Guest Post: The curious confidence in numbers when officials (mis)count the blessings of social policy

A guest post by Bob Edlin:

We should be dismayed not that Treasury number-crunchers erred with child poverty figures but that the numbers which emerge from their economic models can be so blatantly peddled as the stuff of political certainty.

Just before Christmas Radio New Zealand – along with other media – reported on the Government’s income package, to come into effect on 1 July next year and reckoned to cost $5.53 billion over the next five years.

Let’s keep on eye on that number.

More significantly for headline writers, we were told the package “will lift about 88,000 New Zealand children out of poverty”.

Oops. Maybe not.

This morning Radio NZ says Treasury is admitting it made a “deeply regrettable mistake” by likely overstating how many children would be lifted out of poverty as a result of the families package.

The extent of the error – which also affects National’s package – is still being determined.

Officials had estimated Labour’s plan would lift 88,000 children out of poverty by 2021.

Secretary to the Treasury Gabriel Makhlouf said it is currently remodelling the projected impact of both packages and expects the new numbers to available in late February.

The Dominion Post reduced news of the Treasury blunder to a brief on Page 3 but the Stuff website says the child poverty reductions were over-estimated when the previous National Government delivered its Budget in May last year and the error appears to have carried through to affect the current Government.

Reporter Stacey Kirk expects one of Prime Minister Jacinda Ardern’s first big speeches of the year to include a child poverty focus and says the Government is planning to introduce legislation in early February that will require government’s to set child poverty reduction targets.

In Kirk’s report, Makhlouf says a coding error led to flaws in the Treasury simulations. The extent of any change in the projections on child poverty is still being determined.

Kirk nevertheless seems to have astonishing confidence in the robustness of the numbers that eventually will be spat out when fresh figures are fed into the Treasury model. She writes:

“…the exact number of children expected to be lifted out of poverty was not expected to be known until the end of February.

The exact number?

Bollocks.

Let’s just recognise it will be the Treasury’s best bet, in much the same way as officials are obliged to do their best to project the Government’s Budget deficits, then proceed to constantly revise them as updated figures flow into their considerations.