S.C. Johnson & Son, Inc. v. Clorox Co.

930 F. Supp. 753 (1996)

S.C. JOHNSON & SON, INC., Plaintiff,
v.
The CLOROX COMPANY, Defendant.

No. 95-CV-2803 (JS).

United States District Court, E.D. New York.

May 9, 1996. Order Denying Reconsideration July 18, 1996.

Lawrence I. Weinstein, Lisa M. Gigliotti, James W. Kennedy, Paul, Hastings, Janofsky & Walker, New York City, for Plaintiff.

Kim J. Landsman, Mary E. Mulligan, Erik Haas, Patterson, Belknap, Webb & Tyler L.L.P., New York City, for Defendant.

OPINION AND ORDER

SEYBERT, District Judge:

This is an action for false advertising, pursuant to § 43(a)(1)(B) of the Lanham Act, 15 U.S.C. § 1125(a)(1)(B), in which S.C. Johnson & Son, Inc. ("SCJ") seeks a preliminary injunction restraining The Clorox Company ("Clorox") from making certain challenged advertising claims. These claims, which concern the effectiveness of two competing household roach control products — SCJ's RAID Max IV Plus Egg Stoppers® ("RAID Max Plus") and Clorox's COMBAT Super-Bait® roach baits ("SuperBait") — were presented last summer in a 30-second television commercial entitled "Weapon" (the "Commercial") that Clorox intends to resume broadcasting in the near future. A preliminary injunction hearing was conducted during March 1996 that included nine days of live testimony, and the introduction into evidence of over 300 exhibits, and approximately 800 pages of deposition excerpts.

Were it the determinative issue in this litigation, the Court would have little difficulty in concluding that SuperBait is a more effective product than RAID Max Plus. In this regard, Clorox's new testing protocol impresses the Court as a particularly effective measure of the relative efficacy of roach bait products for purposes of product development, insofar as it successfully gauges the products' strengths and weaknesses free of the encumbrances that traditional testing techniques had encountered in attempting to conform testing methods to expected consumer experience, assuming consumer compliance with the product label instructions.

The Commercial, however, does not confine its sales pitch to general statements of product superiority that the evidence amply would support, or for that matter, to a non-quantitative claim that testing proves that SuperBait kills more roaches than RAID Max Plus. Rather, the Commercial states, in virtually unqualified terms, that "testing proves Combat SuperBait kills up to 98%" of consumers' roaches, while RAID Max Plus kills "no more than 60%." These claims, known in the legal vernacular as "establishment claims," therefore seek through advertisement to superimpose the results achieved in the sanctuary created by Clorox's new testing protocol, and in selected laboratory tests, into the variable-ridden world of consumers' homes. Because the Court finds that plaintiff has met its burden of proving that defendant's tests — when applied to the untoward purpose of supporting quantitative declarations of product efficacy in home use — fail to substantiate the Commercial's unqualified propositions with respect to each of the products at issue, plaintiff's motion to enjoin the Commercial is granted.

FINDINGS OF FACT

A. Introduction

1. In the instant action brought pursuant to § 43(a)(1)(B) of the Lanham Act, 15 U.S.C. § 1125(a)(1)(B), plaintiff SCJ seeks a preliminary injunction enjoining defendant Clorox from making certain challenged advertising claims. These claims, which concern the effectiveness of two competing household roach control products — SCJ's RAID Max Plus and Clorox's SuperBait — were presented last summer in a 30-second television commercial entitled "Weapon" (the "Commercial").

2. The Commercial begins with a woman in her kitchen looking at cockroaches. The voice of a male (who later appears on screen as a scientist dressed in a white laboratory coat) tells consumers that "It's you against thousands of them. Choose your weapon wisely." The woman holds up a single package of RAID Max Plus (see Answer ¶ 9), while the scientist declares: "This, only gets rid of some of your roaches." The woman then holds up a single package of SuperBait, while the scientist says: "This, kills just about all of them. And their eggs." The scientist then appears on screen dressed in laboratory attire and states, "In fact, testing proves COMBAT SuperBait kills up to 98%. The other guys [i.e., RAID Max Plus], no more than 60%." When the scientist refers to the SuperBait results, a single package of SuperBait appears on the screen, above which appears the caption "KILLS UP TO 98%." Similarly, when the scientist refers to "the other guys," a single package of RAID Max Plus appears on the screen, together with the caption, "KILLS NO MORE THAN 60%." PX 1, 2.

3. The Commercial began airing on June 26, 1995 (Answer ¶ 1), and it appeared on television stations in the New York metropolitan area and across the United States throughout last summer. SCJ commenced this action on July 13, 1995. On July 20, 1995, this Court denied SCJ's application for a temporary restraining order but granted its request for expedited discovery in advance of a preliminary injunction hearing. The Court conducted an evidentiary hearing on SCJ's preliminary injunction motion over the course of nine days between March 11 and 27, 1996. Unless enjoined, Clorox intends to resume broadcasting the Commercial in the near future. Clorox Pre-Hearing Mem. 1.

B. The Products Featured In The Commercial

4. Both RAID Max Plus and SuperBait are over-the-counter ("OTC") products intended for use by consumers in treating infestations of German cockroaches in their homes and apartments. The German cockroach is recognized to be the most prevalent household cockroach pest species in this country (Owens 93/17-24), and all references to "cockroaches" or "roaches" are to the German cockroach.^[1] There are, at a minimum, "millions" of strains of cockroaches — i.e., individual populations — located in households throughout the United States. Cochran 516/25-517/7. While infestations tend to be somewhat more common in Florida and other southern states whose high temperatures and relative humidity are especially conducive to cockroaches, see Owens 119/21-25; PX 89, Table 33 at 15, roach infestations are not limited to any geographical area, and are found in residences throughout the United States. Owens 97/17-98/2. For that reason, roach bait manufacturers advertise their products in locations throughout the United States. See Shapas 751/22-752/11. Moreover, the cockroach is not a respecter of socioeconomic class and is found in all sorts of housing, from private homes of all sorts to low-income housing. Owens 97/17-98/2. Indeed, according to an internal SCJ marketing analysis conducted in the ordinary course of business, 72% of roach bait users live in single family homes and only 28% are apartment dwellers. PX 89, Table 35 at 17.

5. It is not surprising that U.S. consumers spend vast sums — over $300 million in 1995 alone (PX 91) — in an attempt to control roach infestations in their homes. Cockroaches can contaminate food supplies and kitchenware with bacteria and other microorganisms, and are a source of many potent allergens. Owens 96/17-97/16; PX 18 Chap. 4. The principal forms of OTC roach control products are sprays, foggers and baits. Owens 98/5-16. Roach baits, the subject of the Commercial, consist of a disc (also referred to as a bait tray or bait station) that contains a roach food (referred to as a bait "base") mixed with an insecticide. Roaches foraging for food enter the bait station, eat the toxic bait base, crawl away and die. Owens 102/6-103/7; PX 17.

6. The two Clorox roach bait products marketed to consumers, SuperBait (PX 111) and COMBAT Roach Killing System ("Regular COMBAT") (PX 114), each contain the insecticide hydramethylnon. Hydramethylnon is a relatively slow-acting metabolic poison that "inhibits the ability of the cockroach to have energy to move. So it is like a windup toy that eventually runs to a stop." Koehler 1341/14-1342/5. Regular COMBAT was first introduced by American Cyanamid ("Cyanamid") in the mid-1980s. Shapas 597/9-12. Clorox acquired this and other COMBAT products from Cyanamid in the summer of 1990. During the winter of 1993, Clorox began marketing SuperBait. Silverman Dep. 67/7-11. SuperBait contains slightly more hydramethylnon by weight than Regular COMBAT and the food material in SuperBait is designed to appeal to a broader spectrum of roaches; the toxic formulations of these products, however, are identical. Shapas 606/4-25. Clorox's Dr. Shapas acknowledged that the increased amount of hydramethylnon in SuperBait would not likely have any impact on primary kill, i.e., roach deaths caused by eating the bait directly. Id. 753/5-11. And, while Dr. Shapas theorized that the increased amount of hydramethylnon in SuperBait would increase the amount of secondary kill (id.) — i.e., roaches killed by eating the excrement of other roaches poisoned by the baits (Silverman 1157/20-1158/3) — the only Clorox field test in evidence comparing those products reveals a relatively modest difference in efficacy. PX 60; Owens 205/2-15.^[2]

7. RAID Max Plus (PX 16) is actually two products contained in one package. It consists of (a) 12 roach baits and (b) 3 small containers of the insect growth regulator Hydroprene, which sterilizes certain roaches that come in contact with it. Owens 335/9-336/9; Koehler 887/4-888/17; PX 101, 125. The insecticide in the roach bait component of RAID Max Plus is called chlorpyrifos, also known by the brand name Dursban. PX 17. Dursban is a relatively fast-acting "nerve poison ... that acts somewhere between one and five hours after exposure." Koehler 1340/11-20. Dursban "affects the nervous system resulting in excessive nervous firing," and causes the cockroach to "have spasms, flip over on its back," twitch its legs and "ultimately ... die." Id. 1340/21-1341/1.

8. When SCJ initially introduced RAID Max Plus roach baits in 1989, the product contained an insecticide known as sulfluramid. DX 144. In August 1994, SCJ voluntarily recalled this product nationwide because of EPA concerns about the child resistant package design of the product. See PX 92. RAID Max Plus remained off the market until early 1995, when it was relaunched with Dursban as the active ingredient. See PX 23. Even after SCJ began marketing sulfluramid baits in 1989, it continued to market Dursban baits as well, first under the name RAID Roach Controller (PX 12), and since 1990, under the name RAID Roach Bait (PX 13) ("Regular RAID"). The formulations of Regular RAID and RAID Max Plus baits are identical. Koehler 826/19-828/17.

9. Clorox characterizes SCJ's Dursban baits, including RAID Max Plus, as a "terrible product." Clorox Opening Statement 40/2-5. However, although Dursban baits can, in some situations, be compromised where cockroach strains have developed resistance to that insecticide, consumer tests conducted in the ordinary course of SCJ's business to monitor consumer satisfaction with OTC roach baits show that consumers rate RAID Dursban baits on substantial parity to Clorox's baits. E.g., PX 131 at 1075; PX 132 at 392; DX 70 ("Blind label testing of RAID Roach Bait, RAID Max and COMBAT showed that consumers could not perceive any performance differences between the 3 bait products.").

C. The Inconsistent Performance of Roach Baits In The Real World

10. SCJ is by far the leader in total sales of roach control products (Mar. 27, 1996 Stip.; PX 91), which in part reflects the fact that consumers spend nearly twice as much on sprays as on baits. See Mar. 27, 1996 Stip.; PX 91. In the roach bait sub-category, however, Clorox is the clear market leader (57.7% market share in 1995), with SCJ a distant second (18.5% market share in 1995). Mar. 27, 1996 Stip. at 1 ¶ 3.

11. Unlike sprays, which consumers can target directly on roaches they see and on surfaces where they observe roaches moving, roach baits are passive instruments of roach control. Owens 98/8-99/8. Consumers set out the bait stations in their homes, where they compete with other food sources for the attention of roaches that are foraging for food and water. Id. 102/1-13; PX 18 at 246. Thus, roach baits cannot work unless significant numbers of roaches happen to encounter the bait station and ingest a lethal dose of the bait food. Owens 99/3-8; PX 18 at 239. Even roach baits to which no resistance has been demonstrated, such as hydramethylnon, can frequently be rendered ineffective by the inherent limitations in such a passive system of roach control. See Owens 99/3-8. As Clorox's own expert, Dr. Austin Frishman, observed in a recent article in a trade journal for PCOs, "[e]ven the best of baits fail under field conditions in some circumstances." PX 51 at 23.

12. The efficacy of roach baits in the real world is highly variable and often compromised by factors unrelated to the killing power of the poison. For example, sanitation conditions can greatly impair bait performance: "[b]aits work best when availability of competing food sources in the infested areas is minimized. Therefore, the level of control with baits in apartments or homes with poor sanitation will not be as high as in clean apartments." PX 18 at 295; Owens 256/13-16. Bait depletion can impair control as well. Shapas 628/19-629/1, 783/8-14; Boase 1483/14-19. Other relevant variables include the amount of "clutter" (Koehler 824/7-17); the relatively rigid and often narrow foraging patterns of roaches (Owens 132/3-133/12; see PX 18 at 251-53); and the overall impact of human behavior upon all of these conditions. Owens 70/19-71/3; Koehler 824/18-825/15.

13. Another critical factor in achieving effective control with baits is the placement of the baits. Frishman 713/18-714/20; PX 18 at 249. The objective of effective bait placement is that "[t]he baits must be placed between where the cockroaches harbor and where they eat." PX 51, Item 6. However, this is more easily said than done. Roaches generally live in groups in dark locations that are not readily visible to human beings, such as in cracks and crevices in the walls, wall voids, and inside and under furniture, appliances and cabinetry. (These locations are sometimes referred to as harborages). Owens 65/15-23, 103/4-7. Roaches typically search — or "forage" — for food and water at night when lights in the home are off and there is no human activity. PX 18 at 49; Owens 102/9-13; Koehler 1387/12-20. Moreover, most roaches generally have a very limited foraging range, rarely foraging more than a few feet from their harborages. Owens 132/7-14, 296/2-8; see PX 18 at 251, 253.

14. As a result, the key to effective bait placement is not only to place the baits as close as possible to cockroach harborages, but also to locate all of the harborages in the apartment so that a bait can be placed in very close proximity to each harborage. Frishman 708/10-19; PX 18 at 250. Otherwise, the roach baits — no matter how effective the poison — will not kill roaches living in untreated harborages, and the treatment therefore will not come close to eliminating the infestation. Clorox's expert Dr. Frishman has publicly acknowledged this to be true, noting that "[n]o matter how attractive the bait, if the placement is wrong you may kill 50 percent of a cockroach population, but you need a 96 percent-plus kill to put a permanent dent in their numbers." PX 51, Item 6.

