A common takeaway from an intro to economics course is that a market with many buyers and sellers is good, and a market with a few sellers is bad. One seller is the worst. That damn monopolist!

That’s an important insight, and we shouldn’t neglect it. It helps explain why free entry is often crucial for well-functioning markets. It shows the efficiency-enhancing role that antitrust can play in preventing anti-competitive behavior. Unfortunately, the too-simple punch line of “more sellers are better” doesn’t always hold.

The simplest example is a firm with market power, which we can call a monopolist for simplicity, that produces a good that generates a negative externality. We generally think that the problem with a monopolist is that the firm produces “too little.” However, the problem in an industry with a negative externality is that the firms produce too much. With externalities, monopolists are good. When thinking about efficiency, you want less of the good to be produced—a monopolist, who also wants less of the good produced, may help you get there.

There is a sweet spot where the level of the externality is the right size so that a monopolist produces the efficient amount of a good. One of my favorite exam questions is for the students to solve for this.

Just because there is a possibility that the monopolist is efficient does not mean that fewer sellers are better with a positive externality. If the externality is small, the efficiency gain from increasing the number of sellers will outweigh the loss from the externality. A key result in basic economics is that the optimal amount of an externality is not zero.

What about pecuniary externalities?

If there is a negative externality, then more sellers may generate more externality and therefore be bad overall. This intuition extends to externalities that happen through prices (a pecuniary externality).

The trick with pecuniary externalities is that they have both winners and losers. If a new firm lowers the price of a good in a market, they are generating a negative externality on other sellers but a positive externality on consumers. If the market is perfectly competitive, the positive and negative externalities cancel out. This is why some people think pecuniary externalities don’t matter for efficiency and should be thought of as fundamentally different from “real” externalities.

Outside of perfect competition, pecuniary externalities do not cancel each other. Some pecuniary externalities improve efficiency, and some hurt efficiency. Often, we think of entry by a new firm as being a net positive pecuniary externality. The gain to consumers is greater than the loss to the other sellers.

The result can easily go the other way. In a recent post, I explained why complementary monopolists are worse for consumers than a true, merged monopolist. We can see the inefficiency by looking at the pecuniary externality imposed by each monopolist on the other monopolist. In the example, Josh and I were selling rides on connecting stretches of a railroad track.

When I raise my price, that pushes down people’s willingness to pay for Josh’s stretch. Therefore, since I’m not thinking about my externality on Josh, I raise my price more than would be jointly optimal. Josh does the same thing, and the price goes higher than the monopoly price. If the price is higher, there are fewer trips between Kennesaw and Oxford.

I raise my price too much and inflict both a negative externality on Josh and a negative externality on the consumers.

We can use the externality framework to see the difference between the complementary monopolies model and the standard model with competing firms that are substitutes, such as in a standard differentiated Bertrand model. With substitutes, if Seller A lowers his price, that pushes down Seller B’s residual demand curve, which hurts Seller B. Therefore, Seller A lowers his price more than would be jointly optimal. That negative externality on Seller B is a positive externality for consumers.

A more general lesson is that we are often in the world of the second-best. If there is one distortion in the market (like an externality), then “correcting” another distortion (correcting monopoly status) can backfire. Our intuition doesn’t extend so nicely.

Flipping the intuition shows up again and again in modern papers. If there is adverse selection, more sellers can hurt efficiency, as in Lester, Shourideh, Venkateswaran, and Zetlin-Jones (2019). With network effects (externalities), more competition may hurt consumers, as in Tan and Zhou (2020). The list goes on. Economists like counter-intuitive results; I know I do.

The theory of second-best may mean that we just never know what’s going to happen.

I don’t think that’s true. We just need more information about the case.

When studying a specific situation, we need to understand what is the cause of a single seller or what is driving the change in the number of firms. Is there one firm because of fixed costs? Are there scale economies or network effects? Then it may be optimal for only one firm to operate. A second firm can’t compete profitably. We certainly shouldn’t use policy to prop up a second competitor (looking at you USPS!). Let potential free entry do the work. The flip side is that policy should try to avoid adding more fixed costs through regulation and licensing.

If all other markets are well-functioning and complete so that there are no externalities of any kind, then we can generally think that more sellers will improve efficiency. But where do those firms come from? If you magically make another company that can do what the first company does, that’d be great. We should do that.

That’s my policy proposal. We should encourage more Googles, Apples, Amazons, and Steph Curries to fall from the sky.