The Wolfson Professor of Criminology at Cambridge University, Lawrence Sherman, introduces an unexpectedly sermonic note to the RCT debate, kicking off his keynote lecture with a line from St John’s gospel: “You will know the truth and the truth will set you free”.

He has good reason to make the connection. The caring professions have a lot to say on the subject of truth and much rides on whatever liberating certainties they are able to muster.

An American, Sherman reminds a largely British audience that the UK is probably the world leader in truly knowing what works in relation to health care as a result of establishing the Campbell Collaboration and the National Institute for Clinical Excellence (NICE).

But what of criminal justice approaches and interventions? How do we know that locking up prisoners is an effective way to deal with crime? How do we know what policing practices are effective? What can we do to prevent crime in the first place?

Sherman says the hardest words for the expert are “I don’t know” but they pretty much cover all that police, judges and policy makers are able to say if asked about the effectiveness of criminal justice approaches. And when we don’t know what’s effective, he insists, the admission that we do not know ought to be obligatory.

He reminds the audience of recent European Union legislation that requires providers of services or products to state the evidence of their effectiveness. It even extends to fortune tellers who are required by law to acknowledge that their service is for entertainment purposes only and is not experimentally proven.

He argues that in the case of the criminal justice system not knowing is not good enough. Actions taken without due evidence (for example, prison sentencing and other punitive approaches) have too serious consequences (both economically and in relation to offender and victim outcomes).

He therefore makes the case for a NICE for crime – a National Institute for Criminal Excellence. The evidence for approaches that work and against those that don’t is becoming more compelling. For example, we now know that dangerous and expensive police raids on crack houses may work - but that their impact on drug crime in any particular area lasts for a mere 12 days. The risks are just not worth such transitory benefit.

Other examples of things that don’t work or make matters worse include preventative interventions for young people such as Scared Straight - an attempt to frighten children away from crime which has been found to have the opposite consequence. [For more about what doesn’t work, see Knowing what works isn't altogether working].

What about approaches that do work? Sherman’s c.v. includes rigorous RCT evidence supporting the use of a range of evidence-based policing approaches to reduce crime (for example by increasing patrols in crime hotspots). Prevention Action has also reported on a range of effective approaches working with young offenders. [See, for example: Pennsylvania's prevention gamble comes home at five to one]

Sherman has been in the vanguard of restorative justice approaches which involve reuniting offenders and victims in a safe environment. Usually there is an apology by the offender accompanied by a joint decision-making process (with police) intended to make the response appropriate to the crime and its aftermath. So what is the truth about such calculated entanglements?

Sherman reports on a systematic review of 14 RCTs of restorative justice (six from the UK, four from Australia and three from the US). It appears to work for violent crime (although not for property crime). Effects are small but significant. It appears to be more effective for adults than for young people.

Why are we waiting? Why not roll it out on a large scale?

First, for there to be sufficient confidence in the findings and in order to iron out some of the wrinkles in the approach more evidence is needed. And Sherman reminds us that even when the evidence is clear-cut, it is not unusual for a great deal of time to elapse before uptake.

He offers examples from history including James Lind’s experimental evidence (1747) on the benefits of lime juice in preventing scurvy. It took 48 years to make its way into British Navy practice (just in time to fend off Napoleon). Similarly, Thomas Clarkson’s damning evidence of the cruelty and injustice of the slave trade took 55 years to have effect.

How long will it take to bring change to the criminal justice system? For better or worse, change is inevitable; the rise of RCTs and synthesis of rigorous evidence, as suggested by Sherman’s aspiration for NICE for crime, may help to light the way.