15. Although Clorox's experts testified that consumers have occasionally taken them "by the hand" (Frishman 709/21-25) and "by the arm" (Boase 1456/17-20) to show them where the roaches are, the credible evidence demonstrates that most consumers do not and cannot know the location of all of the cockroach harborages in their homes.^[3] First, since roaches harbor in dark areas that are not readily visible to humans and quickly scurry away when lights are turned on (Koehler 1387/3-20), the sight of roaches foraging for food and water at night will, at best, give the homeowner only a vague clue about the locations of possible harborages. Second, in its cross-examination of Dr. Koehler, Clorox took pains to establish that many of the apartments where the parties have conducted field studies have cockroach infestations of 10,000 or more roaches. Koehler 949/12-952/1; DX 43. Depending on the number of cockroaches contained in any harborage, the number of harborages contained in those apartments would range from approximately 30 to 300 or even more. Compare PX 18 at 62-63 (10 to 30 roaches per harborage) with Boase 1597/1-22 (has seen harborages with as many as 200 to 300 roaches). It is extremely unlikely that any consumer would even appreciate the existence of so many cockroach harborages in his or her apartment, much less be able to locate them with precision. Third, the fact that both Clorox and SCJ use diagrams on the labels of their roach bait products to inform consumers of the "most important locations" (PX 111) and the locations which will produce "the best results" (PX 16) is compelling evidence that both companies recognize that many consumers do not know the optimum placement locations for baits in their homes. See also PX 18 at 249-50 (showing a placement diagram from a COMBAT package and noting that "[p]lacement guides have been developed to assist in bait placement.... These guides accompany most bait products, and generally point out places where bait should be installed in order to maximize cockroach contact.").

16. Finally, there has emerged in the U.S. an industry of professional pest control operators ("PCOs") who, according to Dr. Frishman, sell their services to consumers based in part on their superior training and ability to effectively locate cockroach harborages and to properly lay out roach bait stations in consumers' homes. See PX 51. On cross-examination, Dr. Frishman testified that PCOs are often trained in proper bait placement techniques and have at their disposal tools for locating out-of-the-way harborages, such as mirrors and sticky traps, that are not readily available to consumers. Frishman 708/10-19, 710/3-14. Dr. Frishman conceded that even assuming "consumers have a pretty good idea of the majority of locations where roaches are, it is the peripheral areas that take the extra effort," and "it is the peripheral areas, and the value added from the training [which] the PCOs often have, that often can make the difference between eliminating or virtually eliminating the roach infestation on the one hand and getting only moderate or inadequate control on the other [hand]." Frishman 708/20-709/7.

17. Indeed, as Dr. Frishman has recognized, even trained PCOs often fail to locate and properly place baits in close proximity to all of the cockroach harborages. Thus, as Dr. Frishman wrote in an article directed at PCOs (PX 51):

6. You placed the bait in the wrong area. This is easier to do than you might think. You were not aware of how many locations the cockroaches were hiding in when you initiated your baiting program. The baits must be placed between where the cockroaches harbor and where they eat. No matter how attractive the bait, if the placement is wrong you may kill 50 percent of a cockroach population, but you need a 96 percent-plus kill to put a permanent dent in their numbers....

9. The puck [i.e., professional bait station] was placed in the correct zone, but incorrectly placed within the treat zone. You cannot just place a bait any way you want to.... The pucks should be tucked in corners or along edges where cockroaches most likely congregate or walk. You have to think like a cockroach. That's why you are the professional....

12. You missed a major pocket of cockroaches. This serves as a dispersion point to other areas. The sticky traps can help you find them.^[4] Keep in mind that you are trying to control living organisms. It is not a simple mathematical formula that works every time. You sometimes have to adjust what you did to get the baits to do the job. I call it "tweaking the baits to perfection."

14. You missed the zone by one inch! What? How can that be? Simple. The air currents are moving the bait odors in the opposite direction from where the roaches are.

PX 51.

18. When one steps back from the Clorox's advocacy and looks at how both PCOs and respected academic entomologists view the utility of baits, one gets a very different picture than the one painted by the Commercial. Donald Reierson — a widely respected entomologist at the University of California — Riverside (see Shapas Dep. 231/4-17), who has substantial experience with hydramethylnon (PX 97; DX 56, 131) — recently wrote that except "under optimal conditions ... such as might ... occur in a vacant apartment," baits do not "lead to the [roach] population being eradicated," that while "[b]ait usually reduces the number of cockroaches present ... other measures are needed to keep the population suppressed," PX 18 at 259, and that "[e]xcept where conditions for control of [German cockroaches] with bait are ideal, one should have modest expectations for control with current baits. Greater, and more acceptable levels of control will generally require supplementation of bait use with other control methods." PX 18 at 265.

19. Similarly, since the livelihood of PCOs depends on their ability to effectively treat infestations, and since, unlike sprays and dusts, "[i]n this age of increasing chemical awareness, baits are considered ... especially safe ... [and are] easy to use [and] virtually odorless" (PX 18 at 234), PCOs would have a powerful incentive to use baits exclusively if they provided the type of control the Commercial promises. In fact, however, as Dr. Frishman admits, in many areas of the country as few as 10% of PCOs use baits exclusively (Frishman 693/20-694/10), notwithstanding that Clorox's Maxforce baits, which are identical to SuperBait with respect to their active ingredient (id. 660/21-25), have been available to PCOs for nearly a decade. See PX 66 at 2309; DX 53 at 35; DX 73 at 1280-81. By contrast, Dursban sprays remain today one of the most popular insecticides among PCOs. Cochran 524/3-12, 542/11-19; Frishman Dep. 55/5-20; Boase 1644/17-1646/22. In like manner, despite the prevalence of Clorox's advertising, fewer than 20% of consumers with roach problems use baits exclusively. PX 89 at 37.

20. The clinical evidence confirms that when the baits are used in a manner consistent with label directions, SuperBait rarely even approaches the level of 98% control promised in the Commercial. For example, although Clorox reported that SuperBait reduced roach populations on average by 84.4% in one such test (PX 61), in other Clorox field tests, SuperBait achieved average roach population reductions of only 29.9%, 34.1%, and 44.5% (PX 60, 61). The methodology used in these tests, and the changes in methodology employed in Clorox's new protocol, are discussed at ¶¶ 61-67 infra.

D. The Original Version of the Commercial

21. In the Spring of 1994, Clorox, having become dissatisfied with the results of testing under its existing field test method (Shapas Dep. 136/19-137/12; Bieman 1013/2-12), developed a new field test protocol (Bieman 1015/11-15, 1080/8-1081/5; PX 42), and tried it out for the first time. Bieman 1026/7-16. The products tested were SuperBait and SCJ's then-current RAID Max Plus product, which contained sulfluramid. Id. 1026/17-18. The test sites were two low income public housing complexes in Miami, Florida (the "Miami Field Test"). DX 25, 26. Based solely on the results of the Miami Field Test, Clorox began developing a commercial that was to have stated that "SuperBait kills 98%, RAID Max only 55%." See PX 19. Clorox's Claim Index for this version of the commercial (id.) — a document that identifies what scientific support Clorox has for each of its advertising claims (Ochomogo Dep. 78/7-22) — states that the Miami Field Test was intended to be the substantiation for that proposed commercial's comparative claims. See PX 19.

22. In early 1995, apparently before the 98% versus 55% version of the commercial was ever filmed, Clorox learned that RAID Max Plus would contain Dursban in 1995 instead of sulfluramid (Shapas 740/5-9; PX 23). Clorox quickly appreciated that the change in formulation rendered any comparative claims based on its field test of the sulfluramid product invalid, and so informed its advertising agency, Young & Rubicam ("Y & R"). Thus, in a March 10, 1995 memo (PX 23), Mark Adams of Y & R stated that "Clorox informed Y & R that Raid Max Plus EggStoppers has been reformulated with the active ingredient Dursban — making our `98% versus 55%' claim invalid." Id. (emphasis added).

23. Having planned to shoot the Commercial on April 17, 1995 based on the Miami field testing results, Clorox and its advertising agency were in a bind. As Y & R later observed in a memorandum dated April 6, 1995, "it is important to realize that we have financially committed to our shoot date without proper written substantiation for the majority of competitive and superiority claims made in these [story]boards." PX 24 (emphasis in original). Thus, in April 1995, Clorox conducted an in-house laboratory test (Koehler 811/16-812/13; PX 24, 35, 36) (the "First Laboratory Test") comparing SuperBait, RAID Max Plus with Dursban, and two other RAID formulations, including Regular RAID.^[5] Clorox conducted the First Laboratory Test in tiny arenas the size of a sweater-box. Koehler 813/16-815/16; see PX 31. Each product was tested three times (PX 28 at 3975; Koehler 818/16-19); each time, twenty cockroaches were placed in each arena along with one bait and alternative sources of roach food. PX 28 at 3974; Koehler 816/9-24. The results of the First Laboratory Test reflect the percentage of cockroaches that died as the result of a 20-day exposure to a single bait. PX 28 at 3975; Koehler 816/21-24. In the First Laboratory Test, by the end of 20 days, Clorox reported that SuperBait had killed 58 of 60 roaches, or 98%, while RAID Max Plus killed 36 of 60 roaches, or 60%. PX 36; Koehler 819/15-20. Regular RAID killed 48 of 60 roaches, or 80%, in that same test. PX 36; Koehler 829/5-20.

24. The First Laboratory Test was the only "testing" used to substantiate the Commercial at the time that the storyboard containing the "98% to 60% claim" was formulated in May 1995. Specifically, the Claim Index for the Commercial (PX 22) unambiguously identifies the First Laboratory Test as the sole support for each of the advertising claims in the Commercial. Similarly, a May 18, 1995 Y & R document stated that "[w]e are supporting this claim with the attached substantiation only." PX 28 at 3969. The claim is identified as "In fact, testing proves Combat SuperBait kills up to 98%. The other guys, no more than 60%." Id. The only substantiation attached was a May 10, 1995 memo from Dr. Ochomogo — the in-house Clorox scientist in charge of coordinating advertising claim substantiation (Ochomogo Dep. 48/6-10) — summarizing the method and results of the First Laboratory Test. PX 28 at 3974-76. In deposition testimony, Dr. Ochomogo likewise conceded that the substantiation for the Commercial was a "lab test comparing Combat SuperBait [with] RAID Max Plus." Ochomogo Dep. 48/21-49/10.

E. The Second Laboratory Test

25. Clorox thereafter conducted a second laboratory test (the "Second Laboratory Test") comparing SuperBait and RAID Max Plus with Dursban. PX 39; Silverman 1184/22-1185/7. The Second Laboratory Test was completed prior to June 7, 1995, more than two weeks before the Commercial was first aired. PX 39; Answer ¶ 1. Dr. Jules Silverman, the Clorox scientist who directed that the Second Laboratory Test be performed (Silverman 1184/22-1185/12), testified that he was satisfied with the methodology for the Second Laboratory Test (id. 1239/11-16), and intended to use the results of that test as substantiation for making an advertising claim that a single SuperBait bait would kill more roaches that a single RAID Max Plus bait. Id. 1186/17-23, 1239/3-10. The results of the Second Laboratory Test were reported in a June 7, 1995 Clorox internal memorandum prepared by Cara Drouin, the Clorox scientist who performed the test. PX 39; Silverman 1237/22-1238/19. Ms. Drouin reported that "Raid Max with Dursban kills more roaches than Combat SuperBait." PX 39. This result directly contradicted one of the messages Clorox intended to deliver to consumers in the Weapon Commercial — that "Combat SuperBait kills more roaches than even products with Egg Stoppers." (PX 21).^[6] Furthermore, the internal memorandum reporting the Second Laboratory Test shows that, after 21 days, RAID Max Plus with Dursban, on average, killed 1,112 of the approximately 1,200 roaches in each arena at that time, or approximately 92.67%. PX 39.

26. Defendant Clorox argues that the Second Laboratory Test is entirely irrelevant and unreliable, and therefore may not be considered by the Court in its evaluation of Clorox's establishment claims. The defendant contends that the Second Laboratory Test is invalid because it was conducted with susceptible roaches maintained in the lab for decades with no resistance to any insecticide. Silverman 1186-87. The defendant also asserts that it was not an efficacy test. Silverman 1240-41. In this regard, Dr. Silverman testified that the Second Laboratory Test "didn't measure the number of insects killed versus the total number in there." Silverman 1240-41. Specifically, he testified that no exact records were kept of the precise number of roaches that constituted the initial testing population, Silverman 1241, or of the exact number of roaches that were added to the arenas during that test. Silverman 1240-41. As Dr. Silverman explained "[o]ne cannot derive percent kill data when you don't know what the total population was initially." Silverman 1242.

27. The Court finds unpersuasive Clorox's argument that the Second Laboratory Test is irrelevant to an evaluation of its "no more than 60%" claim. First, Dr. Silverman, Clorox's own scientist, testified on cross-examination that although he didn't know whether there were exactly 500 roaches in the initial roach population, he in fact knew the "[v]ery approximate" number. Silverman 1241/20-22. Dr. Silverman further testified that he regarded Cara Drouin, the scientist who performed the test (and who works for Dr. Silverman) to be very precise. Silverman 1242/11-19. Because the approximate number of roaches in the population at fixed dates, and the precise insect mortality, were reported by the Clorox scientist, it impresses the Court that this experiment permits some quantitative conclusions to be drawn, allowing a reasonable margin for error, concerning the extent to which the SuperBait and RAID Max Plus products reduced the respective roach populations to which they were applied. To the extent that Dr. Silverman suggests otherwise, the Court finds his testimony not to be credible.

28. Second, the Court finds without merit Clorox's contention that it would be inappropriate for the Court to consider the Second Laboratory Test because the roaches that were tested were not resistant to Dursban. As the Court subsequently will discuss, see infra ¶ 54, Dursban resistance among the German cockroach, although occurring in widespread "pockets" throughout the United States, is not pervasive. Thus, although the Second Laboratory Test is clearly not indicative of how the products at issue could be expected to perform in certain low income housing units in Florida, see Koehler 1301, the Court finds this test to be highly relevant to an evaluation of how these products would perform in locations throughout the United States where Dursban resistance or repellency is not as frequent, such as Long Island, where this Court sits.

29. In sum, the Court finds the Second Laboratory Test to be relevant to its evaluation of Clorox's claim that RAID Max Plus kills "no more than 60%" of consumers' roaches, and that Clorox's failure to take this test into account materially undermines the reliability of this claim.

F. Clorox's Field Tests

30. At the preliminary injunction hearing, in addition to the First Laboratory Test, Clorox relied on four of its field tests as substantiation for the Commercial's claims: (i) a 1995 test conducted in Ocala, Florida, (ii) a 1994 test conducted in Miami, Florida, (iii) a 1994 test conducted in Sanford, Florida, and (iv) a 1994 test conducted in Puerto Rico. DX 160. However, in identifying its substantiation for the Commercial in internal documents before this litigation arose, Clorox did not mention any of those tests, even though the Miami, Sanford and Puerto Rico tests were all completed in 1994, many months before the Commercial was filmed. PX 22, 160; see supra ¶ 24. Indeed, a pre-litigation Y & R document specifically noted that Clorox was not relying on the Sanford test. PX 28 at 3969.

