Roty v. Battelle Mem. Inst.

2020 Ohio 4389

• Court: Ohio Court of Appeals
• Decided: September 10, 2020
• Docket: 18AP-956
• Status: Published

Opinion

[Cite as Roty v. Battelle Mem. Inst., 2020-Ohio-4389.]

IN THE COURT OF APPEALS OF OHIO

TENTH APPELLATE DISTRICT

Ann Roty et al., :

Plaintiffs-Appellants, : No. 18AP-956 v. : (C.P.C. No. 14CV-2502)

Battelle Memorial Institute et al., : (REGULAR CALENDAR)

Defendants-Appellees. :

D E C I S I O N

Rendered on September 10, 2020

On brief: Law Offices of Russell A. Kelm, Russell A. Kelm, and Ian M. Kelm, for appellants. Argued: Russell A. Kelm.

On brief: Ice Miller LLP, James E. Davidson, and Catherine L. Strauss, for appellees. Argued: Catherine L. Strauss.

APPEAL from the Franklin County Court of Common Pleas

NELSON, J. {¶ 1} Now claiming a right to information that they didn't request before the end of a generous discovery period and that they didn't seek when they applied for and obtained an earlier order from this court that opened the way to a trial, plaintiffs-appellants Ann Roty and Mary Neff (together, sometimes "appellants" or "plaintiffs") ask us to set aside the jury verdict in favor of defendant-appellee Battelle Memorial Institute and reinstate their case. The record does not justify that course, and we decline to take it. {¶ 2} Ms. Roty and Ms. Neff acknowledge that "[t]his case has a lengthy procedural history." Appellants' Brief at 2. We boil that history down a bit by noting that they were let go by Battelle in 2013 as part of a Reduction in Force ("RIF") and on March 7, 2014 sued Battelle and various supervisors (together, "Battelle"), alleging claims of age discrimination under Ohio law. Discovery proceeded for over a year and a half, concluding pursuant to schedule in September of 2015. The trial court granted summary judgment to Battelle on No. 18AP-956 2

April 5, 2016, finding that under the facts as viewed in the light most favorable to Ms. Roty and Ms. Neff, they had failed to present evidence to show that Battelle had terminated them because of their ages (of 50 and 49, respectively). Ms. Roty and Ms. Neff appealed. {¶ 3} In our decision of December 19, 2017, the panel majority observed that "[d]uring the course of discovery, Roty and Neff moved the trial court for an order compelling Battelle 'to produce discovery requested by plaintiffs relating to the statistics of the ages and positions of those included and not included in [the] reduction in force,' " and that the trial court had limited that discovery to the business unit level and not permitted the collection of those age statistics company wide. Roty v. Battelle Mem. Inst., 10th Dist. No. 16AP-266, 2017-Ohio-9125, ¶ 2, 4. The panel majority further reported that Ms. Roty and Ms. Neff "argue that this limited discovery was insufficient to permit substantially any comparison between the ages of terminated versus retained employees." Id. at ¶ 9. The panel majority agreed with that point, holding:

The decisions of the trial court on the motion to compel and on the motion for summary judgment (which was decided based on a record without the statistics sought by Roty and Neff), must be reversed and the case remanded so that the trial court can order Battelle to produce company-wide employee age statistics from the period of the relevant reduction in force showing both employees who remained at the company following the reduction in force and those who were terminated pursuant to the reduction in force.

Id. at ¶ 18 (emphasis added). The dissent agreed on the formulation of that issue, even while disagreeing as to the result. See id. at ¶ 25 (dissent finds that "the record supports the trial court's decision to deny Roty and Neff's request to compel the production of company-wide age and position data"). {¶ 4} The matter therefore returned to the trial court, which conducted a status conference on January 25, 2018. At that conference, as memorialized in an order issued the next day, the trial court established a trial date of October 22, 2018 and ordered Battelle to produce within two weeks the company-wide age statistics that Ms. Roty and Ms. Neff had sought and moved to compel and that we had said should be turned over. January 26, 2018 Order Regarding Status Conference at 1-2 (also setting other dates and permitting limited further discovery concerning experts and "plaintiffs' updated damages given the age of the case," id. at 2). No. 18AP-956 3

