Though I often disagree with Justin Fox, I’m a fan of his writing. And so I was surprised by his recent discussion of Jake Rosenfeld’s new lament for organized labor’s decline, What Unions No Longer Do. I have yet to read Rosenfeld’s book, and it’s possible that there is a great deal that’s been lost in the translation from the book to Fox’s discussion of it. Just to be clear, I’m reacting to Fox’s brief remarks and not to the book itself.
The decline of unions in the U.S. has often been painted as inevitable, or at least necessary for American businesses to remain internationally competitive. There are definitely industries where this account seems accurate. Globally, though, the link between unionization and competitiveness is actually pretty tenuous. The most heavily unionized countries in the developed world — Denmark, Finland, and Sweden, where more than 65% of the population belongs to unions — also perennially score high on global competiveness rankings. The U.S. does, too. But France, where only 7.9% of workers now belong to unions (yes, France is less unionized than the U.S.), is a perennial competitiveness laggard.
And as for Denmark, Finland, and Sweden, where union density hovers around 70 percent (67.6 percent in Denmark in 2010, 69 percent in Finland in 2011, and 67.7 percent in Sweden as of 2013, the latest numbers available via the OECD for each country), it is important to understand all of them are Ghent system countries. That is, all of them are countries in which the unemployment insurance system is administered by labor unions. In “Paths to Power,” Michael Dimick describes the Ghent system’s distinctive properties, and he argues that ”that collectively-bargained unemployment insurance is efficient and establishes a positive-sum tradeoff between a form of labor-market security for workers and a flexible workplace for employers.” I don’t share Dimick’s enthusiasm for organized labor, but he does an excellent job of explaining the deficiencies of the U.S. approach to labor unions. One of his central points is that labor laws in Denmark and Sweden (he doesn’t specifically address Finland) don’t offer particularly strong protections to unions:
How much better are labor laws in Denmark and Sweden? Both countries have had union densities in the 70 and 80 percents in recent years, much higher than in the US either currently or historically. Given their social-democratic history and politics, one might suppose that these Nordic countries would have untrammeled union-security provisions, effective representation procedures, a strictly-enforced duty to bargain, and high levels of job security, in addition to an elaborate, overarching legal framework for regulating employment relations. In fact, on balance neither Danish nor Swedish labor law is significantly more protective of unions or workers than current labor law in the US. First, in both Denmark and Sweden, union-security agreements are virtually nonexistent. As strange as it sounds, they are essentially “right-to-work” countries.
First, the Ghent system provides an alternative solution to the free-rider problem. Since it is voluntary and administered by unions, it gives workers an incentive to keep and maintain membership without the need for union-security agreements.
By “union-security agreements,” Dimick means the closed shop, in which union membership is a condition of employment.
Second, the Ghent system addresses the recognition problem by separating and reprioritizing employees’ decision to join a union from the employer’s decision to recognize it. Danish and Swedish unions are able to constantly recruit new members through their administration of unemployment insurance, and thence mobilize and build membership support among employees for other labor-movement goals, including recognition from employers, without the need for government certification procedures. These two decisions are confounded and their ordering reversed in US labor law and practice: in order for unions to recruit and build membership support, they must first prevail in a government-administered representation election against a typically intransigent employer. Finally, union participation in unemployment-insurance policy also helps sustain more cooperative labor relations. In Denmark in particular, unions and employers are able to achieve a positive-sum tradeoff whereby workers receive income security in exchange for ceding their demands for job security, which gives employers more flexibility in the workplace. Danish success with the policy—termed ―”flexicurity”—has garnered much attention from European policy makers.
Dimick explains the upshot of these differences for productivity and competitiveness, the subject Fox briefly addresses in his post:
Even defenders of unions usually concede that unions have negative effects on productivity or unemployment, or both. Indeed, I show that when unions and employers bargain over wages and employment- protection rules (such as “just cause”), risk-averse workers will prefer a contract with excessive job security that is production inefficient. However, when unions and employers bargain alternatively over wages and unemployment benefits, leaving to the employer the right to hire and fire, this externality is internalized and the resulting contract causes no loss in productive efficiency compared to the competitive, nonunion benchmark. Understanding these mechanisms can help explain the supportive role that Danish flexicurity plays in solving the adversarial problem. Moreover, as I shall argue, there are good reasons why unions should participate in unemployment-insurance policy in order to make the implementation of flexicurity a success. [Emphasis added]
Dimick concludes by arguing that the U.S. ought to embrace something like the Ghent system, on a state-by-state basis. I haven’t thought very deeply about the idea, but what I can say is that attributing the virtues of Ghent system unions to U.S. unions makes little sense; they are profoundly different, despite the fact that we call them the same thing