Budget papers fudged to boost Rudd's stimulus effect
"It will always be difficult," said Ken Henry in February 2009, after he was asked whether the government would ever be able to judge the success of its stimulus package.
Just over one year later, that caution has completely disappeared.
About a minute into Wayne Swan's budget speech last week the Treasurer said he was "proud to announce this strategy is working, ensuring our economy has far outperformed the rest of the developed world; Without stimulus, we would have gone backwards".
And the Government's budget papers soberly concur: "It appears that the impact on economic growth of the fiscal stimulus that countries, such as Australia, put in place has exceeded expectations."
So: verdict's in? Not quite.
Certainly, Australia's economic performance has been comforting. Our unemployment rate is around 5.4 per cent. Compared to the United Kingdom, which has an unemployment rate of 7.8 per cent, or the United States, which has an unemployment rate of 9.9 per cent, we're doing pretty damn well.
But the question isn't whether we have done well. It's whether we have done well compared to how we would have done if we hadn't had a stimulus, or if it that stimulus had been smaller.
And here the Government has scored an extraordinary own goal.
Slotted within Budget Paper Number One, Statement Two, Box Four is a graph which purports to be the final word on stimulus packages. The graph measures the size of the stimulus packages of 11 countries. It then calculates how their economies have performed, compared to how the International Monetary Fund one year ago predicted their economies would perform. Using this technique, the graph shows a big stimulus package is closely correlated with good economic performance.
Putting aside whether the causation flows the other way (could dire economic forecasts from the IMF have scared countries into implementing bigger packages?) it's an elegant way of showing stimulus works.
Well, it would be - if the Treasury hadn't cherry-picked the data.
In fact, there's a pretty strong case the Government is being deliberately misleading.
That's because the IMF data on which the Treasury graph is built doesn't list just 11 countries. It's actually a list of the G20 countries - 19 states plus the European Union.
RMIT Professor (and my IPA colleague) Sinclair Davidson ran the calculation again, this time with all 19 countries included. And - surprise! - the correlation disappears. There is no statistically significant relationship between stimulus package size and economic performance.
So why did Treasury pick those 11 countries? They're not OECD countries - there are 31 members of the OECD. And they're not the world's "advanced economies", although Wayne Swan keeps using that phrase. Thirty-four countries make up the IMF's official advanced economies list.
Damningly, the Treasury seems to have deliberately ignored those G20 countries (like Russia and South Africa) which implemented massive stimulus packages but have subsequently had terrible economic performance, and those countries (like India, Argentina, and Indonesia) which had small packages, but have still done well.
It's like ignoring patients that died after they took an experimental medicine, or got better without taking it - but applying to be listed on the PBS anyway.
This is not an academic problem.
Whether the stimulus package worked is at the absolute heart of the Rudd Government's re-election strategy. If the Government can't take credit for Australia's economic performance, it loses its only real policy success.
Governments aren't all-knowing and all-seeing. The stimulus package was a crap-shoot as much as anything else; a massive amount of money dumped as quickly as possible into an essentially random smattering of industries, with the hope the Government could lift the economy by brute force alone.
After all, does anybody believe that home insulation is the pivot on which our economy turns?
Few Governments have had so many chips fall in their favour. The economy is performing well, and, at least until late last year, an overwhelming majority of Australians supported the Government's proposed action on climate change. Yet, even with all this, Kevin Rudd has seen a precipitous fall in his popularity. No surprise Rudd and Swan are falling back on a story about performance during the economic crisis.
Rudd is being forced to run on his economic record, and Treasury's argument that countries which pushed through a big stimulus had a big recovery.
So the fact that Treasury had to cherry-pick data to prove the stimulus worked is not a good look.
It makes you nostalgic for the old Ken Henry, who back in February 2009, said "there are sufficient differences among economies that that sort of analysis … would not be particularly useful".
Shame this modesty couldn't last an election cycle.