Subaru of New England, Inc. (SNE), a regional Subaru distributor, appeals an antitrust treble damage award-totalling $51,729.59, 534 F.Supp. 1282, (plus $200,590.18 costs and attorneys fees). The award rests upon the district court’s determination that SNE, as a condition of making Subaru cars available to Grappone, Inc. (Grappone), a local automobile dealer, required Grappone to buy some spare Subaru parts. The court held that this conditioned sale constituted a “tying” arrangement, unlawful under Section 1 of the Sherman Act, 15 U.S.C. § 1 (1982 & Supp. II 1984), and Section 3 of the Clayton Act, 15 U.S.C. § 14 (1982). Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2, 104 S.Ct. 1551, 80 L.Ed.2d 20 (1984); International Salt Co., Inc. v. United States, 332 U.S. 392, 68 S.Ct. 12, 92 L.Ed. 2 (1947). SNE argues that the district court’s determination of liability rests upon an overly mechanical, and legally erroneous, application of the law prohibiting tying. We agree; and, we reverse the judgment.

I

The basic facts are as follows: In 1973, SNE’s supplier, Subaru of America (SOA), was worried that Subaru dealers had not had, and in 1974 might not have, enough spare parts to make repairs to recently purchased Subaru cars. After a review of past buyers’ needs and consultation with Subaru headquarters in Japan, SOA made up a list of spare parts kits that it wished both regional distributors and local dealers to have on hand. In particular, it wanted each local dealer to have several “dealer’s kits,” each with 88 parts for the 1974 cars (plus a few parts for previous models) and also to have “supplemental kits,” each with 44 spare parts, out of a total of approximately 4,000 to 5,000 different Subaru parts. SOA said that readily available spare parts would help Subaru sell its cars, which at that time (1973/74) accounted for a very small share of American, New England, and New Hampshire auto sales. In fact, in 1974, Subaru accounted for 3.4 percent of all auto imports in New Hampshire, 2.8 percent in New England, and 1.5 percent in the United States. (Our own research indicates that in 1974, imports accounted for 15.9 percent of all auto sales, which means that Subarus likely accounted for about one-quarter of one percent of auto sales in that year in the United States, and apparently under one percent of all auto sales in New England and New Hampshire. U.S. Dept. of Commerce, Statistical Abstract of the United States 1988, Chart No 992.)

As part of the “replacement part availability” effort, SNE (the regional distributor) told Grappone, Inc. (the local dealer) that it must have on hand, in 1974, two dealer kits and two supplemental kits. Grappone objected on the ground that the parts kits were overly inclusive. Grappone believed the combined kits should contain fewer parts and sell for about $2,000, instead of about $3,300, leaving it free to obtain the extra parts elsewhere. SNE insisted that Grappone take the kits, and it told Grappone that it could not have its allocation of Subaru cars in 1974 unless it did so. Although Grappone sold AMC, Pontiac, and Jeep cars, and (from another site) Toyotas and Peugeots (indeed the full name on the stationary is “Grappone Pontiac, Inc.”), it wanted the Subarus as well.[*794] Grappone went 10 months without the new 1974 Subarus; it then agreed to take the kits; SNE accepted Grappone’s July 1974 car order; and Grappone brought this lawsuit. After many legal events, which we need not recount, Grappone eventually won the verdict mentioned, on the basis of its tying claim. In our view, however, even that small victory was without adequate legal basis.

II

The case law has long indicated, regardless of whether a plaintiff charges a violation of Sherman Act § 1 or Clayton Act § 3, that a “tying arrangement” is unlawful where (1) the seller has “market power” in the tying product, Jefferson Parish, 466 U.S. at 17-18, 104 S.Ct. at 1560-61; Fortner Enterprises, Inc. v. United States Steel Corp., 394 U.S. 495, 499, 89 S.Ct. 1252, 1256, 22 L.Ed.2d 495 (1969) (“Fortner I”); (2) the tie links “two separate products,” Jefferson Parish, 466 U.S. at 18, 19-22, 104 S.Ct. at 1561, 1562-64; Times-Picayune Publishing Co. v. United States, 345 U.S. 594, 613-14, 73 S.Ct. 872, 883, 97 L.Ed. 1277 (1953); and (3) the tie forecloses a “not insubstantial” amount of potential sales for the “tied” product, International Salt, 332 U.S. at 396, 68 S.Ct. at 15 (Sherman Act § 1, Clayton Act § 3); see also Jefferson Parish, 466 U.S. at 16, 104 S.Ct. at 1560; Northern Pacific Ry. Co. v. United States, 356 U.S. 1, 11, 78 S.Ct. 514, 521, 2 L.Ed.2d 545 (1958) (Sherman Act § 1); Wells Real Estate, Inc. v. Greater Lowell Board of Realtors, 850 F.2d 803, 814 (1st Cir.1988). In our view, the plaintiffs in this case have failed to show the first of these essential elements.

