Any discussion of the labor market’s condition, challenges, and opportunities should begin with a review of the basic data. So many think pieces and cocktail conversations depend upon smart-sounding formulations like “The Big Quit” and “Secular Stagnation” and “The Rise of the Robots” that even quite sophisticated analysts begin to assume some evidence must underlie the painfully self-contradictory pontification. As we all know, so many workers are abandoning the labor force that employers cannot find the needed labor. Also, automation has rendered workers superfluous so quickly that employers have no need for them. And didn’t you know that wages have stagnated because productivity stopped rising? Do try the tuna tartar.
The data tell a much less interesting story, though fascinating in its own right: None of these narratives is true.
Part I: The Post-COVID Labor Market
The COVID-19 pandemic, and the response by governments, caused huge disruptions in the American economy. Millions of workers temporarily exited the labor market, and many were slow to return. In parallel, inflation surged to levels not seen in decades, and economists and employers began warning that a “labor shortage” was both fueling this inflation and threatening the economy’s longer-term prospects.
Clearly, the days of shutdowns and quarantines wreaked all manner of havoc, but whatever trends seemed to be emerging have since faded away. The trend that has persisted is one of wages declining in real terms.
There is no Big Quit. The share of work-eligible American adults who were working or looking for work, known as the labor-force participation rate (LFP), stood at 62.5% in February 2023, down less than one percentage point from before the pandemic. This was slightly above the Congressional Budget Office’s 2018 forecast, which represents a good estimate of what economists anticipated without any knowledge of recent disruptions.
Among prime-age workers, between the ages of 25 and 54, LFP has risen by two percentage points over the past decade and is higher than when COVID struck.
Women’s LFP has received particular attention, on the theory that a shortage of childcare is holding them back from working. In fact, prime-age women’s LFP recovered faster than men’s and in February 2023 stood above its pre-COVID level.
Another popular claim has been that an immigration slowdown is suppressing the labor supply. But while the native-born workforce is unchanged in the three years since the pandemic struck, the foreign-born workforce has grown by more than 2.5 million workers, compared with an increase of only 1.0 million in the prior three-year period.
The actual labor market crisis occurred more than a decade ago, from 2007 to 2013, when overall LFP fell by 3.6 percentage points and prime-age male LFP did as well. Put another way, 2.2 million fewer prime-age men were in the labor force in April 2014 than had their LFP remained at the January 2001 level. That was the era of rapid offshoring and a weak recovery from a massive financial crisis. Employers could find workers whenever they deigned to hire and so had no concern with the state of affairs, or sense that the workers they were abandoning then might be of value at some later date. The real question was why the jobless had not responded promptly by relocating and learning to code.
The ultimate proof of the labor market’s condition can be found in the wage trend itself—representing the price of labor. What sort of wage spike would fairly indicate a labor shortage? The question is an interesting hypothetical one, but irrelevant in practice. The most salient fact about the 2023 labor market is that real wages are falling: lower in February 2023 than February 2022, which was lower than February 2021; roughly the same in early 2023 as before the pandemic struck in early 2020.
This picture remains the same over the past decade for which occupation-level data is available, in a variety of occupations where employers claim to be having trouble finding workers. Economy-wide productivity grew at an annual rate of 1.3% and wages across all occupations grew at 1.1%, yet carpenters’ wages rose by 0.1%, machinists’ by 0.2%, truck drivers’ by 0.6%.
In short, when the business community warns of labor shortages, the correct response should be laughter.
Part II: Unproductive Complaints
Up until the moment when economists and business leaders began complaining of a labor shortage, they advanced a contrary narrative that in fact the labor market’s problem was rapid automation and thus too little demand for workers. But that surge in automation never occurred.
In its most basic form, automation means the substitution of capital for labor, whether that takes the form of new machinery or software. The result is that less labor is needed to produce a unit of output. Put another way, productivity increases. If automation were dramatically disrupting the labor market, productivity should rise quickly. The same amount of labor would be generating much greater output than before. This has not happened.
