillcomebacktotheideaofmulti-siteexperiments,becauseingeneralitadvancestheevidenceratherthandoingsingleexperiments.InthisGMPmodelweshouldhavebothincidentsandcrimereports,andwithanyluckwe’llbeabletotakesomemeasuresofpolice–communityinteractions.

AllofthiscameoutofourgraduateapproachattheUniversityofCambridge,inwhichstudentsin the Diploma course just this year – Chief Superintendents and Superintendents for the most part – have taken the evidence-based policing idea to heart and used it to persuade their chief officersandtheJerryLeeCentreofExperimentalCriminologytosupportnotonlyahotspotsexperimentbutseveralotherexperimentsaswell.WecanhereciteAlfredNorthWhitehead(1861-1947)whosaid:‘Universities[help]createthefuture’–withalittlemoremodesty,byaddingtheword‘help’.FortheseofficerstocometoCambridgeandtalkaboutwhatworksinpolicing,ifnothing else has led to them thinking about policing in a different way. If we can have our graduates leanonevidence,notjustfromManchesterbutfromelsewhere,thatwouldbeameasureofaccomplishment of the programme.

31

Example 3: Domestic common assaultOurthirdexampleofevidence-basedpolicingiscommondomesticassault.UndertheleadershipofMarcusBeale,theActingDeputyChiefConstableandpermanentAssistantChiefConstableinStaffordshire,wehavedesignedanexperimentthatcouldtestpolicediscretionagainstamandatoryarrestpolicy.Mr.BealecamebacktoCambridgelastyeartofinishhisMastersdegree,andforhisdissertation he analysed several hundred arrest-eligible cases for common domestic assault in onebasiccommandunit(BCU)inStaffordshire.Hefoundaclearpatternofofficersoftendoingsomethingdifferentfromwhatthenationalpolicywouldrequire–presumptivearrest.Instead,many of them were consulting victims and asking the victims what they think would be the most effectivepoliceactiontomakevictimssafe.Dotheywantthearrest,dotheynotwantthearrest,dotheywantthefellowtogotohismother’s,ortohisbrother’s?Notin100%ofthecases,butinaverylargeproportion,there’sevidencethatevenwhenthearrestwasbeingmade,itwasbeingmade in conjunction with a victim consultation.

Therecanbelotsofargumentsaboutthis.IthinkwecansayfromAmericanexperimentsindomesticviolenceisthatvictimpreferenceasapolicyhassimplyneverbeentested;wehavenoidea if a victim consultation approach does better than a mandatory arrest policy. Beale’s dissertation raisedthequestionofconductingarandomisedtrialcomparingvictimconsultationtothemandatoryarrestpolicytoseewhichonehasalowerleveloffrequencyorseriousnessofrecidivism.

Someofyoumayknowthatthatproposalbuildsona30yearhistoryofresearch,inwhichthefirstrandomisedtrialsforanyoffenceandalsofordomesticviolencetookplaceinMinneapolisin1980–84.Overallinthatfirsttest,arrestreducedrepeatoffendinginthenextsixmonthsbyabouthalf;butresearchsincethenfoundmorecomplexresults(LawrenceSherman,PolicingDomesticViolence.NewYork:FreePress,1992).

Whattheoverallbodyofresearchevidencehasnothadismuchinfluenceonpolicy.Themostinfluentialresultwastheinitialfindingthateverybodywelcomed,andwhichwaspublishedinThe Times:thatarrestworks,crackingdownworks,gettingtoughworks.Inthose300Minneapoliscasesrandomlyassigningarrest,advisingororderingtheoffenderoffthepremisesfor24hours;arresthada10%recidivismratecomparedto24%forsendingthesuspectawayand19%forpoliceadvice.

Whatyoushouldsayaboutsomethinglikethisis‘veryinteresting’andrepeattheexperiment;thatis,whatyouwouldsayifyouhaveascientifictemperament.Scienceneedsreplication;weneedtoknowifthesameresultscanbefoundrepeatedly,andinthecaseof‘arrestworksbest’,thatisnotwhatwefoundinfivefurthertestsofarrestfromdomesticviolence.But,wedidfindsomethingelsethatwasconsistentacrossthethreetests,andthatis–notwhatpeoplewantedtohear–butthat the effect of arrest depends on the employment of the suspect and the employment level in theneighbourhood.Thismeantthatarresthelpsdeteremployedmen,butitmakesunemployedmenmoreviolent;inareasoflowunemploymentarresthasadeterrenteffect,butinareasofhighunemployment it has a criminogenic effect.

32

Figure 6: Miami: Percentage of offenders with repeat domestic violence

Figure6summarisesthefindingsinMiami;limitedtomarriedcouples,primarilyHispanicrecentimmigrants;whentheywereunemployed,itdoubledtheirriskofrepeatarrestsfordomesticviolenceinthenextsixmonths;whenemployedithalvedthatrisk.ThesamepatternswerefoundinthefrequencyofrepeatviolenceinMilwaukeeinapredominantlyAfricanAmericanpopulation;andinOmahainapredominantlyWhiteworkingclasspopulation.So,threedifferentethnicgroups,threedifferentpartsofthecountry,verydifferentwaysinwhichpolicewerecalled–neighbourscalledpolice,familymemberscalledpolice.Wegotahighlyconsistentfinding,withtheeffectalsodependentuponneighbourhoodandworkingworstinareasofconcentratedpoverty;wheremostcalls to police about domestic violence come from.

TheevidenceraisesaverydifficultquestionaboutthecommonassaultordomesticviolencepolicyinEngland,whichstates:‘theperpetratorshallbearrestedimmediately,ifavailable.Failuretodosomay amount to a neglect of duty and the reasons for not doing so must be clearly documented’ (TheMetropolitanPolice,PolicyStatement,2004;restatedinSeptember2012:http://www.met.police.uk/foi/pdfs/policies/domestic_violence_policy.pdf)

One thing we do know from randomised trials is that presumptive arrest in many cases can cause moreviolencethanifthere’snoarrest;butwecan’tthensaythatdoingnothingisacceptable.Thequestionofwhattodonext,especiallyinacountrywithanationalisedpolicy,wouldbeansweredbytheStaffordshireDomesticViolenceExperiment,whichmightbejoinedbytheManchesterpolice. We invite everybody in this room to consider if they would like to participate in this multi-site,orRandomizedExperimentNetwork(RExNet)comparison,underagovernmentdispensationfromthemandatoryarrestpolicy,thevictimconsultationprocess,definedas:discussingalltheoptionswiththevictim,helpingthevictimadvisepoliceonwhatmightworkbesttostoptheviolenceinboththeshortandlongrun,includingarrest;andthepolicethenfollowthatadvice.Thetrialprotocolinvolvesvolunteerofficers,whodeclarethecaseeligible,calltheUniversityofCambridge,thecaserecordsgointothecomputer,whichthenrandomlyassignsusingamathematicalequalprobabilityprocesstodetermineifthecaseshouldbevictimconsultationorfollowHomeOfficepolicy.

0%

5%

10%

15%

20%

Arrested

Employed

Warned

Unemployed

Arrested Warned

33

Whatwouldbetheresult?Wecan’tknowunlesstheexperimentisconducted.WeproposetocallittheWISDOMexperiment.We’vetrieddeterrence.Nowwemighttrywisdom,andseeifthe Women’s Invitation to Support Domestic Offender Managementworks.This,incidentallyisconsistentwithrecentadvocacywritingbypeoplelikeLindaMills(ProfessorofSocialWork,NewYorkUniversitySchoolofLaw),whoisherselfastudentofdomesticviolenceandavictimofit.The hypothesis is essentially that there would be less harm in the aftermath of a police response guidedbyconsultationwiththevictim,incomparisontothepresumptionofarrest,bothwhentheoffenders are present and when they are not.

Toconclude,itisincreasinglypossibletopursuewhatworksbestinpolicing.Relentlesspursuitof the best evidence from published research can be linked to in-house data to enhance the outcomes that policing can produce. What we know from evidence based policing is that some highriskcrimeispredictable.Usingpoliciesaroundpredictionscan,onaverage,getbetterresultsthannotusingthosepolicies,atleastwhenyousubjectthepoliciestosystematictests.Asaframeworkforgoingforward,usingknowledgetoimprovetheeffectivenessofpolicinghasthebestprospects. We would achieve far less by shutting out the provable facts and going with theories or withthelatestpoliticalgovernmentoftheday.Withevidence-basedpolicing,whatwearedoingis what David Hume asked us to do almost 300 years ago in Edinburgh: to distrust our purely personalexperiences,andseekinsteadamoreobjectivebasisforourknowledge.

Thank you.

34

Discipline and Discretion: The role of discipline and discretion in delivering an improvement agenda5

Professor Nick TilleyJill Dando Institute of Crime ScienceUniversityCollegeLondon,UK

BiographyNick Tilley is Professor at the Jill Dando Institute of Crime Science at UCL and Emeritus Professor ofSociologyatNottinghamTrentUniversity.HehascarriedoutconsiderableworkintheUK,andbetween1991and2003wassecondedtotheHomeOfficeresearchdepartment,developingproblem-orientedpolicingapproaches.TheHomeOfficesetupthe‘TilleyAward’in1999toencourage and recognise best practice in tackling crime and anti-social behaviour using problem oriented-principlesinthelongterm,workingwithlocalagenciestotackletherootcausesofproblems.ProfessorTilley’sresearchinterestslieinpolicing,crimepreventionandprogrammeevaluationmethodology,andheisauthororeditorof10booksandaround150chapters,journalarticles and published research reports. He was awarded an OBE for services to Policing and Crime Reductionin2005andelectedtotheAcademyoftheSocialSciencesin2009.

ForewordGivingpubliclecturesinanunnervingexperience.Inthiscase,itwasespeciallyso.IdeliveredmySIPRAnnualLecturein2010andwasfollowingverydistinguishedpredecessors,whoIadmireandwhoseworkleavesme,forthemostpart,inawe.Moreovermyaudiencewasmadeupofsmart,thoughtfulandinformedpractitioners,policy-makersandresearchers.Inmylecture,Iwastryingtoworkthroughsometentativeideasonpolicingimprovement,whichhadtheiroriginsinobservationsofparalleleffortstofosterbetter,evidence-basedhealthcareintheUnitedStates.Iwasgratefulforthe opportunity to present my arguments and for the attention of those addressed. It was a privilege to talk about research for policing improvement to a group so clearly committed to it.

Professor Nick TilleyMarch 2013

5 Held at the Court Senate Suite, University of Strathclyde, 30 September 2010

35

TheoriginsofthislecturelieinVermont.IwasthereintheSummerof2009atameetingofnurses,physiciansandpharmacistswhowerepartofaburgeoningand,tomymind,impressivemovementforhealthcareimprovementintheUnitedStates(Kenney2008).Myrolewasasoneoftwo‘wizards’ to provide an outsider’s input into their discussions at the end of each day’s deliberations. RayPawsonandIhadbeeninvitedbecauseofourworkonevaluationmethodology(Pawson&Tilley1997;Pawson2006;Tilley2010),whichproposesanalternativetothestandardexperimentalandquasi-experimentalmethods.Thosewhoinviteduswereinterestedinevidence-basedmethodsof improving healthcare and they too had found some standard research methods unworkable. They were attracted to the ideas we had for practical methods of building an evidence-based to improve policy and practice. I’ll come back to that at the end of the lecture.

Forthemoment,letmedescribetheissuethatcroppeduponeday.Itrelatedtohand-washing.EvensincetheworkofIgnazSemmelweis,aHungariandoctorworkinginamaternityunitinaViennesehospitalinthemidnineteenthcentury,hand-hygienehasbeenamajorissueinhealthcare. Semmelweis worked out that the high rates of death from childbed fever in the ward ran by physiciansatthehospitalwasafunctionoftheinadequatewaysinwhichtheywashedtheirhandsfollowingtheautopsiestheyconductedinthemortuary,whichwasadjacenttothematernityward(Semmelweis1983,originally1861).Theywashedtheirhands,butthedifficultyingettingridofresiduesfromthebodiestheyexaminedwasevidentfromthelingeringsmell.Therateofdeathfromchildbedfeverwasmuchhigherintheirwardthantheonerunbymidwives,whodidnotconductautopsies,andhadbeensoforseveralyears.Semmelweisinitiatedaprogrammeofhand-washing in a strong solution of chlorinated lime and the differences in death rates between the two wards disappeared.

TherewasenormousresistancetoSemmelweis’sfindings.Itwasdifficultfordoctorstocometoterms with the possibility that they had been inadvertently killing their patients. But history has vindicated Semmelweis and hand-washing failures are now widely recognised as important causes ofinfections.Notwithstandingthis,gettinghandswashedwellcontinuestobeachallengeandhealthcare improvements still include efforts to achieve more and better hand-washing.

The issue that arose in Vermont was whether it was ever right to break the injunction to wash hands before touching a patient. One view was that the precept was absolute. Under no circumstances should a patient be touched without prior hand-washing – it was non-discretionary. Anotherviewwasthatalthoughhand-washingwasgenerallytobeundertaken,therewerecircumstances in which it would be inappropriate. Those favouring unconditional rules highlighted the risks of failure to wash hands and the dangers of allowing discretion to practitioners. If hand-washing were to be open to judgement-according-to-the-situation then it would cease to be routine,wouldtendtolapse,andtheconsequencewouldbeasignificantincreaseintheratepractitioner-patient infections. Those favouring a less-than-absolute rule could envisage particular situations in which adherence to the rule would not be in the patient’s best interest. We were askedtoimagineasituationinwhichamainarterywassevered,wherethepractitionerpresenthad not washed their hands before the accident but where their prompt action could stem the

36

bleeding. The patient’s chances of survival would be improved by disregarding the injunction always to wash hands before providing treatment.

There is no doubt about the desirability of hand-washing. There is no lack of interest in working out waysofincreasingtherateofhand-washingandofreducingthefrequencyoflapsesinadequatehandhygiene.Thequestionposedwasthatofwhethertheaimshouldbeunconditionallytorequirehandwashingalwaysbeforeanycontactwithapatient.

The Vermont discussion then turned to discretion in medicine and public health more generally. When should behaviours be invariant? When should they be discretionary? How can desirable invariantbehavioursbeproduced?Howcangoodjudgement,incircumstanceswherediscretionisneeded,bestbeinculcated?Andhowcouldaresearchagendahelpanswerthesequestions?Theseare all key issues in improving health care.

Giventhatmydayjobrelatestoresearchonpolicing,crimepreventionandresearchmethodology,unsurprisinglythisledmetowonderabouttheequivalentquestionsinmyfield.TheinvitationtogivethislecturecameasIwasponderingtheseissuesand,perhapsrashly,Idecidedthatitwouldforce me to structure some ideas to share with you and elicit your own thinking.

