Abstract
The $50 billion fraud perpetrated by Bernard Madoff worked for years because his New York-based fund appeared virtually to guarantee high returns. As a result his investors failed to probe the legal plausibility of the exercise. The same phenomenon characterises some crime and reconviction statistics. Which is why the outspoken criticism voiced in early December 2008 by Sir Michael Scholar, Head of the UK Statistics Authority, regarding the premature release by the Home Office of ‘selective’ knife crime statistics was so welcome. Back in 2003, Lord Warner, my predecessor as Chairman of the Youth Justice Board (YJB), trumpeted that the re-offending rate (it should have been described as the reconviction rate) for young offenders subject to non-custodial penalties was 22.5 per cent below that predicted. The announcement prompted Tony Blair publicly to congratulate the YJB on its achievement (YJB, 2003). It seemed implausible (at least to me) at the time and arguably the YJB should have questioned the evidence. But the fault, as a detailed critique initially by Tony Bottoms and subsequently by Bottoms and Dignan (2004) exposed, ultimately lay not with the YJB but with Home Office statistics. There were many missing data. The reconviction rate had subsequently to be revised downwards, first to 7.7 and then to 2.4 per cent (Home Office, 2004; see also Solomon and Garside, 2008).