Stop the rot on Treasury forecasts

Enough is enough. Allegations of political interference in the process by which budget forecasts are put together risk undermining public confidence in Treasury and government.

The idea that Treasury presents the government of the day with a range of forecasts from which it can cherry pick a figure to suit its preference was categorically dismissed by Treasury Secretary Martin Parkinson this week.

The allegation is just plain wrong, reflecting a profound ignorance of the way Treasury forecasts are in fact assembled.

The truth is that Treasury forecasts are assembled according to a very strict and rigorous process. Macroeconomic forecasts are done quarterly, using updated data from the National Accounts, by Treasury’s Domestic Economy Division. The process begins with a business liaison program, contacting up to 120 businesses. Meetings are also held with economists at the Bureau of Resource and Energy Economics to discuss commodity prices. When national accounts are released, Treasury forecasters meet with the Australian Bureau of Statistics to discuss. Perhaps the most important step in the process is a quarterly meeting of the Joint Economic Forecast Group which includes Treasury, the Reserve Bank, the ABS and other central agencies.

A separate unit within Treasury is also responsible for taking the macro forecasts and translating them into revenue forecasts. They do this in close liaison with the Tax Office which can provide confidential information on tax collections.

The forecasts which appear in the budget are the culmination of months of work, consultation and cross checking.

A healthy degree of scepticism when it comes to macroeconomic forecasts is always advisable. 

But the wholesale dismissal of budget forecasts like we’ve seen from the media and the Opposition is unwise and risks undermining confidence in one of the central pillars of our economic success.

In his annual post-budget reply, delivered to a lunch of the Australian Business Economists in the ballroom of the Four Seasons hotel in Sydney on Tuesday this week, Dr Parkinson said it may be desirable for Treasury to join the Reserve Bank in producing graphs showing the confidence intervals around its forecasts.

Then, Dr Parkinson quipped, it may only be necessary for him to explain once every few years why the forecasts turned out to be wrong, not every year like he has to now.

Treasury could do more to educate the public about the way it assembles its forecasts. But ultimately it is politicians who determine the level of confidence the public will have. Taking forecasts and treating them as gospel vows to return to surplus has, by now, proved misguided, to say the least.

A competent Treasurer should be able to explain to a worried public that forecasts are just that, forecasts. Wayne Swan has faltered, and Joe Hockey’s bull-at-a-gate approach to undermining confidence in Treasury forecasts is just as bad.

Until humans invent the capability to see the future, forecast will always be prone to significant error.

In the meantime, we have Treasury. 

And while far from perfect, it’s not doing too bad a job.

A review of Treasury’s forecasting success, released in December last year, found that over the past two decades, Treasury forecasts had exhibited a forecast error of around $8 billion a year in 2011-12 dollars. 

Out of a $350 billion plus budget, that’s not too bad. But it does mean that surplus predictions of less than $1 billion must be taken with a grain - nay a shaker - of salt.

It is true that Treasury underestimated revenue from mining boom mark I and has subsequently overestimated revenue post global financial crisis. But, according to the review, it has been no worse than private sector forecasters, including Access Economics, in this regard. Nor has it done any worse than government forecasting agencies world wide.

Not a bad effort when you remember the world has just suffered its biggest financial shock since the Great Depression. How easily we forget.