A federal court has denied a defendant-employer’s request that plaintiffs sift through and turn over all their social media posts made during their work hours in an FLSA collective action in which the plaintiffs claim their employer failed to give them meal breaks. How did that happen? I thought you’d never ask.
By way of background, Jewell v. Aaron’s Inc., is a nationwide,1,700+ FLSA collective action pending in the Northern District of Georgia. In the suit (Complaint accessible here), the class plaintiffs (current and former employees of Aaron’s) claim they were not paid for their 30-minute meal periods. As you might imagine, with that many plaintiffs discovery has been difficult. So with that, the parties got creative. They were able to work together to narrow the issues and determined that of the 1,700+ class members, discovery would only be served and responded to by 87 of the opt-in plaintiffs (the “Discovery Plaintiffs”). I won’t go into all the details about that discovery (you may read the Defendant’s Memorandum in Support of Motion for Court Approval of Discovery Request to a Small Number of Randomly-Selected Opt-In Plaintiffs here), because I want to focus on the social media portion of the discovery dispute.
Specifically, in one request for production, Aaron’s asked the Discovery Plaintiffs to produce:
“All documents, statements, or any activity available that you posted on any internet Web site or Web page, including, but not limited to Facebook, MySpace, LinkedIn, Twitter, or a blog from 2009 to the present during your working hours at an Aaron’s store.”
To support its argument that the requested information was relevant, Aaron’s produced a Facebook post from the named plaintiff (“Plaintiff Jewell”) that stated that Plaintiff Jewell was taking a lunch break: “At workkkk…on lunch….ready to go home… work two hrs in am then offffff for the day.” The date and time of the post was illegible on the printout Aaron’s received, and Aaron’s argued the date/time stamp should be legible “if provided by Plaintiff.” Aaron’s then argued that other postings (whose date and time stamps were also illegible) also suggested that Plaintiff Jewell was using work time to make his personal posts on Facebook, i.e., take a break from work, including:
- “Hmmm the joys of driving over a hundred miles in three cities to collect payments of non renewed accounts…..life of a cam on Saturdays.”
- “At work til 6pm… lots of work to do today…so very little time to do it”
- • It may be 930 am on a Saturday….but I have this think I do every Saturday between 8 and 9…..its called work lol”
- “About to be off work oh how I can’t wait….”
- Hmmm….ready to be off work….”
- En work blows today…Ready to be at home chillin…”
- Sitting here at work…kinda sleep”
- At work.. hmmm all I am gonna say”
Aaron’s argued it was likely that many of the opt-in plaintiffs made similar posts, that some of the posts may directly show that the poster was taking a lunch break at the time, and that the date and time stamp of other posts may indicate that the poster spent a chunk of 30 minutes or more during the work day engaged in a series of successive personal posts such that there would be a 30-minute period the poster’s work day that would be appropriately excluded from the compensable time the poster could claim he or she could not take breaks during the day because of work pressures. The speculative “it was likely,” “may directly show” and “could indicate” language became significant in the court’s ultimate decision on the issue.
Plaintiffs’ counsel objected to the production, claiming it would be unduly burdensome to review the social media posts of all the Discovery Plaintiffs and determine which ones were made during work hours. Plaintiffs estimated the process of sifting through all the social media in order to retrieve the requested information would take anywhere from 1,323 hours to 26,462 hours, depending on the number of posts each Discovery Plaintiff made per day.
Even though evidence that employees were not engaged in work functions during their scheduled shift is certainly relevant to whether employees are getting breaks — the issue before the court — the court determined that the burden of having to go through 87 Discovery Plaintiffs’ Facebook pages to extract posts made by each one during their working hours was too burdensome and denied the employer’s request. The opinion is available here.
Siding with plaintiffs, the court noted that Facebook does not have a search function (it has Graph Search but that is something different) and that Facebook posts do not contain a timestamp. Therefore, the information Aaron’s wanted, the dates and the times of the posts, was information that could be obtained only by “individually interacting with and clicking each post.”
The court also attempted to verify the accuracy of the plaintiffs’ assertions regarding the potential burden and pointed out Facebook’s applications that allow a user to download Facebook data, including “timeline” information, “wall” postings, activity log, messages and photographs directly from the website, and that once downloaded, the user “may view all posts/activity in a single document in chronological order with a date/time stamp.”
Even with the use of the Downloaded Information function (and other functions Facebook now has that allow users to retrieve their information), the court still determined the request too burdensome and not reasonably calculated to lead to the discovery of admissible evidence:
“Defendant has not made a sufficient predicate showing that the broad nature of material it seeks is reasonably calculated to lead to the discovery of admissible evidence….The court finds that the burden imposed on a class of plaintiffs to produce such an overly broad swath of documents, while technologically feasible, is far outweighed by the remote relevance of the information.”
