Progressive Cas. Ins. Co. v. Delaney
Progressive Cas. Ins. Co. v. Delaney
2014 WL 3563467 (D. Nev. 2014)
July 18, 2014
Leen, Peggy A., United States Magistrate Judge
Summary
Trial court granted defendants’ motion to compel production of 565,000 “hit” documents obtained by using search terms agreed to in the parties’ joint proposed ESI protocol. The court ruled that the production would be subject to the clawback provisions of Federal Rule of Civil Procedure 26(b)(5)(B), Rule 502(d) of the Federal Rules of Evidence and the parties’ Protocol.
Plaintiff insurance company brought declaratory relief actions in several federal courts seeking a determination that its insurance policy with certain failed banks did not cover lawsuits that defendants had filed against former directors, officers and employees of the failed banks. The parties submitted and the court agreed to the joint protocol. The protocol called for plaintiff to produce responsive ESI from the realm of 1.8 million collected documents, after the application of mutually agreed upon search terms. The search terms hit on 565,000.
Plaintiff began manually reviewing the 565,000 documents for relevance and privilege, but realized that it would talk too long and cost too much. Plaintiff unilaterally decided to stop manual review and instead employed the use of predictive coding on the 565,000 documents. Plaintiff claimed that this method was more efficient and less costly than manual review.
Defendants objected to the use of predictive coding and claimed that since the search terms were negotiated and narrowed over the course of months, it would be unfair to allow the plaintiff to unilaterally modify the protocol. Further, using the predictive coding on top of the 565,000 documents would increase the chances that relevant evidence would go undiscovered.
Defendants requested that the plaintiff produce the 565,000 documents to the defendants subject to clawback provisions provided under FRCP 26(b)(5)(B) and FRE 502(d). In the alternative, defendants requested that if the court allowed plaintiff to use predictive coding, plaintiff should have to apply predictive coding to the entire collection of 1.8 documents.
The court agreed with defendants. The court stated that the protocol proposed two options for reviewing the 565,000 documents. The first option, agreed to by plaintiff, called for manual review; the second option called for full production of the 565,000 to the defendants subject to clawback. Since plaintiff unilaterally abandoned the first option, without the consent of defendants or the permission of the court, the court ruled that it was only fair to require plaintiff to produce under the second option. The court ordered plaintiff to produce the entire collection. Plaintiff was permitted to use privilege filters, so that it could “withhold more likely privileged documents identified by use of privilege filters.”
PROGRESSIVE CASUALTY INSURANCE COMPANY, Plaintiff,
v.
Jackie K. DELANEY, et al., Defendants
v.
Jackie K. DELANEY, et al., Defendants
No. 2:11–cv–00678–LRH–PAL
United States District Court, D. Nevada
Signed July 18, 2014
Counsel
Lewis K. Loss, Matthew J. Dendinger, Richard W. Boone, Jr., Loss, Judge & Ward, LLP, Washington, DC, Thomas W. Davis, II, Howard & Howard, Las Vegas, NV, for Plaintiff.Kevin R. Stolworthy, Armstrong Teasdale, Nathan Reinmiller, Sabrina G. Mansanas, Alverson, Taylor, Mortensen & Sanders, Robert R. McCoy, Joni A. Jamison, Morris Law Group, Las Vegas, NV, Andrew M. Reidy, Catherine J. Serafin, Joseph Mark Saka, Dickstein Shapiro LLP, Washington, DC, Eric D. Olson, Lee, Hong, Degerman, Kang & Waimey, Newport Beach, CA, for Defendants.
Leen, Peggy A., United States Magistrate Judge
AMENDED ORDER
*1 This Amended Order corrects the court's previous Order (Dkt.# 143) entered May 20, 2014, which erroneously attributed the publication Top Ten Best Practices of Predictive Coding to United States Magistrate Judge Andrew Peck. That publication was authored by Equivio, a company that has developed text analysis software for e-discovery.
This case is one of nine declaratory relief actions filed by Progressive in various federal district courts arising out of failed banks taken over by the FDIC as receiver. Progressive seeks a judicial declaration that its Directors & Officers Company Liability Insurance Policy for Financial Institutions issued to various failed banks and other named insureds does not provide coverage for lawsuits the FDIC–R has initiated against former directors, officers and employees of the failed banks. Progressive asserts that its insured versus insured exclusion and a “loan loss carve out” from the definition of loss in the policy precludes coverage of the FDIC–R's claims. This action involves the failure of Sun West Bank.
