Good comments

On most blogs the actual posts tend to be quite reasonable and well argued, and it is the comments that can sometimes go off the track and get hysterical.

But I’m pleased to report the opposite has happened on Red Alert. Carmel Sepuloni did a somewhat off the track post claiming:

Todd McClay says that beneficiaries are more likely to murder children

I thought this would set off a round of abuse of Todd, but in fact many of the comments were quite thoughtful, such as:

Rebecca:
Carmel I agree it was incredibly distasteful however, in many ways, while awkwardly put, Todd is right.

The bulk of the physical child abuse, neglect and maltreatment statistics comes out of the quarters of the lower socio-economic groups, most of whom are predominantly on benefits.

Lindsay Mitchell posted a useful link to some CYFS stats which “establishes that there is a statistical association between the two factors examined: benefit receipt and CYPFS notifications”.

Todd himself pops up and posts a draft of Hansard, with what he actually said:

And Lindsay again provides a useful point:

You have to understand that ‘are more likely to’ can co-exist with ‘most don’t’. For argument’s sake;

8 out of 100 beneficiaries abuse their children.
2 out of 100 non-beneficiaries abuse their children.

Therefore beneficiaries are 4 times more likely to abuse their children BUT most don’t.

However, inasmuch as child murder usually occurs at the extreme end of abuse, it is more likely to happen when the parent or other caregiver is a beneficiary.

So stop taking offence and start asking whether there is a link between welfare and child abuse.

Of course not all comments were high quality:

Olwyn said:

Furthermore, we did not have people regularly beating kids to death before we took up neo-conservative economic policies

Blame the neo-cons!

But overall I found the comments thread quite a useful wee debate. Someone also pointed out the difference between correlation and causation.