File this under attempts to clarify my own thoughts. Unions are a tricky topic in economics, particularly for someone of a liberal persuasion like myself. The essence of the issue seems to be that there are two basic freedoms involved, both of which are vital for the operation of a modern economy yet appear to be at odds with each other. The first one is the freedom to collaborate, to join in voluntary cooperation with other market participants if you deem this to be in your interest. The second is the freedom to compete, the need to create a level playing field where everyone has the same opportunities without being hindered by arbitrary market power. Should one be more important than the other? And if not, is there a way to find a middle ground that creates an acceptable balance?
A long time ago economists and politicians correctly decided to break up monopolies on a firm level, effectively putting a limit to the degree to which the freedom to cooperate can be put into practice. The question then comes up: why should we actively proceed against monopoly-like arrangements on the employer-side of the equation while in part encouraging similar arrangements on the employee-side? Different bargaining power seems to be an obvious first answer. There can be no doubt that an individual has, in general, considerably more limited outside-options compared to a company as a whole, particularly in times of economic distress and high unemployment. Of course, in an ideally working economy operating at full employment this issue doesn’t really come up.
Key to a working market for labor thus must be to try and ensure equal bargaining power on both sides of the table, and there is an argument to be made for legally limited the ability of employers to undermine any such arrangement. Game theoretical considerations might require some sort of mandatory unionization, yet it is vital that the power of unions shall not exceed the power of employers. Balance is key. I would argue that the free-rider problem seems to be somewhat exaggerated, and probably only makes a big difference for economy-wide, possibly industry-wide unions, which might be undesirable anyways, but the general train of thought seems to be correct. Mandatory unionization at the firm-level might possibly prove an option, even though I’m uncertain this is needed – laws that ensure unions at this level are not prevented from being formed should prove enough – “right to work” laws appear to be as cynical a name as there can be.
In essence, unions are a method of addressing market failure in a world with unemployment. And while full employment is certainly awesome, it is not only hard to define but seems to be even harder to reach. Unions in a neoclassical framework are at best utterly useless, at worst outright destructive to the economy as a whole. Yet in this world, they may prove a necessity to achieve a second-best outcome if the first-best seems beyond our grasp. Any form of black and white approach to the issue, however, grossly oversimplifies the whole matter and most likely also makes for terrible policy decisions.