The parties submitted a Joint Proposed ESI Protocol (Dkt.# 65), which the court approved in an Order (Dkt.# 67) entered October 24, 2013. As a result of the court-approved ESI protocol, Progressive's ESI vendor collected approximately 1.8 million ESI documents from the following sources:
- Department Shares on Progressive's managing general agent, ABA Insurance Services, Inc. (“ABAIS”) servers: (1) PLGMARKETING; (2) PLGPRODUCT; (3) PLGTEMPLATES; (4) PLG SHARED; (5) Compliance & Licensing; (6) PLGCLAIMS; and (7) PLGUNDERWRITING;
- User Shares, User Desktops/My Documents, and/or Exchange Mailboxes on ABAIS's servers for thirty-eight custodians; and
- Lotus Notes email files at Progressive restored from backup tapes for eight custodians.
The FDIC–R filed a Motion to Compel (Dkt.# 89) December 27, 2013, seeking an order compelling Plaintiff to timely complete production of documents responsive to discovery requests served on June 14, 2013. At the time the Motion to Compel was filed, the FDIC–R had yet to receive the vast majority of Progressive's production and had not received any ESI discovery. Discovery was scheduled to close March 14, 2014, less than three months after the Motion to Compel was filed. As a result, the FDIC–R requested that the court enter an order granting its Motion to Compel Progressive to commence its production of responsive ESI immediately and complete production of documents no later than January 24, 2014. The court set the matter for hearing on January 14, 2014. See Notice of Hearing (Dkt.# 95). Progressive filed a Response (Dkt.# 100) January 13, 2014, the day before the hearing.
At the January 14, 2014, hearing, the court required counsel to have their ESI experts and client representatives in each of the related cases involved in this action to meet and have a meaningful discussion about resolving their outstanding ESI issues and the proposal laid out by the affidavit of counsel for Progressive in response to the Motion to Compel. The court gave the parties until January 28, 2014, to complete the meet and confer process and continued the matter for status conference on February 3, 2014. See Minutes of Proceedings (Dkt.# 101). The parties submitted a Joint Status Report Regarding ESI (Dkt.# 103) on January 29, 2014, indicating the parties had participated in a meeting on January 24, 2014, to discuss Progressive's review and production of ESI in connection with the pending declaratory judgment actions in the various jurisdictions. The parties stipulated to a continuance of the status conference until February 25, 2014, which the court granted.
*2 At the February 25, 2014 status conference, the court reviewed the Joint Status Report (Dkt.# 103) and heard representations from counsel concerning their ongoing ESI disputes and attempts to resolve those disputes. Counsel indicated that they had made progress and were hopeful that a resolution might be reached. The court gave the parties three weeks from the hearing to either submit a joint proposed form of order amending the ESI protocol, or to submit their respective competing proposed forms of order and continued the status hearing until March 25, 2014. See Minutes of Proceedings (Dkt.# 108).
The parties stipulated to an extension of the deadline for submitting their proposals, which the court granted in an Order (Dkt .# 110) entered March 19, 2014. The parties submitted their competing proposals (Dkt.111, 112) on March 20, 2014.
Progressive and the FDIC–R agreed to search terms to be applied to ESI documents. Applying the search terms to the 1.8 million documents collected reduced potentially responsive documents culled by the search terms to approximately 565,000 documents. Progressive's counsel began manually reviewing these 565,000 documents but quickly determined that manual review was too time intensive and expensive. As a result, Progressive began exploring alternative approaches to the review and production of ESI. It consulted with Conor Crowley, a nationally-respected authority on e-discovery, and determined that utilizing predictive coding technology to review ESI would be more effective and efficient. Progressive's selected the Equivio Relevance (“Equivio”) Program recommended by Mr. Crowley, and also used by the FDIC–R, as its predictive coding software package.
Progressive began utilizing predictive coding techniques to review ESI without the Defendants' agreement to amend the parties' stipulated ESI protocol Order (Dkt.# 67), and without seeking leave of the court to amend the ESI Order. Counsel for Progressive, Matthew Dendinger, served as a “subject matter expert” with respect to the issues presented in this matter and other pending declaratory relief actions in which Progressive and the FDIC–R are involved in related lawsuits. Mr. Dendinger reviewed documents selected by Equivio for review as the control or assessment set and reviewed an additional “training set” of documents also selected by Equivio. Mr. Dendinger reviewed 790 documents that constitute the assessment set, and 2,640 documents used in the training set. The Equivio software extrapolated Mr. Dendinger's determination regarding relevance across all of the 565,000 documents culled from the keyword search process. Based on the assessment set, Equivio estimated that 9.87% with a ± 2.09% margin of error of the 565,000 documents, or approximately 55,765 documents, are relevant.
