Competition and cooperation

Michigan Law Review

Volumn 97, Issue 1, 1998, Pages 216-243

  Levmore, Saul ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0039626647     PISSN: 00262234     EISSN: None     Source Type: Journal    
DOI: 10.2307/1290156     Document Type: Article

References (45)

1. Ronald H. Coase, The Nature of the Firm, 4 ECONOMICA 386 (1937), reprinted in RONALD COASE, THE FIRM, THE MARKET, AND THE LAW 33 (1988).
2. Ronald H. Coase, The Nature of the Firm, 4 ECONOMICA 386 (1937), reprinted in RONALD COASE, THE FIRM, THE MARKET, AND THE LAW 33 (1988).
3. I recognize that this expression implies that a firm can be distinguished from a set of arrangements, contracts, or norms. See Steven N.S. Cheung, The Contractual Nature of the Firm, 26 J.L. & ECON. 1 (1983).
4. Without getting too far ahead of the argument, it is perhaps obvious that the firm might make and sell to a competitor even though it is unwilling to buy in a cooperative way. An upscale, elegant retailer might, for example, be unwilling to stock a product bearing the name of a less elegant competitor, but the competitor might be willing to sell a product with Elegant's label, and Elegant might not regard its image as tarnished by the comparison outside of its premises. Elegant could make the product or buy it from an outsider with some conditions attached regarding the outsider's other sales.
5. 4.The discussion below takes account of the fact that municipalities may compete with one another less than most business firms compete against rival firms. Cooperation may therefore be less threatening. In Saul Levmore, Irreversibility and the Law: The Size of Firms and Other Organizations, 18 J. CORP. LAW 333 (1993), I argued that the tradeoff between the agency costs associated with internal expansion and those associated with transacting with an outside entity - the core of the make-or-buy decision - needs to be understood in a way that incorporates the agency problems absorbed by that outside entity. The outside entity expands in order to produce that which the first entity chooses to buy rather than make on its own. Thus, our theory of the firm must understand agency costs in relative, or opportunity cost, terms. X will externalize production and make arrangements with Y when Y can produce more cheaply than X (by enough to offset the greater friction present in X-Y arrangements as compared to monitoring and other organzational matters internal to X). A second step in the analysis is then to see the analogous relativistic character of "irreversibility," a term which refers to a kind of ratcheting, stickiness, or "hysteresis" such that for legal and psychological reasons organizations are slower to contract than to expand. Income tax laws, for example, would seem to be a cause of irreversibility because the well-known lock-in effect of realization rules makes taxpayers disinclined to sell appreciated assets, even to higher-valuing users. But because Y is likely to face the same tax laws as X, there is a serious limit to any claim that the size of X, in terms of its make-or-buy decisions, is much influenced by this factor. Indeed, once we see that previous analyses suffer from the failure to compare alternatives, it becomes obvious that we can often say no more than that the more efficient a manager, the less the need for external price signals. One obvious implication of this approach to irreversibility and the make-or-buy decision is that sole proprietorships are unlike firms with delegate authority. A publicly owned firm may set up barriers to expansion because it fears that its agents will selfishly be biased toward growth - but ex post, when these agents do lunge at opportunities for growth and avoid efficient contractoins, it is difficult to predict whether these firms will make more or buy more than their counterparts with no separation of ownership and control. The decisions are unlikely to be identical, unless the combination of precommitments (against internal and other expansions) and subsequent managerial behavior magically balances at the precise point reached by the owner-manager. Of course, sole proprietors must either absorb the agency costs associated with contracting with outsiders or the costs associated with internal growth and delegation, including the costs of obtaining funds from creditors (a source discussed at greater length below). Not-for-profit organizations, including governmental units, offer fewer obvious implications. I have already suggested, see id. at 356-58, that these organizations may be "stickier" than their for-profit counterparts in both directions. Universities and museums, for example, seem even less likely to shrink than do other organizations but they also seem less quick to expand. There are optimistic and pessimistic explanations for this sort of behavior. I turn below to aspects of this comparison that implicate the idea advanced here of a refusal to cooperate. See infra Part III.
6. I am tempted to concentrate with these examples on law schools and law firms, in part because these are fascinating organizations of great interest to most of my readers. There is a good deal to be gained, however, from suppressing this temptation and proceeding as if the business of law training and practice were not our real focus. In any event, it is useful to see how much law is like other businesses. Impatient readers are of course free to think of law reviews, schools, or firms in place of the manufacturers and retailers named below.
