Site icon populyst

The Futility of Annual Top 10 Predictions

In every recent year, a black swan event has made top 10 lists appear quaintly naive and unimaginative. Our list is probably no better.

This time of year, top 10 predictions are all the rage. These lists can be interesting and entertaining but how useful are they really?

This question goes to the heart of forecasting. How futile or how useful is an attempt to forecast the economy, or technology, or world events for the next twelve months? There are three answers.

First, not futile and somewhat useful. Projecting the trends of 2016 into 2017 is a useful exercise to identify their linear logical trajectories and end points. For example, the automation of many job functions will continue as long as robotics and artificial intelligence make progress. Or, North Korea’s ability to deliver a nuclear warhead on a long-range missile will continue to improve if unchecked.

Second, futile and not that useful. When a desirable trend, say a decline in unemployment, is identified, policy makers will attempt to reinforce it. When an undesirable trend becomes obvious, they will work to counter it. However in both cases, the intervention can be either effective or counterproductive. It can either reinforce or roll back the trend. Human tinkering means that few trends are truly linear or logical beyond the near-term. There may be a slowdown in the spread of automation. There may be an agreement to stop North Korea’s nuclear ambitions.

Third, neither futile nor useful but somewhat irrelevant. While forecasters are focusing their sights on the high probability of a, b and c, there are always bigger low-probability events brewing under the surface. In fact, the most important event in any given year, the one event that shakes things up and that has wide long-lasting ramifications, is usually one that few people foresaw at the beginning of that year.

And so on. If you look at it by decade, the most important events of the 1990s and 2000s were the collapse of the Soviet Union and the 9/11 terror attacks. Neither featured in top ten lists in any year but both had an enormous impact and repercussions that are still rippling around the world.

So instead of a list of top 10 higher probability predictions, we should consider a list of lower probability events each of which, were it to occur, would have a very large impact on the future of politics, economics, science etc. As extensively argued by Nassim Taleb, black swan events often have a much greater impact on the future.

Here is one attempt to compile such a list, with the caveat admission that it is only marginally better if at all than other lists and that the most important event of 2017 will likely be something else.

Low probability high impact events

In no particular order:

These are all low probability but not zero probability events. And the impact of each would be far greater than that of any higher probability event featuring in many top 10 predictions for 2017.

(Photo of The Scream by Edvard Munch, 1893)

Exit mobile version