By Andrew Leigh

Studies show economic forecasters get it right more often than a trained parrot — but only just.

In a quest to predict the cycle of boom and bust, a variety of innovative theories have emerged. Some are practical.

The “cardboard box index” is based on the idea that three-quarters of consumer goods are shipped in boxes. Because companies have to buy the boxes wholesale before they ship the product, the theory suggests that boxes are a good leading indicator for surges or slumps in the economy.

Similarly, the “men’s underwear index” suggests that because your underwear is seen by few people, it might be one of the first things that people start to cut back on when the economy turns bad. One commentator compares men deferring underwear purchases in tough times to the way that others might drive their car an extra 10,000 km between services. 

Other theories are more psychological. The “hemline index” proposes that skirt lengths rise in good times and fall in bad times. Advocates of this theory point out that skirts were shorter in the roaring 1920s than in the hard times of the 1930s and that the miniskirt emerged in the boom times of the 1960s.

The “music index” suggests that in hard times we are more likely to listen to songs that are slower, more comforting and more romantic.

And the “centrefold index” suggests that in tough economic conditions, magazine models tend to be taller, older and have larger waists. 

Alas, when researchers have looked carefully into these theories, there’s little evidence they can tell us anything about the future (indeed, hemlines appear to track economic conditions three years earlier). But what happens when we move from folk wisdom to business commentators? Can business pundits predict the future?

The “men’s underwear index” suggests that because your underwear is seen by few people, it might be one of the first things that people start to cut back on when the economy turns bad.

Somewhat mistaken

About every six months since 1979, Melbourne’s The Age newspaper has been asking a group of prominent commentators to estimate economic growth in the future. Alas, they haven’t exactly covered themselves in glory.

In December 1988, The Age asked 42 commentators to estimate how much the economy would grow in calendar year 1989. Their answers ranged from 0.8 per cent to 3.75 per cent. In fact, the economy grew 4.6 per cent that year.

Less than two years later, in July 1990, The Age asked 38 commentators — many of them the same people who had failed to predict that the economy would grow at 4.6 per cent — to estimate the growth rate for the fiscal year 1990-91. Their answers ranged from 0.5 per cent to 2.8 per cent. Ultimately, the economy shrank by 0.8 per cent, a slump then Australian Treasurer Paul Keating described as “the recession we had to have”. 

As a nation, it’s possible that we might have been able to take quicker action to soften the hard landing of the 1990s recession if only forecasters had seen it coming.

In the space of two years, the nation’s business forecasters had been surprised twice. None had predicted how fast the economy would grow in 1989. None predicted that the economy would shrink in 1990-91. In 1998-99, the economy grew faster than all 28 The Age commentators had predicted.

In 2007, The Age summed up its commentators’ forecasts for the following year as: “The good times are rolling.” Then the global financial crisis hit, and Australia grew more slowly than all 21 commentators had predicted.

Pet parrot

To systematically assess the accuracy of a prediction, we need something to compare it against. In the case of growth, economists typically use as a benchmark the naive assumption that next year’s growth rate will be the same as this year’s growth rate. Think of this as the pet parrot approach, in which you teach your parrot to say the current growth rate. Any time someone asks it to forecast next year’s economy, it will dutifully squawk this year’s number.

Do business forecasters beat the parrot? Yes, but not by much.

When it came to forecasting growth, the typical forecaster was off by 0.9 per cent, while the parrot would have been wrong by 1.1 per cent. In other words, business commentators polled by The Age were only about one-fifth better than a pet parrot — and that’s what you get when you average their forecasts (effectively allowing the pessimists and optimists to cancel one another out). Take any single commentator, and the chance that he or she will beat the parrot is slim indeed. 

To their credit, both the [Australian] Reserve Bank and Treasury have recently published estimates of their own forecasts.

Assessing the predictive power of their economic forecasts over nearly two decades, the Reserve Bank concluded that its forecasts of future unemployment and future growth were generally less accurate than the naive estimate.

Only on inflation was the Reserve Bank able to trounce the parrot.

This is an edited extract from The Economics of Just About Everything by Andrew Leigh, published by Allen and Unwin, RRP A$27, NZ$33.

This article was first published in the December 2014 issue of Acuity magazine.