{¶ 5} Also at that status conference, counsel for Ms. Roty and Ms. Neff "orally moved to expand the required production to include information regarding the 'training/education, special skills, experience, factors used to decide and rank employees, those who were terminated, those who were not terminated, and statistics for the year after' the August/September 2013 RIF." Id. at 2. We will cut through some of the back and forth in the briefing here to express our understanding that by this perhaps rather amorphous request, plaintiffs say they sought not so much to learn what "factors [Battelle claimed had been] used to decide and rank employees"—Ms. Roty and Ms. Neff essentially concede that they already had been provided during discovery the specific factors considered in the matrices by which Battelle ranked employees in their respective departments, including " 'the relevant skills evaluated' " and how Battelle "rated Appellants against their peers," see Reply Brief at 5-6, quoting Appellees' Brief at 21—but, rather, how those factors translated statistically across groups of Battelle employees. They were asking, they tell us, "for statistical evidence pertaining to the independent factors required by" two decades of this court's precedents. Reply Brief at 1-2. {¶ 6} After considering "arguments on the topic," the trial court declined to reopen discovery so broadly. January 26, 2018 Order regarding Status Conference at 2. Counsel for Ms. Roty and Ms. Neff were "unable to point the Court to any portion of the record where plaintiffs had [over the previous four years] requested such information, and the Court * * * was unable to find any such requests itself." Id. The trial court reaffirmed that ruling in its February 5, 2018 Order Regarding 2/1/18 Status Conference, and in an April 3, 2018 Entry and Order (after confirming that Battelle had produced the requested company- wide employee age statistics). And it elaborated on the ruling as trial began, emphasizing that what we in deference to Ms. Roty and Ms. Neff will call the "statistical" request regarding independent factors had not been made for four years and until "long after the discovery deadline had passed." October 22, 2018 Decision and Entry Granting Defendants' Motion in Limine to Exclude the Testimony of William Notz at 2 (emphasis in original). {¶ 7} At that same time, the trial court indicated that it was not inclined to allow the testimony of William Notz, whom plaintiffs put forth as an expert on statistical analysis. Ms. Roty and Ms. Neff had acknowledged, the trial court noted, that under the governing age discrimination framework, " 'statistical evidence needs to account for any other No. 18AP-956 4

independent variable that might explain the association between age and termination rates, including job skills, education, experience, or self-selection.' " Id. at 4-5, quoting October 2, 2018 Opposition to Motion in Limine at 5. Because they "also admit[ted] that Notz did not consider the independent factors," his testimony would not have been helpful to the jury and indeed had been discounted on that very ground in another case more than a decade earlier. Id. at 5, citing Boggs v. Scotts Co., 10th Dist No. 04AP-425, 2005-Ohio-1264 (also same plaintiffs' counsel as here); see also Decision & Entry Granting Motion in Limine at 3, citing Swiggum v. Ameritech Corp., 10th Dist. No. 98AP-1031, 1999 Ohio App. Lexis 4634 (Sept. 30, 1999) (decision to admit statistical evidence was unduly prejudicial where it failed to consider independent variables including skills, education, experience, performance, or voluntary retirement; same plaintiffs' counsel as here). Not only did plaintiffs' counsel and their purported expert both have "established, personal knowledge of the required independent factor analysis," the trial court observed, they had had "substantial time and opportunity to request the independent factor data" in this case but had failed to do so: "to repeat, Plaintiffs' counsel, and their exact same expert, knew the independent factor information was required but their counsel inexplicably did not request it until after the discovery deadline had long since passed." Decision & Entry Granting Defendants' Motion in Limine at 5 (emphasis in original). {¶ 8} Trial began on October 22, 2018. During the course of the trial and outside the presence of the jury, Ms. Roty and Ms. Neff did indeed proffer the testimony of William Notz, Ph.D. They quote from his proffer at some length in their briefing to us. Appellants' Brief at 31-33. Those quotations underscore that he was emphatic that no assessment of factors other than age—regardless of what those other statistics might show—would or could alter his conclusion that age was the statistically significant variable for the RIF. As plaintiffs themselves put it, his proffered testimony was "that any correlation found when considering the independent factors would be between the other factor and age, not in the other factor and selection for termination." Reply Brief at 13. The trial court excluded Professor Notz as a witness. {¶ 9} "On November 1, 2018, the jury returned a verdict, by a vote of 6-2, in favor of Battelle." November 16, 2018 Final Entry. In this appeal, Ms. Roty and Ms. Neff posit two assignments of error: No. 18AP-956 5

[1.] THE TRIAL COURT ERRED IN REFUSING TO COMPEL PRODUCTION OF STATISTICAL DATA REGARDING THE REDUCTION IN FORCE IN WHICH PLAINTIFFS WERE TERMINATED.

