The lead story in Tuesday's Herald read "Murder while on bail - 23 cases revealed". The story reported that between 2006 and 2010, 70,000 defendants committed new crimes while on bail.
The number of people convicted of offending on bail was presented as a five year total - but the number on bail was just a one year total. This mismatch had the effect of magnifying the offending.
Those who offended were about 17 per cent of the total pool of about 390,000 defendants on bail over that five-year period.
The combined total of 9085 people who committed violent offences of some kind also needs to be contrasted with the 390,000.
Many of those violent offences will be for lesser offending, common assaults where there is no injury.
This is not to undermine the importance of the issue. Any further offending while a person is on bail is a matter of public concern. We are right to be outraged if someone on bail commits a murder or other serious offence.
Such concern has led to the Bail Amendment Bill, which is being debated by a parliamentary select committee and which may have prompted the Herald's Official Information Act request.
One of the proposals under consideration is a reverse onus for some serious violent and sexual offenders and for Class A drug dealers. This could be a good idea, but it would not be sensible to jump from the headline numbers on offending on bail to such a policy position.
The rationale for singling out these two groups, is, ostensibly, because they are more likely to offend on bail. It is correct that measured in simple terms higher proportions of those defendants offend on bail, but again it is misleading.
Their offences that take much longer to pass through the courts; they are simply on bail for much longer. The average time on bail for those accused of Class A drug dealing is 371 days. Any fair comparison should adjust for this effect. There will be a number of high volume offences for which the average time on bail is less than 30 days.
Collectively the two groups targeted in the bill account for only about 1.6 per cent (1158 of 69,087) of offences committed on bail. Any reduction in bail for these offenders will not make much difference to total offending.
The maximum possible 1.6 per cent reduction would be achieved only if all the offenders were remanded in custody the whole time.
The Ministry of Justice believes that for Class A defendants, the bill will result in a 6 per cent reduction of those on bail.
If that holds true for the serious violent and sexual offenders, then the actual reduction is likely to be about 70, or one tenth of one per cent.
The chance of avoiding a more serious offence becomes ludicrously small.
These issues are not simple: important human rights are at stake, including the fundamental right to be treated as innocent until proven guilty. Pre-trial detention raises the risk that people who are not convicted are nonetheless punished.
This is not just a theoretical risk. Over the five years from 2006 to 2010, 442 people (36.5 per cent) remanded in custody on Class A or serious violence or sexual offences were not convicted of the offence.
Our collective power as a society was used to detain these people. They are not compensated for the time lost; collectively we expect them to accept it as a cost of being part of society.
It is because of this kind of effect that the bill has a negative cost-benefit. According to the estimates, it will result in annual benefits to the Government and the public of $672,000.
This is the value of the harm avoided by reducing offending on bail.
Conversely the costs are estimated to be $3.77 million.
These are costs of providing additional prison beds for people who are remanded in custody and not convicted and those who are convicted but the offending is determined to not warrant imprisonment.
While the benefits include the avoidance of harm to the public, the costs cover only the cost to the Government and ignore the cost of the harm to the people remanded and their families.
The bill will simply do more harm than good. It is a misplaced reaction to the dreadful crimes that take place. It is very unlikely to avoid any significant number of those crimes, and does so only by the use of unreasonable detention against a substantial number of people who do not deserve it.
Wayne Goodall is doing a PhD in criminology and is a member of Rethinking Crime and Punishment.