Skip to main content

The federal budget lands Tuesday, and Canadians will be watching to see what fiscal treats Finance Minister Bill Morneau offers up as the Liberals look ahead to the October election.

A lot less attention will be paid to the economic assumptions underpinning the budget, particularly the forecast for GDP growth in 2019 and beyond.

That’s unfortunate because forecasters are notoriously bad at predicting the future – in Canada and most other developed countries. Every percentage-point miss means billions of dollars in tax revenue that Ottawa won’t have in order to pay for its promises.

Consider last year’s budget. The government projected the economy would expand by 2.2 per cent in 2018. Instead, it grew just 1.8 per cent. Its forecast for 2019 was 1.6 per cent.

But don’t blame Ottawa for last year’s miss. The Finance Department’s estimates are based on a survey of 14 private-sector economists. Ottawa stopped using its own forecasts in 1994 amid concerns about possible manipulation of data to suit government policy objectives. The government uses the forecast for GDP and inflation to estimate government revenue, expenses, as well as the expected size of the deficit.

The problem starts with the science of forecasting, which is serially unreliable, particularly further out on the time horizon. Forecasters are not very good at predicting what will happen two years down the road, but they get steadily better as they near their target, according to The Economist magazine, which keeps a database of thousands of private-sector forecasts going back 20 years in Canada and 14 other wealthy countries. The data shows that forecasting a year out, they’re off by an average of 0.8 percentage points. And with 22 months of lead time, the miss widens to 1.3 percentage points.

If a recession is lurking this year or next, don’t bet on many forecasters predicting it.

Accuracy problems are not confined to the private sector. Central banks, which operate some of the world’s most sophisticated economic models, also get it wrong – a lot. The Federal Reserve has repeatedly overestimated how fast the U.S. economy would grow since the previous recession, according to the New York Times.

“They keep having to revise their forecasts downward, only to discover that they didn’t go far enough down,” the NYT’s David Leonhardt pointed out.

The gap between reality and the Fed’s forecasts has widened to six percentage points since 2011. That’s US$1.3-trillion of predicted economic activity that never actually happened.

The Bank of Canada has demonstrated similar irrational exuberance. For example, the central bank didn’t predict the recent stalling of the Canadian economy – even weeks before it began. In October, the bank said the economy would expand 2.1 per cent in 2018. That was pared to 2 per cent in January of this year. The final figure came in at 1.8 per cent – largely because the economy barely grew in the final three months of the year.

It is another example of what Bank of Canada Governor Stephen Poloz has called “serial disappointment” in the pace of the current economic expansion, which began in 2009.

Behind the disappointment, however, is serial optimism. Based on an analysis of 20 years of data, a 2017 Bank of Canada staff report found that it is typically 1.6 percentage points off the mark with its first GDP forecast for any given year, but that gap narrows to 0.6 percentage points as new information comes in. The central bank downgrades its forecasts roughly two-thirds of the time, according to the report.

So why are forecasters so often wrong? The main reason is that predicting GDP is an inherently tricky business. There is the lag time on data collection. And no economic cycle is quite like the previous one, particularly as new technology emerges. The Bank of Canada, for example, has repeatedly overestimated how much exports and business investment would contribute to future growth.

There are plenty of other reasons why the performance of the economy continues to let people down. Traditional measures may be failing to pick up the full impact of the shift of the economy online. The aging population is generally slowing growth longer term. Inequality may also be halting the spread of economic gains to lower-income groups, who are more likely to spend them.

All of that is cold comfort for Mr. Morneau, who must base his budget on information he likely won’t have until next year.