[2.] THE TRIAL COURT ERRED IN GRANTING DEFENDANTS' MOTION IN LIMINE AND EXCLUDING THE EXPERT TESTIMONY OF DR. WILLIAM I. NOTZ WITH REGARD TO STATISTICAL ANALYSIS OF THE 2013 REDUCTION IN FORCE TO THE SUBSTANTIAL PREJUDICE OF PLAINTIFFS.

Appellants' Brief at 1. {¶ 10} We review the trial court's conduct of discovery and its rulings with regard to the admission of evidence under an abuse of discretion standard. " 'In discovery practices, the trial court has a discretionary power, not a ministerial duty.' Thus, the standard of review of a trial court's decision in a discovery matter is whether the court abused its discretion." Mauzy v. Kelly Servs., 75 Ohio St.3d 578, 592 (1996) (citations omitted) (adding that " 'an appellate court will reverse the decision of a trial court that extinguishes a party's right to discovery if the trial court's decision is improvident and affects the discovering party's substantial rights' " [citation omitted]; trial court's suggestion there that party's replacement counsel had been provided opportunity for discovery was simply "illusion"); see also Jelinek v. Abbott Laboratories, 10th Dist. No. 17AP-576, 2019-Ohio- 3860, ¶ 20 (trial court with " 'broad discretion in regulating the discovery process,' " quoting Zimpfer v. Roach, 3d Dist No. 17-17-03, 2017-Ohio-8437, ¶ 27). Similarly, as Ms. Roty and Ms. Neff acknowledge, "[d]ecisions regarding the admissibility of evidence are reviewed on an abuse-of-discretion standard." Appellant's Brief at 11, citing Estate of Johnson v. Randall Smith, Inc., 135 Ohio St.3d 440, 2013-Ohio-1507, ¶ 22, citing State v. Hancock, 108 Ohio St.3d 57, 2006-Ohio-160. And as they further note, " '[f]or an abuse of discretion to have occurred, the trial court must have taken action that is unreasonable, arbitrary, or unconscionable.' " Appellants' Brief at 12, quoting Stewart v. Vivian, 151 Ohio St.3d 574, 2017-Ohio-7526, ¶ 34, reconsideration denied, 151 Ohio St.3d 1459, 2017-Ohio-8842; see also, e.g., Estate of Johnson at ¶ 22. {¶ 11} The trial court did not abuse its discretion in declining to allow Ms. Roty and Ms. Neff to engage in the broad new discovery they first sought only after the discovery period had closed and after the particular discovery ruling they contested had been No. 18AP-956 6

