Appellants Waste Management, Inc. ("WMI") and EMW Ventures Incorporated ("EMW") appeal from Judge Griesa's decision, 588 F.Supp. 498 (1983), after a bench trial, that WMI's acquisition of EMW violated section 7 of the Clayton Act, as amended, 15 U.S.C. Sec. 18, and from the resultant order that WMI divest Texas Industrial Disposal, Inc. ("TIDI"), a former EMW subsidiary.

We reverse.

This government antitrust action challenges WMI's acquisition of EMW's common stock. After the district court denied the government's motion for a temporary restraining order, the acquisition was consummated, and the case proceeded to trial before Judge Griesa sitting without a jury.

We summarize those facts that are not in dispute. WMI is in the solid waste disposal business. It provides services in twentyseven states and had revenues of approximately $442 million in 1980. At the time of the acquisition, EMW was a diversified holding company that owned a subsidiary by the name of Waste Resources, which was in the waste disposal business in ten states and had revenues of $54 million in 1980.

WMI and Waste Resources each had subsidiaries that operated in or near Dallas.^[1] WMI has one subsidiary, American Container Service ("ACS") in Dallas, and another, Texas Waste Management, in the Dallas suburb of Lewisville. Waste Resources had a Dallas subsidiary called Texas Industrial Disposal, Inc. ("TIDI"). WMI now operates TIDI as a WMI sub.

Waste collection involves several different types of equipment and serves the needs of various types of customers. For present purposes, it is important to distinguish between "non-containerized" and "containerized" equipment. "Non-containerized" refers to trucks with compactors into which trash cans and bags are loaded by hand. "Containerized" equipment consists of two types of receptacles, "dumpsters" and "roll off," each emptied by different kinds of trucks. Dumpsters typically have a volume of one to eight cubic yards and are emptied by "front-load" trucks that pick the dumpsters up with clamps and empty them into a hopper. Roll-off containers range up to 50 cubic yards in volume and are carried to a dump, emptied and then returned. Trucks that transport roll-off containers are known, not surprisingly, as roll-off trucks. If the customer desires containerized service, the waste hauler provides the dumpster or roll-off container.

There are various relevant classes of customers: (i) single or multiple dwelling residential customers; (ii) apartment complexes of varying size, (iii) "business" customers--stores, restaurants, etc., and (iv) "industrial" customers--construction sites, factories, etc. Customers choose among the kinds of services according to their individual needs, the quantity of trash produced being a critical factor.

The parties strenuously disagree over the proper definition of the relevant product and geographic markets. The government contended in the district court that the product market should be defined in terms of equipment type and that front-load and roll-off waste collection service each constitutes a separate product market. WMI argued that the market includes all forms of waste collection. The district court adopted a definition of the relevant product market that differed from the positions of both parties. Judge Griesa concluded that the product market included all trash collection, except for collection at single-family or at multiple family residences or small apartment complexes. Rejecting WMI's contentions as to the relevant geographic market, the district court excluded Tarrant County, which includes Fort Worth, thus limiting the market to Dallas County plus a small fringe area.

Based on revenue data, Judge Griesa found that the combined market share of TIDI and ACS was 48.8%.^[2] He viewed that market share as prima facie illegal under United States v. Philadelphia National Bank, 374 U.S. 321, 364-66, 83 S.Ct. 1715, 1742-43, 10 L.Ed.2d 915 (1963). Agreeing with appellants that entry into the product market is easy--indeed, individuals operating out of their homes can compete successfully "with any other company"--Judge Griesa nevertheless held that proof of ease of entry did not rebut the prima facie showing of illegality. The district court therefore ordered WMI to divest itself of TIDI. Because we conclude that potential entry into the relevant Dallas market by new firms or by firms now operating in Fort Worth is so easy as to constrain the prices charged by WMI's subs, we reverse on the grounds that the merged firm does not substantially lessen competition.

WMI raises the following claims which we treat seriatim: (1) the district court's definition of the relevant geographic and product markets and its determination of WMI's post-merger market share are erroneous; and (2) the district court erred in rejecting WMI's rebuttal of the prima facie case.

A. Market Definition and Determination of Market Share

In determining the relevant product and geographic market, Judge Griesa considered only firms that presently collect waste in competition with WMI's subs TIDI and ACS. He thus did not consider the effect of potential competition by new entrants upon the market power of the merged firm but rather treated this as part of the rebuttal to the prima facie case. Although potential competition resulting from easy entry can as logically be appraised as part of market definition, see R. Posner, Antitrust Law 132-33 (1976), we will utilize the traditional analysis of defining the market in terms of existing competitors.

