1.8 million American truck drivers could lose their jobs to robots. What then?

So far, discussion of self-driving cars has mostly confined itself to tech geeks and urbanists. But if they live up to their promise, autonomous vehicles could have seismic effects on America’s economy and culture. It’s probably time for a wider circle of participants, including economists, politicians, and social scientists, to start grappling seriously with what’s coming.

Let’s take just one example: long-haul trucking.

Freight trucks (semis, 18-wheelers, tractor trailers, what have you) are so ubiquitous on US highways that we scarcely give them any thought. But they are a big piece of the US economy. According to the American Trucking Association, these vehicles carry 67 percent of the freight that moves within the US — some 9.2 billion tons a year.

There were, as of 2009, some 2.4 million class 8 trucks (semis) and 5.7 million commercial trailers registered in the US. Together they traveled 99.2 billion road miles in 2010.

All that driving employs lots of people. In 2014, according to the Bureau of Labor Statistics (BLS), there were about 1.8 million people driving heavy trucks. It is one of the last jobs available in the US that pays something close to middle-class wages — median annual pay of about $40,000 in 2015 — without requiring a college education.

And those jobs are spread out, too. NPR’s Planet Money made a splash last year with an interactive map of the most common profession in each state over time. It’s based on data from the Census Bureau (which also counts delivery people as truck drivers, somewhat skewing the numbers.) This is for 2014:

This is to say nothing of the other jobs directly related to trucking — building and maintaining the trucks, management and dispatch, sales and clerical, etc. — and the jobs they support indirectly (gas stations, motels, diners). All those jobs are widely distributed as well, and help support lots of small communities.

When it comes to self-driving cars, the hardest problems are in urban environments, where there are complex, rapidly evolving situations involving other cars, bikes, pedestrians, traffic signs, and, well, you never know what might stumble into city streets.

Freeways and interstates are much easier. They feature fewer distractions and decision points than city streets. Many cars already have the capacity to automatically stay within a lane, sustain cruising speed, and maintain a safe distance from other cars — they are becoming self-driving on highways first. For trucks, that’s most driving.

Also, trucks are big, so they can carry more sensors and cameras, enabling them to achieve better situational awareness. So self-driving trucks are likely to arrive sooner than self-driving urban vehicles — at least that’s what the folks working on them say.

How soon? Well, it’s already underway. Trucks have reached level 3 on NHTSA’s scale of autonomy:

Daimler has a level 3 big rig licensed for road tests in Nevada:

The self-driving mode works like this: While traveling along a clearly marked road, the truck’s main display will light up an indicator telling you Highway Pilot is available. You can activate it by pushing a button on the steering wheel. The system is similar to cruise control, except that it also steers the truck. You have to stay behind the wheel, though, in case the software determines that it can’t handle upcoming twists and turns. In that case, the dash starts a 20-second countdown back to human driving.

A convoy of level 3 trucks completed a journey across Europe earlier this year, demonstrating the value of “platooning” to save fuel.

A company called Otto, founded by 15 ex-Google types, is working to build a kit that it can attach to ordinary semis, bringing them up to level 3. They expect to commercialize “soon,“ likely in the next year or two.

Volvo is also heavily into autonomous trucks — they’re working on semis and already have a couple of fully autonomous trucks operating in mines now.

Now, semis are still a ways away from being fully autonomous. That will require that they enter and exit freeways and navigate urban streets, which will in turn require years of software development, testing, and shifts in law and regulation. Human drivers will be needed for the time being.

But for how long? Maybe it’s two years, maybe five, maybe 10, but either way, the trajectory is toward drivers being put out of business, and 1.8 million truck driving jobs (not to mention all the other jobs they support) is a lot to lose in that short a period of time.

Consider: US coal mining employment hit a high of 449 thousand in 1920 and has declined ever since (with some ups and downs — the last peak, in 1980, was 229 thousand). It is now down to 80 thousand. Coal mining has lost around 150 thousand jobs over 30 years or so, around 50,000 in the last five years. And that is considered a social and political crisis worthy of presidential attention.

Truck drivers are a lot more spread out, and 1.8 million, or even half that, is a lot more than 50,000. These will be people (mostly working-class men) losing well-paying jobs, with few alternatives, and they probably won’t be happy about it.

In a great post about autonomous trucks, blogger and independent researcher Scott Santens advocates for a universal basic income to protect truck drivers and the many others who will lose jobs to automation and robotics in coming decades.

That’s an interesting idea — lots of good Vox articles on it — but it seems unlikely to manifest in the US in the next decade.

Until then, what’s the solution to hundreds of thousands of unemployed truck drivers?