If there is one thing many politicians in both parties can agree on, it’s that the middle class is dying. For instance, when she announced her campaign for president in February 2019, Senator Elizabeth Warren (D-Mass.) declared that “over the years, America’s middle class has been deliberately hollowed out.” A few months later, in the summer of 2019, Senator Josh Hawley (R-Mo.) echoed this sentiment, asserting that the “elites” had entered into an “alliance” with multinational corporations that resulted in, among other things, a “hollowed-out middle class.”
The belief that there has been a “hollowing out” of America’s workforce is well founded. Indeed, the decline is clear in the data, and it has caused significant disruption. Populists on the left and right are correct that traditional middle-class jobs are shrinking as a share of total employment. Yet is wrong to view this as the result of deliberate decisions by “the elites” in the sense that the senators mean. Moreover, hollowing out is only half of the story. In recent decades, a new middle is emerging as a result of economic dynamism.
The Decline of the Traditional Middle Class
The hollowing out of the US labor market is clearly shown in the Chart 1 below, using data from Massachusetts Institute of Technology economist, David Autor. The chart captures the change in the share of employment in nine broad occupation categories. The categories are ranked by the average wage of workers in that occupation. Workers in the “managers” occupation are the highest paid, and those in “health and personal services” occupations are the lowest paid.
Autor finds that in 1970, total employment was almost evenly split across low-, middle-, and high-wage occupations, with 31 percent, 38 percent, and 30 percent of employment in each category, respectively. Employment in middle-wage, middle-skill occupations has since fallen significantly, to 23 percent of total employment. In other words, the middle of the labor market has indeed hollowed out.
Think about what workers in those middle-skill, middle-wage occupations do. Many of their job tasks involve repeatedly executing a series of steps with precision and accuracy. For example, a manufacturing worker on an assembly line has to repeat the same steps to put hubcaps on cars. A bookkeeper performs a similar function in carrying out basic accounting procedures. A bank teller must execute the same lists of steps in the same order for every cash withdrawal and deposit. In each case, precision and accuracy are necessary, and rewarded.
It happens that computers and robots are designed to perform exactly this type of function, to execute lists of steps repeatedly and accurately. As the cost of computing has dropped, businesses have increasingly turned to technology to perform the types of rules-based procedures at which computers excel. So robots replaced manufacturing workers on assembly lines. Software replaced bookkeepers. ATMs replaced bank tellers.
Notice that the occupations at risk didn’t include many of the lowest-skilled, least-paying jobs. The tasks in those jobs involve less step-by-step, precise execution than for those in the middle. For example, it turns out that it’s quite difficult for a robot to tell the difference between a pile of papers in a messy office that should not be thrown away at the end of the workday and genuine trash. A human custodian can readily tell the difference. Similarly, low-skill, low-wage workers in food preparation and other service jobs need the ability to adapt to situations and interact with people. These tasks are also not amenable to being codified in step-by-step instructions.
Workers in high-skill, high-wage occupations often have to exercise sound managerial judgment, argue persuasively, be creative, and adapt to situations. As with the tasks required in certain low-wage occupations requiring dynamic decision-making, it’s hard to program computers and robots to do these well.
It is occupations in the middle—which paid better than those at the bottom because their tasks required precision and accuracy, but paid less than those at the top because workers in those occupations are relatively less skilled—that were hit hardest by automation. Those jobs included production and craft workers, machine operators and assemblers—exactly the types of jobs that have political salience today. These are the jobs that President Trump mistakenly argues were primarily affected by globalization (which was a factor, but not nearly as large a factor as automation), the jobs that didn’t require a college degree but did offer a middle-class life.
These changes have had a profound effect on society, culture, and politics. This longer-term, slower-burning, decades-long trend collided with the gut punch of the Great Recession, which contributed to the rise of populism.
Fortunately, we have reason to believe that the middle might make a comeback in the form of a “new middle class.”