I should stress at the start that although discretion comprises an interesting intellectual puzzle and posesaseriesofimportantpracticalquestionsforpolicy-makersandpractitioners,I’mnotatallsure that it is seen as a problem for policing. It may well be that the pattern of rule bound and discretionaryactionthatmarkscontemporarypolicingisdeemedaboutrightor,ifitisnotright,there’s little that can be done about it to improve matters. One of my starting points is curiosity: Howmuchdiscretionisthere?Howisitexercised?Andhowisitcircumscribed?Theotheristhat of good practice: What should be discretionary? In what ways can discretion be improved? What should be mandatory? How can mandatory behaviour be produced? Only having answered thesequestionsinrelationtospecificpracticesdoesitbecomepossibletoaskwhetherthereareshortcomingsnoworwhetherimprovementscouldbemade?Mystartingpoint,however,isthatpolicing would do well to embrace the kind of improvement agenda that is at work in health and that if it does so any change will need to consider issues of discretion.

Much that has been written about discretion in policing comes from the United States. Rather less has come from this side of the Atlantic. In the US until around 1960 it was assumed that the police exercisedlittlediscretion.Iftheysaworheardofanallegationthatacriminallawhadbeenbroken,theywouldinvestigate.Ifacrimehadbeencommittedandtheperpetratorwasidentifiedtheywould arrest and charge the person involved. This was referred to as ‘full enforcement’. As Joseph Goldstein (1960) put it:

37

Within the area of full enforcement, the police have not been delegated discretion not to invoke the criminal process. (P)olice are… charged with the enforcement of laws; success in police administration is directly related to completeness and perfection in the performance of this task.

Lessthanfullenforcement,thatisenforcementwithpolicediscretion,wasdeemeddangerous.Corruption was one possibility where scope for non-enforcement opens the doors for police to be prevailed upon to disregard some criminal offence because of the rewards that are offered. Abuse is also risked where a blind eye is turned to one form of criminal behaviour in return for co-operation inhelpingtosecuretheinformationonthebehaviourofothers.Thepoliceofficersshouldbeblindintheirapplicationofthelaw,showingnofearorfavour;oncediscretionisintroducedblindjusticeissacrificed.Onceblindjusticeissacrificed,confidenceintheruleoflawisjeopardised.

Whateverthetheoreticalmeritsoffullenforcement,inpracticeitwassoonrecognisedthatitisanon-starter. As Joseph Goldstein (1960) again put it:

Limitations of time, personnel, and investigative devices, all in part but not entirely functions of budget, force the development, by plan or default, of priorities of enforcement. Even if there were “enough police” adequately equipped and trained, pressures from within and without the department, which is after all a human institution, may force the police to invoke the criminal process selectively.

Discretion,accordingtothisviewandofmanyothers(forexampleHermanGoldstein1963,1967,1977,1990andGeorgeKelling1999),inescapablycharacterisespolicingandoperatesatanumberoflevels.Ultimatelytheindividualofficercannotroboticallyenforcethelaw.Iftheytriedtodosotheywouldbeparalysed.Toomuchtimewouldbespentontoofewcasesifthepoliceofficerwerealways to go for full enforcement in relation to each instance of law breaking s/he saw or heard about.Onalargerscaleitisnotconceivablethattherewouldeverbesufficientresourcesforfullenforcement to be a practical possibility. Discretion is built in.

Moreover,itisdoubtfulifanyonewouldeverseriouslyadvocatefullenforcement.What’streatedascrimeisoftenamatterofcontext,andthisis,Iguess,almostuniversallyacceptedwhenitcomestopractice.Thelinebetweennormalbehaviourand‘propercrime’,forexample,canbeathinoneand decisions not to crime what could strictly be classed as a ‘crime’ is unavoidably a routine matter insomecontexts.Forexample,inEnglandandWalesalmostaquarterof10to12yearoldboys(23.1%)technicallysufferedassaultsin2008-9,butfewerthatonein15believedthemselvestohavebeenvictims,becausemanyoftheincidentsinvolvedthekindsofsiblingandplaygroundscrapsthatarewidelytakentobeanormalpartofchildhood(MillardandFlatley2010).Fewwouldexpectanysensiblepoliceofficerasamatterofcoursetotreatthesesometimesseriouscrimesas matters calling for the arrest and prosecution of the perpetrators. If full enforcement were to betriedthecriminaljusticesystemwouldnotonlygrindtoahalt,itwouldalsolookridiculous.Furthermore,policingisacomplexsocialinstitution,withmultipleaimsnotallofwhichhavetodowithenforcement(Goldstein1977,1990).

38

Thediverseaimsofpolicing,whichincludesuchmattersaskeepingtrafficmoving,returningmissingchildrentotheirparentsandguardians,dealingwithneighbourcomplaintsofnoisyparties,protectingtherightsofdemonstratorswhilstpreventingthemfromrunningriot,andpersuadingrowdyyouthtobemoreconsiderateorthoseintimidatedbythemtobemoretolerant,furtherreducetheresourcesavailableforfullenforcement.Somepoliceaimsmayalsobeinconflictwithfullenforcement.Inparticularordermaintenancemaybetterbeachievedbyshowingflexibilityin law enforcement than by engaging in it fully at every opportunity. This is not just because full enforcementwouldtaketheofficerawayfromthestreets,whilsttheydealtwiththosearrested.Itis also because of the provocation that such a response would sometimes cause.

Inthelightoftherealconditionsoftheirwork,policeofficershavebeendescribedas‘streetlevelbureaucrats.’ Michael Lipsky says of these that:

Unlike lower-level workers in most organisations, (they) have considerable discretion in determining the nature, amount, and quality of benefits and sanctions provided by their agencies. Policemen decide who to arrest and whose behaviour to overlook. (Lipsky 1980: 13).

Notwithstandingtheinexorabilityofdiscretion,thereareeasilyrecognisedandfrequentlynotedrisksinitsexercise.AsJosephGoldsteinrecognisescorruptionisclearlyadanger.Policeofficersmaybepersuadedtoexercisediscretioninwaysthatfavourfriends,highstatuspeople,thosebelonging to the same ethnic group or those who are prepared to pay bribes. Likewise they may usediscretiontopersecutethosewhotheydislike,thosebelongingtodifferentethnicgroups,lowerclasspeople,orthosewhosedefinitionascriminalsbringsadvantagestothirdpartieswhoarepreparedtobribeofficers.Sometimes,discretionmaybeabusedinthefurtheranceofwhatare deemed important public service goals: planting evidence and the use of violence or threats to elicitconfessionsareexamples.Lipskyevensuggeststhatthistypeofbehaviourmaygoonwiththecollusionofsupervisors,who‘looktheotherway’(Ibid:19).

Discretionmayalsohaveunwantedunintendedconsequences.Ifdifferentpeopleareseentoreceivedifferenttreatmentforostensiblythesameactsthenacasualtymaybelossofconfidenceinthecriminaljusticesystemand,possiblymoreseriously,areluctancetoco-operatewithpoliceofficers(Crawford,2009).Thelatterreluctancemayspilloverintoviolence,wherethediscretionarybehaviouractuallyprovokescrime.Theproblemfortheofficeristhatdecisionshavesometimestobemaderapidlyintheabsenceofcompleteinformation.Moreover,evenifthereweretimeandinformationaplentypredictingtheeffectsofspecificdecisionsishazardous.

Thepoorpoliceofficer,then,hastoexercisediscretion,maybetemptedtoabuseit,mayhavesuperiorswhoareapttoturnablindeyetoitifithelpsachievetheirobjectives,willoftenmaketheirdecisionsinlessthanidealcircumstancesandwillrarelybeableinfullconfidencetoknowwhattheupshotwillbe.And,ifsomeunforeseendisasterfollows,s/herisksbeingblamedforit.Thewisdom of hindsight coupled with scapegoating bites hard.

39

Despiteitsproblems,policeofficersforthemostpartseemtoenjoythediscretiontheyhaveandsee it as one of the rewards of the job. A recent study by Michael Rowe (2007) looked at how policeofficersinthreeEnglishsub-divisionsrespondedtoeffortstoremovetheirdiscretioninresponding to incidents of domestic violence. The police have in the past been criticised for their useofdiscretionindealingwithdomesticincidents,whicharearchetypalrepeatcrimes.Thepolicehadtendedtodealinformallywith‘domestics’,totrytocalmthesituation,andtoavoidinvokingthe criminal justice system if this could be avoided. But this fails to acknowledge that an assault has takenplace,thatprotectionisneededandthatcallstothepolicetendtobemadeonlyafteralotofunreportedincidents.IntheareawhereRoweconductedhisfieldwork,fullenforcementwasthereforebroughtinasthepreferredresponsetothesecases.Thepoliceofficersacquiescedandfollowedthedirectivetheyhadbeengiven.Strictlytheyretainedtheirdiscretioninprinciple,butthe decision not to enforce brought trouble for them from their supervisors: they would have to justifyandacceptresponsibilityfortheconsequencesiftheydecidedtodeviatefromforcepolicy.Theofficersweretrainedtomakearrestsandtoldthatarrestwasinthebestinterestsofthevictim even if that’s not what they wanted. Some were persuaded that the new policy was the right one,butinanumberofparticularcasesthearrestpolicywasfollowedlessbecauseitwasdeemedappropriateandmorebecauseoftheconsequencesfortheofficeriftheydidnottoetheline.Inrelationtoonecaseofaverydrunkcouple,wheretheallegedperpetratorwasasleepupstairsandthevictimtoodrunktomakeastatement,theclinchingargumentinfavouroffollowingthearrestpolicymadebyoneofthetwoofficersattendingwasthat:

…if he commits further violence against her tonight then there will be serious repercussions. The positive arrest policy says we should take him in. It’s a load of bollocks, but if we don’t and someone else gets called back here tonight then we’re going to have some explaining to do. I don’t think it’ll get anywhere, but what else can we do? (Rowe 2007: 286)

InanothercaseRowerefersto,theperpetratorhadcalledtoreporttheincidentandthevictimwanted her partner to be prescribed the medication he needed rather than that he be arrested and charged. She was worried that processing him would delay the treatment he needed. The officerrespondedbysaying,‘Wedon’twanttoarresthim,butwehavenochoiceasitisforcepolicy.’(Rowe2007:290).Hesaid,however,he’dcheckwiththesergeant,whoseadvicewas,‘Youknowthepolicy,arresthim’(ibid),andthat’swhattheofficerdid.

Itisclearthatinparticularcasesofficersbelievedthatreduceddiscretionwasleadingthemtorespondinappropriatelytosomeoftheincidentstowhichtheyweresent.Theysawflawsinthefull enforcement policy even in relation to this particular incident type. In addition to the concerns aboutspecificcases,theyalsohadwiderworries.Inparticulartheyfearedthatanautomaticarrestpolicy might make victims more reluctant to call the police and be open to the kind of help they might provide.

40

Overall,thepoliceofficersRoweobservedwerealsounhappyatfettersontheirdiscretion.Itthreatened their professionalism in that they were no longer able to make their own judgements basedontheirexperience.Theyalsofeltitsuggestedthattheynolongerenjoyedthetrustoftheirseniorofficers.Theycouldnolongerdothejobastheysawfit.Asoneputit:

It’s ridiculous, years ago we were able to make a note in our pocketbook, give words of advice to people and send them on their way. The problem was resolved informally and every one was reasonably happy. Nowadays we have to complete a crime report, take a victim statement and witness statements, do all the paperwork and for what: no further action is taken and the outcome is the same as it would have been years ago, but it took five minutes in those days! (Rowe 2007: 291).

Officers’preferencesfordiscretiondonotmean,ofcourse,thatdiscretionisthebestpolicy.I’vealreadymentionedsomedownsides.Theargumentsbyofficersinitsfavourareclearlyself-serving.The tendency for those at the centre to try to control those at the periphery and for those at the peripherytowanttoavoidcontrolfromthecentreappearstobeubiquitous(Gouldner1959,1960).

Inthecaseofdomesticviolence,criticsofthepoliceresponsehavesuggestedthatuninformed,intuitive discretion might have led to routine practices that do not on balance produce the best outcomes.IndeeditwasthissuspicionthatledtothearrestpoliciesintheforceexaminedbyMichael Rowe. That suspicion had also led to some major studies in the United States of different ways of responding to misdemeanour domestic violence. I want to say a little about them because they may help us gauge better the research agenda that may best provide evidence to improve policing policy and practice.

Intheearly1980sintheUSthekindsofconcernthatledtotheinitiativeRowelookedatwasanimating movements for mandatory arrest policies. In the US there is more research money than thereisinthiscountryanditwasdecidedthatanexperimentshouldbeundertakentoassesstheeffectivenessofthearrestpolicyinstemmingrepeatincidents.ThefirsttrialwasundertakeninMinneapolis. Cases were randomly allocated to arrest or to other responses and rates of repeat incident were compared for each response to see whether the arrest response produced a lower rateofrepeatincidentthanothertreatmentsintheensuingsixmonths(Sherman&Berk1984).Discretionwasremovedfromtheofficersandtheyweretoldwhattodoinindividualcasesona randomised basis. What mattered for this trial was that the measures were implemented as requiredcasebycase.Thearrestresponseoutperformedtheotherresponses.Arrestseemedtowork. Moreover it was possible to gauge by how much it worked – that is to estimate the effect size. Based on police records 10 per cent of those arrested committed repeat violence over the followingsixmonthscomparedto19percentforthoseadvisedand24percentforthosewherethe suspect was simply sent away. Victim reports produced different detailed results but much thesamepicturewithfiguresof19percent,37percentand33percentrespectivelyforrepeatviolenceforarrest,adviceandsendingthesuspectedperpetratorsaway.

41

PartlyonthebasisofthisfindingmandatoryarrestpoliciesinrelationtodomesticviolencebecamethenormintheUnitedStates.LaterstudiesreplicatingtheMinneapolisexperiment,however,cameoutwithdifferentfindings(Sherman1992).Arrestdidnotoutperformotherresponses;indeeditseemedsometimestomakemattersworse.EvenifthefindingsiftheoriginalMinneapolis-StPaulstudywerevalidforthatcity,theycouldnotsafelybegeneralised:inotherwordstheylackedwhatsocialresearcherscall‘externalvalidity’eveniftheyachieved‘internalvalidity’.Thereasonthefirststudynecessarilylackedexternalvalidityisthattherandomallocationwas(asithadtobe)fromaspecificpopulation.Itcouldnotbeofallpopulationsinallplacesatalltimes.Randomallocationstovaryingtreatmentsfromdifferentpopulationsproduceddifferentoutcomepatterns.Ifwefindthatarrest has an effect but we cannot be certain in any community what the direction of the effect will be,weclearlyhaveproblemsinlesson-learning!