Progressive argues that the court should modify the stipulated ESI protocol to allow for predictive coding rather than human review of the 565,000 documents culled from the keyword search of the 1.8 million ESI documents collected from the sources identified in this order. Progressive maintains that predictive coding will achieve a much more accurate measure of relevant responsive ESI documents. Predictive coding using the method described in Progressive's proposal using the Equivio program identified 90,575 potentially relevant responsive documents. Progressive proposes to manually review these 90,575 Equivio documents for privilege and to produce all non-privileged documents to the FDIC–R in discovery.
*3 Progressive applied a set of privilege filters to the 90,575 documents identified by the predictive coding process. Applying these privilege filters identified approximately 27,000 ESI documents “as more likely privileged” and approximately 63,000 ESI documents as “less likely privileged.”
Progressive acknowledges that applying predictive coding on top of search terms “is not always done.” However, Progressive argues that adopting its proposed amended ESI protocol will allow it to complete its review and production of ESI documents as quickly and efficiently as possible, and will result in a 70.80% level of recall. Progressive defines the term “recall” as a measure of how many of the relevant documents processed by Equivio will be included in the subset of documents produced. Progressive maintains that perfect recall is unachievable, and that studies have shown that human reviewers typically obtain recall rates of fifty percent or less. The Equivio software determined that 70.80% of the relevant documents included among the 565,000 documents will be produced after manual review for privilege.
Progressive proposes to perform discrepancy analysis by an attorney at Loss, Judge & Ward, other than Mr. Dendinger, who will examine a random sample of five hundred documents. As the court understands the proposal, the discrepancy attorney would examine five hundred random documents for which there is a lack of agreement on the relevancy determinations made by the Equivio software and Mr. Dendinger, i.e., documents classified as relevant by Mr. Dendinger but classified as not relevant by Equivio, or vice-versa. The discrepancy attorney's determination with respect to these five hundred randomly selected documents would be used to assess the relative quality of Equivio's analysis. A separate attorney would perform the discrepancy analysis to insure independence and validity of the determinations made by either Mr. Dendinger or Equivio. Progressive also proposes to review a statistically valid random sample of the 565,000 documents not selected by Equivio for production to verify these documents are unlikely to contain relevant documents.
Progressive proposes to produce the approximately 63,000 “less-likely-privileged” documents as “confidential” pursuant to the Confidentiality Agreement and Protective Order entered in this case (Dkt.# 63) without conducting any further manual review. Progressive also proposes that it produce these 63,000 documents without reviewing them for privilege. However, in the event Progressive produced privileged documents among the 63,000 documents, Progressive proposes application of Rule 26(b)(5)(B) of the Federal Rules of Civil Procedure and appropriate clawback under Rule 502(d) of the Federal Rules of Evidence. In its March 20, 2014, filing (Dkt.# 111), Progressive indicates that it could produce the 63,000 “less-likely-privileged” documents to the FDIC–R within two weeks of the court's adoption of its proposal.
*4 With respect to the “more-likely-privileged” documents, Progressive proposes manual review and production of all potentially privileged documents, along with a log identifying documents withheld or redacted as privileged or work product. Progressive proposes that any documents identified as privileged should be reviewed and input in an online review platform called Relativity for inclusion in the privilege log. That is, Progressive proposes that the review platform Relativity generate the privilege log rather than a manually prepared privilege log. Progressive requests this amendment to the ESI protocol because the FDIC–R has challenged the sufficiency of its privilege logs in the related action of Progressive Casualty Insurance Company v. Federal Deposit Insurance Corporation, as Receiver of Silver State Bank, 2:12–cv–00665–KJD–PAL. Progressive wants to avoid having its counsel spend significant time reviewing the more likely privileged documents and inputting information into Relativity necessary to generate a privilege log, only for the FDIC–R to challenge the sufficiency of the log. It suggests a twenty percent sample production of non-privileged documents from the 27,000 “more-likely-privileged documents” along with a log identifying any of the sample documents withheld or redacted on the basis of privilege. Progressive's proposal would require the FDIC–R to accept or challenge the sufficiency of the sample log and a resolution of any challenge by agreement or court order before a formal privilege log is prepared and served.
