In re Broiler Chicken Antitrust Litigation

• Court: District Court, N.D. Illinois
• Decided: December 22, 2017
• Docket: 1:16-cv-08637
• Status: Unknown

Opinion

UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF ILLINOIS EASTERN DIVISION

IN RE BROILER CHICKEN ANTITRUST Case No. 1:16-cv-08637 LITIGATION This Document Relates To: All Actions Magistrate Judge Jeffrey T. Gilbert

ORDER Early this year, the parties issued initial sets of Rule 34 requests for production of documents. The parties have since responded and objected to each other’s requests for production, and they met and conferred about those responses and objections. Then, in several status reports and letters, the parties identified and briefed certain disputes regarding the Rule 34 requests for production (and the responses and objections thereto). See [415 at 33-52; 429; 434; 506; 507; 508; 522; 523; 524; 526]. During a telephonic status hearing held on December 11, 2017, the parties stated that a few of those disputes remain outstanding and are now ripe for ruling. In this Order, the Court rules on several of those disputes. 1. Phone Record Redactions Request No. 2(e) of All Plaintiffs’ First Set of Requests for Production of Documents to All Defendants requests the phone records of each document custodian. [508 at 28].' The parties do not actually agree whether there is a dispute with respect to this request for production. Plaintiffs say Defendants Peco Foods, Inc., George’s, and Perdue “have indicated their intent to apply redactions to certain unidentified phone numbers.” Jd. at 11.7 According to Defendants, however, none of them has yet determined whether to redact phone records (or the nature or extent of any such redactions) but all of them (not just Peco, George’s, and Perdue) reserve the right to do so. Defendants do not describe the nature and extent of the redactions they are considering. The Court agrees with Defendants that the parties may not yet have reached an unresolvable impasse with respect to a particular proposed redaction. During the December 11, 2017 telephone conference, however, Plaintiffs indicated it would be helpful to have more guidance on the issue of redacting phone records. In the Court’s view, the lack of a more concrete dispute is not so limiting as to thwart the usefulness of analyzing the parties’ arguments and providing general guidance.

' Request No. 2(e) reads: “For each Document Custodian, all: bills, statements, records, and supporting Documents concerning local, long distance, and cellular telephone calls by such employees, including calls made using telephones not paid for by You (such as home office, fax, and personal telephone numbers, personal cellphones, and temporary pay-as-you go cellphones) if such telephones were used for business purposes[.]” [508 at 28]. George’s and Perdue each includes multiple corporate entities. [212, J 40-42, 58-60].

Plaintiffs argue Defendants should not be permitted to redact phone numbers from their telephone records that are irrelevant to the claims and defenses in this case for any number of reasons. Plaintiffs contend the disclosure of phone numbers contained in telephone records does not implicate privacy concerns and, in any event, the designation of the records as confidential under the Agreed Confidentiality Order [202] will provide sufficient protection since those phone numbers cannot be used for any purpose other than this litigation. Plaintiffs also assert the process of redacting phone numbers, meeting and conferring about the redactions, and then processing and analyzing redacted records will impose substantial undue burden and expense on them. In response, Defendants claim the Order regarding Production of Electronically Stored Information and Paper Documents (“ESI Protocol”) [459] contemplates the possible redaction of irrelevant information. Defendants argue that, because Plaintiffs are seeking production of records from phones used for not only business but also personal purposes, it makes sense to allow Defendants to protect their right to redact irrelevant information. On August 15, 2017, the Court entered the ESI Protocol. [459]. The ESI Protocol provides that, “[a] [p]arty may redact irrelevant information that is highly sensitive, medical, or personal information (e.g., health information or Social Security Numbers).” Jd. at 19° It perhaps is a close question whether “highly” modifies “personal” under the series-qualifier canon.’ Although “highly sensitive” and “highly personal” are common enough concepts, the same may not be said of “highly medical.” There are reasons, though, why “personal” should not be interpreted to include any and all personal information. For one, mundane personal information that does not implicate any privacy interest is unlike highly sensitive or medical information. See United States v. Williams, 553 U.S. 285, 294 (2008) (discussing the canon of noscitur a sociis). Further, the one relevant example given in the order—Social Security Numbers—is not run-of-the-mill personal information. See F.7.C. v. EDebitPay, LLC, 695 F.3d 938, 943-44 (9th Cir. 2012) (““E.g.’ signifies that the subsequent examples” are not “exhaustive” but that they “are illustrative.”). In any event, while this is an interesting riff, it does not necessarily resolve the question at hand for a number of reasons including, most importantly, because the Court does not know specifically what “highly sensitive, medical, or personal” information any Defendant wants to redact from the phone records it produces. The Court addressed the redaction of irrelevant information in phone records during a status hearing held on June 16, 2017. The Court said it “generally” does not “allow redactions for relevance.” [430 at 131]. The Court recognized that, in some instances, it may be appropriate to redact irrelevant information. Jd. The basis of such a redaction, however, would not be irrelevance, but, rather, the highly sensitive nature of the information. Jd. Moreover, the Court indicated designating information confidential, rather than redacting it, generally is “the way to go” on protecting irrelevant information of a sensitive nature. /d. at 135. The Court’s thinking has not changed. The phone numbers dialed from or received by a phone typically are not the type of sensitive information that must be redacted before records are > The Order also addresses the redaction of information contained in a document custodian’s “contacts” that is unrelated to the document custodian’s work. [459 at 16-17]. “ Under the series-qualifier canon, “a modifier at the beginning or end of a series of terms modifies all the terms.” United States v. Laraneta, 700 F.3d 983, 989 (7th Cir. 2012); see generally A. SCALIA & B GARNER, READING LAW: THE INTERPRETATION OF LEGAL TEXTS 147-151 (2012).