We thus begin by determining which firms presently collect waste in actual competition with WMI (a term we use henceforth to include TIDI and ACS). WMI provides waste disposal services to business and industrial customers as do a number of other private haulers. Neither WMI nor these other haulers provides substantial services to residential customers since residential waste is generally collected by private or municipal haulers who provide only limited services to business/industrial customers. Residential and business/industrial customers are thus served largely by different firms.

One reason for the distinction between residential and business/industrial collection is found in customer preferences for collection by particular equipment at existing prices. Thus, residential customers largely prefer non-containerized service while business/industrial customers usually opt for containerized hauling. There is of course overlap, and the district court therefore did not define the market solely in terms of equipment distinctions. Some large apartment complexes prefer containerized service (and were included by Judge Griesa as customers in the relevant market), and some businesses make do with hand collection (also included as customers in the relevant market).

Another reason for the distinction (as modified by the overlaps) between residential and business/industrial collection is that most private haulers provide only containerized service and most municipalities provide only non-containerized service. The largest municipality, the City of Dallas, provides no containerized service. Moreover, the few municipalities that do provide containerized service to business/industrial customers were included as sellers in the relevant product market.

At bottom, however, the distinction between residential and business/industrial service rests on two facts, one economic and the other political. First, customers with large amounts of trash have a greater need for containerized service than do customers with smaller quantities; the former tend to be business/industrial and the latter residential. Second, the latter customers are, for political reasons, served by hand collection provided at low prices by municipalities. At the fringe, of course, some business/industrial customers have sufficiently little trash to be able to choose hand collection at existing prices while some residential customers, such as large apartment complexes, prefer containerized service. Some business/industrial customers haul their trash themselves, although the record indicates that is not a feasible alternative for many at existing prices.

Customer preferences for service, therefore, turn largely on the quantity of trash produced, and different kinds of equipment are most useful for different quantities of trash. It might be that in a market without municipal haulers the larger private haulers would offer a full line of services, and the relevant product market would be all trash collection. However, the decision of the City of Dallas and other municipalities to provide non-containerized service has narrowed the market available to private haulers. The fact that only a small fringe of customers have a meaningful choice at existing prices between hand collection, containerized service, or self-hauling demonstrates that the bulk of business/industrial customers will view the alternatives as economically feasible only when confronted with substantial increases in the cost of containerized service. We regard this as legally sufficient to support Judge Griesa's view of which existing firms presently compete with WMI.^[3]

Nor is there support for WMI's claim that the power of Dallas or other municipalities to provide containerized service is a constraint on the prices charged by private firms serving the market so defined. Such an expansion of services would be essentially political since the City of Dallas is not a profit-seeking entrepreneur, and there is no evidence that such a decision is likely absent very substantial price increases by private haulers.

Indeed, for all of the smoke blown by WMI over Judge Griesa's market definition, internal TIDI documents fully support his findings. In October, 1980, TIDI completed a "Budget Questionnaire" that had been sent by Waste Resources. This required TIDI to list its major competitors by name, percentage of market, and kinds of equipment. TIDI listed only private haulers as competitors, and these provided almost exclusively containerized service. The major competitors included therein were five private firms with 83 trucks suitable for containerized service and two trucks suitable for hand collection. This questionnaire amply documents Judge Griesa's conclusions as to the product market as defined in terms of existing competitors.

The district court found the relevant geographic market to be Dallas County, plus a small fringe, but not Tarrant County, of which the city of Fort Worth is part. WMI argued below, and continues to argue here, that the relevant geographic market consists of the entire Dallas-Fort Worth metropolitan area. We disagree. Dallas and Fort Worth constitute a Single Standard Metropolitan Statistical Area based on the Office of Management and Budget's judgments regarding economic and social integration. Nevertheless, the bulk of existing Dallas and Fort Worth haulers presently operate exclusively in their respective cities. Only a few, situated near the border between Dallas and Tarrant Counties, do business in both. The area between the two cities is not heavily urbanized and some of the towns between them have granted long term exclusive franchises to waste collectors. Since the travel times between the more heavily populated cities is 45 to 50 minutes, daily service between both areas by the same trucks would be costly. Existing firms that presently compete with WMI are thus found for the most part only in Dallas County, a finding consistent with the TIDI Budget Questionnaire.

The district court found that TIDI's and ACS's combined share of the market so defined was 48.8%. This finding was based on a comparison of revenues of the various haulers within this market, and it is to that evidence that we now turn.

In the pre-acquisition Budget Questionnaire filled out by TIDI, it listed its major competitors and gave their approximate market shares. TIDI's responses and Judge Griesa's findings are compared as follows:

 TIDI's Estimate Judge Griesa's
 of Approximate Finding of % of
Name % of Market Market
ACS ............... 26 22.5
Moore Ind.......... 15 11.1
BFI ............... 10 7.2
AIDS .............. 5 3.4
S.C.A.............. 4 2.5

TIDI calculated its own market share as 40-45%, as compared with the 26.3% found by Judge Griesa. The difference between the TIDI estimates and Judge Griesa's findings are accounted for by Judge Griesa's inclusion of municipal revenues from business/industrial waste disposal.