Finally, Progressive indicates that approximately 15,400 ESI documents culled from the 1.8 million documents were not able to be processed by the Equivio software for various reasons. Progressive proposes to manually review the “exception documents” for production of relevant, non-privileged documents, along with a privilege log identifying any discoverable “exception documents” withheld or redacted as privileged or work-product.
The FDIC–R opposes Progressive's amended proposal pointing out that the courtapproved ESI protocol, which Progressive stipulated to, yielded 565,000 “hit” documents from a data set of 1.8 million ESI documents. The court-approved ESI protocol gave Progressive the option to perform a relevance review to make certain only relevant documents were produced. Progressive used contract attorneys who reviewed approximately 125,000 of the 565,000 “hit” documents. However, instead of producing documents the contract attorneys determined were relevant, Progressive abandoned further review and unilaterally employed predictive coding to the remainder of the 565,000 “hit” documents without court approval or the FDIC–R's concurrence.
The parties have met and conferred since January 2014 in an attempt to reach an agreement on Progressive's proposal to incorporate predictive coding to the production of ESI. In the meet-and-confer process, the parties' differences were narrowed, but not resolved. The FDIC–R objects to Progressive's predictive coding protocol because it is complicated and will lead to numerous satellite disputes. Additionally, the courts which have allowed predictive coding, and Progressive's own e-discovery consultant, Conor Crowley, have emphasized the need for cooperation and transparency in adopting predictive coding processes and methods. In this case, Progressive unilaterally developed the predictive coding methodology and implemented it without input or consultation from the FDIC–R's counsel. Progressive's predictive coding methods and processes also deviate from the “Best Practices” recommended by Equivio for the Equivio system Progressive has used.
*5 The FDIC–R's preferred approach is for Progressive to produce the 565,000 culled from the 1.8 million data set using the parties' agreed-upon search terms subject to a clawback for privileged documents. This approach would avoid any cost, expense and delay of any further relevance review by Progressive's counsel, shift the expense of document review to the FDIC–R, allow the parties to proceed with discovery within a month, and eliminate mutual disputes about predictive coding protocols.
If the court is inclined to allow Progressive to incorporate predictive coding into its ESI production process, the FDIC–R proposes the court adopt its proposed amended ESI protocol which it claims is more transparent, collaborative, and evenhanded.
The FDIC–R points out that the ESI protocol entered in this case is the same template protocol the parties agreed to which has been entered by courts in five other related lawsuits involving the FDIC as receiver. The first ESI protocol was entered in the Silver State Bank case on April 18, 2013. The FDIC–R spent months narrowing search terms at Progressive's insistence and the parties eventually agreed to a set of search terms to apply to ESI. Applying the negotiated set of search terms to the data set resulted in collection of the 565,000 “hit” documents.
The ESI protocol required Progressive to review and produce ESI on a rolling basis as soon as practical. Progressive initially advised counsel for the FDIC–R that it would begin ESI production in September 2013, and complete its production by the end of October 2013. However, Progressive did not meet its own schedule, and on December 20, 2013, advised counsel for the FDIC–R that the process of reviewing search term hits was not workable, and Progressive planned to propose an alternative. Counsel for the FDIC–R asked for a written proposal so that it could be evaluated by the FDIC–R's discovery technology experts and other interested parties, but Progressive refused to do so.
At the January 14, 2014, status hearing, the court ordered the parties to have their ESI experts and representatives from each of the related cases to meet and confer to have a meaningful discussion about how to resolve the outstanding ESI issues and the proposal Mr. Dendinger presented in his affidavit just prior to the status conference. The parties met in person and telephonically January 24, 2014. At the February 25, 2014, status conference the parties reported they had had productive discussions, and the court ordered the parties to continue discussions and either submit a joint proposed form of order amending the ESI protocol or their respective competing proposals by March 18, 2014. See Order (Dkt.# 108). The parties requested and received an extension until March 20, 2014, indicating they had continued to meet to discuss the production of ESI and had had productive discussions. However, after weeks of discussions, the parties were unable to agree.
