Domestic violence is, not surprisingly, widespread in Coventry. We were uniformly advised by respondents, both local and national, that there should not be any difficulty in attracting sufficient referrals to a programme for non-convicted perpetrators. Achieving an ethnic and social mix would require work on appropriate publicity and outreach but would be possible on a city-wide basis. The most highly respected services operate a men's programme and a women's support service integrated within one specialist, formally constituted voluntary organisation, with the workers for both located in the same office base. One at least has come to this arrangement in preference to a previous partial separation and considers it the best way of meeting the safety needs of women and children. Local women's organisations, on the other hand, felt that they should take on any support work with women. This issue would need resolving, perhaps through a hybrid model as appears to operate in Colchester. Whatever is decided, there would need to be the closest possible working relationship between any new project and existing local organisations, particularly those working with abused women.

The project, if established, should adopt the minimum standards and good practice guidelines developed by the National Practitioners' Network (now relaunched as Respect) as a baseline for any new development, since they represent agreed best practice (Respect, 2000; see Appendix 1). Beyond this, the practitioners appointed to the potential project should be free to learn about, and to adapt, the detailed content in use in the most firmly established projects elsewhere, taking into account whether or not these have been evaluated in recommended ways.

The biggest problem for such a development might lie in raising adequate resources. A service level agreement with social services, if they were willing, could fill from a third to a half of the available places and, at the same time, provide some security in funding. The other priority beyond what has already been done should be in approaching a greater range of bodies in the health sector (currently in a state of flux) because of the link with both women's and men's mental health, and with the physical well-being of women and children. There is a strong argument that domestic violence work falls within NHS priorities for a healthier nation. A wide range of organisations could be asked to contribute in cash or kind. Assistance from the City Council with low-cost premises would be invaluable, as would a formalised secondment of sessional workers from other agencies to co-lead groups. There would be one-off costs involved in equipping premises and undertaking initial publicity for which a range of organisations and community groups could fund-raise or make contributions. There would still be a need to attract substantial grant funding.

Without some consistency and security of resourcing for its core work, it would be more difficult for the project to recruit experienced workers and to put down roots in its local environment. This would, in any case, be a tall order but experience elsewhere suggests that it can be done. Although the project would be small, it would immediately become part of an impressive and supportive local partnership and national network. The latter would welcome another voluntary sector initiative since there is some feeling that the Home Office is imposing uniformity. Chances of successfully establishing a new organisation would therefore be good. Many of the key agencies in Coventry are already on board. There is more work to be done across the local community where the potential for widespread support is high but where considerable awareness-raising and information-sharing would still be needed. Minority ethnic communities would need to be closely involved, with participation by Asian and African Caribbean perpetrators in the project posing particular challenges for appropriate policy and practice development. The project's accountability to women would need to be reflected in a high level of involvement by local women's organisations in the design and delivery of the programme.

All sectors currently share a frustration with the lack of appropriate responses available in the Coventry area, outside of the criminal justice system, to challenge men's violent and abusive behaviour. There would inevitably be suspicions as to what a new body working with violent men could achieve and how it would operate. Working to agreed national principles would help to answer most of these concerns but all those involved would need to recognise that only a limited proportion of men do change their behaviour, however intensive the efforts of dedicated and skilled professionals, and that the commitment of resources in this direction can only ever form one aspect of a comprehensive raft of services. There has to be a clear decision that it is one priority amongst many, where the alternatives are to do nothing about violent men or to allow interventions to spring up, detached from the specialist domestic violence umbrella groups, which may do more harm than good.

If the project moves to the next step, tasks will include:

maximising support for existing women's services and deciding how to design an integrated women's service to work with partners;

outreach, publicity and awareness-raising across the City;

detailed consideration of how to make the programme appropriate for diverse ethnic participation;

drawing up formalised policies, resourcing agreements and practice protocols including to clarify boundaries with the criminal justice system and to optimise interagency working more generally;

detailed budgeting and fund-raising;

locating suitable premises;

recruiting appropriately skilled and experienced workers;

providing them with adequate initial training, consultancy and supervision;

supporting them in operationalising the programme;

incorporating measures to maximise completion rates;

building in valid and reliable evaluation.

There remains a separate issue of developing appropriate services for children which cannot be subsumed under this current project but which the authors of this report would be pleased to discuss further.