1. We have reached our conclusions after examining with care the Supreme Court’s recent analysis of tying arrangements in Jefferson Parish, supra. Although the Court divided 5-4 over whether courts should treat tying as unlawful per se, we read both majority and concurring opinions as accepting the following basic propositions.

First, in Jefferson Parish, as in previous cases, the Court recognizes that the antitrust laws exist to protect the competitive process itself, not individual firms. Brown Shoe Co. v. United States, 370 U.S. 294, 320, 82 S.Ct. 1502, 1521, 8 L.Ed.2d 510 (1962); Will v. Comprehensive Accounting Co., 776 F.2d 665, 673-74 (7th Cir.1985), ce rt. denied, 475 U.S. 1129, 106 S.Ct. 1659, 90 L.Ed.2d 201 (1986). And, the antitrust laws protect the competitive process in order to help individual consumers by bringing them the benefits of low, economically efficient prices, efficient production methods, and innovation. See Broadcast Music, Inc. v. Columbia Broadcasting System, Inc., 441 U.S. 1, 19-20, 99 S.Ct. 1551, 1562, 60 L.Ed.2d 1 (1979); Interface Group, Inc. v. Massachusetts Port Authority, 816 F.2d 9, 10-11 (1st Cir.1987); Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d 227, 231, 234-35 (1st Cir.1983).

Second, the Court recognized that, typically, the way in which tying may hurt the competitive process itself is by helping to maintain or to augment pre-existing market power, defined by the Court as the power to “raise[] [price] above the levels that would be charged in a competitive market.” Jefferson Parish, 466 U.S. at 27 n. 46, 104 S.Ct. at 1566 n. 46.

Third, both the majority and minority opinions in Jefferson Parish recognized that tying’s anticompetitive mechanism is not obvious. Compare Jefferson Parish, 466 U.S. at 11-12, 14-15, 104 S.Ct. at 1558, 1559-60 (recognizing some combined sales are consistent with rather than compromise “competitive ideal” of Sherman Act), and at 34, 35-36 (O’Connor, J., concurring) (need for “elaborate inquiry into ... [tying’s] economic effects”) with International Salt, 332 U.S. at 396, 68 S.Ct. at 15 (“the tendency of the [tying] arrangement to accomplishment of monopoly seems obvious”); Northern Pacific, 356 U.S. at 6, 78[*795] S.Ct. at 518 (“Where [tying] ... conditions are successfully exacted[,] competition on the merits ... is inevitably curbed.”). Rather, to understand the harm that tying may cause requires a fairly subtle antitrust analysis. One cannot infer an automatic harm to the competitive process simply because a Seller refuses to sell Product A to a Buyer unless the Buyer also buys Product B. That is to say, a tie does not “obviously” hurt a Buyer by making it take a product it does not want. Indeed, the tie itself cannot automatically, in and of itself, force the Buyer to take an inferior Product B or to pay more than it wishes for that product. If the Seller does not have market power in respect to product A, it cannot force buyers to take a more expensive or less desirable Product B, for if the Seller tries to do so, buyers will simply turn elsewhere for Product A. If the Seller does have, and has been fully exercising, market power, it also cannot force buyers to take a more expensive or less desirable Product B, unless it provides buyers equivalent compensation by lowering the price of Product A (or maintaining Product A’s price at a level lower than the Seller has the power to charge), for otherwise buyers, who were already paying as much as the Seller could charge them (with its degree of market power) would also likely switch to other sellers or discontinue use of Product A. See Jefferson Parish, 466 U.S. at 39 & n. 8, 40, 104 S.Ct. at 1572 & n. 8, 1573 (O’Connor, J., concurring). This is simply to make the logical point that “fully exploited” buyers would not take a less desirable Product B without compensation, for otherwise they were not being “fully exploited.” See id.; 3 P. Areeda and D. Turner, Antitrust Law ¶ 725b (1978 & Supp.1987) (hereafter Areeda and Turner); see, e.g., Moore v. Jas. H. Matthews & Co., 550 F.2d 1207, 1213 (9th Cir.1977) (buyers interested in end price of product). Scholars have elaborated and refined this logical point, taking account, for example, of the fact that sellers typically do not set individualized prices and that typically some buyers may have special preferences for a particular seller of even the most competitive product. See, e.g., 2, 3 Areeda and Turner ¶¶ 409d, 410, 725e. The upshot (which the four concurring Jefferson Parish justices called “counterintuitive,” 466 U.S. at 36, 104 S.Ct. at 1570) is that a ‘tie’ does not hurt the typical buyer in any obvious way; one needs a more refined analysis to find the harm. See, e.g., P. Areeda, Antitrust Analysis ¶ 531(a) (3d ed. 1981); 3 Areeda and Turner ¶¶ 725b, 733; R. Bork, The Antitrust Paradox 365-81 (1978); R. Markovits, “Tie-ins, Reciprocity, and the Leverage Theory,” 76 Yale L.J. 1397 (1967); W. Bowman, Jr., “Tying Arrangements and the Leverage Problem,” 67 Yale L.J. 19 (1957).