The FDIC–R requests that the court order Progressive to produce the 565,000 “hit” documents culled from application of the agreed-upon search terms subject to the existing ESI protocol. The FDIC–R argues that Progressive's predictive coding proposal is complicated and will lead to numerous satellite disputes. However, if the court is inclined to allow Progressive to incorporate predictive coding into its ESI production, the FDIC–R proposes an alternative amended ESI protocol which incorporates Equivio's recommended practices for use of the Equivio system. If the court is inclined to allow Progressive to utilize predictive coding, the FDIC–R insists that it be applied to all of Progressive's ESI, i.e ., approximately 1.8 million documents, plus any additional ESI later collected by Progressive and not only to the 565,000 documents collected by keyword searching. The FDIC–R agreed to narrow the search terms to reduce Progressive's burden. The FDIC–R argues Progressive now seeks to take unfair advantage of its agreement to compromise search terms by applying predictive coding to documents limited by the keyword searches. The FDIC–R would not have narrowed its search terms so significantly had Progressive revealed it intended to overlay predictive coding onto the narrowed set. Adopting Progressive's proposal would increase the chance that relevant documents will not be produced.
*6 The FDIC–R also opposes any cost-sharing or cost-shifting and requests that the court adopt an order that requires the producing party to bear its own costs. Because Progressive has not been transparent and has refused to divulge basic information regarding its document review process, it would be inequitable to shift the cost for production to the FDIC–R. It would also be inequitable to shift costs incurred by Progressive because its own missteps and mistakes have resulted in unnecessary cost and delay in obtaining production of ESI in this case.
Finally, the FDIC–R opposes Progressive's proposal to review a subset of the more likely privileged documents and prepare a sample privilege log for the subset of documents withheld or redacted as privileged. The FDIC–R argues Progressive should be required to provide privilege logs for any withheld documents in a timely manner. To lessen Progressive's burden, the FDICR proposes that Progressive provide privilege logs within thirty days after its documents production, or fifteen days after a document is clawed back. In the event there are more than ten thousand documents withheld from production on the basis of privilege, Progressive should be required to produce logs covering the first ten thousand documents within thirty days, the second ten thousand documents within sixty days, and the remainder within ninety days.
The parties submitted a joint proposed ESI protocol (Dkt.# 65) on October 22, 2013, which the court approved in an Order (Dkt.# 67). Pursuant to the parties' stipulated ESI protocol, the parties met and conferred and agreed to search terms to run across ESI which Progressive represented was in its possession, custody, or control. That ESI was described in Paragraph 10 of the parties' Joint Proposed ESI Protocol (Dkt.# 67).
The parties met and conferred and were eventually able to agree on a set of search terms, and on December 17, 2013, Progressive provided the FDIC–R with a “hit report” provided by Progressive's ESI vendor. See Exhibit 3 to Motion to Compel (Dkt.# 89). No ESI was produced, and Progressive did not respond to the FDIC–R's request to indicate when it would begin producing responsive documents, or when it anticipated it would complete production. See Exhibit 4 to Motion to Compel. Instead, Progressive indicated it would provide a new proposal regarding its ESI production at a meeting with counsel in January. Id. The FDIC–R filed its Motion to Compel (Dkt.# 89) December 27, 2013, following a telephone conference with counsel for Progressive in which Progressive again indicated it would make its proposal regarding the production of ESI, but would not provide information as to when its ESI production would be completed.
Pursuant to the parties' stipulated ESI protocol, Progressive was to apply the search terms to email and general documents, review the documents retrieved, and produce them to the Defendants in the format described in Exhibit A to the protocol. It gave Progressive the option to either produce all non-privileged documents captured by the agreed-upon search terms or all non-privileged documents captured by the agreed-upon search terms responsive to the Defendants' document requests, subject to proper objections. Progressive was required to advise the Defendants which option it selected. If Progressive did not produce documents retrieved by the agreed-upon search terms because it determined that documents were not relevant “or an issue arises regarding the timeliness of Progressive's production,” the parties were required to meet and confer to determine ways to resolve the issue without court intervention.” Joint Proposed ESI Protocol at ¶ 13.
*7 The ESI protocol addressed inadvertent disclosure of privileged materials and contained clawback provisions pursuant to Rule 26(b)(5)(B) of the Federal Rules of Civil Procedure and Rule 502(d) of the Federal Rules of Evidence. Id. at ¶ 5. The joint proposed protocol also specified how the FDIC–R and other Defendants would produce ESI. The stipulated ESI protocol contemplated that Progressive would produce documents on a rolling basis as soon as practical after the parties had agreed upon the search terms. Despite agreeing to the detailed protocol, Progressive has yet to produce any ESI.
