Well done to The Australian Financial Review for its “courage” in this morning’s editorial in calling for a 10-year outlook for the federal budget.

“Mr Swan should grasp the opportunity presented by the upcoming Mid Year Economic and Fiscal Outlook to begin a commitment to 10-year budget forecasts like those produced by the bipartisan US Congressional Budget Office. That would make it possible to ascertain what lies beyond the budget’s four-year forward estimates, explain how future governments are going to pay for Labor’s new, big, unfunded social programs, and force policymakers to explain the sensitivity analysis used in their budgetary projections at a time of increased uncertainties about the global economy.

“Australia is fortunate its macro policy levers are not as constrained as in Europe and the US, with Reserve Bank of Australia governor Glenn Stevens noting this week that it still has some ‘ammunition’ left. But with its overnight cash rate already just 25 basis points shy of its 2009 emergency low, the Reserve Bank needs to be careful about using this.”

What the AFR is really calling for is the federal government to make itself an easy target for it and its main rival in national carping, The Australian, for no possible benefit to anyone. To be blunt, it’s a crock of shit.

Perhaps Michael Stutchbury and friends don’t read much news from the US. The rabid Tea Party types in the GOP have been vilifying the “bipartisan” CBO for years now, every time it produces data that doesn’t fit with the tax cuts for the rich/slash-and-burn for everyone else narrative. And the CBO missed the GFC and its terrible impact on the US economy, especially on spending and the size of the budget deficit, not to mention the way the US government debt has ballooned. Nor would the forecasts have picked up the surge in unemployment and the sharp fall in bond yields to record lows.

But then that’s the problem with forecasting when so many variables are outside your control. That’s why under the current budget framework, Treasury estimates revenues and growth for the next two years based on forecasts, but then reverts to trend estimates for the two years beyond that. The AFR wants to see that process extrapolated a further six years into the future.

Doing that would simply magnify the errors that are inevitable in relying on trend estimates, even over a short period. In the 2009-10 budget, for example, Treasury forecast that in the first year of trend estimates, 2011-12, the government would obtain $310 billion in cash receipts. But the actual 2011-12 outcome was $329 billion. So Treasury was out by 6% in just 3 years. How much would it be out over eight years? MYEFO next week is likely to show that even revenue forecasts from May this year were out by a significant margin.

If you go back to 2002 and assume that the then Howard government produced a 10-year forecast, consider what it would have missed: the surge in demand from China and the first and second resources booms, the record investment boom, a global financial crisis and a European depression, not to mention sundry Middle Eastern wars. The destabilising impact of those events on the economy, demand, commodity prices (remember Rupert Murdoch demanding an attack on Iraq because it would deliver $20 oil?) employment and interest rates, and the surge in the value of the Australian dollar, would have all been missed.

The 10-year forecast would have gotten Australia’s economic future radically wrong, projecting the so-called “great moderation” off into the future, when it almost sank the global economy and crippled the financial system of the world beyond repair.

Each subsequent update of that forecast would have missed most of that until around 2006 -07, but a 2007 forecast would have missed the dramatic impact the collapse of Lehman Brothers. In 2008 it would have under estimated the impact of the recovery in China and the re-emergence of the resources and investment boom, and the recovery in the value of the Australian dollar.

And Treasury does have a periodic stab at longer-term forecasting. It’s called the Intergenerational Report. It’s a major undertaking, and tries to grapple with issues like differing population growth trajectories — another key factor that has proven immensely difficult to accurately forecast and which would make 10-year forecasts ridiculous.

This fetishisation of economic forecasts is bizarre. The successive downgrades by the IMG of its global forecasts this year (and previously) should be reason enough to treat all forecasts as a guide only and nothing more. The AFR‘s real purpose is to bash the government, to lash out at “Labor’s new, big, unfunded social programs”, all of which are well beyond the next election and even, in the case of Gonski, extend well beyond even a 10-year forecast.

But here’s a good thought experiment: if the AFR will forecast its average size for each of the next 10 years, as well as its average ad ratio, readership and cover price (and ad rates), then we’ll welcome a 10-year forecast for the federal budget. And, of at the end of the 10 years, when the AFR is out of businesses or radically different, smaller and less important, we will be able to assume that the paper, its editors and executives didn’t know what they were doing, and should have known better.