American news networks really do employ a political commentator called Krystal Ball – but there are good reasons why economists can’t foresee the future. So why do governments and businesses still frame key decisions around their predictions? Alan Shipman provides some answers. 

Now that the Eurozone is threatening to disintegrate, it’s worth remembering why the EU’s core members set it up, and 17 countries eventually joined it. The single currency was designed to strengthen Europe’s single market for goods and services, by reducing transaction costs and removing risks caused by fluctuating exchange rates. It was intended to promote a single European capital market, ending the national fragmentation of stock exchanges, and lowering borrowing costs by giving the EU a bond market comparable to the US’s. Since Europe’s economic space would now be much larger than America’s, there was a very real hope of the Euro replacing the Dollar as a global currency, unlocking further gains through the profits of the European Central Bank. 

To the main objection – that Eurozone members would lose the power to correct a cost disadvantage by devaluing their national currency, and so would have to ‘adjust’ through much more painful wage and price reduction – the Euro architects had three answers.  

First, there would be economic ‘convergence’, causing members’ costs and productivities to move closer together. Second, currency devaluation no longer worked, since it caused inflation that would soon leave ‘real’ costs as high as they were before. Third, as the monetary union deepened, external imbalances between member states would no longer be a policy concern. The persistent trade deficits run by some American states and Canadian provinces (or between Wales and England, or Italy’s Mezzogiorno with Lombardy) have ceased to matter since they adopted the same currency, because of automatic balancing flows through national businesses and budgets.  

'forecasters aren’t measurably better than a random prediction'

Some advocates of EU economic and monetary union even called for the early admission of peripheral countries like Greece and Portugal, without waiting for them to meet all the ‘convergence criteria’ laid down by the European Commission. Their prediction was that convergence would accelerate once countries joined the union, so it made sense to let them in to speed up their move towards German standards of budgeting and inflation, on which sustainable membership ultimately depended. 

It doesn’t augur well… 

Chances of complete accuracy 'infinitesimally small'
Before any cross-Channel laughter at the gap between these predictions and today’s grim reality, it’s worth reflecting on the success of economic forecasts closer to home. As recently as March, the Office for Budget Responsibility (OBR) was forecasting that the UK economy would grow 1.7% in 2011, and 2.5% in 2012. Its revised forecast, issued in November, knocks these numbers down to 0.9% in 2011 and 0.7% in 2012. Because slower growth means a wider budget deficit and higher public debt, it’s a drop that severely affects Treasury plans, forcing the Chancellor to switch from Plan A to Plan A*, Plan A+ and beyond. 

The OBR prepared the ground for this lapse in foresight in its first Forecast Evaluation Report, published in October – when, ironically, it was having to explain why it under-estimated growth and was thus too gloomy about public debt in 2010/11. The Report admits that “the chances of any economic or fiscal forecast being accurate in every dimension are infinitesimally small".

Like any good forecaster, it tries to dissuade users from paying too much attention to the central numbers, reminding them of the margin of error it places around those percentage changes, its demonstrations of how sensitive these are to movements in variables it can’t forecast (like interest on the government’s debt) and unforeseeable shocks (like Eurozone meltdown), and its commentary about downside risks. 

OBR members were recruited because of their distinguished records in economic forecasting. But these were mostly compiled during the long, unusually steady upturn at the start of the century, now known as the ‘great moderation’; none had foreseen the economic step-change in 2007-8. Between its crisis points, an economy’s growth and inflation can be quite reliably forecast by just taking last year’s data as the best guide to next year’s, without using any elaborate models. When crisis strikes, it’s often when the ‘fan charts’ are at their narrowest, so even the forecasters’ worst-case scenario doesn’t capture the cliff-edge ahead.  

Wrong for all the right reasons
 While seldom getting it right, economists have many good reasons for getting it wrong. In the OBR’s case, there have been revisions to the data that feed into their forecast: the Office of National Statistics now identifies a deeper fall in productive capacity in 2008-9, meaning there is less of an ‘output gap’ to generate growth in the forecast period. There have been unforeseeable adverse events in the Eurozone, which changes the scenario for UK exports and financing costs. And, of course, a team that was still unwrapping its computers in temporary Treasury rooms in 2010 has now had a year in which to build and refine its forecasting models. 

On top of this, government and business strategy are affected by announcements from the OBR and other influential analysts, causing modification to the processes they’re trying to predict. In the middle of an upswing or downswing, forecasts may be partially self-fulfilling – reinforcing a belief that things are getting better or worse, so that people change their behaviour to make it so. But when a forecast is especially upbeat or doom-laden, it can spark behaviour-change that undermines it. Businesses get wary of euphoria, reining-in their investment if they’re told it’s about to boom. Governments panic at the prospect of recession, and rush to stimulate growth if the forecast says they’ve stalled it.  

Understandable scepticism over the OBR’s macro-economic forecasts contrasts with predictions of the effects of tax and benefit changes, regularly put forward by the Institute for Fiscal Studies (IFS), from which its main members were recruited. The IFS verdict on who will be better and who worse off after each Budget, and how it will affect the government’s deficit target, is generally accepted by the media and only challenged by ministers whose sums it questions. But the IFS has an easier task, because it only forecasts the effect of a policy change on its tax-and-expenditure model, with all other policies and external variables held constant. The OBR has to forecast the effects of policy change on the actual economy, with all the extraneous events, other policy changes and feedback loops in full flow as it does so.  

Forecasters 'experts at explaining away incorrect predictions'
If economists are ever unsettled by their paucity of foresight, they can take comfort from the track-record of other social sciences. A century on, tribute is still paid to Norman Angell, the political scientist who – in 1910 – published an analysis that appeared to show there could never be another large European war. (In 1921 he published a follow-up, explaining that the Great War’s destructiveness proved his thesis).  Sociologists, meanwhile, have repeatedly predicted,among other things, the rise of a ‘leisure society’ with minimal work-hours, the inevitable rise of comprehensive insurance-based welfare states, and the abandonment of marriage as an institution. And in case technological forecasters are getting complacent, where are our ultrasonic dishwashers and queue-less supermarket checkouts? 

So why do we keep listening to forecasts, and why do governments and businesses still frame key decisions around them? The forecasters would say it’s because theirs is still the best guide to the future, and that what happens next will usually be captured somewhere in the fringes of their fan chart. But psychologists have another explanation, uncomfortably confirmed (well before the financial crisis) by Berkeley’s Philip Tetlock. Tetlock’s long-term studies show that forecasters – whether predicting economic growth, electoral contests or next Saturday’s football results – aren’t measurably better than the average (or a random) prediction. But they’re experts at getting their correct predictions noted, and suppressing the incorrect ones or explaining them away.  

Economists, often frustrated at not being viewed as natural scientists, should be thankful they are not judged by the same exacting standards. In September six Italian seismologists went on trial for playing down the risks of an earthquake, which then struck with a severity that caused over 300 deaths. Many more lives are likely to be blighted by the slowdown that the economic forecasters missed. They won’t be swelling the prison population, any more than they’ll be accurately predicting what it’ll rise to in 2013. 

Alan Shipman 13 December 2011

Alan Shipman is a lecturer in Economics at the Open University. He is responsible for the modules You and your money:personal finance in context and Personal investment in an uncertain world, part of the Open University's foundation degree in Financial Services. 

Cartoon by Gary Edwards