I keep thinking about this graph.
It’s the increase in employment over each business cycle since the 1940s.
[Source: BLS Establishment Survey. I used annual data and chose peak years to avoid recessionary effects; the exception is 2007-2010 to show the impact of the Great Recession.]
In every decade/cycle since the 1950s, employment grew by roughly 20-30%. Except the last cycle, 2000-07, when it increased by a measly 4%, to be followed, of course, by the large losses of the last recession, the effects of which are still with us.
One thing you see here—and it’s the same thing you see when you look at the middle-class income data—is that the recession is really a problem on top of a problem. Previous cycles gave households a perch to fall from in the downturn that ended the cycle. In the 2000s, most working families didn’t climb much of a growth hill at all, so there wasn’t much to cushion the fall.
But the main thing I hear from audiences when I show this graph is the same thing I think about when I look at it and, I suspect, the question you’re asking yourself right now: why?
No one knows what the future of job growth holds, but it is absolutely legitimate to worry that the pace of growth in the 2000s expansion is the new normal. Here are some hypotheses for the lousy job growth of the last cycle:
—fast productivity growth: productivity did grow quickly, 2000-07, relative to previous decades, about 2.5%/yr, but the efficiency acceleration began in the 1990s, when job growth was strong. The difference here is demand—GDP growth was uniquely weak in the 2000s. But this is pretty circular logic—you can approximate GDP growth by adding productivity and job growth, so no new info here. You have to explain why demand was weak in the 2000s.
—lousy policy: regressive tax changes, the wars, asset bubbles, ignoring big imbalances (trade, budget and household budget deficits) surely played a role, but none of this is simple correlation: bubbles often generate growth until they pop (i.e., the dot.com bubble is implicated in the 1990s bar above), and military spending, as Krugman points out this AM, is stimulative too.
—allocative inefficiency: largely a theory at this point, but I think this one probably has explanatory legs. By pouring so many productive resources into financial innovating and engineering as opposed to that of other sectors, too much economic activity amounted to finding and exploiting tiny inefficiencies (e.g., gaining an arbitrage pricing advantage in a flash trade) and regulatory failures (subprime, rating agencies, bad underwriting). And that stuff just doesn’t provide enough economic opportunity for the broad middle.
—technology: yes, it’s been with us forever, and there’s a long history of economists worrying that labor-saving technology gains would wipe out jobs for humans. But some economic analysis and some of what I hear from people in the tech field suggest this one may have explanatory legs too. This piece reviews a new book on the issue by Brynjolfsson and McAfee (seems like author #2 could spare a vowel for author #1). It’s a short read and worth the effort—again, history is littered with predictions that we’re losing the race with the machines, but the fact is they’re getting a lot faster.
—inequality, bargaining clout, a weaker middle class: A bit circular here too, but I’ll bet this is important. My theory is that recoveries that leave out the middle class are less robust, shorter, more vulnerable to credit bubbles, and just generally weak.
More to come on this—it’s one of the most important structural questions in economics.