appealed and resolved. Lacking direct evidence of discrimination, they endeavored to proceed based on statistical inferences, trying to make a "comparison between the ages of terminated versus retained employees" to show a "disparate impact." Roty at ¶ 9, 15. It was in that effort that they requested, during discovery and through their first appeal, to obtain "company-wide employee age statistics." Id. at ¶ 9, 18. But—"inexplicably," in the trial court's view—they made no effort to gather "independent factor" information "until after the discovery deadline had long since passed" and their initial appeal was concluded. October 22, 2018 Decision & Entry at 5. And nothing in the discovery they won through appeal and that was promptly furnished to them uncovered some previously hidden or even unknown need to explore what they call their "requests for statistical evidence pertaining to the independent factors required by Swiggum v. Ameritech Corp., [10th Dist.] No. 98AP-1031, 1999 WL 771022 (1999); Dahl v. Battelle Mem. Inst., 10th Dist. * * * No. 03AP- 1028, 2004-Ohio-3884, (2004); Boggs v. Scotts Co., 10th Dist. * * * No. 04AP-25, 2005- Ohio-1264, (2005); Mittler v. OhioHealth Corp., 10th Dist. * * * No. 12AP-119, 2013-Ohio- 1634, (2013); and most recently Jelinek v. Abbott Laboratories, 10th Dist. * * * No. 17AP- 576, 2019-Ohio-3860 ('Jelinek II') * * *." See Reply Brief at 1-2. {¶ 12} Except for Jelinik II, every one of those cases cited by Ms. Roty and Ms. Neff was decided years before the discovery period here came to an end. Even had plaintiffs' own counsel not been so intimately involved in Swiggum, Dahl, and Boggs and even had their expert not been discounted in Boggs (and then, at roughly the same time as summary judgment was proceeding here, excluded by the trial court in Jelenik II) for his failure to address independent factors in offering a statistical analysis, see October 22, 2018 Decision & Entry at 4-5, plaintiffs would not have been relieved of their responsibility to pursue with some diligence discovery relating to their claimed "disparate impact" case. "It is incumbent on the party seeking discovery to diligently pursue * * * discovery." Levens Corp. v. Aberth, 9th Dist. No. 13927, 1989 Ohio App. Lexis 2845, *9 (July 19, 1989). {¶ 13} Ms. Roty and Ms. Neff fully acknowledge (and complain) that "Swiggum, Dahl, Boggs, Mittler, and Jelinek all require a statistical review of independent factors to be included in the expert's analysis." Reply Brief at 9. They concede that under the longstanding framework, "independent variables are required by this Court to be considered to present a statistical analysis to a jury." Appellant's Brief at 20, citing Swiggum. And they knew the factors by which Battelle said it organized the relevant No. 18AP-956 7

matrices ranking employees for the RIF. See, e.g., Reply Brief at 3 ("Battelle provided" matrices showing the various independent factors on which Ms. Roty and Ms. Neff were evaluated); id. at 5-6; see also, e.g., December 4, 2014 testimony of Marcia Unger in transcript filed February 19, 2015; September 23, 2015 Plaintiffs' Memo Op. MSJ and Exhibits 6-8, 17-22, 25-28. Yet they still offer no explanation (other than their view that the precedents themselves are "flawed," see Reply Brief at 1) of why from the spring of 2014 on they did not seek discovery regarding those "independent variables"—which they call "vital information," Appellants' Brief at 20—until their oral request of January 25, 2018. {¶ 14} Under these circumstances, we cannot label the trial court's 2018 discovery ruling "improvident." From the date they filed their case, Ms. Roty and Ms. Neff to proceed on their planned course had to assess statistical proof; they never undertook, however, to assemble the data they say they needed for that assessment while the discovery period wore on (and as they were seeking part of the equation in the form of the company-wide age statistics that they did request and that a divided panel of this court ordered to be produced). The trial court on the facts of this case did not abuse its discretion in not compelling discovery that was not requested during the lengthy discovery period. Compare, e.g., Cameron v. State Teachers Retirement Bd., 10th Dist. No. 00AP-425, 2000 Ohio App. Lexis 5539, *17-18 (Nov. 30, 2000) ("at the time of the discovery cut-off date [some 42 weeks after the case was filed], the record fails to indicate that plaintiff had made any attempt to depose any other witnesses to establish plaintiff's contention that Cameron was not competent * * *. Under the circumstances of this case, including the length of time plaintiff had to conduct discovery, we cannot conclude that the trial court's decision to overrule plaintiff's motion to compel and extend discovery constituted an abuse of discretion"). {¶ 15} Nor did the trial court abuse its discretion in exercising its evidentiary gatekeeping function to exclude the testimony of Professor Notz. Indeed, our precedent as flagged and contested by Ms. Roty and Ms. Neff required the trial court to do so. {¶ 16} As the trial court thoroughly explained, in Swiggum, "[t]he Tenth District * * * held the trial court's decision to admit Swiggum's statistical evidence [suggesting a correlation between age and job terminations] amounted to prejudicial error because that evidence: No. 18AP-956 8

failed to take into consideration any independent variables that might explain the association between age and termination rates, including job skills, education, experience, performance or self-selection. As noted, Swiggum's evidence did not even attempt to distinguish between employees who were involuntarily discharged and those who voluntarily exited the company * * * *.

