When is a ‘forecast’ not a forecast?

The answer is: when it’s a target or a decision.

A target is what we would like to achieve, even though we may think it’s unlikely. Companies often set sales targets for their staff to motivate them, not because they think the chosen level of sales is the most probable level that will ensue.

We make a decision when we choose a particular outcome from all those that might occur in the future because we think this choice will bring us the most benefit -not because we think it’s the most probable or the expected outcome. If I’m in a marketing department I might think that sales of 500 units are most likely next month, but I choose to present a ‘forecast’ of 400 units. By keeping the forecasts low it’s likely that I’ll be able to boast to senior managers that our brilliant marketing efforts have enabled us to exceed the forecast. If I do this, I’m not forecasting. I’m decision making.

If I’m an economic forecaster and circumstances are changing I might prefer to stick to my original ‘forecast’ of 2% growth, even though I think that 1.6% is now most likely. Changing my forecast too often might be seen as a sign of incompetence. Alternatively, I might play safe and stick to what others are forecasting –even though I think they are likely to be wrong. That way I won’t be exposed if I’m wrong. This is known as herding.

In other circumstances, it might pay me to deliberately make my ‘forecast’ different to others. If I’m the one person who says there is going to be a recession, when everyone else is forecasting growth, I’ll be seen as a brilliant prophet if the economy goes into a slump. I reason that my ‘forecast’ will soon be forgotten if I’m wrong –and anyway I have a catalogue of excuses ready to explain away the blunder.

Decisions masquerading as forecasts are particularly prevalent when forecasting gets mixed up with politics. In organisations, people often have a temptation to exaggerate their forecasts to obtain more funding for their departments. Even the International Monetary Fund (IMF) is not immune from political influence. There is evidence that governments of countries that are politically aligned with the US tend to receive favourable ‘forecasts’ of growth and inflation when they are coming up for re-election. The US is the major funder of the IMF.

Then there are those regular scary weather ‘forecasts’ in tabloid newspapers. ‘Forecasts’ of snowmageddons lasting for three months or summers that will be chillier than winter are outcomes chosen by editors to sell their papers. They know that readers will have long forgotten these headlines by the time the paper hits the recycling box.

The difference between forecasts, targets and decisions is more than a semantic quibble. It can cause confusion and inefficiency in organisations and mislead people in their decisions. Two eminent forecasters, Michael Clements and Sir David Hendry, define a forecast as simply “any statement about the future”. A more specific definition would be helpful. How about: “an honest expectation of what will occur at a specified time in the future based on information that is currently available”. I am sure this can be improved upon, but at least it’s a start.

Paul Goodwin