ORDER

MIRANDA M. DU, District Judge.

I. SUMMARY

This is a class action involving a dispute over the terms of employment for armed guards hired to work in Iraq. Before the Court are Defendants SOC LLC; SOC-SMG, Inc.; and Day & Zimmermann, Inc.'s (collectively, "Defendants") objection (ECF No. 316) to an order issued by Magistrate Judge Peggy A. Leen (ECF Nos. 315 (minutes), 317 (transcript)). The Court has reviewed Plaintiff Karl E. Risinger's response (ECF No. 318). For the following reasons, the Court overrules Defendants' objection.

II. BACKGROUND

Plaintiff plans to use the results of a survey to prove the class's damages at trial. (ECF No. 281 at 4.) The survey essentially asked respondents to estimate how often they worked more than six twelve-hour days per week. (See ECF No. 268-7 at 7-8.) In an earlier motion, Defendants sought to exclude the survey results under Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993). (ECF No. 268 at 2.) The Court did not exclude the survey results altogether but did preclude Plaintiff's expert from extrapolating the survey results to the entire class. (ECF No. 281 at 12.) Defendants once again seek to exclude the survey results, this time based on an alleged discovery violation. (ECF No. 316 at 17.)

As part of damages discovery,1 Defendants requested production of (1) "all documents that relate to or support the statements in the signed declarations" that Plaintiff relied on in moving to certify the class; and (2) "all communications from any Class member concerning the hours any Class member worked while employed in Iraq for SOC." (Id. at 6 (citing ECF No. 244-2 at 5-6).) In response, Plaintiff's counsel produced a Facebook message thread between Plaintiff and various class members that was partially redacted. (See generally ECF No. 244-19.) Plaintiff did not produce any emails between himself and class members. (ECF No. 253 at 7.)

Defendants moved to compel production of the Facebook messages without redactions (ECF No. 244 at 2), and Plaintiff opposed on the basis of attorney-client privilege. (ECF No. 253 at 7.) Plaintiff argued that the Facebook messages were properly redacted because they were copied and pasted from emails class counsel sent to Plaintiff. (See id.; see also ECF No. 266 at 41.) Defendants also moved to compel production of emails between Plaintiff and class members. (ECF No. 244 at 2.) Plaintiff's counsel responded that Plaintiff was unable to locate any emails. (ECF No. 253 at 20-22.) The Magistrate Judge ordered Plaintiff to make his computer available for forensic examination. (ECF No. 266 at 42-43.)

A forensics vendor collected thousands of documents from Plaintiff's computer based on a protocol the parties developed, and Plaintiff produced 155 of them as responsive. (ECF No. 316 at 8; see also ECF No. 310-7 (production email from class counsel).) The Facebook thread that was produced earlier was produced again—this time with fewer redactions. (Compare ECF No. 244-19 ("April Production") with ECF No. 310-8 ("October Production").) Defendants essentially contend that the messages show a class member—Chris Hedberg—was coached to inflate his survey responses. (ECF No. 316 at 11.)

Defendants moved for sanctions, requesting preclusion of the survey at trial or reopening of relevant depositions, and monetary sanctions. (Id. at 8.) The Magistrate Judge denied Defendants' requested form of relief but ordered Plaintiff to pay half of the forensic vendor's fees and to produce 1,206 emails. (ECF No. 317 at 29-30.) The Magistrate Judge did not sanction Plaintiff for redacting the Facebook messages because "on balance . . . what the plaintiff didn't produce is more helpful to the plaintiff than it is to the defense." (Id. at 29.) The Magistrate Judge found that the Facebook messages "do not suggest . . . manipulation by the plaintiff or willful withholding." (Id.)