The Jefferson Parish Court provides that more refined analysis. The majority and concurrence recognized that a Seller, possessing significant market power with respect to Product A, may cause anticompeti-tive harm by tying as follows: by reducing the price of Product A slightly (or by otherwise not fully exploiting its power with respect to Product A), the Seller may induce the Buyer to accept the tie; by doing so, the Seller may build a strong market position in Product B; and that position in Product B, in turn, may increase its power to charge high prices in respect to Product A. If a monopolist of patented can-closing machinery, for example, insists, as a condition of selling his machines, that their purchasers buy his cans, he will likely soon have a monopoly in cans as well as machines. And, that fact — the fact that he controls both cans and machines — may make his monopoly safer from competitive attack when his patent on the can-closing machinery expires. A new competitor would then have to enter both levels of the business (cans and machines) to deprive him of monopoly profits. And, this added security may enable the machinery monopolist to charge higher prices. The tie, by permitting the Seller to extend its market power from one level to two, may thereby[*796] raise entry barriers, providing security that helps a monopolist-seller further harm the consumer. This point is made both by the Jefferson Parish majority, 466 U.S. at 14, 104 S.Ct. at 1559, and the concurring justices, id. at 36-37, 39, 104 S.Ct. at 1570-71, 1572; see also 3 Areeda and Turner ¶¶ 725h, 733e.

Of course, in such circumstances, tying would hurt the Buyer or consumer only when it first hurts firms seeking to sell the tied product. Only if the tie significantly reduces the opportunities to sell Product B, can the tie significantly increase the Seller’s power in respect to Product B, and thereby (i.e., by raising entry barriers) increase the Seller’s power in respect to Product A. See Jefferson Parish, 466 U.S. at 14, 18-25, 104 S.Ct. at 1559. And, insofar as tying impedes “competition on the merits,” discouraging the search for innovation or efficiency, it does so in the tied product markets. Jefferson Parish, 466 U.S. at 14, 104 S.Ct. at 1559; 3 Areeda and Turner ¶ 725g; see, e.g., ILC Peripherals Leasing Corp. v. International Business Machines Corp., 448 F.Supp. 228 (N.D.Cal.1978), aff 'd per curiam sub nom. Memorex Corp. v. International Business Machines Corp., 636 F.2d 1188 (9th Cir.1980), cert. denied, 452 U.S. 972, 101 S.Ct. 3126, 69 L.Ed.2d 983 (1981).

The majority and minority opinions in Jefferson Parish disagree, not in respect to the nature of the link between tie and potential competitive harm, but in respect to the legal conclusions they would draw from the nature of this linkage. The minority would abandon the per se anti-tying rules and analyze tying under a “rule of reason”; it would prohibit tying only when, according to its “demonstrated economic effects[,] ... [tying’s] anticompetitive impact outweighs its contribution to efficiency.” Jefferson Parish, 466 U.S. at 41-42, 104 S.Ct. at 1573-74. The majority would retain pre-existing per se rules; but it also breathes life into the screening function that the preconditions of those per se rules serve. The majority, for example, makes clear that by its requirement of “market power” it means significant market power — more than the mere ability to raise price only slightly, or only on occasion, or only to a few of a seller’s many customers. The Jefferson Parish Court makes this point through its negative holding: the fact that 30 percent of East Jefferson Hospital patients came from Jefferson Parish did not show that the Hospital, because of its geographical location, possessed market power in respect to those who lived in that parish. And, the Court suggests the same point through what it says: “[wjhen ... the seller does not have either the degree or the kind of market power that enables him to force customers to purchase a second, unwanted product in order to obtain the tying product,” the per se rules do not apply. Jefferson Parish, 466 U.S. at 17-18, 104 S.Ct. at 1560-61. To show market power, the seller must have a significant market share or sell a unique, such as a patented, product for which there are not readily available substitutes. Id. at 16-17, 104 S.Ct. at 1560; see also United States Steel Corp. v. Fortner Enterprises, Inc., 429 U.S. 610, 620 n. 13, 97 S.Ct. 861, 868 n. 13, 51 L.Ed.2d 80 (1977) (“Fortner II”) (market power means seller must be able to raise price or impose burdensome terms on an “appreciable” number of buyers) (emphasis added).