The court took issue with the Aaron’s evidence in support of its request, i.e., the Facebook postings from just one of the plaintiffs, and determined that the postings of one plaintiff (especially in the broader context of a large collective action) was simply insufficient to persuade the court to force the plaintiffs to sift through the social media accounts of the 87 Discovery Plaintiffs all in the hopes of discovering that they may have made posts that may indicate that each took a break while at work. Even if a plaintiff made a social media post while at work, it still would not be able to demonstrate whether the plaintiff received a bona fide meal period as defined by the FLSA or not. Thus, Aaron’s’ argument was supported by “nothing more than its ‘hope that there might be something of relevance’ in these plaintiffs’ Facebook, Twitter, and/or MySpace accounts” and hope and wishful thinking are simply not enough for a court to open the doors to this type of social media evidence — certainly not in a case where the requested discovery is for access to 87 individuals’ social media accounts and the evidence, even if discovered, at best, would do little more than chip away at a the damages calculation.
Takeaways. While I think this case is a significant one — as far as I know it is only one of a handful of well-reasoned decisions regarding this type of evidence in an FLSA case — I don’t think it is the death knell for defendant-employers seeking this type of information in FLSA collective actions. I think the problem here came down, in part, to the “open the door” evidence, i.e., the Facebook posts from one plaintiff out of 87 Discovery Plaintiffs. If you are going to ask for access to that much social media information, you need to come with both barrels loaded, not a water gun. Aaron’s simply did not provide the court with enough “open the door” evidence that could allow the court to determine that other evidence existed that was worth the time and burden to pursue.
The second issue was how Aaron’s intended to use the evidence in the case. While the court did not go into full detail, it did note that “whether or not an opt-in plaintiff made a Facebook post during work may have no bearing on whether or not the opt-in plaintiff received a bona fide meal period…” Absolutely. Depending on how you access social media, a user can make a post in a matter of seconds. Heck, you can schedule posts to publish at various times throughout the day and poster does not even have to be near a computer or Smartphone when the posts actually gets posted. These 30 seconds or even few minutes of social media activity are just that, second and minutes. When the issue is whether or not an employee received a 30-minute break period, it simply does not make sense to try to piece together random social media posts in the hopes of demonstrating that on a particular day a user was on social media for 30 minutes during work time. At the end of the day, this type of evidence only chips away — and very slowly mind you — at the damages calculation, but it is hard to see how this would be helpful in a large collective action and a meal break issue.
There is an application on Facebook that identifies when the user is logging in/off and from where. As of right now, this is available through Facebook’s Downloaded Information feature (Facebook changes where information can be found sporadically and typically without notice). This identifies the IP address that accessed the Facebook account and the the date and time associated with the login/off. As I have indicated before, I think this type of information may prove applicable in FLSA cases because it can show where the user is posting and when the poster is logged on and off; however, it may not prove helpful if the user maintains a continual login.
So employers, here are my tips on the issue:
- Get good “open the door” evidence. If you want to argue to a court that relevant evidence is out there, you have to be able to show that the evidence is indeed out there, not hypothetically speaking may be out there and may show something that might be somewhat relevant to the case, though not likely. This typically comes from your employees, or in other words, the employees who don’t like other employees who they believe are taking advantage of the company or being treated more fairly or, in many cases, they just don’t like the person, whatever. Many times, your employees are your evidence gatherers. You’ve already been told you should not to be “friends” or “followers” of your employees, so you aren’t getting this information, right? And with employees only beginning to learn how to use privacy settings effectively, much of the information remains, to some degree, public. What this means is that you should foster an open door environment where employees feel free to share this information and, at the very least, provide an anonymous reporting mechanism, which is how Aaron’s obtained Plaintiff Jewell’s social media posts it used in its evidence, to allow your employees to provide this information anonymously if they so choose. I would not encourage a snitching or tattletale environment because, unless you are a high school principal, you should leave such antics where they belong — buried in the past.
- Make sure the evidence you are seeking will ultimately be relevant to your case. Make sure the evidence you are seeking will, at the end of the day, be relevant to your case. In this case, Aaron’s did a decent job with their discovery requests and in asking the court for social media evidence — the request was narrowly tailored to the issue and identified the specific evidence requested. I just think the evidence the defendant was asking for was, like the court said, too burdensome to extract and, at the end of the day, not really helpful in getting to the ultimate issue of the case. In a smaller FLSA case, maybe it would have been feasible, but not in one with 1,700+ plaintiffs and not based on the postings of a single plaintiff — unless of course it’s a single-plaintiff case.
- Know social media. If you are going to ask for social media, know what is out there. There are three applications through Facebook that contain information that might contain relevant evidence. Don’t limit your discovery requests to just posts and photographs. See what else is available and narrowly tailor your request to get it. For example, in trying to determine whether an employee was really sick at home and not on in another state on vacation, see where the employee has checked in through the “check in” function.