Policy costings: what's real reason Joe's slow off the mark?

So what's really troubling Joe Hockey's belief system and delaying the release of election policy costings - Treasury's estimates or internal Coalition differences on the policies to be priced?

The Coalition has had Treasury's best fiscal guess for 11 days now. As everyone knew, the official PEFO numbers are virtually the same as those used in the economic statement on August 2. A difference of $9 million in the forecast 2014-15 deficit is less than a makeover for a marginal electorate's football ground.

Hockey was getting his denial in early when he declared he didn't believe the economic statement figures, and therefore he can't believe the PEFO figures either. If the man who would be treasurer doesn't believe his department's best efforts, one might assume his first big slash at the Canberra public service will be to sack the entire Treasury and put government bean-counting out to tender.

While the PEFO numbers are apparently not to be believed, education spokesman Christopher Pyne said Joe Hockey and finance spokesman Andrew Robb would consider the numbers and subsequently release their costings "in plenty of time" - but they've effectively been looking at those numbers for the better part of two weeks, and looking won't make the arithmetic change.

The Coalition might also be having problems with its pledge that its costings will be done by the Parliamentary Budget Office. According to The Australian Financial Review, the PBO has costed Opposition Leader Tony Abbott's maternity leave scheme at more than $5 billion a year, or $14 billion over its first three years of operation. Nonetheless, such challenges are sent to be fiddled with - by, for example, excluding public servants from the scheme.

About the only real difference between the PEFO and economic statement is that the PEFO authors include an alternative view of what the unemployment rate might be in the "projection" years of 2015-16 and 2016-17. The ES and the first bit of the PEFO make the assumption that unemployment will suddenly drop from 6.25 per cent at the end of June 2015 to 5 per cent 12 months later. You don't need the computer modelling grunt of Treasury to think such rapid employment growth would be extraordinary - and that the Reserve Bank would react to an economy growing that quickly by boosting interest rates to cool it.

The more credible PEFO "alternative projection" is a more gradual closing of the output gap, with unemployment falling to 6 per cent at the end of 2016 and 5.75 per cent in 2017, which would add $3.2 billion in extra unemployment benefits over those two years. Both sides will no doubt ignore that alternative rather than pluck another $3.2 billion out of the budget air.

The authors signing off on PEFO, Treasury's Martin Parkinson and Finance's David Tune, also go to some effort to explain the difficulty of forecasting the Australian and global economies.

"Against the backdrop of a still challenging global outlook, the Australian economy is expected to transition away from resource investment-led growth towards broader-based growth, although this transition may not occur as smoothly as forecast," the document warns.

In other words, forecasting is a mug's game and even Treasury's best efforts need to be taken, not with outright disbelief, but with some understanding of the craft's limitations. PEFO claims the same sort of confidence band as the Reserve Bank's forecasting efforts. Treasury and Finance are 70 per cent confident that this year's GDP growth will be somewhere between 2 and 3.5 per cent - hence the headline forecast of 2.75 per cent, smack in the middle.

The global factors outside any politician's control could weaken the outlook, but they could also improve it. I don't recall any of the army of commodities forecaster predicting iron ore prices would be enjoying their present lift. Treasury certainly didn't and still doesn't, sticking with a scenario of a steady fall in the terms of trade after being bitten by being too optimistic last year.

The bottom line, then, is that the PEFO numbers may as well be believed as they are likely to be as good or bad as anybody else's. It's always tempting to shop around for the forecast you prefer the sound of, but it's also the folly of a simple mind.