Unions as Bottlenecks?
As I noted in my first post, Bottlenecks suggests that equality of opportunity requires substantial economic equality. I read Fishkin’s second principle as doing a lot of work here. A sufficiently robust set of welfare and income supports would make a far more diverse set of life plans realistic for citizens, and would reinforce families’ and individuals’ diverse views of the good.
What policies does this require? In addition to substantial tax-and-transfer redistribution, Fishkin also notes that opportunity pluralism might support public and private efforts to limit salary differentials within particular firms (p203), though he doesn’t develop the idea in detail. In this post, I’d like to pick up that line of argument, and suggest that opportunity pluralism would support quite robust wage-compressing institutions that limit pre-transfer income inequality. I’m thinking here of unions, which have traditionally been a very effective means to greater economic equality.
While unions’ macroeconomic effects are debated, it is certainly true that widespread unionization enables unionized workers—and at least some other workers—to bargain for significantly higher wages. This would seem to advance Fishkin’s second principle insofar unionized workers and their children will encounter a more open opportunity structure.
Unfortunately, can’t unions themselves become arbitrary bottlenecks? For example, craft unions representing skilled trades workers have often sought to exert bargaining power in part by controlling the supply of qualified workers, creating powerful incentives toward nepotism and even corruption. To be clear, industrial unions (like the UAW) and service unions (like SEIU, which represents mainly building service and health care workers) don’t tend to do this, organizing workers and exerting power based on collective action rather than job skills. Yet even industrial and service unions may become bottlenecks if they drive wages quite high in one segment of the market, sparking a “secondary” labor market with lower wages and/or informal work relationships. Alternately, in heavy industrial sectors like automobile production, strong collective bargaining units in final assembly may become a sort of labor aristocracy, with workers at parts suppliers suffering some of the consequences.
In other words, whether one is working in a unionized or non-unionized sector is in large part morally arbitrary. Surely concerns such as these have helped drive recent arguments that unionized public employees are overcompensated. Unions also raise concerns about compulsory association, and at times advance their members’ interests rather than broader matters of public concern. While both concerns are often overstated, is it any wonder that liberal egalitarians tend to endorse transfer programs rather than unionization as means to achieve distributive justice?
In future work, I’m planning to dig deeply into the idea of and place of unions in liberal distributive justice. For now I’ll just make two observations. First, as I noted in my first post, I strongly suspect that stand-alone redistributive programs—i.e., tax-and-transfer programs imposed upon laissez-faire labor markets—aren’t stable in the Rawlsian sense. Efforts to ensure distributive justice solely through post-earnings tax-and-transfer policies may generate feelings of resentment among those forced to pay higher taxes, and feelings of helplessness among those who rely on public assistance for basic subsistence. Unions can make those systems more stable, both by equalizing pre-transfer incomes, and by ensuring that workers and the poor have a substantial degree of political and economic power. Indeed, in many if not most industrialized nations, unions have often been a prime mover behind welfare state redistributive policies. They create and help sustain egalitarian politics. (For a nice summary of this, see Matt Dimick’s argument in Salon a few years back).
Second, if the problem is a major union/non-union wage differential, the solution may well be more rather than less unionization. The sorts of ugly, drawn-out unionization battles so common in the U.S. are an artifact of a labor law that requires workers to affirmatively opt-in to union representation, generally on a plant-by-plant level. More widespread unionization and different sorts of bargaining regimes could reduce the union/nonunion wage differential. Moreover, such policies would not necessarily lead unions to drive wages to unsustainably high levels if coupled with the sorts of corporatist bargaining that characterize certain Continental European systems. Indeed, US labor law seems to encourage labor/management conflict, while national systems based on national- or sectoral-level bargaining seem to encourage labor-management cooperation.
Legal protections for workers’ collective activity may therefore be a central matter of distributive justice, as well as a practical means of ensuring true equality of opportunity. This strikes me as a more convincing normative argument than the notion that labor law should primarily protect employee choice or basic liberties such as freedom of speech and association.
But achieving this set of goals would require significantly farther-reaching changes to labor law than contemplated in current debates. My gut is that a liberal egalitarian approach to labor law would counsel for strong policies encouraging near-universal representation and national- or sectoral-bargaining, together with substantial democratic reforms to unions’ internal governance to encourage far more fluid forms of union membership, including plural unionism. For what it is worth, I see no reason at all that unions operating on a national rather than a local level could not provide an effective voice and representative for contemporary service workers who are virtually impossible to organize on a firm-by-firm basis.
Whether it is possible to move from the Wagner model to that sort of model is another question. But given the weakness of unions today, this may be an opportune time to begin a conversation about fundamentally different forms of worker representation. Bottlenecks provides an important starting point: if equal opportunity requires substantial distributive justice, then unions or similar institutions may be centrally important to liberal egalitarian social policy.