WMI's principal challenge to these findings focuses upon the admission into evidence of two revenue surveys of various firms prepared respectively by an expert retained by the government and by paralegal employees of the Justice Department. However, so far as the surveys relate to the revenues of the largest private haulers, they are fully corroborated by other testimony or exhibits, and we find no contrary evidence in the record. To the extent the surveys relate to the revenues of smaller private haulers or of municipalities serving business/industrial customers, their exclusion would not significantly affect Judge Griesa's findings as to market share. Accordingly, if error was committed, it was harmless. We have reviewed other evidentiary issues raised by WMI but have concluded that they too are, at best, harmless error.

A post-merger market share of 48.8% is sufficient to establish prima facie illegality under United States v. Philadelphia National Bank, 374 U.S. 321, 83 S.Ct. 1715, 10 L.Ed.2d 915 (1963), and its progeny. That decision held that large market shares are a convenient proxy for appraising the danger of monopoly power resulting from a horizontal merger. Id. at 362-63, 83 S.Ct. at 1740-41. Under its rationale, a merger resulting in a large market share is presumptively illegal, rebuttable only by a demonstration that the merger will not have anticompetitive effects. Id. at 363, 83 S.Ct. at 1741. Thus in United States v. General Dynamics Corp., 415 U.S. 486, 94 S.Ct. 1186, 39 L.Ed.2d 530 (1974), the Court upheld a merger of two leading coal producers because substantially all of the production of one firm was tied up in long-term contracts and its reserves were insubstantial. Since that firm's future ability to compete was negligible, the Court reasoned that its disappearance as an independent competitor could not affect the market.

WMI does not claim that 48.8% is too small a share to trigger the Philadelphia National Bank presumption. Rather, it argues that the presumption is rebutted by the fact that competitors can enter the Dallas waste hauling market with such ease that the finding of a 48.8% market share does not accurately reflect market power. WMI argues that it is unable to raise prices over the competitive level because new firms would quickly enter the market and undercut them.

In discussing ease of entry, Judge Griesa stated:

Defendants in the present case urge that even if the Government has made a prima facie showing on the basis of market share statistics, there are circumstances which undermine the significance of these statistics. Among the arguments in this regard are:

(1) It is so easy to enter the trash collection market that the relative competitive strength of a company cannot properly be measured solely with respect to the existing companies in the market, but must take into account potential new entrants.(2) Over the past several years there has been a trend toward deconcentration, involving steady entry into the market by new companies.

* * *

* * *

[T]he Court agrees that entry into the trash collection business is relatively easy, and the barriers to entry not great. A person wanting to start in the trash collection business can acquire a truck, a few containers, drive the truck himself, and operate out of his home. A great deal depends on the individual's personal initiative, and whether he has the desire and energy to perform a high quality of service. If he measures up well by these standards, he can compete successfully with any other company for a portion of the trade, even though a small portion. If he does not measure up, he is less successful or fails.

Over the last 10 years or so a number of companies have started in the commercial trash collection business, performing containerized service. A few, including TIDI and ACS, have grown to substantial size, presumably because of good management and service, and also as a result of acquiring other companies. The majority of new entrants have either remained relatively small or disappeared as independent entities by going out of business or being acquired by larger companies.

There is no showing of any circumstance, related to ease of entry or the trend of the business, which promises in and of itself to materially erode the competitive strength of TIDI and ACS. With regard to the legal effect of low entry barriers and potential competition in a Sec. 7 case, there is no persuasive authority for allowing such factors to overcome a strong prima facie showing of concentration in the existing competitive structure.

The Supreme Court has never directly held that ease of entry may rebut a showing of prima facie illegality under Philadelphia National Bank. However, on several occasions it has held that appraisal of the impact of a proposed merger upon competition must take into account potential competition from firms not presently active in the relevant product and geographic markets. United States v. Falstaff Brewing Corp., 410 U.S. 526, 93 S.Ct. 1096, 35 L.Ed.2d 475 (1973); Federal Trade Commission v. Procter & Gamble Co., 386 U.S. 568, 87 S.Ct. 1224, 18 L.Ed.2d 303 (1967); United States v. Penn-Olin Chemical Co., 378 U.S. 158, 84 S.Ct. 1710, 12 L.Ed.2d 775 (1964). Falstaff, for example, involved a large national brewer that did no business in the northeast section of the country until it acquired Narrangansett, the largest brewer in New England. Reversing a district court finding that Falstaff was not likely to enter the New England market de novo, the Supreme Court concluded that the very existence of Falstaff as a potential entrant might constrain brewers in the northeast to maintain competitive price levels as a means of forestalling entry by Falstaff. Under that decision, therefore, potential entrants must be considered in appraising a merger.