Thisisitselfanimportantlessonforpoliceofficersandresearcherstryingtobuildanevidencebaseforpractice.Iwant,though,heretoraiseafurtherproblem.Thebasisfortheconclusionabout the effectiveness or otherwise of the arrest policy in Minneapolis was based on comparing pre-ordainedresponsestoincidents.Theofficersinthecourseofthetrialweredirectedtotheresponse they gave on a random basis. They not only had no choice but to arrest if that was the responsetowhichtheywereallocated,theyhadnochoicebuttofollowtheotherresponsesincluded in the studies. The comparison was of one mandated response against another. It was not ofastandard,mandatedresponseagainstdiscretionthatmightsometimesinvolvearrestbutmayalso,case-by-case,involveotherresponsesalso.Whatitcouldinformwasamandatedstandardresponsenottheexerciseofinformeddiscretion.

Thepost-hocexplanationforthevariationsinresponsetoarrestinthedifferentcitieswheretrialswereheldwasthatwherearresttendedtowork,thecommunitieswerestableandemploymentlevels were high. Arrest and its prospect tended to activate shame that reduced the probability that perpetrators would repeat. Where arrest tended not to work the communities were less stable and levels of unemployment were higher. Here arrest and its prospect tended less to activate shameandmoretoactivateanger,whichincreasedtheprobabilitythatperpetratorswouldrepeat.

Thesemechanismsdonotexhaustthepossibilities,assomeofthecommentsofRowe’ssubjectsindicate,andasoutlinedinTable1.Table1derivesfromsomeseminarswheremandatoryarrestwas discussed. They comprise testable theories about how arrest policies may pan out for those in differentcircumstancesbecauseofthewaytheyaffectthedecision-making,bothbythosewhomayact violently and those who may report violence against them.

42

Table 1. Domestic Violence Arrest Possibilities by context

* These hypotheses clearly relate to domestic violence where a man is the perpetrator and a woman the victim

Source: Tilley (2000: 106)

ThewayRowe’ssubjectstalkedaboutthecasestheyattended,wheretheywouldhavepreferrednottofollowthearrestpolicies,indicatedtheirsensitivitytotheparticularitiesofcasesandthewaysinwhichthesamepracticemightproducedifferentintendedandunintendedconsequencesaccordingtotheparticularcircumstances.Theaveragedfindingsforeachofthecitieswherethetrialsofmandatoryarrestmay,ofcourse,encompasssomeindividualcaseswherethearrestresponse increased the risk of repeat incidents as well as others where it reduced it and still others

Mechanism Context Data to text expected outcome pattern

1. Women’s shame Membership of ‘respectable’ knowing community

Reduced level of reporting of incidents amongst those with close attachments to communities valuing traditional family life

2. Women’s fear of recrimination

Historyofviolence;culturallysupportedviolence;alcoholism of offender

Reduced levels of reporting incidents amongst chronically victimized

3. Women’s fear of loss of partner

Emotionalorfinancialdependency on partner

Reduced level of reporting amongst poorer and emotionally weaker women

4. Women’s fear of children being taken into care

Pattern of general domestic violence against whole family

Reduced level of reporting amongst families known to social services

5.Women’sempowerment

Availabilityofrefuges;supportforwomen;financialresources of women

Increased levels of separation where support and alternative living arrangements available

6. Incapacitation of offender

Length of time held Short-term reductions in repeat incidents

7. Offender shame Membership of ‘respectable’ knowing community

Reduced domestic violence within ‘respectable’ communities

8.Offenderanger Cultural acceptability of maleviolencetowomen;what man has to lose from brushes with the law

Increased levels of violence amongst those violence-sanctioning communities marginal to mainstream society

9. Offender shock Offender attachment to partner;self-imageaslaw-abiding respectable person

Reducedlevelsofviolence,andhelp-seeking behaviour amongst short-tempered ‘respectable’ men

10. Changed norms about propriety of domestic violence

Positive publicity Reduced levels of reported and unreported domestic violence

43

where there was no effect either way. The trials measure net effects where some positive and other negativeeffectsmaybecancelledout.Indeedanyfindingof‘noeffects’maydenoteeithernoeffectsatalloracancellingoutofpositiveandnegativeeffects,asTable2shows.

Table 2: Sample ways in which the same net effects can be produced

Thevariabilityineffectscan,ofcourse,beafunctionnotonlyofdifferencesinthecases,intermsof the reactions of those subject to the measures applied. They may also result from different ways inwhichthe‘same’measureisdelivered.Policeofficersdonotalldothesamethinginthesameway.Inadequaciesinimplementationarewellknownandformostprogrammes,notjustthoseinvolving the police.

Demonstration projects in particular tend to outperform successors that are rolled out after them because of the greater enthusiasm and dedication of those involved in the original initiative. Anotherwell-documentedexamplerelatingtoaratherdifferentapproachtorepeatdomesticviolenceillustratesthis.AprojectinKillingbeck,WestYorkshire,addressedrepeatdomesticviolencethroughstandardisedincreasesinintensityandenforcementwithsuccessiveincidents,includingdiscretion that responses could always be ratcheted up earlier if the particular case suggested thatthiswasneeded.Theprocess,however,couldneverbereversed.Thedefaultpositionwastoscaleup;therewasnodiscretiontoscaledown.TheKillingbeckprojectwasdeemedasuccessanditwasdecidedtorollitout(Hanmeretal1999).Whatwasrolledout,however,wasapalereflectionoftheoriginal.Itlackeditsdedicatedleadership,energyandauthority(Hanmer2003).Non-discretionaryscalingupwasnotconsistentlymaintained.JalnaHanmer,whowasinvolvedinstudyingboththeoriginalKillingbeckprojectanditsrollingout,wasaghastatthediscretiontodepartfromtheKillingbeckmodelthatemerged.Forher,clearlymandatedresponsesdefiningthelimits to discretion in accordance with the theory behind the Killingbeck project was crucial.

Cases Positive effect

Negative effect

No effect Net effect

Small positive effect 1 100 30 20 70 10%

Small positive effect 2 100 55 45 0 10%

No effect 1 100 50 50 0 0%

No effect 2 100 0 0 100 0%

Small negative effect 1 100 20 30 70 -10%

Small negative effect 2 100 45 55 0 -10%

Large positive effect 1 100 50 0 50 50%

Large positive effect 2 100 75 25 0 50%

Large negative effect 1 100 0 50 50 -50%

Large negative effect 2 100 25 75 0 -50%

44

Conclusions for improving policing and research aimed at improving policingI want now to draw the threads of the preceding discussion together tentatively to suggest an agenda for improving policing and for research on policing that aims to inform improvements in it.

Letusbrieflyindicatewherewehavegotsofar :1. Policediscretionisinevitable;2. Unfetteredpolicediscretionisdangerous;3. Closelyprescribedpoliceactionisliabletoproducepredictableunintendedharms;4. Assessment methods have tended to focus on measuring the net effects of standardised

policeresponses,ratherthandiscretionaryones.

Points2and3tellusthatrulesanddiscretioncanbothproducepoorpolicing.AsFigure1indicates,unsurprisinglypoordiscretionandpoorrulescancreatepoorpolicing.Improvementscould be created by either improving discretion or by improving rules. Indeed strongly prescribed rule-bound behaviour could lead to improvements in the face of either poor discretion or behaviourthatisdictatedbypoorrules.Similarly,wiseinformeddiscretioncouldproduceanimprovement on both poor discretion and responses dictated by poor rules. If we go for wise and informeddiscretion,howisitbesttobeproducedinpoliceofficers?Ifwegoforrules,howarewebest to determine what the rules should be and how do we best secure behaviour in accordance with them? If improvements in policing overall are to be achieved by having some behaviours thatarediscretionaryandsomethataremandatory,howdowedecidewhichiswhich?Perhapsimprovements could be effected by establishing routines for evidence-based normal best practice asabackclothfortheexerciseofdiscretionwhereitisneeded?

Figure 1: Discretion, discipline and improvement strategies

Problem Diagnosis Improved strategy

Poor performance

Poor rules

Better rules

Poor discretion

Rules

Better discretion

Discretion

45

Theseseemtometobefundamentalquestionsforanypoliceservicebentonimprovementandanyresearchoutfitintentonservinganimprovementagenda.Theanswerswillbeparticulartospecificissuesinpolicingandwillrequireempiricalresearch.I’vefocusedondomesticviolenceinthislecturebutwouldexpect,mutatismutandis,thatsomewhatsimilarissueswouldariseinanyarea.Whatwefoundisthatdealingwithdomesticviolenceiscomplexandachievingsustainedandconsistentimprovementsinoutcomebychangedpolicepracticesisfraughtwithdifficulties.Inrelationtobetterrulesorbetterdiscretionasthebetterroutestoimprovement,thefollowinggeneral principles seem to me to be important starting points:

1. Forsomepurposesruleshavetobeexplicitandtobefollowed.Thoserelatingtothecollection,non-contamination,continuityin,storageofandanalysisofphysicalevidencecomprisesignificantexamples.Thisisnottosaythattheseareentirelymechanicalmatters.Itis,however,tosaythatthereareratherstrictrulesthathavetobefollowed.Pastfailuresin relation to the collection and use of physical evidence suggest that there is a substantial researchagendaonhowbesttosecurecompliance(HMIC2000,2001;TilleyandFord1996).

2. Forsomepurposesitseemslikelythatdiscretionwillbeimportant,forexampleinrelationtomethodsofdealingwithso-called‘wickedissues’:thoseforwhichthereisnoexisting,standard,triedandprovengeneralresponse.Problem-orientedpolicingismostlyrelevanttopolice-relevantcommunityproblemsofthiskind,whichpersistdespitestandardresponses.Inproblem-orientedpolicingdiscretioniscentral,althoughtheactionstakenarenotarbitrarybutfollow analysis and at best apply well-grounded theory. The research and practice agenda here hastodowiththeidentificationandanalysisofhistoricalandemergingproblempatternsandthe assessment of the success with which the tentatively applied theory is applied. The purpose of the improvement-orientated research here is in part to see whether the problem has been successfullyaddressed.Itisalso,however,toprovidefindingsthatenrichtheknowledge-baseofpolicy-makers and practitioners better to use their discretion in dealing with future problems. Oneofthereasonswhycrimeproblems,inparticular,continuouslyevolvehastodowiththeadaptive behaviour of offenders (Ekblom 1997). They adjust their behaviour in relation to those trying to thwart them. They seek out new crime opportunities. They also make use of new toolsforcommittingcrime.Incontrasttothebody,whichbasicallycontinuestofunctioninthesamewayandonlyslowlyandpassivelytoadapttotreatment,insociallifehumansbehavestrategically. They can and do adapt actively and deliberately. The subject-matter of policing is in this sense more malleable than that of medicine and hence what was relevant and effective yesterdaymayceasetobesotomorrow.Itisforthisreasonthatinformed,discretionaryproblem-solving policing has more need of tested general theory than is the case in physical medicine,wherethebody’sreactionsaremorepredictable,althoughevenherevariationsin the ways in which different people respond to the same medication suggests a need for greaterattentiontodiversitythanmayhavebeenthecaseinthepast.Thenewishfieldofpharmacogeneticsisconcernedwithjustthis(Wolfetal2008).

46

3. Having stressed the need for discretion and for a knowledge-base of tested theory to inform itsuse,ithastoberecognisedthatthepoliceoftenhavetoactinconditionsofsubstantialuncertainty. In busy circumstances rules of thumb are inevitably used. But there are better and worserulesofthumb.Identifying,testingandimprovingpoliceofficerrulesofthumbastheydeal with recurrent issues comprises an important part of an agenda for improving policing. The critics of police failures to take domestic violence incidents seriously by using as their ruleofthumbthattheywererubbishworknottobetakenseriously,werealmostcertainlyright.Understandinghowrulesofthumbaredeveloped,whattheyshouldbe,andhowtheycanbechangedwheretheyarefoundwanting,arekeyresearchissuesforimprovement.Oncedecided,ratherthantryingtoimposeuniformbehaviours,itmightbemoresensibletospecifydefaultresponses,departuresfromwhichshouldbereportedandexplained,andtofindwaysofproducingthedefaultbehaviour.Interestingly,althoughoftenthefirstportsofcall,thestrategymaynotalwaysbebestachievedbyeducationandenforcement.Inthecaseofhandhygiene,forexample,therewasaconsensusthathand-washingshouldbethedefaultalthough failures are routine and responsible for many hospital-caused infections. One of the bestexamplesofamethodofimprovingtherateofdefaulthad-hygieneinvolvedrelocatinghand-washingfacilities.Puttingthemattheentrancetosinglewards,wherethedoctorchecksthenotesforthepatient,evidentlyleadstoamuchhigherlevelsofcompliancetothepreferreddefaulthand-washingpracticesthanputtingthematpatient’sbed-head,wherethedoctorhasto walk by the patient to wash his or her hands in the event that they decide that some kind of physicalexaminationisneeded.(IowethisexampletoKathyKirklandofDartmouthCollege.Moregenerallyonsubtlewaysofestablishingdefaultbehaviours,seeThaler&Sunstein2008).There are also strong arguments for and some evidence of the usefulness of well-formulated guidelinesandchecklistsinsecuringgoodpractice(SeeKelling2009;Tilley2006;Gawande2010;Sidebottometal2012).

4. Much work by the police aimed at detecting crime takes place in conditions of considerable uncertainty. There appear to be two ways of achieving relatively high detection rates for volumecrime,suchasburglaryandcartheft(Burrowsetal2005,Tilleyetal2007).Oneinvolves the development of stronger routine (low discretion) procedures where cases are passedfromspecialistunittospecialistunitforprocessing,theothertheuseof(higher-discretion)judgement,wherecasesareseenthroughtocompletionbyasinglepersonorteam. Neither the specialist procedural nor the generalist discretionary arrangements can guaranteehighdetectionrates.Itdependsonhowwelltheyaredelivered,andeithermaybemoreappropriateaccordingtocontext.Themorediscretionarymayworkbestwherethereisastablestaffgroup,includingexperiencedofficers,whodealwithalowvolumeofcases.Themoreproceduralmayworkbestwithrelativelyinexperiencedofficerswithahighturnoverandalargevolumeofcases.Inbothinstancesstrongsupervisionisneeded,intheonecaseto inform sensible use of discretion and in the other to ensure compliance with standardised procedures.Iftheseconclusionsarecorrect,theysuggestthatinthecomplexworldofpolicing,theremaybedifferentmeanstoachievepreferredends,involvingdifferentpatternsofdiscretionordiscipline,andthatthechoiceofwhichwillneedtobecontextuallydependent.

47

Inconclusion,IhopeIhavesaidenoughtopersuadeyou;(a)thatissuesofdiscretionanddisciplineineffectingimprovementsinpolicingareimportant;and(b)thathighlevelsofdiscretionwillalwaysbeneededforsomeareasofpolicing;(c)thatdiscretionneedstobeasinformedaspossible;(d)that the effectiveness of policing may be improved by developing more informed default routine behaviours,alongthegenerallinesshowninFigure2;and(e)thatinculcatingpreferreddefaultactionsisitselfacomplexmatter.Theseconclusionssuggesttomeanimportantsharedresearchagenda for universities and police services.

