Blog

Forecasts, Schmorecasts. People Shape the Future.

Posted by: Tony Dutzik on

Ah, such fond memories of last weekend, when election forecasting wonks were locked in heated debate about whether Hillary Clinton was a complete cinch to win yesterday’s election or merely very likely to do so.

Good times.

Snark aside, the failure of polling and models to anticipate Trump’s victory is worth examining for what it tells us about how we think about – and might choose to shape – the future.

At Frontier Group, we’ve written about the failure of transportation and energy models to accurately anticipate recent events such as slowing growth in driving (until the gas price collapse of the last couple of years at least, which models also missed) and growth in renewable energy. The common theme was that failed forecasting models relied, at their core, on the assumption that the future would be like the past, only more so.

But, of course, that’s not the way the world works. Technological breakthroughs, shifts in the allocation of capital, and the decisions made by real live humans shape the future in ways that no model can anticipate. And those changes are not all linear – history is full of tipping points and critical moments when bells are rung that cannot be unrung and the fate of people and civilizations changes in a heartbeat. Perhaps this is one of them.

The value of models and forecasting is purported to be to help us make intelligent choices. And, in many situations they can be helpful. But when those tools fail to capture the disruptive events that have the greatest long-term consequences for our world, we need to be very, very careful how much we rely on them. And that is especially true in a 21st century America (and world) where disruptive events seem to be the order of the day.

I think there are two responses to this. First, analytically, we need to develop and use tools that can bring disruptive events within our field of vision and play through the multi-step chains that can lead from a seemingly small decision to a major societal change. Scenario analysis and forward reasoning have promise. And we need to try to help fellow citizens understand how to do this kind of thinking for themselves.

And secondly, and more importantly, we need to stop treating the future as if it is predetermined. No model could tell us what would happen in yesterday’s election because the results were the product of collective decisions made by tens of millions of people – people with hearts, minds and God-given free will.

If you care about making social change of any kind, the inclination to worship data and analytics and to downplay vision and values is one you need to fight hard against (and I say this as something of a data guy). For better or for worse, Trump got his supporters to imagine and invest themselves in a vision of American greatness and believe that that future could be created through human will.

Clinton never did that. And so, we are where we are.