produced to the other side. See Corsair Special Situations Fund, L.P. v. Engineered Framing Sys., Ine., 2011 WL 3651821, at *2 (D. Md. Aug. 17, 2011) (“An individual does not have a legitimate expectation of privacy in the telephone numbers that are dialed on his or her telephone.”) (quoting Booker v. Dominion Va. Power, 2010 WL 1848474, at *9 (E.D. Va. May 7, 2010)). During the status hearing on June 16, 2017, the Court indicated that, in police cases, phone numbers of family members commonly are redacted because of the serious security concerns implicated by providing those numbers to a plaintiff suing an individual police officer. [430 at 131]. This exception proves the rule, and no party has articulated such a security concern in this case. Further, the Court agrees with Plaintiffs that the burden involved in redacting phone records line-by-line to remove all irrelevant phone numbers and the delay that likely will be occasioned by that effort significantly outweighs the benefits of doing so. Plaintiffs already have agreed all phone records can be designated as confidential under the Agreed Confidentiality Order, which provides adequate protection by limiting the use and disclosure of confidential information to certain persons and for certain purposes. [202 at 4-5]. For all these reasons, the Court concludes Defendants generally should not redact irrelevant information from the phone records they are producing with a confidentiality designation. If the Court is missing some nuance in Defendants’ position, or there is some particularly sensitive information that one or another Defendant is concerned about, and that concern cannot be addressed by the confidentiality designation or it falls outside of the Court’s discussion of the issues above, then that Defendant can bring the issue to the Court’s attention and the Court can address it in the context of a specific dispute. 2. Time Period Issues The parties have agreed that, as a general matter, the appropriate time period for discovery in this case is January 1, 2007 through September 2, 2016. [459 at 12]. Some of Plaintiffs’ Rule 34 requests seek documents that pre- or post- date that time period. Defendants have objected to this expanded temporal scope with respect to certain requests. The disputes now before the Court can be grouped into three categories. A. Documents Related to the Georgia Dock Index and the Georgia Premium Poultry Price Index Request No. 21 of All Plaintiffs’ First Set of Requests for Production of Documents to All Defendants requests documents related to the Georgia Dock Index (“GDI”). [508 at 28]. Plaintiffs want Defendants to produce responsive documents through at least February 28, 2017. Plaintiffs say this extended time period is justified because the Georgia Department of Request No. 21 reads: “From January 1, 2007, through the present, all Documents relating to the Georgia Department of Agriculture and Broilers, Broiler supplies, or Broiler pricing, including the Georgia Dock Advisory Committee, information collection, verification, auditing, and reporting methodologies or procedures, inaccuracies, errors, or vulnerabilities to manipulation of the Georgia Dock price index, any agreement with other Defendants or coconspirators concerning reporting of information for the Georgia Dock, any responses or Communications relating to any investigation (whether formal or informal) concerning the accuracy of information reported to the Georgia Department of Agriculture (including the United States Department of Agriculture’s investigation disclosed publicly in early November 2016), and documents relating to the proposed Georgia Premium Poultry Price Index.” [508 at 33].