That the “market power” hurdle is moderately high — that it cannot ordinarily be surmounted simply by pointing to the fact of the tie itself or to a handful of objecting customers — makes sense in light of the harms the anti-tying rules seek to avoid. After all, sellers typically set fairly uniform prices designed to attract a large number of buyers, not simply a handful of buyers who have some unusual and special preference for its products; a seller who has the power to raise prices only in respect to that special handful is a seller who cannot easily cause harm in tied product[*797] markets; and therefore one who cannot easily harm consumers. Of course, virtually every seller of a branded product has some customers who especially prefer its product. But to permit that fact alone to show market power is to condemn ties that are bound to be harmless, including some that may serve some useful social purpose. See, e.g., United States v. Jerrold Electronics Corp., 187 F.Supp. 545 (E.D.Pa.1960) (“tie” between equipment and installation aided success of system and growth of industry), aff'd per curiam, 365 U.S. 567, 81 S.Ct. 755, 5 L.Ed.2d 806 (1961). Though some members of the Supreme Court might go further, the holding of Jefferson Parish convinces us that the entire Court means the “market power” requirement to be serious enough to screen out this class of harmless tie.

2. The plaintiffs here cannot meet the significant “market power” requirement of Jefferson Parish. They have not provided evidence that SNE could raise prices significantly above the competitive level; in fact, the evidence in the record indicates the contrary. For one thing, Subaru’s market share, whether measured in terms of sales of all autos or of imports or in any other reasonable way, is miniscule. See Kenworth of Boston, Inc. v. Paccar Financial Corp., 735 F.2d 622, 623-24 (1st Cir.1984) (tying product is Paccar trucks, relevant market is that for all heavy trucks); Kingsport Motors, Inc. v. Chrysler Motors Corp., 644 F.2d 566, 571 (6th Cir.1981) (tying product is Dodge cars, relevant market is that for all medium priced automobiles sold in U.S.); Anderson Foreign Motors, Inc. v. New England Toyota Distributor, Inc., 475 F.Supp. 973, 986 (D.Mass.1979) (tying product is Toyotas, relevant market is that for all new foreign and domestic cars); see also A.I. Root Co. v. Computer/Dynamics, Inc., 806 F.2d 673, 675 (6th Cir.1986) (tying product is specialized computer equipment, relevant market is that for all small computers). National sales of Subarus during the relevant years (as suggested by the import figures in the record) were likely about a fraction of 1 percent of all autos sold. Subaru at most accounted for 3.4 percent of auto imports sold in New Hampshire in 1974. (The record contains references to something called “competition” cars, in respect to which SNE’s 1974 market share was, at most, two or three percentage points higher (a maximum of 5.6 percent in the New Hampshire market.)) If the fact that Jefferson Parish Hospital recieved 30 percent of all hospital patients living in East Jefferson Parish did not show market power, it is difficult to see how these far smaller figures could show the contrary here. Cf. A.I. Root, 806 F.2d at 675 (2-4 percent market share insufficient); Kenworth of Boston, 735 F.2d at 624 (18 percent share insufficient); Phillips v. Crown Central Petroleum Corp., 602 F.2d 616, 629 (4th Cir.1979) (10 percent share “probably ... very close to the minimum permissable”), ce rt. denied, 444 U.S. 1074, 100 S.Ct. 1021, 62 L.Ed.2d 756 (1980); U.S. Dept. of Justice, Vertical Restraint Guidelines § 5.3 (1985) (market shares less than 30 percent reated as insufficient), quoted in Will, 776 F.2d at 672. Were we to try to measure SNE’s market share, in terms of the market for auto franchises, the plaintiff fares no better. See Mozart Co. v. Mercedes-Benz of North America, Inc., 833 F.2d 1342, 1346-47 (9th Cir.1987) (where franchise agreement required spare parts to be purchased from distributor, tying product market is that for automobile dealerships). The record shows that Subaru was but one of many automobile franchisors, and that Grappone itself held not only Subaru, but also AMC, Pontiac, Jeep, Toyota, and Peugeot franchises.