31. The results of the four field tests upon Clorox relies are summarized below:

The 1995 Ocala Field Test: This was the only field test performed by either party on the currently marketed formulas of SuperBait and RAID Max Plus Egg Stoppers containing Dursban. Mr. Bieman, a PCO employed by Clorox, was assisted in that test by Jose Tomeu of Alachua Pest Control, who chose the test site from locations suggested by the Ocala Housing Authority, collected the sticky traps, and counted the roaches at the test's conclusion. Bieman 1042. At the completion of the test, SuperBait had achieved a mean kill rate of 97.05% versus 50.72% for RAID Max Plus Egg Stoppers with Dursban. DX 41 at 3009A.

The 1994 Miami Field Test: The results of the Ocala field test are consistent with the results of a field test conducted a year earlier in Miami, Florida, in which SuperBait achieved a 99.2% mean reduction in roaches compared to a 50.9% mean reduction in apartments treated with RAID Max Plus Egg Stoppers containing sulfluramid, the product SCJ was marketing at that time. Bieman 1035-36; DX 26.

The 1994 Sanford Field Test: In July 1994, after the Miami test, Clorox conducted a field test of SuperBait in Sanford, Florida. Even though the infestation at the site was substantial, SuperBait killed, on average, 94% of the cockroaches at the end of twelve weeks. Bieman 1038; DX 29.

The 1994 Puerto Rico Field Test: Clorox subsequently tested its product in apartment blocks in San Juan, Puerto Rico. After thirteen weeks, the cockroaches in untreated apartments had increased by 63%, but the apartments treated with SuperBait obtained a 97% reduction. Bieman 1040; DX 31.

32. Each of these four field tests involved SuperBait. However, only one, the Ocala test, also involved RAID Max Plus with Dursban. The Ocala test did not begin until after the Commercial was filmed (compare PX 24 at 3010 with PX 79 at 1) and was not completed until nearly two months after the Commercial began running in late June 1995. Compare DX 41 with Clorox Answer ¶ 1. These four field tests each used a methodology developed in March 1994 by Clorox employee Donald Bieman. Bieman 1015/11-15; DX 17. This new Clorox methodology resembles other, established field testing methods in only the most general terms: Clorox conducts its field tests in low-income public housing; it uses sticky traps to sample the roach population within an apartment; it then applies roach bait treatments to infested apartments; and it uses sticky traps at regular intervals throughout the test (i.e., post-treatment traps) to measure changes in cockroach populations, expressed in terms of mean percent reduction. Owens 113/25-115/1, 124/13-126/15, 131/6-24; see Boase 1458/3-1462/17; DX 17.

33. In implementing this general approach, however, the Clorox protocol not only relies on the skill of a trained professional to locate cockroach harborages and strategically place the baits in close proximity to the harborages, but also deviates substantially from label directions in the number and placement of baits. Koehler 847/11-848/11; see DX 3; DX 17 at 88. Under the new Clorox protocol, Mr. Bieman, a licensed PCO (Bieman 991/14-17), first conducts a detailed and systematic inspection of each apartment, either by himself or with another PCO assisting him. Id. 1017/20-1018/7. This inspection can last up to two hours, factoring in the time attributable to documenting the location of the traps and setting the traps. Bieman Dep. 72/9-73/6.^[7] The inspection includes the use of sticky traps to locate suspected roach harborages. See DX 17 at 87. As Mr. Bieman acknowledged, this usage of sticky traps differs markedly from traditional field testing methods, in which sticky traps are employed only to monitor changes in roach populations, not to locate harborages. Bieman 1100/24-1101/6.

34. Under the new Clorox protocol, Mr. Bieman places a sticky trap at each location where cockroaches are seen in the initial inspection. Bieman 1017/20-1018/25; DX 17. If the trap catches four or more roaches, baits are placed at that location "in a position right next to where the trap is" in order to "monitor what is going on where you put the baits." Bieman 1020/4-18. Thus, in the new Clorox protocol, the number and location of baits may vary from one apartment to another, depending upon the results of Mr. Bieman's inspection and sticky trapping. Id. 1020/4-1021/1. By contrast, in the traditional roach bait field test design discussed at ¶¶ 61-67 below, the number of baits is the same and the bait placement locations are the same or similar in each apartment in the test.

35. The new Clorox protocol calls for at least 3 grams of bait — the equivalent of 2 bait stations (Bieman 1083/6-14) — to be placed at each location where a bait is placed. DX 17 at 88. However, Mr. Bieman placed three baits at each location in the Ocala test (Bieman 1137/2-5) and one large bait station for American roaches — the equivalent of three German cockroach bait stations — in the Sanford and Puerto Rico tests. Bieman 1121/4-1123/12. Clorox claims that this much bait is placed at each location due to concerns about bait depletion over the course of the test. Bieman 1083/6-1084/5; Shapas 628/6-629/2. Bait depletion refers to the situation in which all the bait in particular stations is eaten before a roach infestation is eliminated, such that there is no longer any bait at certain locations where roaches are present. Shapas 783/8-14.

36. Clorox maintains that its new protocol is a scientifically valid means of testing the effectiveness of roach bait products, not only for internal testing purposes, but also for purposes of gauging the effectiveness of these products under the real-world conditions of consumers' homes. This proposition has not been adopted by any member of the urban entomology community outside of Clorox, other than by Clorox's paid expert witness in this case, Mr. Boase. Although Clorox maintains that Dr. Richard Brenner of the United States Department of Agriculture endorsed aspects of the new protocol (Bieman 1022/4-8; Boase 1469/18-21, 1623/19-1624/1), Dr. Brenner's deposition makes it clear that he did not intend "to communicate that [he] approved or endorsed Clorox's field test protocol as a valid and appropriate method for determining the percent of cockroach population reductions in infested residences." Brenner Dep. 40/10-17. Dr. Brenner specifically declined to sign a letter proposed by Clorox which stated that "[Clorox's] methodology makes the results more accurate and precise than previous methods." PX 55.^[8]

37. Even more importantly, abstract debates over whether or not the new Clorox methodology may be a scientifically valid for some purposes are of no relevance to SCJ's motion. In assessing an advertised claim to consumers based on scientific tests, one cannot evaluate the appropriateness of the test method without also considering whether the test in fact measures what is being claimed. Here, the Commercial unambiguously informs consumers how SuperBait and RAID Max Plus will perform when consumers use those products in their homes. Thus, irrespective of whether the new Clorox protocol is a valid and appropriate method for a PCO to treat a roach infestation, or for comparing roach baits for product development purposes, it can only substantiate the claims in the Commercial if it measures how the products will work when used by consumers in their homes following product label directions.^[9] In three critical respects, the new Clorox protocol fails to substantiate the claims made in the Commercial.

38. First, Mr. Bieman's role in inspecting and baiting the apartments interjects an entirely unrealistic element as compared to how these products are applied in the real world. It cannot seriously be disputed that professional PCOs are far better trained than the average consumer — and, with the use of sticky traps, much better equipped — to locate cockroach harborages and place baits in a manner that will optimize performance. Koehler 850/10-12 ("Basically, the average consumer would not have anywhere near the capabilities that Mr. Bieman has in locating harborages."). And, while Mr. Bieman testified that consumers are even more capable of locating harborages than are PCOs (Bieman 1101/24-1102/11), that testimony is not credible. If Mr. Bieman were correct, there would be little need or justification for PCOs, a point which Dr. Frishman recognized when he conceded that he "wouldn't be making a living" if professionals were not more skilled than consumers "in determining precisely where to place the baits...." Frishman 716/4-9. Dr. Frishman further conceded in cross-examination (712/2-16):

Q. And you would agree in many situations [Mr. Bieman's] skills and your skills in locating cockroach harborage areas are better than the homeowners?

A. Yes. In terms of the out-of-the-way places. But there are times when we rely on that homeowner — in other words, part of the training I do, which is standard procedure, is to ask the people living in the facility where do you see the cockroaches?

Q. But wouldn't you agree, Doctor, that many times when you go into apartments for treatment, and you see evidence of somebody's consumer baits lying around, it doesn't matter who it is, one of the most frequent reasons for failure is that the baits are not located in sufficient proximity to the harborages, to all of the harborages?

A. In many cases, yes.

Frishman 712/2-16. Because precisely placing baits as near as possible to roach harborages is critical to the success of baits, see supra ¶¶ 13-18, Mr. Bieman's superior expertise in locating roach harborages is highly likely to have influenced significantly the percent reductions SuperBait has achieved in tests conducted under the new protocol. Koehler 848/15-852/1.

39. Second, the number of baits used in the new Clorox protocol is completely inconsistent with product label directions. Both SuperBait and RAID Max Plus are sold in packages of 12 baits. In plain and unambiguous language, the SuperBait label states, "For thorough roach control, use all 12 bait stations at the same time." DX 3. In similarly clear terms, and in three separate places, the RAID Max Plus label states "[f]or thorough roach control" and "for best results," "use all 12 bait trays at one time." DX 4. Yet in three apartments in the Ocala field study, including 2 of the 8 RAID Max Plus apartments, Mr. Bieman used only 6 bait trays (PX 113), or precisely half the number recommended on the label for "thorough roach control." DX 4. One such apartment treated with RAID Max Plus achieved only a 20.7% roach population reduction after 12 weeks, which substantially reduced the mean percentage reduction achieved by RAID Max Plus in this test. PX 113.

40. Moreover, in many apartments in both the Miami and Ocala tests, Clorox used substantially more than 12 baits, and, in fact, averaged more than 2 boxes of baits. In seven apartments in the Ocala study, Mr. Bieman used 3 or more boxes of baits: 45 baits (apt.2904); 57 (apt.1630.4); 36 (apts. 1640.1 and 1628.1); 42 (apt.3032); 63 (apt. 1628.2); and 42 (apt.1630.6). PX 113. In fully half of the SuperBait apartments in the test, more than 2 boxes of baits were used. Id. Similarly, Mr. Bieman used at least 30 baits in nine apartments in the Miami study, including 98 SuperBait trays — more than 8 packages — in one apartment. PX 77.

41. Clorox argues that the SuperBait label instructs consumers to use "at least" 24 baits in heavy infestations. Clorox Pre-Hearing Memo. 23 (emphasis in original). That is not true. In fact, the SuperBait label instructs consumers that "[h]eavy infestation may require 2 boxes (24 bait stations)," DX 3, and does not suggest that more baits would ever be required. Moreover, the product labels do not define what a "heavy" infestation is and there is no evidence of what U.S. consumers perceive to constitute a "heavy" infestation. The evidence does suggest that consumers rarely use more than a single package: on average, consumers apply much less than a full package at a time — 4.9 baits — and only 10.5% applied 11 baits or more at a time. PX 142 at Question 6. This is counterbalanced to some extent by SCJ consumer survey data indicating that approximately 24% of the respondents replace their baits more frequently than every three months. PX 142, Response to Question 7. In any event, whether it would be proper to conduct a field test using up to 24 baits instead of 12 is not in issue here. There is no language on the SuperBait label (DX 3) that instructs consumers to use more than 2 boxes of baits under any circumstances, even in treating "heavy infestation." Furthermore, the Court expressly finds — both with and without factoring in the aforementioned consumer survey evidence — that no fair reading of the SuperBait label instruction would suggest to a lay consumer that, under certain circumstances, more than 24 bait stations would be required at a given time. But that is precisely what Mr. Bieman did. As Clorox's own contemporaneous document admits: "[t]he total number of bait stations used in the test [for] both SuperBait and Raid were more than the number recommended on the label." PX 20 (emphasis added).

42. Third, nothing on the label even remotely suggests that consumers should place multiple baits at each location where baits are placed, as Clorox does in its new protocol. This practice cannot be justified by Clorox's concern that bait depletion will affect the results of its field tests. Bait depletion is a fact of life when consumers use these products, in large part because consumers cannot readily tell when baits are depleted. These products only contain one-and-one-half grams of bait — a tiny fraction of an ounce — encompassed in an enclosed opaque case. Shapas 787/4-12; Koehler 877/9-16. Hence, "[i]f [a bait station] was depleted, it would be very difficult for a consumer to tell whether it was empty or not." Koehler 877/14-16.^[10]

43. As Dr. Koehler testified, by using a PCO to locate harborages and place baits, by extensively using more than 24 baits, and by placing multiple baits at each location, the new Clorox protocol artificially boosts the performance of SuperBait far above the levels that would likely be achieved if the product were used by a consumer in accordance with label directions. Koehler 851/9-852/1, 860/16-861/8, 862/21-863/9, 871/6-872/17, 874/18-875/14.^[11] The clinical evidence corroborates Dr. Koehler's opinion regarding the "boosting" effect of the new Clorox methodology on the efficacy of SuperBait. While in the Clorox field studies using the new method, SuperBait was reported to consistently achieve average percent reductions in the high 90s, in Clorox field tests using its previous methodology, SuperBait often reduced roach populations by less than 50%. See supra ¶ 20.

44. Furthermore, Mr. Boase, Clorox's own witness, conceded that the results of the Ocala field study would have been different if Mr. Bieman had placed only one bait at each bait location, rather than three:

Q. Mr. Boase, ... I want you to make the following assumption: In [the] Ocala test, one bait was placed at each location rather than three baits for SuperBait, all right? I want you to assume that all other aspects of the protocol were followed by Mr. Bieman as he followed them in the actual test. Is it your position that the results in terms of percentage reduction would have been the same, different or you don't know?

A. It is my position that the results — and we are talking about the results for the trial as a whole — would have been different, or likely to [have been] different, had one used one bait station for either product, because there is a likelihood of bait [depletion] in some locales in some apartments, and this is what that three-bait-perlocation issue is trying to address and remove. ...^[12]

Boase 1638/23-1639/16.

45. The remainder of Mr. Boase's answer to that inquiry, and other evidence from the mouths of Clorox's own witnesses, demonstrates that Mr. Bieman and Clorox developed the new protocol with the objective of eliminating real-world variables that cause SuperBait and RAID Max Plus to perform more evenly when used in accordance with label directions. Mr. Boase continued:

And I believe bait depletion and the experimental response of — the researcher's response of dealing with bait depletion, or dealing with heavily infested apartments like this is an interesting challenge for researchers. Some choose to ignore it and follow the label with twelve or occasionally 24 baits, or don't recognize that these infestations are heavy.

