Mr Swan's boast is that he has come up with savings worth $7.3 billion in the coming financial year, whereas his new spending totals only $5.3 billion.

That's true, but the $7.3 billion in "savings" includes $1.9 billion in tax increases.

The tax measures involve the increased excise on alcopops, the increased luxury-car tax, removal of the exemption from oil excise for Woodside's petrol condensate, tightening of the fringe benefits tax and an increase in the airport passenger tax.

So Mr Swan's total spending cuts only just cover the cost of his new spending, and only in the first year.

In the following year they fall well short of covering it.

In which case, how did Mr Swan manage to come up with a budget surplus for the coming year as high as his target of at least 1.5pc of gross domestic product, without relying on the $3 billion increase in expected tax collections since the previous estimate was made during the election campaign?

By falling back on no less than $2.4 billion in non-tax "parameter and other variations" since the election. This is mainly a big increase in the dividend received from the Reserve Bank.

The point is that, given the assist from tax increases and changes in non-tax parameters, Mr Swan was able to achieve his 1.5pc target without bothering to cut spending by more than needed to just cover his new spending plans.

That's how hard these guys tried. What's more, the target they set themselves - 1.5pc of GDP - was never going to be a stretch.

The budget outcome hasn't fallen below 1.5pc for the past four years.

Believing their rhetoric, I had assumed the emphasis in their target of at least 1.5pc was on "at least".

In fact, they got above 1.5pc only by means of the $3 billion upward revision in their estimate of tax receipts. With that assist they got to 1.8pc of GDP - hardly a stratospheric figure. That's how hard they tried.

When you look through the list of their spending cuts you find a few big ones but thousands of rats and mice. They got a lot of their savings through their imposition of a 2pc efficiency dividend on government departments, their decisions not to proceed with the Howard government's Access Card and to cancel the Libs' Opel contract for rolling out broadband (while not yet having any cost to include for the contract that will replace it).

Now get this: their spending cuts have been pumped up to look good in 2008-09, but fall back sharply in the following years.

Next financial year, tax increases account for a quarter of their total "savings" of $7.3 billion.

By three years later, however, the tax increases account for almost half the total savings. The tax increases grow while the spending cuts contract.

In the budget papers' list of their savings measures, spending cuts too minor to list separately total $2.4 billion, or almost half their total spending cuts. By the following year, however, the $2.4 billion has collapsed to just $800 million.

Why? Because so many of the cuts are once-only measures. Consider this list, starting with a once-only "special dividend" from Australia Post of $150 million. (This is a non-tax revenue measure, not a spending cut, but even so it was a decision of the razor gang.)

They save $100 million in the budget year by deferring to the following year a measure to align the pay-as-you-go instalments of certain firms.

The decision to raise the Medicare levy surcharge's thresholds will save $230 million in the budget year in reduced health insurance rebate, but the $200 million cost of lost surcharge payments doesn't hit until the following year.

In Defence, they'll make a once-off saving of $370 million in the budget year because of a decision to establish an operations reserve. They'll save $190 million in the budget year because of "provisioning for future defence requirements", but this will increase their costs in subsequent years.

They'll save $300 million in the budget year by reallocating funding of the Higher Education Endowment Fund to the following year. They'll save $50 million in the budget year by pushing into the following year spending on starting an inland rail corridor.