III. LEGAL STANDARD

Magistrate judges are authorized to resolve pretrial matters subject to district court review under a "clearly erroneous or contrary to law" standard. 28 U.S.C. § 636(b)(1)(A); see also Fed. R. Civ. P. 72(a); LR IB 3-1(a) ("A district judge may reconsider any pretrial matter referred to a magistrate judge in a civil or criminal case pursuant to LR IB 1-3, where it has been shown that the magistrate judge's ruling is clearly erroneous or contrary to law."). "This subsection . . . also enable[s] the court to delegate some of the more administrative functions to a magistrate, such as . . . assistance in the preparation of plans to achieve prompt disposition of cases in the court." Gomez v. United States, 490 U.S. 858, 869 (1989). "A finding is clearly erroneous when although there is evidence to support it, the reviewing body on the entire evidence is left with the definite and firm conviction that a mistake has been committed." United States v. Ressam, 593 F.3d 1095, 1118 (9th Cir. 2010) (quotation omitted). A magistrate's pretrial order issued under 28 U.S.C. § 636(b)(1)(A) is not subject to de novo review, and the reviewing court "may not simply substitute its judgment for that of the deciding court." Grimes v. City & County of San Francisco, 951 F.2d 236, 241 (9th Cir. 1991).

IV. DISCUSSION

The Magistrate Judge did not clearly err in finding that class counsel lacked bad faith. The messages that were redacted in the April Production but produced in the October Production do not suggest that class counsel manipulated or willfully withheld information harmful to Plaintiff's case.

In one set of previously redacted messages, the class members were attempting to determine whether an attorney who contacted a particular class member—Jose Villa— represented Defendants or Plaintiff. (See ECF No. 310-8 at 10.) Risinger noted that he could not confirm "if this is our side or theirs" because "Gizer's office says he is out on trial at the moment." (Id.) Villa responded that the attorney claimed to represent Plaintiff. (Id.) Risinger then requested that the class members refrain from talking to any attorneys until they spoke with class counsel: "Okay, here's the deal . . . these subpoena's [are] coming from SOC's lawyers! DO NOT talk to anyone until our lawyers reach out to you and prepare you for the deposition." (Id.) A few days later, Risinger reiterated that point: "Notices went out from our side to opt-in. Their side has been sending out subpoenas for depositions. Don't do the depo until you have been contacted by our side prior to the deposition. Not everyone will be asked. They are trying to find loop holes. You MUST consult with our lawyers before giving a statement!" (Id.) These messages are innocuous—they do not suggest that class counsel deliberately withheld information harmful to Plaintiff's case.

Another set of previously redacted messages is similarly innocuous. (See id. at 3-4.) A class member asked whether he should physically mail a completed survey questionnaire to class counsel, and Risinger responded that he did not know. (Id. at 3.) Related messages simply indicate that some class members left voicemails for class counsel with no discussion of the substance of those voicemails. (Id. at 4.)

The last set of previously redacted messages also does not necessarily suggest bad faith. A class member—Chris Hedberg—asked for clarification about the survey: "Hey Karl,[ ]just got this soc questionnaire from our lawyers. Are they counting from pickup time to drop off time as "shift time"? From weapon pickup to drop off or what? Filling it out now and if it's pickup to drop off, it'll be 100%, 13-14 hours. Lol." (Id. at 8.) Risinger confirmed Hedberg's assumption: "Yes, pickup to drop off." (Id.) Another class member—Matt Olson—also confirmed: "Yep 14 hours is about right. And then include training which easily extended. Also I remember Karl specifically having to come in on a `day off' for training. As I'm sure many of us did." (Id.) Hedberg wondered if he was underestimating his survey responses: "Yup[.] I said between 30-40% of the time I worked 7 days a week, I was there for 15 months. I know for a fact I worked the first 5 weeks straight! Is that lowballing it Karl?" (Id.) Risinger opined: "Definitely lowballing. Considering the time sheets I turned in. We went 3-4 months straight without a day off and then sporadic week to week with a day off here and there." (Id.) Hebderg asked: "So more like 60-70%?" (Id. at 7-8.) Risinger responded: "Yeah, 2/3d." (Id. at 7.) Risinger then sent out the Word version of the questionnaire and said: "If you have not received this yet, please fill out and follow the instructions to return to our lawyers. Please pass it along to everyone you know that is not in this group. Thanks!" (Id. at 7.)