The Clorox protocol recognizes they are heavy infestations and there is the possibility of bait depletion. And there is the possibility that this may induce unwanted variance within the trial.

Boase 1639/17-1640/2 (emphasis added); see also Shapas 639/18-24 ("We think by putting the bait close to the insects, by eliminating bait depletion as a variable, and by using our sticky trap regime, where we are actually sampling where the roach population is, we have been able to minimize this sort of variability, and really tighten up our data sets so we can make really good product comparisons.").

46. In his testimony, Dr. Shapas tried to portray the elimination of real-world variances as scientific progress, stating that the new protocol is "like focusing the Hubble Telescope, in that it now allows us to see things much more clearly." Shapas 635/8-10. However, as Dr. Koehler testified, the new Clorox methodology cannot be justified on that basis:

[T]he purpose of scientific tests is not necessarily to reduce variance. And, basically, if consumers read the label, place the product out according to label directions, sometimes the product will work, sometimes the product won't work. And that is inherent variation as a result of that kind of treatment. Of course, that's exactly what some of these tests show. And any attempts to apply the bait away from the label directions in order to reduce variation, then, of course, that is inappropriate when it is characterized to consumers that that is the way the product will work when it is applied in their homes — when they apply it in their homes, according to label directions.

Koehler 878/5-21.

47. In sum, the new Clorox methodology not only fails to test the effectiveness of household roach bait products when used by consumers in their homes in accordance with label directions; moreover, it affirmatively eliminates real-world variables that can affect the actual performance of roach bait products in consumers' homes.

G. Low Income Public Housing Is Not Fairly Representative Of How RAID Max Plus Will Generally Perform In Other U.S. Households

48. Both Clorox and SCJ have routinely conducted roach bait field efficacy trials in low-income public housing, often in Florida. The principal reasons for selecting such locations are practical ones, such as having enough infested apartments in close proximity that will provide adequate replication and that are similar in construction. Owens 114/12-115/5; Shapas 620/18-25. However, there is no dispute that low income public housing presents an unusually challenging environment for roach bait products due to a variety of factors, including poor sanitation and clutter that provide roaches with a plentiful array of alternative food sources, and make it more difficult for roaches to encounter the baits. Shapas 621/1-6; Owens 119/8-20; see PX 18 at 295 ("Baits work best when availability of competing food sources in the infested area is minimized. Therefore, the level of control with baits in apartments or homes with poor sanitation will not be as high as in clean apartments."); id. at 244. Indeed, Clorox acknowledged in its opening statement that the reason it tests in low-income housing is that "[y]ou test your products in the most challenging environment." Tr. 37/7-9.

49. Clorox no doubt has sound reasons to conduct its basic scientific research in the most rigorous environment available. However, the Commercial states, without qualification, that RAID Max Plus kills "no more than 60%" of your (i.e., consumers') roaches. Thus, tests in low income public housing provide a questionable basis for telling consumers that RAID Max Plus will not achieve higher levels of reduction in less demanding environments, such as private homes and apartments located in Long Island and elsewhere in the United States.

50. The Commercial is likewise unqualified and unequivocal in its comparative aspect, i.e., its claims that SuperBait kills just about all (up to 98%) of your roaches, while RAID Max Plus only kills some (no more than 60%) of your roaches. There is no doubt, however, that resistance to chlorpyrifos, also known by its brand name — Dursban, substantially contributed to RAID Max Plus's erratic performance in Ocala. In fact, Clorox knew when it approved the Ocala site that RAID Max Plus would perform poorly there. As Mr. Bieman wrote in a June 17, 1995 report to Drs. Ochomogo, Silverman and others, "[f]rom our own experience (mine and Mr. Tomeu's) with chlorpyrifos in multiple housing settings in Florida, we anticipated its poor efficacy; chlorpyrifos has not been effective in more than 10 years in such locations." PX 79 at 3.^[13] And, while Mr. Bieman was somewhat grudging in his acknowledgment at the hearing that Dursban resistance was the reason he expected RAID Max Plus to perform poorly (see Bieman 1125/19-1130/7), resistance was the only explanation offered in a Clorox document prepared by Dr. Ochomogo before this litigation began and shortly after Mr. Bieman wrote his June 17 report. PX 29 ("An explanation for this data is resistance in German roaches to [Dursban]...."). Moreover, convincing testimony was provided at the hearing that Dursban resistance accounted for RAID Max Plus's performance in the Ocala test. Cochran 538/9-540/6.

51. There is no evidence that Clorox made any attempt to ascertain whether the level or extent of Dursban resistance at the Ocala complex is fairly representative of other households throughout the U.S. Indeed, Mr. Bieman testified that he had never had any discussions on this subject with anyone at Clorox, and never even considered holding the test anywhere but in public housing in Florida (Bieman 1130/8-12), notwithstanding his prior knowledge of how the SCJ product would likely perform there and that, when the Ocala test began in May 1995 (see PX 79 at 1), Bieman had been aware for months that Clorox was considering an advertising claim based on its new protocol comparing SuperBait to RAID Max Plus. Bieman 1113/11-24. Thus, in asserting that the Commercial's advertising claims are substantiated by the Ocala test, Clorox is turning a blind eye to whether Dursban resistance at public housing sites in Ocala, Florida is at all typical of the scope and level of Dursban resistance in other types of residences throughout the United States, such as private homes and apartments.

52. SCJ, meanwhile, has presented affirmative evidence to establish that the Dursban resistance encountered in the Ocala, Florida field test in not representative of the level of resistance to be expected in other regions of the United States that have a different demographic, or climatic, environment. This evidence took the form of the testimony of SCJ's expert, Dr. Donald G. Cochran, Professor Emeritus of Entomology at Virginia Polytechnic and State University, and a leading authority on German cockroach resistance. Dr. Cochran has studied, written about and taught the subject for the major part of his nearly 40 year career as an entomology professor (Cochran 500/20-502/17; PX 6); both sides have retained him to conduct resistance testing (Cochran 512/5-11); and Clorox's expert, Dr. Frishman, recommends that PCOs who suspect resistance problems contact Dr. Cochran (PX 153 at 2). Dr. Cochran testified that in his opinion, the extent of Dursban resistance in the Ocala study is not at all indicative of the extent of Dursban resistance generally in the U.S. Cochran 539/17-22. His opinion is consistent with other credible evidence. As Dr. Cochran testified:

[R]esistance is a genetic characteristic, and, therefore, for resistance to develop in any given population, the gene for resistance to that insecticide has to be present in that population.... [I]n addition, there has to be an extensive and continued exposure to that insecticide. Because what has to happen is that the population basically has to be treated completely so that the insecticide has a chance to kill off the susceptible individuals within that population.

Cochran (524/13-525/5) (emphasis added). This process generally takes a period of two or three years, during which the entire roach population "has to be exposed on a more or less continuous basis" to Dursban. Id. 525/13-526/1. As such, resistance is most likely to occur in commercial establishments and hospitals, which are heavily treated by professional PCOs on a monthly basis, and in low-income public housing complexes where state and federal regulations require regular pest control activities by professional PCOs. Id. 527/20-529/4. See also PX 123 (internal SCJ memo stating that "it would be reasonable to expect that [Dursban resistance] would be limited to areas subject to repeated heavy P.C.O. pesticide use"); PX 18 at 190. Predictably, the low-income public housing complex used in the Ocala test had long received regular cockroach treatments from PCOs. See Koehler 853/18-855/11.^[14]

53. Moreover, Dr. Cochran's opinion is corroborated by the testimony of Dr. Frishman, who acknowledged that he has extensive familiarity with the experience of PCOs with Dursban resistance on Long Island. Frishman 695/6-11. Dr. Frishman conceded that Dursban resistance on Long Island is rare and that in many areas of the country, PCOs are still able to control roach infestations with Dursban. Id. 682/11-15, 688/3-12, 690/8-692/19.

54. In addition, Dr. Cochran's opinion is buttressed by the history of insecticide resistance, the continued substantial use of Dursban by PCOs. PCOs are in business to effectively treat insect infestations and it would be contrary to reason to believe that an insecticide would continue to be heavily used by PCOs if it were generally ineffective. See Cochran 523/4-10. Not surprisingly, history has shown that when high resistance levels become so pervasive that PCOs cannot achieve effective roach control with an insecticide, PCOs have quickly abandoned the insecticide, as they did with chlordane in the 1950s and malathion thereafter. Cochran 523/4-10.^[15] By contrast, Dursban continues to be used extensively by PCOs today. E.g., Cochran 524/3-12; Frishman Dep. 55/5-20; Boase 1643/23-1646/22. Thus, as Dr. Cochran testified, while Dursban resistance may be "geographically widespread," it is not "pervasive." Cochran 541/14-542/10.

55. Clorox's principal response is that more than half the roach strains Dr. Cochran tested in the last few years proved to be resistant to Dursban, from which Clorox extrapolates that Dursban resistance is pervasive in the United States. However, Dr. Cochran persuasively testified that the roach strains he tested represent a "skewed sample" (Cochran 588/1-12), because Dr. Cochran has made it well known to the "cooperators" who supply him with roaches that he is only interested in testing strains where control failures indicate resistance has occurred. Id. 519/1-22. This testimony is corroborated by an article published by Dr. Frishman in a PCO trade journal (PX 153 at 2), where Dr. Frishman urged PCOs to send roaches to Dr. Cochran for testing where control failures had occurred due to suspected resistance.^[16]

H. Field Testing Results Ignored by Clorox

The Commercial's Comparative Claims

56. The Commercial tells consumers not only that SuperBait consistently kills virtually all of their roaches, but also that it consistently kills far more roaches than RAID Max Plus. PX 1, 2. However, a different result was achieved in what appears to be the largest field test Clorox ever conducted comparing SuperBait to Regular RAID, which contains the same formulation as the roach bait component of RAID Max Plus. That test was a five product comparison conducted by Clorox in 1993, prior to the implementation of its new testing protocol, involving SuperBait, Regular COMBAT, RAID Max with sulfluramid, Regular RAID and Black Flag (the "Five Product Test"). One year earlier, in 1992, Clorox had concluded that a few roach strains had become averse to the glucose in Clorox's Regular COMBAT baits and that those strains were not averse to a bait base which contained fructose as the sole sugar. Silverman 1164/22-1165/5. Beginning in 1992, Clorox began field testing the presently-marketed formulation of SuperBait known as Krystar, which contains fructose and no glucose. See Bieman 1077/7-1078/14 (acknowledging that Krystar is the presently-marketed formulation of SuperBait); PX 61 (showing that beginning in 1992, all Clorox field tests of SuperBait used the Krystar formulation). Building on the 1992 field studies of SuperBait with Krystar, Clorox conducted the Five Product Test, which involved an average of 17 apartments for each product tested. At 12 weeks in the Five Product Test, SuperBait and Regular RAID performed at substantial parity: SuperBait achieved a mean reduction of 29.9% and Regular RAID a mean reduction of 26.7%. PX 60; Owens 204/15-205/15.^[17] These results are a far cry from the Commercial's absolute and comparative claims about SuperBait's efficacy.^[18]

57. Moreover, RAID Dursban bait with Egg Stoppers and SuperBait performed statistically equally in the only SCJ field test comparing those products — the 1993 study conducted by Dr. Gold of Texas A & M. PX 70. At 12 weeks, the RAID product achieved a mean reduction of 7.32%, while SuperBait achieved a mean increase of 26.3%. PX 70 at 115, 117; Owens 206/19-207/22. The RAID Dursban bait with Egg Stoppers also numerically outperformed SuperBait at the 2, 4 and 8 week intervals. PX 70 at 115, 117; Owens 206/19-207/22.^[19] After conducting a statistical analysis, Dr. Gold concluded that "Old Combat (Maxforce) seemed to do better than the new formulation [i.e., SuperBait], and the new Dursban bait was equal to either Combat formulation tested." PX 70 at 111; Owens 208/17-209/14.

58. Although Clorox began selling SuperBait during the winter of 1993, see supra ¶ 6, Dr. Shapas testified that the new Krystar formulation did not become commercially available until 1994. Shapas 610/18-611/2. Clorox elected not to produce any documentary evidence to corroborate this testimony, even though such evidence was likely available. Dr. Shapas was not a credible witness (see Shapas 770/20-780/7, 791/7-792/17) and this testimony is rather dubious, in light of Dr. Silverman's acknowledgment that Clorox had isolated the glucose problem by 1992 and the documented fact that as early as 1992 (Shapas 773/1-9), Clorox's field tests of SuperBait used the Krystar formula. PX 61.

59. In any event, even assuming that the Gold test involved the glucose version of SuperBait, its results are still relevant. Dr. Silverman and Mr. Bieman have acknowledged in a publication they co-authored that the problem of glucose aversion is "sporadic." PX 124 at 34. Clorox has never determined how widespread this phenomenon is (Silverman 1248/22-1251/8) and has never found any glucose-averse roaches in Texas, where Dr. Gold conducted his test, or in any other state besides Florida and California (and then only a single strain in the latter state). Silverman 1250/23-1251/15. Moreover, in the only field test in evidence comparing the efficacy of the glucose and fructose versions of SuperBait, there were no statistically significant differences. PX 64; Owens 210/4-212/19.

The Commercial's "Just About All" And "Up To 98%" Claims

60. In contrast to the Commercial's claims that SuperBait kills "just about all" and "up to 98%" of your roaches, Clorox's field tests using its previous protocol regularly achieved mean percent reductions of 50% or less. For example, at 12 weeks in Clorox's Five Product Test, the mean reduction for SuperBait was only 29.9% for the 18 apartments tested. PX 60; Owens 191/14-193/10. And, in two 1992 Clorox tests in Orlando, SuperBait with Krystar achieved mean reduction scores of 23.7% and 29.4% at 2 weeks; 50.3% and 12.4% at 4 weeks; 49.6% and -29% at 8 weeks; and 34.1% and 44.5% at 12 weeks. PX 61 at 910; see Owens 188/22-191/4. As further discussed at ¶¶ 61-67 infra, the field test methodology used by Clorox in those tests contained standard elements commonly used by urban entomologists for field testing OTC roach baits. As recognized by these entomologists and by pre-litigation documents, the standard elements are designed to, and do, follow package label directions. These and other relatively modest test results for SuperBait are confirmed by Clorox's Dr. Silverman who, in contrast to the exaggerated claims in the Commercial, represented to his fellow scientists in a peer-reviewed article that using the new bait formulation (i.e., Krystar), Clorox achieved population reductions of from 42-85%. DX 84 at 627.

