AP Photo/Keith Srakocic, File
The unemployment rate has been stuck at a low level of 4.1 percent for the last six months. The papers are replete with stories of labor shortages, of companies fighting over high-school students and hiring folks right out of prison. While pockets of labor-market weakness always exist somewhere, there’s little question that in broad terms, the U.S. job market is closing in on full employment.
And yet, wage growth has been underwhelming. I'll lay out the facts of that case in a moment—but how can that be? Don't tight labor markets boost workers' bargaining clout? Don't employers have to bid up paychecks to keep their workers from just going around the corner to better jobs? Isn't one of the key benefits of full employment that—as is not the case when labor markets are slack—employers must compete for workers? What's blocking that crucial economic dynamic?
Here's what I think is going on. First, the job market isn't as tight as the low unemployment rate suggests. Second, slow productivity growth is a constraint on wage growth. Third, a set of institutional and cultural factors—some old (too little union power), some new (monopoly power by megafirms)—are reinforcing the inequality that's long been embedded in our economy and are pushing back on wage growth, even at low unemployment.
It's trickier than you might think to establish what's up or down with wage growth. There are dozens of different wage data series, and they don't always agree. As well, wage growth comes in two flavors: nominal and real (inflation-adjusted). Even so, enough consistent patterns emerge that we can conclude the following:
Tight labor markets still boost nominal wage growth, just not as much as they have in past economic expansions.
A few simple figures tell this story. The first figure plots nominal hourly wage growth for middle-wage workers against the unemployment rate, showing that in the last few economic expansions, as the unemployment rate came down, wage growth picked up. In fact, the pickup was remarkably consistent: in each case, wage growth peaked at around 4 percent.
But not this time, at least so far. Though unemployment is just as low as it was in the past few expansions, nominal wage growth is barely hitting 2.5 percent.
Source: BLS
That 2.5 percent happens to be roughly the rate of inflation, meaning real earnings, at least by this series, haven't gone much of anywhere for mid-wage workers in recent years, as I've argued here.
President Trump's economists falsely claim that the tax cut is already showing up as real wage growth. Not so: real hourly and weekly earnings for mid-wage workers have hardly budged since the end of 2016.
The next figure shows two scatterplots of subsets of the data from the first figure. The plot on the left, covering the 1990s, shows a tight correlation between lower unemployment and faster wage growth. In the second plot, however, from 2010 through today, the dots are all over the map.
Source: BLS
To be clear, some other series show faster wage growth, and there’s some compelling evidence that minimum wage increases have helped boost low wages in states that raised their them. But in every series I’ve checked, the correlation between unemployment and wage growth is diminished.
Which begs the question: Why?
First, the job market is simply not as tight as the unemployment rate might lead you to believe. The media accounts cited above are not wrong, but they don't represent the national economy. The tame wage growth shown above is itself one indicator of the existence of slack (as is the absence of price pressures), but another important one is the employment rate of prime-aged (25-54) workers. That rate—a proxy for labor demand—fell sharply during the recession.
It's been climbing back, but, especially for men, it still has some ways to go. Prime-age men, whose employment rates have suffered a longer-term decline, have made back 76 percent of their pre-recession loss. Economists have warned that some unknown number of these workers, by dint of their skill or health deficits, are out of the reach of the labor force—their joblessness is structural, not cyclical. But I say: be careful about writing people off. The power of a piping-hot labor market to pull people in may be greater than you think.
Second, slow productivity growth is a real constraint on pay gains. From the mid-1990s through the mid-2000s, productivity, or output per hour, grew more than twice as fast as it has since then. But what does that have to do with wages?
To understand that, we need to invoke our third causal factor: inequality, and the high and growing imbalance in power between wage setters and wage earners. As noted, diminished union power is very much a factor, as is the increased concentration of dominant employers in certain industries, like Walmart in retail, which gives such employers the power to set wages below where the tightening labor market would predict. Then there’s labor market discrimination, housing segregation, unequal access to education and training opportunities, persistent trade imbalances, unwarranted Federal Reserve rate hikes, and the chipping away at labor standards designed to protect workers against exploitative employers (wage theft, misclassifying employees as contractors, eroded minimum wages and overtime rules).
When output-per-hour, or productivity, is high within a firm, employers can handily respond to tighter labor markets with higher wages, and yet still maintain their profit margins. But when productivity is low, then higher pay means lower profits (or higher prices, which also dings sales and profits), so employers will do whatever they can to stave of pay raises.
Put it all together, and you'd expect to see, even at low unemployment, precisely what we are seeing: flat real paychecks for mid-level workers amidst soaring profits, with the regressive tax cuts contributing much to profits and nothing to paychecks.
That means the job market is going to have to tighten up a lot more before workers have the clout to push up their paychecks. What 4 percent unemployment helped to accomplish in previous recoveries might well take 3 percent unemployment today. Higher wages will also require union power, more balanced trade, a patient Fed, the restoration of labor standards, regulation of monopoly power, and workplace voice for those who continue to suffer discrimination.
If that sounds like daunting agenda, that's because it is. But it is also the path to economic justice. And nobody said that path wasn't going to be steep.