7. There are, of course, important examples of joint ventures in the auto industry, although it is perhaps notable that GM and Toyota share in such a venture while GM and Ford do not. There is also the interesting question of identified versus anonymous suppliers. Generally speaking, if components are supplied by a firm with something of an independent reputation, then they are more likely to be identified as such to the consumer. But cooperation itself may be more likely in some industries if the supplier is anonymous. Competing auto makers do, for instance, purchase a variety of auto body parts from common, unidentified suppliers. This choice, between identified and unidentified components, may have a great deal to do with the "original equipment tie." If consumers will buy replacement parts on their own, as is often true for tires but not for running boards, then an auto maker may profit in the form of lower prices from its suppliers by buying things like tires and installing them in new vehicles with the name of their outside maker clearly identified; the idea is that consumers may be overinclined to match the original equipment on their cars. Thus, Firestone will "pay" Ford in the form of prices that may be even lower than Firestone's marginal cost, but the maker of running boards will not offer a similar discount to Ford. One reason not to pursue this matter is that it seems to have little bearing on the central puzzles here. GM might also offer parts to Ford at low prices in order to take advantage of consumer perceptions but we do not observe GM parts on Ford vehicles. The text concedes, however, that supply arrangements between competitors are found in other industries. I return to this mixed evidence below. See infra text accompanying notes 28-29.
8. The contrast between implicit and explicit cooperation is less sharp than it might be even in the auto industry; there are components, ranging from engines to stereo systems, regarding which competitors seem unlikely to cooperate even through the use of a third-party supplier.
9. On the other hand, undiversified, single-brand stores do not seem less important a part of that industry than are single-brand catalogues in mail-order retailing. The distinction based on search costs may therefore be unhelpful.
10. I will not try to solve all these puzzles of where we do and do not find cooperation -even of the explicit kind. Consider, for example, the case of a delicatessen, Z, which makes fabulous bread in its bakery. If Z's bread is also sold at local grocery stores or even at other delicatessens in the same town as Z. we might be unsurprised by the explicit cooperation, figuring that Z profits from its superior product. Any disinclination of these other stores is overcome by their recognition of the fact that they would lose customers who would go elsewhere, or simply to Z itself, to buy bread and then might not return for other groceries or foodstuffs. On the other hand, if Z refuses to sell to competitors, we might conclude that Z hopes to attract bread-loving consumers who will then buy other products from Z. The claim must be that Z may not be able to discriminate as successfully by simply raising the price of its bread because it cannot raise the price just for these bread lovers. In short, many arrangements and practices seem to be consistent with rationality, and Z's actual practice (which was once to hold all bread as its own, later to sell a few of its many varieties of bread to competing stores, and now apparently to sell all of its varieties to other parties who sell the bread at the same retail price as Z) may not cast much light on the various puzzles posed by other actors and in other industries.
11. Schools do occasionally share the travel costs of applicants for faculty positions or itinerant presenters of papers at faculty workshops. These examples of cooperation seem trivial if only because an outsider's request for reimbursement from both hosts would amount to profitable and conventionally unethical overrecoveries. More interesting and more common sharing arrangements involve library collections because conventional wisdom would not have suggested exceptional efficiencies in this area.
12. At the personal, individual level there is much more cooperation. Faculty members at competing law schools invite each other to intellectual events, call on each other as referees and colleagues, and share information about candidates on the job market. These explicit examples of cooperation do not generate charges of disloyalty but they are also remarkably informal, avoiding for example any explicit terms of exchange. A remarkable example in the opposite camp is that there is little if any co-authorship between faculty members at neighboring law schools.
13. And this is the case even though economic thinking suggests rather strongly that specialization in the area of one's comparative advantage can lead to mutual advantage so that trading with another party does not imply that the other has an absolute advantage, or superiority, in producing that which it makes and sells.
14. Lands' End may seek to raise the average quality of the goods it impresses upon customers. It offers superior products but does not offer less expensive, lower quality products attached to a specific outsider's label. Presumably, the gain from improved selection is more than offset by the fear that consumers will either attribute the lower quality to Lands' End itself or simply decline to read the catalogue as thoroughly once they see items they do not like.
15. Competing outsiders are best because they eliminate the fear of collusion between one's competitor and supplier. The fear is not, however, a great one because an outsider who colludes with A at B's expense may eventually have some incentive to seek a better deal with B.
16. The example assumes that these department stores compete as neighbors in one or more locations.
17. One of the interesting things about final offer arbitration is that it has been put in place by explicit agreement of adversarial parties and, even where imposed on them, its strategy of giving the outsider less flexibility than other forms of arbitration draws on the efficiency we might associate with explicit rather than implicit cooperation.
