Kawas v. Spies

IN THE UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF GEORGIA BRUNSWICK DIVISION ANNE KAWAS; and PAUL KAWAS, Plaintiffs, CIVIL ACTION NO.: 2:20-cv-138 v. JAMES SPIES; DARLENE SPIES; and DUDLEY DO SSI LLC, Defendants. O RDE R This matter is before the Court on Defendants’ Motion to Exclude Expert Testimony of H. Hal Sigman, a real estate appraiser. Doc. 37. Plaintiffs filed a Response, and Defendants filed a Reply. Docs. 41, 49. For the following reasons, the Court GRANTS Defendants’ Motion. BACKGROUND Plaintiffs’ claims concern alleged defects in a home they purchased in October 2019 from Defendants at 105 Dudley Lane on St. Simons Island, Georgia. Plaintiffs contend Defendants fraudulently concealed home defects prior to the sale—primarily yard erosion, defects with a bulkhead/sea wall, water intrusion, and mold damage. Doc. 1. Plaintiffs intend to offer testimony from a real estate appraiser, H. Hal Sigman, who opines on the fair market value of their home both with the alleged defects and without. Doc. 37-1 at 77–95. To determine the value of the home without the alleged defects, Sigman analyzed the fair market value of the home at the time of sale, assuming no defects were present. Id. at 78. Defendants do not challenge this portion of Sigman’s opinion (i.e., appraisal of the fair market value of the home assuming no defects). Doc. 37 at 3 n.1. Sigman also performed a “paired sales” analysis to determine the discounted value of the home with the alleged defects.1 In his paired-sales analysis, Sigman identified one “defected property” and one “non-defected property” and, after comparing the two properties, derived a 32% discount rate. Id. at 79. Sigman then applied the 32% discount rate to the fair market value of Plaintiffs’ home to determine the effect of the alleged defects on the home’s value. Defendants seek to exclude Sigman from testifying to the discounted value of the home—that is, the value of the home with the alleged defects. Defendants argue this testimony should be excluded because Sigman’s paired-sales analysis is not based on sufficient data and is not reliably applied. Docs. 37, 49. DISCUSSION I. Federal Rule of Evidence 702 and Daubert The United States Supreme Court’s holding in Daubert v. Merrell Dow Pharmaceutical, Inc., 509 U.S. 579 (1993), and the text of Rule 702 require trial judges to serve as gatekeepers in determining the admissibility of expert testimony; however, any decision regarding admissibility is not a position on the strength or weight of the testimony. Fed. R. Evid. 702; Kumho Tire Co. v. Carmichael, 526 U.S. 137, 141 (1999). In this Circuit, courts routinely look to three elements to determine if an expert is qualified under Daubert and Rule 702. As the Eleventh Circuit Court of Appeals has stated, the elements for consideration are whether: 1 “In a paired sales analysis, the appraiser identifies two properties that are as similar as possible except for one factor.” Atlantic Coast Pipeline, LLC (“ACP”) v. 0.07 acre in Nelson County, Virginia, 396 F. Supp. 3d 628 , 648 (W.D. Va. 2019). When the sales are compared, the difference in price is best explained by the one factor that is different. McCann Holdings, Ltd. v. United States, 111 Fed. Cl. 608, 626 (2013). (1) the expert is qualified to testify competently regarding the matters he intends to address; (2) the methodology by which the expert reaches his conclusions is sufficiently reliable as determined by the sort of inquiry mandated in Daubert; and (3) the testimony assists the trier of fact, through the application of scientific, technical, or specialized expertise, to understand the evidence or to determine a fact in issue. United States v. Frazier, 387 F.3d 1244, 1260 (11th Cir. 2004) (citations omitted). “[A]lthough there is some overlap among the inquiries into an expert’s qualifications, the reliability of his proffered opinion and the helpfulness of that opinion, these are distinct concepts that courts and litigants must take care not to conflate.” Quiet Tech. DC-8, Inc. v. Hurel-Dubois UK Ltd., 326 F.3d 1333, 1341 (11th Cir. 2003). The trial court has broad latitude in evaluating each of these three factors. As to qualifications, an expert may be qualified “by knowledge, skill, experience, training, or education.” Hendrix ex rel. G.P. v. Evenflo Co., Inc., 