Agriculture’s investigation of the GDI and its efforts to replace, potentially, the GDI with the proposed Georgia Premium Poultry Price Index (“GPPPI”) go through at least February 28, 2017. Defendants, on the other hand, propose an end date of December 31, 2016. Defendants note the GDI ceased to exist in the middle of that month and argue the GPPPI is not relevant to this case because Plaintiffs’ claims are based on alleged manipulation of the GDI only. The alleged manipulation of the GDI is one of the core components of Plaintiffs’ claims in this case. See [489 at 14] (noting documents relating to the GDI “are at the core of Plaintiffs’ allegations in this case’); [541 at 15-16] (describing the alleged manipulation). Defendants are correct that Plaintiffs have not alleged a conspiracy to manipulate the GPPPI. Common sense, however, supports the notion that information concerning the investigation of the GDI and the attempt to create a replacement index (namely, the GPPPI) likely could be relevant to Plaintiffs’ allegations about a conspiracy to manipulate the GDI. For instance, during the effort to create the GPPPI, there may have been discussion about how the GDI operated and what went wrong (or right) with the GDI. Although the Federal Rules of Civil Procedure do not countenance wholesale fishing expeditions, the common sense considerations that could implicate a potential nexus between documents relating to the GDI and the GPPPI are sufficient to distinguish Plaintiffs’ request from a fishing expedition. Moreover, Defendants have not raised and developed a burden objection, and the Court does not see why two additional months of discovery in this vein would be disproportionate to the needs of the case. Therefore, the time period for Request No. 21 may be extended to February 28, 2017. B. Pre-Class Period Structured Data Related to the Production, Capacity, Processing, Sale, and Distribution of Broilers Request Nos. 48 through 53 of All Plaintiffs’ First Set of Requests for Production of Documents to All Defendants request a wide array of structured data related to the production, capacity, processing, sale, and distribution of Broilers. See [508 at 40-42]. The parties agree that discovery of this structured data should not be limited to the class period, but they disagree on what the start date should be for the production of that material. Plaintiffs propose January 1, 2002, while Defendants propose January 1, 2004. Plaintiffs say they are willing to agree to Defendants’ shorter period if Defendants stipulate that they will not use information from before January 1, 2004, for expert analysis or otherwise to oppose class certification. Defendants are not prepared to stipulate that time period is adequate without knowing anything about the econometric techniques of Plaintiffs’ experts. It seems implicit in Plaintiffs’ argument, however, that if Plaintiffs want that stipulation from Defendants, then their experts would not be using information from an earlier time period to support class certification. There is no inherently appropriate period of time between the parties’ proposed look-back periods of 4 and 6 years. Defendants cite decisions in which courts set windows shorter than 4 years and Plaintiffs cite decisions in which courts set windows longer than 6 years. The parties, however, neither describe what economic analyses their experts may conduct in this case nor do they explain why those analyses require data for the periods of time they are proposing. Because Plaintiffs want the structured data sought in Request Nos. 48 through 53 to show the impact, if any, of the alleged conspiracy and their damages, if any, this lack of discussion about how the experts in this case may use the data largely thwarts the Court’s ability to determine the relative benefits of producing two additional years of data. On the other hand, it essentially is undisputed

