Why This Particular Recovery Is So Bad at Creating New Jobs
The good news: Economists are starting to come up with some decent theories as to why this recovery is so bad at generating employment. Now here's the bad news.
As Walmart goes, so goes the nation. With 1.3 million workers, the retail giant is the country’s largest private employer, so its hiring patterns tell us something about the health of the larger U.S. jobs market. And recent signs have not been encouraging. A June 13 Reuters survey of 52 Walmart stores (out of about 4,600) in all 50 states found that more than half were hiring only temporary workers, a practice Walmart has never adopted on so large a scale outside the busy holiday shopping season. Walmart spokesman David Tovar confirmed to Reuters that "flexible associates" had come to comprise nearly 10 percent of Walmart’s U.S. workers—compared to only one or two percent before 2013. A large chunk of the nation’s biggest private workforce, in other words, is just barely employed.
If we were in the midst of a normal economic recovery, Walmart wouldn’t be acting this way. To be sure, the nation is clearly on some kind of rebound: Four years after the Great Recession ended, consumer confidence is up, household debt is down, and residential construction is stirring back to life. But the unemployment rate is still above seven percent, and isn’t expected to drop below that level until next year.
The “great American jobs machine,” which once boosted employment during recoveries as quickly as it shed them during recessions, has rusted up. If the machine were operating properly, Walmart would be hiring permanent employees to meet rising consumer demand and stay competitive in a tightening labor market. For that to happen, though, the labor market would have to be tightening a whole lot more than it is. Why is it so stubbornly slack?
The consensus among most economists is that the U.S. economy doesn’t have a jobs problem so much as a gross domestic product problem. Boost GDP growth to a healthier level, the conventional wisdom says, and the jobs will follow. But there are a few economists out there who suggest that something deeper is out of joint. "Jobless recoveries" like the ones we’ve experienced from the last three recessions (1990-91, 2001, and 2007-09) may be the new norm, they say, because GDP growth just doesn’t create jobs as reliably as it used to. In effect, they argue, the economy is learning how to grow without creating jobs—or at least without creating nearly as many as it once did.
Businesses are reluctant to invest their record profits in new workers because they aren't convinced people can afford to buy new goods and services. Too many of them are unemployed!
THERE'S NO QUESTION THAT GDP growth has been weak since the Great Recession—even as corporate profits, weirdly, have reached historic highs. Annual GDP growth has been only about two percent during the recovery; a normal recovery would have started out with growth closer to five percent before falling perhaps to three percent.
There are three standard explanations for why this is so. First, there’s a strong argument that the government simply never provided enough economic stimulus to revive the recession-wracked economy. Second, and relatedly, there is ample data to suggest that Washington’s pendulum swing away from stimulus and toward policies of austerity—like the recent uptick in the payroll tax, which removed an estimated $115 billion out of workers’ paychecks, and the $85 billion in budget cuts that came with "sequestration" earlier this year—has created what economists call "fiscal drag" on the economy. And third, there’s the fact that this was no ordinary recession. Before they became known for an Excel spreadsheet error in a paper bolstering the case for austerity, the economists Carmen Reinhart of the University of Maryland and Kenneth Rogoff of Harvard argued in a still highly regarded 2009 book that recovery from a banking crisis—like, say, the 2008 financial meltdown—is qualitatively different from recovery from more a typical economic downturn. Specifically, housing prices stay depressed longer (check), as do output (check) and employment (check).
Of course, at bottom, economic growth is slow simply because demand is weak, and herein lies the chicken-and-egg relationship between unemployment and GDP. Businesses are reluctant to invest their record profits in new workers, and instead, they’re hoarding cash—because they aren’t convinced people can afford to buy the goods and services they sell. Too many of them are unemployed! This paradox isn’t unique to the current recovery, but it seems to be taking longer than it has in the past to resolve itself. It’s as if the gear ratio between GDP and unemployment, once stable, had unexpectedly slipped.
According to the standard view, first articulated in 1962 by the economist Arthur Okun of President John F. Kennedy’s Council of Economic Advisers, the relationship between unemployment and GDP growth never changes. In its simplest form, "Okun’s Law" states that a two percent drop in GDP growth will create a one percent increase in unemployment.
During most of the recession the economy obeyed Okun’s Law, but toward the end and during the recovery that followed it did not. One recent paper, released in May by the University of Wisconsin’s Menzie Chinn and the French economists Laurent Ferrara and Valérie Mignon, concluded that U.S. employment was about one percent below where it should be according to Okun’s theory. By their calculations, 1.1 million more Americans were out of work than should be the case.