609 F.3d 1183, 1193 (11th Cir. 2010). The expert need not have experience precisely mirroring the case at bar to be qualified. Maiz v. Virani, 253 F.3d 641 , 665 (11th Cir. 2001). However, where an expert does have experience directly applicable to an issue at bar, experience alone may provide a sufficient foundation for expert testimony. Frazier, 387 F.3d at 1261 . “The burden of laying the proper foundation for the admission of the expert testimony is on the party offering the expert, and the admissibility must be shown by a preponderance of the evidence.” Allison v. McGhan Med. Corp., 184 F.3d 1300 , 1306 (11th Cir. 1999). However, “it is not the role of the district court to make ultimate conclusions as to the persuasiveness of proffered evidence.” Quiet Tech., 326 F.3d at 1341 . Instead, “[v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.” Daubert, 509 U.S. at 596 . II. Analysis Defendants acknowledge a paired-sales analysis is generally an accepted methodology. However, Defendants argue Sigman’s paired-sales analysis in this particular instance was not based on sufficient data and was not reliably applied, and, therefore, his testimony regarding that analysis should be excluded. Docs. 37, 49. Specifically, Defendants contend Sigman’s pairedsales analysis does not rely on sufficient data because he only considered two properties (i.e., a single pair) to derive the 32% discount rate. Additionally, Defendants assert the properties used in the paired-sales analysis are not comparable to Plaintiffs’ home, so the analysis is unreliable. A. Sigman’s Paired-Sales Analysis Is Based on Insufficient Data Defendants argue Sigman’s analysis is unreliable because he used a single pair of properties—one defected property and one non-defected property—which were not materially similar to the subject property in his paired-sales analysis. Doc. 37 at 6–8; Doc. 49 at 2–8. Defendants contend Sigman’s analysis was not based on sufficient data, and, therefore, he should not be permitted to offer testimony about the analysis at trial. Plaintiffs contend this issue goes to the weight of Sigman’s testimony, rather than its admissibility. Doc. 41. Plaintiffs argue cross-examination at trial is the proper way to test Sigman’s testimony, and exclusion is not warranted. As an initial matter, as the proponents of the expert witness, it is Plaintiffs’ burden to lay the proper foundation for the admission of Sigman’s testimony. Stewart v. Johnson, No. 5:18CV-37, 2021 WL 3030174 , at *3 (S.D. Ga. July 19, 2021), aff’d, 2021 WL 6752312 (S.D. Ga. Aug. 10, 2021) (quoting Allison v. McGhan Med. Corp., 184 F.3d 1300 , 1306 (11th Cir. 1999)). Accordingly, Plaintiffs must show Sigman’s testimony meets the requirements of Federal Rule of Evidence 702 and Daubert. Notably, Rule 702 expressly requires that an expert’s testimony be “based on sufficient facts or data.” In performing his paired-sales analysis, Sigman extrapolated the 32% discount rate from considering a sample of paired sales. Sigman used only one pair of sales as his sample (i.e., one defected property and one non-defected property). Where an expert forms his opinion based on a sample of data, challenges to the size and sufficiency of the sample are common. In some circumstances, courts have considered challenges to the sufficiency of an expert’s sample size and concluded the precise challenge raised goes only to the weight of the expert’s opinion and does not render the opinion in admissible. See, e.g., Platypus Wear, Inc. v. Clarke Modet & Co., No. 06-20976-CIV, 2008 WL 4533914 , at *6 (S.D. Fla. Oct. 7, 2008) (concluding challenge to expert’s review of a sample of agreements went to the weight of the expert’s testimony, not its admissibility). However, in circumstances where the expert’s data sample is excessively small or fundamentally flawed, courts have determined the deficiencies render the expert’s opinions unreliable, and, therefore, inadmissible. See, e.g., Brown v. Electrolux Home Prod., Inc., No. 1:08-CV-30, 2017 WL 2952281 , at *2 (S.D. Ga. July 10, 2017) (concluding expert’s reliance on a sample size of one rendered the expert’s methodology unreliable and inadmissible); ChenOster v. Goldman, Sachs & Co., 114 F. Supp. 3d 110, 124 (S.D.N.Y. 2015) (“[I]t is self-evident that [the expert’s] sample size of two pairs . . . is too small