18. The idea is that if not for risk aversion they might be happy to compete with one another for a better price in the marketplace.
19. The outsider is subject to competitive pressure from new entrants.
20. In other words, if there were relatively more cooperation in the for-profit sector, then implicit cooperation could be easily associated with price competition, and risk aversion regarding the division of gain or relative cost advantages.
21. And the inclination not to cooperate with respect to permanent fundraisers or development programs would need to be explained on simpler transaction cost or signaling grounds.
22. Analogously, of all the problems and puzzles posed by real estate agents, there does not seem to be a fear that they will unequally divide the gains from cooperation. See Saul Levmore, Commissions and Conflicts in Agency Arrangements: Lawyers, Real Estate Brokers, Underwriters, and Other Agents' Rewards, 36 J.L. & ECON. 503 (1993).
23. Note that while real estate agents cooperate explicitly in many regions by publishing magazines in which competing real estate brokers buy pages (and bypass local newspapers), competing auto makers do not advertise in this explicitly parallel manner although, to a limited degree, some do cooperate through single dealerships.
24. See infra Part III.
25. Note that the theories advanced in the present essay are necessarily incomplete because I make no further attempt to explain which firms or industries will fail to display a disinclination to cooperate. For every airline-industry sort of example there is a mail-order-retailer kind of counterexample with plenty of competition but little cooperation. Indeed, the airline industry itself offers a mixed message because inter-carrier booking was also the norm when that industry was much less competitive than at present.
26. See generally Vincent R. Johnson & Virginia Coyle, On the Transformation of the Legal Profession: The Advent of Temporary Lawyering, 66 NOTRE DAME L. REV. 359 (1990).
27. The modest cooperation among law firms that is observed with respect to outside libraries highlights the absence of coventures (or rentals) by firms occupying a single office building where they might, for example, maintain a single library. An optimist would say that law firms fear the inappropriate exchange of information if their lawyers use the same tables and books, but then it is mysterious that these lawyers are permitted to use university or bar association libraries where such inadvertent exchanges might also take place. It is possible that implicit cooperation through not-for-profit organizations reflects a bias in favor of this sort of limited cooperation generated by the tax system.
28. Indeed in some of these cases, such as that involving travel expenses, it is almost difficult to see how each firm could proceed without some sharing of expenses.
29. There are very few cases that can be described in this manner where explicit cooperation is actually found. I do not dwell on these cases because the very nature of the mistrust raises the problem of interpreting cooperation, which can simply be rationalized as an example of two enemies wishing to keep a close eye on one another. To take a political example, the Democrats and Republicans might sometimes entrust a sensitive matter to an independent outsider but at other times choose to appoint a joint venture with personnel drawn from the ranks of the party faithful but with an equal number from each party.
30. The example thus weakens any claim that cooperation is avoided where there is confidential information at stake.
31. For what it is worth, I have not yet encountered a faculty member or administrator at Columbia or NYU who thought those schools' behavior could be explained by anything but unilateral or mutual disinclinations to cooperate.
32. See Levmore, supra note 4, at 356-58.
33. But the lack of cooperation when both states have these schools does not necessarily reflect a disinclination to cooperate or a belief that cooperation sends negative signals. The potential gain in consumer choice or specialization may simply be more than offset by the transaction costs of the arrangement. In most instances where there is no explicit cooperation between neighboring states, residents of one state can apply as would residents of any other state to the host state's school. The question of cooperation arises only where tuition and admission preferences in a nearby state might reduce the political pressure for duplicative investments in facilities.
34. A topic I touch on in Part IV but consider more seriously in a companion paper, Saul Levmore, Efficiency and Conspiracy: Conflicts of Interest, Anti-Nepotism Rules, and Separation Strategies, 66 FORDHAM L. REV. 2099 (1998) (1997 Levine Lecture).
35. The consumer ignorance explanation might be salvaged with some painstaking identification of consumers' perceptions regarding the core function of an enterprise, with the idea being that cooperation in this core is what threatens to transmit a negative signal. The inquiry is then similar to that mentioned earlier in connection with firm pride. Cooperation with respect to tires but not engines is consistent with either of these theories.
36. Unless the perception is that Harvard gives up any claim in sciences and engineering while MIT shrugs off a variety of areas. There are in fact schools with catalogues that suggest credit for courses taken elsewhere only where the home school offers nothing comparable, but Harvard and MIT do not fall into this set and many of their faculty and administrators would strongly resist the extreme specialization claims that outsiders might make about these institutions.