I. Other Issues Raised By Clorox

Clorox's Criticisms Of Standard Field Testing Elements

61. In this litigation, Clorox has repeatedly criticized the SCJ field test protocol, which SCJ has referred to in its previous Court papers as reflecting an industry "standard methodology." In putting the SCJ methodology on trial, Clorox has tried to show that there is no one "standard" field testing methodology, and that the SCJ methodology is inconsistent with the label directions for the SuperBait and RAID Max Plus products. In the final analysis, however, Clorox has devoted considerable effort to proving a point that its own witnesses reject as untrue.

62. Drs. Owens and Koehler each testified that while there are variations in certain aspects of the field testing methodology employed by U.S. urban entomologists, there are certain key elements that have long been standard, including: the use of the same number of baits (generally 12) in each apartment; the placement of those baits at the same or equivalent bait locations in each apartment; the placement of one bait only at each bait location in each apartment; where trapping instead of visual counting is employed, the use of the same number of traps in each apartment; and the placement of those traps at the same or equivalent locations from apartment to apartment (collectively, the "standard elements"). Owens 127/7-137/17; Koehler 878/22-881/9; PX 160. Drs. Koehler and Owens did not adopt this position for this litigation. Rather, the documentary evidence shows that virtually every U.S. entomologist whose name was mentioned during the trial has used these standard elements in field efficacy trials. E.g., DX 46 at 109 ¶ C, 110 ¶ E (Professor Roger Gold); PX 143 at 62 (1st and 2d paragraphs) (Professors Clyde Ogg and Roger Gold); DX 52 at 0056 (Dr. Shripat Kamble); DX 47 at 0027 (Dr. Philip Koehler); DX 44 at 2742 (Dr. William Robinson); DX 74 at 1156 (Dr. Arthur Appel).

63. Furthermore, before Mr. Bieman developed the new Clorox protocol, Clorox used these same standard elements in its field efficacy trials (Bieman 1067/13-1068/12), as did Cyanamid before it. Id. Indeed, Mr. Bieman conducted 50 to 60 field tests using those elements as an employee of Cyanamid and Clorox. Id. Not surprisingly, Clorox has referred to these elements as "standard" or "established" in contemporaneous documents created before the commencement of litigation. PX 45; PX 53 (stating that the new method differs from "established, existing procedures"). And, when Mr. Bieman was asked why he continued using Cyanamid's testing methodology for seven years despite his alleged belief that the methodology flew in the face of label directions, he explained his thinking at that time: "Listen, I have been doing it for all those years, and everybody else is doing it for all these years. No one — I mean, how is what I am doing any different from anyone else? This is supposedly the standard...." Bieman 1068/22-1069/23 (emphasis added). See also Boase 1611/3-1612/19 (Standard elements are "common themes" used by researchers "throughout the USA.").

64. Although never proffered as an expert, Mr. Bieman testified that in his view the standard elements are inconsistent with label directions. Bieman 1068/18-21. Even if he were correct, that would largely be beside the point, because this action does not involve a challenge to advertising claims made by SCJ based on its field testing results. However, the view espoused by Mr. Bieman in court is directly at odds with the pre-litigation position adopted by Dr. Silverman, Mr. Bieman's boss (Bieman 1132/12-14), who wrote that Clorox's previous method "followed label directions" and thus contrasts with Clorox's new protocol. PX 45; Silverman Dep. 78/12-13. Moreover, Clorox has itself relied on tests conducted by SCJ using these standard elements (see Clorox Pre-Hearing Br. at 16-18; DX 161), and Clorox's expert witness expressly declined to "condemn" as scientifically inappropriate field tests using these standard elements when offered the opportunity to do so by SCJ's counsel. Boase 1609/22-1610/23.

65. In contrast to Mr. Bieman, Dr. Owens and Dr. Koehler each testified that field testing using these standard elements is faithful to product label directions. Owens 128/1-18; Koehler 852/8-25. This testimony is corroborated not only by Dr. Silverman's pre-litigation chart (PX 45), but also by the pre-litigation reports of other eminent U.S. entomologists having no connection with this litigation. See Reports of Dr. Gold (DX 46 at 110 ("Bait stations were placed in a manner ... modeled after label recommendations.")) and Dr. Bennett (DX 55 at 174 (bait placements "based on the label recommendations")). Of course, the concessions of Messrs. Bieman (1067/20-1069/24) and Boase (1611/21-1612/23) that the standard elements have been consistently followed by U.S. entomologists is itself strong evidence that field testing following those elements is consistent with label directions. It is hardly credible to suggest, as Mr. Bieman apparently does, that the entire U.S. urban entomology community had adopted field testing procedures for OTC roach baits that were inconsistent with label directions for those packages.

66. Furthermore, a simple comparison between the product labels on the one hand, and the standard elements respecting the number and placement of baits, on the other, proves that Drs. Owens and Koehler, and the rest of the U.S. entomology community, have every reason to conclude that the standard elements are consistent with product label recommendations. For example, use of 12 baits is consistent with the oft-repeated label recommendations to "use all 12 baits" for "thorough roach control." PX 16; PX 111. Although the labels recommend the use of additional baits for "heavy infestation," the level of infestation is in the eye of the beholder. As Dr. Koehler testified, low-income apartment residents often will not see large numbers of roaches even in apartments where thousands of roaches live behind the walls. Koehler 1386/22-1387/9. Moreover, Clorox field tests only in kitchens, see DX 17, while the packages recommend that in heavy infestations at least some of the additional baits be placed in rooms other than the kitchen. PX 16, 111; see PX 18 at 250 (Clorox placement guide showing placement of baits in bedroom, living room, etc.); Koehler 856/15-857/18. And, the record evidence indicates that consumers rarely apply more than 12 baits at a time. See supra ¶ 41. Thus, although it might not be inappropriate to use up to 24 baits in a field test, the use of 12 baits per apartment can hardly be viewed as contrary to label directions.

67. In like manner, while the placement of one bait station at each bait location is consistent with label recommendations, the use of multiple baits is not: nothing on the product labels even remotely instructs a consumer to place multiple baits at any, let alone all, locations. Koehler 871/17-872/3. Similarly, the placement of baits at the same or equivalent locations within each apartment closely adheres to label directions. Consumers are far less skilled than professional PCOs in locating cockroach harborages, especially in peripheral areas of a kitchen. See supra ¶ 38. As a result, researchers implementing the standard elements will place the majority of baits in those areas designated as bait locations on label diagrams, while placing other baits in locations where consumers may be expected to see roaches. Owens 128/1-130/7. This approach harmonizes label language instructing consumers to place baits "where cockroaches are seen" with label diagrams designating the "Most Important" (PX 111) and "For best results" (PX 16) bait locations. Koehler 1374/6-16.^[20]

Repellency and Bait Shyness

68. Clorox argued at length that one reason RAID Max Plus is, allegedly, an ineffective product is that Dursban is repellent and/or causes "bait shyness." Bait shyness refers to the preferential feeding habits of the cockroach caused by unpalatable bait, while repellency refers to the effect of the vapor phase of the active ingredient (Dursban) which causes roaches to stay away from the bait. Boase 1585-87.

69. The Court agrees with Clorox's position that RAID Max Plus's active ingredient of Dursban produces roach repellency and bait shyness. The incidence of roach repellency, in fact, was acknowledged by SCJ's own scientists, in 1989, at the time that RAID Max was introduced with sulfluramid as its active ingredient. See DX 143 at 38 (quoting SCJ scientist Dr. Keith Kennedy as stating that "[o]ld style roach killers were roach-repellent...."). Indeed, numerous studies by a variety of researchers, including SCJ employees, have concluded that Dursban has a problem of repellency. See DX 74 at 1154 (Appel study concludes that "Relative repellency ... was greatest for chlorpyrifos formulations"); DX 75 (University of Nebraska study concludes that chlorpyrifos was repellent); DX 73 at 1285 (Dr. Koehler and others state that chlorpyrifos and certain other toxicants "are known to be low level repellents to cockroaches" and therefore less effective).

70. The issues of repellency and bait shyness, however, are not of themselves determinative of the parties' dispute. The Commercial does not make any claims about the inherent characteristics of SuperBait and RAID Max Plus, or why they are, or are not, effective. Instead, the Commercial makes explicit, numerical comparisons of the efficacy of those products that are allegedly based on testing. Thus, this case turns on whether or not the testing substantiates these claims, and not upon a qualitative analysis of the products' active ingredients to the extent that such analysis is collateral to the ultimate question of efficacy.

71. A related issue that is potentially determinative of the parties' dispute is whether the effects of Dursban repellency and bait shyness are of sufficient magnitude as to nullify the prejudicial effects of Dursban resistance, which the Court has found not to be pervasive throughout the United States. See supra Findings of Fact ¶¶ 48-55. In this regard, the Court expressly finds that the repellent properties and bait shyness attributable to Dursban fail to nullify, in any material sense, the testing prejudice sustained by RAID Max Plus through Clorox's failure to take into consideration the non-pervasive nature of Dursban resistance throughout the United States.

The Volatility of Hydroprene and Clorox's Failure to Consider the 80% Reduction in Roach Population Achieved by Regular RAID in the First Laboratory Test

72. As noted above, in the First Laboratory Test, Regular RAID (with Dursban but without Egg Stoppers) killed 80% of the roaches, while RAID Max Plus (with Dursban and with Egg Stoppers) killed 60%. See supra Findings of Fact ¶ 23. Plaintiff SCJ claims that Regular RAID's performance in this test should not have been disregarded by Clorox. Clorox, in turn, contends that it correctly disregarded the results for Regular RAID in making a claim about RAID Max Plus because they are different products that could perform differently.

73. In their post-hearing briefs, the parties direct considerable attention to the testimony of Dr. Koehler, one of SCJ's witnesses. In his testimony, Dr. Koehler offered a possible explanation for the disparity in the results between regular RAID and RAID Max Plus: in studies conducted by him, the active ingredient in hydroprene has been shown to impair the toxicity of Dursban in heavy concentrations but to enhance its toxicity in low concentrations. Koehler 888/13-889/10. While the plaintiff contends that Dr. Koehler's reasoning should be adopted by the Court to suggest that the results in the First Laboratory Test concerning regular RAID undermine the reliability of Clorox's "no more than 60%" claim, defendant Clorox asserts that Dr. Koehler's testimony is scientifically invalid because he testified that hydroprene "is not very volatile at all," Koehler 1411, and that it spread by a process of sublimation, which he presented as akin to crawling across a surface. Id. According to the defendant, this testimony is contradicted by SCJ's internal marketing documents, the published scientific literature, the dictionary definition of the chemical process of sublimation, and SCJ's own field tests of hydroprene conducted by well-respected urban entomologists. (Koehler Trial Tr. 1414-16); Boase Trial Tr. 1517-18; DX 71 at 1365 ["Hydroprene's LC Vapor-phase activity"], DX 46 at 111 ("high volatility of hydroprene"); PX 18 at 269 ("hydroprene is an acyclic sesquiterpenoid .... acyclic sesquiterpenoids are volatile and breakdown in ultraviolet light."). SCJ further argues that, at the time that the First Laboratory Test was conducted, Clorox had no way of knowing whether hydroprene would in fact affect the results and, if it did, the extent or direction of the effect of hydroprene from an Eggstoppers dispenser.

74. Upon considering the evidence, the Court finds that the 80% kill rate achieved by Regular RAID in the First Laboratory Test substantially undermines the defendant's claim that RAID Max Plus kills "no more than 60%" of consumers' roaches. As previously discussed, see supra Findings of Fact ¶ 7, RAID Max Plus is actually two products contained in one package. It consists of (a) 12 roach baits and (b) 3 small containers of the insect growth regulator Hydroprene. The hydroprene merely sterilizes certain roaches that come in contact with it. Owens 335/9-336/9; Koehler 887/4-888/17; PX 101, 125.

75. The Court observes, as a preliminary matter, that the record does in fact support the proposition that exposure to hydroprene negatively impacts upon roaches's susceptibility to chlorpyrifos (i.e., Dursban). This result obtains because sterilization causes their body mass to increase, and thereby presumably increases their ability to survive exposure to lesser amounts of poison. See PX 101 at 2315. Nevertheless, Clorox's own internal document, which was prepared to provide support for the original version of the Commercial, states that "[i]t takes 1-6 months for hydroprene to start controlling roaches. The effect of [hydroprene is to] render[] exposed females sterile. Therefore, [RAID Max Plus] takes longer than [Combat SuperBait] to see effectiveness in controlling roaches." PX 19 at 3872 ¶ 3. In view of these circumstances, the Court finds that the 80% result achieved by Regular RAID in the First Laboratory Test provides circumstantial evidence supporting the inference that had the First Laboratory Test been allowed to proceed for a period in excess of 20 days, the effects of the hydroprene component in RAID Max Plus eventually would have become operative, with the result that aggregate roach mortality would have increased beyond the 60% cap articulated in the Commercial. Thus, inasmuch as Clorox's own documentation strongly suggests that the 20-day duration of the First Laboratory Test did not provide enough time for hydroprene to begin controlling roaches, see PX 19 at 3872 ¶ 3, the Court regard this 20-day cutoff to constitute an uncompensated disadvantage to RAID Max Plus, that skewed the results of this test unfairly against it.

76. Correlatively, the Court finds that other evidence in the record independently impels the conclusion that the 20-day cutoff in the First Laboratory Test unfairly prejudiced RAID Max Plus. Specifically, in an inter-office correspondence from Maria Ochomogo, a Clorox scientist, to Mark Adams, concerning technical support for the Weapon Commercial, Dr. Ochomogo notes that complex arena tests are ordinarily assessed over periods between 1 day to 4 weeks. PX 28 at 3974. The rationale provided in this correspondence for the shorter period of time employed for a laboratory test as compared to a field test was that in a laboratory test, because the number of roaches is fixed, the test properly should end when all roaches that are likely to die have indeed died. PX 28 at 3974. The failure to use the outer duration of 4 weeks is especially unwarranted in this case — in which an absolute claim that RAID Max Plus kills "no more than 60%" of the roaches has been made — because at the time of the 20-day cutoff, there was little support for an unqualified conclusion that all roaches that would have been killed by RAID Max Plus in fact had died. Indeed, the basis for this conclusion is undermined by Dr. Ochomogo's acknowledgment that laboratory tests can take up to 4 weeks, or 28 days. See PX 28 at 3974.

