With U.S. unemployment at a 50 year low, why is job quality so poor?

The numbers tell one story. Unemployment in the US is the lowest it’s been in 50 years. More Americans have jobs than ever before. Wage growth keeps climbing.

People tell a different story. Long job hunts. Trouble finding work with decent pay. A lack of predictable hours.

These accounts are hard to square with the record-long economic expansion and robust labor market described in headline statistics. Put another way, when you compare the lived reality with the data and it’s clear something big is getting lost in translation. But a team of researchers thinks they may have uncovered the Rosetta Stone of the US labor market.

They recently unveiled the US Private Sector Job Quality Index (or JQI for short), a new monthly indicator that aims to track the quality of jobs instead of just the quantity. The JQI measures the ratio of what the researchers call “high-quality” versus “low-quality” jobs, based on whether the work offer more or less than the average income.

A reading of 100 means that there are equal numbers of the two groups, while anything less implies relatively lower-quality jobs. Here’s what it looks like:

So, what is this newfangled thing telling us? Right now the JQI is just shy of 81, which implies that there are 81 high-quality jobs for every 100 low-quality ones. While that’s a slight improvement from early 2012—the JQI’s 30-year nadir—it’s still way down from 2006, the eve of the housing market crash, when the economy regularly supported about 90 good jobs per 100 lousy ones.

Or, in plainer English, the US labor market is nowhere near fully recovered from the Great Recession. In fact, the long-term trend in the balance of jobs paints a more ominous picture.

“The problem is that quality of the stock of jobs on offer has been deteriorating for the last 30 years,” says Dan Alpert, an investment banker and Cornell Law School professor who helped create the index. (Along with Alpert, the index is built and maintained by researchers at Cornell University Law School, the Coalition for a Prosperous America, the University of Missouri-Kansas City, and the Global Institute for Sustainable Prosperity.) The “whole story” told by the index, he adds, is “the devaluation of American labor.”

How low can you go?

To grasp the JQI’s significance, it’s worth understanding how it’s made. The researchers take private-sector non-manager jobs—which make up 82% of all private-sector jobs—and divvy them into two groups: the “high quality” jobs that pay more than the average weekly wage and tend to have more hours per week, and the “low quality” ones that pay less and offer fewer hours. The index is a weighted ratio of the two, averaged over the previous three months to cut out the noise. 1

The prevalence of low-quality jobs suggests that there’s still a lot of slack in the labor market—meaning, people could be working more or using their skills more fully than they currently are. This is pretty much the opposite conclusion you’d draw from the ultra-low unemployment rate, robust job creation numbers, and other conventional headline data.

That’s why, starting next month, the JQI will be released on the same day as the closely watched monthly jobs data published by the Bureau of Labor Statistics. The hope is to offer policymakers a more granular picture of America’s economic health, says Alpert.

Take, for instance, the Federal Reserve. “Had they viewed the market through this lens, they might not have [started raising interest rates in 2018],” says Alpert. “That’s the type of policymaker we’re trying to address. It’s a reality check on the conventional data.”

America’s shift to a crummy-job economy

Since 1990, America has cumulatively added some 20 million low-quality jobs, versus around 12 million high-quality ones. In short, the US economy has shifted toward creating more bad jobs than good.

This has significant implications for the economy. For instance, inflation-adjusted income for workers with high-quality jobs has grown at a brisker clip than it has for workers in low-quality ones.

That point might seem a little tautological; after all, the researchers’ definition of “high quality” rests on average weekly wage, a big factor in determining income. However, wage growth isn’t what’s really driving that divergence. Rather, it’s a sharp split in hours worked. Those in low-quality jobs now clock a mere 30 hours a week, on average, down from 31 in 1999. That compares with an average 38 hours a week for high-quality jobs.

On an individual basis, that seven-hour gap might not sound like a lot. Add it up, however, and it reveals the stunning scale of America’s “underemployment” problem. If the average low-quality job offered the same number of hours as the typical high-quality job, it tallies up to about 480 million hours per year. Those unworked hours—representing labor that could have contributed to the US economy—represents around 12 million forgone jobs every year.

Of course, some of that decline may be due to workers seeking to curb their hours voluntarily. Another key reason is that employers sometimes limit hours to keep from having to pay benefits (a problem that, by the way, European workers are facing too).

Overall, the hours-worked deficit reflects a much larger problem. The swelling share of jobs that offer lower-than-average incomes means that job growth, as reflected by a super-low unemployment rate, implies less and less spending power than in the past. “As you create these jobs in lower quality areas,” says Alpert, the economy gets “a lot less bang for its buck.”

The dramatic rise and fall of a couple key industries brings this point into sharper focus.

The collapse of US manufacturing

Throughout the second half of the 20th century, US manufacturing supported around 12.8 million production jobs. Then, all of a sudden, America started hemorrhaging factory jobs. By 2010, the number had shrunk to levels last seen in 1939, when the US population was two-fifths of its current size. Although it’s recovered a bit since, manufacturing jobs are still down more than a quarter since 2000.

Nearly all of the manufacturing jobs lost, say the JQI researchers, have been replaced by just four service-sector industries: retail, administration and waste services, healthcare, and leisure and hospitality. In all of these industries, jobs with lower-than-average income predominate.

“If you travel around this country and manufacturing areas, it is not hard in a restaurant to find somebody who was once a manufacturing worker,” says Jeffrey Ferry, an economist at the Coalition for a Prosperous America, a nonpartisan group representing labor and manufacturing interests.

