As I’ve written before, it’s important for our policymakers and elected leaders not to lose their heads over technological advances in the economy. This recent headline—“The driverless truck is coming, and it is going to automate millions of jobs”—does not help.
While it is true that new trucks are coming, the author exaggerates when he predicts a “400 percent price-performance improvement in ground transportation networks.” Without citing his source, he asserts that 75 percent of long-haul trucking costs are labor costs. That figure is nonsensical: depreciation, fuel, insurance, profits, interest costs, and maintenance consume a much larger share of these shipping costs. The Bureau of Economic Analysis in its industry studies shows that just 26 percent of transportation industry costs in 2014 were due to labor compensation.
Furthermore, not all of the 1.6 million long-haul drivers will lose their jobs. First, many drivers are part of the unloading process. Second, automated driving technology is not very good in unusual circumstances (e.g., changing lanes in traffic jams, taking directions for construction and other detours, fixing flat tires, etc.). Third, the notion of convoys of trucks driving in close proximity at a fixed speed of 45 miles an hour (to save gas) creates risks for cars trying to exit the highways.
For these reasons, just as there are extra workers on trains, humans will have to be present to respond to these out-of-the-ordinary events. Total job losses will therefore be considerably less than the total number of driver jobs.
But losing jobs in one area does not have to lead to overall job losses. Ever since the Luddites tried to smash the machines that reduced their numbers, total employment hasn’t gone down—it has only been reorganized into different areas. In the hundred years from 1860 to 1960, agricultural employment went from 60 percent of the work force to five percent without leading to massive unemployment, which tended to move up and down in response to the business cycle.
In the 1950s and 1960s, many feared that the introduction of new manufacturing machines would drive up unemployment; in November 1961, Time ran a major story on the “automation jobless.”
Indeed, the share of manufacturing employment declined from 35 percent in 1950 to 15 percent in 1990 (before the latest rounds of trade agreements). And, over the last several decades in the United States, containerization reduced dock side employment and did away with many clerical jobs. Yet, before the 2008 financial crisis, we had incorporated huge numbers of women and immigrants into the labor force while maintaining high employment among non-elderly adults.
In this presidential electoral cycle, there have been many proclamations about how badly the economy is performing. While few candidates have addressed technological unemployment itself, claims that the economy will no longer provide enough jobs in the future are part of a broader narrative that instills worry about the present and the future.
Finally, while many people profess that the economy as a whole isn’t doing well, a recent survey from the Federal Reserve Board found 69 percent of Americans said that they were “living comfortably” or “doing okay” and only just over 9 percent were “finding it difficult to get by.” To me, none of this suggests that we should worry about driverless 18-wheelers any time soon.