MEMORANDUM AND ORDER

NATHANIEL M. GORTON, U.S. District Judge.

I. Background

Pending before this Court is the motion by plaintiff, American Honda Motor Co., Inc. ("Honda"), for summary judgment against the defendant, Richard Lundgren, Inc. ("Lundgren"), a Honda dealership (Docket No. 72). On December 7, 2001, this Court entered an Order allowing Honda's motion for summary judgment against Lundgren with respect to Count II of Lundgren's counterclaims alleging retaliation under M.G.L. c. 93B § 4(1) (Docket No. 96). The Court took under advisement the remainder of the plaintiff's motion f or summary judgment in order to hear expert testimony (which it did at a hearing on December 18, 2001) regarding how a dealership's relevant market area should be measured under M.G.L. c. 93B § 4(3)(1). The only remaining issue for the Court is, therefore, whether Honda's proposed new dealership in Westborough, Massachusetts ("the add point") is located within Lundgren's relevant market area thereby giving Lundgren standing to protest the establishment of the new dealership.

Honda seeks to establish a new Honda dealership in Westborough, approximately 11 miles east of Lundgren's dealership in Auburn, Massachusetts. In March of 1998, Honda notified Lundgren of its intent to award the new dealership to Mark P. Ragsdale, George M. Ragsdale and Robert J. Avolizi. Lundgren duly notified Honda of its intent to protest the award of the new dealership pursuant to M.G.L. c. 93B § 4(3)(1).

On April 21, 1998, Honda filed the instant suit seeking a declaratory judgment that Lundgren does not have standing to protest the appointment of the new dealership or, in the alternative, that the proposed establishment of a new dealership is lawful and proper. Lundgren counterclaimed alleging retaliation and violation of M.G.L. c. 93B § 4(3)(1) and seeking a permanent injunction against the opening of the new dealership.

Until this point in the litigation, the parties have disputed whether Lundgren has standing to protest Honda's establishment of a new dealership at the add point. Initially, the dispute about standing concerned the shape of the relevant market area as that term is defined in the statute more fully discussed in the "Analysis" section of this Memorandum and Order below. Honda argued that the controlling statute requires a dealership's relevant market area to be defined as a circle with the existing dealership at the center while Lundgren responded that the statute allows for an irregular, polygonal shape to define the relevant market area. For the first two years of the litigation, there was little dispute regarding the methodology to be employed in computing and platting the relevant market area. Such is no longer the case.

The First Circuit Court of Appeals certified the issue of the shape of the relevant market area to the Supreme Judicial Court of Massachusetts ("SJC") which ultimately declared that, under Chapter 93B § 4(3) (1), a relevant market area is "a geographic area circular in shape, although not a perfect circle, and contiguous to the existing dealer's location." American Honda Motor Co., Inc. v. Bernardi's Inc., et al., 432 Mass. 425, 427 (2000). The Court went on to explain that:

A "circular" shape may differ from the shape of a perfect "circle." A "circle" is a "closed plane curve every point of which is equidistant from a fixed point within the curve," . . . [whereas] a "circular" shape has "the exact or approximate form or outline of a circle: ROUND."

Although the parties now must agree on the shape of the relevant market area, they now disagree with respect to the definition of "service sales" as that term is used in the controlling statute and the methodology to be employed in determining the relevant market area.

II. Analysis

Massachusetts General Laws Chapter 93B was enacted to prevent unfair methods of competition and unfair or deceptive acts among motor vehicle manufacturers, distributors and dealers. American Honda Motor Co., Inc., 432 Mass. at 427. Specifically, it prohibits, inter alia, the improper granting of

a franchise or selling agreement to or with an additional franchisee who intends or would be required by such franchise or selling agreement to conduct its dealership operations from a place of business situated within the relevant market area of an existing franchisee or franchisees representing the same line make.

M.G.L. c. 93B § 4(3)(1) (1977).

An existing dealer has standing to contest the establishment of a new dealership if the add point is within the existing dealer's relevant market area. American Honda Motor Co., Inc., 432 Mass. at 428. Prior to the 1977 amendments to Chapter 93B, the statute required courts to determine a motor vehicle dealer's relevant market area using equitable principles. Ricky Smith Pontiac. Inc. v. Subaru of New England. Inc., et al., 14 Mass. App. Ct. 396, 403 (1982). Equitable principles included such factors as demographics, access to public transportation and brand popularity. Id. at 410-11.

In 1977, the Legislature amended the statute and substituted a formula definition of "relevant market area" for the equitable principles. The goal of the Legislature was to construct a bright-line test for determining a motor vehicle dealer's relevant market area. Id. at 414. The amended statute defines "relevant market area" as:

the more narrowly defined and circumscribed geographical area immediately surrounding its existing dealer location within which it obtained, during the period of time the dealership business has been operated from said location or the three-year period immediately preceding the date of said notice of intent to grant or enter into an additional franchise or selling agreement, whichever is the lesser, at least two-thirds of (i) its retail sales of new motor vehicles of said line make or (ii) its retail service sales . . .

