Price-match Guarantees and Monopoly

The anti-competitive implications of competitive practices

Why do firms engage in price-match guarantees? What might seem to be a benefit to the consumer need not be – the adoption of price-matching in fact leads to the monopoly price being charged, no matter how many firms are in the market, or how search costs are modeled. This contrasts with standard models where firms are restricted to setting one price, which generally argues that deviations from the competitive price are an increasing function of search costs.

To quickly review search cost models, they must find a way between the Scylla and Charybdis of “all firms offer the monopoly price”, and “all firms offer the competitive price. If consumers search sequentially and pay some positive cost to search again, then (as in Diamond 1971) the price converges to the monopoly price. (The intuition here is that everyone expects everyone else to offer the monopoly price, and so no one pays the cost to search). If consumers see two or greater firms, then we have the standard result that two firms competing on price is sufficient to reduce prices to marginal cost, which is the competitive outcome. Burdett and Judd (1983) get around this by showing that if consumers have the choice between searching for one or two firms at a cost, the optimal response of firms is to randomize their prices. Price dispersion falls (and converges to the competitive price) as search costs decrease. I covered this all in greater detail on my blog before, and you can read that if you want to learn more.

There are two firms considering providing a service. Each can adopt a price, and choose whether to match lower prices or not. We suppose that consumers care only about the price which they will actually pay for the good, and not whether they were offered a higher or lower price initially. Suppose that each firm chooses to enter sequentially, and chooses a price, and whether or not to match lower prices. There is no cost to enter. Each firm is strictly better off if they choose to match lower prices, and each firm will start by charging the individually profit-maximizing price. Nobody ever offers the lower price, so it never gets matched lower.

It is easy to see that it doesn’t matter how many firms the consumer queries. The consumer will get the same offer from either of them. Nor does it matter how many firms there are in the market – each firm which enters will choose the same pricing structure.

This is not, unfortunately, a totally novel idea. At least as far back as Salop (1986), economists have remarked upon price matching as an arrangement that facilitates collusion. It is useful to think of what prevents coordination to begin with, even between two firms with nobody else able to enter. Firms are in an iterated prisoner’s dilemma, which means that they can temporarily profit if they defect and cut their prices. If there is a long enough lag between the cut and the other firm’s retaliation, then defecting is optimal. A price-match guarantee is simply cutting the lag down to zero.

One of the more interesting extensions can be found in Edlin and Emch (1998). Price-matching will lead to temporary profits, but eventually firms will enter until average costs equal the supracompetitive price. Thus, nobody gains at all from the higher prices, and the new equilibrium is even worse than if there were simply a monopoly granted to one firm. The new equilibrium has all the duplication of capacity of monopolistic competition, and all the distortion of consumption from monopoly pricing. The key statistic is whether fixed costs are large. If they are, then the pure monopoly is indeed worst, because few firms would enter and there isn’t that much waste from duplication. As fixed costs get smaller, more and more firms enter

So why aren’t all firms, all the time, at the monopoly price? Thankfully for us, the result is sensitive to a bit of friction. If consumers have to pay a small amount in order to redeem the discount, then price competition is restored. Consumers are no longer indifferent, in other words, between a high and low initial price. Adding some asymmetry between firms leads to price matching leading to some, if smaller, rises in price.

However, this suggests that when hassle costs fall, price-matching will once again lead to more distortionary outcomes. Online shopping is far more common now, and it is trivial for firms to switch their prices. Consider Amazon, which does not explicitly match prices. What it does do is automatically change their prices in response to its competitors prices. By punishing defectors, it prevents anyone from lowering their prices to begin with.

I have not found any adequate work covering the actual effects of these algorithms. However, there is theoretical work showing that multiple firms experimenting will converge upon supracompetitive prices if the informational value of a given experiment is high, like Hansen, Misra, and Pai (2021).

I would like economists to consider the implications of firms choosing menus of prices, rather than a single price at which they sell all units at. I covered the Bornstein and Peter paper “Nonlinear Pricing and Misallocation” on the blog recently; a simple change to a model, which is absolutely ubiquitous in the real world, radically changes the policy implications. This is, however, not something that I believe I am equipped to add to, at the moment.