Drive to prevent homelessness needs better evidence of ‘what works’

4 April 2000

A growing understanding of the circumstances that place single people at risk of homelessness is creating new scope for early intervention and prevention. But policy makers and housing professionals must avoid jumping to conclusions about causes of homelessness and ‘good practice’ without sound evidence to support their assumptions.

The call for a more systematic and evidence-based approach to tackling single homelessness and rough sleeping comes in the first-ever, comprehensive overview of research commissioned by the Joseph Rowntree Foundation and CRASH, the construction and property industry charity for the single homeless. It finds that attempts to explain single homelessness as purely a housing problem are giving way to more complex analyses that take account of social, economic and personal risk factors.

  • Major economic and social forces

likely to affect the level of homelessness include shortages of suitable or affordable housing, availability of jobs, levels of poverty and social security and trends in relationship breakdown and family re-structuring. However, there has been little rigorous research to show how national trends affect patterns of single homelessness at local level.

  • Individual factors

Linked to an increased risk of homelessness include low income, unemployment, sexual or physical abuse, family breakdown, school exclusion, poor mental health, drug and alcohol misuse, service in the armed forces and experience of local authority care or prison.

  • Events

Most likely to ‘trigger’ homelessness among high-risk individuals include leaving the family home after arguments, relationship breakdown, eviction, bereavement, a sudden deterioration in mental health or increase in drug / alcohol abuse, leaving care, release from prison or discharge from the armed forces.

The report argues that the need to focus on prevention of homelessness is now widely accepted. Knowledge concerning potential trigger points can be used to target individuals experiencing a range of risk factors with support services.

Possible preventive interventions would include:

  • Housing advice and aid, mortgage rescue schemes and rent deposit guarantee schemes.
  • School education programmes on leaving home and homelessness, mediation for family disputes, befriending and mentoring to tackle social isolation.
  • Tenancy support for young people and those at risk of eviction, resettlement programmes for people leaving the armed forces and extended support for care leavers.

Suzanne Fitzpatrick of the University of Glasgow, a co-author of the study, said: “One of the main themes to emerge from this review is the need for policy makers and practitioners to be flexible and holistic in the way they respond to the needs of homeless people. We need lasting solutions based on early prevention and long-term resettlement initiatives, not just crisis intervention. That means paying more attention to evaluation and to the evidence of ‘what works’ - and avoiding the tendency to simply follow fashions in service delivery.”

The review of single homelessness research was carried out to assess the strength of existing knowledge and identify gaps in the evidence available. CRASH is making the review available to key agencies working with single homeless people free of charge. It has also urged organisations to send in findings from any future research. Studies of importance will be summarised in a web-site bibliography, to be hosted by the University of Glasgow. CRASH will also arrange for the research to be reviewed and published on-line once a year.

Louise Casey, Director of the Government’s Rough Sleepers Unit, has welcomed the report. She said: “This research review will be of considerable value to government and all concerned with delivering better services to homeless people. With CRASH’s offer to continue to monitor and review single homelessness research, we will be able to identify continuing research gaps and raise research standards.”

Recommend to a friend via email: