Opinion
CIV 04-4172.
July 3, 2008
MEMORANDUM OPINION AND ORDER
Pending before the Court is Plaintiff, Henry Kayongo-Male's, motion to strike testimony on statistical analysis by Defendant, South Dakota State University's ("SDSU"), expert, Dr. Neumann. In the event the Court denies Plaintiff's motion to strike expert testimony, Plaintiff requests that the Court order Defendant to produce certain materials that he contends are necessary in order for him to adequately prepare for trial. Plaintiff also seeks to depose Dr. Neumann and additional witnesses. The discovery deadline in this case passed on September 29, 2006, and the case is currently set for a jury trial on July 22, 2008. Plaintiff requests an oral hearing on this motion.
BACKGROUND
Plaintiff moves to strike the testimony of Defendant's expert, Dr. Neumann, on two grounds. First, Plaintiff alleges that Dr. Neumann did not base his findings in his expert report on his own regression analysis, but rather on a regression analysis produced by Defendant, SDSU. Accordingly, Plaintiff argues that Dr. Neuman's testimony should be excluded under Fed.R.Evid. 702. Second, Plaintiff seeks to exclude Dr. Neumann's testimony pursuant to Fed.R.Civ.P. 37(c)(l) on the basis that Plaintiff has suffered prejudice as a result of Defendant's willful failure to disclose Dr. Nuemann's regression models and the raw data that he used to produce these models in violation of Fed.R.Civ.P. 26(a)(2)(B).
In order to remedy the prejudice Plaintiff has allegedly suffered as a result of Defendant's alleged discovery abuses, Plaintiff seeks to depose Dr. Neumann in order to ascertain the basis for his expert opinion. Plaintiff requests that the attorney's fees, expert fees, and costs arising from such deposition be borne by Defendant. Plaintiff also seeks to depose three SDSU personnel responsible for generating the data set which Dr. Neumann used to develop his regression models.
In the event that the Court denies Plaintiff's motion to strike Dr. Neumann's testimony, Plaintiff requests that the Court issue an order compelling Defendant to produce the regression models he relied upon in his report as well as an electronic copy of the raw, unfiltered data from Defendant's human resource database which Dr. Neumann used to conduct his regression analysis.
Plaintiff also seeks the Court's permission to depose three alleged key witnesses who Defendant's counsel failed to depose during the initial discovery period.
Finally, Plaintiff requests an oral hearing on this motion.
DISCUSSION
I. The Court denies Plaintiff's request to exclude Dr. Neumann's testimony under Fed.R.Evid. 702 as well as Plaintiff's request to depose Dr. Neumann.
Plaintiff alleges the Defendant's expert did not base his findings in his initial, September 2006 expert report on his own regression analysis, but rather on a regression analysis produced by Defendant. Accordingly, Plaintiff argues that Dr. Neumann's testimony should be excluded under Fed.R.Evid. 702 because he has failed to "apply the principles and methods reliably to the facts of the case."
There are several reasons why Plaintiff believes that Dr. Neumann based his findings in his initial expert report on regression models produced by Defendant. First, Plaintiff contends that the initial regression models that Dr. Neumann disclosed to Plaintiff included the same variables used in the regression models produced by Defendant about a week prior to the time Dr. Neumann disclosed his initial report, but were inconsistent with the variables that Dr. Neumann stated as having used in his initial report. Specifically, Plaintiff contends that while Dr. Neumann's initial report states that he controlled for tenure status, current rank, and total years of service at SDSU, only Defendant's regression models, and not the models initially disclosed by Dr. Neumann, control for these variables. Second, Plaintiff contends that while the coefficients and t-ratios for "white income differential" and "African American income differential" in Table 1 of Dr. Neumann's initial report differ from the models disclosed by Dr. Neumann on April 9, 2008 ( See Cutberth Aff. ¶ 19), they are consistent with the regression models produced by Defendant. Finally, as further evidence that Dr. Neumann did not base his findings in his initial report on regression models independently created by him, Plaintiff contends that both of the regression models produced by Dr. Neumann included controlled variables for "squared years in current rank" and "squared total years of service at SDSU" whereas there is no mention of controlling these variables in Dr. Neumann's initial report.
The Court does not find sufficient evidence on the record to support Plaintiff's allegation that Dr. Neumann based his finding in his initial expert report on regression models produced by Defendant. The fact that there are discrepancies between the variables used in the initial regression models produced by Dr. Neumann and the variables he reported having used in his initial report is wholly consistent with the explanation given by Dr. Neumann in his Affidavit (Doc. 96). In his Affidavit, Dr. Neumann stated that in April 2008, he realized that he failed to include "years of service" and "years in rank" in the initial regression analysis as a result of a computer programming error and thus conducted another regression analysis in which he added those missing variables. It appears that these new models were disclosed to Plaintiff on April 9, 2008, with little explanation by Plaintiff's counsel regarding the omissions in the previous models. According to Dr. Neumann, although the addition of these variables changed the coefficients and t-ratios for the "white income differential" and "African American income differential," these changes were not statistically significant and thus did not alter the opinion contained in his initial report-that there is no statistically significant difference in the salaries of whites and African Americans at SDSU. On May 16, 2008, Dr. Neumann filed a revised expert report that reflected in Table 1 the results of the new regression analysis conducted in April 2008.
Finally, although Plaintiff believes that there is a further inconsistency between Dr. Neumann's expert reports and his regression models since the models use "squared years in current rank" and "squared total years of service at SDSU," it appears that Plaintiff is incorrect in his assessment. As stated by Dr. Neumann in both his initial and revised reports, these are the two variables that were used to control for tenure status.
Overall, Plaintiff has no basis upon which to exclude Dr. Neumann's testimony under Fed.R.Evid. 702.
II. The Court denies Plaintiff's motion to strike Dr. Neumann's testimony as a discovery sanction under Fed.R.Civ.P. 37(c)(l) as well as Plaintiff's motion to compel production of the regression models produced by Dr. Neumann. However, the Court grants Plaintiff's request that Defendant produce an electronic copy of the raw data from Defendant's database that Dr. Neumann used to conduct his regression analysis. The Court finds that the raw data will permit Plaintiff to adequately prepare for trial and therefore denies Plaintiff's request to depose Dr. Neumann or the persons at SDSU who were responsible for compiling the information for Dr. Neumann.
Plaintiff also seeks to exclude Dr. Neumann's testimony as a sanction under Fed.R.Civ.P. 37(c)(l) on the basis that Defendant, in violation of Fed.R.Civ.P. 26(a)(2)(B), failed to provide the regression analysis and raw data upon which he based his opinion in his expert report.
Fed.R.Civ.P. 37(c)(l) authorizes a variety of sanctions for failure to disclose information and supplement disclosures, including exclusion of testimony. In order to impose a discovery sanction under Rule 37, there must be an order compelling discovery, a willful violation that order, and prejudice to the other party. Mems v. St. Paul, Dept. of Fire Safety Servs., 327 F.3d 771, 779 (8th Cir. 2003) (citing Chrysler Corp. v. Carey, 186 F.3d 1016, 1019 (8th Cir. 1999)). The district court has a large amount of discretion in imposing an appropriate remedy under the statute and its decision will not be reversed absent an abuse of discretion. Chrysler, 186 F.3d at 1019.
Regression Models
The Court finds that Plaintiff suffered little prejudice as a result of Defendant's failure to provide Dr. Neumann's initial regression models along with his expert report which was due on September 6, 2006, since those regression models were essentially incorrect due to the fact that Dr. Neumann's failed to include two important control variables in his analysis. Moreover, once the error was discovered, Defendant provided to Plaintiff the new regression models (Ex. 16) and revised expert report (Ex. 20) within the period designated by the Court to make supplemental disclosures and reports.
Accordingly, the Court denies Plaintiff's motion to strike Dr. Neumann's testimony as well as Plaintiff's motion to compel production of the regression materials.
Raw Data
Plaintiff requests that the Court order Defendant to produce the raw, unfiltered data from its database from which it compiled the information that Dr. Neumann used to conduct his regression analysis. Plaintiff's expert, Dr. Toutkoushian, states that this information is critical to his analysis of Dr. Neumann's opinion because "[t]he results from any faculty salary investigation can be greatly affected by decisions such as how to calculate variables in the dataset, what variables to leave out of the final dataset, and what individuals to include in the final dataset." (Toutkoushian Aff. ¶ 7.)
Defendant opposes producing this information on two different grounds. First, Defendant argues that it gave the Plaintiff in hard copy format all the information that Dr. Neumann relied on in creating his regression models. The Excel spreadsheets in Exhibits 6 and 23 show the name of each person and their affiliated salary data as well as race and affiliated salary data. The spreadsheets that Plaintiff has, however, does not show the attachment of names to the race identifiers. Dr. Neumann states that names did not factor into his analysis since at SDSU, the process of gathering race data is purely one of self-identification and all that mattered to the analysis was how each person identified themselves as shown in the database.
The second ground upon which Defendant declines to produce this information is that it claims that how personnel choose to identify themselves with a particular race is confidential, private information. Accordingly, Defendant is opposed to producing this information absent a court order. In response, Plaintiff notes that although Defendant has not produced names and affiliated race data on one spreadsheet, Defendant has essentially already disclosed this information since the data contained in Exhibits 6 and 23 permits the Plaintiff to identify each faculty member with his or her affiliated race. ( See Ex. 24.)
While the Court finds that excluding Dr. Neumann's testimony at trial for failing to produce the raw data is an unnecessary sanction, the Court orders Defendant to produce the raw data from its database in electronic format from which it compiled the information Dr. Neumann used to conduct his regression analysis. The Eighth Circuit permits liberal discovery of "all information disclosed to a testifying expert in connection with his testimony, whether or not the expert relies on the documents and information in preparing his report." Kooima v. Zacklift Int'l, Inc., 209 F.R.D. 444, 446-47 (D.S.D. 2002) ( quoting In Re Pioneer Hi-Bred Int'l, Inc., 238 F.3d 1370, 1375 (Fed. Cir. 2001) (applying Eighth Circuit law)); see also Monsanto Co. v. Aventis Cropscience, N. V., 214 F.R.D. 545 (E.D. Mo. 2002). Moreover, Plaintiff is correct that the manner in which someone filters raw data can drastically affect the results of a regression analysis. See Cooper v. S. Co., 390 F.3d 695, 717-18 (11th Cir. 2004) ( quoting Eastland v. Tenn. Valley Auth., 704 F.2d 613, 621-24 (11th Cir. 1983) (stating that if the tested disparity in multiple regression analysis is based on erroneous assumptions or suffers from flaws in the underlying data, then standard deviation analysis is foredoomed to yield an equally faulty result.").
The Court further concludes that Plaintiff will be able to adequately prepare for trial with the electronic information from Defendant's database and need not depose Dr. Neumann nor the persons at SDSU who were responsible for compiling the information for Dr. Neumann.
III. The Court denies Plaintiff's request to depose three additional witnesses.
Plaintiff's attorney seeks to depose certain key witnesses whom Plaintiff's prior counsel failed to depose. The first person Plaintiff seeks to depose is Thomas Cheesbrough who is the decision-maker in rating Dr. Kayongo-Male. Plaintiff's attorney also seeks to depose Dr. Carol Peterson who was a decision-maker on Plaintiff's discrimination complaint and who made key statements regarding SDSU underpaying Plaintiff. Finally, Plaintiff's attorney seeks to depose Karyn Weber, the Director of Human Resources, who has offered affidavit testimony in the matter and played a key role in decision-making in this matter.
The Court denies Plaintiff's request to depose these additional witnesses. The discovery deadline has long since passed and the Court need not reopen discovery in order to accommodate the different litigation tactics of Plaintiff's current counsel.
IV. The Court finds that the facts are clear enough from the briefs to issue a ruling on Plaintiff's motions and denies Plaintiff's request for an oral hearing.
For the foregoing reasons, the Court hereby ORDERS:
1. Plaintiff's Motion to Strike Expert Testimony or in the Alternative Motion to Compel (Doc. 91) is GRANTED in part and DENIED in part. a. The Court DENIES Plaintiff's request to strike the testimony of Defendant's Expert, Dr. Neumann. b. The Court DENIES Plaintiff's request to depose Dr. Neumann. c. The Court DENIES Plaintiff's request to depose the persons at SDSU who were responsible for compiling the information from Defendant's database the Dr. Neumann used to create his regression models. d. The Court DENIES Plaintiff's request to compel disclosure of Dr. Neumann's regression models since the models Plaintiff received on April 9, 2008, accord with Dr. Neumann's revised expert report. e. The Court GRANTS Plaintiff's request to compel the production of raw data from Defendant's database from which Defendant compiled the information that Dr. Neumann used to conduct his regression analysis. The Court ORDERS that this information be disclosed to Plaintiff in electronic format no later than Thursday, July 10, 2008, at 5:00 p.m. f. The Court DENIES Plaintiff's request to depose Thomas Cheesbrough, Dr. Carol Peterson, and Karyn Weber. 2. Plaintiff's request for an oral hearing (Doc. 98) regarding this motion is DENIED.