The Rise and Fall of the Labor Force

In keeping with today's emerging "labor" theme, I now turn to David Leonhardt's article on the mystery that is puzzling economists:  why is American GDP recovering robustly, while employment growth lags other nations, and our own history?

This is not a new mystery.  You may be old enough to remember the ponderous articles on the "jobless recovery" that appeared like clockwork to punctuate the passage of months of the early Clinton administration.  You are almost certainly old enough to recall similar fare under Mr. Bush.  Before the 1990s, recessions seemed to have a clear pattern--jobs would be lost during the downturn, then roar back along with the rest of the economy.  Since then, we've seen a different pattern: the jobs go away, and the workers take a long time to find another.  In 2003, Erica Groshen and Simon Potter offered a possible explanation for this, in a paper for the Fed.  Unemployment was lagging, they suggested, because while previous downturns had mostly featured cyclical unemployment--employers laid off, then rehired when demand picked up--the more recent recessions were producing mostly structural unemployment: whole industries and job categories were going away, which meant that a lot of laboriously acquired human capital (both skills and networks of contacts) was being destroyed.  Those workers were taking a lot longer to find new jobs.

There's been an interesting back and forth over this theory in the blogosphere recently; like Leonhardt, I particularly recommend Tyler Cowen and Paul Krugman on the topic.  Leonhardt's suggestion for the quandary, heartily seconded by most of the left-leaning blogs in my RSS reader, is simple: as the labor movement has declined, workers have lost bargaining power.

Relative to the situation in most other countries -- or in this country for most of the last century -- American employers operate with few restraints. Unions have withered, at least in the private sector, and courts have grown friendlier to business. Many companies can now come much closer to setting the terms of their relationship with employees, letting them go when they become a drag on profits and relying on remaining workers or temporary ones when business picks up.

Just consider the main measure of corporate health: profits. In Canada, Japan and most of Europe, corporate profits have still not recovered to precrisis levels. In the United States, profits have more than recovered, rising 12 percent since late 2007.

For corporate America, the Great Recession is over. For the American work force, it's not.

I think I can tell an intuitively compelling story where this is true (though Adam Ozimek disagrees):  unions fight layoffs harder than single workers can, so more powerful labor means fewer firings.

However, there are (ahem) some problems with this theory, and with the conclusions that others have drawn from it--leaping into lamenting the loss of labor protections.  For one thing, if it is true that labor prevents companies from firing during downturns, it almost certainly means that there is less hiring during upturns--if you can't fire workers when you're losing money, you tend to hire as few as possible in the first place.  I can, with a lot of squinting, read the Europe/US data disparities to confirm this slightly less inspiring story of unionization lowering the volatility of employment--but you have keep your head moving in a sort of figure eight to prevent Spain's unemployment data from crossing your field of vision.

But the larger problem, I think, is that the decline of the labor movement is not an uncaused cause.  I can tell a pretty plausible story where the decline of the labor movement isn't itself causing the changes in employment trends--rather, it's a symptom of other forces which cause both the decline of unions, and more volatility in the labor market.

You can tell a lot of these "just so" stories, some more plausible than others.  Maybe the entry of women into the workforce undercut unions (bigger labor supply=less labor bargaining power, plus anecdotally, wives used to exert a lot of pressure on their husbands to stick at steady, boring union jobs; the benefits were designed to give the biggest boost to families).  Perhaps it also made it easier, psychologically, to lay people off who weren't the sole support of their families. Or perhaps the more rapid pace of structural change in the economy has decreased the returns to unionization: there's not much point investing a lot of time in a union if your job description will be obsolete in five years, and you can't expect much loyalty from a firm if you're planning to leave whenever someone else offers you more money.  Firms try to avoid layoffs whether or not there's a union because of the effect on morale; perhaps all these changes have lowered the morale cost of firing workers.

I'm not saying that all the bloggers who wrote about this are unaware of these trends; indeed, many cited them.  But they tended to assign major causal weight to the decline of the labor movement, or worker bargaining power, when those might simply be side effects of some third factor.

The most plausible third factor, to my mind, is one that not many bloggers did mention: companies are simply more competitive than they used to be.  The mighty labor-industrial complexes of the postwar era were mostly cosy oligopolies; there was a lot of value for labor to extract because they didn't have to worry so much about losing their customers.  Those cosy oligopolies had cosy relationships with bankers, who lent them money to keep overstaffed in downturns without much thought of how the depositors would feel, and the depositors didn't care because their heavily regulated accounts paid exactly the same interest rates as every other bank, and were federally insured.    The CEOs, cosily safe from interference by meddling outsiders like shareholders, could amass vast piles of cash to keep workers on even when there was little work for them, as well as building their conglomerated empires.

There were upsides and downsides to this arrangement, of course, but this is not the blog post in which to discuss them.  Rather, the point is that when this regime ended, it made it much harder for labor unions to bargain: whatever arrangements they struck suddenly had to take customers, competitors, and shareholders into account, rather than simply arguing with management over their share of a relatively fixed pie.  This made unions much less valuable to workers; it also, independently, put pressure on companies to fire workers during downturns.  Do we have any evidence that a more unionized workforce would have prevented the layoffs made necessary by a much more competitive environment?

This story--the story of more competitive industries, not less-organized ones--is arguably the best fit with the data.  It explains why our recessions now look different from our recessions in 1960, and also why our recessions look different from recessions in Germany or France, where domestic firms enjoy a lot more protection from competition on various levels than American firms do.  Maybe it even explains why someplace like Spain, where the markets are much more open than they used to be, are having so much trouble.

Or maybe not; I don't want to claim too much for any one theory.  But it's at least worth thinking about.