What’s different now? One theory that’s gotten some attention lately is that automation is finally reducing the size of the workforce, just as dystopians have been predicting for a century. Ordinarily, productivity growth through technological innovation actually increases employment, because more output creates more wealth, which creates more worker demand. Automation eliminates certain jobs, sure, but as the economy grows other jobs are created. Increasingly those new jobs require greater skill levels than the displaced workers possess, and the net result is greater income inequality. That’s a problem. But the overall number of jobs still increases. That’s the way it’s worked thus far. But it’s no longer true, according to the economists Erik Brynjolfsson and Andrew McAfee, both of MIT’s Sloan School of Management. Their argument boils down to the observation that growth in productivity—economic output per total hours worked—and growth in employment stopped rising in tandem at the start of this century. Robots, in other words, have started to eliminate more jobs than they’re creating.
It’s an interesting hypothesis, but entirely speculative; Brynjolfsson and McAfee simply don’t have the data to prove that the fall in employment is in any way connected to the rise in productivity from automation.
A better theory concerns state underemployment. About one-sixth of the U.S. workforce consists of government employees, most of whom work for state and local governments. Governments have to balance their budgets, so when recessions lower tax receipts they lay off state and local workers. During recoveries, the process is typically reversed. As tax receipts rise, state and local workers are rehired. But that doesn’t seem to be happening in this instance. Princeton economist Alan Blinder pointed out in a June Wall Street Journal op-ed that during the past three years, all net employment growth came exclusively from the private sector. Private business created 6.56 million new jobs, while government lost about 1.14 million jobs. This is terra incognita. “Never before in postwar history,” Blinder wrote, “has government employment declined during a recovery.”
Why aren’t state governments creating jobs at the same pace they did during earlier recoveries? Josh Bivens of the Economic Policy Institute, a Washington non-profit, thinks the changing composition of state legislatures is "probably an important fact." In 2010 Republicans wrested control of the majority of state legislatures away from Democrats, achieving a level of GOP dominance in state politics not seen since the 1950s. Despite President Obama’s election victory, Republicans maintained that dominance in 2012. At the same time, state legislatures, which traditionally have been more collegial and less partisan than the U.S. Congress, have gotten significantly more polarized ideologically. The result can be seen in pitched battles over public-employee unions, taxation, and the size of government.
Perhaps the most dismaying theory about what’s different comes from Harvard economist Richard Freeman, in the form of a history lesson. In a forthcoming paper, Freeman suggests that, from the 1950s through the 1970s, American firms tended to hoard labor during recessions rather than lay people off. He arrives at this interpretation by looking at two streams of economic data: In those years, he points out, America’s loss in output during recessions was, percentage-wise, greater than its loss in employment. Logically, the two should decline at the same rate, as firms “right- size” to meet lesser demand. But in those years firms held onto their workers—not because they were softhearted but because “they expected short recessions and wanted experienced workers on board to work in the recovery.” And since the employees who weren’t laid off had less work to do in a down economy, output shrank.
That pattern started to change—and ultimately flipped on its head—Freeman argues, after the 1970s. Instead of decreasing during recessions, productivity (i.e., output per worker) started increasing during them. In the 1990-91 recession, employment loss bottomed out at minus 1.7 percent while output loss bottomed out at minus 1.2 percent. That meant productivity increased by 0.5 percent.
In the 2001 recession, employment loss bottomed out at minus 1.2 percent while output bottomed out at minus 0.3 percent. That meant productivity increased by 0.9 percent. And in the 2007-09 recession, employment bottomed out at minus 6.3 percent while output bottomed out at minus 4.7 percent. That meant a 1.6 percent increase in productivity. In modern downturns, fewer and fewer workers do more and more work.
In turn, Freeman writes, recoveries have become more “jobless,” with rebounds in employment increasingly lagging behind rebounds in GDP. Freeman suggests this change is the result of declining union power and a growing tendency to “lay off workers first and ask questions later.” Why worry, after all, if the result is rising productivity?
Which brings us back to Walmart. Asked to explain why the company was hiring so many "flexible associates," the company’s spokesman denied it was a cost-cutting move and said merely that it was to ensure that stores were "staffed appropriately." It’s hard not to conclude that Walmart just doesn’t want to hire permanent workers, even if the result of underhiring is that it will find itself struggling next year to hang onto employees as the economy heats up.
One possible explanation, then, for why this recovery is so bad at creating jobs is that businesses are more resistant than they used to be to hiring, period—and that they can afford to be because unemployment takes less of a toll on GDP growth than it once did. An anti-employment stance by American business would not appear a successful long-term strategy, since at some point unemployment really is bound to be an impediment to economic growth. But if it’s not as big an impediment as it used to be—and if corporate profits are already sky-high anyway—then perhaps American business has become less impatient to reach full recovery. A poky one might do just fine.