Given all the economic subjects syndicated columnist Robert Samuelson could be outraged about, the claim that the current U.S. recovery is creating mainly low-wage jobs seems a strange choice. In his May 20 column, he writes emphatically, “It turns out that this is wildly misleading and that the economy’s employment profile — the split between high- and low-paying jobs — hasn’t changed much since the recession or, indeed, the turn of the century.”
Even if you look at the numbers more closely than the recent research cited by Samuelson, you find that he’s correct – but only in the most technical sort of way. Much more important, though, a detailed examination of wage data shows that low-wage jobs represent a share of total U.S. employment with which no one should be happy, and that this percentage has been growing steadily over time. Moreover, this increase is especially disturbing considering some other labor force developments that have hardly been secrets.
Samuelson reports that he asked Elise Gould of the progressive Economic Policy Institute to “examine whether the recession has shifted the economy’s job distribution.” Her main finding: Hardly at all. According to Gould, in 2000, low-wage jobs (those that paid less than today’s average wage of $25 per hour, made up 24.4 percent of American workers. In 2007, the year the recession officially began (at the very end), this figure had risen to only 25 percent. And this year’s latest numbers show it only increasing to 25.7 percent. So the trend is going the wrong way, but as Samuelson emphasizes, “It’s striking how little has changed.” Moreover, he rightly observes that the recession can’t be the only cause.
To his credit, Samuelson mentions another economist who comes up with considerably larger absolute numbers for the low-wage share of total jobs (about 40 percent) but who contends that their increase has been “somewhat greater.”
I was pretty confident that Gould’s figures for absolute levels of low-wage employment were too low mainly because the categories she examined were too broad. As I’ve pointed out, even super-categories like “professional and business services,” which overall pays better than average, contains a large share of low-paid workers (in its case, 43 percent as of the latest – April – Labor Department monthly jobs report). I also suspected that much the same held for health-care services, which Gould and Samuelson regard as middle-wage.
When I ran my own numbers, breaking out some of the obvious low-wage sectors of health-care services (like home health-care aides), I did get higher absolute numbers of low-wage jobs than Gould – 36.55 percent of private sector workers as of April. But I was surprised to find that this share hadn’t risen much over the last few economic cycles.
In December, 2001, when the previous decade’s economic expansion began, low-wage workers comprised 34.38 percent of all private sector workers. By November, 2007, when that (bubble-y) expansion officially ended, percentage had actually dipped to 34.33 percent. When this recovery began, it stood at 35.50 percent. And by this April, low-wage workers represented 36.55 percent of the private sector workforce.
My numbers, therefore, also indicate that the current recovery has indeed seen outsized employment growth in low-wage industries, but that this problem has not gotten significantly worse.
So what’s the problem? I can think of two big ones, neither of which should have escaped Samuelson’s notice . First, since the Great Recession struck, American workers have been dropping out of the workforce like flies. As widely noted, the Labor Force Participation Rate – the share of Americans either employed or actively looking for work – fell sharply during the downturn and through the recovery, and now stands at near-four-decade lows.
Although that’s just as widely – and correctly – seen as economically damaging, one logical byproduct should be a greater share of Americans holding better jobs. Unless we really think that most of those who have exited the labor force are enjoying cushy retirements? As a result, even the modest rise that’s been witnessed in the low-wage share of total private sector jobs should be alarming given how few Americans relatively speaking are still working.
Second, a large, growing (even modestly) share of U.S. workers in low-paying jobs would make sense in an economy with strongly growing productivity. In principle, the better-paying sectors, which tend to be the most productive, would see their job gains at least restrained by dint of being able to generate ever more output with ever smaller payrolls. But America’s labor productivity (the type of productivity for which we have the timeliest data) has been growing weakly throughout the current recovery, and has actually fallen for the two straight quarters.
Moreover, even the traditionally high productivity manufacturing sector has been faring poorly lately in this critical respect – meaning that if its output keeps expanding (which has been the case), its employment increases should be especially robust. Yet job-creation in the low-wage sectors – which tend to be low-productivity, too – continues to outperform.
So let’s give a cheer-and-a half to Samuelson for dispelling the belief that America has lost the ability to create meaningful numbers of jobs that don’t involve asking “May I take your order?” or something similar. But as a result, we’re entitled to ask why he – and others – aren’t investigating why these jobs remain so prevalent, especially when so many other economic indicators tell us that the labor market should be looking much healthier.