The Arcane Art of Predicting Recessions -- By Prakash Loungani, Assistant to the Director, External Relations Department, IMF

December 18, 2000

Views and Commentaries for 2006 | 2005 | 2004 | 2003 | 2002 | 2001 | 2000 | 1999 | 1998

The Arcane Art of Predicting Recessions

By Prakash Loungani
Assistant to the Director
External Relations Department, IMF

Financial Times
December 18, 2000

Spotting the arrival of the next recession—still considered a distant possibility by most—is once again becoming a spectator sport in financial circles. But if past performance is any guide, a recession tends to arrive before the forecast.

Even though the US has avoided a recession for nearly a decade, many other countries around the world have been less fortunate. How well did private forecasters do in predicting recessions in these cases? Quite simply, the record of failure to predict recessions is virtually unblemished.

Only two of the 60 recessions that occurred around the world during the 1990s were predicted a year in advance. That may seem like a tough standard to impose on forecast accuracy. Maybe so—but two-thirds of those recessions remained un-detected seven months before they occurred.

If this still seems unreasonably harsh, we can lower the bar even further. As late as two months before each recession began, about a quarter of the forecasts still predicted positive growth for the country concerned. In addition, they were too optimistic in 50 out of the 60 cases.

This "predictive failure" is well known in the case of US recessions. Even Alan Greenspan, chairman of the Federal Reserve, told his colleagues in late August 1990—a month into a recession—that "those who argue that we are already in a recession are reasonably certain to be wrong".

Earlier US recessions were also missed by forecasters, as Victor Zarnowitz, an expert in the assessment of economic forecasts, pointed out in a 1986 paper. The evidence for other countries shows that US forecasters are not alone in their inability to predict recessions. Indeed, in absolute terms, the magnitude of forecast errors tends to be larger for developing than for industrialised countries. However, it must be pointed out that the economic growth of developing countries is much more variable than that of industrialised ones.

Why are recessions so difficult to forecast? One theory is that the information needed is lacking: forecasters either do not have access to real-time information or lack reliable models for translating information into predictions of a recession.

Another is that there are few incentives for producing an "outlier" forecast-a recession or a period of strong growth. For example, some researchers and private forecasters argue that the incentives are tilted against predicting a recession.

Mr. Zarnowitz wrote that "predicting a general downturn is always unpopular and predicting it prematurely—ahead of others—may prove quite costly to the forecaster and his customers". Gary Shilling, a private forecaster, stated in an interview this year with Euromoney magazine that "whistle-blowers are unemployable".

This may also explain why forecast revisions—updating an original forecast in the light of new data—are reluctant to depart radically from the original forecast. In other words, the revisions are often too "smooth". This smoothness suggests that incentives are such that smooth forecasts are preferred to "jerky" ones because the former may be more palatable to customers and bosses.

If private sector growth forecasts are of little use in spotting recessions, why not use the forecasts provided free by the official sector? The International Monetary Fund's World Economic Outlook, the World Bank's Global Economic Prospects, released in early December, and the Organisation of Economic Co-operation and Development's Economic Outlook are the main providers of global growth forecasts. Yet there is not much to choose between private sector and official sector forecasts.

Statistical "races" between the two tend to end up in a photo-finish in most cases. This finding casts a new light on allegations made by some that the growth projections of international organisations tend to be over-optimistic. For instance, a 1999 study by the Heritage Foundation, the research think-tank, argued that the IMF's growth forecasts are too optimistic in cases where the countries have IMF programmes. But since the private sector is not subject to the same pressures, it is puzzling that its forecasts end up so close to those of international organisations. At the same time, the pressures acting on private forecasters to lead them towards over-optimism are not faced by forecasters in international organisations.

One possible explanation for the similarity of the predictions is that private and official sector forecasters feed on each other and—in many countries—are heavily reliant on government forecasts. Exuberance on the part of governments may affect both private sector forecasts and those of international organisations.

So, when it comes to forecasts, it is very much a case of 'buyer, beware!' Read the fine print: often the 'story' being told to explain the outlook and the associated risks is much more important than the forecast itself.


Public Affairs    Media Relations
E-mail: E-mail:
Fax: 202-623-6278 Phone: 202-623-7100