On the left, there are two prominent worries about American jobs. One is that new jobs are terrible: Low-level service and retail positions, or “McJobs,” are replacing the middle-class occupations of the past. The second is that there soon won’t be any jobs at all: Robots, artificial intelligence, and similar technologies will produce so much labor displacement that even crummy jobs won’t be available.
That’s pretty depressing if these two (somewhat contradictory) views are accurate. Fortunately, neither one is. Start with the idea that jobs will disappear. This is, to say the least, highly unlikely. Concerns about the disappearance of jobs have recurred again and again as technology has advanced, and they’ve always been wrong. The general pattern is that technological transformations put workers out of jobs in one sector only to have more jobs created in others as demand for new products and services grow. Thus, technology advances but the availability of jobs does not decline.
Take the 1950s, when postwar concerns about automation and technological control (the so-called cybernetic revolution) were peaking. The Nation magazine remarked in 1958: “The problem we shall have to face some time is that the working force is expansive, while latter-day industrial technology is contractive of man-hours.” Sound familiar? Somehow we’ve soldiered on without mass unemployment even as the labor force has expanded from 68 million then to 160 million now.
But perhaps today is different? This technology — robots, artificial intelligence, the computerization of everything — is so powerful that it will hoover up most of the jobs and leave very few for actual humans to do. The futurist Martin Ford, in his best-selling 2015 book, Rise of the Robots: Technology and the Threat of a Jobless Future, predicted half of US workers would be replaced by robots in the next 20 years. In an influential article for Mother Jones magazine, with the striking title “Welcome Robot Overlords: Please Don’t Fire Us?”, the blogger Kevin Drum argued that robots’ capabilities are growing so fast that there will not be much need for human workers by, say, 2040, and the inevitable result will be mass unemployment and social dysfunction.
If the robots were coming, we’d see accelerating productivity. We don’t.
But what is the evidence that technologically driven labor displacement is taking place at a particularly fast — much less accelerating — rate? To put it bluntly: There isn’t any. If Ford’s “rise of the robots” were taking place, we would be seeing very rapid productivity increases today (fewer workers, larger output). We’re not. Instead, productivity increases have been abominably slow in recent years — a mere 1.3 percent per year, just over a third of the rate at the end of the last century.
Another indicator that the robots are gaining on us would be an exceptionally high rate of “occupational churn,” the rate at which the job structure is changing as some occupations decline and others grow. In a study of Census data going back to 1850, economists Robert Atkinson and John Wu found instead that the rate of churn in recent decades has been exceptionally slow — slower, in fact, than at any other period in their study. Indeed, the rate of occupational churn in the new century has been less than a third of that in the dynamic 1950 to 1980 period.
These results are completely inconsistent with the “jobless future” hypothesis. We are in no danger of having humans crowded out by incredibly productive robots. If anything, given the anemic productivity growth, we should worry that robots and related technologies are being underused, thereby holding back the advance of new economic sectors.
OK, the jobs pessimist might reply, maybe there will be jobs in the future but they’ll be jobs barely worth having. After all, isn’t that how American jobs have been trending for 50 years?
The pessimist has it right that past trends may continue, but the pessimist is wrong about what those past trends have been. Since “blue collar” and “white collar” have become hopelessly muddled terms, the best way to look at changes in the job structure over time is to sort workers by skill levels and work performed.
In The Economy Goes to College, by economists Anthony Carnevale and Stephen Rose, the authors sort workers into three occupation tiers: managerial-professional, middle-skill, and low-skill. In this categorization, mid-skill jobs include not only mid-level supervisors, skilled craftspeople and clericals but also service occupations such as medical assistants and police; the low-skill category includes not just factory workers but also retail sales clerks and service workers including security guards, janitors, and waitresses.
