Exclusive Dealing and Foreclosure
Someone asked me about the past few years’ papers on exclusive dealing, so I did some thinking. I’ll lay it out in my own way.
Suppose we have 100 buyers, each with 1% of the market, one upstream incumbent seller who charges the monopoly price. Everybody knows that in a year a potential entrant seller will arise, and that he will need 15% of the market to achieve the necessary scale economies.
Ramseyer-Rasmusen-Wiley (AER), with elaboration by Sigel and Whinston, pointed out that the incumbent would be willing to offer the buyers 1 dollar each to sign an exclusive-dealing contract with him. Each buyer knows that the incumbent will then charge the monopoly price, and the entrant would charge marginal cost, but they also know that if the other buyers all sign up, the entrant won’t bother to enter. Thus, one equilibrium is for them all to sign up. (If a buyer expects other buyers to refuse to sign, he won’t sign, though. And we can complicate the story with sequential offers, high offers to the first 86 buyers, and so forth.)
But what if the buyers are not final consumers, but retailers in competition with each other? Fumagalli and Motta (AER 2006) note that then the entrant can survive serving just one retailer, because that retailer can sell to the entire consumer market, as much as the other 99 were selling and plenty to get the necessary economies of scale. Thus, exclusive contracts won’t work to create a monopoly.
Simpson and Wickelgren’s comment of June 2005 and their own paper (forthcoming AER?) also analyzes retailers, but concludes that exclusion *will* work. Their main argument is that since retailers compete, they will get little advantage from not signing the exclusionary contract. Suppose 99 retailers sign, but one does not. Will that 1 seize the entire market in conjunction with the low-priced entrant seller? No. The incumbent will be forced to choose a low price for his captive retailers, since otherwise they won’t be able to compete against the holdout. Thus, none of the retailers, including the holdout, would make any profit. Foreseeing this, no retailer would want to give up the one-dollar signing fee and be a holdout.
This argument is not as powerful as it sounds, though. It is valid in the model as I described it, but let’s think about the real world. If the incumbent has 99 captive retailers, would he really lower the price to marginal cost to match the entrant’s price to the holdout retailer? No. He’d do better with a higher price. He’d sacrifice volume, but the 99 retailers would not, in fact, lose all their customers if they charged more than the holdout. Because of product differentiation, differing locations, switching costs, or poor consumer information about prices, the 99 retailers would keep some of their customers. In that case, exclusion won’t work.
It’s a twisty argument, because it is the profitability of the captive retailers that is the undoing of the incumbent’s scheme. Because they are so profitable, being a holdout will be profitable for a retailer, so there will be a holdout.
I think there’d actually be a mixed-strategy equilibrium, though, and the exclusionary contracts *would* be profitable in the end. That is because it would not be an equilibrium for nobody to sign the exclusionary contract either. If there are 2 or more holdouts, then they will compete retailer profits to zero, in which case the captives, who at least get the signing fee, will have higher payoffs. What will happen is that the incumbent will choose the signing fee that maximizes his profit (which will be bigger then epsilon), retailers will be indifferent about signing, and exclusion will often but not always work. Sometimes all 100 retailers will sign, sometimes there will be 1 holdout (which will be profitable for that retailer), and sometimes there will be more than one holdout (which will yield zero payoffs to the holdouts).
Simpson and Wickelgren have a separate argument, though: fancier contracts will restore exclusion. Suppose the incumbent offers price- matching as part of his exclusionary contract: if the entrant enters with a lower price, he will match it. That will deter entry. Or suppose he uses a two-part tariff where he offers to sell a limited amount at marginal cost in return for a sign-up fee, with the proviso that if entry occurs he will sell unlimited amounts at that price.
That is a good point. Below I’ve excerpted Simpson and Wickelgren’s comment and some testimony of Joe Farrell’s. Farrell wonders whether courts, or anybody, can make much use of these ideas at the moment, since the conclusions are so tangled. I think they can, without much difficulty. Here’s what I conclude:
1. Exclusive-dealing contracts are something to worry about if scale economies are big enough, but not otherwise.
2. What matters is whether the amount of the market foreclosed is big enough to entirely deter or expel a rival, not whether it hurts his profit. If it can’t kill him, it doesn’t fit the monopolizing motive of these models.
3. Courts should look at the details of the contracts and contracting process if the prosecutor leaps hurdles 1 and 2 and proves possible monopolizing. If the excludor seems to be trying to panic buyers, or makes discriminating offers in sequence, for example, that looks bad. If he includes a price-matching clause, that’s bad. If he uses some kind of conditional two-part tariff, that’s bad.
These models are, in the end, story-telling. Courts can understand that. The stories depend crucially on assumptions which are relatively easy to check. The most common problem for antitrust authorities is deciding what mergers to allow. That is not a hard theoretical problem, but it is a very hard practical problem, because it requires estimating elasticities of demand and predicting cost savings and product improvements from a merger. That’s a lot harder than checking on whether a firm can survive with 30% of the market or whether a contract includes a price-matching clause.