Progressive's Opposition to the Motion to Compel (Dkt.# 100) indicated that approximately 565,000 documents had been retrieved by using the FDIC–R's search terms. Before learning of the total number of ESI documents retrieved in the process, Progressive had intended to host the documents in house and review them using the Summation software program. The Opposition represented that the number of documents retrieved made this “infeasible from a technical perspective.” Opposition (Dkt.# 100) at 4:20–22. As a result, Progressive and its counsel switched to an online review platform, Relativity, and employed contract attorney reviewers to manually review the ESI documents for relevance. Id. at 4:22–25. However, after a month, it became apparent that it would take approximately six to eight months “at an unacceptably high cost to Progressive” to complete the review, and counsel began exploring alternative approaches. Id. at 5:1–2. Mr. Dendinger has told the court on a number of occasions that Progressive employed eight contract attorneys to manually review these 565,000 “hit” documents for production in all of the related declaratory relief actions Progressive has filed.
Progressive retained an eminently qualified ESI discovery consultant and now proposes a predictive coding-based ESI protocol. Progressive intended to discuss its planned approach to ESI with the FDIC–R before filing its opposition to the motion to compel. However, the parties did not personally meet and confer in advance of the response date because the FDIC–R insisted on information in advance of the meeting about Progressive's new approach to ESI. Progressive determined that “short of throwing open its doors to the FDIC, both physical and electronic, and allowing the FDIC and its counsel to rummage through its files at will, there is nothing Progressive can do with respect to discovery that will not lead to still more complaints from the FIDC, still more demands for additional documents and information, and still more motion practice .” Thus, Progressive advised the court of its new approach to ESI in opposition to the Motion to Compel.
The FDIC seeks an order compelling Progressive to produce the 565,000 “hit” documents that resulted from the previously agreed-upon search terms within fourteen days, or alternatively, an order adopting the FDIC–R's ESI amended protocol attached as Exhibit A to its submission (Dkt.# 112).
*8 The FDIC–R's proposed amended ESI protocol would amend Paragraphs 10–17 and 19 of the current ESI protocol (Dkt.# 67). It would also include additional terms and definitions applicable to the amended ESI protocol. The FDIC–R's proposal would require Equivio review of the entire 1.8 million ESI documents collected for review rather than the 565,000 “hit” documents.
Equivio has published a white paper identifying ten best practices in predictive coding. See Equivio, Top Ten Best Practices of Predictive Coding (2012), available at www.equivio.com/resources.php?type=product (last visited July 15, 2014). In addition, Magistrate Judge Andrew Peck authored an ESI decision in Da Silva Moore v. Publicis Groupe, 287 F.R.D. 182 (2012), which addressed the use of predictive coding. Predictive coding, or technology assisted review, uses software that can be trained by a human being to distinguish between relevant and non-relevant documents. See Top Ten Best Practices. However, the quality of its product depends on the quality of the information used to “train” the software. Id. A human being, usually an experienced attorney involved in the litigation, referred to as the “expert” trains the software to identify relevant documents from the universe of ESI collected for review for production in discovery. Id.
Progressive's own e-discovery consultant, Conor Crowley, is a co-author of a paper on predictive coding. William P. Butterfield, Conor R. Crowley & Jeannine Kenney, Reality Bites: Why TAR's Promises Have Yet to be Fulfilled, available at http://www.umiacs.umd.edu/ oard/desi5/additional/Butterfield.pdf. Mr. Crowley is the founder of Crowley Law Office and the Chair of The Sedona Conference Working Group on Electronic Document Retention and Production, where he served as the Editor–in–Chief of the Sedona Conference Commentary on Proportionality in E–Discovery (2013). He is also the Senior Editor of the Sedona Principles (2d ed.2007), and the Sedona Conference Commentary on Legal Holds (2010). Mr. Crowley sits on the Advisory Board of Georgetown University's Law Center Advanced E–Discovery Institute and Bloomberg BNA's Digital Discovery and E–Evidence.
Mr. Crowley, his co-authors, and many others have argued persuasively that the traditional ways lawyers have culled the universe of potentially responsive documents for production—manual human review, or keyword searches—are ineffective tools to cull responsive ESI in discovery. Predictive coding has emerged as a far more accurate means of producing responsive ESI in discovery. Studies show it is far more accurate than human review or keyword searches which have their own limitations. See, e.g., Maura R. Grossman & Gordon V. Cormack, Technology–Assisted Review in E–Discovery Can be More Effective and More Efficient Than Exhaustive Manual Review, XVII Richmond J.L. & Tech. 11 (2011); see also Nicholas M. Pace & Laura Zakaras, RAND Inst. For Civil Justice, Where the Money Goes: Understanding Litigant Expenditures for Producing Electric Discovery 55–58 (2012), available at http://: www.rand.org/content /dam/rand/pubs/monographs/2012/RAND—MG1208.pdf. See, e.g., The Sedona Conference, Best Practices Commentary on the Use of Search and Information Retrieval Methods in E–Discovery, 8 Sedona Conf. J. 189, 201–02 (2007), available at http://www.thesedonaconference .org/; Maura R. Grossman & Terry Sweeney, What Lawyers Need to Know About Search Tools: The Alternatives to Keyword Searching Include Linguistic and Mathematical Models for Concept Searching, Nat. L. J. (Aug. 23, 2010).