For another thing, Grappone concedes that out of 64 Subaru dealers who accepted the parts kits, only three seriously protested SNE’s efforts to require purchase of the kits. As far as the record is concerned, the remaining 61 may well have believed SNE’s explanation, presented at a meeting where even a protesting dealer admitted that SNE “drummed up” interest in the kits, that the kits would help maintain customer satisfaction; the dealers then would have wished to obtain both cars and kits. Supp.App. at 70. And given the plausible business rea[*798] son for accepting both products, their having done so without protest makes that acceptance virtually worthless as evidence of SNE’s market power. See Fortner II, supra (acceptance of competitive deal is not evidence of coerced tie). Those few instances in which protests suggest that SNE took advantage of a special preference for its product cannot by themselves demonstrate significant SNE market power — the power to coerce an “appreciable” number of buyers. See Fortner II, 429 U.S. at 618 n. 10, 620 n. 13, 97 S.Ct. at 867 n. 10, 868 n. 13 (to show market power, plaintiff must show that “an appreciable number” of buyers accepted the tie in “the absence of other explanations for the[ir] willingness ... to purchase the package”) (emphasis added).

Further, the amount of money that best measures the extent to which SNE may have coerced Grappone — the extra amount of charge for the kits — was small. It amounted to the $1,000 worth of “unnecessary” parts in each of two combined kits supplied for one year. Given the fact that Grappone could expect to sell about 24 Subarus in 1974 according to the district court’s findings, SNE’s ability to extract so small an amount of “extra” money by itself shows very little power, even in respect to Grappone, particularly given the fact that Grappone distributed many other cars as well. Given Subaru’s tiny market share, why, if SNE or Subaru sought to raise prices significantly or to impose other serious barriers, would Grappone (which calls itself “Grappone Pontiac”) not simply have allocated more time, space, and effort to selling its AMC, Pontiac, Jeep, Toyota, and Peugeot cars instead of Subarus?

Finally, plaintiff has made no showing that Subarus had any special or unique features, such as patents or copyrights, that might demonstrate market power. See Digidyne Corp. v. Data General Corp., 734 F.2d 1336, 1341-42 (9th Cir.1984), ce rt. denied, 473 U.S. 908, 105 S.Ct. 3534, 87 L.Ed.2d 657 (1985); cf. Jefferson Parish, 466 U.S. at 16-17, 104 S.Ct. at 1560. The most one can say is that Subaru has a brand name and sells through “authorized” Subaru dealers. But, we find no Supreme Court case law suggesting that such features by themselves are sufficient to show “market power.” They ■ do not automatically demonstrate any “economic” or “cost advantage” or any other advantage “not shared by ... competitors,” Fortner II, 429 U.S. at 620-21, 97 S.Ct. at 867-868 (quoting in part Fortner I, 394 U.S. at 505 n. 2, 89 S.Ct. at 1259 n. 2), particularly in an industry where many larger competing firms all do business in this way. Will, 776 F.2d at 672 (no uniqueness argument unless plaintiff shows barrier preventing rivals from offering same package at same cost); Spartan Grain & Mill Co. v. Ayers, 581 F.2d 419, 426-27 (5th Cir.1978) (same), cert. denied, 444 U.S. 831, 100 S.Ct. 59, 62 L.Ed.2d 39 (1979); cf. Ungar v. Dunkin’ Donuts of America, Inc., 531 F.2d 1211, 1224-25 (3rd Cir.) (proof of coercion needed, otherwise factors “inherent” in voluntary franchise agreement would suffice to show tie), cert. denied, 429 U.S. 823, 97 S.Ct. 74, 50 L.Ed.2d 84 (1976). Grappone points to Digidyne, supra, as contrary authority, but in that case the Ninth Circuit found a tying product, a computer operating system, “unique” because (1) it was copyrighted, and (2) many of its buyers had previously invested millions of dollars in software design that worked only with that system. We find no directly analogous circumstances here. Grappone does not offer similar evidence of such large dealer investments in Subaru dealerships by those who protested the tie; Grappone’s multiple brand representation suggests the contrary, at least for some significant number of franchisees; and, given the strong competition by other brands, we do not see how such dealer investment (if it existed) could easily translate into Subaru market power of a kind that, through tying, could ultimately lead to higher than competitive prices for consumers. Regardless, the product here in question, unlike that in Digidyne, is not copyrighted; and therefore does not benefit from whatever legal market power “uniqueness” a copyright still offers in the tying contest. Cf. Jefferson Parish, 466 U.S. at 16-17, 104 S.Ct. at 1560.

