Stuck in an IMF forecasting fix

The IMF’s recent growth estimates aren’t as dire as media has made out, but they do highlight serious problems with the fund’s forecasting process.

Lowy Interpreter

When the IMF  updated its forecasts for global growth, the headlines were dramatic.  The Financial Times reported that the IMF has 'slashed its forecasts' and that 'the downgrades highlight the gathering clouds around the world economy'. The Wall Street Journal opened its reporting on the update like this: 'investor fears that the end of easy money is at hand are ricocheting around the globe, slamming financial markets and squeezing economic growth prospects from Brazil to Turkey.' 

This sounds like a big deal, but the latest IMF update trims just 0.1 per cent* off the global growth forecast it made in April for the current year, and has taken 0.3 per cent off its forecast for next year.

Emerging economies' growth forecasts, which feature in the dire media reporting, were trimmed by exactly the same amount, and are still comfortably in excess of 5 per cent.

If the IMF's current forecasts come to pass, growth this year will be one-third faster than last year (3.5 per cent against 2.6 per cent) and 2014 should be a bit stronger still. This is certainly a very weak recovery (upswings are often around twice this pace), but this is hardly news.

That said, there is a problem with the fund's forecasting record during this recovery. The IMF has been consistently too optimistic, forcing it to revise its forecasts as reality catches up. Each downward revision, even if modest, creates a melodramatic headline.

The IMF shared the widespread optimism that the economy was recovering satisfactorily after the stimulus spending in 2009. For 2012, its initial forecast for global growth was just over 4 per cent; the outcome was 2.6 per cent. Similarly, the initial forecast for this year was just over 4 per cent; it is now 0.6 per cent lower at 3.5 per cent. Next year started at 4 per cent and now stands at 3.7 per cent. 

It is intrinsic in the IMF forecasting process — in principle independent of national authorities but in practice influenced by official forecasts — that it will be too optimistic during slow times, being careful not to 'talk the economy down',

But if the IMF had correctly forecast Europe's 1 per cent fall in GDP in 2012, could policy makers have left their austerity policies unmodified? For the current year, the initial forecast of 1.4 per cent for Europe made in April last year took pressure off policy makers; would they have appreciated it if the IMF had, at that stage, produced its current 2013 forecast of 0.3 per cent?

Of course the context of the initial forecasts was different: the IMF shared a widely held view that the global economy was recovering. This, however, doesn't absolve the fund. Official growth forecasts (including the fund's) are different from, say, weather forecasts, because they influence the outcome.

Even without political pressures, it's hard to get growth forecasts right. As a recent review of (Australian) Treasury macroeconomic forecasting found:

In common with many forecasters in Australia and overseas, Treasury has experienced mixed success forecasting over this period. In particular, the identification and prediction of major turning points in economic activity recently has been a failing of most forecasters around the world.

The models cannot capture the complexity of the real world and in any case don't incorporate the sorts of shocks which throw forecasts far off track. But forecasting is still the basis of policy, so we have to go on doing it.

So are we at a forecast inflection point now, needing a significant revision to the baseline projection, justifying the alarmist headlines?

Much has been made of Fed Chairman Bernanke's statements on quantitative easing, while others point to China's liquidity glitch. QE was always going to be tapered. After all, this tapering just means a slowing down of the increase in the Fed's holdings of government securities, leaving an actual winding back of its holdings to a much later date. Bernanke's commitment remains firm: leave the policy interest rate (which is the instrument that matters most) at zero until unemployment falls to 6.5 per cent (compared with the current 7.6 per cent). The prospect of a somewhat faster US recovery next year is good. Growth this year is being held back by a 2 per cent of GDP fiscal contraction. Easing off on this austerity next year would give the US recovery the opportunity to follow a more normal pattern of acceleration. 

The liquidity glitch in China was, in itself, unimportant. For all the talk of China slowing, the step down from 10 per cent growth to the current rate of around 7-8 per cent occurred five years ago, and the overwhelming majority of forecasters still have a forecast starting with '7' as their central probability, matching the government's target. Credit tightening might reduce growth to the bottom of the 6-8 per cent band, which is now the normal underlying pace of growth, but slower growth will provoke a policy response.

Europe is still mired in the problems of the periphery, with chronic gloom firmly established as the 'new normal.' UK growth is still constrained by austerity. If financial markets continue to panic over QE, capital outflows from emerging countries will present some nervous moments for policy makers, but they are better prepared than in the past.

In summary, the IMF was too optimistic when it released its April forecast, and has belatedly trimmed this back. The global recovery is lamentably slow, but there seems no good reason for the dramatic headlines.

*The Fund presents its growth figures in two ways: as year-on-year growth rates and as fourth-quarter-on-fourth-quarter. The latter are used here as they give a better representation of the shape of the cycle (for discussion, see this RBA appendix).

Originally published on the Lowy Interpreter. Republished with permission.

Related Articles