Economic forecasting: is Treasury losing its memory?

By David Donaldson

March 9, 2016

Treasury building
Treasury building

The federal government relies on Treasury’s macroeconomic forecasting to create a sustainable budget, but those predictions have been out for several years now.

High turnover, less experienced staff and an over-reliance on formal modelling have contributed to these repeated misses, according to a report released on Tuesday.

While Treasury’s formal macroeconomic models “provide a sound base for forecasting”, they “are applied by less experienced staff in an environment of high staff turnover and short tenure”, warns the review, conducted by Treasury Macroeconomic Group principal adviser Warren Tease.

Treasury has been criticised for having consistently overestimated GDP growth in forecasts since the global financial crisis, causing problems for the federal government, which relies on the predictions for the budget. It also underestimated  growth during the Howard mining boom years, though this of course caused less consternation. This is the third review into macroeconomic forecasting since 2002.

“… models should be a complement to well-informed judgement, not a substitute for it …”

Not enough input from sector experts and senior staff has meant forecasters have come to rely too heavily on models without deeper analytical context. “This is an institutional risk,” Tease wrote.

Judgement is a crucial aspect of the forecasting process, especially “when forecasts are generated in an iterative way by inexperienced staff … models should be a complement to well-informed judgement, not a substitute for it, as now often seems to be the case.”

Models “cannot of themselves incorporate economy-wide themes in a consistent way”, he argues, and despite the importance of the forecasts, there “seemed to be little emphasis on bringing to bear deep sectoral expertise or a deep analysis of major macroeconomic themes” that could be contributed by senior staff.

Sector analysts have short tenure and high turnover, which “limits the staff’s ability to develop true sector experience or technical expertise”, according to the report:

“In this environment it is important that senior staff provide technical guidance to the less experienced staff and, from a forecasting perspective, provide guidance on the major macroeconomic themes over the forecast period. At present, senior staff take a more informal and iterative role in the production of the forecasts compared with other institutions.”

It’s a warning that echoes comments made by the former head of the Treasury and now-secretary of the Department of the Prime Minister and Cabinet, Martin Parkinson, in Laura Tingle’s Quarterly Essay on the erosion of capacity in public sector organisations that “today, in some institutions, smart people look around at their colleagues and find there is no one to talk to, to learn, from, who has experience in delivering real reform”, leading to “a decline in quality of advice and an erosion of capability, to the detriment of good government”.

Treasury’s models’ bias towards assuming stability inevitably make it difficult to factor in fluctuations in commodity prices and trade, which in themselves can be hard to foresee. The general feeling among economists who spoke to The Mandarin last year was that the primary blame for the inaccurate predictions did not lie with Treasury, but with the unpredictability of some of these economic drivers. The report states:

“Commodity prices and the terms of trade are hard to forecast. Treasury has developed a rigorous framework for analysing commodity prices over the long term. However, over the two-year forecast horizon, Treasury now assumes that prices will be unchanged. This is common and reasonable but leaves its forecasts vulnerable to large changes in commodity prices.”

About the author
Inline Feedbacks
View all comments
The Mandarin Premium

Insights & analysis that matter to you

Subscribe for only $5 a week


Get Premium Today