to serve as the basis for any statistical inference.”); Saxon Glass Techs., Inc. v. Apple Inc., 393 F. Supp. 3d 270 , 287 (W.D.N.Y. 2019) (concluding experts use of small sample size in conducting a survey rendered expert’s opinions unreliable and inadmissible), aff’d, 824 F. App’x 75 (2d Cir. 2020); Kilpatrick v. Breg, Inc., 613 F.3d 1329, 1337 (11th Cir. 2010) (finding no abuse of discretion where an expert was excluded because he did not explain “whether it was statistically meaningful to extrapolate from such a small sample size”). Courts have recognized a sample size of one “is rarely, if ever, sufficient.” Am. Honda Motor Co. v. Allen, 600 F.3d 813, 818 (7th Cir. 2010) (citing DeKoven v. Plaza Assocs., 599 F.3d 578 , 581 (7th Cir. 2010)); Brown, 2017 WL 2952281 , at *2. As a result, an expert’s opinion based on an incomplete, imperfect, or even weak sample is admissible and subject to cross-examination—even if the resulting opinion is “shaky”—but an opinion based on an insufficient sample is unreliable and inadmissible. See Am. Honda Motor Co., 600 F.3d at 818 (“‘Shaky’ expert testimony may be admissible, assailable by its opponents through crossexamination, but the testimony proffered here is not merely shaky: it is unreliable.” (cleaned up)). Here, the Court must determine whether Sigman’s use of a single pair of sales is sufficient to conduct his paired-sales analysis. Plaintiffs point primarily to Atlantic Coast Pipeline, LLC v. 0.07 acre in Nelson County, Virginia (“ACP”), 396 F. Supp. 3d 628 (W.D. Va. 2019). In ACP, the plaintiff argued the damages calculated from the expert’s paired-sales analysis should be excluded because, among other reasons, the expert only used seven pairs of properties, which plaintiff contended was too small of a sample size to reliably estimate the impact of the defect to the property. Id. at 648 . The court expressed concern about the sample size but ultimately declined to exclude the expert’s testimony. Id. at 649–50. The court observed the plaintiff (the party moving to exclude the expert) failed to cite any “authority providing guidance on the numbers of paired sales considered sufficient or insufficient.” Id. at 650 . In the absence of such guidance, the court concluded using a small sample size—only seven pairs—went to the weight of the expert’s testimony, rather than its admissibility. Id. ACP does not strongly support Plaintiff’s position. In ACP, the expert relied on seven pairs of sales—far more than Sigman relied on here. Even then, the court in ACP expressed concern about size of the expert’s sample. The court’s concern in ACP demonstrates there are circumstances where sample size for a paired-sales analysis is so small, or the sample so flawed, that the court should exclude an expert’s testimony under Rule 702. The ACP court did not conclude every sample size challenge is properly addressed through cross-examination. Rather, the ACP court indicated a sample size of seven paired sales was concerning, but the moving party had not shown the sample size was wholly insufficient. Defendants, on other hand, point to Jackson v. United States, 155 Fed. Cl. 689 , 715–16 (2021). Doc. 37 at 7. In Jackson, the court considered the valuation of a parcel of property in a condemnation action. One expert explained he could not conduct a paired-sales analysis because there were not sufficient sales of similar properties to compare. The court considered the expert’s position and agreed insufficient comparable sales would render any paired-sales analysis invalid, and, therefore, alternative valuation techniques were warranted. Jackson, 155 Fed. Cl. at 715–16 (crediting an expert’s opinion that described the problems with a small sample size when performing a paired-sales analysis). Although the discussion in Jackson did not involve Rule 702 or Daubert, Defendants contend the case demonstrates a paired-sales analysis is invalid if it is not based on consideration of sufficient comparable properties. Plaintiff has not shown Sigman based his paired-sales analysis on sufficient data, and, therefore, his testimony on that topic is unreliable and inadmissible. Sigman used a single pair of sales, and thus, a single data point, in extrapolating his 32% discount rate. As explained above, a sample size of one “is rarely, if ever, sufficient.” Am. Honda Motor