This third set of messages does not suggest bad faith because—as the Magistrate Judge noted—the messages are helpful to Plaintiff's case. They show that the class members recalled working more than six twelve-hour days per week and that Hedberg merely sought confirmation about his estimates. Neither Risinger nor Olson encouraged Hedberg to fabricate or inflate his responses. Rather, Risinger reassured Hedberg that Hedberg's own estimates were accurate. Olson reminded Hebderg that he may have had training on a "day off." Even if Risinger and Olson's messages could be viewed as coaching Hebderg, the Court is not free to disturb the Magistrate Judge's implicit determination that this was not the case. The Magistrate Judge did not clearly err in finding that the previously redacted messages—considered collectively—did not suggest bad faith.

Defendants argue that the Magistrate Judge mistakenly relied on messages that were produced in both the April Production and October Production (ECF No. 316 at 11-12), but the examples that the Magistrate Judge offered to support her finding were just that—examples (see ECF No. 317 at 29). The Magistrate Judge expressly noted that she "really did read every single one of these entries" and that "on balance, I will tell you my impression is that what the plaintiff didn't produce is more helpful to the plaintiff than it is to the defense." (Id.) The Magistrate Judge's determination was based on the entirety of the previously redacted messages—not just the examples that the Magistrate Judge cited.

As additional evidence of bad faith, Defendants rely on a purported contradiction in class counsel's explanations for redacting the Facebook messages. According to Defendants, class counsel initially represented that the Facebook messages were redacted because Plaintiff was parroting class counsel's emails but later represented that the messages were redacted because they revealed class counsel's theory of the case. (ECF No. 316 at 12-13.) Plaintiff responds that the messages were redacted based on a good faith belief that the communications about the survey were privileged regardless of whether they were copied and pasted from emails between Risinger and class counsel. (See ECF No. 318 at 6.) Class counsel's conduct does not necessarily suggest bad faith— legal strategies evolve. And even if class counsel's behavior were susceptible to an inference of bad faith, there is insufficient evidence in the record to demonstrate clear error on the part of the Magistrate Judge. At most, class counsel's representations permit multiple inferences, and the Court is not free to disturb the Magistrate Judge's selection among those inferences.

The Magistrate Judge's ruling also was not contrary to law. Defendants argue that Judge Leen failed to mitigate prejudice to Defendants and stop Plaintiff's pattern of repeated discovery violations (ECF No. 316 at 13), but there is insufficient evidence to connect any prejudice Defendants suffered to bad faith conduct on the part of Plaintiff's counsel. Rather, the evidence before the Court permits the inference that class counsel redacted the Facebook messages in good faith, an unintended consequence of which was to potentially limit Defendants' evidentiary basis for seeking exclusion of the survey. And while Defendants contend that Hedberg was coached to inflate his survey responses, the Court explained supra that the messages may also be viewed as innocuous. Defendants also argue that Plaintiff must be deterred from future misconduct (id. at 16), but the Magistrate Judge's sanction requiring Plaintiff to pay half the forensic vendor's fees does that.

The only binding authority Defendants cite in support of their argument that the Magistrate Judge ruled contrary to law are Trulis v. Barton, 107 F.3d 685 (9th Cir. 1995), and Wanderer v. Johnston, 910 F.2d 652, 656 (9th Cir. 1990). But in Trulis sanctions were warranted because "[t]he uncontroverted facts demonstrate[d] subjective bad faith as a matter of law." 107 F.3d at 694. Here again the facts support Judge Leen's finding of a lack of bad faith. And in Wanderer sanctions were appropriate because "[t]he existence of prejudice [was] palpable"—defendants failed to appear at their depositions and repeatedly flouted court orders to produce documents. 910 F.2d at 656. Plaintiff has not engaged in such egregious conduct here.

Accordingly, the Court overrules Defendants' objection.

V. CONCLUSION

The Court notes that the parties made several arguments and cited to several cases not discussed above. The Court has reviewed these arguments and cases and determines that they do not warrant discussion as they do not affect the outcome of the objection before the Court.