[*799] III

Without the use of per se rules the plaintiff cannot show a violation of the antitrust laws, for the “tie’s” anticompetitive effects do not outweigh its legitimate (here proeompetitive) business justifications. Jefferson Parish at 31, 104 S.Ct. at 1568. See 7 Areeda & Turner § 1500, at 362-63. For one thing, the record does not demonstrate an actual anticompetitive effect in the tied product market. See pp. 797-98, supra. Rather, the district court found a potential effect. It said that, without the tie, replacement dealers capable of supplying some of the parts might decide to design those spare parts themselves and to enter the “Subaru part” replacement market. But, we could find no evidence (and given the tiny numbers it is unreasonable to think) that any decision by others to enter that market for a year was likely or that their failure to do so could have had a significant anticompetitive impact. That any such effect was of little significance is reinforced by Grappone’s claim that some of the extra parts were unnecessary; to that extent, anticompetitive impact is impossible, for purchases from SNE could not have replaced purchases from any other actual, or potential, SNE competitor. Jefferson Parish, 466 U.S. at 16, 104 S.Ct. at 1560 (“when a purchaser is ‘forced’ to buy a product he would not have otherwise bought even from another seller in the tied product market, there can be no adverse impact on competition because no portion of the market which would otherwise have been available ... has been foreclosed”). Indeed, we are not certain whether the evidence of tied product harm would have been strong enough to invoke per se antitying rules even had our decision in respect to the tying product been different (though we need not decide the matter). See Jefferson Parish, 466 U.S. at 22, 104 S.Ct. at 1563 (per se antitying rules apply only where the plaintiff can “identify a distinct product market in which it is efficient to offer [the tied product] separately from [the tying product]”). We are certain, however, that the tie shows no more than trivial effects when judged under a “rule of reason.”

For another thing, the record contains evidence of legitimate, proeompetitive, business justifications for the tie. The parties refer to portions of the record that indicate that SOA, the American Subaru importer, developed its spare parts package only after dealers complained of part supply shortages that made it more difficult to build a market for Subarus. An executive from one of Subaru’s oldest distributors also testified without contradiction that such parts supply problems had cut Renault’s sales of 100,000 per year in the 1960’s to 5,000 to 6,000 by 1977. And, the district court found the availability of sufficient spare parts “basic” to the successful marketing of automobiles. SOA determined which particular spare parts customers had needed in the first year after purchase, and which spare parts had not been readily available. It used this list to develop its spare parts kits. SNE insisted that kits be available only for the first year after the cars were sold. In other words, the record indicates that Subaru was a small firm attempting to break into the industry; it shows that the tie plausibly served the proeompetitive purpose of helping the firm develop or maintain sales. Given the absence of serious anticompeti-tive impact, this evidence of justification is sufficient to qualify the agreement as lawful under a rule of reason (a question that the district court did not directly consider). Indeed, it is conceivable (though we need not decide), that the “tie” was “efficient” enough a way to do business that the agreement could have escaped per se condemnation under the lines of cases that have created certain exceptions to the per se rule where economic, or proeompetitive justifications are particularly strong. See, e.g., Jerrold Electronics Corp., supra; Will, 776 F.2d at 672-74 (package sale); see also Jefferson Parish, 466 U.S. at 22, 104 S.Ct. at 1563 (applying per se tying rules only where “it is efficient” to offer the tied and tying products separately); International Salt Co., Inc., 332 U.S. at 397, 68 S.Ct. at 15 (lessor can impose “reasonable restrictions ... to minimize mainte[*800] nance burdens and to assure satisfactory operations”).

For these reasons, we conclude that the plaintiff has failed to prove that the per se anti-tying rules apply in this case; nor has the plaintiff shown a trade practice, the anticompetitive effects of which outweigh its legitimate business justifications. The judgment of the district court is

REVERSED.