Increases in wage inequality and the fall of unions: chicken and egg?

A few posts ago, I promised I would say more on the issue of unions and wage inequality.

Many economists suggest that the declining fortunes of unions are the main driver behind the increase in income inequality over the past few decades. While I agree that decreases in union membership and increases in wage inequality go hand-in-hand, I believe the standard narrative is somewhat misguided when it comes to causality. Rather than declining unions leading to increased wage inequality, it is the increasing variance of individual worker productivity that lies behind both.

In a pattern familiar to union watchers, the union will usually stand for wage equality rather than 'meritocratic' pay. It is easy to see why this is so: unions are democratic institutions. As is the case with elected governments, they will redistribute from the productive (potentially high-wage) to the unproductive (potentially low-wage) workers in order to please the median voting member. In the presence of a dominant union, all workers are awarded the wage that corresponds to the average productivity of all members.

In contrast to the state, however, membership of a union is not compulsory (at least in most cases). Why do high-productivity individuals choose to join the union then? To understand this, it is important to grasp a subtle point: even if a worker has higher productive potential than average and would thus be paid more in the absence of collective wage bargaining, what matters to the employer is marginal productivity - the product this additional worker will generate given the existence of all other workers.

Marginal productivity tends to fall as the number of workers increases. As a result, a large, unionised workforce will in many cases mean that the marginal productivity of the high-ability worker may actually be below average productivity. When that is the case, the high ability worker has an incentive to join the union, as the average wage represents the best attainable outcome.

When the differences in individual workers' productivity are relatively small, joining the union is the most attractive option for low and high ability individuals alike. However, as the nature of jobs changes and the productivity potential of some workers increases far above that of the others, opting for the average productivity wage offered by the union is no longer the most appealing proposition for some very productive workers - their marginal productivity is above average, and they can command a higher wage by choosing to go it alone.

As more and more workers break away from the pact in terms of individual productive potential, the average productivity of the union's membership falls; and the existence of a large pool of high-ability non-unionised workers restrict its ability both to redistribute from high to low ability workers as well as halt its own demise. At the equilibrium, the only sectors where unions are dominant are those where workers are of relatively uniform ability: for example, industries employing the low-skilled.

The fall in the power of unions was not a random event that then led to the increase in wage inequality we observe. This development should be attributed to the changing nature of production and the subsequent increase in potential wage inequality: unions were merely a short-run obstacle to achieving that new equilibrium.