As a proportion of total employment, low-skill jobs have dropped by 10 percentage points since 1967
Back in 1967, just 21 percent of jobs were managerial-professional, the authors found, while 39 percent were low-skill and about the same proportion were mid skill. Today, 35 percent of jobs are managerial-professional, 36 percent are middle-skill and only 29 percent are low-skill. Thus, managerial-professional positions are up 14 percentage points as a share of jobs since 1967; low-skill jobs have actually dropped by 10 percentage points
In a related analysis, Carnevale and Rose classify all workers by the content of the work they perform. This breakdown produces a division of the workforce into five basic categories: office work (across all industries); high-skill services (non-office work in health care, education, and communications); low-skill services (retail, personal, and food services); manual labor in industry (manufacturing, construction, etc.); and primary production (mining, farming, fishing). They found that 44 percent of US jobs today are in offices, with another 20 percent in high-skill services. Just 15 percent of jobs are manual labor in industry, another 19 percent in low-skill services.
Since 1967, the big change has been the rise in office work and high-skill services (up 14 points), while the big decline has been in industrial manual labor, down 13 points. Interestingly, and very importantly, the share of low–skill service jobs is just about the same today as it was back in 1967. (Fast food jobs — the quintessential low-skill gig — are stuck at around 2.3 percent of US jobs, with no sign that that share is going up since the turn of the century.)
Statistics like those put paid to the notion that middle-class jobs are disappearing and being replaced by “McJobs.” This view equates the decline of low-skill, relatively well-paid jobs like those in manufacturing — which, indeed, has been going on since 1948 — to an overall decline in middle-class jobs, which is not merited.
The middle-class jobs of today are in the growth areas of offices and high skill services. These two areas of the economy now provide 64 percent of all jobs. And they’ve expanded more as a share of jobs since 1967 than manufacturing and related jobs have declined. Middle-class jobs are not disappearing; rather, they have moved to different sectors that require higher levels of education and cognitive training.
Of course, that middle-class jobs are not disappearing does not mean inequality is not rising. It is. That trend is a product of many forces, including a lack of support for low-income workers, stagnating wages, and the skew of economic rewards toward the rich. The availability of new middle-class jobs in offices and high-skill services is not sufficient, by itself, to counteract these powerful forces. But that is not fault of these jobs. In part, it has to do with unwise policy choices.
We should expect current trends in the job structure to continue. The jobs of the future will likely require more skill, not less skill, than those of today. Take one of the chief obsessions of today: the projected rise of self-driving cars and the presumed decline of driver jobs. As experts convened by the National Academy of Sciences note in a recent report, Information Technology and the US Workforce: Where Are We and Where Do We Go from Here, the story is a lot more complicated than drivers losing their jobs to autonomous Ubers. The report notes that “there will be a long period of mixed-use road use, with both autonomous driving and manual driving cars sharing the roads.” It continues:
It is possible that ongoing development of these technologies, including infrastructure, will create more jobs than are lost in the wake of self-driving vehicles, but it is likely that the skills required for such jobs will be quite different from those currently possessed by today’s truckers and taxi drivers. The new jobs are likely to rely more heavily on analytic, cognitive, and technical skills. Indeed, even in the near term, as self-driving technologies are being developed, the occupation of trucking is likely to be transformed.
So it was ever thus. Continuing technological advance, as in the past, is unlikely to produce a future of no jobs. It will lead instead to a future of different and more highly skilled jobs. Rather than bemoaning a chimerical disappearance of work, or seeking to somehow reinvent the manufacturing economy of the past, the left should seek to increase access to high-skill and growing sectors of the economy, particularly in regions where the decline in low-skill industrial labor has significantly eroded the job base.
Analysis of Census data indicates that 84 percent of noncollege workers who make $50,000 a year or more work outside of manufacturing. Count on that trend to continue.
Above all, we must keep the focus on running a high-pressure, full employment, rising-wages economy — and expanding the benefits available to all workers. American jobs will continue to evolve toward higher skills over time; let’s make sure those jobs pay as well as possible and that they provide solid economic security.
Ruy Teixeira’s new book is The Optimistic Leftist: Why the 21st Century Will Be Better Than You Think. He is a senior fellow at the Center for American Progress.
The Big Idea is Vox’s home for smart discussion of the most important issues and ideas in politics, science, and culture — typically by outside contributors. If you have an idea for a piece, pitch us at email@example.com.