Figure 2: Discretion, discipline and improvement strategies and embedded improvement

Problem DiagnosisImproved strategy

Embedded improvement

Poor performance

Poor rules

Poor discretion

Rules

Better rules

Better discretion

DiscretionSupervised:

Guidelines

Checklists

Defaults

48

ReferencesBurrowsJ,HopkinsM,RobinsonA,SpeedM,TilleyN.2005.Understanding the Attrition Process in Volume Crime Investigation.HomeOfficeResearchStudy295.London:HomeOffice.

Crawford A. 2009. ‘Governing through Anti-social Behaviour: Regulatory Challenges in Criminal Justice’. British Journal of Criminology.49(6):819-831.

Ekblom P. 1997. ‘Gearing up against crime: a dynamic framework to help designers keep up with the adaptive criminal in a changing world.’ International Journal of Risk, Security and Crime Prevention. 2 (4): 249-65.

Gawande A. 2010. The Checklist Manifesto.London:ProfileBooks.

Goldstein H. 1963. ‘Police discretion: the ideal versus the real’. Public Administration Review 23(3): 140-148.

GoldsteinH.1967.‘Administrativeproblemsincontrolingtheexerciseofpoliceauthority’.The Journal of Criminal Law, Criminology, and Police Science.58(2):160-72.

Goldstein H. 1977. Policing in a Free Society.Cambridge,Massachussetts:Ballinger.

Goldstein H. 1990. Problem-Oriented Policing. New York: McGraw-Hill.

Goldstein J. 1960. ‘Police discretion not to invoke the criminal process: low-visibility decisions in the administration of justice’. The Yale Law Journal.69(4):543-594.

GouldnerA.1959.‘Reciprocityandautonomyinfunctionaltheory’.In:L.Gross(Ed.).Symposium on Sociological Theory.Evanston,IL:RowPeterson.

Gouldner A. 1960. ‘The norm of reciprocity: a preliminary statement’. American Sociological Review. 25:161-78.

Hanmer J. 2003. ‘Mainstreaming solutions to major problems: reducing repeat domestic violence.’ In: K. Bullock and N. Tilley (Eds). Crime Reduction and Problem-Oriented Policing.Cullompton,Devon:Willan.

HanmerJ,GriffithsSandJerwoodD.1999.Arresting Evidence: Domestic Violence and Repeat Victimisation.PolicingResearchSeriesPaper104.HomeOffice;London.

Her Majesty’s Inspectorate of Constabulary (HMIC). 2000. Under the Microscope. London: HMIC.

Her Majesty’s Inspectorate of Constabulary (HMIC). 2002. Under the Microscope Refocused. London: HMIC.

49

Kelling G. 1999. ‘Broken Windows’ and Police Discretion. National Institute of Justice Research Report. WashingtonDC:USDepartmentofJusticeOfficeofJusticePrograms.

KenneyC.2008.TheBestPractice:HowtheNew Quality Movement is Transforming Medicine. New York: Public Affairs.

LipskyM.1980.Street-Level Bureaucracy.NewYork:RussellSageFoundation.

MillardB,FlatleyJ.2010.Experimental Statistics on Victimisation of Children aged 10-15: Findings from the British Crime Survey for the year ending December 2009.HomeOfficeStatisticalBulletin11/10.London:HomeOffice.

Pawson R. 2006. Evidence-Based Policy: A Realist Perspective. London: Sage.

PawsonR,TilleyN.1997.Realistic Evaluation,London:Sage.

RoweM.2007.‘Renderingvisibletheinvisible:policediscretion,professionalismanddecision-making’. Policing and Society. 17(3): 279-294.

SemmelweisI.1983(1861).Etiology, Concept and Prophylaxis of Childbed Fever. Madison: University of Wisconsin Press.

ShermanL,BerkR.1984.The Minneapolis Domestic Violence Experiment. Washington DC: PoliceFoundation.

Sherman L. 1992. Policing Domestic Violence.NewYork:FreePress.

SidebottomA,TilleyN,EckJ.2012.‘Towardscheckliststoreducecommonsourcesofproblem-solving failure’. Policing. 6(2): 194-209

ThalerR,SunsteinC.2008.Nudge: Improving Decisions about Heath, Welfare and Happiness. New Haven: Yale University Press.

TilleyN.2000.‘Doingrealisticevaluationofcriminaljustice’.In:V.Jupp,P.DaviesandP.Francis(Eds).Criminology in the Field: the Practice of Criminological Research. London: Sage.

Tilley N. 2006. ‘Knowing and doing: guidance and good practice in crime prevention’. In: R. Clarke and J. Knuttson (Eds). Putting Theory to Work. Crime Prevention Studies Vol. 20.Monsey,NY:CriminalJusticePress:217-252.

50

TilleyN.2009.‘What’sthe‘what’in“whatworks?”?Health,policingandcrimeprevention’.In:J.Knutsson and N.Tilley (Eds.) Crime Prevention Studies Vol. 24.Monsey,NY:CriminalJusticePressandCullompton,Devon:Willan:121-145.

TilleyN,FordA.1996.Forensic Science and Crime Investigation. Crime Prevention and Detection Series paper 73.London:HomeOffice.

TilleyN,RobinsonA,BurrowsJ.2007.‘Theinvestigationofhighvolumecrime’.In:T.Newburn,T.Williamson and A. Wright (Eds). Handbook of Crime Investigation.Cullompton,Devon:Willan:226-254.

WolfC,SmithG,SmithR.2008.‘Pharmacogenetics’.British Medical Journal 320:987-990.

51

Observations from a Decade on the Inside: Policing Cultures and Evidence Based Policing6

Professor Betsy StankoAssistantDirector,CorporateDevelopment,MetropolitanPoliceService,UKandRoyalHolloway,UniversityofLondon

BiographyProfessorBetsyStankoisAssistantDirector,CorporateDevelopment,MetropolitanPoliceService(MPS).Previously,shewasaProfessorofCriminology,teachingandresearchingatClarkUniversity(USA),BrunelUniversity,CambridgeUniversityandRoyalHolloway,UniversityofLondon(whereshe remains a visiting Professor). She has published over 70 books and articles over her academic career.ThemostcitedoftheseworksisIntimateIntrusions:Women’sExperiencesofMaleViolence,publishedin1985.ShehasbeenawardedanumberoflifetimeachievementawardsfromtheAmericanSocietyofCriminology,mostnotablytheVollmerAward(1996),recognisingoutstandinginfluenceofheracademicworkoncriminaljusticepractice.From1997-2002shewastheDirectoroftheESRCViolenceResearchProgramme.In2002,shejoinedtheCabinetOffice,inthePrimeMinister’sOfficeofPublicServicesReform,andinSeptember2003,shejoinedtheMPSasafulltime employee. She is a fellow of the Royal Society of the Arts.

What I would like to do today is to use three particular observations about evidence based policingthatIhavelearnedwhileworkinginsidethepoliceserviceoverthepastdecade,mapthoseontosomecontemporarydiscussionsincriminology,andbechallengingabouttheslownesstochange.Oneofthethingsacademicsaskmeallthetimeis,‘whydon’ttheytakemyresearchseriously?’ ‘Why can’t we get police to use research in a different kind of way?’ The last ten years ofmycareerhasbeenspentintryingtofindawaytomakeresearchrelevanttocontemporarypolicing,andtoactionchangeinaverylargepolicedepartment.Mypoliceserviceemploysover55,000people.Thatnumberwillreduce,infouryears’time–wearesubject,likeeveryoneelse,tomajorcuts–andneedtoloseabout£600mwithinthenextfouryears.Whatimpactthishasondelivering policing to London remains to be seen. I am convinced that it is possible to mitigate the impact by embracing the use of evidence based practice.

Ibelievetheimportantthingaboutefficiencyisworkingbetterthroughresearchknowledge.Thereisalotofdebatetodayabout‘whatworks’,whatis‘evidence’,andhowdoesitenablepolicingtodosomethingdifferently?Inmyjobwespendalotoftimetalkingabouttactics,andsometimeswe talk about strategic tactics.WhatIdo,however,isstrategy;tryingtogetasenseofwhatthe‘bigpicture’ is and how a ‘big picture’ mission would enable policing to work better. I will be arguing

6 Held at Scottish Police College, Tulliallan Castle, 20 October 2011

52

todaythatwhilethere’salotofdebateabout‘whatworks’andenergyexpendedlookingatwhatanofficerdoeswhenheorsheleavesthestation;dotheyturnleftorright?Thereislessenergyongetting police to harness ‘what works’ as a routine. I want to push the boundaries of ‘what works’ into the areas of theory and concept. My ambition while working inside the police service is not to justifyaparticulartactic,buttobegintochangehearts,mindsandbehaviouraboutknowinghowtochoose a tactic and why.

Firstlyinthistalk,Iwillusemyworkaroundconfidencetotalkaboutadifferentwayofthinking;necessarybecausetherearesomanychangesinpolicingthatrequireabetterunderstandingofthepublic.Totakeoneexample,inLondon,aMayor’sOfficeofPolicingandCrime(MOPAC)hasbeencreatedandbeginsinJanuary2012,andthatwillbeamajorchange.

The governance of policing is about to change in England,becomingmorelocalandinScotlandwiththecreationasinglepoliceservicewith17,000policeofficers.Itissometimesverydifficultwhenworking in a big space for change to cascade down learning to the front line. Under the UK coalition governmentconfidenceinpolicingisnolongerthesingle,overarchingtarget.Butitwouldbewrongtoforgetaboutconfidence,becauseitiscriticaltoput the public at the heart of policing in order to ensure its legitimacy. With three days of disorder in London this August we know what happens when itbreaksdown(Stanko&Dawson2013).

Secondly,Iwouldalsoliketogiveyousomeinsightfrom seven years’ longitudinal research offering a conceptualinsighttothepolicingofrape,acrimetype where throughout the UK and the world the justiceoutcomeisextremelychallenged.Overthe

past eight years I have been asking ‘what do the police know about rape from the reports they receive?’Westillneedtothinkquitedifferentlyabouthowthatcrimetypehappens.Itispossibletothinkdifferentlyaboutacrimetype;inthe30yearsthatIhavebeeninthisprofessiontherehasbeenashiftinthewaythepublicthinksaboutdomesticviolence,thewaywepoliceitandlegislatearoundit;suchashiftinthinkinghasnotbeenreached,Ihavetosay,inthewaywethinkaboutrape.

Finally,Iwillconsiderhowthetwoexamplesabovegiveussomeinsightintopoliceimprovement,andwhat I will term a ‘neo-tribal resistance’ to both outsider and insider evidence which impedes our use of research based evidence in policing. It’s been a long sustained approach to get the Met to think differentlyabouttheinformationtheyhave,soeverythingI’mgoingtobetalkingaboutisourowninformation;it’snotsomebodyelse’s,it’sours,andthat’simportant.Thisisnot‘academic’information

Mayor of London, Boris Johnson (The Daily Mail)

53

whichisnotownedbyus,itisourinformationandthereforeourresponsibilitytotakeitseriouslyandunderstandexactlywhatpeoplearetellingus.It’snotalwaysthefastcarthatgetsthevillain;wehave to be reminded that we don’t need to do policing in the traditional way to be successful.

Overthepastdecadewehavehadafastandfuriousdebateabout‘whatworks’,andifcrimescience can help. Inside the Met I try to offer systematic use of knowledge and to suggest that therearewaysofbeingsystematicaboutthestrategytoreducecrime.Weknowforexample,thatpolicingofproblemplacesmakesadifference;it’stakenmetenyearstohavethatadoptedas a strategic point of view and we are now policing strategically by identifying the worst places in Londonandfocusingdifferentlyuponthem.NewYorkhasusedimpactzonesforthelast15years,with20outof75precinctsbeingdesignatedimpactzones.Fifteenofthose20haveneverchangedoverthattime;butthat’sOK,becausethestrategyofNewYorkisthattheyputmoreresourcesintothose20mostchallengedzones,includingalloftheirnewrecruits,thanintotheother55;because they are the accepted as the worst areas. People spend a lot of time saying that crime datainhighcrimeareasisuseless;it’sonlyapartialreportofwhathappensoutthere.Absolutelytrue;Ihavenodoubtthatpeoplewholiveinhighcrimeareasreportlessthanpeoplewholiveinlowcrimeareas;iftheyactuallyreportedeverythingthattheyexperienced,thosehighcrimeareaswouldjustbehigher.Observing624wardsinLondon,Iknowwherethetop20highcrimeareasare,Iknowwhichstreetsremain‘highcrimestreets’;there’sapersistenceandconsistenceinthewaythatrecordedcrimesstackintoandontoplacesthatweneedtopayattentionto(Braga&Weisburd 2010).

WhenSaferNeighbourhoodscommunitypolicingwasintroducedin2004inLondon,oneofthethingsIwantedtodoforthethennewCommissioner,IanBlair,wastoprovidehimwiththemostrobustinformationpossible,sothatwhenhestoodupinpublicheknewpreciselywhatwasgoingon.Idesignedaquestionnaire,ahybridpublicattitudesurveythatwasn’tquiteliketheBritishCrimeSurvey;itwasmuchmoreaboutpolicingitself.Thereforesince2005wehavealongterm tracker that enables us to take a look at people’s opinions about the police. In 2004 we startedwith8,000surveyinterviews,we’vebeenashighas20,000,andgiventhecurrentbudgetconstraintswe’renowat12,000interviewsayear ;3,000peopleeveryquarterareaskedhowthey feel about policing in London. The bit that’s important to understand about me is that I was a qualitativeresearcherwhenIstarted;Iamnowrollingaroundinmoredatathananyonehaseverhadintheirentirelife–800,000crimesayearreports,12.5millioncallsfromthepublic–wecanuseallthedatathattheMethas,analyseandstatisticallyunderstandit.ThissurveydataisnowarchivedattheUniversityofEssex.

Working with three academics from the Methodology Institute of the London School of Economics –JonathanJackson,BenBradford,andKatrinHohl–webegantoplaywiththedata(StankoB,JacksonJ,BradfordB,HohlK2012).Whatthisworkhasdeliveredisamanagementboardaccepted,signedoff,approved,has-the-T-shirtmodelofpublicconfidenceinLondon,createdbyBenandJon.Thirty-twoboroughsareperformance-managedbasedonthismodel,andonaquarterlybasistheyget information about how well they are doing in the eyes of the public. The thing that police like

54

themost–effectivenessindealingwithcrime–isveryimportant;howdowesupportvictims,howdowetackleandpreventcrime.Butwefindthathowwellweengagewiththepublic,listentoandunderstandthekindsofconcernsthattheyhaveisthekeydriverofpublicconfidence.Wealsoask:arepolicereliableifyoucallthem,howwelldopolicedealwithantisocialbehaviour,andhowfairlydopeoplefeeltheyaretreated.Fairnessisabsolutelycriticalintermsofthinkingaboutpublicengagement and legitimacy of policing (Myhill and Quinton 2011). These are the main drivers of people’s opinions of how good a job police are doing locally.

