Colleagues,

A while ago (around our disagreements about May Day) I had an exchange with Fernando about labor unions. I’ve been thinking about this again recently, partly as a result of my ongoing work on intermediate groups, partly due to some union issues in Canada, and partly in light of Jessica’s exchange with Chris Bertram about employment relations.

Here are some views which seem to me to hang together, but I don’t lknow whether their joint coherence would survive serious scrutiny. Neither do I know how idiosyncratic I am in holding all of them.

1) Labor unions often harm the well-being of the worst-off by making entry into low-wage work more difficult, raising the bottom rung of the wage ladder above the marginal productivity of outsiders.

2) This provides the unions institutionally with a kind of control over access to labor market rents, which contributes to hierarchical power relations within unions that can be highly problematic.

3) Unions have often contributed to making their firms or their industries uncompetitive; they lack a built-in check that would keep them from killing the goose.  This may be an even worse problem when they are faced with temporarily oligopolistic or monopolistic firms– the big traditional airlines, the Big Three automakers, traditional AT&T— which are themselves prone to become vulnerable overcommitted dinosaurs over time, and which are happy enough to lock in medium-term labor peace with contracts that put the firm one transformative competitor away from disaster.

4) Public sector unions have, under some conditions, a considerable ability to capture both sides of the bargaining table and seriously weaken public sector budget restraint.

5) Unions bureaucratize work relations in ways that can be both unpleasant and counterproductive for all concerned; they create rigidities and inflexibilities that can prevent cooperative improvements.

6) Labor unions form the organizational core of political parties and movements that seek to do more than bargain in the workplace; they’ve been a major force for social-democratic political economic organization.  And the politics they shape is not especially prone to be liberal along cultural, social, or constitutional dimensions.

7) Unions– their leadership, their membership, their supporters– are capable of significant bullying that easily spills into immoral coercion against outsiders, nonmembers, and dissenters; the charge of “scab” is often an accusation of punishable treason against a cause that the accused never signed up for.

And yet:

8. Unions have been and continue to be absolutely indispensable for the mitigation of workplace power imbalances between managers and employees; they provide due process protections, protection against favoritism and nepotism and retaliation and harassment sexual and otherwise, protection against unjust dismissal and against the countless ways that managers can use the threat of dismissal to gain personal advantage.  Now, there’s a sense in which many or most of these abuses are contrary to a firm’s interest.  They amount to managers acting on their own behalf instead of on the firm’s, and so represent a type of corruption.  (Sexual harassment is the clearest case here; the manager demanding sexual services is seeking something that is only for his own benefit, not for the firm’s.  But the same is true when someone is promoted or fired because of a personal relationship– favoritism or retaliation.)  But the principal-agent problems involved in firms trying to manage their own managers to prevent all of these things are serious; it’s easy enough for higher-level managers to think that a mid-level manager is more indispensable than is a low-level worker.  It’s certainly possible to generate these protections without unions; many professionals have them.  But for many non-professional employees, it has only ever been unions that fought for these protections; it has only ever been unions that provided them; and they do not survive well when unions die out.

Note that I did not say, in the above, “these protections are demanded by libertarianism.”   But I also did not say “these protections are prohibited by libertarianism.”  I think the general commitment to human freedom that motivates libertarianism should also motivate concern for these kinds of issues.  (See my previous posts here and here on this kind of claim.)  But I’m not arguing for that here; just collecting my various views.  The empirical claims listed above are also neither demanded nor prohibited by libertarian ideas.

9) Collective wage bargaining on the offer curve within the Edgeworth box between employees and employers is not only legitimate but (at least often) a positive good.  That is, there is bargaining room to maneuver in any given employment setting; there are lower and higher wages that are both compatible with a given efficient level of firm output.  And one doesn’t have to be very committed to anything like diminishing marginal utility or the difference principle to say: if there are rents to be had or surpluses to be gained thanks to the strictly strategic aspects of bargaining– because “equilibrium wage” actually describes a range– then workers ought to have a reasonable chance of winning the negotiation and gaining that surplus.