Consequently, the Court concluded the statistical evidence lacked probative value and should not have been admitted because it did not include the independent factor analysis." October 22, 2018 Decision & Entry at 3-4, quoting Swiggum, 1999 Ohio App. Lexis 4634 at *45. {¶ 17} The trial court also correctly quoted from Dahl to the same analytical effect. Id. at 4, quoting 2004-Ohio-3884, ¶ 18 ("raw statistics presented by [Dahl] simply do not present a sufficient level of specificity to support the inference of discrimination. Appellant's statistics do not account for variations among employees in skill level, job function, and education as variables that necessarily must be accounted for in establishing a probative value of the raw numbers given."). {¶ 18} And it also properly cited Boggs, id., which upheld a trial court's determination that statistical evidence propounded by Professor Notz in that RIF case was not probative of discriminatory intent because "Notz's affidavit indicates that he only looked at the number of persons above and below the age of 40 who were selected for termination. * * * * It may be that some of the departing employees over the age of 40 were terminated for legitimate reasons, such as poor performance or lack of appropriate skills. * * * [T]hey may have accepted early retirement. Other factors may have been involved, but looking only at the ages of those affected by the reduction in force does not indicate what those other factors were. Thus, [Professor Notz's] statistical evidence was not probative of whether [the defendant there] decided to terminate [plaintiff's] employment based solely upon her age, and we agree with the trial court that summary judgment for [defendant] on this question was proper," 2005-Ohio-1264, at ¶ 18. {¶ 19} We observe that Professor Notz reappeared in Jelinek II, where he again sought to opine on statistical significance between age and RIF terminations and where he again acknowledged that his analysis of the statistics "only looked at the ages of the individuals terminated." 2019-Ohio-3860, ¶ 33, 35. In a March 2016 decision—a decision made before plaintiffs' took even their first appeal in this case—"the trial court relied on No. 18AP-956 9

Boggs and its progeny to find there was 'no probative value in the presentation of raw statistics,' and the court [held] that * * * '[a]ny testimony to this effect that fails to take into account the variations among employees in skill, etc., should be prohibited.' " Id. at ¶ 38. There, too, the plaintiffs urged that the precedents are wrong (" 'flawed' ") and "that multiple regression analysis often gives the wrong answers." Id. at ¶ 36. And there, too, we found "no error with the trial court's determination that the statistical evidence offered by appellant did not meet the criteria required by this court's prior decisions, including Boggs": we "adhere[d] to this court's precedent that 'unelaborated statistics that fail to consider independent variables' are not probative of discrimination." Id. at ¶ 38. {¶ 20} In sum, the trial court here had no choice but to exclude the proffered Notz testimony. Ms. Roty and Ms. Neff concede that without the independent factor information (that they had failed to seek during the discovery period), "Dr. Notz was unable to analyze the data as required by precedent in this Court." Reply Brief at 4. That fact entirely undermines their second assignment of error. {¶ 21} Ms. Roty and Ms. Neff do contend in their briefing that the serial precedents that guided the trial court are "flawed," and we will briefly address their concern even while noting that the trial court was (a) obliged to follow that precedent, and (b) would have been well within its discretion to exclude the proposed testimony of Professor Notz in any event. We are constrained to observe that appellants are not helped in this regard by the notion we understand to have been suggested at oral argument that what a company ought to do during a RIF is simply make "adjustments" to terminations "so there's no disparate impact" by age, without regard to further relative consideration of education, skills, voluntary retirements, or the like. In contrast to that approach, the guiding precedents help ensure that statistics used are of some " 'probative value' " to provide " 'a sufficient level of specificity to support the inference of discrimination' " as required for a case under the Ohio statutes: "looking only at the ages of those affected by the reduction in force does not indicate what those other factors were * * * * [and] [t]hus * * * [is] not probative of whether [an entity] decided to terminate * * * employment based * * * upon [an employee's] age." Boggs at ¶ 16, 18 (citations omitted). {¶ 22} First, and despite any contrary implication, see Reply Brief at 13-14, our precedents are not in conflict with Ahern v. Ameritech Corp., 137 Ohio App.3d 754, 773-74 (8th Dist.2015). There, in a case not involving a RIF and presenting a mix of direct and No. 18AP-956 10