Thismodelisakindof‘nudgetheory’whichsomeofyoumayknowalittleabout;whatyouaretryingtodoisgetpeople,i.e.policeofficers,tochangetheirbehaviour.ItisimportanttounderstandthatthemaindriverofpublicconfidenceinthepoliceinLondonisengagementwiththecommunity;secondtothatisfairtreatment;third,iseffectivenessindealingwithcrime.Whenyoutalktopoliceofficers,thefirstthingtheysayis‘wehavegottobeeffectiveinthewaywedealwithcrime;thatisthewaypeoplejudgeus’.However,mostpeopledon’tdealwiththepolicedirectly,theythereforejudgeyoubyhowwellyoureachouttothemandiftheythinkyoutreatpeople fairly.

Figure 1: Contributors to confidence in policing.

WhenIstartedthisworkbackin2005Ihadconcernswiththeconceptof‘fearofcrime’,andtheaimofthisworkwaspartlytochallengetheBlairgovernment’sobsessionwiththatconcept.Fearofcrimeisadriver,butnotamajordriver,ofpublicconfidenceinpolicing;aswehaveseen,themaindriverofpublicconfidenceinpolicingishowwellpeopleunderstandthepoliceasengagingwith,listeningto,andunderstandingtheconcernsofthecommunity.Ifyoudon’thavecontactwithpolice,thenhowdoyoumeasurethat?NotthroughtheDaily Mail.

Perception of how well the police are:• responding to emergencies,• tackling and preventing crime,• supporting victims and witnesses,• providing a visible presence and• policing public events Views that the police are:• commited to and engaged with the community,• by listening, understanding and dealing with their concerns

People feel that the police are:• fair and treat them with respect, regardless of who they are or the reason for contact

Effectiveness in dealing with crime

Information about local policing activities shapes public opinion

Engagement with the community

Fair treatment

Alleviating Local ASB

People percieve that there has been a decrease in ocal disorder and ASB, e.g. noisy/nuisance neighbours, teenagers hanging around, drinking in the street and vandalism

Confidence in Policing

55

Figure 2: Drivers of public confidence in policing, from the MPS survey, 2005 – 2012.

It’sadifficultthing,butinLondongettinginformationdirectfromthepolicewascriticaltopublicconfidence,intheformoflocalfeedbackfrompoliceonaneighbourhoodbasistoresidentsthroughtheirletterboxes(Wünsch&Hohl2009).Wehave630SaferNeighbourhoodTeamsinLondonwhoprovidedinformationvianewsletters;werecommendthattheinformationprovidedtoresidentsgroupsimplyaroundthreeareas:‘thisiswhatyousaid’arepriorities,thisiswhatwe’vedoneaboutit,andthisishowyoucontactus’.Wenowcanshow,statisticallyandthroughanexperiment,thatifyouproactivelyfeedbacktopeople,publicconfidenceincreases(Hohl,BradfordandStanko2009).Becauseresidentsknowwhatyouknow,andyouknowwhatresidentsknow,and if they have anything else to tell you then residents know how to contact you – that’s basically the structure of the newsletter.

Bynomeanswasthisgoodnewstomypoliceofficers.Theydidn’tlikethefactthatwewereaskingthepublicdirectlyaboutpolicing,particularlythosewhodidn’thaveanycontactwithus.Itwasn’tpeopleyouknew,butarandomselectionofaweightedrepresentationofthepopulation.Thismodelistestedeveryquarterwithover3000interviews,andsofarthestrengthofthemodelremains.

So the lesson here is how do we think very differently about the police’s engagement with people? Becauseengagementwithpeopleisvery,veryimportant;notjustengagementwiththe‘usualsuspects’,nottheoneswhocometoyourmeetings,buteverybody.Andit’sarealchallenge.Howyoureachouttothosepeoplewhodon’tnecessarilycometothosemeetings,andparticularlythehardtoreach?Wesuggestthatletterboxcommunicationisagreatequaliserintermsofengagement.

Worrying about crime

Alleviating local ASB

Effectivenessin dealingwith crime

Fairtreatment

Engagement with the

community

56

OneofthethingswelearnedfromafiveyeartrackerofSaferNeighbourhoodsisthat,interestinglyenough,peoplebegintoconvergearoundwhatthekindsofcrimethataremostimportantintheirlocalareas.Whatwefoundisthatyoucanaskalotofdifferentpeoplealotofdifferentquestionsaboutcrime;somepeopleknowaboutseriousandorganisedcrime,theylivenexttoit,itaffectstheirlives.It’sverylocal,veryspecific.Theyunderstandthedifferencebetweenvolumeandviolentcrime;again,someevidenceabouthowitexistsverylocally;andtheyunderstandincivilities.Butwhenyougetpeopletogetherwetendtofindthatopinionsconvergenotaroundwhetheritshouldbeabouttrafficordog‘mess’–andyoucanhaveendlessdiscussionsaboutthat–butwhether or not there’s a serious problem in the neighbourhood. People tend to be more altruistic aroundseriouscrimethanaroundthepettystuff.ThatfindingcomesoutverystronglywhenwelookedateightwardsinLondonoverfiveyearsandinterviewedasampleof400peopleonayearly basis.

Generallythepublicdocareaboutcrime,andtheydohaveopinions,butthoseopinionsaredrivenbywhathappenstofamilyandfriends.Theymaynotbethedirectvictimsofcrime,theymaynotliveinhighcrimeareas,butifanybodytheyknowhashadanycontactwiththepolicethispersonalcontactwithotherpeopleinformsthewaytheyanswerthequestionnaire.Itisn’tjustfingerintheair ;it’swhattheyhearaboutpolicingfromotherpeople,oftenfamily,friendsandneighbours.Sohearing about police contact is just as important as having contact.

ThefiveyeartrackertoldusthattheMet’sownleafletisthepreferredformoffindingout aboutlocalpolicing.Wedidanumberoffocusgroupsaboutwhatformthatwouldactuallytake,sotheleafletitselfisevidencebased.Peopletakethatleafletasadirectformofcontact,anditispositive. We also know that fair and respectful treatment – and I will keep coming back to fair and respectfultreatment–predictswhetherthepublicviewthepoliceaslegitimate.Publicconfidencecanbeinfluenced.

ButIhavelearnedtonever,everunderestimatethevalueofthelegitimacyofpolicing.WhatweknowfromthisworkandfromotherworkbyBenBradford,JonJacksonandKatrinHohlisthatissues around public trust are absolutely critical to positive policing and justice outcomes. The lessonforthisisthatweneedtothinkcontinuouslyonhowtomaximisepubliccooperation,andthis is perhaps where we have not done very well. The issue of fair treatment is critical in terms of cooperation in policing. We know that detections often arise from information provided by victims andwitnesses;soyouwanttokeepthepeopleonside.Weknowthatthosewhobelieveinstatelegitimacyaremorewillingtocomplywiththelaw,andaforthcomingpieceofworkshowsthatpeople are less likely and less willing to act outside the law if they believe that the forces of the state are legitimate (Jackson et al 2012).

57

Figure 3: Confidence in Local and London-wide Policing (MPS PAS 09/10-11/12 Rolling 12 month data)

SomethinginterestinghashappenedinLondon,asyoucanseeinFigure3.WhenwestartedSaferNeighbourhoods,ifyouaskedpeoplegenericallyaboutpolicinginLondon,yougotamorefavourable response compared to local policing – what I call the ‘Scotland Yard Effect’ – seen in pink here.Thathaschanged;there’ssomethinggoingonabouttheScotlandYardbrandthatisdiminishingintheeyesofthepublic.Itmayonlybeareductionby5%,butbecausealongstandingpatternhasbeguntochange,itisaninterestingfinding.MygoalwhenwestartedSaferNeighbourhoodswastoincreasepeople’sconfidenceinlocalpolicingtothelevelofLondonpolicing;itseemsifithasjustsurpassedthat.Thatissignificant,andwe’rekickingoffanumberofpiecesofworkaroundwhatthatmeans for the London Metropolitan Police service as the ‘corporation’.

InthelasttenyearsIhavebeenintheMet,therehasbeenacompletetransformationintechnology–thereareforinstance20,000imagesonYouTubeofpeopleintheriots.Therearepagesandpagesand pages of images on the Metropolitan Police website of the people they are trying to arrest fordisorder.Ifyoulookattheimagerythatthepublicthemselvesshare;thisimageryisinterestingfor our observations about legitimacy and the state. Technology itself has become an instrument ofdemocracyandtransparency,andyouwouldnotbesurprisedifIsuggestthatimageryaroundthe police and policing of the riots was particularly strong. I was in the United States at the time of thedisorderandwatcheditontelevision.Forthoseofyouwhohaven’tbeentoLondon,itwasalocaliseddisorderaffectingafewpostcodes.Itwasafirework,butlocalised.Lotsofpeoplecameandwent,travelledaboutLondonusingtheir‘Oystercards’touchinginandtouchingout.TheywenttoworkonMondayjustafterthedisorderbegan,andtheywenttoworkonTuesdayabitmorenervouslybuttheystillwenttowork.OneofthecentralquestionsweaskedintheMPSiswhydidthe riots take place? Why did they happen? Many of the reports on the disorder found a common theme in the accounts of those who participated: feeling unfairly treated by police.

58

Metropolitan Police Service Officers during the London disorders, August 2011.

IshowtheaudiencethisleafletwhichwasfoundinTottenhamandputonYouTube.Theleafletadvisespeoplenottogotothepolice.It’simportanttoseethewayinwhichdifferentpresentationsofwhathappenedareportrayedbytheleafletauthors.‘Getahaircutandcolour,growabeard,wearglasses…DOkeepyourhouseclean.Getridofspraycans,demorelatedstuff,dodgytexts/photosonyourphone.Don’tmakelifeeasyforthem[thepolice]byhavingdrugs,weaponsoranythingelseillegalinthehouse.’Theleafletacknowledgesanappearanceoflegitimacythroughquiteconventionalbehaviour.Evenriotersknowabouthowtogetamessageouttoalocalcommunity:leafletsthroughtheletterbox.

Drawingfromproceduraljustice,wecanlink the disorder to people’s willingness to comply with the law. The analysis of the Public Attitude Survey in Jackson et al. (2012) suggests that there is a link between policing andtheinfluenceofthestateitself.Previouslypoorlyunderstood,legitimacyitselfandtheway that people feel about policing intersects as a form of social control/compliance. Attitudes to violence are also a part of how people feel about legitimacy of the state. The more inclined people are to view the state aslegitimate,themorenegativetheirviewof

Leaflet found in Tottenham advising people against handing themselves into the police.

59

extra-legalviolenceis.FrominterviewswithpeopleinLondon,whatwe’refindingisactuallyquitesignificantconceptualchallengesinthewaypolicetreatpeople.

I now turn to the second observation. I am using victims’ allegations of rape as a victim-centred lookattheneedanduseforpolicing.We’velookedatrapeinLondonoverthelastsixyearsbycodingupeveryallegationofrapeintheMetinthefirsttwomonthsofeachfinancialyear,and this study is continuing. What is striking about those who report rape to the police is that anoverwhelmingmajorityofvictimsarevulnerable.In1985IpublishedabookcalledIntimateIntrusionsinwhichItalkedaboutvulnerabilityandrespectability;Ihadnoideahowstronglythedatainsideapoliceservicewouldreflectthatobservation.WhatwefindinthemajorityofcasesofreportedrapeinLondon–and92%ofthemarewomen–isthattheimageofrapeandthewaythat it is investigated assumed invulnerability. The approach to the investigation by the police is that the victim was so invulnerable that the violence she confronts was so strong that he overpowered her;previouslydescribedasthe‘realrape’victim(Estrich1987).

However,themajorityofthecasesofrapeinLondon,alargecityinwhichyoumayexpectahighernumber of stranger rapes – and I would suggest that this is probably true across the country and theworld–arethatin75%ofrapesthevictimknowstheoffender.(ThemostrecentUSNationalInstituteofJusticereportonfemalevictimsofsexualviolencebythewayshowsthat78%ofsexualviolenceinvolvesanoffenderwhowasafamilymember,intimatepartner,friendoracquaintance.)Thisisthen‘evidence’andrequiresthatwethinkdifferentlyabouthowrapeisinvestigated.Twenty-fiveyearsagoIwouldneverhaveguessedthatthepercentageofknownrapistswouldbeashighinreportstothepolice.Themajorvulnerabilitieswithinwhichrapestakeplaceandare,importantly,reportedtothepolice,areyouth(victimisunder18atthetimeoftheassault),mentalhealth,alcohol and drug misuse prior to the rape and domestic violence.

The vast majority of those who report rape to the MPS then can be considered to be vulnerable insomewayorother(87%in2005;84%in2010).

• Onethirdvictimswhoreportrapeareagedbelow18atthetimeoftheoffence (33%in2005;34%in2010)

• Oneinsixofvictimswhoreportrapehaveamentalhealthissue(18%in2005;14%in2010)• Onethirdofvictimswhoreportrapehaveconsumedalcohol/drugspriortotherape

(35%in2005;35%in2010)• Onequarterofvictimswhoreportrapeareorhavebeeninanintimaterelationshipwith

theirassailant(24%in2005;26%in2010)

Overtime,lookingatsixyears’ofdata,youcanseethattheoutcomeoftheallegationshasnotmuchchanged.Wenowhavethelargestraperesponseunitintheworld,andtherehavebeenchangestopolicepractice.IfyoulookatthekeyattritionpointsinFigure4,theseshowthedifferences over time:

60

• Moreallegationsareclassifiedascrime(67%in2005;91%in2010);notingachangeintheinitial contact and the recording of the crime of rape

• Moresuspectsareidentified(53%in2005;72%in2010)• Moresuspectsarearrested(39%in2005;53%in2010)• Fewersuspectsarecharged(44%ofallarrestedsuspectsin2005;26%ofallarrestedsuspects

in 2010) • Oncecharged,moresuspectsareconvicted(31%in2005;44%in2010)

Theoveralloutcomeisthatin2005,5%ofallallegationsleadtoaconviction.In2010,itis6%.

Figure 4: Attrition rates by percentage of all allegations, MPS 2005 – 2010.

Whenwethinkaboutevidencechangingpoliceoutcomes,progresscanbereally,reallyslow.Someof the outcomes may not be easily altered when the view of rape still heavily relies on imagining ‘realrape’asthemainproblem.Ibelievethesefindingsrequireustothinkaboutwhetherwearefocusing the problem in investigation in the right place. I suggest that the problem is in ‘evidencing’ howanoffenderexploitedvulnerability.Sowhyisthissodifficulttoseeorexplaininsidethepoliceservice? (The Jimmy Saville investigation makes this point even more strongly than I ever could. BS, 2013).

