Lawson v. Spirit Aerosystems, Inc.

Opinion

IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF KANSAS

LARRY A. LAWSON, ) ) Plaintiff, ) ) v. ) Case No. 18-1100-EFM-ADM ) SPIRIT AEROSYSTEMS, INC., ) ) Defendant. )

MEMORANDUM AND ORDER

This matter comes before the court on Defendant Spirit AeroSystems, Inc.’s (“Spirit”) Motion to Shift Costs of Technology Assisted Review of ESI to Plaintiff Larry A. Lawson (“Lawson”). (ECF 133.) At Lawson’s request, the parties spent months engaged in an ESI discovery process regarding the issue of business overlap between Spirit and non-party Arconic, Inc. (“Arconic”) using traditional ESI methods involving custodians and search terms. When that process repeatedly yielded low responsiveness rates, the court allowed the parties to proceed— again, at Lawson’s request—with a technology-assisted review (“TAR”) of approximately 322,000 documents, with the caveat that the court would decide whether to allocate the TAR expenses to Lawson. Spirit now moves the court to require Lawson to pay Spirit its costs and expenses for the TAR process pursuant to Federal Rule of Civil Procedure 26(c). As explained below, Spirit’s motion is granted. The court is mindful of the default rule that the producing party should ordinarily bear the costs of production, but the court finds good cause to allocate the TAR expenses to Lawson in order to protect Spirit from undue burden and expense. Early in the case, Lawson pursued a scattershot ESI approach on the issue of Spirit’s “Business,” and the court repeatedly cautioned Lawson to better focus his ESI custodians and search terms because the court would, at some point, begin shifting costs. Spirit has already borne its fair share of expenses providing discovery on this subject matter by accommodating Lawson’s ESI requests for the custodians and search terms he selected, by running court-ordered sampling exercises, and by making targeted document productions on a separate path than the ESI process. That ESI process repeatedly yielded low responsiveness rates. But Lawson was unwilling to abandon the largely non-responsive ESI dataset and instead sought continued review via TAR that

unnecessarily perpetuated and exacerbated ESI/TAR expenses. The TAR process ultimately yielded a responsiveness rate of only 3.3%. Even the documents that were technically responsive were of marginal (if any) relevance above and beyond what Spirit produced outside of the ESI/TAR process. Thus, the ESI/TAR process became disproportionate to the needs of the case. The parties are directed to meet and confer to try to reach agreement on the amount of the TAR expenses. In the event they are unable to reach agreement, the court orders further briefing as to what dollar amount the court should award, as set forth below. I. BACKGROUND The background of this lawsuit is more thoroughly set forth in this court’s prior orders,

familiarity with which is presumed. Briefly summarized, Lawson is Spirit’s former chief executive officer. He retired on July 31, 2016. His Retirement Agreement contained non-compete obligations for two years, until July 31, 2018. In early 2017, non-party investment firms Elliott Associates, L.P. and Elliott International, L.P. (collectively, “Elliott”) hired him to provide consulting services in connection with a proxy contest Elliott launched to replace five Arconic board members. When Spirit learned about this, Spirit notified Lawson that his involvement with Arconic constituted a breach of his non-compete, and Spirit stopped paying Lawson and demanded that he repay what the company had already paid him under the Retirement Agreement. Lawson disputes that he breached the non-compete. The disputed issues in this case largely involve interpreting and applying the non-compete provision in Lawson’s Retirement Agreement. That provision prohibited Lawson from being involved with “any business that is competitive with the Business or any portion thereof.” Lawson v. Spirit AeroSystems, Inc., No. 18-1100-EFM, 2018 WL 3973150, at *2 (D. Kan. Aug. 20, 2018). The Retirement Agreement defined the term “Business” as follows:

We [Spirit] are engaged in the manufacture, fabrication, maintenance, repair, overhaul, and modification of aerostructures and aircraft components, and market and sell our products and services to customers throughout the world ( . . . the “Business”). Id. Lawson’s theory of the case focuses on his allegations that Spirit is a tier-one manufacturer of aerostructures and aircraft components (i.e., it builds and sells large structures and components like fuselage, propulsion, and wing systems) whereas Arconic is a tier-three or tier-four manufacturer of lightweight engineered metal components (e.g., small fasteners, connectors, bolts, engine components, fan blades, etc.) that end up in airplanes because they are used by tier-one suppliers like Spirit. Id. at *7-*9. Lawson therefore contends that Spirit and Arconic are not in the same “Business” because they do not provide, market, or sell the same “specific products and services.” Id. Furthermore, Lawson contends that Spirit and Arconic do not regard each other as competitors in their SEC filings or otherwise. Id. Spirit does not seem to dispute its market positioning vis-à-vis Arconic—namely, that Spirit is primarily a tier-one supplier whereas Arconic makes and sells smaller aerostructures and aircraft components. In fact, Arconic is one of Spirit’s suppliers. Spirit instead relies on the business overlap between Spirit and Arconic in light of the non-compete language prohibiting Lawson from being involved with “any business that is competitive with the Business or any portion thereof” (emphasis added) and defining “Business” to include “manufacture, fabrication, repair, overhaul, and modification of aerostructures and aircraft components.” See generally, e.g., Lawson v. Spirit AeroSystems, Inc., No. 18-1100-EFM-ADM, 2020 WL 2101251, at *1 (D. Kan. Apr. 30, 2020) (discussing Spirit’s motion to compel Arconic to provide discovery on business overlap); Lawson v. Spirit AeroSystems, Inc., No. 18-1100-EFM-ADM, 2020 WL 243598, at *1 (D. Kan. Jan. 16, 2020) (same). Spirit contends that both it and Arconic manufactured, fabricated,

maintained, repaired, overhauled, modified, marketed and/or sold the same or similar aerostructures and aircraft components; marketed similar relevant machining capabilities; competed for employees; committed capital and other resources for research and development; maintained relationships with, submitted proposals or bids to, and contracted with the same or similar customers; and pursued strategic initiatives to try to expand their respective market shares. (ECF 281-1, at 6-8.)1 Spirit also contends that Arconic sought to expand its aerospace business via its relationship with Spirit by extracting more of the aerostructure and aircraft components business for itself as a supplier to Spirit (i.e., attempting to move up the value chain). (Id.) Lawson filed this lawsuit seeking to recover what he believes Spirit owes him. Elliott’s

role in the current lawsuit is in some respects germane to the current motion, and it is more thoroughly explained in one of the court’s prior orders. See generally Lawson v. Spirit AeroSystems, Inc., 410 F. Supp. 3d 1195, 1201-02 (D. Kan. 2019). Briefly summarized, Lawson and Elliott entered into two agreements on January 31, 2017. The first was a Consulting Agreement for Lawson to provide Elliott with consulting services in connection with the Arconic proxy contest. By the time Elliott and Lawson entered into the Consulting Agreement, Spirit had already notified them that Spirit believed Lawson’s consulting arrangement with Elliott would

1 Where the court cites ECF documents throughout this order to include page numbers, the cited page numbers are the ones assigned by the CM/ECF system that appear at the top of the page. violate Lawson’s non-compete. So Lawson and Elliott also entered into an Indemnification Agreement by which Elliott agreed to indemnify Lawson if Spirit failed to pay him under his Retirement Agreement, in which case Elliott would become subrogated to the extent of those payments to Lawson’s rights of recovery from Spirit. Id. at 1201-03. Elliott paid Lawson tens of millions of dollars pursuant to the Consulting and Indemnification Agreements (ECF 152-1, at 16-

18), and retained Lawson’s litigation counsel at Elliott’s expense, Lawson, 410 F. Supp. 3d at 1203. Elliott is now funding this lawsuit to recover the amounts Spirit allegedly owes Lawson pursuant to his Retirement Agreement. A. The Parties’ Initial Discussions Regarding Spirit’s ESI By the time this case was reassigned to the undersigned on March 26, 2019, Lawson had already filed a motion to compel Spirit to produce ESI directed to the issue of whether Spirit and Arconic are in the same “Business.” (ECF 57, at 23-24.) Lawson’s motion to compel was based on its Requests for Production (“RFPs”) seeking various documents related to Spirit’s relationship with Arconic and the overlap between their businesses. (Id. at 12; ECF 58-2, at 9-14 (RFPs 19,

25-30, 34-38, 40).) The parties had not been able to agree on ESI custodians or search terms and had difficulty meeting and conferring productively. So Lawson moved to compel Spirit to produce documents according to Lawson’s list of search terms and custodians. (ECF 57, at 24-30.) Spirit responded, arguing Lawson’s ESI demands were “nothing short of a fishing expedition,” disproportionate to the needs of the case, and “abusive,” and that Lawson was “using discovery for the sake of creating obvious burden.” (ECF 72, at 2-3.) Spirit explained that Lawson had demanded that Spirit search 69 custodians’ ESI plus each custodian’s assistant’s ESI. (ECF 72-9, at 1; ECF 136-2, at 1.) Lawson had also demanded that Spirit run these searches using about 90 search terms. Many of these terms contained one or more “OR” connectors, and therefore the effective number of search terms far exceeded 100. (See ECF 136-2, at 2-4.) None of the search terms were tailored to specific custodians. Many of the search terms like “Manufactur! OR Sell OR Sale OR Sold” and “Catalog OR Catalogue” and “Aerostructure OR component” (id.) were not tailored to the issues in the case. Another broad search term was “Elliot OR Elliott OR Eliott OR Eliot OR singer.” (Id. at 2.) Spirit explained to Lawson that this search term was too broad

because Spirit has more than 60 current or former employees with “Elliott” or “Elliot” in their names. (ECF 136-3, at 2.) Other search terms were generic terms in the aviation industry such as “paint,” “fuselage,” “spoiler,” “pylon,” “bulkhead,” and “fabricat!” without corresponding limiting terms. (Id. at 2-3; see also ECF 136-2, at 2-3.) In February of 2019, Spirit had identified four individuals (out of the dozens of custodians Lawson had proposed) that Spirit believed would be most likely to have relevant and responsive information, and Spirit ran searches on their ESI using Lawson’s proposed search terms. (ECF 72-13 ¶¶ 10-11, at 2-3.) These searches returned more than 320,000 documents, of which Spirit reviewed approximately 400 and determined that 85% were irrelevant. (ECF 136-3, at 3.) Spirit

viewed Lawson’s proposed search terms as ineffective, and therefore told Lawson that Spirit would craft its own search terms. Spirit also suggested limiting the ESI searches to the ten custodians it believed were most likely to have relevant information. (Id. at 3-4, 6.) B. Hearing on Lawson’s Motion to Compel

On April 23, 2019, the court held a hearing on Lawson’s motion to compel. (ECF 182-2, at 160-265 (hearing transcript).) At that time, Spirit explained why the sampling exercises revealed that traditional ESI techniques involving custodians and search terms had proven unworkable: “we keep viewing the issue of the business of Spirit and search terms and custodians as being very challenging because everyone is a custodian of documents related to the business, and every document is related to the business.” (Id. at 171-72.) So it was “a challenge . . . to come up with search terms on the issue of ‘business’ that will actually be workable.” (Id. at 173.) Spirit had therefore conducted custodian interviews and produced documents such as fabrication, R&D, collaboration, and capital expenditure slide decks and reports to try to avoid “searching through everyone’s email for things like ‘fuselage’ or ‘wing kit,’ et cetera.” (Id. at 179-80, 221-25.)

But Lawson professed to have knowledge about how he wanted the searches run by virtue of his role as Spirit’s former CEO. His counsel argued as follows: There are a couple of things that are a little different, here. One is that my client is the former CEO of Spirit Aerosystems, which means he does have some knowledge, but incomplete knowledge of where the documents are buried, if you will, and which people are important to talk to. . . . . . . . And our document requests and our initial search terms and custodian lists and meet and confer points were based, in part, on our conversations with Mr. Lawson.

(Id. at 168.) The court consulted with the parties about a proposed plan for tailoring ESI custodians and search terms. (Id. at 209-60.) Beginning first with the issue of custodians, the court rejected Lawson’s request for 69 custodians and encouraged Lawson to prioritize his list of custodians because at some point the court would start shifting costs. (Id. at 217, 225-26.) In consultation with the parties, the court developed the following ESI protocol:  Lawson would first identify up to seven categories for which he was seeking ESI;  For each category, Spirit would list the top three custodians most likely to have relevant ESI, from the most likely to the least likely, along with a brief explanation as to why Spirit believed the custodian would have relevant information;  Lawson would then serve a list of five custodians with proposed search terms for each, and a second set of five custodians and search terms a week later; and  Spirit would search those custodians’ ESI using Lawson’s search terms, conduct a sampling to determine responsiveness rates, and suggest modified search terms if the sampling revealed an unreasonably large number of non-responsive or irrelevant results. See Lawson v. Spirit AeroSystems, Inc., No. 18-1100-EFM-ADM, 2019 WL 1877159, at *2-*3 (D. Kan. Apr. 26, 2019). The court directed the parties to work together on search terms to try to achieve an 85% responsiveness rate. (ECF 182-2, at 242-43, 249-51.) The court explained that “we’re not going to run these really broad search terms that end up in ridiculous numbers of unresponsive documents. You might lose some responsive documents somewhere along the way, but there’s got to be a tradeoff, at some point. . . . 85 percent to me seems like a pretty fair cutoff.” (Id.) Lawson said he was “fine with that.” (Id.) The parties initially proceeded according to this protocol. Lawson identified seven categories for which he sought ESI. On May 20, Spirit identified the top three custodians for each

category. They included the following eleven custodians, about half of which overlapped across multiple categories: Kevin Matthies, Katie Wesbrooks, Alan Young, Zach Zimmerman, Wendy Crossman, Eric Hein, Ron Rabe, John Pilla, Bob Skinner, Susan Shook, and Reid Jackson. (ECF 136-4, at 1-3.) On May 23, Spirit served first amended initial disclosures that disclosed all of these custodians as potential witnesses. (ECF 147-2, at 7, 10-16.) On May 23 and 30, Lawson selected his first and second sets of five custodians. (ECF 136-5, at 1; ECF 136-6, at 1.) Unbeknownst to the court at that time, only three of the ten were custodians Spirit had identified as most likely to have relevant ESI—Matthies, Crossman, and Rabe. The remaining seven ESI custodians were not on Spirit’s list—Duane Hawkins, Michelle

Lohmeier, Jim Cocca, Krisstie Kondrotis, Thomas Gentile, Vic McMullen, and Bill Brown. However, Spirit’s suggested ESI custodian list in February 2019 had suggested Hawkins and Gentile as proposed ESI custodians. Also unbeknownst to the court at that time, Lawson provided Spirit with 803 search terms (counting terms with “OR” as multiples) and asked Spirit to run those search terms on all ten custodians’ ESI. (ECF 136-5, at 4-5; ECF 136-6, at 3-4.) Spirit harvested the ten custodian files. They consisted of 1.8 million documents, or 1.2 million after deduplication. (ECF 135 ¶ 7, at 2; ECF 136-7.) Spirit ran the search terms. They returned 304,272 documents, or 468,595 documents including families, for a total of

approximately 200 GB of data. (ECF 135 ¶ 8, at 2; ECF 136-7.) Spirit reviewed a 384-document sample and determined that only 7.8% were responsive. (ECF 135 ¶ 9, at 2; ECF 136-7.) Of those, many were technically responsive but were actually irrelevant to the claims and defenses in this lawsuit. (ECF 152-1, at 27 n.1.) Spirit provided Lawson with hit reports for the first five custodians. (ECF 136-7; ECF 152-1, at 29-42.) Spirit also proposed revised search terms with corresponding hit reports for those custodians. (ECF 152-1, at 43-56.) C. The Modified ESI Protocol

On June 6, 2019, the court convened a telephone conference to discuss various discovery issues, one of which was Spirit’s concerns about the lack of efficiencies in the ESI process. (ECF 182-2, at 291-423 (hearing transcript).) At that time, Spirit told the court that Lawson had selected only three custodians from the list of custodians Spirit identified as most likely to have ESI responsive to the categories Lawson had identified; that the number of Lawson’s search terms had increased to 803; and that all of this had resulted in only a 7.8% responsiveness rate. (Id. at 348- 51.) Turning first to Lawson’s selected custodians, the court remarked that Lawson’s decision to pick seven custodians that were not on Spirit’s list would be at his own peril. (Id. at 359.) Turning next to search terms, the court limited Lawson to 25 search terms and instructed him to tailor them according to custodian rather than running the same search terms across all custodians. (Id. at 359- 67; see also ECF 88, at 1.) The court again told the parties to work together to try to achieve an estimated responsive hit rate of at least 85%. (ECF 88, at 1.) On June 28, Lawson sent Spirit revised proposed search terms. (See ECF 136-8, at 1.) Many of Lawson’s revised terms were again common aviation-related terms, such as “nacelle” or “nut,” as well as verbs commonly used in many industries such as “fasten*” and “procure*.” (See

id. at 4.) Spirit conducted new searches of the ten custodians’ ESI using Lawson’s revised terms, and this returned approximately 322,000 documents. (ECF 135 ¶ 12, at 3.) A sampling exercise revealed that the responsive rates for each custodian ranged from 0.5% to 13.5%, with an average across all custodians of 5.1%. (ECF 136-9; ECF 135 ¶ 13, at 3.) Spirit again characterized many of the responsive documents identified in the sampling exercise as “technically responsive” but “largely irrelevant to this dispute.” (ECF 136 ¶ 13, at 3.) By August 9, the parties met and conferred about search terms again, and Spirit stated that it believed continuing to discuss individual search terms and custodians would not be productive. (See ECF 147-2, at 133.) On August 19, Spirit produced responsive documents from its July

sampling exercise, totaling only 173 documents. (ECF 136-10, at 2; ECF 136 ¶ 13, at 3.) On September 9, Spirit produced 77 non-responsive documents in an effort to assist Lawson in determining why his search terms were resulting in such few responsive documents and next-to- no relevant documents. (Id.) Spirit further advised Lawson that it believed reviewing the remaining approximately 322,000 documents “is not proportional to the needs of this case and will likely result in a small number of relevant documents.” (ECF 136-10, at 2.) Spirit predicted that, “[b]ased on the most recent sampling exercise, it is likely that only 5% of these documents are responsive to outstanding discovery requests, and that these technically responsive documents are largely irrelevant to the dispute.” (Id.) D. Technology Assisted Review Around that time, the parties abandoned efforts to refine search terms to meet the 85% responsiveness-rate goal, and they began discussing the option of conducting a TAR of the 322,000-document set identified in July 2019. (ECF 147-2, at 136.) Spirit’s ESI vendor Legility offers a TAR tool called “Predict.” On September 12, Spirit provided Lawson with information

on this tool. (Id. ¶ 15, at 3.) After an initial set of documents is coded for responsiveness, the Predict tool uses continuous active learning to code additional documents. Predict ranks coded documents from the most likely responsive to the least, and then humans review the top-ranked documents. When Predict determines the pool of responsive documents is depleted such that the effort of continued review is disproportionately outweighed by the possibility of additional gain, review ceases. Legility then conducts a statistical validation of the TAR’s results. (Id. at 155-56.) On September 17, the court convened a telephone conference to discuss a number of discovery issues, including the parties’ progress on ESI and the impact of the status of discovery on the case schedule. (ECF 182-2, at 456-507 (hearing transcript).) Lawson explained that the

parties had discussed the TAR process, and that Lawson wanted to proceed in that fashion. (Id. at 463-67.) Spirit explained that it had been proceeding with document discovery on two different paths: (1) the ESI protocol and the process Lawson had discussed, and (2) separately, the “old- fashioned way” of targeted productions via custodian interviews and collections. (Id. at 468-73.) According to Spirit, the second method had proven to be more efficient and effective. Using that method, Spirit had already produced about 39,000 pages of documents primarily on the issue of the “Business,” and Spirit wanted to continue to proceed down that path. (Id.) Meanwhile, the ESI process was costly and yielded exceptionally low responsiveness rates. Spirit explained that the issue of business overlap between Spirit and Arconic is a “unique area of discovery that’s so incredibly broad that we just aren’t seeing that the electronic discovery processes have been working,” and therefore Spirit was “not so sure we are in absolute agreement . . . that the Predict plan is the way to go and we should be doing that.” (Id. at 471-73.) By that time, Spirit had already spent hundreds of thousands of dollars on document collection, processing, and hosting, as well as the sampling exercises, and the parties had not even

achieved a 15% responsiveness rate. (Id. at 473-75.) Given this, the court raised the possibility of adjusting the case schedule in order to allow the parties to proceed with TAR, with Lawson bearing the TAR costs. (Id. at 476-85.) The parties did not agree as to the allocation of costs at that time, but they agreed to move forward with the TAR process subject to Spirit filing a motion to shift those costs to Lawson. (Id.) On September 19, Spirit reached out to Lawson in a last attempt to try to avoid another lengthy and expensive ESI review via the TAR that was not likely to yield many responsive documents. (See ECF 136-10.) Spirit reiterated that the sampling exercise it conducted in July suggested that only 5% of the 322,000 documents would be responsive to Lawson’s discovery and

those “technically responsive documents are largely irrelevant to the dispute.” (Id. at 2.) Spirit once again proposed that, in lieu of TAR, it continue to engage in its own efforts to identify custodians who likely had information responsive to discovery requests, reviewing that information, and producing it rather than incurring the TAR costs. (Id.) Through this process, Spirit had already produced approximately 4,700 documents, totaling approximately 40,000 pages. (ECF 134, at 7; ECF 147, at 7.) Spirit explained that the TAR could cost “$250,000-$400,000 in eDiscovery and document review costs, and $40,000-$60,000 in outside counsel time, as well as additional costs not yet identified.” (ECF 136-10, at 2.) The letter concluded that, if Spirit did not hear from Lawson by 5:00 p.m. on September 20, Spirit would proceed with the TAR. (Id.) Lawson responded with a letter on September 20 that in no way suggested that Lawson did not want to proceed with the TAR. (ECF No. 136-11, at 1-2.) To the contrary, Lawson wanted to meet and confer to work out the TAR protocol. And Lawson reiterated that “we believe the TAR Protocol is an effective and efficient means to review the documents from the custodians we have selected, and the Court indicated that it expected Spirit to proceed with the TAR Protocol in order

to move quickly to the substantial completion of document discovery in this case.” (ECF 136-11, at 1.) Lawson stated that he expected Spirit to produce documents located through the TAR on a rolling basis to be completed by November 1. (Id.) On September 26, the parties met and conferred regarding the TAR protocol, including first-level review by contract attorneys and a second-level quality-control review by Spirit’s counsel’s law firm. (ECF 136-14, at 1.) 2 E. Spirit’s ESI Discovery Expenses and the Instant Motion Spirit incurred approximately $108,000 in ESI vendor costs and approximately $31,000 in attorneys’ fees conducting the three sampling exercises in February, May, and July of 2019. (ECF 136 ¶ 14, at 3.) This does not include the time Spirit spent conferring with counsel, reviewing

correspondence relating to the issue of custodians and search terms, speaking with custodians, harvesting data, or coordinating with Legility. (Id.) When Spirit filed the instant motion, Spirit estimated that its expenses for the TAR process would total about $325,000 to $500,000, broken down as follows:  $150,000 to $300,000 for costs incurred by Legility’s document review team;  $25,000 to $30,000 in data promotion costs to migrate the ESI corpus into Legility’s review platform;

2 Lawson disputes that he directed Spirit to proceed with the TAR. (ECF 147, at 9.) But the only reasonable inference to be drawn from Lawson’s September 20 response letter and the parties’ subsequent meet-and-confer on September 26 to discuss TAR logistics is that Lawson wanted to proceed with the TAR and that time was of the essence.  $6,000 per month through the date of trial3 in hosting costs; and  $40,000 to $60,000 in attorneys’ fees to provide a quality control review of Legility’s document review, coordinate with Legility and its review team, and review the documents prior to production. (ECF 134, at 9.) By late October 2019, Spirit estimated that only 3.9% of the documents subject to the TAR would be responsive. (ECF 152-2 ¶ 7, at 2.) At a discovery conference on November 8, Spirit reported that it estimated ending the TAR upon achieving a 65% recall rate4 and, in view of that, substantially completing document production by December 6. (ECF 169, at 2.) At a discovery conference on January 10, 2020, Spirit reported that it had reached a 68.5% recall rate, but that Lawson did not believe that was sufficient. (ECF 227-1, at 35-125 (hearing transcript).) So Spirit agreed to keep reviewing to an 80% recall rate to accommodate the then-upcoming deposition schedule, but with the understanding that continued review would be subject to this motion to shift costs. (Id. at 60-65.) Spirit eventually completed production of the TAR documents in mid-January after reaching an 85% recall rate. See Lawson v. Spirit AeroSystems, Inc., No. 18-1100-EFM-ADM, 2020 WL 1813395, at *1 (D. Kan. Apr. 9, 2020). Only 3.3% of the documents in the TAR set of

322,000 documents were responsive. Id. at *8. Of those responsive documents, Spirit produced (or withheld as privileged) 23,951 documents, consisting of 170,083 pages. Id. Of the documents Spirit produced, only 9,128 were responsive and the rest were non-responsive, irrelevant family members. Id.

3 At the time, the court would have projected a trial about 18 months later, so data hosting costs would have been about $108,000 ($6,000/month x 18 months). Since then, the schedule has been extended further, so the total data hosting costs may be higher. 4 “Recall rate” refers to the percentage of responsive documents within the TAR dataset that have been correctly identified. Lawson filed a motion to compel Spirit to produce the remaining TAR documents beyond the 85% recall rate—i.e., the “residual TAR documents.” See generally id. at *1. The court denied this motion because Lawson refused to bear Spirit’s costs to review and produce the residual TAR documents, no authority supported what Lawson was effectively seeking (a 100% recall rate), and further review was not proportional to the needs of the case. Id. at *9. By late January 2020, Spirit

estimated its TAR expenses to be approximately $400,000 in vendor costs and $200,000 in law firm fees. Id. at *5. F. Spirit’s Motion to Shift Costs Spirit now asks the court to shift all costs and attorneys’ fees associated with the TAR to Lawson under Rule 26(c). Spirit argues that it spent months collecting, processing, hosting, and searching millions of documents from custodians selected by Lawson and using search terms selected by Lawson; that this process cost hundreds of thousands of dollars and yielded only a small percentage of responsive documents “and far fewer relevant documents.” (ECF 134, at 3 (emphasis in original).) Meanwhile, Spirit’s separate path of conducting custodian interviews and

gathering targeted files resulted in far more significant and fruitful productions on the issue of business overlap between Spirit and Arconic. Spirit therefore urges the court to shift the TAR expenses to Lawson in order to enforce proportionality standards. Lawson opposes the motion primarily on the grounds that cost-shifting is only available for ESI that is not reasonably accessible, which is not the case here. (ECF 147, at 9-11.) Lawson also contends that Spirit “refused to cooperate to reduce its own burden” by failing to help craft search terms to reach the court’s 85% responsiveness target. (Id. at 5-6, 7-8.) And Lawson argues that, to the extent the court considers cost-shifting factors, they weigh in favor of denying Spirit’s motion. (Id. at 11-14.) II. APPLICABLE LEGAL STANDARD Lawson’s argument that the court may only shift costs for ESI that is not reasonably accessible misapprehends the applicable legal standards. In support of this argument, Lawson correctly points out that it is ordinarily presumed that the responding party bears the expense of complying with discovery requests. Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 358

(1978). But Lawson erroneously relies on Rule 26(b)(2)(B), Zubulake v. UBS Warburg LLC, 217 F.R.D. 309 (S.D.N.Y. 2003), and Semsroth v. City of Wichita, 239 F.R.D. 630 (D. Kan. 2006). Spirit’s motion does not rely on Rule 26(b)(2)(B), which is the rule that applies to non-reasonably accessible discovery. Rather, Spirit’s motion invokes Rule 26(c). (ECF 133, at 2 (“Pursuant to FED. R. CIV. P. 26(c) . . . .”).) That rule is not limited to non-reasonably accessible discovery, and it was amended in 2015 to make clear that the court may allocate discovery expenses for good cause in order to protect a party from undue burden or expense. The court begins with a brief history of the development of the law on this issue in order to put Lawson’s argument in context. A. Even Before the 2015 Amendments to Rule 26(c), the Court’s Authority to Shift Costs Was Not Limited to Non-Reasonably Accessible Discovery Courts have long recognized that cost-shifting may be appropriate where ESI is not reasonably accessible (e.g., data stored on backup tapes) because of the burden and expense involved in restoring or reconstructing the data into a usable format. This is reflected in the seminal case of Zubulake, 217 F.R.D. at 318-20 (setting forth factors to consider when determining whether to shift non-reasonably accessible ESI costs). In 2006, Rule 26(b) was amended to reflect the

principles articulated in Zubulake and subsequent cases on shifting costs for non-reasonably accessible ESI. See U.S. ex rel. Carter v. Bridgepoint Educ., Inc., 305 F.R.D. 225, 239-40 (S.D. Cal. 2015) (recognizing that Rule 26(b)(2)(B) “embrace[s] the logic in Zubulake”); see also Maria Perez Crist, Preserving the Duty to Preserve: The Increasing Vulnerability of Electronic Information, 58 S.C. L. REV. 7, 15 (2006) (discussing the 2006 amendments); FED. R. CIV. P. 26(b)(2) advisory committee notes to the 2006 amendment (whether to require production of non- reasonably accessible ESI turns on whether the burdens and costs are justified by the circumstances of the case and listing considerations similar to the Zubulake factors). That rule now provides that “[a] party need not provide discovery of electronically stored information from sources that the

party identifies as not reasonably accessible because of undue burden or costs.” FED. R. CIV. P. 26(b)(2)(B). But, the court may still order discovery of non-reasonably accessible ESI “if the requesting party shows good cause, considering the limitations of Rule 26(b)(2)(C)”; the court may specify conditions for the discovery, including requiring the requesting party to bear the costs. Id.; see also FED. R. CIV. P. 26(b)(2) advisory committee notes to the 2006 amendment (“[C]onditions may also include payment by the requesting party of part or all of the reasonable costs of obtaining information from sources that are not reasonably accessible.”). Although the court in Zubulake stated that “[a] court should consider cost-shifting only when electronic data is relatively inaccessible,” 217 F.R.D. at 324 (emphasis in original), that

approach is no longer accepted. Even around the time Zubulake was decided, other courts focused on Rule 26(b) proportionality factors to determine which party should bear the costs of discovery without regard to whether ESI was reasonably accessible or not. See, e.g., Thompson v. U.S. Dep’t of Hous. & Urban Dev., 219 F.R.D. 93, 97-99 (D. Md. 2003) (noting the court can shift burdensome or expensive ESI costs, in whole or in part, under Rules 26(b) and 26(c)). Zubulake author Judge Scheindlin herself later recognized that cost-shifting may be appropriate even for accessible ESI “so long as the factors set forth in Rule 26(b)[] evidence cost-shifting’s suitability.” Carter, 305 F.R.D. at 240 (citing SHIRA A. SCHEINDLIN & DANIEL J. CAPA, ELECTRONIC DISCOVERY AND DIGITAL EVIDENCE 314 (2009)). Indeed, when another judge in this district considered the issue of cost shifting in Semsroth in 2006, the court found that it was “questionable” whether the ESI at issue was “not reasonably accessible because of undue burden or cost,” but nevertheless proceeded to analyze whether cost-shifting was appropriate based on the Zubulake and Rule 26(b) proportionality factors. See 239 F.R.D. at 638. Since Zubulake and the 2006 amendments to Rule 26(b), the practice of storing data in

“inaccessible” formats (e.g., backup tapes) has declined. See The Sedona Principles, Third Edition: Best Practices, Recommendations & Principles for Addressing Electronic Document Production, 19 SEDONA CONF. J. 1, 41-42 (2018) (noting “the quickly diminishing role of backup tapes”). Corresponding with this decline, a “preference for cost-shifting has been extended beyond merely inaccessible ESI.” Carter, 305 F.R.D. at 240. Considering the proportionality factors in Rule 26(b), “the cost of even accessible ESI’s production may be shifted to a party that has not shown its peculiar relevance to the claims and defenses at hand.” Id.; see also F.D.I.C. v. Brudnicki, 291 F.R.D. 669, 675-77 (N.D. Fla. 2013) (requiring the parties to share the costs of producing ESI that the court noted were “largely marginal documents”).

B. Rule 26(c)(1) Was Amended in 2015 to Expressly Authorize Allocating Discovery Expenses, Even for Non-Reasonably Accessible ESI Against this backdrop, Rule 26(c)(1) was amended in 2015 “to include an express recognition of protective orders that allocate expenses for disclosure or discovery” in order to “forestall the temptation that some parties may feel to contest” a court’s authority to issue such orders. FED. R. CIV. P. 26(c)(1) advisory committee’s note to the 2015 amendment. That rule now expressly authorizes a court to issue an order for good cause to protect a party from undue burden and expense including specifying the terms of discovery such as “the allocation of expenses for the disclosure or discovery.” FED. R. CIV. P. 26(c)(1)(B). Thus, the rule confirms the court’s authority to allocate expenses, even where ESI is reasonably accessible. See Oxbow Carbon & Minerals LLC v. Union Pac. R.R. Co., 322 F.R.D. 1, 11 (D.D.C. 2017) (noting that “courts have, over the years, looked beyond accessibility to determine whether to shift discovery costs” and that Rule 26 was amended in 2015 to reflect this). Although a court has this authority, the amendment “does not imply that cost-shifting should become a common practice. Courts and parties should continue to assume that a responding party ordinarily bears the costs of responding.” FED. R. CIV.

P. 26 advisory committee’s note to the 2015 amendment. Lawson’s argument that cost shifting is appropriate only for non-reasonably ESI is also contradicted by the parties’ own agreed ESI protocol. The parties agreed that they would generally limit ESI discovery to reasonably accessible ESI. (ECF 40 ¶¶ I(D), II(A)(1), at 2.) They also agreed that identifying and producing ESI would be “subject . . . to the development of reasonable and appropriate cost allocation agreements.” (Id. ¶ VII, at 8.) Therefore, the parties’ own agreement about the possibility of allocating costs was not limited to non-reasonably accessible ESI. The parties’ agreed ESI protocol also stated that any such cost-allocation agreements would be “tailored to give the parties incentives to use cost-effective means of obtaining information and

disincentives to use cost-effective means of obtaining information and disincentives to engage in wasteful and costly discovery activity.” (Id.) This appears to be designed to further similar goals as Rule 26(c)(1)(B). In sum, Lawson’s argument that the court may shift costs only for non-reasonably accessible discovery is contrary to both the Federal Rules and the parties’ own agreed ESI protocol. The court applies Rule 26(c)(1)(2) to determine whether to allocate the TAR expenses to Lawson. C. Legal Standard for Allocating Expenses Under Rule 26(c)(1)(2) Under Rule 26(c), Spirit has the burden to demonstrate good cause. See Brave Law Firm, LLC v. Truck Accident Lawyers Grp., Inc., No. 17-1156-EFM-JPO, 2019 WL 3740594, at *2 (D. Kan. Aug. 8, 2019). To establish good cause, the moving party must make “a particular and specific demonstration of fact, as distinguished from stereotyped and conclusory statements.” Gulf Oil Co. v. Bernard, 452 U.S. 89, 102 n.16 (1981). The court has broad discretion “to decide when a protective order is appropriate and what degree of protection is required.” Seattle Times Co. v. Rhinehart, 467 U.S. 20, 26 (1984). This is because the court “is in the best position to weigh fairly the competing needs and interests of the parties affected by discovery. The unique character of

the discovery process requires that the trial court have substantial latitude to fashion protective orders.” Id. Courts evaluate the Rule 26(b)(1) proportionality factors to determine whether discovery imposes undue burden or expense such that allocating expenses under Rule 26(c)(1)(B) is warranted. See Oxbow Carbon, 322 F.R.D. at 11 (noting “the Rule 26(b) proportionality factors . . . are essentially identical to the factors courts have considered in determining whether to shift discovery costs under Rule 26(c)”); McClurg v. Mallinckrodt, Inc., No. 4:12-CV-00361-AGF, 2016 WL 7178745, at *3 (E.D. Mo. Dec. 9, 2016) (evaluating whether to allocate expenses under Rule 26(c)(1)(B) by considering the proportionality factors); see also Carter, 305 F.R.D. at 240

(S.D. Cal. 2015) (discussing that the proportionality factors may evidence the suitability of shifting “the cost of even accessible ESI’s production”); Brudnicki, 291 F.R.D. at 676 (recognizing courts may order cost-shifting to enforce proportionality limits). Thus, expenses may be allocated where “discovery presents an ‘undue burden or expense’ relative to the prospective benefit of the discovery.” United States ex rel. Bibby v. Wells Fargo Bank, N.A., No. 1:06-CV-0547-AT, 2016 WL 7365195, at *1 (N.D. Ga. May 26, 2016) (citing FED. R. CIV. P. 26(b)(1) and (c)(1)). III. GOOD CAUSE WARRANTS ALLOCATING THE TAR EXPENSES TO LAWSON TO PROTECT SPIRIT FROM UNDUE BURDEN AND EXPENSE “Parties may obtain discovery regarding any nonprivileged matter that is relevant to any party’s claim or defense and proportional to the needs of the case . . . .” FED. R. CIV. P. 26(b)(1). The court already determined that the discovery Lawson initially sought through the ESI process regarding whether Spirit and Arconic are in the same “Business” is relevant. The court will therefore analyze the proportionality factors to determine whether to allocate the TAR expenses to Lawson. Those factors include “the importance of the issues at stake in the action, the amount in controversy, the parties’ relative access to relevant information, the parties’ resources, the

importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit.” Id. A. Importance of the Issues at Stake in the Action In a case seeking damages, the first proportionality factor looks at whether the issues at stake implicate broader “public policy spheres, such as employment practices, free speech, and other matters [that] may have importance far beyond the monetary amount involved,” or if the claims seek to “vindicate vitally important personal or public values.” FED. R. CIV. P. 26(b)(1) advisory committee notes to the 2015 amendments; see Nyberg v. Zurich Am. Ins. Co., No. 15- 1359-EFM, 2016 WL 11671468, at *3 (D. Kan. June 21, 2016) (analyzing these considerations). The parties undoubtedly believe the issues at stake in this action are important and, as Lawson

emphasizes, they involve tens of millions of dollars. But the breach-of-contract claim here does not implicate any broader societal impact. This is a case between private parties seeking money damages. The claims relate to the terms of Lawson’s employment, but the claim itself does not involve matters of public policy such as prohibited forms of discrimination. This factor weighs against the need for far-reaching discovery measures. Compare Cratty v. City of Wyandotte, 296 F. Supp. 3d 854, 860 (E.D. Mich. 2017) (finding this factor weighed in favor of discovery where constitutional rights were at stake), with Lakeview Pharmacy of Racine, Inc. v. Catamaran Corp., No. 3:15-290, 2019 WL 587296, at *3 (M.D. Pa. Feb. 13, 2019) (finding this factor weighed against discovery where the case involved claims for breach of contract and breach of the implied duty of good faith and fair dealing between private parties); Nyberg, 2016 WL 11671468, at *3 (same, in case that involved payment of dismemberment benefits); see also, e.g., L. Zingerman, DDS, PC v. Nissan N. Am., Inc., No. 14 C 7835, 2016 WL 4206062, at *4 (N.D. Ill. Aug. 10, 2016) (finding this factor weighed against in camera review where the case involved a vehicle’s features relating to entertainment and social media, not safety or functionality).

B. The Amount in Controversy In evaluating the next proportionality factor, courts compare the cost of the discovery at issue to the amount in controversy. Oxbow Carbon, 322 F.R.D. at 7. If the cost of the discovery is close to the amount in controversy, a court is more likely to find that a protective order is warranted. See, e.g., Alley v. MTD Prod., Inc., No. 3:17-CV-3, 2018 WL 4689112, at *3-*4 (W.D. Pa. Sept. 28, 2018) (finding this factor weighed in favor of a protective order where the cost of responding to the discovery was “roughly equivalent” to the amount in controversy). According to Lawson, the amount in controversy is $39 million to $53 million. Spirit does not dispute this. The TAR expenses reportedly total approximately $600,000. Lawson contends

these projected costs are not unreasonable given the amount in controversy. In support, Lawson cites In re Broiler Chicken Antitrust Litigation, No. 16 C 8637, 2018 WL 3586183, at *8-*9 (N.D. Ill. July 26, 2018) (finding no undue burden where discovery would cost $1.2-1.7 million in antitrust case involving multi-billion dollar industry); U.S. ex rel. Guardiola v. Renown Health, No. 3:12-CV-00295-LRH, 2015 WL 5056726, at *10 (D. Nev. Aug. 25, 2015) (same, where discovery would cost $136,000 and plaintiff was claiming multi-million dollars in damages); and Xpedior Creditor Trust v. Credit Suisse First Boston (USA), Inc., 309 F. Supp. 2d 459, 466-67 (S.D.N.Y. 2003) (rejecting cost-shifting where discovery would cost $400,000 and potential recovery was between $68.7 million and $7 billion). The court agrees that the $600,000 in TAR expenses are not necessarily unreasonable considering the fact that the amount in controversy far exceeds those TAR expenses. However, the TAR expenses are not Spirit’s only document production expenses in the case. Indeed, they are not even Spirit’s only document production expenses on the issue of business overlap between Spirit and Arconic. To the contrary, Spirit has already borne significant discovery expenses on

that subject matter via the resources it devoted to (1) the ESI process that led up to the TAR review, (2) its separate path of producing documents on this subject matter by the “old-fashioned method” of targeted productions via custodian interviews and collections, and (3) the discovery it has pursued from Arconic. Unlike the cases Lawson cites, this is not a case in which Spirit refused to provide discovery on the subject matter at issue. See, e.g., In re Broiler Chicken Antitrust Litig., 2018 WL 3586183, at *1-*2, *4, *7-*8 (refusing to allow a defendant to simply reproduce custodial searches from a DOJ investigation rather than running new search terms covering a broader time period because the documents from the DOJ investigation were not calibrated to the issues in the antitrust price-fixing conspiracy at issue); Guardiola, 2015 WL 5056726, at *10

(granting motion to compel gap-period emails that were “highly relevant and contain information not likely available in other discoverable documents”). The fact that a plaintiff seeks millions in relief does not give him or her license to conduct fishing expeditions that run up the cost of discovery. See Murphy v. Deloitte & Touche Grp. Ins. Plan, 619 F.3d 1151, 1163 (10th Cir. 2010) (“Rule 26(b) has never been a license to engage in an unwieldy, burdensome, and speculative fishing expedition.”). To the contrary, “monetary stakes are only one factor” that must be balanced against the other proportionality factors. FED. R. CIV. P. 26(b)(1) advisory committee notes to the 2015 amendment. C. The Parties’ Relative Access to Relevant Information The court turns next to the parties’ relative access to relevant information. “In considering this factor, courts look for ‘information asymmetry’—a circumstance in which one party has very little discoverable information while the other party has vast amounts of discoverable information.” Oxbow Carbon, 322 F.R.D. at 8 (quoting FED. R. CIV. P. 26(b)(1) advisory

committee’s note to the 2015 amendment). “[T]he burden of responding to discovery lies heavier on the party who has more information, and properly so.” FED. R. CIV. P. 26(b)(1) advisory committee’s note to the 2015 amendment. At first blush, it might seem that Spirit has superior access to relevant ESI. But this is true only to a certain extent. Sprint has superior access to ESI regarding its “Business.” Indeed, as Spirit pointed out over a year ago, every document generated by a Spirit employee relates somehow to Spirit’s “Business.” (ECF 182-2, at 171-72.) But the relevant issue for discovery purposes is not solely Spirit’s “Business.” Instead, it is the competitive overlap between Spirit and Arconic. To that end, Spirit has superior access to documents regarding that subject matter only to

the extent that Spirit knows the areas in which it believes there is competitive overlap with Arconic. And Spirit already produced documents and information that bear on that issue outside of the ESI/TAR process. Spirit’s answer and interrogatory responses provided Spirit’s contentions regarding the alleged business overlap between Spirit and Arconic. (See ECF 29 ¶ 8, at 2-3; ECF 147-2, at 141-42, 150-53.) As early as the April 2019 hearing on Lawson’s motion to compel, Spirit had already conducted custodian interviews and made targeted productions of documents such as fabrication, R&D, collaboration and capital expenditure slide decks and reports. (ECF 182-2, at 180, 221-25.) At the time Spirit filed the instant motion, it had already made 16 productions amounting to approximately 4,700 documents totaling 40,000 pages located through “targeted searches and collection of ESI from custodians that Spirt [sic] determined—through custodian interviews—are most likely to have information relevant to this dispute.” (ECF 134, at 2, 7.) This included documents bearing on the issue of business overlap between Spirit and Arconic. Ud.) On September 12, 2019, before the TAR process began, Spirit served interrogatory responses regarding the area of business overlap that referenced a detailed, three-page list that identified documents Spirit produced to support its contentions. (ECF 147-2, at 139-53.) Spirit’s document production was so fulsome by that time that Spirit was able to list the bates numbers of selected documents that it claimed support both it and Arconic’s business for each allegedly overlapping aerostructure or aircraft component. (/d. at 150-53.) As but one example, following is the excerpt for floor beams: Spirit & Arconic Selected Selected Documents Re: Spirit's Business Aerostructures/Aircraft| Documents Re: a Spirit000020608 — (Picture of Triple Bay Floor Panel - closeup) Spirit000020609 — (Picture of Triple Bay Floor Panel) Spirit000020610 — (Picture of Floor Grid Panel) Spirit000020621 — Picture of Floor Grid Spinit000024664 — Monolithic Floors (Spirit AeroSystems Fabrication Overview) loor beams Spint000021322 Spirit000024667 — Monolithic Floors (Spirit AeroSystems Fabrication Overview) Spirit000024668 — Beams (Spirit AeroSystems Fabrication Overview) Spirit000025049 — Joined Body Section (737) (Boeing Amdt No. 30) Spirit000025050 — Joined Body Section (737) (Boeing Amdt No. 30) Spirit0000251 15 — Aft/Fwd Lower Lobe (747) (Boeing Amdt No. 30) Spirit000025151 — Lower Lobe (767) (Boeing Amdt No. 30) pirit000025162 ~ Lower Lobe (777) (Boeing Amdt No. 30 The flip side of Spirit’s contention is ascertaining the nature of Arconic’s business. That information would largely come from third-party discovery from Arconic and, to that extent, there is no “information asymmetry” because that discovery is equally available to both Lawson and Spirit. Yet it appears that Spirit is the only party that vigorously pursued discovery from Arconic. The docket sheet reflects that Lawson served Arconic with a document subpoena (ECF 108, 117)

and a deposition subpoena he later withdrew (ECF 191-92, 305).5 Spirit also served Arconic with multiple document and deposition subpoenas, and Spirit pursued that discovery. These included two apex-level depositions of Arconic fact witnesses Arthur Collins and Rodney Heiple. (ECF 225 & 235 (deposition notices) and 303-2 & 303-3 (deposition transcripts).) In addition, Spirit moved to compel Arconic to produce documents (twice) and a Rule 30(b)(6) witness. See, e.g.,

Lawson, 2020 WL 243598, at *1 (granting motion to compel Arconic to produce documents); Lawson, 2020 WL 2101251, at *1 (same, and Rule 30(b)(6) designee). This is therefore not a case in which Spirit has superior access to ESI that it is withholding regarding the issue of its “Business” overlap with Arconic. Spirit has every incentive to produce discovery evidencing its competitive overlap with Arconic because such documents would support Spirit’s theory of the case. And that is exactly what Spirit did by producing discovery on this subject matter separate and apart from the ESI/TAR process. In contrast, Lawson must prove a negative—namely, the lack of any such competitive overlap. See Lawson, 2018 WL 3973150, at *6 (holding Lawson’s compliance with his non-compete obligations was a condition precedent to

Spirit’s payment obligations under the Retirement Agreement). Yet Lawson has not shown any way in which perpetuating review through the TAR process was important to resolving the issues of overlap in specific products and services between Spirit and Arconic. The court is therefore not persuaded that Spirit has superior access to documents regarding the competitive overlap between Spirit and Arconic above and beyond what Spirit already

5 Lawson’s withdrawal of his deposition notice to Arconic in the midst of Spirit twice moving to compel discovery from Arconic is curious, particularly considering Arconic’s intransigence in providing that discovery. According to Spirit’s prior filings in the case, during the relevant time period, Elliott was Arconic’s largest shareholder, owning 8-11% of Arconic’s stock valued between $600 million and $1.46 billion. (ECF 355, at 5 n.7.) In May of 2017, Elliott and Arconic entered into a settlement agreement to end the proxy contest, pursuant to which Elliott placed multiple board members on Arconic’s board. (Id.) produced. Spirit has repeatedly lamented the uselessness of the ESI/TAR batch of documents compared to the targeted collections it produced via custodian interviews. (See ECF 136-3, at 2- 3; ECF 182-2, at 171-73, 468-73.) And Lawson has not provided the court with any information from which the court could find otherwise. This factor is therefore neutral. D. The Parties’ Resources

Lawson contends the parties’ resources weigh against shifting costs because “Spirit is a Fortune 500 company with $6.8 billion in revenue” and he is merely an individual. (ECF 147, at 13.) But “consideration of the parties’ resources does not foreclose discovery requests addressed to an impecunious party, nor justify unlimited discovery requests addressed to a wealthy party.” FED. R. CIV. P. 26(b)(1) advisory committee’s note to the 2015 amendment; see also Nyberg, 2016 WL 11671468, at *4 (noting that “consideration of the parties’ resources ‘does not . . . justify unlimited discovery requests addressed to a wealthy party’”). Courts must apply the proportionality factors “in an even-handed manner that will prevent use of discovery to wage a war of attrition or as a device to coerce a party, whether financially weak or affluent.” FED. R.

CIV. P. 26(b)(1) advisory committee’s note to the 2015 amendment. Spirit may be a Fortune 500 company with billions in revenue, but it also faces real business pressures. This year alone, Spirit has endured massive layoffs as a result of the Boeing 737 MAX grounding and the impact of the COVID-19 pandemic. See, e.g., Michael Stavola, Spirit AeroSystems in Wichita to Lay Off 1,450 Employees, WICHITA EAGLE, May 1, 2020 (discussing layoffs due to “the grounding of the 737 MAX and . . . the COVID-19 pandemic”)6; Michael Sainato, ‘The Only Ones Not Paying for Boeing’s Mistakes is Boeing’: Laid-Off Supply Workers

6 Available at https://www.kansas.com/news/business/aviation/article242446291.html. Voice Their Anger, GUARDIAN, Feb. 1, 2020 (discussing Spirit’s layoffs of 2,800 Wichita employees due to Boeing halting production of the 737 MAX).7 (See also ECF 221, at 1 (granting in part Spirit’s oral motion to amend the scheduling order “to allow Spirit executives . . . to focus on imminent issues facing the company as a result of Boeings’ directive to Spirit to stop all 737 MAX deliveries”).) The court will not require Spirit to shoulder needless litigation expenses

simply because it is a big company. Lawson’s argument that he should not bear any of the TAR expenses simply because he is an individual is unpersuasive. Lawson and Elliott entered into a multi-million-dollar agreement pursuant to which Elliott is funding this lawsuit with the goal of helping Lawson recover the millions Spirit allegedly owes him under the Retirement Agreement, a significant portion of which Elliott already paid to Lawson. As explained above, Elliott and Lawson entered into a Consulting Agreement and an Indemnification Agreement on January 31, 2017. See Lawson, 410 F. Supp. 3d at 1201-02 (providing more details regarding this arrangement). By that time, Lawson and Elliott already knew Spirit would contend this arrangement violated Lawson’s non-compete. See id. Yet

they moved forward and entered into their business arrangement anyway, knowing it would trigger this dispute. In the following months, Elliott paid Lawson $5,303,412.01 pursuant to the Consulting Agreement relating to the Arconic proxy contest and $26,373,893.80 pursuant to the Indemnification Agreement. (ECF 152-1, at 17-18.) Lawson is therefore amply funded. He has already recovered over $26 million from Elliott that he claims Spirit owes him, plus Elliott paid him over $5 million extra via the Consulting Agreement. And Elliott, which is now funding this

7 Available at https://www.theguardian.com/business/2020/feb/01/boeing-workers-spirit- layoffs-future-unsure. lawsuit for Lawson, was willing to roll the dice knowing its multi-million-dollar arrangement with Lawson would fuel this dispute with Spirit. Ultimately, this factor is neutral. It does not weigh in favor of or against shifting expenses. Both parties have sufficient resources to bear their fair share of litigation expenses. E. The Importance of the Discovery in Resolving the Issues

In analyzing the importance of the discovery in resolving the issues in the case, the court looks to whether the discovery seeks information on issues “at the very heart of [the] litigation.” Oxbow Carbon, 322 F.R.D. at 8 (alteration in original); see also Lakeview Pharmacy, 2019 WL 587296, at *4 (stating that, in weighing this factor, the court “looks to whether the discovery request goes to a central issue in the case”). “A party claiming that a request is important to resolve the issues should be able to explain the ways in which the underlying information bears on the issues as that party understands them.” FED. R. CIV. P. 26(b)(1) advisory committee notes to the 2015 amendment. Where a party can establish only marginal relevance, courts are less likely to determine that the discovery sought is proportional. See, e.g., In re Bard IVC Filters Prod. Liab.

Litig., 317 F.R.D. 562, 566 (D. Ariz. 2016) (concluding defendants were not required to search ESI that “appear[ed] marginal” to resolving the issues in the case). 1. Documents Produced Through the ESI/TAR Process Appear to Be of Marginal Benefit Given the Other Information Spirit Produced Regarding the Alleged Areas of Competitive Overlap Throughout this litigation, Spirit disclosed specific areas of alleged competitive overlap with Arconic. This began as early as Spirit’s answer, which identified the alleged areas of competitive overlap as manufacturing aerospace parts and components and assembling aerospace structures (listing specific parts, components, and structures); machine fabrication and chemical processing for the aerospace business; sourcing, purchasing, and utilizing equipment in connection with the aerospace business (listing equipment and brands); maintaining the same or substantially similar certifications in the aerospace business (providing examples); and having overlapping customers or potential customers in the aerospace business such as Boeing, Airbus, and Rolls- Royce. (ECF 29 ¶ 8, at 2-3.) By the hearing on April 23, 2019, Spirit had already conducted custodian interviews and produced documents separate and apart from the ESI process including fabrication, R&D, collaboration, and capital expenditure slide decks and reports. While the parties

moved forward with the ESI process, Spirit continued to pursue its separate path of producing targeted collections via custodian interviews and collections. By mid-September 2019, Spirit had produced about 39,000 pages of documents primarily on the issue of Spirit’s “Business” and served interrogatory responses that described the overlapping products and components that both Spirit and Arconic manufactured, produced, or sold in an attached exhibit listing forty-nine overlapping aerostructures and aircraft components, along with citations to selected supporting documents regarding both Spirit’s and Arconic’s business. (ECF 147-2, at 139-53.) The list included bates- numbers for documents that Spirit had already produced relating to each product or component. (Id. at 150-53.) For most of the products or components, the list cited to multiple documents,

including citations to specific pages of Spirit’s contracts with Boeing and amendments thereto, fabrication overviews, pictures, and Spirit capabilities documents. (Id.) By the time Spirit filed the instant motion, it had collected, reviewed, and produced 40,000 pages (approximately 4,700 documents, in 16 separate productions), including documents bearing on the issue of the “Business” overlap between Spirit and Arconic. Spirit had been undertaking those efforts since December 2018 and they were ongoing at that time. And, as mentioned previously, Spirit has also worked diligently to shore up a robust discovery record from Arconic to prove the areas of overlap. Meanwhile, the ESI/TAR process consistently yielded exceptionally low responsiveness rates. Spirit’s sampling exercises in February, May, and July of 2019 yielded responsiveness rates of 15%, 7.8%, and 5.1%. At the time Spirit filed this motion, it predicted that the TAR review would “likely yield only 5% technically responsive documents (and far fewer relevant documents).” (ECF 134, at 3 (emphasis in original).) By the time Spirit filed its reply brief, it estimated that only 3.9% of the TAR corpus would be responsive. (ECF 152, at 3.) In the end, the actual responsiveness rate was only approximately 3.3%. See Lawson, 2020 WL 1813395, at

*8. Spirit ultimately produced 23,951 documents via the TAR process—9,128 were responsive documents and the rest were non-responsive, irrelevant family members. Id. at *8. But even more to the point than the exceptionally low responsiveness rate, Lawson does not articulate any way in which the documents produced as a result of the ESI/TAR process added anything of meaningful value to the documents and information Spirit produced separate from the ESI/TAR process on the issue of “Business” overlap. As noted previously, Spirit has repeatedly lamented the uselessness of this batch of documents compared to the targeted collections it produced via custodian interviews. (See ECF 136-3, at 2-3; ECF 182-2, at 171-73, 468-73.) And Lawson has not provided the court with any information from which the court could find otherwise.

The only concrete information in the record that speaks to the marginal benefit of the TAR documents is briefing on Lawson’s January 2020 motion to compel Spirit to produce the residual TAR documents. (ECF 226-29, 240, 247.) By the time Lawson filed this motion, Spirit had completed the TAR production after reaching an 85% recall rate. Lawson sought the residual TAR documents but refused to pay the estimated costs to produce them, approximately $40,000. Lawson, 2020 WL 1813395, at *5. It speaks volumes that Lawson did not consider the residual TAR documents to be important enough to cover a relatively small amount in costs. Id. at *9. On that record, Lawson did not even argue that the residual TAR documents were important to resolving the issues in the case. To the contrary, his briefing cited the importance of the issues at stake, the amount in controversy, the parties’ relative access to information, and the parties’ resources, but notably not the importance of the discovery in resolving the issues. (ECF 227, at 4.) He argued in conclusory fashion that three documents previously produced through the TAR and attached as exhibits (ECF 228, at 3-24) were “relevant” (ECF 227, at 7), but he did not articulate any way in which they added anything of meaningful value above and beyond other

documents Spirit had already produced and identified in discovery. Thus, his argument was more to the effect that Spirit must produce every relevant document. But the fact that Spirit may have more documents regarding the nature of its “Business” is unsurprising. Spirit is not obligated to turn over every relevant document it possesses on this subject matter. The court is therefore unable to find that the ESI/TAR documents were important to resolving the issues in the case. 2. Lawson Did Not Tailor the ESI/TAR Process to Target Relevant Discovery About Spirit and Arconic’s Competitive Overlap Lawson contends that the TAR process was tailored to discover relevant information about the “Business” overlap between Spirit and Arconic. (ECF 147, at 12.) In support of this argument, Lawson argues the court already granted his motion to compel on the RFPs at issue, finding he tailored his initial RFPs during the meet and confer process. (Id.) This argument mischaracterizes the record. The court granted Lawson’s motion to compel subject to the court’s April 2019 ESI protocol that gave Lawson a blueprint to select custodians that Spirit had identified as likely to have responsive information and required Lawson to narrow his search terms to achieve an 85% responsiveness rate. Lawson did neither.

The court’s April 2019 ESI protocol allowed Lawson to dictate the discovery topics, custodians, and search terms. Generally, “the party who will be responding to discovery requests is entitled to select the custodians it deems most likely to possess responsive information and to search the files of those individuals.” In re EpiPen (Epinephrine Injection, USP) Mktg., Sales Practices & Antitrust Litig., No. 17-MD-2785-DDC-TJJ, 2018 WL 1440923, at *2 (D. Kan. Mar. 15, 2018) (quotation omitted). But, during the hearing on Lawson’s motion to compel, Lawson claimed he had knowledge of where relevant ESI would be located because he was Spirit’s former CEO. (ECF 147-2, at 183 (arguing Lawson “is the former CEO of Spirit Aerosystems, which means he does have some knowledge . . . of where the documents are buried”).) The court

therefore permitted Lawson to choose the ESI custodians and search terms, but the court placed limits on those choices (ten custodians) and search terms (achieve an 85% response rate) in order to incentivize Lawson to focus on meaningful ESI rather than going on a fishing expedition. Turning first to Lawson’s selected ESI custodians, Spirit followed the court’s April 2019 ESI protocol by devoting the resources to identify the custodians it believed were most likely to have relevant documents that would be responsive to the categories of information Lawson said he was seeking. These custodians included (1) Kevin Matthies, (2) Katie Wesbrooks, (3) Alan Young, (4) Zach Zimmerman, (5) Wendy Crossman, (6) Eric Hein, (7) Ron Rabe, (8) John Pilla, (9) Bob Skinner, (10) Susan Shook, and (11) Reid Jackson. (ECF 136-4, at 1-3.) When Lawson

selected his custodians, only three of the ten were on Spirit’s list—Matthies, Crossman, and Rabe. (See ECF 136-5, at 1; ECF 136-6, at 1.) Lawson’s other seven were not on Spirit’s list—Hawkins, Lohmeier, Cocca, Kondrotis, Gentile, McMullen, and Brown—although Spirit had previously suggested Hawkins and Gentile as ESI custodians in February of 2019 and Spirit ultimately disclosed them as potential witnesses. So at least half of Lawson’s ESI custodians—Lohmeier, Cocca, Kondrotis, McMullen, and Brown—were not tailored to discover relevant information. This was confirmed when Lawson later filed a motion to exceed the ten-deposition limit in which Lawson stated that he intended to depose Matthies, Crossman, Rabe, Gentile, and Hawkins (the ESI custodians Spirit had identified in some way) but not Lohmeier, Cocca, Kondrotis, McMullen, or Brown (the ESI custodians Spirit never identified). See generally Lawson v. Spirit AeroSystems, Inc., No. 18-1100, 2020 WL 1285359, at *1 (D. Kan. Mar. 18, 2020). Lawson’s search terms also were not tailored to discover relevant information. Lawson’s list of search terms in May of 2019 included such broad terms as “Lawson” and “Arconic” without any limiting modifiers, as well as terms apparently designed to target nearly all of Spirit’s business

such as “Boeing AND (buy OR sell OR sale OR sold OR design* OR manuf* OR fab* OR suppl*)” and similar searches for Airbus, GE, Rolls Royce, and Bombardier. (ECF 136-5, at 4-5; ECF 136-6, at 3-4.) These terms made no attempt to target the specific products and services that Spirit had identified as areas of competitive overlap with Arconic. Lawson’s search terms also included numerous combinations of common aerostructures and aircraft components like “Fuselage” and “Spoiler” and “Bathroom” and “Flap” and “Aileron” paired with general business terms like “Boeing,” and “Fab*” and “Design*” and “Component.” (Id.)8 Lawson did not tailor the search terms to individual custodians. Spirit ran the search terms on the first five custodians and provided Lawson with a

custodian-level hit report and proposed revised search terms, including a custodian-level hit report with the revised search terms. (ECF 147-2, at 47-77.) Spirit’s revised search terms were more targeted to the issues in the case. For example, rather than simply searching the word “Fuselage” paired with a general aerospace business term like “Boeing,” Spirit suggested targeting specific portions of the fuselage like “fuselage frame” and “fuselage kit” and “fuselage panel” and “fuselage skins” and “fuselage stringers” and “fuselage-to-wing connection.” (Id. at 64-65.) Spirit

8 The second page of Lawson’s proposed search terms listed 39 different aerostructures and/or aircraft components (and variations thereof) in a column called “Term 1” and a second column titled “Term 2” that contained 20 modifiers (some of which contained multiple variations like “(Sell OR Sale OR Sold)”). The instructions required Spirit to pair each of the search terms in the first column with each of the first terms in the second column. also proposed replacing certain “AND” connectors with “w/i 25” connectors. For example, Lawson proposed the search “(‘Tier’ 1 OR First-Tier OR ‘Tier one’ OR ‘First Tier’ OR ‘1st Tier’) AND (suppl* OR Compet* OR acquir* OR Consold* OR merg* OR aerostructure*).” Spirit proposed replacing the “AND” connector with a “w/i 25” connector and dropping “acquir*” OR. “Consod*” OR “merg*”—terms that are not relevant to the areas of competitive overlap. By making simple changes like these, Spirit’s suggestions substantially dropped the hit count as follows: Po Lawson’s Search Terms Spirit’s Search Terms Total Hit Family Hit Total Hit Family Hit Count Count Count Count 29,738 48,543 7,250 16,426 104,872 147,817 21,593 47,128 27,087 40.018 5,548 10,148 Michelle Lohmeier 36,432 63,051 9320 20,875

SLT EWAN Woy) CAKE During the discovery conference on June 6, 2019, after Spirit told the court that Lawson had served 803 search terms, the court limited him to 25 terms per custodian and ordered him to tailor them to each custodian. (See ECF 88, at 1-2.) But Lawson’s next round of search terms were even broader than the previous set. He proposed terms like “compet*” and “nut” and “snub*” and “Procur*” and “fuselage” and “bulkhead” and “Fasten*” and “flap” and “pylon” and “nacelle” and “wing” and “Boeing” with no limiting modifiers. (See ECF 136-8, at 4.) In view of the above, the court disagrees that Lawson’s custodians and search terms were targeted to obtain relevant discovery about Spirit and Arconic’s competitive overlap. Spirit’s answer alleged that its “core products” include fuselages, nacelles, and wing assemblies; that it also sells aerospace structures like pylons and bulkheads: and that Boeing is one of its customers.

(ECF 29 ¶ 8, at 2.) Spirit presumably knows its own business and produced documents in discovery outside of the ESI/TAR process proving these allegations. (ECF 147-2, at 150-53 (citing selected bates numbered documents that Spirit produced in discovery to support these contentions).) It is therefore unclear why Lawson would also require all documents from ESI custodians containing the word “fuselage” or “nacelle” or “wing” or “pylon” or “bulkhead” or

“Boeing.” Lawson is entitled to take reasonable discovery related to certain aspects of Spirit’s “Business,” but he does not need every single document relating to Spirit’s business, even from selected custodians. 3. Spirit is Not Responsible for the Low Responsiveness Rate Lawson attempts to blame Spirit for the failure of the ESI/TAR process. Lawson argues that Spirit refused to participate meaningfully in the iterative search term protocol ordered by the court; that this led to the parties’ inability to craft search terms to reach the court’s 85% target; and that Spirit did not provide Lawson with the names of alleged overlapping products until serving an amended interrogatory answer in September 2019. (See ECF 147, at 5-8.) This argument is unpersuasive for several reasons.

First and foremost, the district judge already determined that Lawson bears the burden of proving compliance with his non-compete as a condition precedent to recovering the amounts he claims Spirit owes him under the Retirement Agreement. See Lawson, 2018 WL 3973150, at *6. And Lawson instigated the ESI/TAR process by filing a motion to compel in the spring of 2019 seeking a court order that Spirit produce ESI from 69 custodians that Lawson selected using search terms that Lawson selected. At the hearing on that motion, Lawson professed to have superior knowledge about how he wanted the searches run by virtue of his role as Spirit’s former CEO. (ECF 182-2, at 168 (“[M]y client is the former CEO of Spirit Aerosystems, which means he does have some knowledge . . . of where the documents are buried, if you will, and which people are important to talk to . . . .”).) The court granted the motion as it related to the overlap in Spirit and Arconic’s business, but subject to the court-ordered ESI protocol that allowed Lawson to select custodians and search terms to achieve an 85% responsiveness rate. The problem is not that Lawson never achieved an 85% responsiveness rate. The problem is that he never got anywhere close to that. Instead, he unilaterally selected five custodians Spirit had never identified as key

ESI custodians and who apparently did not have enough relevant documents to even warrant deposing them. And he continued to pursue unnecessarily broad search terms despite the court’s admonition at the April 2019 hearing that “we’re not going to run these really broad search terms that end up in ridiculous numbers of unresponsive documents. You might lose some responsive documents somewhere along the way, but there’s got to be a tradeoff, at some point. . . . 85 percent to me seems like a pretty fair cutoff,” to which Lawson responded that he was “fine with that.” (ECF 182-2, at 250-51.) Thus, the onus was on Lawson to craft effective ESI searches that would be important to proving his theory of the case—i.e., the lack of competitive overlap between Spirit and Arconic. But he has not explained any way in which the TAR was important to resolving that

issue. See FED. R. CIV. P. 26(b)(1) advisory committee notes to the 2015 amendment (“A party claiming that a request is important to resolve the issues should be able to explain the ways in which the underlying information bears on the issues as that party understands them.”). Second, Spirit had every incentive to produce documents as economically and efficiently as possible because of the default rule that the producing party ordinary bears the costs of production. But, during the April 2019 hearing, Spirit explained that traditional ESI techniques involving custodians and search terms were unworkable: “we keep viewing the issue of the business of Spirit and search terms and custodians as being very challenging because everyone is a custodian of documents related to the business, and every document is related to the business.” (ECF 182-2, at 171-73.) So it was “a challenge . . . to come up with search terms on the issue of ‘business’ that will actually be workable.” (Id.) Yet Spirit participated in the court-ordered ESI protocol. Spirit devoted the resources to identify relevant ESI custodians, which Lawson ignored. Spirit harvested the data from Lawson’s requested ESI custodians and ran the search terms Lawson requested. And, following the May 2019 sampling exercise, Spirit suggested revised search terms

that would have substantially reduced the hit count. But Lawson came back with revised search terms that were so broad that, as the sampling exercises continued, the responsiveness rates dwindled. By August 2019, even Lawson abandoned efforts to refine search terms. By the discovery conference on September 17, Spirit again explained that the issue of business overlap between Spirit and Arconic is a “unique area of discovery that’s so incredibly broad that we just aren’t seeing that the electronic discovery processes have been working,” and therefore Spirit was not necessarily in agreement that the TAR process should move forward. (ECF 182-2, at 471-73.) So this is not a case in which Spirit refused to cooperate to reduce its own burden. Rather, it is a case in which Spirit’s predictions proved accurate: the issue of “Business” overlap is a discovery

area that the parties’ best efforts showed was simply not amenable to traditional ESI processes. Lastly, this brings the court to what is perhaps the most important point about the TAR process as it bears on the importance-of-the-discovery-in-resolving-the-issues proportionality factor—that is, what Lawson knew by the time Spirit embarked on the TAR process. Even if the court were to accept for the sake of argument that Spirit did not fairly participate in the ESI process up to September of 2019 (a proposition that the court does not accept), the key point in time for purposes of the present motion was when Lawson insisted on moving forward with the TAR process in and around late September of 2019. By that time, Spirit had already made substantial document productions primarily on the issue of its “Business” separate and apart from the ESI process. (ECF 182-2, at 468-73.) And, on September 12, Spirit had served detailed interrogatory responses that explained the alleged areas of competitive overlap, including the multiple-page Exhibit A that cited bates-numbered documents that allegedly supported Spirit’s contentions as to both its and Arconic’s businesses. (ECF 147-2, at 150-53.) Lawson knew that only 5% of the TAR set was likely to be responsive and, according to Spirit, even those documents would be

“technically responsive” but “largely irrelevant.” (ECF 136-10, at 2.) Yet Lawson still has not shown the court any way in which documents produced as a result of the TAR are important to resolving the issues in this case. In making this finding, the court wishes to emphasize that it is not making this determination because of the low responsiveness rates throughout the sampling process. Rather, the court makes this determination because, to this day, Lawson has not articulated how documents produced through the TAR process were not just relevant (and hence duplicative for evidentiary purposes), but uniquely relevant in such a way that they were important to resolving the issue of “Business” overlap. This factor weighs heavily in favor of allocating the TAR expenses to Lawson.

F. Whether the Burden or Expense of the Discovery Outweighs Its Likely Benefit The final factor the court must consider in the proportionality analysis is “whether the burden or expense of the proposed discovery outweighs its likely benefit.” FED. R. CIV. P. 26(b)(1). “A party claiming undue burden or expense ordinarily has far better information -- perhaps the only information -- with respect to [this] part of the determination.” FED. R. CIV. P. 26(b)(1) advisory committee notes to the 2015 amendment. Spirit estimates its TAR expenses total about $600,000, including $400,000 in fees and costs to Legility and $200,000 in fees to its law firms. Lawson, 2020 WL 1813395, at *5. These expenses were in addition to Spirit’s substantial ESI vendor costs and attorneys’ fees to conduct the three sampling exercises in February, May, and July of 2019 (ECF 136 ¶ 14, at 3; ECF 134, at 2) plus whatever expenses Spirit incurred in making targeted productions separate and apart from the ESI process. Furthermore, the issue of competitive overlap between Spirit and Arconic is only one area of discovery. Outside of the ESI/TAR process, Spirit collected, reviewed, and/or produced documents from over thirty custodians and spent hundreds of thousands of dollars producing documents. (ECF 152, at 4-5.)