This isn’t just a problem for hard-on-their-luck rust belt residents; it hurts the whole economy.

Factory jobs are good for economies. For one, they tend to pay well and offer steady hours—and not just because of unions or tradition, but because they produce a lot of value. Mechanization boosts the worth of people’s labor by allowing them to produce way more than they otherwise could. In the grand scheme of things, this increased productivity—the ability to produce more, faster—is what drives long-term growth and, ultimately, raises people’s living standards.

Of course, technology can boost human productivity in service-sector jobs, too—for example, how computers let bankers process payments much faster than when checks were cleared by hand. So, it might not matter if the lost factory jobs were replaced by high-tech service-sector jobs—the “knowledge economy” and all that.

But, alas, this hasn’t been the case. Waiting tables, selling shoes, mopping floors—there’s not much technology can do to improve on human labor required for these tasks. So, all else equal, swapping out a factory position for one of these jobs strips the technological value-boost from a person’s exertions—and, in aggregate, leaves the economy producing less than it otherwise could. And, indeed, sluggish labor productivity growth since the end of the Great Recession tracks with the trend in JQI (though there’s not yet enough data to determine anything conclusive).

With a growing share of jobs producing less value from human labor, it’s hardly surprising that more workers are earning offers lower-than-average incomes. The problem is, workers are also consumers. The less they earn, the less they’re able to spend, the poorer they feel, and the more they save. This gets to how the JQI might help explain why the unemployment rate can slip to near-record lows without powering faster growth and whipping up inflation.

What’s worse, this dynamic reinforces itself. Weaker spending discourages companies from investing in new equipment or technology. As incomes stagnate, workers grow increasingly desperate and accept lower wages. The combination of cheap labor and ever-fickler demand pushes firms to replace high-quality jobs with low-quality ones—causing the economy itself to lose vitality. The longer this goes on, the harder it is to switch back because, eventually, the structure of the economy itself prevents demand from reviving growth.

The key to understanding how to escape this perilous feedback loop, says Alpert, is to grasp what caused it to begin with.

A devastating double-whammy

On January 20, 1981, in his first-ever remarks as America’s new president, Ronald Reagan uttered one of his most famous phrases: “Government is not the solution to our problem; government is the problem.” More than simply a memorable line, those words became a whole ideology—one that Alpert argues underlies many of today’s economic woes.

The rise of this worldview nixed the US government’s traditional role of supporting demand and investing in boosting productivity (an authority it has yet to reclaim). America’s infrastructure took a big hit. The damage began to appear as early as the 1990s, when the US’s infrastructure deficit started to emerge, says Alpert. Jobs building and maintaining that infrastructure have long provided high-paying, full-time jobs for those who might otherwise struggle to find decent work, which is why he suspects this is a key reason for the JQI’s steady decline.

As unwise at this publish spending-aversion was, the timing of this shift was tragically unlucky, coinciding as it did with something unprecedented in human history.

The initial blow came from the Asian Tigers—South Korea, Taiwan, Hong Kong, and Singapore—that emerged as manufacturing powerhouses in the late 1980s and early 1990s. Then, following a fluky trio of events—the fall of the Berlin Wall, Tiananmen Square, and India’s sweeping economic reforms—the opening up of Eastern Europe, China, and India added more than a billion workers to the world’s labor pool under a single global capitalist system (depending on how you count them; Alpert puts it closer to 3.5 billion). This “Great Doubling” of the global workforce, as Harvard economist Richard Freeman wrote in 2006 (pdf), “presented the US economy with its greatest challenge since the Great Depression.” (On top of all that, advanced economies experienced a steady upsurge of women joining their workforces throughout the 1980s and 1990s.)

At the same time, a series of US-led trade and investment agreements—most notably, permanently normalizing trade relations with China in 2001—lifted the floodgates on foreign investment. As US companies moved factories overseas en masse, American manufacturing employment imploded. (Though plenty of politicians, economists, and media still argue that automation, and not trade, caused most of this collapse, research by economist Susan Houseman shows that trade was indeed the chief driver.)

This would have been bad enough as it was. But without public spending to replace the high-quality jobs lost, the mix in jobs shifted so severely that even near-record-low unemployment rates aren’t enough to revive the US economy’s former might.

Quality time

Many think the “China shock” is mercifully over. Not Alpert. “I’m still looking at billions of people who aren’t urbanized around the world who will continue to put downward pressure on global labor,” he says. So, what can be done to revive high-quality job creation? There are two possible approaches, he says. The first is protectionism and other more supply-side measures—basically, the tack Donald Trump has taken (this includes his trade wars as well as his tax bill and, to a lesser extent, deregulation).

As you can see in the chart above, the president’s policies—or, perhaps in some cases, confidence in his policies—managed to juice the JQI at three distinct points since Trump won the election. But those effects were fleeting, likely because the problem isn’t that businesses had too little money to spend. Rather, the lack of demand gives them no good reason to spend it.

This brings us to the second option: valorizing the traditional role of the government as a creator of demand. And the ticket here, argues Alpert, is big, sustained increase in infrastructure spending.

Consider the housing bubble-related bounce in the JQI in the mid-2000s. That was in no small part driven by the surge in construction spending. And, sure, building McMansions in empty suburbs was never going to be terribly productive in the long term. But it’s a different story with new bridges and railways. Those sorts of projects tend to boost overall productivity, by making it easier and cheaper for capital to move around the economy. And, of course, they also require putting America’s vast, neglected pool of desperate workers to far better use.