A new institute aims to underpin social work with research - no easy task, says Terry Philpot.

Social work has shown a notorious resistance to look at the evidence for what it does. Some argue that it is an art; that it works with the unpredictable material of humanity; that it is geared to getting results and not to reflection; and that social workers are, as child welfare expert Olive Stevenson once wrote, "brokers in shades of grey".

This month, the ?2 million Social Care Institute for Excellence opens. This younger sibling of the National Institute for Clinical Excellence aims to raise standards and iron out inconsistencies, review research, set up a database of good evidence-based methods and services and produce and disseminate best practice guidelines. Covering England and Wales, it will feed into the work of the emerging General Social Care Council, the Care Standards Commission and the performance-monitoring work of the Social Services Inspectorate.

At one level, social work can hardly complain. Some evidence-based initiatives have beavered away for years only to find social services generally impervious to their findings. Exeter University's Centre for Evidence-Based Social Services alone has held 163 dissemination conferences.

ÖйúAƬ

A CEBSS survey earlier this year found that only one in two managers and frontline staff believe that research informs day-to-day work. Brian Sheldon, CEBSS director, says only "a tiny proportion" could identify or describe evidence for what they did.

But sometimes evidence is not enough. It is 25 years since Jane Rowe and Lydia Lambert showed in Children Who Wait the deleterious effects of children waiting in care. That their findings have had so little impact probably has more to do with the chronic underfunding and organisational neglect of residential childcare than with any resistance to take to heart what was found.

ÖйúAƬ

Also, there is often agreement to disagree about what works. To that extent, social work is not a science. Take same-race adoption: how does SCIE hope to cut through the tangle of years of debate and value-based assertions? Social work employs a vast range of methods, including casework, crisis intervention and family therapy: will SCIE try to "prove" which "works" based on a performance-related utilitarianism that could easily become a hallmark?

SCIE also wants to find out what works for social work consumers - easier said than done. Ironically, given that social work accords almost iconic status to users' views and that "solutions" must be tailored to individuals, the user's voice is far weaker in social care than in medicine.

To tackle such problems, June Thorburn, dean of the School of Social Work at the University of East Anglia, has proposed forming time-limited working groups of people with different views to examine evidence, find areas of agreement and pinpoint where more research is needed.

Despite all the questions, SCIE has won general approval from academics, managers and practitioners. By focusing on the value of research to social work and creating a common vocabulary for academics and practitioners, it may yet prove to be one of the most useful, if least trumpeted, of the government's plethora of initiatives.

ÖйúAƬ

Terry Philpot is co-editor, with Chris Hanvey, of Practising Social Work , Routledge 1994.