Eli Dourado

Obligatory unions post

The kerfuffle over public-sector unions in Wisconsin has captured the imagination of the chattering class to the point that unless I wish for my blog to lapse further into irrelevance I must break my silence on this topic. I am not a labor economist, and I have never had the privilege of visiting Wisconsin, which strikes me as an exotic land of people who walk around with cheese on their heads. So I am clearly an expert.

The labor economics literature on unions is as vast as it is boring. Instead of troubling with it, let’s try to think systematically about market power more generally, and then we can apply our findings to unions specifically.

A simple model of market power has P > MC, a deadweight loss triangle, and a monopoly rent rectangle. Most economists freak out about deadweight loss triangles and consider the monopoly rent rectangle a transfer. Transfers don’t matter for efficiency, but may matter for reasons of equity.

The simple model is wrong because it’s static. In a dynamic model, the monopoly rent rectangle is a prize for winning at some competition. People expend resources to capture supranormal profits.

The question, therefore, is not “how can we eliminate deadweight loss triangles?” or “how can we equitably divide surplus between management and labor?” but “what sorts of competition do we want to reward?” The monopoly rent could be a prize for shrewd risk-taking, for jumping through hoops, for being politically connected, etc.

In general, I’d say that to the extent that people are risk-averse, set in their ways, lazy, and ignorant, we want to reward things like shrewd risk-taking, innovation, productivity, and skill acquisition. We don’t want to reward things like political clout, corruption, or credentials. We want to dismantle monopolies that are based on political privilege (e.g., your local cable company), and tolerate market power that is based on investment of the kind that we want to encourage.

Back to unions. Unions have market power, but we can’t say whether a particular union is good or bad until we ask what its supranormal profits are rewarding. On what margins do people compete to get into the club? I don’t know enough to say for sure, but it seems to me that public-sector unions reward meaningless credentials, low ambition, luck, and getting along with other public-sector workers. On the other hand, if employees in other industries are bearing the costs of investing in firm-specific skills, the monopoly rent that comes with unionization can incentivize efficient investment (though of course, the employer can also bear the cost of investment, or the employees can buy the firm).

Most people who favor unions don’t do so for the reasons I’ve outlined. They favor them for expressive reasons: yay, labor! boo, management! But in my model, there is no reason why generic, unskilled labor (or generic, unskilled management) should ever earn more than its marginal product. We want to reward the human capital and risk-taking components of labor. We can debate whether laissez-faire and freedom of association get us close enough to optimal rewards, but it seems pretty clear that busting public-sector unions in Wisconsin is unlikely to hurt anything.