Skip to main content


Link Copied

The domestic violence protection order – evidence based policy or policy based evidence?

Penelope Gibbs
31 May 2018

I am beginning to wonder why the government does any research on justice issues. They either fight tooth and nail to keep their best reports secret, bury them, misinterpret them or fail to follow them up.

I’m afraid the report on unrepresented defendants in the Crown court (about which the ICO is taking criminal proceedings against the Ministry of Justice) is only the tip of the iceberg of secret research and analysis carried out by MoJ. Many reports never see the light of day. When I FOI them, my requests always get rejected.

A number of good published reports are ignored, maybe because their findings are inconvenient. The excellent 2010 evaluation of virtual courts is online, but it has never been quoted in any government communication since publication. Its main finding – that virtual courts cost more than physical courts – did not suit the narrative. A Justice Data Lab report on accommodation for those on bail/HDC suggested that those housed there were more likely to reoffend. This has been ignored too.

Misinterpretation of reports is also a problem. The MoJ process evaluation of pre-recorded video cross-examination of vulnerable witnesses has been used by judges, ministers and civil servants to claim that the pilot was a resounding success. But the evaluation didn’t say that. A process evaluation is the cheapest possible, and merely determines whether a programme has been implemented as intended. Only an outcome evaluation can determine the success of any measure. Witnesses who had undergone a pre-recorded video cross examination were interviewed as part of the process evaluation, but their views were not compared to those of witnesses interviewed on video “live” or to witnesses who appeared in the physical court. And we don’t know the justice outcomes of one versus another. Pre-recorded video cross-examination of vulnerable witnesses seems a good idea, but we still have no research to prove it.

The biggest mistakes made in relation to the evaluation of the Domestic Violence Protection Order were in the design of the research and the failure to follow it up – neither of which are the fault of the original researchers. The Domestic Violence Protection Order is a civil order designed to protect a domestic abuse victim. It imposes restrictions on the perpetrator (such as keeping away from the home of the victim, for 28 days), thus giving the victim “breathing space” to access services, and work out what to do. The government piloted DVPOs in 2011 in three police forces. The order was hailed as a resounding success on the basis of the evaluation, but is this justified?

  • The DVPO was compared in reducing recidivism with arrest followed by no further action. This is an incredibly low bar. Why not compare with court conviction and sanction, with out of court disposals or with some other approach? Just because DVPO had some positive effect, this doesn’t mean another approach/sanction might not have worked better.
  • The success in terms of recidivism was based on only 123 matched samples, and no evidence was gathered of long terms effects (follow up was 9-19 months post DVPO).
  • There were only statistically significant improvements in recidivism if the DVPO was imposed after 3+ police call outs. If the DVPO was imposed in cases where there had been only one or two call outs, it had no impact on recividism. In fact, for cases where the police had been called out only once, it looks as if imposition of the DVPO might have made things worse.
  • “Not one perpetrator had engaged, via DVPO, in an intervention programme to begin to address their abusive behaviour”.
  • Victims said nearly half of DVPOs were breached (in terms of the perpetrator contacting the victim) but the police stats said only 1% were breached.
  • When factoring in the costs of imposition, reduction in recidivism etc: “Overall, considering both costs and benefits associated with DVPOs, the analysis indicates that the net economic and social impact of DVPOs was -£896,518 across the three police forces. This is equivalent to a return of 23 pence for every pound spent on DVPOs – i.e. a negative return on investment”.

Despite repeated suggestions in the report for it to be followed up, and despite the limited sample size, there has been no follow up research or monitoring of DVPOs. We know how many are imposed, but not how many are breached, nor how many victims access support services, nor how many perpetrators access programmes. Crucially, we have no idea how DVPOs are currently affecting the behaviour of perpetrators ie whether they are reducing domestic abuse.

The clear steer from the evaluation that DVPOs have no beneficial (and possibly a harmful) effect in cases where there have been two or fewer police call outs seems to have been completely ignored. I can’t find reference to it in police guidance, in inspection reports, or in the Home Office consultation on domestic abuse.

Victims, police and practitioners are quite positive about DVPOs. But given the lack of data and research, and indications they don’t work that well, I’m wondering why the government are proposing extending DVPOs (to be re-named Domestic Abuse Protection Orders or DAPOs) so they can be applied for by a wider group of individuals and agencies, and imposed by civil and family, as well as criminal courts? This looks like policy based evidence rather than evidence based policy. And responses to the current consultation on domestic abuse will only help a bit, since they’ll be based on personal experiences. Such testimony is valuable, but without well resourced, comprehensive, long term studies of the current DVPO, why risk extending its use?

PS If you have any experience of the domestic violence criminal justice process, and/or views on the DVPO/DAPO please fill in our survey