*9 In this case, the parties negotiated an ESI protocol which was adopted by the court as an order in October of last year. Counsel for the parties advised the court that the ESI protocol in this case was the same basic template used in the related declaratory judgment actions filed in various federal district courts. Had the parties worked with their e-discovery consultants and agreed at the onset of this case to a predictive coding-based ESI protocol, the court would not hesitate to approve a transparent, mutually agreed upon ESI protocol. However, this is not what happened. Progressive agreed to search the universe of documents identified in the stipulated ESI protocol using search terms. Progressive had the option to produce all non-privileged documents retrieved without manually reviewing them to determine if they were responsive to the FDIC–R's discovery requests. Alternatively, Progressive had the option of manually reviewing the “hit” documents for production in discovery as appropriate, subject to proper objections, including privilege objections. Progressive chose the latter and began manually reviewing the ESI using eight contract attorneys.
On October 1, 2013, Mr. Dendinger communicated with counsel for the FDIC–R by email. See October 1, 2013, email, attached as Exhibit B to Memorandum (Dkt. # 112). In the email, Mr. Dendinger acknowledged that he had earlier communicated to the FDIC–R's counsel that he expected to make an initial production of ESI documents by the end of September 2013. Id. However, Mr. Dendinger indicated that given the large number of documents retrieved by the search terms, it took longer than anticipated. Id. Mr. Dendinger indicated that Progressive had a team of people dedicated to the review of the documents, and that an initial production of ESI could be expected by the end of the following week “assuming there are relevant documents in those they review between now and then.” Id. If no relevant documents were identified in the review, Mr. Dendinger indicated he would let counsel know. Id. Mr. Dendinger also affirmed that Progressive would make additional rolling productions of relevant, non-privileged documents on a regular basis. Id. This did not, however, occur.
The FDIC–R filed its Motion to Compel on December 27, 2013, when Progressive had yet to produce ESI based on the parties' stipulated ESI protocol. The Motion to Compel was also filed after counsel for the FDIC–R attempted to obtain detailed information about Progressive's alternative proposal for producing ESI. The FDIC–R asked for detailed information so that it could consult with its own ESI discovery experts and consultants to intelligently respond. Progressive's counsel did not provide the information requested in advance of an in-person meeting before the motion to compel was filed. Progressive's counsel has repeatedly told the court that Progressive believed that the FDIC–R has engaged in a pattern of making unreasonable discovery demands in each of these related declaratory relief actions. Progressive believed the FDIC–R's request for detailed information about its alternative ESI proposal was part of this pattern, and that no amount of information other than unfettered access to all of its files would satisfy the FDIC–R
*10 Progressive's proposal would relieve it of the burden of manual review of ESI according to the ESI protocol it stipulated to and allow it to adopt a predictive coding alternative to a small subset of the universe of ESI collected. Its proposal would also give its counsel exclusive responsibility for training the predictive coding software, and it fails to comply with all of the best practices recommended by the Equivio software program. Progressive proposes a “do-over” of its own invention that lacks transparency and cooperation regarding the search methodologies applied.
The article written by Progressive's e-discovery consultant and his co-authors mentioned earlier cogently and comprehensively explains why technology assisted review has not received widespread acceptance by litigants, their counsel, and the judiciary. The cases which have approved technology assisted review of ESI have required an unprecedented degree of transparency and cooperation among counsel in the review and production of ESI responsive to discovery requests. As the authors point out, typically, courts give deference to a producing party's choice of search methodology and procedures in complying with discovery requests. In the handful of cases that have approved technology assisted review of ESI, the courts have required the producing party to provide the requesting party with full disclosure about the technology used, the process, and the methodology, including the documents used to “train” the computer. See, e.g., Da Silva Moore, 287 F.R.D. at 191; In re Actos (Piglitazone—Prods.Liab.Lit.), No. 6:11–md–2299, slip op. (W.D.La. July 7, 2012). As Mr. Crowley and his co-authors also point out, litigators are loathe to reveal their methodological decisions for various reasons including assertions that: methodological decisions reveal work product; discovery about discovery exceeds the scope of Rule 26 of the Federal Rules of Procedure; revealing documents non-responsive to discovery requests exposes the producing party to unnecessary litigation risks; and the Federal Rules of Civil Procedure only require parties to conduct a reasonable search for responsive documents.