Moreover, under General Dynamics, a substantial existing market share is insufficient to void a merger where that share is misleading as to actual future competitive effect. In that case, long-term contracts and declining reserves negated the inference of market power drawn from the existing market share. In the present case, a market definition artificially restricted to existing firms competing at one moment may yield market share statistics that are not an accurate proxy for market power when substantial potential competition able to respond quickly to price increases exists.

Finally, the Merger Guidelines issued by the government itself not only recognize the economic principle that ease of entry is relevant to appraising the impact upon competition of a merger but also state that it may override all other factors. Where entry is "so easy that existing competitors could not succeed in raising prices for any significant period of time," the government has announced that it will usually not challenge a merger. United States Department of Justice 1984 Merger Guidelines, 46 Antitrust & Trade Reg.Rep. (BNA) No. 1169 Spec.Supp. Sec. 3.3, at S-6. If the Department of Justice routinely considers ease of entry as relevant to determining the competitive impact of a merger, it may not argue to a court addressing the same issue that ease of entry is irrelevant. We conclude, therefore, that entry by potential competitors may be considered in appraising whether a merger will "substantially lessen competition."^[4]

Turning to the evidence in this case, we believe that entry into the relevant product and geographic market by new firms or by existing firms in the Fort Worth area is so easy that any anti-competitive impact of the merger before us would be eliminated more quickly by such competition than by litigation. See R. Bork, The Antitrust Paradox 222 (1978). Judge Griesa specifically found that individuals operating out of their homes can acquire trucks and some containers and compete successfully "with any other company." The government's response to this factual finding is largely to the effect that economies of scale are more important than Judge Griesa believed. As with his other findings of fact, however, this one is not clearly erroneous, as there are examples in the record of such entrepreneurs entering and prospering.

In any event, entry by larger companies is also relatively easy. At existing prices most Fort Worth and Dallas haulers operate within their own cities, but it is clear from the record that Fort Worth haulers could easily establish themselves in Dallas if the price of trash collection rose above the competitive level. Although it may be true that daily travel from Fort Worth to Dallas and back is costly, there is no barrier to Fort Worth haulers' acquiring garage facilities in Dallas permitting them to station some of their trucks there permanently or for portions of each week. The risks of such a strategy are low since substantial business can be assured through bidding on contracts even before such garage facilities are acquired, as one Fort Worth firm demonstrated by winning such a contract and then opening a facility in a Dallas suburb. That example can hardly be ignored by WMI or other Dallas haulers (not to mention their customers) in arriving at contract bids. The existence of haulers in Fort Worth, therefore, constrains prices charged by Dallas haulers, much as Falstaff constrained pricing by northeast breweries.

The fact that such entry has not happened more frequently reflects only the existence of competitive, entry-forestalling prices, as contemplated in Falstaff. 410 U.S. at 532-537, 93 S.Ct. at 1100-1103. We have no doubt that, if WMI were now to seek to acquire a major hauler in Fort Worth, the Falstaff decision would compel us to treat the Fort Worth hauler as a potential competitor of WMI. We perceive no reason why Fort Worth haulers should be treated as potential competitors in those circumstances but not in the present case.

Judge Griesa's conclusion that "there is no showing of any circumstances, related to ease of entry or the trend of the business, which promises in and of itself to materially erode the competitive strength of TIDI and ACS" is consistent with our decision. TIDI and ACS may well retain their present market share. However, in view of the findings as to ease of entry, that share can be retained only by competitive pricing. Ease of entry constrains not only WMI, but every firm in the market. Should WMI attempt to exercise market power by raising prices, none of its smaller competitors would be able to follow the price increases because of the ease with which new competitors would appear. WMI would then face lower prices charged by all existing competitors as well as entry by new ones, a condition fatal to its economic prospects if not rectified.

The government argues that consumers may prefer WMI's services, even at a higher price, over those of a new entrant because of its "proven track record." We fail to see how the existence of good will achieved through effective service is an impediment to, rather than the natural result of, competition. The government also argues that existing contracts bind most customers to a particular hauler and thereby prevent new entrants from acquiring business. If so, they also prevent the price increases until new entrants can submit competitive bids.

Given Judge Griesa's factual findings, we conclude that the 48.8% market share attributed to WMI does not accurately reflect future market power. Since that power is in fact insubstantial, the merger does not, therefore, substantially lessen competition in the relevant market and does not violate Section 7.

Reversed.