Implementation Science in Criminal Justice – Making Evidence Work in Practice
Podcast: Play in new window | Download
Subscribe: Apple Podcasts | Spotify | Amazon Music | TuneIn | RSS
The latest IN-CJ podcast brings together the closing roundtable in the series on implementation science in criminal justice. Chaired by Vivian Geiran, the discussion builds on earlier European and American perspectives, and asks a deceptively simple question: once criminal justice agencies know what works, how do they make it work in practice?
The roundtable brought together Professor Ioan Durnescu of the University of Bucharest, Dr Nick Powell of the Georgia Department of Community Supervision, Dr Frances Chen of Georgia State University, Dr Tonya Van Deinse of the University of North Carolina at Chapel Hill, and Lena Canti of the Irish Probation Service. Each contributor approached the question from a different standpoint, including research, frontline probation practice, staff development, implementation evaluation, and organisational leadership.
A clear theme running through the discussion is that evidence does not implement itself. Criminal justice systems have become more skilled at identifying evidence-based practices, but the harder task is embedding them in real services, with real staff, in real organisations, under real pressures. As Nick Powell put it, evidence-based practice is a verb, not a noun. It is not something that can be lifted from a shelf, launched, and then left to run.
That observation challenged a common organisational habit: the assumption that implementation is mainly a roll-out, a launch date, or a go-live moment. The contributors suggested that this model may be too strongly influenced by technical project management and ICT implementation. Human services do not work in the same way. Criminal justice practice depends on judgement, relationships, supervision, trust, capacity, and local conditions. The work of implementation is therefore continuous, adaptive, and often slow.
Tonya Van Deinse cautioned against making “implementation science” sound more complicated than it needs to be. The practical task is often to ask clear questions: what is being introduced, why is it needed, who has to use it, what will make it workable, and how will we know whether it is being used well? The language of implementation science can be useful, but only if it helps practitioners and agencies make better decisions. If the terminology becomes a barrier, it risks becoming another technical discourse imposed on staff rather than a practical aid to change.
This point was reinforced by Frances Chen, who emphasised fidelity, design, and user involvement. If a new tool, programme, or practice model is intended to help officers, then officers need to be involved from the beginning. They should not simply be handed a finished model and asked to comply with it. They need to help shape the design, test whether it fits their work, and identify the barriers that will affect uptake. In this sense, implementation cannot be separated from the experience of the people who are expected to make it happen.
Lena Canti described how the Irish Probation Service has approached the implementation of the Irish Probation Framework. The service has moved away from the idea that staff can simply be trained, given a manual, and expected to deliver change. Instead, it has developed communities of practice, feedback loops, reflective spaces, and support for managers. These help staff to understand the changes, discuss barriers, share experience, and identify what further support is needed. This is implementation as organisational learning rather than one-way instruction.
Ioan Durnescu brought the issue to life through the metaphor of the kitchen. A recipe is not the same as a meal. Two people may follow the same recipe, but the outcome will depend on the quality of the ingredients, the equipment available, the skill of the cook, and whether the cook is exhausted. In the same way, an evidence-based model does not produce the same results automatically in every setting. Implementation science helps organisations understand what needs to be in place so that practice can be delivered with sufficient fidelity while still being adapted to context.
The roundtable also explored how success should be measured. The contributors were careful not to reduce implementation to one final outcome, such as recidivism or revocation rates. These outcomes matter, but they often appear much later and are shaped by many factors. Implementation also needs to be assessed through intermediate measures, including acceptability, appropriateness, adoption, reach, staff confidence, fidelity, sustainability, and the quality of feedback from practitioners. The question is not only whether a policy has been introduced, but whether it has become part of practice in a meaningful and durable way.
A related issue is the balance between evidence-based practice and practice-based evidence. Several contributors noted the risk of top-down imposition, especially in criminal justice organisations that are already highly structured and sometimes coercive. Staff resistance should not simply be dismissed as obstruction. It may contain important information about whether a proposed change is workable, credible, or aligned with the purpose of the service. Listening to staff is not a concession to resistance; it is part of the evidence base for implementation.
The discussion also addressed the risk of managerialism, or what some contributors described as “McJustice” or “McProbation”. Standardisation can be valuable where it helps people receive a fair and consistent service. However, it becomes problematic when it strips practice of judgement, discretion, and humanity. Nick Powell argued that implementation must “keep the human in human services”. Evidence-based practice should serve the mission of improving lives, supporting communities, and creating better outcomes, not simply give organisations another system to impose.
Coaching, supervision, and peer learning were identified as crucial parts of implementation. Staff need more than initial training. They need opportunities to practise, reflect, receive feedback, and develop confidence over time. This is especially important where implementation involves complex practice, such as specialised mental health supervision, motivational work, or new supervision frameworks. The discussion suggested that good implementation builds the conditions in which staff can do the work well, rather than blaming them when a poorly supported model fails.
The contributors were broadly optimistic about the future of implementation science in criminal justice. It may not always be called implementation science, and the terminology may change, but the underlying principles are likely to become more embedded. These include starting small, piloting, learning from practice, adapting to context, using data intelligently, supporting staff, and treating implementation as a long-term process.
One of the strongest closing messages was the need for clearer, simpler communication. Frances Chen summed this up directly: speak in plain words. If implementation is to succeed, people need to understand what is being changed, why it matters, what is expected, and how their experience will shape the process. Overly elaborate language can distance people from the work. Direct communication can bring them into it.
The value of the IN-CJ series lies in making these conversations international and comparative. Different jurisdictions face different legal, organisational, and cultural conditions, but many of the implementation challenges are shared. The details vary, but the central problem is familiar: how to move from knowing what works to making it work in practice.
This podcast shows why that question cannot be answered by research alone, or by policy alone, or by management alone. It requires dialogue between researchers, practitioners, leaders, trainers, and the people who carry the responsibility of frontline delivery. Implementation science, at its best, is not a technical fix. It is a disciplined way of thinking about how evidence, people, organisations, and practice can be brought into better alignment.