The Rise of a New Middle Class
Split up middle-wage jobs into two groups: an “old middle” and a “new middle.” The old middle is what we think of when we talk about traditional middle-class jobs. These are the types of jobs—certain production, construction, and clerical occupations—that have been disappearing largely owing to automation and that have played a major role in the public debate these past few years, fueling the rise of populism. But over the past two decades, as jobs in the old middle have declined, jobs in the new middle have increased.
Using data from the Bureau of Labor Statistics, I sort occupations based on the median wage earned by workers in that occupation. I then look at the middle-wage occupations. Specifically, I examine the middle third of occupations by pay—the occupations for which one-third of all jobs pay more and one-third pay less. I do this for every year from 2000 to 2018, and I look at how the jobs that are included in the middle third change over time. I divide jobs into two groups: the old middle and the new middle.
I assign three broad occupational categories to the old middle: production occupations, construction and extraction occupations, and clerical occupations. The specific jobs in those broad occupation categories include assemblers of machines and equipment, metal fabricators and fitters, meat packers, poultry and fish cutters and trimmers, machine setters, furnace operators, metal pourers and casters, tool and die makers, brick masons, welders, carpenters, construction laborers, telephone operators, payroll and timekeeping clerks, file clerks, typists, and secretaries and administrative assistants.
Jobs that are not in the old middle but that are still in the middle one-third of the occupation-wage distribution I classify as belonging to a new middle. These jobs are found in several broad occupational categories: healthcare support occupations, transportation, education and training, and personal care and service, to name a few.
As Chart 2 above shows, jobs in the old middle are indeed declining as a share of overall employment. Overall, middle-wage production, construction, and clerical jobs—the old middle, represented by the yellow line—have fallen from around ten percent of total employment to eight percent of total employment over the past two decades.
Production jobs that fell in the middle third of the occupation-wage distribution constituted about three percent of total employment in the year 2000. By 2018, that had fallen by one-third, to two percent. Middle-wage clerical jobs made up six percent of total employment in 2000 but now only make up four percent.
But as that has happened, middle-wage jobs that are not in what we think of as traditional middle-wage occupational categories have seen considerable growth. Over the past two decades, the share of total employment in new middle jobs—represented by the teal line—increased by 25 percent, from around 10 percent to 12.5 percent.
Over the past half-decade, the fastest-growing jobs in the new middle include sales representatives, truck drivers, managers of personal service workers, heating and air conditioning mechanics and installers, computer support specialists, self-enrichment education teachers, event planners, health technologists and technicians, massage therapists, social workers, marriage and family counselors, audiovisual technicians, paralegals, healthcare social workers, chefs and head cooks, and food service managers.
These jobs probably require a little more education, skill, and experience than jobs that made up the old middle class. They require more situational adaptability, social intelligence, customer service and interpersonal interaction, low-end managerial skills, and administrative, technical, and communication skills that are resistant to automation.
Mostly a Story of Upward Mobility
Recall that MIT economist Autor calculated that over the four-and-a-half decades since 1970, middle-skill employment fell from about 38 percent to 23 percent of total employment. But this fall in the middle was offset by a rise in employment at the top. Employment in high-skill occupations grew from about 30 percent to 46 percent. At the same time, low-skill employment did not increase. On the whole, the hollowing out of the middle is a story of middle-skill employment being replaced by high-skill employment. It is a story of upward mobility and an affirmation that the American Dream is very much alive.
Political and opinion leaders have done an excellent job in recent years of focusing on the negative effects of the dynamism wrought by factors like automation and globalization, with scant mention of its benefits. The populist response to dynamism is wrong. We shouldn’t attempt to turn back the clock, seeking to revive shrinking occupations and to reverse the productivity gains made possible by technological change. Instead, we should be working to ensure that the institutions that impart worker skills—schools, public programs, private businesses—are up to the challenge of equipping workers with what they need to flourish in the 21st century economy. Workers themselves need to be willing to adapt to change and to do what is necessary, though often very difficult, to flourish in a dynamic economy.
Businesses have an incentive to find ways to use those workers who are adjusting to the new economy. As traditional middle-class jobs—the old middle—become a smaller share of the economy, a new middle is rising. Economic dynamism is much more a blessing than it is a curse.