I summarise through some observations about police culture and its relationship to changing policingthroughevidencebasedfindings/research.Iamanon-uniformed‘Professor’andIhavebeenabletoimprovepoliceanalysisofourowninformationfromtheinside.Butmyexperienceregardinganalysisandevidencebothforconfidenceandrapeisthatsomepeople(thatis,policeleadersandfrontlineofficers)getit(thelinkbetweentheresearchandchangingwhatonedoes),andsomepeopledon’t.Myexperienceisthattheresearchobservationshaveabetterchanceofbeing successful when you have a senior uniformed sponsor using the analysis in decision making. I canputonthebestshowpossibleandwithouthavingsomeonesay‘yes,we’regoingtodoitthatway’–withouthavingmyconfidencemodelapprovedbythemanagementboard,putthroughtheperformance regime and printed on the t-shirt – nobody would understand it. It is still a struggle todefinefromapolicingpointofview‘whatyoudodifferentlyinpolicing’asaconsequenceofknowingwhatdrivespublicconfidence.

61

Policescholarsandevenfictionwritershavewrittenandresearchedintopolicecultureandhowuniformedofficersbehave,thinkandact.Thestudiedbehavioursofthepolicehaveachievedakindof‘folklore’;forexample:theuseofforce,coercion,racism,sexism,anti-intellectualism,corruption,biasedlawenforcement,secrecy,exclusivity,loyalty,brotherhood,guardedtrust,competitive,authoritarian,entitled,bullying,abusive,goodhumoured,brave,helpful,reliable,selflesshavebeenpart of studies of police culture.

Thereisanotherpartofpoliceculture,andthatisthat‘onlyuniformscanmakedecisions’.EventhoughIamininsiderintheperformanceregime,ithastakenanextraordinaryamountoftimetohave new cutting edge academic research and police information practically used inside my own policeservice.Thisismychallengetoyou:thinkaboutverifiedevidence,lookatwhatoutsidersdo,itofferssomeinterestingwaysofseeingtheworld.It’snotjustabouttacticalevidence,helpingyoutoturnrightorleftonyourwayoutofthestation.Usingresearchrequiresachangeinthewayofficersthinkabouttheworld.

Letmeofferacoupleofpiecesofadvice.Usingconceptualevidencerequiresshiftingtheway‘wedothingshere’,andconceptualisingthingsdifferently–suchaslocatingrapevictimsalonga continuum of vulnerability – may also challenge the way investigation is routinely approached. Evidenceisnottheonlythewayoneadaptsto‘turningrightorleftoutofthestation’,theevidencepresentedhererequireschangingthewayone‘sees’anddoesthings,firstlybyacceptingandincorporatingevidencewhichhasbeenverifiedbythoseoutsidetheservice,asacorepartofimprovement.ThetwothingsIhavetalkedabouttoday–confidenceinpolicingandrape–dorequirechangesinthewayyouseesomethingandthewayyoudosomething;itdoesrequirechanging behaviour.

Intermsofconfidence,Iwouldsuggestthatthechangerequiresmovingbeyondtheuniformtounderstandpublicneed.Iregularlyencounterpolicehavingdifficultyinunderstandingwhatthepublic ‘want’. We still need active translation and transition. We’re still as a police service not very goodatdealingwithallofthepublic.Wedealwith‘theusualsuspects’,wedealwithpoliticians,butordinarypeopleonthestreetandtryingtounderstandwhattheywantandrequire?Iamstillencounter a ‘uniformed’ view of the world as opposed to a ‘public’ view of the world. I think it’s reallyimportant(thoughdifficult)tothinkandfeellikeacitizen,becausethenyouunderstandtheimportanceofthejobintermsofservingthestateandthepeople.Ifirmlybelieveindemocraticpolicing.PeopleintheMiddleEastaredyingfordemocracy;andweneedtounderstandthatpeople are dying for democratic and legitimate state power.

Intermsofthinkingaboutrape,weneedtorethinkhowthereportingofrapeanditsinvestigationcouldfeatureondocumentingtheexploitationofvulnerability,asthepolicingofdomesticviolence is now being seen through the lens of ‘power and control’. We need to think about what wouldchangeifwewrappedtheinvestigationaroundtheoffender’sexploitationasopposedto invulnerability of the victim. I believe this changes the entire approach to investigation. I can guaranteeitwon’talwaysleadtoasuccessfulprosecution,becausethecasesandrapeallegations

62

thatIhavereadaremessy.Astheevidenceconsistentlytellsus,theseallegationsarisebetweenpeoplewhoknoweachother,theyareallegationswherepeoplewithmentalhealthissuesareexploited–ofcoursethey’redifficulttomanageinacriminaljusticesystem!Buthowelsearewe–inthisadvanceddemocraticstate–goingchallengesexualexploitation?Sexualviolencehappenssooftenasexploitation.Thedominantimageofraperemainsofthebushlurkeroverwhelminginvulnerability. Actually what we are talking about – what the data overwhelmingly tell us – is that wearedealingwithcrimethatisaboutanintentionalexploitationofvulnerability.

Yet,strangebuttrue,acrosscontinents,acrosstheworld,I’mstillbridging,translating,challengingwithhumour and respect and some impatience – and optimism. I hope that you join me in the journey to make evidence accessible and used by the police service today.

Thank You.

63

BibliographyBraga AA and Weisburd DL. 2010. Policing Problem Places: Crime Hot Spots and Effective Prevention. Oxford:OxfordUniversityPress.

EstrichS.1987.Real Rape.Cambridge,Massachusetts:HarvardUniversityPress.

HohlK,BradfordBandStankoEA.2009.‘InfluencingTrustandConfidenceintheLondonMetropolitanPolice:Resultsfromanexperimenttestingtheeffectsofleafletdropsonpublicopinion’ British Journal of Criminology.50(3):491-513.

JacksonJ,BradfordB,HohlK,Farrall,S.2009.‘Doesthefearofcrimeerodepublicconfidenceinpolicing?’ Policing: A Journal of Policy and Practice. 3 (1): 100-111. DOI: 10.1093/police/pan079

JacksonJ,BradfordB,HoughM,MyhillA,QuintonP,TylerTR.2012.‘Whydopeoplecomplywiththelaw?Legitimacyandtheinfluenceoflegalinstitutions’.British Journal of Criminology.52(6):1051–1071. doi:10.1093/bjc/azs032.

MyhillA,QuintonP.2011.It’safaircop?Policelegitimacy,publiccooperation,andcrimereduction:An interpretative evidence commentary.London:NationalPolicingImprovementAgency(NPIA),September 2011.

MyhillA,BradfordB.2012.‘Canpoliceenhancepublicconfidencebyimprovingqualityofservice?Results from two surveys in England and Wales’. Policing and Society: An International Journal of Research and Policy.22(4):397-425.DOI:10.1080/10439463.2011.641551

MyhillA,QuintonP,BradfordB,PooleA,SimsG.2011.‘Itdependswhatyoumeanby‘confident’:operationalizingmeasuresofpublicconfidenceandtheroleofperformanceindicators’.Policing: A Journal of Policy and Practice.5(2):114-124.http://policing.oxfordjournals.org/content/5/2/114.short

StankoB,JacksonJ,BradfordB,HohlK.2012.‘Agoldenthread,apresenceamongstuniforms,andagooddealofdata:studyingpublicconfidenceintheLondonMetropolitanPolice’.Policing and Society: An International Journal of Research and Policy. 22(3): 317-331. http://www.tandfonline.com/doi/abs/10.1080/10439463.2012.671825

StankoB,DawsonP.2013.‘Reflectionsontheoffendinghistoriesofthosearrestedduringthedisorder. Policing: A Journal of Policy and Practice.7(1):3-11.doi:10.1093/police/pas054

StankoEA.1985.Intimate Intrusions – Women’s Experience of Male Violence. New York: Methuen.

WünschD,HohlK.2009.‘Evidencinga‘goodpracticemodel’ofpolicecommunication:Theimpactoflocalpolicingnewslettersonpublicconfidence.’Policing: A Journal of Policy and Practice. 3(4): 331-339.

6May2008.‘BorisJohnsondeclares‘truce’withpoliceasheputscrimeattheheartofmayoralmanifesto. The Daily Mail.http://www.dailymail.co.uk/news/article-1018290/Boris-Johnson-declares-truce-police-puts-crime-heart-mayoral-manifesto.html#ixzz2MkTwSSwt

64

The New Police Professionalism – Democracy, Evidence Based Practice and a 21st Century Profession7

Peter Neyroud CBE QPMProfessionalResearchConsultantandProjectDirector, CrimeHarmExperiment,JerryLeeCentreofExperimentalCriminology,CambridgeUniversity

BiographyPeterNeyroudisaformerChiefConstableofThamesValleyPolice,ChiefExecutivefortheNationalPolicingImprovementAgency(NPIA),whereheledthe£3bnAirwaveprogrammetoprovidepoliceradios,includingtheinstallationofradiointheLondonundergroundandfortheOlympics and the IMPACT programme to create the Police National Database. He is also a former Vice-PresidentoftheAssociationofChiefPoliceOfficersandawidelypublishedauthorofbooks,articlesandpapersonpolicingandEditoroftheOxfordJournalofPolicing.Hislastmajorpieceofworkinpolicingwasa‘FundamentalReviewofPoliceLeadershipandTraining’fortheHomeSecretary,publishedinApril2011.SinceleavingthePoliceServiceattheendof2010,Peterhascreatedaprofessionalresearchconsultancy,iscompletingaPhDanddirectingamajorresearchprogrammeatCambridgeUniversityintoCrimeHarm,andisprovidingtraining,leadershipdevelopment and consultancy nationally and internationally.

I am very pleased to be here and feel very privileged to be addressing you in this stunning venue. I amgoingtotryandcoverquitealotofground,becauseIdon’tgetaudienceslikethisveryoften;andyouareaveryimpressiveaudience,agreatcross-sectionofthosetowhomthistopicmatters,beingpoliceofficers,thoseinterestedintheirgovernance,andproperperformance.Soholdontoyour hats – or not as the case may be.

Iwanttostartbytalkingalittlebitaboutthe‘oldprofessionalism’;theonethatispassingaway,butwhichhasn’tquitegoneyet;andfromthe‘old’tothe‘new’,now12yearsintothe21stcentury.Iparticularly want to demonstrate the evidence that dismantles the myths that most of us – and I wouldsayallofus–thinkwork.Perhapsunsurprisingly,Ithenwanttotalkalittlebitaboutpoliceprofessionalismandpolitics,andhowwemightwemight,inthemidstofallthischange,developa‘new professionalism’ for the future.

7 Held at the Playfair Library, University of Edinburgh, 25 October 2012

65

Tobeginbygoingbackalittle;mostofyouwon’tknowAugustVollmer(1876-1955),butheisworthlookingup;hewasChiefofPoliceinBerkeley,Californiabetween1909and1932.Vollmerwasanunusualcharacter,particularlyintheearly part of the twentieth century because he was seriously interested in professionalizing policing at a time when it most definitelywasnot.Itwashethatfirsttoputradiosinpolicecars,itwashe–andbearinmindthiswasin1920–thatfocused on recruiting women and African-Americans into policing,becausehefeltitmatteredinbuildingrelationshipsbetween the police and the community. He professionalised crimeinvestigation,andlastly,andprobablymostimportantly,headvocatedcollegeeducationforcops,atatimewhenthatwasdefinitelynotthewaythattheywererecruited.EarlytwentiethcenturypoliceofficersintheUnitedStateswere in most cases recruited because of who they voted for andwhotheyknew;notwhattheyknewandwhattheydidwhen they got on the ground.

Vollmerleftalegacy–anditisimportantforChiefstoleavealegacy–intheformofadisciple,O.W.Wilson(1900-1972),whowaschiefinWichitaandChicago.ItwasWilsoninthe1950swhodescribedthe‘oldprofessionalism’;aschemaofpolicingbuiltaroundrapidresponsetocallsanddeterrentpatrol;theomnipresentcop,whowouldreducecrime,andreactiveresponsetocrimeprevention.It’sthe‘oldprofessionalism’thatalotofusstilldo,andIwillgoontowhythatmightbeamistake.HewasthefounderofoneofthefirstpolicesciencedegreesatWichitaUniversity,andnotonlyapoliceofficer,butthroughouthiscareera‘clinical’Professor.

August Vollmer (1876-1955)

66

TheBritishapproachisacontrastingtradition,oronemightsay,theEnglishapproach–andIshallbecarefulto distinguish where appropriate between the English andScottishapproaches.InEngland,whereapoliceservice was deemed to be in need of reform you got a directentrysoldier,AirMarshallLordTrenchard(1873-1956);MetropolitanPoliceCommissioner1931to1935.He founded the Hendon Police College to educate an‘officerclass’whocould‘control’thepoliceofficers,who after the police strike of 1919 were deemed to be slightly suspect. Trenchard’s approach to cure and create professionalism among the police was to bring ‘chaps’ in fromthebestpublicschools;andhavingbeentooneofthose,WinchesterCollege,Icanimaginejustthesortofchaps whom he brought in. The response to this initiative was interesting in that it wasn’t just the police service thatdislikedthisapproach;itwasalsounpopularwiththeright-wing press. There was the most astonishing selection of cartoons in papers like the Daily Mail lampooning Trenchard’s‘genstables’;onedepictsaconstableservingteatoaseniorofficerwithasilvertray,thetitle‘Butler’across his lapel. It is a remarkable sign of the cross-over betweenprofessionalismandtheexpectationofpolicing.Thestatusofthepoliceofficer,theirknowledgeofsociety,andthewayinwhichpoliceinfluencesocietyandareinfluencedbyithasmanyresonancesacrossalongperiod;Icouldhaveshownyouupto50politicalcartoonsonthistypeofsubjectgoingbacktothe1930s;itisremarkablehow strong this theme is.

AnAmericanacademic,SteveHerbertarguedinanarticle(Herbert2006),drawnfromhisresearchintocommunitypolicinginSeattle,thattherelationshipbetweenpoliceprofessionalsindemocracycentresaroundthreeconcepts:subservience,separationand‘generativity’.Subservienceheseesasanessentialpartofwhatthepoliceshouldbe;theyshouldbeservantsofdemocracy,andtheirveryresponsestothepublic’scallsaretestamenttothat.Therequirementthattheyconsultandlisten,thattheyareheldtoaccountbyavarietyofmechanisms,thatthereisnosuchthinginpolicingan‘independentprofessionalism’,it’snotthetypeofprofessionalismthatyoucanself-govern. He also distinguished two aspects of separation – looking across primarily American policing but it undoubtedly translates across the Atlantic – that we also need to be independent underthelaw.Thatmattersnotjustintheindependenceofpoliceofficers,butalsointhelegitimacyofthepoliceservice;thatdecisionsthatthepolicemakeareseentobeindependent,fair,andnotunderothersubjectiveinfluences.Alsothere’sanaspectwhichisprobablymoreproblematic,and

Air Marshall Lord Trenchard (1873-1956)

67

thatisthatthepolicethemselveswanttobeculturallyseparate;it’seasiertoexerciseauthorityinsomerespectswhenyoufeelcomfortableandremotethanwhenyouareproximateandpartofthe community.