M.G.L. c. 93B § 4(3)(1) as amended.

To determine whether a dealer has standing to protest the establishment of a new dealership, the court analyzes hard statistical data regarding dealership sales and after the relevant market area is platted using that data, simply decides whether the add point falls within or without the existing dealer's relevant market area. Ricky Smith, 14 Mass. App. Ct. at 415. Although the Legislature intended to provide courts with an easily-applied bright line test, courts have fundamentally disagreed with respect to the shape of the relevant market area, the definition of sales and the expert methodologies.

The SJC has now determined and defined the shape of the relevant market area but, in this Court's humble opinion, has left more confusion than enlightenment. That is because its definition of relevant market area is inconsistent with the bright-line test enunciated by the Legislature and impracticable to apply. By defining the relevant market area as a circular shape, as opposed to an exact circle, there is no way to determine the relevant market area on the basis of anything other than subjective or equitable principles. If the trial court is to substitute an "approximate circle" for an "exact circle" on what basis is it to so approximate? If the area is not a circle then for what reason is an alternative shape to be chosen? Faced with this conundrum, we nevertheless labor on to apply the law as announced.

In this case, the plaintiff and defendant utilize different data and methodologies to arrive at two diametrically opposed conclusions. Not surprisingly, Honda's expert concludes that the add point falls just outside Lundgren's relevant market area, while Lundgren's expert finds that it falls on the line forming the outer boundary of Lundgren's relevant market area.

The parties agree that "retail service sales" form the more narrowly defined geographical area immediately surrounding the Lundgren dealership but they disagree with respect to the definition of "service sales." Lundgren contends that the total number of service customers, the total number of repair orders and the total dollar value of repairs are all valid indicia of service sales. It further asserts that if either the total number of customers or the total number of service dollars is used to determine service sales, the add point falls within Lundgren's relevant market area using a method called "geocoding."

Honda, on the other hand, argues that the number of service customers and the dollar volume of sales are not reasonable definitions of "service sales" and that geocoding is not a reliable method for computing and platting the data. It is undisputed that if service sales is defined by the number of repair orders (as opposed to the dollar volume or number of customers), the add point falls outside of Lundgren's relevant market area.

The statute is silent both with respect to how the term "service sales" should be defined or measured and what methodologies are acceptable. Nor did the SJC opine on those subjects. The expert reports and testimony in this case, however, support a finding that both dollar volume and number of repair orders are acceptable definitions of service sales. Massachusetts trial courts have considered both definitions of service sales when determining a motor vehicle dealer's relevant market area. See Subaru of New England, Inc. v. Subaru of Wakefield, Inc., 1999 Mass. Super. LEXIS 564, *13-14;Richard Lundgren, Inc. v. American Honda Motor Co., Inc., 1994 Mass. Super. LEXIS 497, *10.

The number of service customers, on the other hand, is not a reasonable definition of service sales. This Court is unable to find a single case in Massachusetts that considers such a measurement to be a valid definition of retail service sales. Moreover, the plain language of § 4(3)(1) does not support the use of such a definition. The statute expressly provides that relevant market area is to be measured by "sales," not "customers."

The use of service customers as opposed to service sales also results in an inaccurate relevant market area in this case. The data used by the experts consisted of customer addresses, repair order numbers, and the dollar amount of repairs. Defining service sales by the number of customers would result in the omission of service sales because a given customer may own two Honda vehicles that have been serviced or have obtained service for the same vehicle on more than one occasion.

The lack of guidance from any source with respect to methodologies to be used in determining the relevant market area is exemplified by not less than six expert reports that have been filed in this case espousing various methods for computing and platting a dealership's retail service sales. The SJC, in its response to the certified questions submitted to it by the First Circuit Court of Appeals, declined to address the issue:

[W]e decline to enunciate a particular methodology. . . As illustrated by the parties' experts, several methods may be employed. Significantly, the statute does not indicate how two-thirds of new vehicle sales or service sales should be computed and platted. We conclude, depending on the methodology employed, the method of computing and platting two-thirds of a dealer's new vehicle sales or service sales, is one more appropriately left for expert testimony.

In its most recent expert report, Lundgren's expert, Dr. Ernest Manuel, concludes that the add point falls within its relevant market area of retail service sales measured by the dollar volume of those sales using a method of platting the data called "geocoding."

Honda's expert responds, to the contrary, that 1) the method of "geocoding" is not sufficiently developed to be an acceptable method for determining a dealer's relevant market area under the statute and 2) a method called "zip code distribution" is more accurate and reliable.

