Economic forecasting – voodoo or science? Part 4: The bottom line

Economic forecasts are not complete nonsense or impossible per se. Indeed, it is the way they are usually drawn up these days that can give rise to substantial problems.

• The quality of the underlying data is often poor. Or it is manipulated or used selectively to support a previously fixed statement.
• Forecasters try to sound like experts by formulating their predictions more precisely than is actually possible. Indeed, in order to maintain their image as experts who hold superior knowledge, they avoid mentioning possible margins of error, although this additional information would be very helpful for their target audience.
• Economic forecasts are often made for a period of 1-2 years. However, during this period of time, the complexity of the economic system can often lead to surprises and therefore considerable deviations from forecasts.
• A decisive factor for the success of some predictions is their ability to trigger a feedback loop with the economy.

These shortcomings are also partly linked to unrealistic public expectations, which are considerably reinforced by economic forecasters themselves. Erroneously, the general public tends to believe that a precise forecast is also a reliable one. If an economists were, for example, to express a prediction in this way, “Based on certain assumptions, I expect economic growth to be between 0.5% and 1.5% with an 80% probability for next year”, his competitors, the press and clients would all react in the same way – it is not precise enough, cannot be taken seriously. They want statements like: “the gross domestic product will grow by 1% next year”, even if this expresses a clarity, which does not exist in reality.

This favours ‘hedgehog forecasts‘, which may be formulated in precise terms, but are incorrect. Any addressee of a forecast who expects a degree of precision that is not possible should not be surprised if forecasters put on a “prediction show”, which has absolutely nothing to do with real-life developments. However, forecasts of this kind are of absolutely no use for making investment decisions. For this reason, decision-makers who are evaluated on the basis of their investment results generally make ‘fox forecasts’ to form their expectations for an uncertain world. These are relatively inexact statements about trends, which are adapted to current developments on an ongoing basis.

Criticism – often formulated in a populist manner – about the poor forecasts produced by analysts and economists is very often unfair. On the one hand, they are expected to do the impossible, when they are expected to deliver precise statements for the medium term on the basis of incomplete data. And then when they fail to achieve the impossible, they are portrayed as incompetent idiots. Nevertheless, we can reproach the forecasters for taking part in this game and for feigning more expertise than realistically possible.

Economic forecasts could be drawn up and presented in a more scientific manner than is currently the case. However this would require forecasters to admit that the outcome is strongly influenced by the quality of the data and by complex interactions. Currently, a theatre production is being staged for the general public, in which the experts place themselves in the limelight, certain opinions are reinforced and the forecasters are the actors. Thus, the majority of published forecasts are more like a voodoo show than like impartial science.