A critical perspective: why Crime Science needs better frameworks

Implementation failure is pervasive – but is dumbing down the solution?

Many of the reports on major national initiatives like the Crime Reduction Programme in England & Wales (prematurely ended 2002) acknowledge implementation failure (e.g. see Maguire 2004; Bullock and Tilley 2003; Ekblom 2011c, chapter 2). Common explanations for such failure have been ‘poor project management skills’, or ‘short-term and overcentralised funding regimes’. But important though these are, another significant factor is arguably dumbing down.

Senior managers and policymakers in the field of crime prevention, security and community safety, together with many practice-oriented researchers, often seem to believe something along the lines of ‘the only information you can ever hope to get into practitioners’ heads is a slogan or two – if you’re lucky.’

Crime prevention reality – complex and messy

Paul Ekblom couldn’t disagree more. Certainly, the basic idea of crime prevention and related activities is simple: cut the cause and you cut the risk of crime. But in practice, prevention is complex and messy. Reliable and valid data must be collected; patterns analysed to reveal important subtleties and avoid misleading conclusions; evidence of what works assessed, selected and adapted (cookbook copying won’t work); in-principle ideas turned into practical reality; people and organisations mobilised or engaged in partnerships; evaluation, feedback, adjustment and learning undertaken; and knowledge shared for collective benefit.

There is numerical complication, too – one excellent project on underage drinking (see Project Moonshine) had 13 different interventions – all had to be realised, and customised to context. Until this and the accompanying practical information was winkled out of the local project team in the course of a three-hour interview they simply didn’t realise the extent of their own achievement… and more importantly, neither did anyone else.

Here’s the main point. Given this complexity, there is no point in dumbing down the communication of crime prevention or community safety knowledge into immediately graspable slogans and rapid-read case studies to aid communication, if the information thereby imparted cannot then deliver actions that are sophisticated and subtle enough to have a chance of working, and of avoiding harmful side-effects. Ashby’s Law of Requisite Variety is spot-on – our mental models and representations have to be sufficiently complex and sophisticated in themselves to control the complexity of the real-world problems they are intended to tackle.

The severest instances of oversimplification are the frameworks most practitioners are given, to help organise their knowledge of crime causes and preventive interventions, and the process of doing crime prevention and community safety. Two excellent instant introductions to the problem-oriented approach to prevention are the Problem Analysis Triangle (PAT – crime problems caused by combination of target/victim, offender, location, as clues to intervention) and SARA (the process of Scanning, Analysis, Response, Assessment). But once practitioners have grasped these basics, they rapidly find little depth to structure, store and retrieve knowledge, and guide action. Simple introductions become constraints on advanced action.

Nor can the PAT and SARA inspire the creativity and generate the innovations which are vital to take crime prevention to new contexts and to keep up with social change and adaptive criminals. The PAT is incomplete and under-represents the richness of the offender side in particular: even situational interventions may work better if based on a more subtle understanding of offenders, and what influences their behaviour. SARA’s amorphous ‘Response’ stage, in particular, confuses several quite distinct activities, as this discussion makes clear.

Oversimplification in research and practice

Looking beyond the development of guidance for front-line practice of crime prevention, towards research and conceptualisation within Crime Science, fragmentary part-models of causes and theories abound. Each individual model (such as Routine Activities Theory) is simple – but the models are in effect dumped in a pile and it’s left to each individual researcher (and worse, practitioner) to work out how to relate one to the other (how does Routine Activities relate to the Rational Offender? And where does Environmental Criminology or Crime Prevention Through Environmental Design fit in?). How are researchers and practitioners to cope with gaps, overlaps and explanations which may be implicitly competing, or operating at different levels? Taken as a whole, this is not simple – just chaotic.

Any plans put forward to improve crime prevention, security and community safety performance will require a decent framework and language to describe and structure knowledge of good practice if they are to succeed in getting the best out of existing research, stimulating new studies and facilitating the training and guidance of practitioners. This is basic good practice – in knowledge management.

What lies behind this drive to oversimplify?

Several mutually-reinforcing and self-fulfilling factors come to mind:

Misunderstanding of practice and disdain for detail. Senior policymakers prefer high-level abstractions such as the Systematic Reviews produced by the Campbell Collaboration, delivering messages such as ‘Street lighting is a cost-effective way of reducing crime and fear’. This is completely right for strategic policy (and great, too, for political slogans), but once we shift perspective to the infrastructure needed to deliver the policy or the practical detail needed to implement it on the ground, it’s not remotely detailed enough.