Discovery in this case has been contentious, and the court has resolved many, many discovery disputes in this case and the related Silver State Bank case. The parties spent months meeting and conferring and advised the court that they believed they would be able to resolve this dispute only to advise the court no agreement had been reached. As indicated, counsel for Progressive estimated it would begin producing responsive ESI culled from the parties' stipulated ESI protocol by mid-October 2013. Progressive spent a month, having eight contract attorneys review the “hit” documents only to abandon the process. Progressive does not claim that manual review of the documents revealed no responsive ESI. Rather, Progressive abandoned the manual review option it selected because it was taking too long and was not cost effective.
*11 Progressive is unwilling to engage in the type of cooperation and transparency that its own e-discovery consultant has so comprehensibly and persuasively explained is needed for a predictive coding protocol to be accepted by the court or opposing counsel as a reasonable method to search for and produce responsive ESI. Progressive is also unwilling to apply the predictive coding method it selected to the universe of ESI collected. The method described does not comply with all of Equivio's recommended best practices. The court agrees with the FDIC–R that approving Progressive's predictive coding proposal, or for that matter, the FDICR's competing predictive coding protocol, will only result in more disputes. It will also further delay completion of discovery in this 2011–filed case.
At the March 25, 2014, hearing, the court asked Mr. Dendinger why Progressive was opposed to applying predictive coding to the 1.8 million search documents retrieved and collected. Mr. Dendinger responded that doing so would likely vastly increase the number of more-likely privileged documents which would require manual review and preparation of privilege logs which would only result in further delay in production.
Under these circumstances, the court will require Progressive to produce the “hit” documents to the FDIC–R within fourteen days without further review. The court recognizes that requiring production of all of the “hit” documents will likely result in the production of documents not responsive to the FDIC–R's discovery requests. However, the parties' stipulated ESI protocol adopted this approach as one of two alternatives for Progressive's production. Progressive elected and then abandoned the second option—to manually review and produce responsive ESI documents. It abandoned the option it selected unilaterally, without the FDIC–R's acquiescence or the court's approval and modification of the parties' stipulated ESI protocol. Adopting the FDIC's proposal of producing the “hit” documents will shift the cost of review to the FDIC–R. The FDIC–R has committed to devote the resources required to review the documents as expeditiously as possible and estimates the process could be completed in about a month by tapping into the resources of the numerous firms involved in these related actions who also have a substantially similar ESI protocol entered in their actions. It will allow discovery, which has been stalled for many months while this dispute is pending, to move forward, and reduce future disputes about Progressive's ESI production.
Progressive may apply privilege filters to the 565,000 documents retrieved by use of the search terms and withhold more likely privileged documents identified by use of privilege filters. Progressive shall serve a privileged document log which fully complies with Rule 26(b)(5)(A) for any documents withheld or redacted as privileged or subject to protection as trial-preparation material. To reduce the burden of preparation of a privileged document log, the court will permit Progressive to serve three privileged documents logs for one-third of the documents withheld or redacted at thirty day, sixty day, and ninety day intervals.
*12 IT IS ORDERED that the FDIC–R's Motion to Compel (Dkt.# 89) is GRANTED to the extent that:
1. Progressive shall have until June 2, 2014, to produce the 565,000 “hit” documents culled from the use of the search terms to the FDIC–R subject to the clawback provisions of Federal Rule of Civil Procedure 26(b)(5)(B), Rule 502(d) of the Federal Rules of Evidence, and the court's existing ESI protocol.
2. Progressive may apply privilege filters to the 565,000 documents and withhold documents identified as more likely privileged as a result of application of privileged filters. However, Progressive shall serve the FDIC–R with a privileged document log which fully complies with Rule 26(b)(5)(A) for any documents withheld or redacted on the basis of privilege. Progressive shall produce the first privilege log for one-third of the total documents / / / withheld or redacted by June 19, 2014, the second privilege log by July 17, 2014, and the final privilege log by August 14, 2014.
NUNC PRO TUNC May 19, 2014.