Lawson argues the 3.3% responsiveness rate for the TAR process is irrelevant because the nature of the TAR means that Spirit did not have to review all of the 322,000 documents in the dataset for review. (ECF No. 147, at 5.) This argument misapprehends the major cost drivers of the TAR process. First, the volume of data subjected to the TAR process materially impacts technology costs such as data processing and hosting. Maureen O’Neill, Order Highlights Potential Costs of Predictive Coding, DISCOVERREADY (March 19, 2013).9 Thus, a more voluminous dataset drives up TAR expenses. Second, the “richness” of the dataset—i.e., the prevalence of responsive documents—can also be a key driver of TAR expenses. Id. This is because TAR is not as simple as loading the dataset and pushing a magic button to identify the

relevant and responsive documents. Rather, the parties must devote the resources (usually a combination of attorneys and contract reviewers) necessary to “educate” or “train” the predictive algorithm, typically through an ongoing process that involves multiple rounds of reviewing selected documents and providing feedback to the software. As this occurs, the software continuously analyzes the entire document collection and re-ranks the population based on relevancy. See generally BOLCH JUDICIAL INST. & DUKE LAW, TECHNOLOGY ASSISTED REVIEW

9 Available at https://discoverready.com/news-insights/insights/order-highlights-potential- costs-of-predictive-coding/?cn-reloaded=1#gref. GUIDELINES (TAR) GUIDELINES 2-5 (Jan. 2019).10 Where a very small percentage of the TAR set is relevant, that means the TAR set has extremely low “richness.” Id. at 27. When “the TAR set’s richness is extremely low, human reviewers may have a difficult time training the software on what is relevant, because examples may be scarce or difficult to come by in the TAR set.” Id. Third, another cost driver of the TAR process is the recall rate—a completeness metric that

measures the proportion of truly relevant documents that TAR has identified. Id. at 5-6. The higher the recall rate, the more resources the team must devote to train the TAR algorithm. Id. at 24-25. So, for example, if the target recall rate is set too high, “it may require unreasonable and disproportionate human review to train the computer to be able to achieve that targeted recall.” Id. In this case, all of these factors increased the burdens of the TAR. To begin, the TAR set was unnecessarily voluminous because it consisted of the bloated ESI collection that was assembled using Lawson’s selected custodians and search terms. Next, the 3.3% responsiveness rate reflected an exceptionally low level of richness that meant, in simple terms, that the TAR technology and team had to work harder to process the TAR set. And, Lawson prolonged the TAR

when Spirit wanted to cease review at a 65% recall rate but Lawson insisted that Spirit continue to an 80% recall rate. All of this drove up TAR expenses. Therefore, Lawson’s argument that the 3.3% responsiveness rate was irrelevant to the burden of the TAR process is without merit. Lawson also argues he tried to reduce the burden by conducting the TAR himself. Specifically, on September 30, 2019, Lawson’s counsel suggested taking the ESI/TAR dataset on an attorneys-eyes only (“AEO”) basis, reviewing the documents using TAR, producing the responsive documents back to Spirit, and then destroying the remainder. (See ECF 136-14, at 2.)

10 Available at https://judicialstudies.duke.edu/wp-content/uploads/2019/02/TAR-Guidelines- Final-1.pdf. Spirit had sound reasons for rejecting this proposal. As early as the April 2019 discovery hearing, the court raised the possibility of Spirit producing ESI on an AEO basis, but the parties agreed that it would be more efficient for Spirit to review ESI for privilege, confidentiality, and relevance once at the outset, as opposed to potentially having to conduct multiple reviews of the same documents. (ECF 182-2, at 243-48.) These concerns were exacerbated by the time Lawson made his

alternative AEO proposal on September 30 because 95% of the TAR corpus was expected to be non-responsive and/or not relevant, and the documents were expected to contain privileged and work-product communications, Spirit proprietary information unrelated to the case, and confidential third-party and customer information. (ECF 136-15, at 1.) Lawson is not entitled to have access to documents that are non-responsive, irrelevant, and/or privileged or work product. And the fact that Lawson filed a motion to compel re-production of documents that Spirit clawed back as privileged suggests that giving Lawson access to such documents probably would have only generated more disputes. See Lawson v. Spirit AeroSystems, Inc., No. 18-1100-EFM-ADM, 2020 WL 708021, at *1 (D. Kan. Feb. 12, 2020) (finding Spirit’s clawed-back documents were

privileged). Lawson’s alternative AEO proposal was also too late. In terms of supposed cost savings, Spirit pointed out that Lawson sent this proposal the day before Spirit’s motion for cost/fee shifting was due, eleven days after Spirit asked whether Lawson anted to proceed with TAR review, and four days after Spirit indicated the TAR process had already begun, for which Spirit had already invested significant costs. (ECF 136-15, at 1.) Indeed, a significant portion of Spirit’s estimated TAR-related costs were front-end costs relating to migrating data into the TAR platform and training document reviewers. (See ECF 135 ¶ 17, at 4.) Furthermore, the court is not persuaded based on the present record that the TAR review protocol (ECF 136-14) or Legility’s rates are unreasonable, particularly given Legility’s experience working on other matters for Spirit. (See ECF 136-15, at 2.) In the court’s experience, Spirit’s estimated expenses for the TAR are not out of line for a document production of this magnitude.11 The court therefore does not find that Lawson’s proposal to conduct the TAR himself would have practically relieved Spirit of the burdens associated with the TAR.

In sum, the substantial burden and expense of the TAR process far outweighs the likely benefits. This factor weighs heavily in favor of allocating TAR expenses to Lawson. G. The Proportionality Factors Weigh in Favor of Allocating the TAR Expenses to Lawson By mid-September 2019, Lawson’s continued pursuit of the ESI dataset via TAR was not proportional to the needs of the case. This lawsuit involves a dispute over one executive’s severance package. Although the TAR expenses are not unreasonable compared to the amount in controversy, Spirit has already borne its fair share of other discovery expenses. Both parties have adequate resources to bear their fair share of discovery expenses. Spirit produced fulsome discovery separate and apart from the ESI/TAR process regarding the subject matter on which it has superior access to discovery—namely, what it believes to be the areas of competitive overlap between Spirit and Arconic. The remaining discovery that bears on this issue would come from Arconic, and Lawson has equal access to that discovery. Lawson has not articulated any way in which documents sought through the TAR process were actually important to resolving the issue of competitive overlap (as opposed to merely being relevant and duplicative for evidentiary

11 The court acknowledges Lawson’s arguments and supporting declarations regarding the reasonableness of Spirit’s estimated $600,000 in TAR expenses, but the court is not deciding that particular issue at this time. At this procedural juncture, the court is only assessing the proportionality factor involving the burden of the TAR expenses compared to the likely benefit of the TAR process. The court will determine the reasonableness of Spirit’s TAR expenses if the parties find it necessary to brief the amount of expenses the court should award. purposes on the issue of Spirit’s “Business” generally) above and beyond the discovery Spirit produced outside of the ESI/TAR process. To the contrary, by the time the TAR process was complete, Lawson was not even willing to pay approximately $40,000 for Spirit to review and produce the residual TAR documents. Meanwhile, Spirit spent approximately $600,000 on the TAR over and above the hundreds of thousands of dollars it spent on sampling exercises and

discovery outside the ESI/TAR process. This substantial burden far outweighs any marginal benefit of the TAR process. The court long ago warned Lawson that it would allocate ESI costs if he continued to pursue needlessly overbroad discovery. (See ECF 182-2, at 217, 226 (encouraging Lawson to prioritize his list of custodians because the court would, at some point, start shifting costs).) In June 2019, after learning that Lawson had mostly ignored Spirit’s input as to the custodians most likely to have relevant information, the court again warned Lawson that his decision would be at his own peril. (Id. at 359-67.) By mid-September 2019, Lawson knew that Spirit had already produced ample discovery on this subject matter and that the TAR set would be overwhelmingly

non-responsive and irrelevant. He also knew the TAR would be expensive. On September 17, the court warned Lawson that it was inclined to allow the parties to proceed with TAR with Lawson bearing the TAR costs (id. at 476-85). See Oppenheimer Fund, 437 U.S. at 358 (recognizing the court may condition discovery on the requesting party’s payment of the costs of discovery in order to protect the producing party from undue burden or expense); see, e.g., Stryker Corp. v. Ridgeway, No. 1:13-CV-1066, 2015 WL 4425947, at *1 (W.D. Mich. July 20, 2015) (affirming magistrate judge’s decision to require defendant to pay for a search for further documents); Moody v. Aircastle Advisor, LLC, No. 3:13CV575, 2014 WL 1761736, at *1 (D. Conn. Apr. 30, 2014) (denying motion to compel further ESI unless plaintiff reimbursed defendant for the additional expenses). The court is mindful of the default rule that the producing party should ordinarily bear the costs of production, and therefore finds good cause to require both parties to bear some portion of the expenses for the overall ESI/TAR process on the issue of competitive overlap between Spirit and Arconic. Spirit has already borne approximately $150,000 through the ESI sampling exercises. Because Lawson is the party that wanted to proceed with the TAR process at a point in

time when it was disproportional to the needs of the case, the court will allocate the TAR expenses to Lawson to protect Spirit from undue burden and expense. See FED. R. CIV. P. 26(c)(1)(B). This results in the parties splitting the overall ESI/TAR expenses roughly 20%/80%. See, e.g., Wiginton v. CB Richard Ellis, Inc., 229 F.R.D. 568, 577 (N.D. Ill. 2004) (weighing the proportionality factors and shifting 75% of the discovery costs to the plaintiff where sampling showed responsiveness rates of 4.5%-6.5%). IV. CONCLUSION The court will not at this time determine a specific dollar amount to allocate to Lawson because Spirit only had projected expenses available when the parties briefed the instant motion.

Spirit should now be able to assemble its actual expenses incurred in connection with the TAR process, including vendor costs and attorneys’ fees. Because the court grants Spirit’s motion in full, it will also consider whether to award Spirit its reasonable expenses, including attorneys’ fees, incurred in filing the motion. See FED. R. CIV. P. 37(a)(5)(A) (stating that the court “must” impose fees where a motion is granted unless the movant filed the motion before conferring in good faith; the opposing party’s response was substantially justified; or other circumstances make an award of expenses unjust); see also FED. R. CIV. P. 26(c)(3) (stating that “Rule 37(a)(5) applies to the award of expenses” with respect to motions for protective orders under Rule 26(c)). The court recognizes that litigating the appropriate amount of expenses too often results in the parties spending as much time and resources as they did litigating the underlying discovery motion. For this reason, the court orders Spirit to serve a notice by June 25, 2020, informing Lawson of the dollar amount Spirit is requesting, including whatever expenses Spirit has already incurred and estimates it will incur in further briefing on this matter. Thereafter, the parties must

confer to attempt to reach an agreement regarding the issue of expenses on or before July 2. If the parties have not reached agreement by that date, Spirit may file a motion seeking expenses by July 10, with the memorandum in support limited to ten pages. Lawson’s response brief is due by July 20, and is limited to ten pages. Spirit’s reply brief is due by July 24, and is limited to five pages. IT IS THEREFORE ORDERED that Spirit AeroSystems, Inc.’s Motion to Shift Costs of Technology Assisted Review of ESI to Plaintiff (ECF 133) is granted. IT IS SO ORDERED. Dated June 18, 2020, at Topeka, Kansas.

s/ Angel D. Mitchell Angel D. Mitchell U.S. Magistrate Judge