Simpson and Wickelgren, “Exclusive Dealing and Entry, when Buyers Compete: Comment,” June 2005, says:
Vigorous downstream competition affects the ability of exclusive contracts to deter entry in two ways. First, vigorous downstream competition substantially reduces the benefit a buyer obtains from securing a lower input price: Because downstream competition drives price toward marginal cost, most of the benefit from lower input costs are passed on to final consumers. This means that an upstream incumbent can induce downstream buyers to sign exclusive contracts by offering them a small side payment even where upstream scale economies are absent (See Simpson and Wickelgren (2004). Thus, the Chicago School argument (Richard Posner 1976; Robert Bork 1978) that exclusion is unlikely because upstream competition maximizes the joint surplus of the incumbent and buyers does not apply when buyers compete vigorously in a downstream market.
Second, vigorous downstream competition reduces the importance of upstream scale economies as a barrier to entry. Focusing on buyers as final consumers, Eric Rasmusen et al. (1991) and Ilya R. Segal and Michael D. Whinston (2000) (henceforth RRW-SW) found that exclusive contracts can deter entry when an entrant must sell to multiple buyers in order to attain scale economies. Given the entrant’s need to attain scale economies, a buyer imposes a negative externality on other buyers when it signs an exclusive contract. In the RRW-SW model, the incumbent monopolist gets buyers to sign exclusive contracts that inefficiently deter entry by exploiting this externality. Where buyers are downstream competitors rather than final consumers, a downstream firm with lower costs than its rivals can expand its market share and thus its purchases of the upstream input. Consequently, given sufficiently intense downstream competition, a single buyer can enable an upstream entrant to attain scale economies. FM’s paper focuses on this second effect while ignoring the first effect.
Joe Farrell speaking in “UNITED STATES FEDERAL TRADE COMMISSION 2 and 3 UNITED STATES DEPARTMENT OF JUSTICE 7 SHERMAN ACT SECTION 2 JOINT HEARING 8 UNDERSTANDING SINGLE- FIRM BEHAVIOR: 9 EXCLUSIVE DEALING SESSION 10 WEDNESDAY, NOVEMBER 15, 2006″, which is on the web at a 4- line address (google it) says:
Now, what about the other side of the courtroom, divide and conquer exclusion, Rasmussen and Ramseyer and Wiley, 1991, corrected, beefed up and radically improved by Segal and Whinston in the American Economic Review, 2000, show that exclusion can profitably and harmfully work against end users; however, although I think that is very well understood and accepted, the fact is their models involve buyers who are end users.
In most cases that I am aware of, exclusive dealing is not a deal struck with end users. It is a deal struck with retailers or distributors or someone else intermediate in the value chain between the manufacturer and the end users. That makes a lot of difference.
So, interestingly, a year or two ago, there appeared to be economics literature, two broadly parallel articles, papers, one by Fumagalli and Motta, which I believe has been published or is about to be published in the American Economic Review, and one by John Simpson and Abraham Wickelgren, and within the last 24 hours, I have learned about other articles by Yong and Shaffer that may be somewhat along the same lines, and both of these articles address the question, how does the RRWSW theory of anticompetitive exclusive dealing change when you recognize that the buyers in the model, in practice, should be replaced by buyers who are not end users?
Well, there are two forces, okay? One force is that intermediate buyers, nonfinal buyers, actually do not care that much if the price goes up or stays high, provided it goes up or stays high to all of them, because then it gets passed through downstream, okay? How much that is true depends on the details of the market structure and so on, but that tends to be true. That lowers their resistance to things that maintain monopoly upstream relative to what it would be if they were end users. So, that you would expect would make anticompetitive exclusive dealing easier.
Another force, however, is that if you have a nonfinal buyer who holds out and does not sign the exclusive deal, then an entrant can come to him and say, “Aha, I will give you a lower price than all your tied up rivals will be getting. You can expand. You and I can meet my scale requirements, and you will make a bundle of money.” So, that dynamic potentially makes it harder to have anticompetitive exclusive dealing. Well, Fumagalli and Motta found conclusively that it went one way, and Simpson and Wickelgren found conclusively that it went the other way, and which way Yong and Shaffer come out, I do not know yet. Which of them is right and when? Well, I attempted to diagnose this in my Antitrust Bulletin article last year. My attempted diagnosis is that it depends on whether in that last situation where you had one hold-out buyer, the incumbent is then able to or does adjust the price that it charges the tied buyers. So, I believe Fumagalli and Motta assumed that it does not, and Simpson and Wickelgren assumed that it does, or maybe it is the other way around, okay?
When I put this tentative diagnosis to one of the four economists — and I will not say which one — the response I got was, “Ah, that is interesting, I am not sure.” That is telling, I think, because it says that it is kind of unlikely that a court is going to do a very good job of disentangling all of these difficult concepts. Now, the optimistic view is this is just the beginning of the economic exploration of this topic, and come the year 2010, we will understand it well and in a way that is good enough for us to brief courts on it, and maybe that will happen, …