SCJ's Decision Not to Conduct its Own Field Test

77. Clorox contends that SCJ should have conducted its own field test of SuperBait versus RAID Max Plus. SCJ did, however, cause such a test to be conducted by Dr. Gold, which was reported upon to SCJ in April 1994. PX 70. As Dr. Owens testified without contradiction, SCJ did not become aware of Clorox's contention that the Gold test involved the glucose version of SuperBait until late September or early October 1995. Owens 477/25-479/20. Thus, even if a location for the test could have been found very quickly — and Mr. Bieman admitted that finding appropriate sites is a major problem (Bieman 993/22-994/18) — the test would have been ongoing in December 1995 and January 1996. Putting aside that such a test thus would have been completed and reported on after expert discovery was over under Magistrate Judge Lindsay's discovery orders and would likely have further delayed the hearing, Clorox, despite having provided evidence of SCJ's past testing history, fails to persuade the Court that those months were optimal times for conducting field tests in the United States. Indeed, Mr. Bieman conceded this very point. Bieman 1074/10-1076/15; Owens 479/21-481/20.

Additional SCJ Tests

78. Defendant Clorox asserts that certain field tests commissioned by SCJ showed results for the two products that are comparable to those claimed by Clorox in the Commercial. A test performed under the SCJ protocol in Omaha, Nebraska in 1990-1991 found that at the twelfth week Maxforce (which is the same product as the old regular COMBAT®) achieved a 92% mean kill, while RAID Roach Controller, a Dursban bait product, achieved only 42%. DX 52 at 61. A test performed in Orlando, Florida in 1994, with SuperBait and RAID Max Plus with sulfluramid included an experiment in a smaller number of apartments using 24 bait stations for each product. Although, it was not a full-fledged, replicated test, the defendant contends that it supports the reliability of the Clorox tests by showing that SuperBait achieved a 95% mean kill rate versus 52% for RAID Max Plus with sulfluramid. DX 45.

79. Moreover, the reliability of Clorox's test results for the RAID Dursban product is supported by similar results for Dursban baits in tests performed for SCJ, including tests by Dr. Koehler. As shown below, Dursban roach baits never achieved better than a 60% mean kill in any test performed in this country since 1991:

                                                                     Mean Percent Kill
                                                                         RAID®
 Ex.     Test location and year              Coordinator             Dursban roach bait
DX44     Roanoke, Va. (1995)               William Robinson               37.96%
DX46     Houston, Texas (1993)             Roger Gold                     -64.3%
DX49     Gainesville, Fla. (1992)          Philip Koehler                 48.71%
DX53     Gainesville, Fla. (1991)          Philip Koehler                  20.5%
DX52     Omaha, Neb. (1991)                Shripat Kamble                 42.12%
DX61     Torrance, Cal. (1991)             Donald Reierson                 35.8%
DX50     Gary, Ind. (1991)                 Gary Bennett                     46%
DX50     Texas (1991)                      Roger Gold                      35.6%

80. Although the above tests clearly support the defendant's "no more than 60%" claim, the Court does not regard them to be determinative of whether the plaintiff has shown that defendant Clorox fails to substantiate this claim. Specifically, in view of the unqualified nature of the Commercial's claim, the Court finds the probative value of these tests to be overridden by those tests discussed supra in which testing results for RAID Max Plus in excess of 60% are readily inferable. Further, upon careful consideration of all of the tests in evidence, the Court does not regard the above-referenced tests to offset the probative value that the Court has attached to the First and Second Laboratory Tests — which in this Court's view materially undermine the reliability of the Commercial's "no more than 60%" claim.

The University Of Florida Internet Document

81. At the hearing, Clorox introduced several pages from a University of Florida sponsored Internet document in an attempt to attribute to Dr. Koehler the statement that chlorpyrifos is "slow-acting and cannot be expected to control heavy infestations of cockroaches." DX 179; see Koehler 921/2-9. The Court finds, however, that Dr. Koehler neither wrote nor approved this statement. Koehler 1338/22-1339/5. Moreover, there is no basis to believe that the anonymous author was even a competent entomologist. Clorox's own witness, Dr. Silverman, corroborated Dr. Koehler's testimony that no competent entomologist would characterize Dursban as a "slow-acting" poison. Silverman 1202/3-8; Koehler 1340/1-5. By all accounts, Dr. Koehler is not only a competent entomologist, he is universally well respected in his field. PX 9; Koehler 807/18-810/12 (publications and awards); Owens 89/1-90/2; Shapas Dep. 231/4-232/20.

J. The Egg Kill Claim

82. As previously noted, the Commercial tells consumers that SuperBait kills "just about all of them [i.e., your roaches]. And their eggs." PX 2. The parties do not dispute how cockroaches reproduce or the effect of insecticides, including hydramethylnon and Dursban, on cockroach reproduction. Within about a week after mating, the female cockroach forms a purse-shaped egg capsule — called an ootheca — that contains about 30 to 35 embryos. Under ordinary circumstances, the female carries the egg case for three to four weeks, and the new cockroaches, called nymphs, will hatch just after the female drops the egg capsule at the end of the gestation period. Owens 226/10-227/8; Boase 1528/22-1529/1. However, if the pregnant roach (called a gravid roach, Owens 96/12-16) is killed by ingesting a lethal dosage of hydramethylnon during her pregnancy, nymphs may or may not hatch from the egg capsule, depending on three factors: the gestation period, the temperature, and the humidity. Owens 227/9-228/7.

83. SCJ does not dispute that SuperBait kills some of the eggs of the roaches it kills, and that it kills a lot of those eggs where the gravid female is killed in the early stages of pregnancy. However, the very studies relied on by Clorox demonstrate that when the gravid roach dies after the sixth day of pregnancy, eggs will hatch from a substantial portion of oothecae. Thus, in a study by Dr. Silverman (DX 96), eggs hatched from 30% of the oothecae of gravid females poisoned by hydramethylnon. DX 96. Similarly, in a SCJ study, eggs hatched in 26% of the oothecae of females poisoned with hydramethylnon at days 7-8 of the gestation period, and in 96% of the oothecae where the poisoning occurred on days 13-14 of the gestation period. DX 100 at 1381. Although these studies do not measure the number of eggs that hatched at any given time, oothecal hatching normally does not occur unless more than half of the eggs are viable. Boase 1529/2-9. Moreover, yet another SCJ study demonstrates that hydramethylnon's effect on nymphal production decreased substantially as the gestation period progressed, and that hydramethylnon had a relatively insignificant effect on the number of nymphs produced when the gravid roach ingested it on days 7-8, and 13-14 of the pregnancy. DX 97. In that study, by days 7 and 8, gravid roaches who ingested hydramethylnon had produced an average of 32.6 nymphs per egg capsule (as compared with 39.5 for the control). By days 13 and 14, these numbers had increased to an average of 41.7 hatched nymphs per egg capsule (or 96% of all eggs) of gravid roaches killed by hydramethylnon compared to 41.8 for the untreated control. Id. at 150. Further, the Court does not regard the published articles that Clorox relies upon to direct a contrary result. See Def.'s Proposed Findings of Fact ¶¶ 103-107.

84. The Court recognizes that the above-referenced studies involved regular Combat, which has a lesser concentration of hydramethylnon than Combat SuperBait. Specifically, at the time that these studies were performed, regular Combat had a 1.65% concentration of hydramethylnon, see DX 97 at 147; DX 98 at 8; DX 100 at 1379, as compared to the 2% concentration of hydramethylnon present in SuperBait. Nevertheless, the Court finds the concentrations of hydramethylnon among these two Combat products to be sufficiently similar to warrant the conclusion that, even assuming SuperBait kills "just about all" of the cockroaches in a given infestation, it would not kill "just about all" of their eggs in many instances. In any event, gravid females are more likely to eat the SuperBait food while foraging for food and water in the middle and latter stages of gestation than in the early stages because they remain somewhat inactive during the first week of the gestation period. Owens 452/5-455/2. In the case of gravid roaches killed during the middle and later stages of gestation, a substantial percentage of eggs (from 26% to 96%) hatch. DX 97 at 150. This is a significant problem in terms of reinfestation, because if a significant number of viable eggs survive insecticide treatment, infestations will reoccur within a few months. See Owens 95/9-96/3.

CONCLUSIONS OF LAW^[21]

To obtain a preliminary injunction, SCJ must establish "(a) that it will suffer irreparable harm if relief is denied, and (b) either (1) a likelihood of success on the merits or (2) sufficiently serious questions going to the merits to make them fair ground for litigation and a balance of hardships tipping decidedly in [its] favor." Castrol, Inc. v. Quaker State Corp., 977 F.2d 57, 62 (2d Cir.1992) ("Castrol I"). Where the "plaintiff demonstrates a likelihood of success in showing literally false defendant's comparative advertisement which mentions plaintiff's product by name," irreparable harm will be presumed. Id. (citing McNeilab, Inc. v. American Home Prods. Corp., 848 F.2d 34, 38 (2d Cir.1988)) (other citations omitted).

I. Likelihood of Success on the Merits

A. Standards under § 43(a) of the Lanham Act

Section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a), provides in pertinent part:

Any person who, on or in connection with any goods or services ... uses in commerce any word, term, name, symbol, or device, or any combination thereof, or any ... false or misleading description of fact, or false or misleading representation of fact, which ...

(B) in commercial advertising or promotion, misrepresents the nature, characteristics, qualities, or geographic origin of his or her or another person's goods, services, or commercial activities,

shall be liable in a civil action by any person who believes that he or she is or is likely to be damaged by such act.

15 U.S.C. § 1125(a)(1)(B).

To succeed under § 43(a), "a plaintiff must demonstrate that `an advertisement is either literally false or that the advertisement, though literally true, is likely to mislead and confuse consumers. Where the advertising claim is shown to be literally false, the court may enjoin the use of the claim without reference to the advertisement's impact on the buying public.'" Castrol I, 977 F.2d at 62 (quoting McNeil-P.C.C., Inc. v. Bristol-Myers Squibb Co., 938 F.2d 1544, 1549 (2d Cir.1991)); see Coca-Cola Co. v. Tropicana Prods., Inc., 690 F.2d 312, 317 (2d Cir.1982). This is to be contrasted with allegations asserting that an advertisement is merely "ambiguous," which occurs when an advertising claim is alleged to be "literally true," but nonetheless "tends to confuse or deceive consumers." Tripledge Prods., Inc. v. Whitney Resources, Ltd., 735 F. Supp. 1154, 1164 (E.D.N.Y.1990). In the latter situation, consumer survey evidence is required. See Hertz Corp. v. Avis, Inc., 867 F. Supp. 208, 213 (S.D.N.Y.1994) ("Hertz II").

In the instant case, SCJ challenges the Commercial's advertising claims as being literally false and unsubstantiated. Accordingly, the Court must "determine, based on its own review of [the] advertisement, [whether] a literally true or a literally false message is conveyed...." Hertz II, 867 F.Supp. at 212.

A plaintiff's burden in proving literal falsity varies depending on the nature of the challenged advertisement. Where the defendant's advertisement, without making either explicit or implicit reference to tests or studies, "claims that its product is superior, plaintiff must affirmatively prove defendant's product equal or inferior." Castrol I, 977 F.2d at 63.

"Where, as in the current case, defendant's ad explicitly or implicitly represents that tests or studies prove its product superior," id., a plaintiff may meet its burden of proving literal falsity in either of two ways. First, "[i]f the plaintiff can show that the tests, even if reliable, do not establish the proposition asserted by the defendant, the plaintiff has obviously met its burden." Id. In such a situation, whether or not the tests are "sufficiently reliable" is entirely irrelevant. Id.

Second, the "plaintiff satisfies its burden by showing that the tests did not establish the proposition for which they were cited." Id. Specifically, this requires the plaintiff to show that the tests referred to by the defendant "were not sufficiently reliable to permit one to conclude with reasonable certainty that they established the proposition for which they were cited." Procter & Gamble Co. v. Chesebrough-Pond's Inc., 747 F.2d 114, 119 (2d Cir.1984). In making this determination, a fact-finder "should consider all relevant circumstances, including the state of the testing art, the existence and feasibility of superior procedures, the objectivity and skill of the persons conducting the tests, the accuracy of their reports, and the results of other pertinent tests." Id. Further, the fact that the challenged tests are flawed, in and of itself, is insufficient to meet this burden. Rather, the plaintiff bears the burden of demonstrating that the flaws in the defendant's tests were sufficiently material as to render the defendant's reliance thereon objectively unreasonable. Thus, for example, in Procter & Gamble, Chesebrough challenged the reliability of Procter & Gamble's tests on the grounds that they did not compare the currently advertised products at issue and were the product of highly questionable data manipulated to reach a predesigned conclusion. See Procter & Gamble, 747 F.2d at 117. The Second Circuit affirmed the district court's denial of a preliminary injunction notwithstanding the finding that Procter & Gamble's tests "`were far from perfect and are subject to various infirmities.'" Id. (quoting the district court); see also id. at 118 n. 5 (listing weaknesses in both sides' tests). Likewise, in L & F Products v. Procter & Gamble Co., 845 F. Supp. 984, 1001 (S.D.N.Y.1994), aff'd, 45 F.3d 709 (2d Cir.1995), the court found that the plaintiff failed to demonstrate affirmatively that defendant's tests were not sufficiently reliable even though it determined that they "were not perfect." In that case, the court recognized that a plaintiff cannot meet its burden of proof by merely exposing the methodical weaknesses of defendant's tests or by simply alleging that the tests are unpersuasive. See id.

B. The Challenged Advertising Claims Are Establishment Claims

In Castrol I, the Second Circuit held that "[a] plaintiff's burden in proving literal falsity ... varies depending on the nature of the challenged advertisement.... Where, as in the current case, defendant's ad explicitly or implicitly represents that tests or studies prove its product superior, plaintiff satisfies its burden by showing that the tests did not establish the proposition for which they were cited." Castrol I, 977 F.2d at 63 (emphasis added). In the case at bar, the challenged advertising claims communicate both explicitly and implicitly that the claims are based on scientific testing. This message is communicated explicitly by the express statement in the Commercial that "[i]n fact, testing proves Combat SuperBait kills up to 98%. The other guys, no more than 60%." Equally important, the test-proven message is communicated implicitly, yet powerfully, by the presence of a scientist in a laboratory coat who makes each of the challenged claims in the Commercial. See L & F Prods., 845 F.Supp. at 1000 ("[A] commercial may imply that tests or studies support a superiority claim. For example, a product comparison performed by actors dressed as scientists on a set appearing to be a laboratory may imply that tests or studies were conducted.").

