The first domestic abuse policing strategy in UK history to be trialled under experimental conditions has found that an inexpensive two-day course in behaviour management for first offenders resulted in 35% fewer men reoffending against their partner, and reduced further harm to victims by over a quarter.
Researchers at Cambridge University's Institute of Criminology worked with Hampshire Constabulary to conduct the study using the recently developed CARA (Cautions and Relationship Abuse) programme: small-group discussion workshops for men who received conditional cautions for first arrests for low-harm domestic abuse.
The researchers say that, in just this initial study of hundreds of Southampton-area offenders over a 12-month period, the CARA programme prevented significant harm to victims, hundreds of prison days, and consequently saved thousands of pounds.
The findings are published in full in this week's print edition of the Cambridge Journal of Evidence-Based Policing.
The team behind the study say that several police forces want to replicate the use of the CARA course, developed by the Hampton Trust domestic abuse charity. However, they say that current guidance from the Crown Prosecution Service restricts the use of conditional cautions for domestic abuse across the country.
"Dealing with high volumes of low-harm common assault cases against intimate partners is a significant issue for police forces across the UK, particularly in times of continued austerity," said study lead author Professor Heather Strang, Director of Research at Cambridge's Jerry Lee Centre for Experimental Criminology.
"No other programme to our knowledge now has such strong evidence of yielding a substantial reduction in harm to victims of domestic abuse.
"The CARA programme should be approved for general use with low-harm first offenders, preferably with further randomised trials to ensure it works for different communities across England and Wales."
The study only involved adult men who admitted their offence, were not judged 'high risk', and had no record of any violence in the preceding two years. All victims agreed to their partners' participation.
To be eligible for the experiment, the offence had to be classified as either common assault/battery, criminal damage, harassment, threatening behaviour, or domestic theft.
Of the 293 offenders who fit the strict criteria between August 2012 and November 2015, around half were randomly assigned to attend CARA workshops, run by experienced facilitators from the Southampton-based Hampton Trust.
The CARA programme consists of two five-hour group discussions of between four and seven men, held on weekends one month apart, in which facilitators raise questions that cause attendees to reflect upon their behaviour and how they might change it.
Offenders in the other half, the control group, were given 'conditional cautions': meaning any repeat offence within four months would see prosecution in court. This is a commonly deployed police response to first arrests for low-harm domestic abuse.
Professor Strang and colleagues - including several Hampshire police leaders enrolled on the Cambridge Police Executive Programme - followed up with offenders a year after the first arrest. They found that 35% fewer men in the CARA group had committed any further offence against their partner.
However, Cambridge co-author Professor Lawrence Sherman describes such simplistic 'crime counts' as unhelpful when determining the real cost of crime: harm caused to victims. "The key result for the team came when we analysed all reoffending in both groups using the Cambridge Crime Harm Index," he said.
This Harm Index, or CCHI, is a new tool that measures harm by weighting the severity of each crime in sentencing guidelines for different offences, rather than just totting up overall crime figures. The Office of National Statistics credits the CCHI as the stimulus for its own (modified) version of a harm index, introduced earlier this year.
Overall, those in the CARA group caused 27% less harm per offender to their partners than the control group.
Using the CCHI, the team calculated that the recommended number of prison days under English sentencing guidelines for reoffenders in the year following the first arrest was an average of 8.4 days for the CARA attendees, compared to an average of 11.6 days for offenders not sent to CARA.
"This would mean that, for every thousand first time offenders sent to CARA workshops, 380 days of recommended imprisonment would be saved, and victims would be spared the inflicted harm equivalent to 380 common assaults, or 19 assaults with actual bodily harm," said Sherman.
Men who participated in the CARA workshops described having a greater understanding of the impact of their behaviour on partners and children, and when to walk away from a fight. Some talked of going on to attend support groups such as Alcoholics Anonymous as a result.
Chantal Hughes, Chief Executive of the Hampton Trust, said: "We know from consultations with victims that they want help for their partners. Those choosing not to remain in an intimate relationship often have children, and this means child contact arrangements. Victims have advised us that workshops such as CARA are a positive and much needed intervention."
Study co-author Scott Chilton, Assistant Chief Constable of Hampshire Police and Chair of the Society of Evidence Based Policing, said: "CARA is an outstanding example of evidence based innovation that can influence national police policy and practices.
"This type of research, where professionals from law enforcement, working with academia and charitable organisations has proved to be extremely promising."