37. I have avoided implicating the literature on exclusive dealing. The arrangements puzzled over here do not generally involve exclusive dealing, but rather something less than exclusivity. Moreover, the clever explanations advanced in such works as Howard P. Marvel, Exclusive Dealing, 25 J.L. & ECON. 1 (1982), and Victor Goldberg, Enforcing Resale Price Maintenance: The FTC Investigation of Lenox, 18 AMER. BUS. L.J. 225 (1980), are of little application to these puzzles of cooperation. Lands' End does sell belts other than Coach's and if Coach is happy to see its products alongside Lands' End's, it ought to be pleased to be seen in competitors' catalogues.
38. I have avoided implicating the literature on exclusive dealing. The arrangements puzzled over here do not generally involve exclusive dealing, but rather something less than exclusivity. Moreover, the clever explanations advanced in such works as Howard P. Marvel, Exclusive Dealing, 25 J.L. & ECON. 1 (1982), and Victor Goldberg, Enforcing Resale Price Maintenance: The FTC Investigation of Lenox, 18 AMER. BUS. L.J. 225 (1980), are of little application to these puzzles of cooperation. Lands' End does sell belts other than Coach's and if Coach is happy to see its products alongside Lands' End's, it ought to be pleased to be seen in competitors' catalogues.
39. Note that neither pride nor negative signaling offers much in the way of an explanation of the real estate example.
40. Other examples include milk delivery (where duplication and noncooperation may have accelerated the decline of the industry in favor of more frequent trips to grocery stores) and express mail. Federal Express and its competitors could cooperate by having a single messenger visit each small business location. In the long run, we may observe a kind of implicit cooperation of just this sort by way of businesses offering short-term monopolies to the lowest bidder from among the competing express services. There would still be some duplication of routes, unless the bids perfectly segregated the market geographically, but much less duplication of elevator trips and the like. Note that the United States Postal Service - much more than Federal Express - does in fact explicitly cooperate by contracting for trucks and aircraft in a manner that permits cooperation with other shipper and travelers. See generally Leonard Merewitz & Mark A. Zupan, Franchise Bidding, Contracting Out, and Worksharing in the Production of Postal Services, in REGULATION AND THE NATURE OF POSTAL AND DELIVERY SERVICES 69 (Michael A. Crew & Paul R. Kleindorfer eds., 1993).
41. A tempting explanation for noncooperation is that retailers might mislead customers and suppliers by connecting hoses to underground tanks in a manner that simply delivered the cheapest gas to the customer regardless of the latter's expressed preference a the pump. But this explanation standing alone is unconvincing because cooperation might still be worth-while through a dealer who recommitted to buy gas from competing suppliers at a uniform price, because occasional monitoring with severe penalties ought to discourage dealer fraud, and because we find single-brand stations offering several grades of gasoline even though a similar fraud problem presents itself with these tanks. The brand-name supplier's incentive to monitor means that inter-brand fraud is probably less serious a problem than the fairly substantial problem of fraud as to grade. See I.P.L. Ping & David Reitman, Why are Some Products Branded and Others Not?, 38 J.L. & ECON. 207, 213 (1995) (noting that in one study 15.4% of premium gasoline samples were substandard). Note that the same unsatisfactory explanation might be tried for the question of why restaurants do not offer both Coke and Pepsi while grocery stores do. In this case, however, the fraud explanation is even weaker because of the presence of restaurant patrons with refined taste buds.
42. The equal division explanation is somewhat more successful, although cooperation and equal division would seem manageable through the use of a cost-sharing formula based on gallons sold. It is difficult to think of reasons why gasoline should be such a prominent example of an item sold through stand-alone retailers. A store that sells only Coach leather products or Krispy Kreme donuts can focus its marketing efforts and enjoy other advantages, but then it loses some gains from cooperation in terms of off-peak hours and attracting consumers who care about reducing their shopping time. Stand-alone donut stores have in fact given way somewhat to the creation of outlets in supermarkets and convenience stores, even though those stores carry competing products, but the combination seems sensible because these stores have serious overcapacity in the morning, donut hours. Gasoline is sometimes sold by convenience stores, but not alongside competing products.
43. For an extended discussion of just this topic, see Levmore, supra notes 4, 33.
44. Alternatively, the law may cause contraction if internal production is too costly.
45. I have intentionally advanced examples where the role of law is quite minimal. I have, for instance, avoided the case of cooperation in the form of competitors patronizing a single law firm for such things as environmental or insurance defense work. The companion paper, cited in note 33, discusses lawyers, bankers, trustees, government employees, and other third-party suppliers whose roles may be guided or constrained substantially by legal rules.