Productivity growth fell to its lowest level on record in the 2010s. Throughout 2022, with labor shortage complaints reaching a fever pitch, presumably the conditions in which incentives to automate would be strongest, economy-wide productivity fell for four straight quarters. Will robots someday take the jobs? Maybe. Are they doing so yet? No. If employers are trying desperately to make do with fewer workers, they have a funny way of showing it.
One problem underlying the failure to boost productivity is that business investment has been in steady decline for decades. When managers pull money out of their firms and return it to shareholders, rather than invest in capital equipment and growth, productivity gains are harder to achieve.
A concrete illustration of this failure to invest comes in the manufacturing sector, where American producers deploy a fraction of the robots used in Korea or Singapore and fewer even than in China. The United States looks more like Slovenia or Italy than Germany or Japan.
Thus, while economists have attempted to blame automation for the decline in manufacturing employment, the truth is that automation is occurring no faster than in the past. From 1947 to 1972, manufacturing productivity increased at 3.4% per year and output increased by a faster 4.2% per year. Thus, employment increased. Even at higher productivity, more workers were needed to serve the growing demand.
By comparison, from 2000 to 2022, manufacturing productivity grew more slowly than in the earlier era—no rising robots in sight. The problem, though, was that demand stalled thanks to globalization and offshoring. Output rose only 1.4% per year, thus many fewer workers were needed, and employment collapsed.
Incredibly, and in defiance of any economic theory of a well-functioning capitalist system, slowing investment and demand has created a situation where manufacturing productivity actually declined over the past decade. Factories needed more workers in 2022 than they did in 2012 to produce the same output.
Part III: The Big Picture
Low investment and slow productivity growth are symptoms of a larger problem in the American economy. Automation is one way to improve productivity, but anything that allows a firm and its workers to produce higher output, whether capital investments or process improvements or training, offers the same promise. No one ever worries about training or education “destroying jobs.” Historically, when productivity increases, the result is not job losses but rather output gains. Technology tends to augment what humans can do, not render them superfluous. Firms inevitably introduce it gradually, as they deploy investment and develop new processes, and it feeds their endless hunger for growth. Even when firms merely maintain output, and thus require fewer workers, the higher output allows them to better compensate the remaining workers and an economy in which productivity and prosperity is rising tends also to be one in which new and better opportunities will be available.
By contrast, in an economy where job losses result from firms offshoring productive capacity, or where gains from lower labor costs and other cost-cutting accrue to management and shareholders without reaching a firm’s remaining workers, new opportunities will be scarce and often worse. Overreliance on these latter mechanisms to increase profits in recent decades has harmed workers greatly.
For more than half a century, productivity, GDP, and profit have risen together. Wages have not followed suit.
Look at this 50-year snapshot. GDP per capita up more than 130%, productivity up more than 140%, profits per capita up more than 180%. Wages up 1%. How could this happen?
The reality is that wages do not rise naturally with these other measures. Wages rise under only one condition: when employers cannot hire the workers they need at the existing wage. The average wage went from $28 in 1972 (in 2022 dollars) to… $28 in 2022, because employers did not have to offer $29. They could offer $29—after all, the corporate sector had roughly tripled its profit even after accounting for population growth. But they did not have to, so they did not.
One reason that employers did not have to offer higher wages is that the labor market experienced an enormous influx of additional workers. More than half of the total increase in the American workforce over the past 15 years is attributable to immigration. Or, put another way, immigration caused the labor force to grow twice as fast as it otherwise might have.
Of course, as economists eagerly point out, immigration need not suppress wages if it creates new demand as quickly as it creates new supply. The question is the skill composition of the immigration.
Unfortunately, American policy has skewed the skill composition of its immigrant workforce toward low-wage jobs, the effect of which is to increase the supply of such workers much faster than the demand for such work. Immigrants account for roughly one-third of all workers in occupations that pay less than $30,000 per year, but only one-sixth of workers in occupations that pay more than $60,000 per year. These figures likely undercount illegal immigrants, which would skew the results further.
Policymakers have it within their control to alter this balance and create economic conditions in which employers must pay low-wage workers more, so that they finally begin to share in the gains that high-wage workers and shareholders have enjoyed for decades. What’s missing is the political will.