Despite increased funding and a new community justice strategy, reoffending rates in Scotland have failed to improve over the past decade.

It is a continuing and costly problem, with abut one-third of offenders being found guilty of a further offence within a year.

The most damning evidence that services to reduce offending are not working is that one in five offenders sentenced in 2010/11 had at least 10 previous convictions. There is an urgent requirement to find an effective way to halt this vicious circle, whose social and economic costs are around £3 billion a year. An attempt was made five years ago when eight Community Justice Authorities (CJAs) were set up to allocate money between the 32 councils to deliver community-based sentences and services such as addiction treatment for offenders.

Last year the CJAs paid £99 million to service providers but effectiveness varied widely across the country. Audit Scotland, in a critical report into reoffending, has identified "a mismatch between what is delivered and what is known to work". That is putting it mildly. Despite more than 1300 services being provided for offenders in prison and the community across Scotland, the system is failing. It is clear that changes are required to avoid squandering money on schemes that fail to tackle the factors that lead to a life of crime. The report says there is plenty of evidence of measures that are effective. It is essential, therefore, that funding is directed towards successful outcomes. These involve helping offenders to find jobs, improve relationships with their families and manage their lives.

The present inflexible funding arrangements for CJAs must be changed. These are based on factors such as the unemployment rate and previous offending rates rather than actual need. Auditor General Caroline Gardner found this does not encourage CJAs to reduce offending. Their role across different councils also appears to be causing difficulties. Yet they were set up to improve joint working between the multiplicity of organisations involved with offenders. It seems the need for co-ordination was correctly identified but the wrong remedy applied.

When resources are scarce they must deliver good value. Of the £419m spent dealing with people convicted of offences in 2010/11, less than one-third was used to reduce reoffending. The need for better outcomes has long been recognised, not least by the Scottish Prison Service. Efforts to extend rehabilitation in jail, however, have been hampered by the increase in the number of prisoners. Despite the Scottish Government's policy of a presumption against sentences of less than three months, the prison population has risen by 27% in 10 years. The 50% increase in the number serving between six months and four years is especially worrying because this is the group most likely to reoffend.

More support is the key to break the spiral of offending. If these less serious offenders can be helped to gain employment and form stable relationships, thus gaining a stake in society, they could be prevented from falling into the group with three or more previous convictions who cost the taxpayer £5.4bn over a 10-year period. The cost in wasted lives is incalculable. We know what works; we must ensure the money follows best practice.