There is No Labor Shortage

by Arnold Kling

"Two main lessons derived from [my experience as an economist] are:

1. Economists do not know very much

2. Other people, including the politicians who make economic policy, know even less about economics than economists do."
--Herbert Stein, Washington Bedtime Stories, p. xi

Late in the fall of 1997, I attended an annual meeting of Management Recruiters International, a headhunting firm that focuses on mid-level technical, management, and sales positions. They were a sponsor of a web site that I had launched a few years earlier, which offered relocation information.

A keynote speaker addressed the conference on the outlook for the executive recruiting industry. He forecast a 20-year shortage of workers, which meant that the staffing industry would not have to worry about finding positions for individuals--it would only have to worry about finding individuals to fill an abundance of positions.

The speaker made his presentation using fancy charts and graphs. He obviously had studied the data extensively, and he cited a recent study by the United States Department of Commerce that also forecast a long-term labor shortage.

I had never considered the issue before. I had been putting all of my thought and energy into my own business, as any entrepreneur can appreciate. I was not familiar with the data or the previous studies that the speaker referenced. Everything about the presentation, in an auditorium packed with hundreds of rapt listeners, suggested that he was an expert on the subject. And I was absolutely certain that he was wrong.

I knew he was wrong because the notion of a long-term shortage is a violation of elementary economics. Economics professors reveal most of our knowledge to first-year students. Not much is held back for later courses. Certainly, no one has to wait for an advanced course to be exposed to the proposition that supply and demand curves intersect--that the price system adjusts so that neither demand nor supply is in excess. Nonetheless, in the media and government, non-intersecting supply and demand curves are drawn with disturbing frequency.

For example, in the 1970's the public was told to expect a chronic shortage of oil. Graphs depicted "demand" for oil increasing indefinitely, getting further and further away from "supply," which was shown increasing less rapidly, or perhaps falling. Thus, an "energy crisis" was proclaimed. Billions of dollars of taxpayer money were wasted on research into synthetic fuels and pouring oil down a hole in Louisiana called the "strategic petroleum reserve."

In contrast, the proposition that supply and demand curves intersect implied that there was a price that would eliminate the oil shortage. When the market mechanism was tested by the Reagan Administration, it proved to work.

Late in 1997, the economy was booming, which led to talk of a "labor shortage," or a "skilled labor shortage." At the MRI conference, the keynoter showed graphs of "demand" for workers growing faster than "supply."

The causal factor in his analysis is the proposition that demographic trends imply slow labor force growth in the U.S. relative to overall population growth. The baby boomers will reach retirement age, while a smaller cohort enters the working age. A smaller cohort of working-age Americans implied a labor shortage.

Pose the question to any first-year economics student: what happens to the new equilibrium in the labor market if there is a decline in the working-age population? The answer will be that employment is reduced and the wage rate increased. The increase in the wage rate is what chokes off the incipient shortage. At higher wages, more people are willing to work. As the late-1990's boom continued, elderly people came out of retirement and young people dropped out of college to take advantage of the available opportunities. Also, at higher wages, businesses try to get by without hiring new workers. Greater efficiency and outsourcing (using overseas workers) were responses to the tight conditions of the labor market.

Fortunately, a foolish prediction of a 20-year labor shortage probably did not cause the staffing industry to make any mistakes over its relevant time horizon, which probably is closer to 20 days. More ominous were the rumblings from the Department of Commerce, which had published its study purportedly showing that the U.S. faces a future scarcity of skilled labor.

The prediction of a skilled labor shortage gave me the feeling that we were being softened up for some foolish policy initiatives: perhaps a crash program to develop "synthetic nerds," or maybe an underground facility to store network engineers as a "strategic reserve."

In 1997, one of the biggest shortages was for computer programmers with proficiency in a new language called Java, developed by Sun Microsystems. I posed this question:

Suppose that we took a survey of all employers and asked, "At a salary of $30,000 per year, how many Java programmers would you hire?"
Adding up the answers, one might obtain a somewhat larger total than the number of people willing to supply Java programming skills at that salary. But I conjectured that a shortage of Java programmers would disappear if we surveyed employers and programmers assuming a salary of $300,000 a year. Somewhere between $30,000 and $300,000 is a salary that will balance demand and supply.

My prediction was that wages for computer programmers would rise too slowly in the "Dilbert" sector of the economy (government agencies, financial services companies, telecommunications companies, and other firms whose primary business is not software, but who currently maintain large software development organizations). Consequently, the best programmers would leave the Dilbert sector to join companies whose sole focus is software development. I viewed this as a healthy economic process, but it was widely interpreted as a shortage.

The natural economic tendency is for resources to move to where they are most productive. Market prices act as signals to bring about these shifts. The market price for computer programmers has soared in recent years. Small, specialized software companies are more effective than the Dilbert sector at using programmers. Thus, they can afford to pay programmers higher wages, while the Dilbert sector cannot. The more complaints that one hears from such firms of a shortage of technical workers, the more confident one can be that market forces are working to allocate technical workers to their most productive uses.

Late in 1997, I wrote that

Although most economists would share my confidence that the market can take care of a labor shortage, there is much that we do not know. We do not know how far away the current wage rate is from the one that is consistent with no excess demand for labor. We do not know if the process of wage adjustment will be inflationary (nominal wages rising) or deflationary (prices falling relative to wages). We do not know how long the process may take.

Noneconomists, who forecast a chronic labor shortage as if there were no equilibrium wage rate, know even less than economists do.

As it transpired, it took a little over two years for the shortage in technical workers to evaporate. Once the Internet Bubble burst in March of 2000, the demand for programming began to slip. By 2002, the shortage was not in workers but in jobs. There were computer programmers complaining that they had been out of work for months, and they talked about sinister forces, including the outsourcing of work to India and elsewhere.

The shortage of computer programming jobs, like the shortage of workers that preceded it, had no permanent basis. One could be confident that wage rates will adjust, leading some workers to retire, return to school, or switch fields, and leading some firms to increase their hiring of software developers.

Herbert Stein was right. Economists may not know much, but we know more than non-economists. We know that wherever the price system is allowed to operate, shortages cannot last long.