The usual remedy is for practitioners wishing to emulate a successful project, to be given the team’s contact details or if lucky, to be bussed out to visit them. But the architect of the project may no longer be in post, or may have little time to talk to them. Or they may simply be unable to articulate what they had done in clear, transferable terms that make the underlying principles and ‘troublesome trade-offs’ with other goals, and interactions with contextual factors, overt rather than tacit. Although there is a very significant place for peer-to-peer learning, undertaken in isolation and without structure it is hardly efficient or quality-assured.

A patronising attitude to practitioners. There is concern that the poor practitioner won’t understand and use the information. Allied to this is a vision of practitioners as technicians with limited diagnostic skills and limited repertoire of responses. Yet it is common experience from working alongside practitioners that they can well handle the complexity of knowledge if they see it as relevant to their task and reflecting the necessary complexity of what they are daily dealing with. And they certainly aspire to be less like technicians, and more like consultants with principles and methods at their fingertips ready to configure and innovate in each new context.

Top-down stops too soon. Top-down initiatives for improving performance are important, but the programme planners somehow always leave their grand edifice miraculously suspended without a ground floor, let alone foundations. To fill the gap they rely on barely adequate toolkits and minimalist peer-to-peer sharing of good practice – namely superficial entries in ‘knowledge bases’ (an unfortunate example was described by Bullock and Ekblom 2010).

A reluctance to grasp the nettle of improving the frameworks for organising knowledge and action. Fear prevails, that practitioners could never be prised out of their comfort zones of the PAT and SARA to accept something that was initially more challenging – but ultimately much more functional. (Would you want your doctor’s training to skip over relevant theory and knowledge of diagnosis and treatment, to rely instead on four very generalised steps and a mixed-up set of theories?)

The problem is compounded by the ‘pick’n’mix’ approach to learning and the reinvention of (often very imprecise) terminology and structure in every new practice guidance document or toolkit. The slogans and superficial definitions used in major programmes give no clear and systematic grounding to practitioners’ work.

A stance of ‘low investment in professional education and career development, low yield in performance’. It’s understandable perhaps that employers and maybe employees too are unwilling to devote significant resources to training if staff are dipping in and out of community safety posts or doing them alongside other work. This is an especial problem in smaller crime prevention/ community safety organisations and partnerships. But the consequence is wasted opportunity for safer communities, and limited professional careers.

So what’s to be done?

To be avoided is trying to pack lots of complex knowledge into an unready and soon-to-be-aching head, by reliance on learning from knowledge-bases and toolkits while the practitioners struggle to do their job. The investment in practitioners has to be front-loaded – give them a professional start to their career by imparting a rich and sophisticated schema for understanding crime, security and community safety and knowing how to tackle it. With such a schema embedded in their minds they can comfortably take in much more information, make their own professional judgement about how valid or useful it is, assimilate it (they know exactly which mental files to store the information in and retrieve it from) and start to apply it.

Paul Ekblom has spent much of his career working with others including practitioners in developing frameworks and definitions as ‘precision tools for thinking, communication and action’ that are designed to handle appropriate complexity in the knowledge and practice of crime prevention and community safety.

One of these tools is the Conjunction of Criminal Opportunity Framework (CCO) which provides a one-stop-shop to identify causes and design interventions to reduce crime. The CCO is an integrated map of 11 generic causes of criminal events, covering both situational and offender perspectives, and 11 equivalent families of intervention principle. It incorporates the PAT, Routine Activities, Rational Choice and much more, on both situational and offender sides.

Another tool is the 5Is Framework for capturing, consolidating and transferring knowledge via structured and detailed project descriptions and a prescriptive process model. It avoids oversimplification by extending and deepening the SARA process (as Intelligence, Intervention, Implementation, Involvement and Impact, and many subsidiary levels beneath these main task streams). Other frameworks, definitions and glossaries on this Crime Frameworks website help researchers and practitioners handle, not hide, the complexity they encounter in understanding and controlling crime.

More advanced and sophisticated, yes – but not exactly quantum mechanics. Both 5Is and CCO Frameworks zoom from top-level headings down into organised and analytic detail allowing a wide choice of intervention methods, innovation and configuration of action to context; and allow users to find their own comfort level of sophistication.

Good research and practice? Don’t dumb down – invest in frameworks!

The above text is adapted from Ekblom (2006a). Fuller exposition of this viewpoint is in chapters 3-5 of the 5Is book.

Presentations

The presentations below have developed the above themes.