Using the method of geocoding, an expert takes the information from each service sale, consisting of the customer's address, repair order number and dollars spent, and inputs it into a software program that sorts the data geographically by producing a latitude and longitude for each repair order so that it can be plotted on a map. In this case, the software was able to geocode 89.4% of the service sales data, i.e. that percentage of the addresses were positively and exactly located. The remaining faulty data were plotted at the geographic center of their respective zip codes.

Under the zip code distribution method, the dealership's service sales are summarized and sorted according to zip code. Then the sales within each zip code are distributed randomly as dots on a map and the closest two-thirds of the service sales immediately surrounding the dealership in question are circumscribed to constitute its relevant market area.

It is difficult to comprehend how a method that distributes all of the service sales randomly is more reliable than a method that distributes most of the service sales precisely. Upon consideration of the expert testimony and reports, this Court finds that geocoding sales data is an acceptable and the preferable method for determining relevant market area under the statute. See Subaru of New England, Inc. v. Subaru of Wakefield, Inc., 1999 Mass. Super. LEXIS 564, *13 (rejecting zip code analysis in favor of geocoding).

Even if we accept Lundgren's definition of service sales and use its methodology, it has failed to sustain its burden of proof on the issue of standing. See Lundgren v. American Honda Motor Co., Inc., 1994 Mass. Super. LEXIS 497, *8, vacated on other grounds by 45 Mass. App. Ct. 410 (1998) (stating that the dealer bears the burden of proving that it has standing to protest the establishment of a new motor vehicle dealership).

Dr. Manuel calculated the dollar value of sales for each repair order and determined, by use of the sequence of geocoded repair orders, the radius of a circle (with the dealership at the center) which would circumscribe two-thirds of the total repair order dollars. He then compared the length of that radius to the distance between Lundgren's and the add point and found that the two distances were the same if he rounded to the nearest onetenth of a mile. Both distances were, however, measured from the outer property line of the Lundgren dealership.

Dr. Manuel repeated his analysis using drive time rather than linear distance in order to account for existing topography. He again concluded that the proposed new dealership falls within Lundgren's relevant market area.

There are four problems with the analysis of Lundgren's expert:

1. He misconstrues the SJC's opinion as expressly authorizing him to take into account topography in computing the geographical area within which two-thirds of the dealer's total repair order dollars is contained. To Lundgren's misfortune, the SJC did no such thing.

The SJC referred to topography in the process of rejecting a construction of the definition of a dealer's relevant market area that results in a perfect circle. 432 Mass. at 431. It ominously warned that any formulation of relevant market area will result in factual disputes and that even if a "perfect circle" were used, experts would dispute whether its radius "follows existing topography." Id. at 432. Thus, the SJC mentioned topography as an example of a dispute it sought to avoid in its formula for determining relevant market area. Adding the equitable consideration of topography to the analysis would simply undermine the "bright line test" which was said to be the purpose of the amendment to the statute. Id.

2. By measuring distances in terms of radii in its most recent expert report, Lundgren analyzes its relevant market area in terms of a circle. In earlier rounds of this litigation, Lundgren conceded that if the relevant market area is defined as a circle, it would not have standing. Notwithstanding that concession made after this Court defined "relevant market area" as a circle rather than a circular shape, Lundgren's expert now uses a radius to circumscribe its relevant market area thereby presupposing a circle and not a circular shape. Its new position is fatally inconsistent with its old one.

Lundgren attempts to avoid the inconsistency by a) stating that it now has software that was unavailable prior to the SJC's opinion and b) analyzing service sales using the number of service customers and repair order dollars rather than the number of repair orders. If Lundgren considered such definitions of "service sales" to be reasonable, it is logical to assume that it would have conducted an analysis of service sales using repair order dollars and number of customers after this Court first defined relevant market area as a circle rather than conceding its lack of standing. Lundgren's change of heart now takes on the appearance of equivocation.

3. Lundgren measures the distance from its dealership to the add point and the radius of the circle intended to circumscribe two-thirds of its total repair order dollars from the outermost boundary of its property rather than from the center point of the dealership. Measuring the distance in such a way is, again, inconsistent with the SJC's opinion which requires that:

the circular boundary be drawn around the smallest geographical area . . . closest to the existing dealership, with the existing dealership in the center of the circular boundary.

4. Finally, Dr. Manuel rounds off the distance between Lundgren and the add point in order to incorporate the add point within its relevant market area. Even after rounding, the new dealership apparently falls on the line forming the circumference of the circle and not within the area. The defendant's expert arbitrarily rounds off the distance to include the add point because it advantages his client. Such an approximation is insufficient to support a finding by a preponderance of the evidence that the point falls within the defendant's relevant market area.

ORDER

For the reasons set forth in the Memorandum above, the plaintiff's motion for summary judgment (Docket No. 72 in Case No. 98-10690 and Docket No. 37 in Case No. 98-40061) is ALLOWED.

So ordered.