Lots that could be said about Ruth Milkman's op-ed on labor unions in the New York Times today, but let's start with this sentence: "Employers have abandoned the paternalistic job security measures, pensions and fringe benefits of which they boasted a few years ago." That's quite right, but it also seems to oversimplify the story of how firms have restructured and reorganized quite drastically over the past 30 years, as well as how that change has affected job security, mobility, and overall economic inequality in America. This story's obviously a complicated one, but I think it needs to be spelled out briefly, because Milkman's claim that workers are in the same plight they were back in 1935 strikes me as misleading in fairly important ways.


In 1935—in fact, through the post-World War II era and up to the mid-1970s, American industry was largely organized along mass-production lines. Large firms dominated, growth depended mostly on market stability and expansion, and, in the workplace, control was divvied up between unions and management. Labor negotiated wages and benefits, agitated for better working conditions and grievance procedures, leaving management to the planning and decision making. For unionized workers, wages and advancement were determined largely by seniority, and workers often stayed with a single firm. (Lay-offs would happen, especially for those low on the union totem pole, but many would be quickly re-hired by the same firm during a recovery.) For higher-level workers, too, promotion within a firm was quite common. Most workers started at an entry-level position, acquired a lot of firm-specific knowledge and skills, would learn on the job, would train for higher positions, and would receive a fatter paycheck with each promotion. For workers of all skill levels in these firms, job security was solid and the pay was good. (Obviously, as Ed Kilgore has noted, the 1950s and 1960s was terrible for those who didn't get to latch onto the new industrial bonanza, but leave that somewhat aside.) The large firms, on the other hand, were quite constrained: it was difficult to fire workers, it wasn't easy to cut pay (since pay was linked with one's rigorously-defined position), and it was more difficult to hire from the outside than to promote from within. In this sense, then, mobility was a relatively natural process—workers could "advance" fairly easily within their firms, whether they were unskilled or skilled.

Eventually, of course, American firms decided that they could no longer afford this practice. In the mid-1970s, productivity growth was sagging, and bold new managerial models became all the rage. Many companies decided that they couldn't maintain their rigid "internal labor markets," or the job security and well-defined job classifications that went with them, and decided that labor flexibility was the yellow brick road to competitiveness. This theory applied to both high-skilled workers and low-skilled workers, although the latter were hurt far more, and businesses' attempts to create more "flexibility" among low-wage workers have been much-aided by the loosening of labor legislation over the last three decades. Meanwhile, the weakening of internal labor markets has had a major effect on mobility, wage growth, and inequality: Median wage growth by mid-career fell by 21 percent in recent years; there are now 40 percent fewer workers in the central part of the wage growth distribution; etc., etc. (See this book for a long string of evidence that insecurity has increased and mobility declined over the past 30 years.)

At the same time, not all businesses emphasize finding profits by cutting wages—i.e. avoiding unions, subcontracting, making better use of unskilled workers—so maniacally. Some firms emphasize innovation as the path to success, and focus on investing in workers, having employees perform a variety of tasks, offering new incentives to workers, etc.—which can in turn lead to high wages, job security, and a decent shot at advancement and upward mobility. Of course, you have to be lucky enough to get those jobs. In practice, many firms use both methods of organization—the "high road" and the "low road," which means that if you're on the high end, life is still pretty good. If not, then not. (One argument for raising the minimum wage is that it will give more firms incentive to take the "high road" invest in more of its workers, since "low road" practices, like subcontracting and rapid hiring and firing will become less profitable.)

At any rate, as one can see, there's a rather vicious "two-tiered" system being put into effect within firms. I've been reading a new book from the Russell Sage Foundation, Moving Up or Moving On, which finds that low-earners do much better if they "move on" to different employers rather than "move up" in the same firm. It's not quite like the old days. Stronger unions would ameliorate many of the ill-effects here—especially declining wages—but not all; things wouldn't go back to the post-WWII "glory days" by any means. Meanwhile, although I've argued in the past that the education gap doesn't explain the rise in economic inequality over the past few decades, it is the case that the change in firm structure has affected workers of different education differently. Firms no longer invest quite as much in on-the-job training—why bother, if a worker isn't going to be with the firm for life, as used to be the case? To remedy this problem, policy wonks usually propose variants on worker training or education. Meanwhile, Timothy Bartik has proposed "labor-demand policies"—government incentives to "induce employers to provide more or better jobs"—which seem promising.

One other key question is, To what extent has the change in firm structure been dictated by actual economic necessity—by the "market"—and to what extent by changes in the legal landscape? Changes in labor regulation have obviously had an effect, but it's also not entirely obvious that American firms could have continued to thrive with the more restrictive labor policies of the 1950s and 60s. Maybe they could have; I just don't know. Same with the deregulation of unionized industries since the 1970s—to what extent has this affected mobility and wages? And to what extent was it necessary? Meanwhile, the shift in governance towards shareholder control, and the greater emphasis on the short-term "bottom line," has also affected how firms operate. This is a difficult story to untangle, and I'm not up to it right now.

At any rate, while unions are not obsolete by any means, it's not clear that they alone can counteract the various trends that have caused businesses to reorganize and restructure in ways that have, quite dramatically, increased job instability and decreased mobility over the past thirty years. On the other hand, one hugely important thing unions can do, as Matt Yglesias pointed out in a post that disappeared somehow, is provide a counterbalance in Washington D.C. in the push for more labor-friendly government policies, and defeat the business lobbyists. I've discussed before how the AFL-CIO is badly overmatched on the lobbying front nowadays, and it's a real problem.