that producing data for two additional years will impose a meaningful burden on Defendants, particularly because it may be more costly for them to recover older data. Four years of pre-class period structured data may be sufficient for all parties to conduct their economic analyses. In fact, no party, including Plaintiffs, says 14 years of structured data (i.e., 2002 through 2016) is necessary in this case. The Court understands Plaintiffs’ concern that Defendants could switch horses down the road by arguing Plaintiffs do not have enough data at the class certification stage. The Court will not require Defendants to stipulate to the adequacy of a given time period now. If Defendants, however, decide to use data from before 2004 in their own analyses, then Plaintiffs also will be entitled to discover that pre-2004 data from Defendants. In other words, if Defendants expand the time period used in their analyses, they will have to supplement their production to Plaintiffs. For now, though, the Court is not convinced it is proportional to the needs of the case to require the production of documents that pre-date January 1, 2004 in response to Request Nos. 48 through 53. At some point before expert disclosures, the parties and, if necessary, the Court should revisit this issue to make sure that any need for the production of more information does not slow down expert discovery or class certification briefing. C. Documents Produced to Government Authorities and Concealment of Antitrust Violations Request Nos. 34 through 37 of All Plaintiffs’ First Set of Requests for Production of Documents to All Defendants request documents produced to government authorities or related to actual or potential violations of laws, regulations, or rules. See [508 at 36-38]. Plaintiffs argue the appropriate time period for these requests is January 1, 2000 to present. Plaintiffs assert evidence related to previous anticompetitive conduct involving broiler chickens or similar products/markets is relevant to their conspiracy claims and to prove absence of mistake. Plaintiffs say it should not be burdensome to locate documents related to previous government investigations. In response, Defendants contend Request Nos. 34 through 37 are so vague and broad in scope that they amount to a fishing expedition. Given the breadth of the requests, Defendants maintain the burden of searching for responsive documents up to 8 years before the alleged conspiracy is disproportionate. Defendants thus propose limiting Request Nos. 34 through 37 to the general time period for discovery. With respect to Request Nos. 34 through 36, the Court agrees with Defendants.° Request No. 34 encompasses all documents “relating to competition in the Broiler industry produced” to any government agency or regulator. Request No. 35 seeks all documents relating “to any possible or alleged violation” of a seven categories of federal, state, or international laws (e.g., anti-monopoly laws, unfair competition laws, and anti-bribery laws) “and any similar laws, rules, or regulations” by “any . . . Broiler-related business or person.” Request No. 36 requests documents relating to “actions to conceal or avoid detection of any potential violations” of such laws, rules, or regulations. Words and phrases such as “competition, “possible,” “similar laws, or regulations,” “conceal or avoid detection,” and “potential” are vague. Moreover, the scope of these requests is incredibly broad and sweeps in large swaths of potentially (and very Plaintiffs indicate they have “offer[ed] additional clarifications regarding the scope” of Request Nos. 34 and 35. [508 at 9]. Plaintiffs do not describe the clarifications or claim they narrowed the scope of the requests,

likely) irrelevant information. For instance, Defendants presumably have produced to government agencies and regulators many documents “relating to competition” that are not relevant to this lawsuit. The cases cited by Plaintiffs actually illustrate the overbreadth of the requests as those cases involved prior conceded price-fixing, guilty pleas, and formal claims or lawsuits involving antitrust issues. See In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 661 (7th Cir. 2002); In re Static Random Access Memory (SRAM) Antitrust Litig., 580 F. Supp. 2d 896, 903 (N.D. Cal. 2008); New Park Entm’t L.L.C. v. Elec. Factory Concerts, Inc., 2000 WL 62315, at *7 (E.D. Pa. Jan. 13, 2000). Because of the expansive scope of Request Nos. 34 through 36, though, requiring Defendants to produce all responsive documents since January 1, 2000 would be unduly burdensome and not proportionate to the needs of the case. Therefore, without more specificity from Plaintiffs about what they want to be produced and a more fulsome explanation as to relevance, the Court will not extend the time for Request Nos. 34 through 36 beyond the default period already agreed to by the parties. The Court views Request No. 37 differently. Plaintiffs now have limited that request to a copy of the consent decree and/or agreement with the U.S. Department of Justice in a prior antitrust case, United States v. Nat'l Broiler Marketing Assoc., et al. (N.D. Ga.). See Nat’l Broiler Mktg. Ass’n v. United States, 436 U.S. 816 (1978). Although the National Broiler Marketing Association case was filed decades ago, the consent decree and/or agreement may at least have some marginal relevance to this case to the extent it played a role in shaping the development of the broiler chicken industry or informed any Defendant’s conduct. Because Defendants have not raised any burden or other objection (beyond relevance) to production of that consent decree and/or agreement, Defendants should produce the document if they have it. 3. Trade Associations Defendants want Direct Purchaser Plaintiffs (““DPPs”) to produce documents related to their participation in trade associations. According to Defendants, DPPs will argue the finder of fact should infer Defendants formed a conspiracy based in part on Defendants participation in trade associations. Defendants contend evidence of DPPs own participation in trade associations is relevant to rebut this argument—at least rhetorically—by showing there is nothing improper about participating in a trade association. DPPs are willing to produce documents concerning trade associations related to broiler chickens or other protein products. But DPPs argue that information regarding non-protein trade associations is irrelevant and that wide-ranging discovery concerning Plaintiffs’ participation in those trade associations would be unduly burdensome and disproportionate to the needs of the case. The Court understands Defendants’ desire to make the rhetorical point, either at summary judgment or during trial, that there is nothing nefarious or improper about participating in a trade association because, among other things, DPPs also participated in such organizations. This theory of relevance, though, is at least somewhat tangential to the claims and defenses in this case. DPPs have not alleged that Defendants mere participation in a trade association (including communication with a competitor at a trade association meeting) shows Defendants conspired to fix prices or manipulate the GDI. Instead, DPPs allege trade association meetings provided Defendants with opportunities to conspire and Defendants took advantage of those opportunities. In other words, DPPs’ claims turn not on the fact that Defendants were members of and participated in trade associations but, rather, on what Defendants did through their participation in trade associations. Of course, it would not be a defense to the alleged illegal anticompetitive