statistical evidence, the court of appeals found that there had been no showing of "material adverse prejudice" from any error in admitting statistical evidence at trial. "A trial court's ruling concerning the admissibility of evidence will not be reversed on appeal absent a clear abuse of that discretion which results in material adverse prejudice," the court noted. Id. at 773 (citations omitted). That standard and the appeals court's deference to the trial court in no way benefits Ms. Roty and Ms. Neff. {¶ 23} Next, the federal Older Workers Benefit Protection Act doesn't have anything to do with this case. Compare Reply Brief at 15. Ms. Roth and Ms. Neff cite 29 U.S.C.(f)(1)(F)(ii) and (H), but those provisions (1) are not the (Ohio) statutes under which plaintiffs brought their claims; (2) relate exclusively to employee "waiver[s]" unrelated to any issues in this appeal; and (3) apply only to the release of claims in exchange for severance pay, whereas Ms. Roty and Ms. Neff concede that "Battelle provided severance without requiring a release of claims." See Appellants' Brief at 28. {¶ 24} And the federal Seventh Circuit case from 2000 that appellants cite in their opening brief, Adams v. Ameritech Servs., Inc., 231 F.3d 414 (7th Cir.2000), naturally does not purport to interpret Ohio law or call into question our analysis in Dahl (2004), Boggs (2005), or Jelenik II (2019). We have affirmed the need for independent factor assessment several times since Swiggum, and recently, and we discern no persuasive reason to depart from the well established analytical framework. {¶ 25} We also are not persuaded by appellants' contention that Professor "Notz's proffered evidence did consider other factors and found those factors to have no bearing on his ultimate conclusion that age was a determining factor in selection for the 2013 RIF at Battelle." Reply Brief at 17 (providing no citation). Professor Notz frankly conceded that he had "looked just at age." October 25, 2018 (proffer) Tr. at 462. So we take it that what appellants mean here is that he dismissed out of hand any evaluation of the independent factors invoked by Battelle, even without being privy to the relevant data on those matters. {¶ 26} Thus, for example, when he was asked during his proffer whether "if you ranked all of these employees by job skills and did a statistical analysis of that, would that change your assessment that the age was a statistically significant issue in this RIF?", he insisted: "No, no." Appellants' Brief at 31, quoting Tr. at 459 (further suggesting that if [potentially low?] job skills were found statistically significant for the job termination, "what that would tell me is that age and job skills are very highly correlated," perhaps No. 18AP-956 11

inversely in that hypothetical). Nor would other factors such as "performance" or "self- selection" sway his analysis of "the factor of age in this reduction of force." Id. at 32-33, quoting Tr. at 461-62. Although he had not examined them, those other factors, in his estimation, would all have to be "a proxy for age" to the extent that they correlated with termination. Tr. at 462. We suppose that perhaps this is all true enough in a highly generalized sense, but Ms. Roty and Ms. Neff do not begin to argue that the independent factors invoked by Battelle somehow would be inappropriate considerations in a RIF, and we in no way could say that the exclusion of the proffered testimony as confusing or unhelpful would even have approached being an abuse of the court's evidentiary gatekeeping function in a case of this nature. See, e.g., Columbus v. Swanson, 10th Dist. No. 18AP-524, 2020-Ohio-357, ¶ 26 (trial court did not err in excluding expert testimony that posed substantial danger of confusion). {¶ 27} In any event, and as Ms. Roty and Ms. Neff put it, "[t]his Court requires the expert opining on a statistical analysis to consider several 'independent factors' to rule out the most common nondiscriminatory reasons for termination." Reply Brief at 15; see also id. at 8 (cases "require the statistician to consider in his or her analysis the statistics of the independent factors in concluding that age was a determining factor in the RIF"), 9 (multiple precedents "require a statistical review of independent factors to be included in the expert's analysis"). Professor Notz's analysis did not do that, and the trial court did not abuse its discretion in excluding it. {¶ 28} We overrule the two assignments of error and affirm the trial court's judgment as based on the jury's verdict. Judgment affirmed. KLATT and BRUNNER, JJ., concur.