Unions and redistribution

Last week, Demos released a study that proposed raising wages for workers at big box retailers. According to the study, establishing a $25,000 wage floor at big box retailers would cost big box retailers $20.8 billion per year. If that additional cost were made up through price increases alone, it would amount to a 1% price increase, or about $35.46 per year for the average household.

Critics of establishing this wage floor — in addition to the usual minimum wage criticisms — point out that Walmart primarily caters to low income customers, and so a price increase would be a price increase on the poor. I find these points generally annoying because of the kinds of people who make them, but they do raise legitimate questions worth considering about the role of unions in redistributing income.

Imagine for example that all of the big box retailers in the country became unionized, and the unions successfully negotiated a contract for a $25,000 wage floor for workers at the stores. According to this study, such a contract would impose $20.8 billion of new costs on these retailers. This money must come from somewhere. Some set of people who are currently receiving a combined $20.8 billion will not any longer. The obvious question then is: who?

Answering that question can get really complicated, but there is a simplistic treatment of it which should suffice here. Roughly speaking, that $20.8 billion can come from three sources: 1) other employees of the company (management) through lower incomes, 2) shareholders of the company through lower profits, or 3) consumers through increased prices. That is to say, one or all of those groups of people will have less money so that the workers at these big box retailers can have more.

Ideally when you redistribute, you want money that was going to go to wealthier people to go to poorer people instead. If the big box retailers choose option one, then that will be achieved. Cutting executive pay and management pay to make up the difference (ignoring that they may quit the company if that happens) would shift money from the relatively higher-paid to the relatively lower-paid.

If the big box retailers choose option two, the same is true, but with one caveat: all sorts of people probably hold shares of these big box retailers. Even people with very modest incomes have retirement accounts or are covered by a pension fund, and no doubt some of that money is tied up in big box retailers. If they absorb the costs by reducing profits, then those stocks will take a hit. While it is probably true that the shares are disproportionately held by wealthier individuals, it is at least somewhat of a mixed bag. Overall I am sure it would be redistributive, but it’s kind of messy.

Finally, if big box retailers choose option three, then we start to have a real problem. Some consumers of these stores will have more money than the workers while others will doubtlessly have less. Redistributing through price increases redistributes from all classes, with a likely skew towards the lower incomes. You end up then, in part at least, taking money from poor people in order to give it to other poor people. That is certainly not ideal.

In practice, it is hard to know what the big box retailers (or any other company for that matter) will do. It really just depends on what decisions their executives make. This lack of control over where the money comes from drives some of the left-neoliberal (and even socialist) criticism of unions as agents of redistribution. When a union wins a wage increase, the person who decides how to reallocate the distribution to account for that wage increase is a corporate executive, and that decision can have very sub-optimal distributive consequences.

If the goal is to redistribute from the rich to the poor, there is a legitimate point to be made that this is not exactly the most efficient way of going about it. It might redistribute from higher income to lower income, and then again it might not. In the big box retailer case, the workers make so little that it is more likely than not that the aggregate effect would be redistributive, but it becomes much more fuzzy when we start talking about wage increases at middle and upper-middle income workplaces. Wage increases in those workplaces could very possibly have an ultimately regressive distributive effect if the costs are passed on through prices and the costumer base for the firm is primarily low income.

It is important to note that unions are not merely redistributive agents. They do other things as well, some of which are actually more important, I think (e.g. protecting workers from arbitrary firing and favoritism). But redistribution is certainly one of their main appeals, and to be fair, that is a bit of a tricky affair. All unions can really do is make sure their members get more; they have no control over who ends up getting less. The executives ultimately make that call.