While Clorox concedes that the "up to 98%" and "no more than 60%" claims are establishment claims, it maintains that the first three claims of the Commercial are general efficacy claims that do not refer to tests: (1) RAID Max Plus Egg Stoppers "only gets rid of some of your roaches;" (2) SuperBait "kills just about all of them;" and (3) SuperBait kills "their eggs." DX 2. According to the defendant, the Commercial makes a clear and intentional distinction between the general efficacy claims and the particular establishment claims. The general superiority claims are made in the first half of the Commercial, depicted by the first five frames of the storyboard. Those frames show a consumer, not a scientist or technician. The person doing the voiceover is not identified. By contrast, the establishment claims are made in the second half of the Commercial, depicted by the last six frames of the storyboard, showing a scientist or technician in a laboratory setting. Clorox moreover contends that a clear and intentional distinction between the two types of claims is further drawn by the words "[i]n fact" that precede the "testing proves" claims. Thus, Clorox asserts that the categories of claims are distinct and that therefore SCJ's burden of proof differs accordingly.

Clorox's argument is without merit. When the Commercial is viewed in context, it is unmistakable that all of the challenged claims are closely interrelated and that the "[i]n fact, testing proves" claims are intended to validate and amplify the earlier "kills just about all of them" and "only gets rid of some of your roaches" claims. When the presence of the scientist is factored in, there is simply no doubt that each of the challenged claims is premised on tests or studies.

C. The Applicable Legal Standards For Assessing The Meaning Of The Challenged Claims

In assessing the meaning of the Commercial, the Court is not limited to the spoken words: the visual images can be equally important. See, e.g., Coca-Cola, 690 F.2d at 318 (court examines "visual component" of ad and concludes that the visual image constitutes an explicit and false representation that defendant's packaged orange juice "is produced by squeezing oranges and pouring the freshly-squeezed juice directly into the carton"); Tripledge, 735 F.Supp. at 1164 (print advertisement false where ad showed photographs of complete windshield wipers including superstructure but where only a portion of the wiper not including the superstructure was offered for sale); Avis Rent A Car System, Inc. v. Hertz Corp., 782 F.2d 381, 385 (2d Cir.1986) ("Hertz I") (court held that picture of airport rental lot in advertisement is critical to understanding that commercial is comparing the number of cars that the litigants have available for rent, rather than the number of cars they own). Thus, the Court must examine the Commercial in context, taking into account both the audio and visual messages it communicates. See, e.g., Castrol Inc. v. Pennzoil Co., 987 F.2d 939, 946 (3d Cir.1993) ("Castrol II") ("in assessing whether an advertisement is literally false, a court must analyze the message conveyed in full context"); Hertz I, 782 F.2d at 385; Cuisinarts, Inc. v. Robot-Coupe Int'l Corp., No. 81 Civ. 731-CSH, 1982 WL 121559, at *2 (S.D.N.Y. June 9, 1982) ("in determining facial falsity the court must view the face of the statement in its entirety, rather than examining the eyes, nose, and mouth separately and in isolation from each other").

Both the intended audience and the intended product usage are relevant in understanding the meaning of the Commercial and its truth or falsity. See, e.g., Hertz I, 782 F.2d at 384-85 (in determining that advertisement compared number of cars available for rental rather than number of cars companies owned, court found relevant that both parties "have made their reputations as companies that rent cars" and that advertisement was pitched at car renters); Smithkline Beecham Consumer Healthcare, L.P. v. Johnson & Johnson-Merck Consumer Pharm. Co., 906 F. Supp. 178, 186 (S.D.N.Y. 1995) (court found intended users of advertiser's and competitor's stomach acid products — OTC rather than prescription purchasers — relevant in assessing literal meaning of advertising claim and finding it to be literally false), aff'd, 1996 WL 37325 (2d Cir. Jan. 30, 1996). Thus, the Court should use logic and common sense in deciding the literal meaning of the Commercial. See Hertz II, 867 F.Supp. at 212 ("The Court may rely on its own common sense and logic in interpreting the message of the advertisement."). In doing so, this Court has the power to determine the Commercial's "necessary implication" and meaning by considering the Commercial's spoken words and visual images, as well as the type of products it features and the intended audience for those products. See, e.g., Castrol II, 987 F.2d at 947 (finding commercial literally false based on the commercial's necessary implication); Playskool, Inc. v. Product Dev. Group, Inc., 699 F. Supp. 1056, 1060 (E.D.N.Y.1988) (same); Tambrands, Inc. v. Warner-Lambert Co., 673 F. Supp. 1190, 1194 (S.D.N.Y.1987) (defendants' advertisements "are facially false ... by necessary implication"); Cuisinarts, Inc., 1982 WL 121559, at *1 (same).

D. The Unambiguous Meaning Of The Commercial

Applying these principles, we begin with the recognition that the products featured in the Commercial are OTC roach baits that consumers apply to treat roach infestations in their homes. That the Commercial informs how SuperBait and RAID Max Plus will work when consumers use these products in their own homes is obvious not only from the Commercial's plain language but also from its visual images.

The Commercial begins with a shot of a woman in her kitchen, shortly followed by the voice of the scientist stating, while a package of RAID Max Plus is being shown, that "this only gets rid of some of your roaches" — an unmistakable reference to how the product will perform when the consumer uses it against the roaches in his or her home. Indeed, from the opening line of the Commercial — "It's you against thousands of them. Choose your weapon wisely" — to the closing shot of the scientist visiting the woman in her kitchen and advising her that "[y]ou've got to outsmart them.... [w]ith Combat SuperBait," the Commercial clearly compares how SuperBait and RAID Max Plus will perform when consumers use these products in their homes.

Clorox did not use any terms of qualification in informing consumers that SuperBait will kill "just about all of" their roaches. Accordingly, the Commercial unambiguously tells consumers that SuperBait will consistently kill nearly all of the roaches when consumers use that product in their homes. Equally unequivocal is the phrase "no more than 60%." This advertising claim unambiguously tells consumers that RAID Max Plus is incapable of reducing more than 60% of the roaches in their homes.

Turning first to consider those claims in the Commercial which standing by themselves read as general efficacy claims, it cannot be overemphasized that the context in which the Commercial makes those claims — both in its spoken words and in its visual images (including the depiction of a scientist) — is crucial to the determination of whether those claims operate as establishment claims. SCJ's post-hearing brief can be construed, however, to request that these claims should be enjoined as literally false irrespective of their context. See Pl.'s Post-Hearing Br. at 68. Consistent with the Court's finding that SuperBait is a more effective product than RAID Max Plus, the Court therefore DENIES SCJ's request to enjoin any advertised claims that (1) RAID Max Plus Egg Stoppers "only gets rid of some of your roaches;" and (2) SuperBait "kills just about all of them." DX 2. In reaching this determination, the Court is unable to conclude that SCJ has met its burden, in respect of these two claims, of affirmatively showing "that the tests, even if reliable, do not establish the proposition asserted by the defendant," Castrol I, 977 F.2d at 63, or moreover, that these statements are literally false. See id. For completeness of record, the Court further finds that, even assuming that these two claims operate as establishment claims to the fullest extent urged by SCJ, the plaintiff has failed to meet its burden of showing that the tests referred to by the defendant, in respect of these two claims, "were not sufficiently reliable to permit one to conclude with reasonable certainty that they established the proposition for which they were cited." Procter & Gamble, 747 F.2d at 119.^[22]

The analysis that follows will focus principally upon the Commercial's "up to 98%" and "no more than 60%" claims. To the extent that these two explicit establishment claims are enjoined, the parties' dispute would appear to be resolved pending a trial on the merits — at least for now—because the Commercial, as presently constituted, would not be broadcasted. This, among other things, would change the equitable landscape with respect to the requirement of irreparable harm. For completeness of record, however, the Court also will analyze, in Part F infra, the Commercial's claim that SuperBait kills "[roaches'] eggs" as it relates to the context of the Commercial.

E. The Commercial's Explicit Quantitative Claims

Upon careful consideration of the evidence, the Court concludes that SCJ has demonstrated, by the preponderance of the evidence, that it is likely to succeed on the merits of its claim that the Commercial's advertising claims that "Combat SuperBait kills up to 98%" of consumers' roaches, and RAID Max Plus kills "no more than 60%," are unsubstantiated, and therefore literally false.

1. Clorox's Failure to Substantiate the Commercial's Claim that "Combat SuperBait Kills Up to 98%" of Consumers' Roaches

Upon consideration of the evidence, the Court concludes that SCJ has met its burden of proving that the tests relied on by the defendant in respect of the Commercial's claim that "Combat Superbait kills up to 98%" of consumers' roaches "were not sufficiently reliable to permit one to conclude with reasonable certainty that they established the proposition for which they were cited." Procter & Gamble, 747 F.2d at 119. In its Findings of Fact, the Court has addressed at length why Clorox's new protocol fails to measure, with any semblance of accuracy, how SuperBait can be expected to perform in consumers' homes. It suffices to say that although the Court finds SuperBait to be a more effective product than RAID Max Plus, and considers Clorox's new testing protocol to be an effective tool for product development, these findings carry little weight in view of the ultimate issue at hand in this litigation.

It is clear that the field tests conducted under Clorox's new protocol do not reenact how consumers are instructed to use these products. To the contrary, Clorox's own documents, as well as the testimony of its witnesses, confirm that the new protocol purposely eliminates important real-world variables, such as bait depletion and the difficulty consumers have in locating hard-to-find roach harborages, that significantly affect the performance of over-the-counter roach baits when consumers use them in their homes. Indeed, in at least three separate respects — the use of a trained pest control operator employing sticky traps to locate roach harborages, the number of baits applied, and the use of multiple baits at each location where baits are placed — the new Clorox protocol differs markedly from how consumers ordinarily are capable of using, or are directed by label instructions to use, these products. These differences have a critical impact on the results SuperBait is capable of achieving; indeed, Clorox's own expert witness, Mr. Clive Boase, conceded that the new protocol significantly improves SuperBait's performance.

In addition, the single laboratory test that Clorox relies upon to corroborate the results achieved in its field tests under its new field protocol was conducted in a tiny, sweaterboxsized "complex" arena. However, S.C. Johnson's experts, other prominent entomologists not associated with this case and even Clorox's own scientists agree that laboratory tests are incapable of determining the percent reductions roach baits will achieve in home use. Indeed, Clorox's own documents confirm that its "complex" arena laboratory test is not a replacement for field studies, and merely enables Clorox to examine the basic properties of experimental products so as to determine which ones to test in the field.

In short, Clorox's laboratory and field studies do not purport to measure the accuracy of, and thus cannot establish, the Commercial's claim to consumers that when consumers use SuperBait in their homes, it will consistently kill "up to 98%" of their roaches.

Furthermore, Clorox's own field studies under its previous protocol prove the Commercial's "up to 98%" claim to be false. According to a telling pre-litigation document prepared by Clorox witness Dr. Jules Silverman (PX 45), Clorox's previous field-testing protocol does "follow[] label directions," while its new protocol does not. In Clorox field tests using the previous protocol, Superbait did not even come close to achieving near total roach elimination, and performed only at parity with RAID baits. Thus, Clorox's advertising claims could be deemed truthful only if this Court concludes that Clorox's new protocol is, to borrow Clorox's phrase, the "One True Scientific Way." See Clorox Pre-Hearing Mem. at 2. However, even Mr. Boase refused to go that far, admitting that field tests under Clorox's previous method are not scientifically invalid. Indeed, given Mr. Boase's concession that the core elements of Clorox's previous field test method are practiced by virtually all U.S. entomologists, it would be nothing short of preposterous to disregard results using that method.

In sum, the Court therefore finds that plaintiff has met its burden of proving that defendant's tests fail to substantiate the Commercial's claim that "Combat SuperBait kills up to 98%" of consumers' roaches.

2. Clorox's Failure to Substantiate the Commercial's Claim that RAID Max Plus Kills "No More Than 60%" of Consumers' Roaches

The Court also finds the Commercial's claim that RAID Max Plus kills "no more than 60%" of consumers' roaches to be unsubstantiated. The Court bases this conclusion upon Clorox's unjustifiable failure to consider data from the First and Second Laboratory Tests. Although the Court — in view of the record evidence of the parties' field tests showing the roach reduction rates of RAID Max Plus with Dursban to be consistently below 60% — is unable to conclude that this claim in fact is contradicted by the record as a whole, the Court has no difficulty in concluding that SCJ has met its burden of proving that the tests relied on by the defendant "were not sufficiently reliable to permit one to conclude with reasonable certainty that they established the proposition for which they were cited." Procter & Gamble, 747 F.2d at 119. The Court regards each of Clorox's own laboratory tests, conducted in April and June of 1995 respectively, independently to warrant this conclusion.

a. The First Laboratory Test

As previously discussed, see supra Findings of Fact ¶¶ 72-76, in the First Laboratory Test, Regular RAID (with Dursban but without Egg Stoppers) killed 80% of the roaches, while RAID Max Plus (with Dursban and with Egg Stoppers) killed 60%. As earlier stated, RAID Max Plus is actually two products contained in one package. It consists of (a) 12 roach baits, and (b) 3 small containers of the insect growth regulator Hydroprene. The hydroprene merely sterilizes certain roaches that come in contact with it. Although the Court observes that exposure to hydroprene negatively impacts upon roaches' susceptibility to chlorpyrifos (i.e., Dursban), the Court — in view of the toxic similarity between Regular RAID and RAID Max Plus — finds these effects to be insufficient to offset the circumstantial evidence suggesting that the 80% result achieved by Regular RAID substantially undermines the propriety of advertising a 60% cap for RAID Max Plus.

