Will the last human worker please turn out the lights?

Capitol Hill tends to agree on the importance of a vibrant American middle class, with people able to support their families and maintain enough economic and political power to keep our democracy honest and responsive to its citizens. Yet the U.S. now faces economic shifts at a faster tempo and of higher scale than before, and if Congress doesn’t prepare accommodations, then entire sectors of middle-class jobs may convert into unemployment.

{mosads}The evolving economy — called “new,” “digital” or otherwise — may not be fundamentally different from what we’ve seen before. It is natural for industries to crumble over time and be replaced by others, a normal process long called creative destruction or “Schumpeter’s gale” by economists. But today, this gale is blowing more quickly and intensely than before, with more dramatic consequences for American jobs at the middle and lower ends.

Take Uber. By connecting riders and drivers, the company provides an innovative and welcome service (especially here in Washington, where Uber’s presence quickly jolted taxis to begin taking credit cards and providing better service). But by all accounts, serving as an Uber “driver-partner,” or contractor, is not the same as holding a traditional job with a taxi company. Uber’s drivers can flexibly choose their own work hours, but there is no prospect for wages or insurance benefits of the type associated with traditional jobs. The net result for the economy is a substantial decline in traditional jobs available in the driver sector — taxi drivers and chauffeurs put out of business by Uber — and a much smaller growth of traditional jobs in another sector — the computer programmers, administrators, managers and other higher-skilled labor in Uber’s corporate offices.

This job shift is no reflection on Uber, which is simply a company that identified a consumer demand and satisfied it quite well. Over time, similar shifts will be seen through the lens of other sectors both low and higher on the value chain. For instance, massive open online courses, or MOOCs, seem likely to one day obviate the need for today’s number of high-level teaching jobs. Why do we need thousands of expensively trained economics professors across the country if everyone can stream video of the best economics professor at the same or less cost?

It is the defining trajectory of an evolving economy that the lower ends of an industrial value chain are increasingly automated and mechanized. But it may in fact be different this time, because so few sectors seem able to absorb displaced workers with their current skill sets. And the skills fueling the growth industries and superstar companies of this decade — programming, data mining, project management — are far from those required to be a driver, and are not learned quickly or easily.

This rapid economic change will only continue. In March, Uber hired 40 of Carnegie Mellon University’s best robotics experts in order to develop driverless-car technology, and has doubled down with interest in purchasing Tesla’s driverless cars. This implies a future where nearly 180,000 taxi drivers and chauffeurs (May 2014 numbers), plus Uber’s 160,000 “active” drivers (January 2015), are looking to apply their skills elsewhere. If you add self-driving trucks into this vision, there are another 3.4 million Americans who will be looking for work. While a 20-something who finds that his or her industrial sector has disappeared can likely learn new skills and recover, it is hard to imagine what a 52-year-old whose industry disappears will do for the final decade before his or her Social Security benefits kick in.

These displaced workers will not quickly obtain the high-level skills needed in growth industries (say, computer programmers at our new driverless-automobile companies); even if they could, the number of jobs available there would be a fraction of what we’ve lost.

There’s no question the growth of customer services, MOOCs and driverless technology push the world forward and provide once-unimaginable benefits, lowering transport and education costs and allowing users like myself to save money and spend it elsewhere in our economy. But where will I spend that money I’ve saved? If it does not go toward sectors in which those displaced workers can find jobs, we will find ourselves with a problem. It is easy to imagine a future improved by robotic automobiles and other tech wizardry; it’s more difficult to imagine what U.S. policies will successfully shepherd citizens through jarring industrial shifts.

A range of active policy options are available to Congress, though none are in vogue at the moment. Data on jobs training programs do not inspire confidence, and redistribution remains a dirty word in U.S. politics, evoking visions for some of rightfully earned wealth being stripped away, or more dryly, subsidizing less productive areas of the economy at the expense of more productive ones.

Earlier this summer, a friend of mine who used to courier urgent, signed financial statements on a daily basis not only lost his job — his entire sector ended overnight. All banks now have the requisite encryption to make such couriers unnecessary. When this happens to larger sectors of employment — first the taxi drivers, eventually the economics professors — it will be clear that the U.S. has an interest in smoothing these economic transitions in a way that keeps the middle class strong and prevents a generation of lost jobs. Whatever solution at which Congress arrives, it is important it begins the debate now and has a plan in place before the last driver is replaced by technology.

Rettig works for the Global Economy and Development program at the Brookings Institution. Read more of his work at www.mikerettig.com. The views presented here are entirely his own.