MUNCIE – One feature of the past two or three decades is what economists call the polarization of labor markets. This is a fancy way of saying that we are seeing growth in high- and low-wage jobs, but a decline in middle-wage work. By any definition, the U.S. still has a large middle class, but three decades is a long time to be in decline, and there is no evidence that this trend is about to turn around.

There is growing concern about this phenomenon, and much of the COVID relief and infrastructure proposals hold policies designed to bolster middle-wage jobs. Unfortunately, there’s little reason to hope that short-term spending policies will have much long-term effect on this trend. The reason for this is that the root cause of these problems is not likely to respond to direct subsidies or stimulus.

Over the short run, say a few months or even a few years, an economy can experience too few jobs of one sort or another. For example, during COVID the declining demand for restaurants, hotels and recreation caused a big decline in low-wage work. During the financial crisis, there was a big decline in higher-wage financial services. Still, over the long term, the mix of jobs and salaries are almost wholly caused by the mix of workers and skills in a region. So, the types of jobs and pay in jobs in your town is mostly the result of who lives and works there. Nationally, the consumer demand for goods also plays a role, but it is secondary to the available workforce.

This means that the growth in high- and low-wage jobs is generally caused by an expansion of available workers in both categories. This can be attributed to a couple different factors — automation and changing consumer demand. These two factors are connected because automation reduces the price of some goods relative to others. This causes consumers to spend more in one area and less in another.

Automation is the most obvious of the two factors leading to job polarization. Most people develop useful skills over a work life. But, if those skills can be replicated more quickly by a machine, the demand for employees with those skills will disappear over time. Most of us experience change incessantly over our careers. But, over the past few decades, the jobs most susceptible to automation are those involving routine tasks, like assembly plants or warehousing.

The second cause is the significant shift in consumer spending from goods to services. Over the past half century, this shift has involved more than 30 percent of household spending. This is a massive adjustment that won’t reverse itself in the coming decades or even centuries.

Service sector jobs are not low-wage jobs. In fact, almost all the growth in high-wage jobs is in the service sector. But, unlike manufacturing of a half century ago, wages in the service sector are closely connected to education. In highly capital-intensive jobs, like those in a large auto factory, the variation in wages between workers were modest compared to their skill differences. The assembly line cancelled out any individual ability to do more or less on the job, thus erasing pay differences. It wasn’t unions that built the middle class, rather it was the huge factories of post-war America.

In today’s labor-intensive jobs, individual skill and talent drives wages far more than occupation. The wage variation among actors, physicians or financial advisors are mostly determined by how good they are in that job. This holds even in construction trades and in jobs like food servers, chefs or yoga instructors. These are market outcomes, not primarily the result of government policies or business practices.

The combination of automation and changing demand has proved especially good for well-educated workers in service sector occupations. Technology tends to complement the skills of better-educated workers. But, for less well-educated workers, the results have been poor. Technology is typically a substitute for the skills of less well-educated workers. For example, better computing has boosted wages for accountants, but it has clobbered the demand for bookkeepers.

There are many factors at work in altering labor market outcomes, such as regional monopolies, rapid urbanization and occupational licensing. But, the large and persistent polarization of labor markets is incessantly remaking the American economic landscape. This issue isn’t necessarily good or bad, just different from what we have been accustomed to experiencing.

Along with the polarization of jobs, regional economies are also polarizing. Since the 1980s, metropolitan areas and states have become less alike in productivity and wages. The rich places are growing richer and the poor places have stagnated, or even declined. More than 70% of the wage differences between states is explained solely by the share of adults with a college degree. Regional economic prospects in the 21st century will be almost wholly caused by differences in educational levels. Labor market polarization will amplify regional polarization.

I write this again because too few really understand that this one single factor drives almost all regional growth. Individually, college graduates have enjoyed three decades of wage growth, an unemployment rate that is half that of high school grads, and a labor force participation rate 25% higher. There are many more adults without a college degree than those who have them, yet 80% of all job growth for three decades have been among those with a college degree.

Labor market polarization will continue, and we will observe three important trends. The first of these will be increasing adoption of technology to cut workplace costs. The second will be loss of jobs in easily automatable occupations, such as warehousing, transportation, manufacturing and construction. An astute reader will identify this as most of the workforce training Indiana offers.

The future will also bring continued growth in job openings for both high- and low-wage workers. However, the demand for these jobs will vary across location according to how many workers of each type are available in each region. The most jobs will migrate to places that are better educated, which in turn will attract better-educated workers. Of course, this trend won’t continue forever, just another 50 to 100 years. 

Michael J. Hicks, PhD, is the director of the Center for Business and Economic Research and the George and Frances Ball distinguished professor of economics in the Miller College of Business at Ball State University.