Further, the clear weight of the evidence suggests that Clorox's termination of the First Laboratory Test at 20 days artificially reduced the results that RAID Max Plus was able to achieve, and that had the First Laboratory Test been allowed to proceed for even a few more days, RAID Max Plus would have obtained material additional roach mortality. In this regard, the Court observes that Clorox's own documentation strongly indicates that the 20-day duration of the First Laboratory Test did not provide enough time for hydroprene to begin controlling roaches. See PX 19 at 3872 ¶ 3. Further, in view of Dr. Ochomogo's acknowledgment that laboratory tests can last up to 4 weeks, or 28 days, Clorox's failure to continue it for such duration is nothing short of astonishing. See PX 28 at 3974. The Court therefore regards this 20-day cutoff to constitute an uncompensated disadvantage to RAID Max Plus, that skewed the results of this test unfairly, and materially, against it, thereby rendering the defendant's tests significantly unreliable.

b. The Second Laboratory Test

In addition, the defendant's failure to consider RAID Max Plus' impressive performance in the Second Laboratory Test substantially undermines the reliability of the defendant's tests. As previously discussed, see supra Findings of Fact ¶¶ 25-29, the internal Clorox memorandum reporting the Second Laboratory Test shows that, after 21 days, RAID Max Plus with Dursban, on average, killed 1,112 of the approximately 1,200 roaches in each arena at that time, or approximately 92.67%. PX 39. The Court rejects Clorox's argument that the Second Laboratory Test is irrelevant to an evaluation of its "no more than 60%" claim because (i) the approximate number of roaches in the population at fixed dates, and the precise insect mortality, were reported by the Clorox scientist, and (ii) the Court has found that Dursban resistance among the German cockroach, although occurring in widespread "pockets" throughout the United States, is not pervasive. See supra Findings of Fact ¶ 54. Accordingly, the Court regards the Second Laboratory Test to be relevant to its evaluation of Clorox's claim that RAID Max Plus kills "no more than 60%" of consumers' roaches, and concludes that Clorox's failure to take this test into account materially undermines the reliability of this claim.

F. The Commercial's Egg Kill Claim

SCJ is also likely to prevail on the merits of its claim that the Commercial's "egg kill" claim is unsubstantiated and false when viewed in the context of the Commercial. The Commercial tells consumers that SuperBait kills "just about all" of their roaches. "And their eggs." Because the parties are relying on the same tests in relation to this issue, there is little dispute over testing results or methodology. Moreover, SCJ has conceded that SuperBait kills some of the eggs of the roaches it kills, and, in some instances, many of those eggs.

Hence, resolution of this claim turns upon the Court's interpretation of the message which the Commercial conveys to consumers. Viewed in context, the words "just about all" in the phrase "kills just about all of them. And their eggs" must be read to apply both to the number of roaches and to the number of eggs killed. In other words, the Commercial clearly tells consumers that SuperBait kills "just about all" of their roaches and "just about all" of their eggs. This is a materially different claim than the carefully worded statements that have appeared over the years on the parties' roach bait packages, with EPA approval, to the effect that roach baits kill roaches and their eggs. Unlike those claims, the Commercial creates the necessary implication that SuperBait will ultimately eradicate cockroach infestations in consumers' homes by killing virtually all of their roaches and virtually all of their eggs as well.

The studies relied on by both parties, however, prove that where the gravid female ingests a lethal dosage of hydramethylnon at 7-8 days into her pregnancy, egg hatch occurs in a substantial percentage of the oothecae (from 26% to 30%), and that hydramethylnon has no discernible effect whatsoever on egg hatch where the gravid roach eats the bait poison in the latter stages of pregnancy. Furthermore, an SCJ study — relied on by Clorox — achieved these results under environmental conditions which are well within the range of ordinary for cockroach populations: temperature of 80 degrees Fahrenheit and a relative humidity of 50%. DX 100 at 1379. Thus, since the testing proves that SuperBait will not consistently kill "just about all" of the eggs of the roaches it kills, the Commercial's "egg kill" claim is both unsubstantiated and false. Further, because the claim made in the Commercial is different in kind from the claims made by Clorox on package labels over the years, Clorox's contention that SCJ unduly delayed in challenging the egg kill claim is without merit. Finally, because the claim made in the Commercial is different in kind from the claims made by SCJ since at least 1987 that its Dursban roach baits without hydroprene "kill roaches and the eggs they carry," Clorox's contention that SCJ is barred from challenging Clorox's egg kill claim on the basis of "unclean hands" likewise must fail. See Nikkal Indus. v. Salton, Inc., 735 F. Supp. 1227, 1238 (S.D.N.Y.1990); Haagen-Dazs, Inc. v. Frusen Gladje, Ltd., 493 F. Supp. 73, 76 (S.D.N.Y.1980).

II. Irreparable Harm

Under the case law in the Second Circuit, SCJ has easily proven a likelihood of irreparable injury. In McNeilab, Inc. v. American Home Products Corp., 848 F.2d 34, 38 (2d Cir.1988), the Second Circuit held that where a comparative advertising claim is false, "irreparable harm will be presumed" because "[a] misleading comparison to a specific competing product necessarily diminishes that product's value in the minds of the consumer." Id.; see Castrol I, 977 F.2d at 62 (Where the "plaintiff demonstrates a likelihood of success in showing literally false defendant's comparative advertisement which mentions plaintiff's product by name," irreparable harm is presumed.) (citing McNeilab, 848 F.2d at 38). The presumption of irreparable injury from a false comparative advertisement is extremely powerful because, as the Second Circuit has recognized, such false comparisons are highly likely to diminish the plaintiff's product in consumers' eyes, see McNeilab, 848 F.2d at 38, and "eventually," though not necessarily immediately, "result in loss of sales" for the plaintiff. Vidal Sassoon, Inc. v. Bristol-Myers Co., 661 F.2d 272, 278 (2d Cir.1981).

Further, even assuming the presumption of irreparable harm does not apply, the Court still concludes that plaintiff has proven irreparable harm by a preponderance of the evidence, thereby entitling it to injunctive relief in respect of each of the Commercial's claims for which the Court has found it likely to succeed on the merits. Specifically, the evidence reflects that in the COMBAT heavy-advertised markets, SCJ has lost market share since the Commercial began airing last June. PX 90. The Court infers that this lost market share is substantially related to Clorox's Commercial, yet sufficiently speculative to call into question SCJ's ability to recover money damages at trial. In addition, the Court's finding of irreparable harm is further reinforced in view of Clorox's announced intention to begin airing the Commercial again shortly. Because national television advertisements are very expensive, there is little doubt that Clorox would not air the Commercial if it did not believe that it has been and will continue to be successful in achieving its goal. Accordingly, even if the presumption of irreparable harm did not apply in this case, SCJ still would be entitled to injunctive relief.

CONCLUSION AND ORDER

For all of the foregoing reasons, the Court enters the following orders in this action:

1. Plaintiff's motion for a preliminary injunction enjoining the Commercial from being broadcasted is GRANTED.

2. Pending the earlier of a trial on the merits or such time at which the following claims can be substantiated as demonstrated upon application to the Court, defendant is enjoined from making the following claims in any of its advertisements:

(a) COMBAT SuperBait kills up to 98% of consumers' roaches;

(b) RAID Max IV Plus Egg Stoppers kills no more than 60% of consumers' roaches; and

(c) its "egg kill" claim in the context set forth in the Commercial.

3. In view of the importance of context, the Court's preliminary injunction does not extend to the following advertised claims to the extent that the context in which these claims are made materially differs from the Commercial enjoined herein, considering the prospective advertisement's surrounding verbal and visual messages:

(a) RAID Max IV Plus Egg Stoppers only gets rid of "some" of consumers' roaches; and

(b) COMBAT SuperBait kills "just about all" of consumers' roaches.

4. The parties shall settle a proposed order on five days' notice. This preliminary injunction shall become effective upon the plaintiff's posting of an unlimited undertaking, the prerequisites for which the Court finds to exist.

5. The Court wishes to commend all counsel involved in this litigation for their professionalism and the extraordinary quality of their written submissions.

SO ORDERED.

ORDER ON RECONSIDERATION

SEYBERT, District Judge:

Defendant Clorox's motion to reconsider Paragraph 58 of the Court's Opinion and Order dated May 9, 1996 is DENIED for each of the independent reasons stated in plaintiff's opposing memorandum of law. The Court notes for the record that its statement in paragraph 58, that "Dr. Shapas was not a credible witness," does not purport to make any finding concerning said witness' character. Rather, this statement merely manifests the Court's determination not to credit Dr. Shapas' testimony with respect to the specific subject matter addressed in paragraph 58, as the Court found this testimony's probative value to be outweighed by the conflicting evidence presented at the hearing, and the inferences that the Court drew from said conflicting evidence.

SO ORDERED.

NOTES

[1] Citations to testimony at the hearing identify the name of the witness and the transcript page and line numbers corresponding to that testimony. Citations to deposition excerpts designated by a party are in the form "[witness name] Dep. [page and line number]." Citations to exhibits are in the form "PX ___" or "DX ___."

[2] Similarly, SCJ's comparative testing of Superbait and Regular COMBAT showed the former performing in substantial parity with the latter. PX 70 at 111; PX 69 at 139.

[3] Neither Dr. Frishman nor Mr. Boase were offered as experts in consumer behavior, and Mr. Boase expressly disclaimed any knowledge of the cockroach treatment experience of U.S. consumers. Boase 1634/2 — 1634/5, 1635/14 — 1635/24.

[4] Nothing on the RAID Max Plus or COMBAT SuperBait labels advises consumers to use sticky traps in order to locate cockroaches or their harborages.

[5] The designation "O.R.F." in PX 36 refers to Regular RAID. See Koehler 826/19-830/9. Although it does not possess the sterilizer ("eggs-topper") component hydroprene which is found in RAID Max Plus, regular RAID contains the identical toxic formulation as the bait component of RAID Max Plus. Koehler 827/1-829/3.

[6] In preparing his expert report for SCJ, Dr. Koehler calculated the mean and variance based on the raw data for the Second Laboratory Test and determined that RAID Max Plus killed significantly more cockroaches than Superbait in that test. The results were statistically significant at the 95% confidence level. Koehler 837/9-838/22.

[7] In an apparent attempt to minimize the difference between the comprehensive nature of his inspection and consumer capabilities and practices, Mr. Bieman testified on direct that his inspections only take 15 minutes. Bieman 1018/8-10. The Court does not regard his explanation for the very different answers at his deposition and on direct at the hearing — that he made a mistake at his deposition (Bieman 1094/25-1097/23) — to be credible.

[8] Despite the unequivocal nature of Dr. Brenner's sworn testimony, Mr. Boase insisted that Dr. Brenner in fact endorsed Clorox's new protocol and characterized Dr. Brenner's deposition testimony as "convoluted". Boase 1627/17-18. Unlike Mr. Boase, Dr. Brenner participated as a witness in this proceeding under subpoena and without compensation. He is not aligned with either side, and his testimony that Clorox's earlier brief did not present "a fair characterization of the opinion [he] intended to express in the January 6 letter" (Brenner Dep. 43/5-44/8), that he had no idea that he was being asked to lend support to an advertising claim which used the new protocol to compare two OTC roach baits (id. 47/2-12) and that he would not have signed the January 6 letter if he had known this (id.), is entirely credible. Mr. Boase's continued reliance on Dr. Brenner, however, is not.

[9] This scrutiny is underscored where, as in the instant case, the product label instructions themselves have been expressly incorporated into the testing protocol. See DX 17 at 88 ("The use of at least three grams of bait at a location is not a violation of the label. Labels do not limit how much bait can be used and indicate large infestations may need additional bait.").

[10] The Court finds that Dr. Koehler's testimony on this subject is far more persuasive than Dr. Shapas's testimony for Clorox that consumers could balance one bait station against another "on a pencil" to ascertain the amount of bait depletion. Shapas 791/14 — 791/20. While Dr. Shapas may not "really know" whether consumers "do that often" (id. 791/21-23), this Court can safely assume that they do not. Indeed, Clorox's Mr. Bieman agreed with Dr. Koehler on this point. Bieman 1137/13-16.

[11] By contrast, these factors did not have a similar "boosting" effect on the performance of RAID Max Plus because of the presence of resistant roaches in some of the apartments tested in Miami and Ocala. As Dr. Koehler testified, "it doesn't matter how close you place [the RAID Max Plus] baits to the harborages; in the apartments where the cockroaches are highly resistant, there would be no boost in performance." Koehler 854/15 — 854/18.

[12] See also Bieman 1145/13-1146/4 (multiple baiting to avoid bait depletion has "a tremendous impact on the variance and the ultimate statistical measures for that product") (emphasis added).

[13] Mr. Tomeu is the PCO who initially selected the Ocala testing site. Bieman 1041/14-1042/4.

[14] PCO use by consumers with roach problems is not very extensive. Approximately 21.3% of households with roach problems reported any use of PCOs during 1995 (PX 89, Table 8; PX 142 at Question 9), let alone the type of repeated, uninterrupted treatment with a single insecticide that is necessary to develop resistance to that insecticide. Cochran 524/13-525/5.

[15] Professor Cochran's testimony on the history of resistance is completely corroborated by his pre-litigation writings, where he made precisely the same point. PX 18 at 171-72.

[16] Clorox also elicited testimony from Dr. Frishman that Dursban use by PCOs is principally in the form of sprays, rather than baits, a fact Dr. Frishman attributed to Dursban being repellent in bait form. Frishman 658/25-659/20. The Court, however, does not regard Dr. Frishman's testimony to cast doubt on the validity of Dr. Cochran's conclusion that the continued pervasive use of Dursban by PCOs (in whatever form) is strong evidence that Dursban resistance is far from the norm and that in most types of housing, Dursban resistance is still generally not a problem.

[17] At the same time interval, RAID Max Plus with sulfluramid was also at statistical parity with Superbait, with an even higher mean reduction of 36.7%. PX 60. The plaintiff concedes that testing results involving RAID's sulfluramid product have no relevance in determining the efficacy of RAID Max Plus with Dursban (Cochran 536/19-538/8). See Pl.'s Proposed Findings of Fact ¶ 55 n. 24.

[18] Clorox offered no evidence to rebut Dr. Owens' understanding that the SuperBait product used in the Five Product Test contained the Krystar sweetener. Owens 205/6-8.

[19] Clorox relied on this test in its Prehearing Memorandum (p. 17 ¶ b) for the proposition that RAID Max Plus with Dursban achieved "a mean reduction of only 7.32%" at 12 weeks, but failed to mention that in the same test, SuperBait performed no better statistically and even worse numerically.

[20] Although Clorox takes the SCJ protocol to task for granting the cooperator some discretion in the placement of a few of the baits, the discretion is entirely consistent with label directions, as both labels counsel consumers to use more baits (12) than the number of recommended placement locations in the diagrams — 10 in the kitchen on the RAID Max Plus label, see PX 16, and 8 on the SuperBait label. See PX 111.

[21] To the extent that any part of the Court's Conclusions of Law may be construed to state a Finding of Fact, the Court hereby adopts it as such.

[22] SCJ does not argue that it should prevail in this litigation under that prong of the preliminary injunction standard which considers whether it has established "sufficiently serious questions going to the merits to make them fair ground for litigation and a balance of hardships tipping decidedly in [its] favor." Castrol I, 977 F.2d at 62.