The Prime Minister, Tony Blair, has launched a major new intiative to help cut the number of children who runaway from home or care.

By the age of 16, one in nine children have run away from home or care for at least one night, new research revealed yesterday.

The Prime Minister has now asked the Cabinet Office's Social Exclusion Unit to lead a project on young runaways working closely with other Government departments and outside organisations.

Launching a report on the subject, Mr Blair said: "We must find a better way of tackling this issue.

"We have to make it less likely that young people run away in the first place and, if they do run, ensure their immediate safety. And we must not just turf them back where they came from without finding out why they ran away and addressing these issues."

The report found that there are 129,000 runaway incidents each year, involving 77,000 under 16-year-olds. Nearly half - 45 per cent - of young people living in care have run away compared with 9.5 per cent of young people not living in care.

Running away, in turn, increases the likelihood of entering care and becoming homeless later in life. A recent study of young people missing from care found that nearly half had first run away from home prior to admission to care.

And homelessness studies suggest that running away at a young age is a strong predictor of later homelessness.

Children in care are also more likely to run away repeatedly, with one study finding that a third of the young people had gone missing ten or more times in the past.

There is also a strong link between truancy from school and running away. Nearly a third of young people who truanted often and a fifth of those who sometimes truanted had run away overnight, compared with 6 per cent of those who had never truanted.

Similarly young people who had been excluded from school were also more likely to run away, with a quarter doing so overnight on at least once occasion.

For young people who go missing from care, the running away rate was 1.9 for those attending school and 8.6 for those regularly or persistently truanting.

More than half the runaways from care leave return to the parental home or to be with friends, often because they are distressed by family separation, feel torn between loyalty to parents and foster carers or cannot adjust to firmer rules of behaviour set by foster carers.

Other key reasons for going missing from care included bullying among peer groups, a culture of non-attendance at school and involvement in offending, says the report.

Some young people run away from bullying, while others take part in 'group escapes' to gain acceptance by a dominant peer group. Two out of five runaways from care run away with a friend.

Running away from care is also strongly associated with not having a stable care placement.

One survey showed that, where a person had made no more than one move during their time in care, the average number of absences was four, but for those who had moved ten times or more, the average number of absences was 21.

Research also shows the wide variations between runaway rates in different children's homes. These ranged from 25 per to 71 cent of four local authorites surveyed and of 32 children's homes surveyed, seven accounted for two thirds of incidents.

Generally, the majority of runaways are aged between 13 and 15, but a quarter first run away before the age of 11. Evidence suggests that those who first go missing at an early age are more likely to run away more often.

Girls are more likely to run away than boys, but boys are more likely to run away at earlier age and on more than one occasion.

For many youngsters, running away is largely unplanned and can be dangerous and frightening.

One young runaway said: "I didn't take anything with me not even a coat. I didn't have much money and I couldn't afford to buy any food. So I couldn't even sit inside a cafe to get warm."

Another said: "I had no money, I was upset constantly. I was in bad health and I became a thief and a prostitute."

Most runaways do not come to serious harm, but research shows that a quarter sleep on the streets, with some surviving through begging, stealing, drug dealing and prostitution.

And as many as 10,000 runaways suffer physical or sexual abuse while they are away from home.

Often it is abuse they are are running from in the first place.

One runaway said: "I don't know what you mean about choice. If I stay I get smacked around. If I run off then I might get beaten up or robbed but at least I might not. At home I know for sure that I will What choice is that?"

The Government's efforts to tackling the issue stems from the fact there is currently no national policy on the best way to deal with runaways - so how effectively it is dealt with varies widely from area to area.

The background report by the Social Exclusion units says runaways tend to receive services depending on whether they fit into a particular category such as 'missing person' for the police, 'child in need' for social services, or 'person estranged from parents' for the benefits agency.

Most runaways return home or are returned but receive little help in sorting out the problems they ran from. Meanwhile, national helplines find it hard to refer callers to local help available because there is no national network of services.

The police tend to focus on missing people whose lives may be endangered and social services tend to focus their efforts on those runaways who are already in care or on the child protection register.

And runaways who run across boundaries, or who have moved home or care placement since previous incidents, often receive a poor response from various agencies.

Wherever possible, social services will try to ensure that the young person returns and remains at the family home. But if returning the child to the family home would endanger the child, social services should provide emergency access to foster or residential care.

However, says the report, "anecdotal evidence suggests that this may not always happen".

At the moment, there are no reliable national figures on the numbers of unauthorised absences from care and social workers, carers or local authorities may not know of a person has run away before because records may not have been passed on from previous incidents.

However, from April 2000 local authorities in England have been required to record absences of more than 24 hours from agreed placements for a sample of one third of their looked-after children.

This will provide the first national estimate of the incidence of running away from care. The Department of Health is also sponsoring the Missing from Care Project which involves collating detailed information from six local authorities.

The new review of runaways suggest that each runaway should be interviewed to identify and address the reasons for their behaviour.

It also recommends setting up a network of refuges and support so that under-16s have an alternative to sleeping rough.

Other ideas include having a named person in charge of co-ordinating services for runaways in every area and establishing national standards to address the huge variations in the way local authorities respond to runaways.

The report pointed to two areas of good practice in England - the ASTRA project in Gloucester - which police say has cut the number of repeat runaways by 61 per cent, saved 75 working days dealing with missing person cases and reduced the number of runaways who have been arrested by 21 per cent.

This project has involved having specialist workers who have interviewed each runaway and referred them to specialist services if needed.

Meanwhile the way that Coventry Police deals with runaways, including having a named person in charge of co-ordinating services for runaways, has managed to halve the number of runaways from care.

The report on runaways is available on the Cabinet Office website

The Social Exclusion Unit is seeking the views and opinions of interested parties on runaways. They must be receieved by June 22 either in writing to: Ana Cavilla, Social Exclusion Unit, Room 1.3, 5 Great Smith Street, London SW1P 3BQ or by e-mail to Ana.Cavilla@cabinet-office.x.gsi.gov.uk