Given the salubrious set of economic circumstances the government faces , this was a shamefully loose budget. The economy is growing apace, unemployment is below 5% and falling, and the terms of trade are at highs not seen since the late 19th century. Inflation has reared its ugly head too, already running above 3%.

Yet the budget anticipates a large deficit this coming financial year of $22.7 billion, $10 billion bigger than the government forecast just six months ago.

Whether you’re a Keynesian or a liberal (the real sort), those figures scream “cut” and “save”.

The government’s vaunted $22 billion in saving are over four years, and even include the flood levy, which is patently not a “saving” but an increase in revenue.

The government has also been prattling on about how the real growth in spending is low over the next few years. It is true that forecast annual growth in real spending is only 0.5% this coming year and below 2% for the out years, which in ordinary circumstances would be relatively impressive.

But even if we believe these lower growth rates will eventuate, should spending be growing at all? In 2008 and 2009, to combat the North Atlantic financial crisis, the government increased spending (with borrowed money) by about 18%, more rapidly than any time since 1975.

Yet far from receding, the level of real spending has flourished from its newly elevated position. I thought Keynesian stimulus was supposed to be “timely, targeted and temporary”! That’s the mantra, yet 2½ years after Lehman brothers collapsed, we’re still “stimulating”. As Japan has found, once countries latch onto the Keynesian teat it’s difficult to get off.

Notwithstanding the paltry size of the “savings”, the government has promised to spend $18 billion on new “initiatives”.  Some of these are obviously bad, like extra welfare for “families” and set-top boxes for pensioners. But commentators have embraced the new spending on mental health and “skills”.

I agree spending money on health and skills is better than giving it in welfare to people earning $150,000 a year. But we need to ask questions about the effectiveness of such spending, especially when the functional areas receiving the new money have titles equivalent to motherhood.

The $2.2 billion on mental health, for example, includes many “prevention and early intervention” measures with little rigorous testing of the evidence of their effectiveness. Also, it is not clear to me why mental health needs to be treated differently from other ailments. The services of GPs, psychologists and psychiatrists are already publicly funded.

Pouring money into “skills” is also popular, but again little hard evidence has been provided. Governments have been spending ever more on skills and training for decades yet the quality appears to have declined. For the vast majority of jobs, it is not credentials that matter but grasp of English, basic mathematics and common sense. The most important training occurs “on the job”, and the sooner that can start the better.

It should be mandatory for the budget to contain footnotes that point to evidence, academic or otherwise, that supports the new spending programs. Ideally, programs should have goals that are observable. By how much will new spending on mental illness reduce its incidence and by when, for example?

Ultimately, the test for new spending should always be whether the money is not better off in the hands of low-income earners though a higher tax free threshold, say. But this alternative is hardly ever mooted.

If we’ve learnt anything over the past few years (indeed past six months) it is that forecasts are hopelessly inaccurate. Yet we continue to obsess over whether a surplus is scheduled in two years.

This is partly the fault of the Howard government, whose Charter of Budget Honesty mandated a four-year forward estimates period. This development has distracted attention from what is happening now and next year, vastly more relevant and informative pieces of information.

So how might the process be improved? Only projects that start in the coming financial year should be included in the forward estimates. That would stop the endless games of shuffling expenditure between years to massage whatever particular outcome the government of the day desires.

The government has blown its last chance to have the sort of budget we should have got in 2008. Just imagine how excruciatingly pusillanimous the two budgets are going to be in the years leading up to the next election.