It might be difficult– might be impossible– to get institutions that would enable collective bargaining within that range and still check against bargaining that pushes outside the range and hurts competitiveness.  I have no idea.  But that range does exist, in many negotiations; many labor markets are (at least locally in time and space) characterized by a small number of firms bargaining with a large number of workers, and that gives the firms strategic power that is unrelated to efficiency; unions can mitigate or counteract that.

10) The fact that workers have so often and in so many places sought to organize, and the fact that firms have so often and in so many places resorted to illiberal restrictions on freedom of association if not outright violence to prevent them from doing so, itself looks like prima facie evidence from the world in unions’ favor.  Whatever one’s complaints against the regime of employment relations created by positive legislation such as the Wagner Act, unionization comes first, before the state action and initially in spite of state action.  When I first started researching freedom of association, I was annoyed that so much of the literature was dominated by the case of unions, since that wasn’t a case I was interested in.  I now think that was the wrong reaction, and that I should take the centrality of unions to fights for freedom of association as indicative.  There is widespread revealed-preference demand for associations of something like this form, and as a libertarian I should pay attention to that, not sweep it under the carpet.

There’s nothing approaching a theory here.  I look at some industries, some sectors, some places and say “there’s too much unionization there; the bad effects are getting crippling.”  But I no longer have the basic hostility to them that many people (including many libertarians?) who hold to positions 1-7 above do, and I look at other industries, sectors, and places and say “there’s too little unionization here,” especially when workers evidently suffer from managerial workplace domination and abuse and lack procedural protections against them.  In the world as it exists (please, no arguments about how free currency competition would get us to macroeconomic full employment and so eliminate bargaining imbalances; maybe you’re right, but I don’t really care) the reality of workplace domination means that some cost in the bureaucratization of the workplace is worth paying.  If I’m right about (8), then as far as I’m concerned that’s a strong reason to support unions and unionization much of the time, even if considerations 1-7 suggest needs to balance against that.

As I said: not yet a theory, just a collection of views. I’m curious how they strike you.  Are they jointly incoherent in a way that I don’t see?  If they’re right, do you join me in thinking that BHL has to adopt a different stance toward unionization from more traditional libertarian views?

Update: one more thought about points 7 and 8.  In my view one of the hardest problems within liberalism– which includes libertarianism– is how to balance concerns about in-group domination and abuse against worries that central states will suppress groups far more than is justified, that they will treat concerns about local domination opportunistically as excuses to constrain or crush counterbalancing associations and organizations, and that even if the constraints they place on groups in the short term do enhance freedom, it will be at the cost of weakening future organizational sources of resistance to state power.  This dilemma recurs all the time: in debates about federalism and how much central state intervention is warranted to protect the liberty of local minorities; in debates about freedom of religion and cultural freedom when women are disadvantaged by group practices; in (a special area of interest for me) disputes about the authority of Indian reservation governments to govern non-tribe members; in disputes about universities’ internal restrictions on speech or association, or their internal procedures for disciplinary cases.  There’s a temptation to look at the person in front of you whose freedom is being violated and respond with a massive restriction on the authority of the group doing the violating.  And sometimes that’s warranted.  But, even when it is,  you need to pay attention to later dynamic effects of the change, and whether they leave people even more vulnerable to future violations of freedom.

Note that this can apply both to the problem of unions’ treatment of dissenters, and to the problem of firms’ treatment of workers. But in my view it’s a more important corrective to worries about 7 than to worries about 8– because collective bargaining and union protections restricting the internal actions of a firm are not the same as direct state regulation of the internal actions of a firm.  I worry about 7 a lot; the abuse of the never-signed-up-for-this supposed scab offends me, and I think should be called out for what it is.  But I also worry a lot about using it as an excuse for state restriction of union activity, for the same reasons I worry about such excuses in the kinds of cases listed in the previous paragraph.