AccordingtoHerbert,it’snotgoodenoughtohavejustabalanceofsubservienceandseparation,youalsohavetohavewhathedescribesas‘generativity’;thepolicenotonlyhave,butshouldhavesomeinfluenceonsociety.Forexample,thatthepoliceactuallystructurethecitizens’understandingofcrimebythewaysinwhichthepolicerespondtoit,andbythewaysinwhichthepolicerecordit;thepolicealsoprojectanincrediblyimportantmoralimageinthewayinwhichtheygoaboutthings.Totakeanimportantrecentexample,thecaringimageoftheDyfedPowyspoliceandthewaythattheyhandledthelittlemissinggirl(thecaseoffive-year-oldAprilJonesabductedfromMachynllethinWaleson1October2012)wasquiteextraordinary.I’veneverheardsuchpositivefeedbackfromcrimejournalists,whoaregenerallynotthemostcompassionatebunch;theyweredeeplymovedinthewayinwhichtheforceandsurroundingforcesacted,demonstratingacarefrom the community with the police at the forefront.

Nowbacktotheoldprofessionalism,wherewasthebalancebetweensubservienceandseparation? In Vollmer’s case it was about cutting free from the political corruption of the nineteenthcentury;inO.W.Wilson’scaseitwasaroundanoperationalstrategy,butunfortunatelyO.W.Wilson’sstrategydidn’tstackup.Duringthefirstwaveofresearchintothepoliceitwasestablished that random patrol did not produce the kind of effective deterrence that Wilson had argueditwould;rapidresponsedidnotleadtomorecrimesprevented;andreactivedetectionproduced little more than information from the public.

Soweenteredtheeraof‘nothingworks’attheendofthe1970s;thiswasthepointwhenIjoinedthepoliceservice.IntheUSArandompatrolbecameassociatedwithracialprofiling,producingacrisisofconfidencebetweenthepoliceserviceandthepublic,particularlyamongstAfrican-Americans;justthinkbacktotheRodneyKingbeating,andtheriotsthatfollowed.IntheUKtherewasthewatershedoftheConfaitcase(themurderofMaxwellConfaitin1972)whichraisedquestionsaroundthetreatmentofvulnerablesuspectsandtheemphasisondetectionbasedonwhatprovedtobefalseconfession.ThatcaseledtotheRoyalCommissiononCriminalProcedure,leadinginturntothePoliceandCriminalEvidenceAct(PACE)1984.ThesubsequentBrixtonriotsalsoproducedaPublicInquiry(Scarman1981)thatwascriticalinUKpolicinginforcingustoreconsiderthebalancebetweencrimefightingandordermaintenance.Fundamentaldistinctionsbetweenwhatarethepoliceheretodo;areweherejusttodetectcrimes,oraretherewholeranges of things to be held in balance?

Oneresponse,outofasearchforamission,whennothingworkedoutofthosethree‘R’s,wascommunity policing. It was a response partly plucked from Herman Goldstein’s ground breaking book,Problem-OrientedPolicing,publishedin1990,althoughGoldsteinhadbeentalkingaboutaproblem-solvingmodelthroughoutmostofthe1980s.Hisworkinproblem-solvinghadevolvedfrom his earlier concerns about the way in which police use discretion. The community policing

68

modelwasgivenahugeboostinthe1980swhenthefirstHarvardsession,aconveningofAmerican,BritishandAustralianpolicechiefs,discussedthewatershedinpolicingandwheretogonext.Therewasahugedichotomyinthatdebatebetweenthe‘oldprofessional’DarylGates(ChiefoftheLosAngelesPoliceDepartmentfrom1978to1992)andLeeP.Brown,whohadbeenthechiefinHouston(thecity’sfirstAfrican-AmericanPoliceChief,1982to1990),andservedasthecommissionerinNewYork,wherehewasanadvocateofcommunitypolicing.Communitypolicingundoubtedlyemergedfromthosesessionsasthedefaultstandard,themodelthatmostforceswouldfollowoverthenext30years.Thereareveryfewpolicechiefsorpoliticiansthatwouldsaythat their force is not doing community policing.

Therewasasecondresponse,slightlylater,butneverthelessveryBritish,whichhasalsobecomeaworldmovement.Thereisnobody,atleastinmostpartsoftheworld,whowouldnotclaimtheyweredoingintelligence-ledpolicing,becauseyouwouldn’twanttobeunintelligent.ItwasadistinctivelyBritish contribution which evolved – and I was involved in the early development of it – as a response to concerns that our pursuit of detection had led to some very serious miscarriages of justice. Police werealsowerespendingfartoomuchtimetryingtodetectcrimesbyavarietyofexcitingmeans,such as pitching up at major prisons and getting people to confess to crimes that which they had not necessarilycommitted.Intelligence-ledpolicingwasalsoanattempttomanageadeficitineffectiveness;in contrast with community policing which was an attempt to ‘do’ the frontline role better.

The alternative approach to improving the management of policing began with William J. ‘Bill’ Bratton’s(PoliceCommissionerinNewYorkCity,1994to1996)developmentofCompstatinNew York at the turn of the 1990s. The critical thing about this is in some ways not so much the mechanicsofCompstat,butthefactthatwhatunderpinsitistherediscoveryofpoliceconfidencethat they couldfightcrime;inmarkedcontrasttothe1980swhentherewasacomprehensivebelief that the police could not. Very few police chiefs in America or this side of the Atlantic stood upinthe1980sandsaidthattheycouldreducecrime,becausetheydidn’thavealotofevidencetosupportthatfact.ThereweretwobigchangesthatcontributedtoBillBratton’ssuccess,oneofwhichwastheopportunitytodealwith‘bigdata’.Incontrasttothe1970sand1980s,henowhaddataonwherecrimewastakingplacethathecouldusetoputpoliceofficersintherightplace.Thesecondbigchangewastheemergenceof‘evidencebasedtactics’,thefirstbeing‘copsonthedots’–hotspotpolicingwasn’tquiteassophisticatedasitisnow,butneverthelessitwasacriticalcomponent.Compstat,determinedleadership,andachangeoftactics,combinedwithotherfactors,producedaverysignificantshift;there’sabigdebatestillabouthowfarthepoliceaffectedthatbutI don’t think there’s any doubt that they did affect it and they continue to affect it.

Andthenwecometothemostrecentdevelopment,evidencebasedpolicing,whichunderliesthe30-yearperiodfrom‘nothingworks’,toknowing‘whatworks’.Firstly,ifyoufocuspolicinginhighcrimeplaces;secondly,ifyoufocusontacklingthemostharmfulandpersistentoffenders;thirdly,ifyoufocusonprotectingvulnerablevictims;andfourthlyonincreasingthelegitimacyoftheservice,youwillhaveaverysignificantimpactoncrime.ThesearewhatIcall‘thefourfacesofevidencebasedpolicing’,supportedbyconsistent,credibleevidencefromover30systematicreviewsonthe

69

CampbellCrimeandJusticewebsite.Yetdespite25years’ofresearchintothepolice,politiciansandthepublicarestillthinkingaboutpolicingandcrimebasedonintuitionratherthanresearchedfacts,the ‘I’m a police chief and these things work’ argument.

Totakesomeexamplesofthese‘myths’andtoanswerthemwiththe‘facts’fromthosesystematicreviews:Thefirstmythisthat‘morepoliceonpatrolmeanlesscrimesolet’sget“morebobbiesonthebeat”’.Wouldthatitworkedlikedthat–wewouldallhavesolvedtheproblemsandgonehome.Fact:morefocusedpatrolofhotspotsandtargetedproblem-solvingwillmostdefinitelydelivercrimereductions,butthereisaconsiderableamountoffinesse;it’sthesameaskeyholesurgery–youdon’tjustdoit,thereisaconsiderableamountofpracticeandartrequiredtogetitright. The same goes with the increasing evidence that targeted problem-solving will deliver an even greater reduction and probably a more sustainable one.

Asachiefofficer,theresponseyouwillgetwhenyousay‘I’mgoingtodohotspotpolicing’is:‘butcrimewilljust“movearoundthecorner”’.No;andIwouldguessifIdidastrawpollofthisaudience,andifIaskedyouthatquestionbefore,that90%ofthisaudiencewouldhaveagreedwiththatstatement,andthat90%wouldbewrong.Farfromdisplacement,moststudiesnowshowadiffusionofbenefitsfromtargetedhotspots,thatcrimedoesn’tgoroundthecorner.Yougetabenefitaroundthehotspots,becauseactuallycrimehappensinthehotspotsbecausetherearefactorsofbothpeopleandplacetherewhichcontributetoit.So,thereforetargethotspots.

Thethirdmythisthatmostofusfeelthatprosecutingoffendersincourtisaneffectivedeterrent;andmostofuswouldbewrong.Formalprocessingofoffenders–andtherehasbeenstudy,afterstudy,afterstudy,includingawholeseriesofrandomizedtrials–thattendstosuggestanegativeimpactonoffending,particularlyamongyoungoffenders.Notleastofwhich,howmanyofyouractuallyoffencesactuallyendupincourt?It’sabout2%to3%ofthetotaloffencesthatarecommittedinyourcommunities,soifyouthinkyouaregoingtorelyoncourtstoreducecrime,forgetit;it’snotgoingtobeeffective.‘Tough’approachesshouldonlybeusedcarefullyinordertotarget the most persistent and harmful of offenders.

Anothermythisthatcommunitypolicingpreventscrime,butthereisverylittleevidencetosupportthis,andthereisasystematicreviewthatgoesintothatevidenceinsomedetail(Mazerolleetal2012).However,communitypolicingdoeshaveasignificantimpactonpolicelegitimacy,andlegitimacy matters. Increasingly we know that legitimacy is a crucial component in helping people tokeepthelaw;soinsteadoffocusingonthosewhobreakit,focusonstrengtheningcommunityefficacyandonthewayinwhichpeoplekeepthelawanddon’trespondwithhostility.

OurfinalmythisonewhichI’vehadthroughoutmycareerfrommycolleagues,andthatisthattraditionalmethodsworkbest.Inburglarydetection,themostrecentrandomisedcontrolledtrialsshowthatthemosteffectivewayofidentifyingsuspectsisthroughDNA;moreeffectivethanevenfingerprints,letalone‘traditionaldetectivework’suchasgoinghousetohouseandarrestingandinterviewingsuspects,whichisvirtuallyuselessinsolvingthecrime.

70

Why do these myths matter? I conducted an analysis of a cross-section of the campaign pledges of PoliceandCrimeCommissionersinEnglandandWales;andfoundarelianceon‘increasingarrestsandprosecutions’;pledgesto‘putmoreofficersonthebeat’,withonlylimitedreferencetopoliciesthataresupportedbyevidence.Itmattersthatwegetthisright,becauseotherwisewhatweareaskingthepoliceservicetodoissomethingthatdoesn’twork,andthenholdingthemtoaccountforsomethingthatdoesn’twork,bymeansandmethodsandmeasurementsthatdon’twork;whichisn’t very professional.

WheneverIteachpoliceofficers–andItaughtagroupofhighpotentialofficerslastweekinaUKforce–IgothroughsomeofthisevidenceandIaskthemifhavetheyreadit,andtheyhaven’t.ItisanextraordinarypartofpolicingthatweknowthatmostpoliceofficerscanrecitetoyoutheTheftAct,buttheydon’tknowhowtopreventtheactoftheft;itisnotpartoftheireducation.Thecurrentinfluencesonpolicingarethese:firstly,the‘clinicalexperiencemodel’,whichisbasedonjustcomingupto200years’worthofclinicalexperience;secondly,professionaltraditionsindifferentforces–andhereinEdinburghweareinoneoftheoldestpoliceforces,olderinfactthantheMetropolitanPolicewhichistraditionallythoughtofasbeingthefirst;thirdly,lawandbureaucracy,andfourthly,politics.

Thestandardsetofpolicetacticsarepreventivepatrol,arrest,prosecutionandpostreleasesupervision.EverywhereIgo–fromEdinburgh,toIndia,toAustralia–thesearethestandardtactics;andwhenIaskthequestion,‘sohowdotheywork?’Igetblanklooks.Ifyouwereembarkingonamedicaltreatment,youwouldwanttoknowhowitworked,thepossiblesideeffects,andifthedoctor–orinthiscase,thepoliceofficer–wasactuallyqualifiedtodeliverthetreatment.PersonallyI’mnotupforhavinganoperationbysomeonewithclinicalexperiencewhohasn’treadthescientificevidence.

Policeofficersoftendon’tevenknowtheactiveingredientsofdeterrence,buttheyare,accordingtoworkbyProfessorDanielS.Nagin(Durlauf&Nagin2013):severity,certaintyandcelerity.Eachoneoftheseingredientsmatters,andyouwouldthinkpoliceofficerswouldwanttoknowthis,andknowwhichoftheseelementswouldapply;youwouldthinkthismightbepartofthebasictrainingforpoliceofficers,muchmorethansubsection3.1oftheLicensingAct.RathermoreimportantthanthequestionsIfoundinthePoliceInspectionexaminationinEnglandWales,whichwas:‘youarestandingontheforecourtofalicensedpremisesonahotday,andthelicenseeoffersyouaglassofwater,whatoffencehastakenplace?’Towhichmyresponsewas‘Idon’tcare!’ItisanoffenceinEnglishlawforalicenseetoofferrefreshmenttoapoliceofficeronduty,butgoodnessme;I’veneverquitethoughtthatweweregoingtoenforcethatone.

71

Likeallmedicines,deterrencewillonlyworkifyoufollowtheinstructionsonthepacket;meaning,if you don’t implement it properly it won’t work;ifyoudon’tdiagnosetheconditioncorrectlyitwon’twork,andsomepatientswillexperiencesomeserioussideeffects.Forexample,ifyoufocusonlyon stop and search you are highly likely to end up with a high degree of disproportionate stop andsearchofminoritycommunities;itisaworldwidephenomenon,notlimitedtotheUK.Andwithsomechronicconditions,youmayneedtorepeatthetreatment.Soletmeillustratethisbytaking a problem like burglaries. The traditional response and methods of detection are typically floodingareaswithextrapatrolsandgeneralcampaignstoencouragecrimeprevention.Theyhavepermeatedsomuchintopolicingthatwecanseetheminthebureaucracyofthecrimereport;thesearetheboxesthateveryinvestigatingofficerfeelstheyhavetotick,becausetheyhavebeenthere–andIhavetrackedthembackinhistoricaldocuments–for150years.

Nowtheevidencebasedapproach;firstlypredictive prevention patrols – you want to prevent burglaryandwecandoso;butdoyouknowwheretheyaregoingtohappenratherthanwheretheyhavehappened?Andtheanswerisyoumightbesurprisedtoknowthatwitharound50%to60%accuracywecanpredictwhereburglariesarelikelytotakeplacebyusingstatisticaltechniques.If that is the case then surely we would be putting resources where they are about to take place? This is being tested at the moment in Birmingham.