conduct in this case to say DPPs engaged in similar misconduct through their own trade associations. The Court thus finds DPPs’ participation in trade associations is only marginally relevant to the claims and defenses in this case, if it is relevant at all.’ The burden of conducting wide-ranging discovery into DPPs’ participation in non- protein trade associations cannot be justified by the minimal marginal benefit it would provide. As already noted, DPPs have agreed to produce documents, if any, regarding their participation in protein trade associations. They should do so. The rhetorical point that Defendants want to make, however, does not differ with respect to protein and non-protein trade associations. To the extent some discovery related to non-protein trade associations is appropriate, though, there are more proportional and less burdensome discovery vehicles than requiring the production of large volumes of potentially irrelevant documents. For instance, one Rule 34 request served by Defendants asks DPPs to identify documents sufficient to show the trade associations to which they belong and the meetings or conferences they attended. Similar information also could be obtained, perhaps more appropriately and with less burden, with an interrogatory that asks DPPs to identify the trade associations to which they belong or belonged, when meetings occurred, and who attended the meetings. Whether as a document request or an interrogatory, that is the scope of discovery that the Court believes, based on the information presented in the parties’ filings to date, should be permitted on this topic consistent with the scope of Rule 26(b)(1) of the Federal Rules of Civil Procedure. If the parties cannot agree on an appropriate framework for this discovery without more guidance, they can present their competing proposals to the Court. 4. Downstream Data The parties dispute whether Plaintiffs must preserve and, in due course, produce their “downstream data.” The Court understands “downstream data” to encompass information concerning Plaintiffs’ sales of broiler chickens or, in the case of the Commercial and Institutional Indirect Purchaser Plaintiffs (“CIIPPs”) consistent with the document they filed on June 23, 2017, products derived from broiler chickens (429].8 See also [415 at 33-40, 43-50; 434; 506 at 2-4; 522 at 1-2]. As an initial matter, the Court is aware that the parties are meeting with the Special Master about Defendants’ request that Plaintiffs identify and preserve this downstream data, including related structured data sources, and that Plaintiffs will be responding to Defendants’ inquiries about this data by January 12, 2018. Those discussions should continue. Among other things, they may be relevant to the questions raised below. On the merits, the Court has reviewed the parties’ filings to date concerning Defendants’ request that DPPs and CIIPPs identify custodians of and produce downstream data because it

’ To be clear, the Court is not ruling or even implying that evidence about DPPs’ membership or participation in any trade associations would be admissible at any trial of this case. The only point being made is that the Court understands how that evidence could be relevant to a claim or defense in this case for the purposes of discovery. * The CIIPPs’ stipulation that they do not sell broiler chickens seems, at first blush, to be not entirely consistent with the putative classes identified in the CIIPPs’ Third Amended Complaint which speaks in terms of classes of entities “who indirectly purchased Broilers from Defendants or co-conspirators in the United States during the Class Period for use or resale in their business or organization.” [253 at 112 — 115]

