The Foresight Paradox
In every foresight or forecasting exercise, there are two overarching tensions:
- The more certain and detailed the forecast, the more people will accept it and believe it to be useful.
- The more certain and detailed the forecast, the less likely it is to happen.
This is the foresight paradox: you can be completely accurate, or you can be completely engaging, but you can't be both. As a result, every forecast (or scenario, or prediction) has to find the right balance between the two, trading off likelihood for believability.
As a simple example: a forecast that says "the next decade will see continued economic disruption" is very likely to be true, but of limited utility and almost no capacity to inspire innovative thought; conversely, a forecast that says "the Eurozone will collapse in the Summer of 2013, leaving EU countries scrambling to find usable currencies, with many temporarily adopting the dollar" is almost certainly not going to happen as described, but offers clear guidance for action, and can inspire novel business and political strategies. If the latter forecast is given by someone in a suit and tie, with a very serious sounding title from a very serious sounding institution, many people will accept it as being much more than informed conjecture -- and will reject the more general forecast as being useless.
This shouldn't come as a surprise. Precise and detailed forecasts offer structure for thinking, giving the listener a framework upon which to build strategies or make concrete rebuttals. Moreover, there appears to be a psychology of belief that makes people more likely to listen to detailed predictions, offered with certainty and clarity, than to listen to general forecasts, or those offered with plenty of hedges and caveats -- even though the detailed predictions are almost always wrong. (This isn't helped by a media culture that favors the spectacular over the thoughtful, and the adamant over the hesitant.)
It's not hard to find pundits and self-described futurists who will gladly accept the visibility and attention that comes from making detailed, spectacular predictions, no matter the eventual accuracy. If confronted, they'll mumble something about timing or unpredictable events; such confrontations are vanishingly rare, however, especially for high-profile pundits. It doesn't matter how wrong you are if you get good ratings.
Ethical futurists have a bit more of a dilemma here, however. A forecast needs to be vivid and engaging enough to trigger action, yet general and cautious enough to engender restraint. Or, as I put it in one interview, should be wrong in useful ways.
The simplest approach is to keep forecasts as general as possible, using detail only when well-supported by evidence. With this method, the emphasis is on the present-day and near-term drivers that lead towards the (more general) future. There is a temptation to over-emphasize the visible, and not leave enough space for wild cards and "black swans," however. The core quandary remains knowing how general and cautious one can be while still offering useful insights, and how specific and detailed one can be while still not leading the audience astray.
Another fairly straightforward method is to use a more detailed forecast, but emphasize the uncertainty from the outset, being clear to the audience that the real outcomes will vary. The given forecast should be considered an example, not a certainty, a possible future that fits within a broader framework. Audiences don't always respond well to that approach, however; in some cases, they'll still take the example future to be the "real prediction," and in others can interpret an emphasis on caution to mean that the futurist really doesn't know what she or he is talking about.
My preferred approach is to use scenarios, essentially giving multiple examples within the general framework. This illustrates the shape of the broader framework better, and makes clear that no one specific forecast is the "real prediction." Yet the problems with this approach are manifold: coming up with three to five internally consistent forecasts is significantly harder than just coming up with one; audiences will gravitate towards preferred scenarios, sometimes ignoring those that don't turn out in ways they like; and it's difficult to encapsulate multiple scenarios into a short presentation or statement without rendering them meaningless.
This last problem is one that I've encountered quite a bit recently. There seems to be a trend in conferences right now (especially in Europe) to limit presentations to 15 minutes. Although there are definite benefits to this approach (most notably in maintaining audience interest), it means that any foresight-based presentation is crippled. A speaker simply doesn't have the time to offer multiple scenarios in anything other than a bullet point/headline format, surrounded by lots of big idea framing to give the scenario headlines some context (the talk I gave at the Guardian Activate Summit in London last year is probably my best effort at doing this).
Unfortunately, audiences don't respond as well to multiple scenarios as they do to single, detailed forecasts, even when they know the detailed forecasts will inevitably be wrong. Moreover, appearances limited by time (such as, in particular, television) make even the headline scenario approach difficult. The best one can do -- in my experience, at least, and I'd love to hear better suggestions -- is to be sure to offer caveats and use cautious language such as "appears to," "likely," and especially "one possibility" (or similar statements underlining that different outcomes are possible).
The modern spectacle-driven media loathes uncertainty, and will almost always give more attention to aggressive certitude (no matter the accuracy) than caution. Many business audiences feel the same way. Sadly, the foresight paradox boils down to this:
The futurists who get the most attention are usually the least accurate.