The second evidence based approach is a focus on solvability,whichhasbeenaroundforawhilewithsomeworkbyJohnE.Eck(ProfessorofCriminalJusticeattheUniversityofCincinnati,USA),buthasneverreallybeenfollowedthrough,lookingatwhatfactorsleadtothesolvingofburglary;andbytheway,theyhavechangedhugelyin30years.WhenJohnfirstlookedatthis,youdidn’thaveDNA,evenfingerprinttechnologywaspapercopiesandlocaldatabases.Now,wehaveagreat opportunity and it would be a good idea to look at how and why we are successful.

Figure1showsapieceofanalysisusinga‘Forestplot’;thekeywithaForestplotistolooktotherightofthelineandthatwhichistotherightofthelineshowsthatitisofsignificanteffect,showingthatithasbeendetected.Itemsontheleft,whichinterestinglywashousetohousecompleted,showsyouthatitisnotonthewholeleadingtodetection.Lookingatthetwofurthesttotheright,oneisDNAwhichissignificant,andtheother–onewedon’tinvestenoughtimeandenergytowards–isfootwearmarksrecovered;footwearislikeafingerprintifusedproperlyand is remarkably effective at identifying perpetrators. You would like to know this if you were developingastrategytotackleburglary.Moreover,thispieceofresearchwascarriedoutbyaservingpoliceofficer.

72

Figure 1: Forest Plots – Burglary Dwelling Completed Offences (Paine, 2012)

IamcompletingapieceofresearchwithProfessorLarrySherman(Cosma,ShermanandNeyroud:forthcoming);we’vetaken100,000criminalrecordsfromthePNCwiththesupportofACPOandforthefirsttimeanalysedthemstatistically,inordertoidentifywhichoffendersandhowaccuratelywecanpredictthoseatahighriskofcommittingahighharmoffence–seriousassaults,serioussexualoffence,robbery,andarson.

73

Figure 2: High, Medium and low risk offenders (Cosma, Sherman & Neyroud: forthcoming)

Ourlatestdatashowsthat2%arehighharm,andthatpredictionismorethan80%reliable.Butifthat’sthecase,ifyoucanbemuchmoreaccuratethanyouusedtobebyusingverylargedatabases,logicallyyouwouldfocusmoreofyourresourcesonthehighharmgroup,andthinkabout strategies in the low harm group to free up time and focus. This would also apply to focusing around the likelihood of victims being victimized again and the strategies that can be used. This is a pieceofUKresearchgoingback20yearswhichwestillhaven’tusedsufficiently;wherewefocusonvictimsandareasthathavebeenburgled,‘targetharden’aroundvictimsinacocooningstrategy,and use of Neighbourhood Watch in high crime areas.

OperationTurningPointisapieceofmyownresearchinBirmingham,whereyoumightjusttakethatgroupoffirsttimeorlowharmoffendersandcombinedeterrenceanddesistancetreatment,taking them out of the system altogether. In which case given the legitimacy issues about just takingoffendersoutofthesystem,particularlyburglaryoffenders,whatwouldyoudoifyoudidn’t prosecute them? In this case you might decide to put them into some sort of deferred prosecution,forexampleusingcurfews,drugstreatmentandnon-association,whereyouholdprosecution over them and test whether you can encourage them in desistance. That is being testedatthemomentinarandomizedtrial,whichIthinkisthefirsttimeinthiscountrythatprosecutions have ever been randomized.

Sothat’showyoumightdealwithburglary;Ihaven’tgonethroughtheentireevidencebasedresourcelist,butthosearesomeofthetoolsthatareemergingfromtheresearchandbeingdevelopedandbeingtestednow.Andbytheway,thatpieceofworkonsolvabilitywasnotdonebyanacademicbutbyaservingchiefinspector,aspartofhisMaster’sprogramme.

High Risk (2.2%)

Neither High nor Low Risk (36%)

Low Risk (61%)

74

Itisperfectlypossibleforproperlytrainedpoliceofficerstousedata,inanewprofessionalmodel.We should be using evidence to test practice and practice innovation to generate research. We shouldberequiringallpractitionerstoqualifyandkeepqualifiedagainstthattypeofevidence;becauseifwedon’tknowit,wecan’tuseit,andifyoudon’tknowit,youareresistanttousingitbecauseyouarefrightenedofit.Weshouldrequiremanagersandleaderstobequalifiedtomanageandlead.Butwealsoneedsomecontinuityinthis;aprofessionalbodythatcanensurethatknowledgeismadeavailable,maintainedinacarefulway,continuallytestedanddisseminatedtotheprofessionandbeyond.Policeofficershaveahabitofleavingtheprofessionearly;wethereforeloseknowledgerapidly,withoutanymeansofprofessionalretention.

There have been some very important developmentsinthenewprofessionalism,andthatisthat we have had a huge increase in the number of Systematic Reviews from the Campbell CollaborationonCrimeandJustice(hostedbyNorway);thisinvolvestakingalloftheevidenceinaparticularsubjectanddoingameta-analysisacrossit,lookingforproblemsandeffects;whatworksacrossawholeseriesofstudies.Fiveyearsagotherewereonlyfourorfive;wenowhaveover35reviews.WealsohavethedevelopmentthisyearatGeorgeMasonUniversity(Virginia,USA)oftranslationtoolsliketheEvidenceBasedCrimeCenterMatrix,developedwiththeNPIA/NationalCollegeofPolicing,andPOLKA,UK;acomprehensivesetofcurrent‘highqualitystudies’whichhaveevaluated policing using a control versus treatment. There is a police online learning application which I was responsible for developing at NPIA – again it’s about trying to make the evidence systematicallyavailabletopoliceofficers.

But we also need a revolution in training.Wehavetrainedourselves,toourselvesinthesamewayforaroundabout180years;andthatsimply has to stop.Whatwehavetrainedfor180yearsisnotgoodenoughtosustainusforanother200years;policinghastoreinventitself,butaboveallithastoreinventitstraining,byturningNationalPoliceTrainingCollegesintoPoliceUniversities,andintegratingpolicetrainingwithhighereducation.Inthe‘AgeofAusterity’,whichIdoubtisgoingtogoawayanytimesoon,thepressureformorecost-effectivepolicingbecomesevermoreimportant.Weneedtoprovide,astheGeorgeMasonUniversityCrimeCenterMatrixdoes,atoolforpoliceofficerstousetoshowthemtheevidence.

75

Figure 3: George Mason University Crime Center Matrix

Ifyouwereanofficeraccessingthematrix,youcouldclickononeofthesedots,forexample‘individuals’ifyouwerefocusingonindividualoffenders,thenclickonthelefthandboxandseewhatworks.Andthatwouldstartatrailofevidenceastowhatthingswouldwork,andthesearesummarisedinawaythatanypoliceofficercouldaccess,understandandapply.Itgivesyouastarterforten,thesortofthingsthatmightworkinyourcircumstances.Wedostillneedprimaryresearch,andwealsoneedtoensurethattheevidencethatwehaveisavailableaswidely as possible. I think one of the things that SIPR is doing which is particularly important is thepractitioner–academiccrimeresearchwhichiscritical,andsharingthatiscritical.Notenoughresearchisbeingdoneinruralareas,andtherearealotofruralareasinScotland.IwasinNorwaylastweek,whichhasvastruralareas,andtheydon’thaveahotspotstrategyforthetundra!

InthecaseofNorway,theNorwegianPoliceCollegehasevolvedintoaPoliceUniversity,andisawardingitsfirstPhDs.Itisalsocrucialthatthereisanewrelationshipbetweenthepoliceandhighereducation;notthetypeofrelationshipthatLordTrenchardwouldhaveapprovedof,butagrownuprelationshipofthetypethatO.W.Wilsonalludedto,wherehighereducationandpolicingcometogetherandwherewetrainandeducate.Furthermore,wherewemoveawayfromtrainingabout policing which is all about ‘push’;wepushthetrainingintopeopleastheyarriveandtheyspendtherestoftheircareerswishingtohavetheirtraining‘pushed’intothem,anddon’ttake

76

responsibility for ‘pulling’ the training themselves. ‘Pull’ training is the type of training that Universities seektoencourage;theyseektopreparethepeopletheyareeducatingforalifewheretheycangoandfindthetrainingandtheevidence,notalifewheretheycanexpecttobespoonfed.

I’mafraidtosaythatthepoliceservicehasgottogrowup,notjustbecauseofthefinance,butbecause you cannot be professional in a ‘push’ model of education. That means that new entrants shouldbeproperlypreparedforarolewithapre-entryqualification,becauseunlessyouarepre-qualifiedwearespoon-feedingyou.WehavetogetoverournervousnessabouthavingLevel4foundationdegree,becausepolicingisacomplicatedprofession,andpeopleneedtobecompetentandpreparedforit.Similarly,managersshouldbequalifiedtomanagewithamanagementqualificationatLevel5,andIpaytributetotheScotsforhavingstarteddownthatroad;andseniormanagement should be prepared to lead and manage at strategic level with a Senior Management qualificationatLevel7;aMaster’sdegreeofanappropriateform.

Soit’snotatraditional‘policecollege’thatIproposesouthoftheorder,butaprofessionalbodywhich is not only responsible for delivery but for the educational and professional practice standards,rresearch,knowledgegeneration,disseminationandaccreditationofpractice.ItstartsonthefirstofDecember2012,andIamdelightedthatAlexMarshall,previouslythechiefconstableofHampshireConstabulary,willbethefirstchiefexecutiveofthenewNationalCollegeofPolicing.

We are at a point when O.W. Wilson’s vision of police and Universities coming together to deliverabettersocietyisessentialforpolicinginthe21stcentury,andSIPRhasplacedScotlandinthevanguard;youhavethebasisforsomethingreallyspecialinScotland,andthewayinwhichProfessorNickFyfeandhisteamandthechiefofficersinScotlandhavecometogetheronthisisdeeplyimpressive.Butthisisthepointatwhichtogofurther ;thenewScottishPoliceServicecanmaintain that lead by developing clinical–academic partnerships in teaching and research. If Tulliallan carries on being a place where people do drill and where training is old fashioned training then I’m afraidwe’reintrouble.Therehastobeamomentofdramaticshift,whereTulliallanbecomeseitherthe campus of a University or moves to become part of a Scottish University.

There is a tendency for education in leadership and management to be divorced from science. There is a lot of good evidence based management literature out there (and I don’t mean the kind ofthingyoufindinairportbookshops)whichistryingtoensurethatscientificmanagementtakesplace.AndhereinScotland,whereyouhavebeenhugelyinfluentialinthefieldofsciencethatshould be particular message is understood.

77

Figure 4: Evidence based decision making.

WhenItalkaboutevidence,Iamnotsayingthatyouleavebehindthemanagerialexperienceandprofessional knowledge that you have – of course you don’t – but you also need to have a clear understandingoftheorganisationalfactsfromyourowndata;there’snopointgoingdownaroutewhich is completely a result of your stakeholders’ values and concerns. Evidence doesn’t sit in the backroom;itsitsinasetofrequirementsthatlookabitlikeFigure4.

It does mean for those of you who are leading the ‘new profession’ of policing – and I mean that withregardnotjusttothosewhoareintheprofession,butthoseinvolvedinthegovernanceoftheprofession – that the greater focus is moving away from inputs and outputs. ‘How many have I got and how many detections have they produced?’ is not particularly interesting. What’s interesting is: ‘what did theyproduceintermsofoutcome?Howmanyfewercrimesarethere?Howconfidentisthepublic?’ Leaders therefore need to challenge practice with evidence and provide a vision to translate policeactivityintooutcomes;it’snotgoodenoughtosay‘crimewentup’,or‘crimewentdown’;youneedamuchmoresophisticatedapproach.Onealsothatpaysattentiontovaluesandethics;andifyouthinkoftheHillsboroughcase,failuretopayattentiontobasicvaluesandethicshashadacatastrophiceffectonthepublic’sconfidence;anditdoesn’tmatterifits20yearslater,theimpactis now. Leaders need to be transformational but also ‘authentic’ (Adolf Hitler was a transformational leader,buthecertainlywasn’tauthentic);andinternationalintheiroutlook.Policingisasmallworldinternationally,andthereisgoodresearchgoingoneverywhere,muchofitistranslatable.‘Notinmybackyard’isnotagoodexcusefornotlisteningtogoodpracticeproperlygeneratedelsewhere.

Professionals,ledbywellqualifiedleaders,withevidencebasedstrategies.Whilesocialstatusandbeingrecognizedasprofessionalsdomatter,whatmattersmoreistheoutcomethatweachieve.BeingmoreprofessionalinthewaythatIhavesuggestedis,Ifirmlybelieve,thewayforwardforapoliceservicewhichmustreinventitself,almost200yearsonfromitsfoundation.

Thank you.

Best available scientific evidence

Organizational facts and characteristics

Mangerial expertise and judgement

Evidence-based decision

Stakeholders’ values and concerns

78

BibliographyDurlaufSN,NaginDS.2013.‘HardTimes’.Policing UK 2013: Priorities and Pressures: A Year of Transformation.London:ThePoliceFederation.http://www.polfed.org/PolicingUK_2013.pdf

Goldstein H. 1979. ‘Improving Policing: A Problem-Oriented Approach’. Crime & Delinquency.25(2):236-258.

Goldstein H. 1990. Problem-Oriented Policing. New York: McGraw Hill.

Goldstein,H.1960.‘Policediscretionnottoinvokethecriminalprocess:low-visibilitydecisionsintheadministration of justice’. The Yale Law Journal:69:543-594.

HerbertS.2006.‘Tangledupinblue:Conflictingpathstopolicelegitimacy’.Theoretical Criminology, 10(4):481-504.DOI:10.1177/1362480606068875.

MazerolleL,BennettS,DavisJ,SargeantE,ManningM.2012.‘LegitimacyinPolicing’.Campbell Systematic Reviews.Oslo:TheCampbellCollaboration.http://johnjayresearch.org/ccpc/files/2012/08/Campbell-Review-Legitimacy-Policing-Final_2012-11-7.pdf

MurderofMaxwellConfait.http://en.wikipedia.org/wiki/Murder_of_Maxwell_Confait

Neyroud,P.2011.Reviewofpoliceleadershipandtraining.London:HomeOffice.http://www.homeoffice.gov.uk/publications/consultations/rev-police-leadership-training/

PoliceandCriminalEvidenceAct1984(PACE).(http://www.opsi.gov.uk/acts/acts2002/ukpga_20020030_en_6#pt4-ch2-pb3-l1g51

Scarman,L.1981.The Brixton Disorder April 10-12 1981. London: HMSO.

79

80

The Scottish Institute for Policing ResearchSchool of the EnvironmentUniversity of Dundee Dundee DD1 4HN Scotland

t:01382384425/384205f:01382388588

www.sipr.ac.uk

Contact Details