Skip to main content

Around this time last year, Stephen Poloz and the Bank of Canada were confidently predicting a pretty decent 2015 for the Canadian economy.

Gross domestic product would grow at an annual rate of 2.5 per cent in the first three months of the year, according to the central bank's April 2014 Monetary Policy Report.

Over the next four forecasts, as the oil price collapse worsened, the bank would cut its forecast to 2.4 per cent, then to 1.5 per cent, and finally to zero by the time the first quarter was over.

The central bank still got it wrong. Statistics Canada reported in late May that the economy actually shrunk by 0.6 per cent in the first three months of the year – a quarter Mr. Poloz famously dubbed "atrocious."

The Bank of Canada missed the mark by more than three percentage points.

It wasn't alone. Most private sector economists also failed to predict the depth and speed of the oil price collapse, and the resulting hit to Canada's resource-dependent economy.

The episode is a reminder about the severe limitations of economic forecasting.

Even armed with the most sophisticated models, economists are more wrong more than they are right. A Ouija board might produce more reliable forecasts.

Too often, it's a one-way error. Research suggests that forecasters are serially overoptimistic, especially when the economy is heading south.

Even the best aren't particularly good. Both the International Monetary Fund and the Organization for Economic Co-operation and Development did extensive analysis on why they missed the 2008-09 global recession.

Their conclusions are disquieting.

"There are periods where forecast errors have been substantial and relatively widespread," according to a 2014 background paper by the IMF's Independent Evaluation Office. "Large forecast errors tend to be particularly clustered around regional or global recessions."

The OECD reached a similar conclusion. The organization's research staff "didn't predict the scale of the collapse in economic activity between 2008-09 and then overestimated the speed of the subsequent recovery," according to an OECD summary of its postmortem.

Forecasters, it seems, let us down when we need them most. The forecasts that governments, central banks, businesses, investors and individuals rely on to make sound decisions may be completely unreliable – right up until the moment when the future becomes the past.

The easy explanation for why economists are chronically wrong is that forecasting is extremely hard work. Econometricians are trying to track how a vast array of variables interact, and then predict how those movements will affect future economic output.

The U.S., for example, produces 45,000 economic indicators every year – everything from housing starts to crop plantings. The private sector generates millions more.

Forecasters will typically use many different tools, continually searching for the best predictors of the future. A 2013 Bank of Canada study found that combining forecasts from different models, and then giving more weight to the ones that worked well in the past, "tends to improve accuracy."

It's simply too easy to blame an unknowable future for chronic missed forecasts.

Subjectivity seeps into the craft, even for those who do it best.

Private-sector forecasters may sometimes have a disincentive to acknowledge the economy is poised to tank. Downturns are bad for business.

That shouldn't be the case for elite governmental and quasi-governmental organizations such as the Bank of Canada, the IMF and the OECD.

And yet the IMF report acknowledged that staff at the fund "may lack incentives to predict recessions." At the IMF, economists who cover particular countries or regions must constantly justify their forecasts, particularly when they differ from the consensus.

"Desk economists can minimize the amount of scrutiny their forecasts will get by not differing significantly from mainstream opinion," according to the internal IMF paper. "The cost of forecasting a recession that does not materialize may be perceived as higher than that of having wrongly predicted a boom."

The OECD similarly warned that "groupthink" and a herd mentality may be influencing their economic and financial analysis.

Thankfully, forecasters have a secret weapon to fix past wrongs – revisions.

Stay tuned for a whole new set of forecasts from Mr. Poloz and the Bank of Canada, due out July 15.

Just don't expect them to hold true.