Reader Questions: On Price-Taking and Competition

We will switch things up today and answer some questions that I received from an astute reader, Ashwin Varma. Ashwin is a medical student, and I’m pumped to have non-economists, not just reading Economic Forces but also seriously engaging with the ideas. Thanks, Ashwin!

Ashwin’s email is indented (and slightly edited).

As I understand my college IO textbooks, the fundamental characteristic of 'competitive' markets is that they consist of price-taking firms. This doesn't necessarily mean there's no profit for the firms; so long as the firm isn't the highest-cost producer it earns a 'rent' (ie, supply curves slope up). However, when I look at the most profitable companies around me, notably any company on the Fortune 500 for example, I would be hard-pressed to suggest that they are price-takers… However, my hypothesis is that the absence of price-taking behavior does not imply that competition is absent in these markets.

Price-taking is an easy, simplifying assumption. It’s great to use when teaching how to solve a model and find a competitive equilibrium, whether in partial equilibrium or general equilibrium. That is why we often conflate price-taking with competition

We should not confuse a simplifying assumption for a necessary condition or the essence of competition. I’m guilty myself, especially when teaching. In Chicago Price Theory, the authors also say that one condition for competition is that buyers and sellers take prices as given. (We will touch on the second condition later.)

I don’t ever like to disagree with Kevin Murphy….

To go back to a simple example that I’ve discussed before, the standard Bertrand model. In that model, we assume the firms can pick their own prices. The firms are not price-takers. Yet, the equilibrium is competitive in the sense that the outcome is the same as it would be if the firms were price-takers. Sometimes it is a useful shortcut to just assume the firms are price-takers since it brings you to the same outcome.

The problem is that the interpretation is different between price-taking and Bertrand. While they are not “price” theorists, Louis Makowski and Joseph Ostroy are the right people to read on this. In the price-taking framework, prices must exist before people can maximize utility/profits; prices proceed maximization. Instead, Makowski, Ostroy, and I argue that we should think of (and model) prices as arising from bargaining individuals. In that world, maximization precedes emergent prices.

In some cases, the prices are the same as in the price-taking world. In other cases, they are not. For example, a Bertrand model with 1 seller is just a monopoly model. In that case, the equilibrium we focus on has a monopoly price and is inefficient. However, there is a price-taking equilibrium for that economy which is efficient. If the monopolist was forced to be a price-taker by assumption, she would choose quantity such that price equals marginal cost, just like a competitive seller. We ignore that equilibrium because it seems weird.

In both cases, the reasonable approach that gets the “right” equilibrium is not to assume price-taking ex-ante but instead to assume no one is a price-taker.

Instead of price-taking, the essence of competition is the substitutability of the consumers/producers. Closer substitutes for a firm’s product flattens its residual demand curve, but so does more intense competition from buyers, say if we double the number of buyers. Closer substitutes on either side of the market increases competition.

In the Bertrand model, each firm is a perfect substitute for the other. When you have to compete against a perfect substitute, you have no market power to extract surplus above your marginal product. I’ll save the meaning of a person’s marginal product for another newsletter. For now, read another excellent piece by Makowski and Ostroy.

Through the lens of substitutability, it becomes clear that competition is not on or off, as it seems when we compare a standard model of supply and demand to the monopoly model.

The monopolist faces competition too! The demand curve they face is not infinitely high. If the monopolist raises prices, some sellers leave the monopolist and go to the competition. This interpretation further highlights how the competition that the monopolist faces doesn’t come directly from similar sellers of the same homogenous good, since they are a monopolist and there are no sellers of the same good. Instead, competition comes from any other use for the consumer’s dollar. Competition is pervasive. Perfect competition with a perfect substitute is an idealized situation that never obtains.

Ashwin continues:

I reference this paper (a relatively well known one in the empirical IO literature) by Amil Petrin (BA: my econometrics professor at Minnesota) on the minivan. Chrysler introduced the minivan, which was very popular, and was able to maintain differentiation from others attempting to copy it. Hence it earned considerable profits and maintained a considerable market power (price ~15% over marginal cost; no price-taking here) for about 15 years! But despite the market power, the market for minivans seemed incredibly competitive by other metrics. Many firms entered the market in the first 5 years after its initiation, and all firms in the market - including Chrysler - were forced to actively invest to improve their products over time, improving consumer welfare. It seems to me that firms were responding to some form of competitive pressure even if there was a lot of market power exercised by firms participating in it and hence lots of profit earned.

We’ve already seen that in a monopoly model, the price can be above marginal cost and the seller can still face competition, still need to innovate, etc. The question is whether there is competition but how intense that competition is and how closely other firms can become substitutes for Chrysler.

Given that price is consistently above marginal cost in this case, most economists would conclude this market does not approximate perfect competition. We all “know” that in perfect competition, price equals marginal cost.

Let’s go back to the second condition given in Chicago Price Theory for competition. They say that with competition the marginal entrant earns zero profit. Notice they don’t say that price equals marginal cost. I think this distinction is on purpose and the correct way to think of things.

The distinction between profit and price minus marginal cost matters whenever firms have fixed costs. I’m not sure if auto manufacturing has fixed costs but let’s imagine it does. For simplicity, let’s assume marginal costs are constant so we don’t need to confuse ourselves with the confusing distinction between rents and profits. Then clearly, firms will need to cover their fixed costs, and the price will be above marginal cost. This will be true even if there is a perfect substitute waiting in the wings for the firm to raise prices by a penny. Baumol and Bradford show markups are efficient relative to marginal cost pricing.

Another way to see that confusion about marginal cost pricing is to remember that in the long run all costs are marginal costs. We don’t want to wrongly conclude P>MC simply because we have the wrong notion of MC.

I’m not saying the minivan market was perfectly competitive. All I’m saying is that we need to be careful about these things. We need more info, such as measures of fixed costs or demand elasticities, which the referenced paper and much of modern IO spend lots of time fussing over.

If you have questions/topics you’d like either of us to address, feel free to comment on any post or send us an email.