could be relevant to class certification. As a general rule, courts have not required anti-trust plaintiffs to produce this kind of information at this stage of the case, a point that Plaintiffs emphasize in their filings. /n re Air Cargo Shipping Servs. Antitrust Litig., 2010 WL 4916723, at *1 (E.D.N.Y. Nov. 24, 2010); Jn re Aspartame Antitrust Litig., 2008 WL 2275528, at *1 (E.D. Pa. Apr. 8, 2008); Jn re Vitamins Antitrust Litig., 198 F.R.D. 296, 301 (D.D.C. 2000). Defendants maintain there are nuances in this case that distinguish it from cases in which courts have not ordered the production of downstream data. They note, for example, that Plaintiffs have brought claims in this case under state laws that permit Defendants to raise a pass-through defense that is unavailable in cases that involve only Sherman Act claims. Defendants’ motions to dismiss those claims recently were denied. [541]. They also point out that End User Consumer Plaintiffs (“EUCPs”) already have served document preservation subpoenas on Defendants’ largest customers for transactional data that EUCPs say will be “used to analyze the extent to which the overcharges are passed through the chain of distribution.” [415 at 46]. Defendants contend that is precisely the information they are seeking from Plaintiffs. Defendants also contend downstream data could be relevant to “issues of commonality, typicality, and adequacy” in the context of class certification.” After reviewing the parties’ submissions, the Court still does not have a good handle on a number of issues. That is not surprising because the downstream data issue was raised initially in a joint letter brief filed on June 14, 2017 [415], in preparation for a scheduled status hearing concerning a number of other matters and then supplemented to some extent by the parties’ letters addressing Rule 34 requests for production and objections discussed above. Thus, fulsome briefing on the specific issue of the production of downstream data has not yet occurred. The parties’ submissions to date on this issue are very general, perhaps because the procedural context in which they were presented lent itself to general, 30,000-foot presentations. The Court is by no means criticizing the parties’ submissions with this comment. For example, however, the Court has questions such as (1) What specific categories of information are encompassed within Defendants’ downstream data discovery requests; (2) How is this information relevant to class certification specifically in this case, in Defendants’ view, other than with respect to generalized issues of commonality, typicality, and adequacy; (3) How, if at all, does CIIPPs’ clarification of the scope of their proposed class [429], which was served after the parties’ submitted their positions on the downstream data issue on June 14, 2017 [415], impact the relevance or discoverability of those Plaintiffs’ downstream data; (4) To what extent is the fact that the EUCPs may purchase broilers from entities other than CIIPPs relevant to this inquiry; (5) Are Plaintiffs raising a burden objection to producing downstream data and, if so, what is the basis for that objection; (6) Is there a less burdensome method by which Defendants can obtain the type of information they say they are seeking; (7) Are there alternative sources from which Defendants can obtain the same or similar information with less burden on Plaintiffs and assuming appropriate stipulations concerning how that information can be used in this case;

Defendants make this argument very generally and without any specifics. The Court notes that similar arguments have been rejected by other courts in terms of the relevance of downstream data to class certification and the balancing of the burden and benefit of producing the information. Jn re Skelaxin (Metaxalone) Antitrust Litig., 292 F.R.D. 544, 553 (E.D. Tenn. 2013); In re Aspartame, 2008 WL 2275528, at *4—6.

and (8) What distinguishes this case, specifically, from other anti-trust cases in which courts did not order the plaintiffs to produce downstream data. The Court would like Plaintiffs and Defendants to file supplemental briefs, not to exceed 15 pages, addressing the discoverability of Plaintiffs’ downstream data and the questions the Court has raised above. In these briefs to be clear, the parties are not limited to addressing only the questions identified above by the Court. Given the approaching holidays, however, and the ongoing discussions with the Special Master about the identification of structured data that may have a bearing on the parties’ positions on at least some of the questions raised above, the parties should confer about an acceptable briefing schedule and file a short report proposing that schedule on or before December 29, 2017. If the parties would prefer to propose a briefing schedule after they have reviewed Plaintiffs’ response to Defendants’ November 22, 2017, letter, which response is due to be served on January 12, 2018, as noted above, then that also is acceptable. The Court would suggest simultaneous briefs from each side, followed by simultaneous responses, but if the parties would prefer a different format, they are free to propose one. It is so ordered.

ffrey T. Gilbert United States Magistrate Judge

Dated: December 22, 2017