Co., 600 F.3d at 818 . Plaintiff has failed to show Sigman’s sample size of one is sufficient. Additionally, the one pair of sales Sigman relied on involved sales of properties that were not comparable to Plaintiff’s property—e.g., condominiums as opposed to single-family homes.2 The disparate nature of these properties exacerbates the unreliability of the extraordinarily small sample size. This data is not the kind usually relied on by experts performing a paired-sales analysis. Experts typically use multiple pairs of properties and comparable properties in performing paired-sales analyses. Jackson, 155 Fed. Cl. at 715–16. In sum, Plaintiff has not shown Sigman’s opinions are “based on sufficient facts or data.” Plaintiffs try to defend Sigman’s lack of data by pointing to his testimony on the matter. Doc. 41 at 9–10. Sigman testified he found no other suitable properties to use as comparators. Doc. 37-1 at 65. The fact Sigman used the only data available does not mean that data is sufficient. Nat’l R.R. Passenger Corp. v. Heritage Forest Prod. LP, No. 404CV161, 2005 WL 6008102 , at *5 (S.D. Ga. Oct. 6, 2005) (“[J]ust because [an expert] has drawn his opinion from all of the available data does not mean his opinion is based on ‘sufficient facts or data’ as required by Rule 702.”); Player v. Motiva Enterprises LLC, No. CIV. 02-3216, 2006 WL 166452 , at *7 (D.N.J. Jan. 20, 2006), aff’d, 240 F. App’x 513 (3d Cir. 2007) (excluding an expert where he only selected the comparator property not because of its similarity to the plaintiff’s land, but because he did not know of any other cases with readily available data). Beyond these justifications for a lack of data, Plaintiffs offer little other argument that Sigman’s methodology was reliable. Thus, Plaintiffs fail to establish that Sigman’s use of a single pair—a data set of one—is sufficient data. Victory Recs., Inc. v. Virgin Recs. Am., Inc., No. 08 C 3977, 2011 WL 382743 , at *3 (N.D. Ill. Feb. 3, 2011) (excluding exert testimony where the plaintiff failed 2 The dissimilarities between the properties used in the paired-sales analysis and Plaintiffs’ home are further discussed in Section B, infra. establish to that a sample size of one is appropriate). Therefore, Sigman’s opinions based on his paired-sales analysis are unreliable and inadmissible. B. Sigman’s Paired-Sales Analysis Does Not Use Comparable Properties Defendants also challenge the reliability of Sigman’s paired-sales analysis because he did not use properties similar to Plaintiffs’ home. Doc. 37 at 8–13. Defendants point out the properties in Sigman’s paired-sales analysis do not have a bulkhead or seawall, which is an alleged defect in Plaintiffs’ case. Additionally, the properties used in Sigman’s analysis are beachfront condominiums, whereas Plaintiffs’ home is a single-family marsh-front home. Defendants argue these differences and Sigman’s failure to adjust for them renders his pairedsales analysis unreliable. Id. Plaintiffs argue Defendants’ challenge goes to the weight the testimony should be assigned rather than its admissibility, pointing again to ACP, 396 F. Supp. 3d at 628. Doc. 41 at 6–9. Plaintiffs also assert Sigman did adjust for a lack of bulkhead in the sample properties. Id. at 10 (citing Doc. 37-1 at 57–58). The court in ACP considered whether the properties in the paired-sales analysis were sufficiently similar to one another. Atlantic Coast Pipeline, LLC, 396 F. Supp. 3d at 649. The court ultimately concluded the paired-sales analysis was sufficiently reliable because the expert accounted for differences in the paired properties, such as whether they were cleared of woodlands or contained wetlands. Id. The issue in this case is different from the issue raised by Defendants. Defendants do not challenge whether the paired properties are sufficiently similar to each other, but rather, whether the paired properties and the subject property (Plaintiffs’ home) are sufficiently similar. Thus, ACP offers little guidance. As explained in Sabal Trail Transmission, LLC v. 3.921 Acres of Land in Lake Cnty., Fla., an expert must account for differences in the subject property and the paired properties for his paired-sales analysis to be reliable. No. 5:16-CV-178, 2017 WL 9939115 , at *6 (M.D. Fla. Oct. 23, 2017) (“In order for the results from a paired sales analysis