We all know that we don’t know what the future will hold but we all plan for it anyway. We plan our holidays, our retirement, the sales of a new product and the state of the economy. We pull wool over the eyes of our ignorance because someone, somewhere needs a view.

Just occasionally, there is the potential to have more than one view. Most of us find this quite hard. We are not programmed to believe several impossible things before breakfast. We are happy to talk about risks, but these are merely up and down (usually down) divergences from that ‘view’. It is often described as a central view but this is not often really the case; it is very unusual when a commentator has as many risks on the upside as on the down.

These views are stories. Stories have coherence and rationality. They tell why this will happen as a result of that. They give comfort of an evening that we can make sense of the world we live in. Scenarios both increase and decrease this comfort. Once, long ago, when I still engaged with large macroeconomic models we set up a scenarios club. With our clients, we proposed to develop scenarios of alternative futures of the world economy. (My macro model was so large that it covered the whole world!) The club sat down to envisage the scenario. We talked of oil shocks and market shocks – it was 1987 – and the future of international agreements. Every scenario that we imagined was run through the model and presented to the club. And then the reactions would set in.

Each participant imagined the policy reaction to the shocks. And when we ran these through the models, we got swiftly back to the business as usual which was our baseline forecast. A world in which things go on pretty much as they have done before. Our scenario club, perhaps not surprisingly, did not last long. I didn’t last long either. I quickly came to the conclusion that these models were a waste of time and money and told me nothing I could not write on the back of an envelope or fag packet (I was still smoking in those days).

Moreover of course we utterly missed the real scenario of the rise of new economies and the spread of globalised markets. Nor could our model have handled this, since it had no data on China and India and very little on Brazil or Russia.

This shows how hard it is to imagine alternative futures, in which the world is really different. And even if we can imagine the causes of these, the outcomes are utterly unpredictable. Not so much stories as fairy stories. So why bother?

One reason to bother is that it might help to be ready when things change. The story of Shell’s success in reacting to rising oil prices in 1973 is supposed to reflect how it had considered this scenario in advance. And this is certainly why the military play war games: when things move fast it helps to have figured out your reaction in advance rather than trying to do the analysis in real time.

On the other hand, playing war games convinced some highly intelligent analysts during the Cold War that pre-emptive strike was the best solution. Fortunately more cautious counsel prevailed. So scenarios are not an unmixed blessing. In this case, they were being used to counsel extreme reaction. A similar use, potentially similarly by vested interests, has been provided by some climate scientists. A scenario based on the risk of greenhouse gas increases, subsequent temperature rises, sea level increases and tipping points of various physical relationships has been used to create a scare which at one level induces the desire to play hard because it is all too late to save the planet.

At other levels, the scenario is simply a story. Facts are uncertain, their implications still more so. Our reactions can stabilise or destabilise and to be frank in the current state of knowledge either could be the case. One reaction to this is to look for probabilities. With enough history we can estimate the probability distribution of any variable in which we are interested. In this case, upside and downside risks are much more likely to be balanced and human tendencies to be either gloomy or optimistic might be minimised. But probabilities are not stories. They carry no freight of rationality or causation, nor of policy reaction implications. The Bank of England has carried probability ranges of inflation and output outcomes for some years now. They are not very wide (not nearly as wide as the data suggests) and in any case have not succeeded in getting much traction. People prefer stories.

An important reason for this is a sense that we face uncertainty. Uncertainty means we cannot attach probabilities to any given outcome. We simply do not know what is most likely. So it is not irrational to prefer the story. Choosing stories to tell and writing stories helps us learn about possibilities. It may help us react to events as they unfold. But it does not help us know what will actually happen. That is much more about defining a probability distribution and considering whether it is possible to do this.