BeyondDC has an absolutely fantastic post up about what he expects the D.C. area to look like in 2040. On any given point, I think it’s hard to disagree (except maybe that they’ll close National Airport—businesses outside the Northeast would flip out).

But I think that it highlights the most dangerous and difficult question of urban planning. The timeline for so much of this thinking requires predictions about what the region will be like in 2040. Take the Purple Line. By the time they finalize plans for the Purple Line, build it (with delays, of course,) and then build new developments around the stations, and then fill those developments, what year are we talking? And that’s a project that’s been in the works for a decade already.

But at the same time, history doesn’t just glide along incrementally and predictably; in so many ways it is punctuated by radical, unimaginable shifts. History moves catastrophically. It destroys long-term predictions.

In Doug Rae’s City: Urbanism and its End, Rae reproduces a graph that the best and brightest of New Haven made predicting population growth:

As you can see, they just expected the population to continue rising. It didn’t. The car came. And just like that, the city population started to decline as people, jobs and shopping flooded into the suburbs.

No one could have predicted it.

The growth of cities a century earlier was just as sudden. In 1860, Chicago had 109,000 residents. That grew to 503,000 by 1880 (despite the city burning down in 1871) and then to a million in 1890 and 2.2 million by 1910. The railroad came. No one could have predicted it. Between 1950 and 1970, Orange County grew from 216,224 residents to 1.4 million on the backs of the FHA and the defense industry. To go international, Shenzhen, China went from being a fishing village in 1979 to a city of 8.6 million people after Deng Xiaoping designated it a Special Economic Zone.

At a less dramatic level, would anyone have guessed in 1991 that there was about the be an urban Renaissance in this country? I don’t think so; if you reread Mike Davis’ City of Quartz, from 1990, it feels apocalyptic, not hopeful. Or today, Matt Yglesias argues that Las Vegas could be ruined by something as simple as California deciding that legalized gambling would help its budget crisis.

In other words, I’m very skeptical of any prediction that doesn’t account for some sort of disruptive technology or other exogenous shock. In D.C., maybe a conservative government will actually shrink government or a liberal one will actually cut out special interests, laying off thousands of federal employees or lobbyists and having a ripple effect through the economy. Maybe gas prices will triple, due to either market forces or taxes. Maybe, as has been predicted ad nauseum for 50 years, new telecommunications technology will eliminate the benefits of working near each other and drive a new round of sprawl. Maybe there will be a terrorist attack in the District.

The problem with this critique is that it’s impossible to know what the unexpected, transformative moment will be. It’s a known unknown. But you can’t build it into your predictions. So I’m not sure how I would change BeyondDC’s predictions beyond the caveat that no matter how smart a look at current trends, the future will look wildly, completely, unimaginably different from what we guess.