to be reliable, the two properties being compared must be similar in all respects except for the characteristic being analyzed . . . .”). In Sabal, the properties used in the paired-sales analysis and the subject property were not similarly sized and were zoned differently, and the expert made arbitrary adjustments.3 Id. The court concluded an expert’s opinion based on a paired-sales analysis was not reliable because of these issues. Similarly, the paired properties Sigman considered are different from Plaintiffs’ property in nature (condominium versus single-family home) and geographic features (beachfront versus marsh-front). Sigman fails to account for the differences between the paired properties and Plaintiffs’ home. Sigman’s failure to adjust or account for these differences makes his paired-sales analysis unreliable. Id,; Mountain Valley Pipeline, LLC v. 1.40 Acres of Land, Owned by Rives, No. 7:19-CV-160, 2021 WL 1235024 (W.D. Va. Mar. 31, 2021) (excluding an expert from testifying on his paired-sales analysis where there was no evidence the properties in the paired-sales analysis were comparable to the subject property). The minimal accounting Sigman did make for the differences between Plaintiffs’ house and the paired properties does not establish reliability. Plaintiffs argue Sigman accounted for differences in the properties by adjusting for the value of the bulkhead, but Sigman’s adjustments for the bulkhead are arbitrary and unreliable. As explained above, the sample properties Sigman considered in his paired-sales analysis were two condominiums. Sigman extrapolated a 32% discount rate based on what he contends were comparable defects in those properties. Plaintiffs contend the bulkhead at their property was defective because it lacked “helical anchoring to the upland.” Doc. 37-1 at 78. But the two condominiums Sigman considered had no bulkheads. In 3 The expert in Sabal considered 45 pairs of sales in conducting his paired-sales analysis. forming his opinions about the value of Plaintiffs’ property, Sigman assigned $131,000 of the overall value of the property to the bulkhead, but that figure was based solely on Sigman’s own “knowledge of the cost to build a bulkhead.” Doc. 37-1 at 56, 78. Then Sigman applied the 32% discount rate to $131,000 figure to determine the value of the purported defect in the bulkhead (concluding the value of the bulkhead was reduced by $42,000 to $89,000). Thus, Sigman offered an ostensibly arbitrary value of Plaintiff’s bulkhead, then reduced that value for purported defects using a discount rate derived from properties without any bulkhead. This “adjustment” does not account for the differences in the properties, and the basis for conducting the calculations is arbitrary and unreliable. “A district court cannot simply accept that an opinion is reliable because the expert says that his methodology is sound.” United States v. Azmat, 805 F.3d 1018, 1042 (11th Cir. 2015) (citations omitted). An expert can extrapolate to some degree based upon available data, “[b]ut nothing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert. A court may conclude that there is simply too great an analytical gap between the data and the opinion proffered.” Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997). Sigman’s say-so as to the value of the bulkhead is not enough. Frazier, 387 F.3d at 1261–62; FTC v. Nat’l Urological Grp., Inc., No. 1:04-CV-3294, 2017 WL 6759868 , at *41 (N.D. Ga. Oct. 10, 2017), aff’d, 786 F. App’x 947 (11th Cir. 2019). Plaintiffs fail to establish reliability because the paired properties and Plaintiffs’ home have significant differences, Sigman did not account for those differences, and Sigman’s $131,000 figure for the bulkhead is unsupported. Thus, Sigman’s analysis is unreliable. Because Sigman’s analysis is not based on sufficient data and is unreliable, the Court declines to address Defendants’ other arguments for exclusion. CONCLUSION For the foregoing reasons, the Court GRANTS Defendants’ Motion to Exclude. Sigman is prohibited from testifying on his paired-sales analysis, the discount rate he derived from his analysis, and the damages he calculated to Plaintiffs’ home as a result of that analysis. SO ORDERED, this 7th day of June, 2022. BENJAMIN W. CHEESBRO UNITED STATES MAGISTRATE JUDGE SOUTHERN DISTRICT OF GEORGIA