ORDER
GREGORY G. HOLLOWS, Magistrate Judge.
Introduction and Summary
This case involves candles which allegedly explode or otherwise combust in such a violent fashion that persons using them can be burned by hot wax smattering exposed areas of their bodies—hereafter "flashovers." Plaintiff made a discovery request, and later a motion, attempting to acquire information relating to all of the candles distributed by defendant Bath & Body Works, including design, manufacture and similar information, as well as an attempt to ferret out information pertinent to all flashover incidents for candles of any types which defendant placed in the stream of commerce. Defendant opposed in the main with a burden argument with respect to all requests arguing, in essence, that the sheer number of candles (hundreds of millions over the years) distributed by defendant made compliance with plaintiff's "any candle" discovery very unreasonably burdensome. The undersigned accepted, in significant part, the proffer of burden and significantly narrowed the scope of the discovery requests, i.e., to only candles employing three wicks, the type of candle with which plaintiff claims injury.
One of the important series of requests by plaintiff involved requests for incidents reports of candle flashovers involving other persons. These particular incident requests were also opposed on burden/lack of proportionality grounds, and in particular, opposed because of the necessity of manually identifying the "relevant" flashover incidents.
After obtaining the initial, limiting discovery order, plaintiff went about taking depositions of various of defendant's personnel, including persons working in a related corporation who engaged in consumer interaction for defendant's products as well as products of sister corporations. Through efforts other counsel may not have employed, plaintiff discovered the computerized incident "reason codes" for the consumer interaction complaint data base, and much potentially relevant discovery about flashover incidents was eventually disclosed. The irony of this sanctions motion is that, as it turned out, ferreting out flashover incidents for any candle was easier to identify as opposed to ascertaining those incidents for just three wick candles or Eucalyptus Spearmint three wick candles. Defendant had created incident codes for product complaints, and specific codes had been developed for flashovers and other abnormal candle flame incidents for all candles distributed by defendant. Further manual investigation of the text of complaints, if any was indeed necessary, was a defendant self-created burden.
Plaintiff ultimately may not have been harmed in terms of discovery production substance, given that purposeful misinformation was initially given counsel with respect to flashover incidents in pre-hearing discovery resolution efforts and later in court, i.e., plaintiff ultimately received a comprehensive list of flashover incidents for at least three wick candles. Nevertheless, plaintiff has moved for sanctions asserting that defendant and counsel deceived counsel and the court in the discovery process. At the very least, plaintiff asserts that defendant was not substantially justified in maintaining the position that identification of flashover incidents was burdensome.
For the reasons set forth herein, sanctions are appropriate for 50% of plaintiff's counsel's costs in the initial motions to compel and the motion for sanctions.
The Discovery Requests and Responses
As stated above, the fairly comprehensive discovery requests, primarily requests for Production of documents, sought to discover much information about all of defendant's other distributed candles. For example:
Request No. 14
Please produce all DOCUMENTS EVIDENCING the design of the CANDLE, including but not limited to design drawings, shop drawings, and manufacturing specifications.1
Defendant opposed producing any information other than that for the specific candle which plaintiff alleged injured her, i.e., a three wick candle with a Eucalyptus mint fragrance.
A good portion of the discovery requests sought to ascertain information about flashover incidents:
Request No. 6
Please produce all DOCUMENTS, including but not limited to all COMMUNICATIONS, EVIDENCING any investigation of other incidents in which a candle marketed, sold, or distributed by DEFENDANT allegedly caused personal injury or property damage.
Request No. 12
Please produce all DOCUMENTS EVIDENCING any sudden, unexpected flaring of the CANDLE, including but not limited to the OTHER CANDLE INCIDENTS.
Request No. 13
Please produce all documents EVIDENCING the cause of any sudden, unexpected flaring or any alleged sudden, unexpected flaring of the CANDLE, including but not limited to OTHER CANDLE INCIDENTS.
As set forth in footnote 1, defendant responded only as if each request specified the specific type of candle which injured plaintiff:
"Objection. Overly broad, unduly burdensome and not reasonably calculated to lead to the discovery of admissible evidence. Additionally, this request is vague and ambiguous and key terms are not defined. Plaintiff's definition of "the CANDLE" is so broad that it includes a great number of different and dissimilar products. . . . Defendant agrees to search for and produce documents or communications evidencing any investigation relating to claims involving allegations of personal injury or property damage as a result of a high flame or sudden high flaming, associated with the use of the candle at issue."
ECF No. 36 at 14-15 specifically responding to Request 6 (emphasis added).
The Motion to Compel Further Responses
Although it is not precisely clear, defendant appeared to have initially produced only flashover incident information for seven three wick Eucalyptus Spearmint flashover incidents, and sixteen Consumer Product Safety Commission public noted incidents for any candle. Plaintiff moved to compel discovery for the universe of flashover incidents.
Defendant opposed the "flashover" incident requests as broadly defined by plaintiff on the following grounds —
1. The requests were burdensome and not proportional to the needs of the case in that plaintiff had not made any showing that all candles had a similar flashover potential;
2. Privacy interests of third parties would be impinged;
3. Post-incident information was not relevant at all.
Given that the primary thrust of the discovery was plaintiff's desire to obtain information about other flashover incidents, defendant proffered two declarations — Steven Smith, who discussed the millions of candles distributed by defendant, as well as the difficulties in extrapolating flashover tendencies of different type of candles, and Tabitha Vaughn who focused on the difficulties of supplying flashover information even of just three-wick Eucalyptus Spearmint candles. Importantly, the Vaughn declaration (ECF No. 36-8) proffered:
6. Unfortunately, the database does not allow for Boolean searches to be conducted. In order for a particular product to be tracked, a filter is applied for that product's SKU number.
7. If a product is not identified by a consumer, or if the SKU is not known, then the consumer or customer service representative sometimes refer to the product in the "case text" section of the VOC.
8. The search process is therefore part automatic and part manual. First, a filter search is applied to narrow the retailer (i.e. BBW) and product type. Then, a manual review of all "case text" sections occurs to look for a product that is similarly named as what is being searched for. This process is time intensive and requires significant resources.
9. By way of illustration, I searched the database for all consumer concerns identifying "Eucalyptus Spearmint." The preliminary results included all types of candles, lotions, sprays, soaps, wall plug-ins, car fragrances and body washes with that fragrance.
10. A manual review was then required to separate the non-candle concerns.
11. It is important to note that the database captures all consumer concerns and comments. This varies from positive experiences with a product to potentially negative incidents. By way of illustration, a customer may report that they don't like the strength of a fragrance, or that they referred an old container that is now discontinued.
12. When the consumer's concern is entered into the database, the system tracks each consumer's comment separately. This means that any given phone call can produce several entries for one consumer. If a consumer calls about more than one product, the system runs a report for each concern or comment the consumer detailed. By way of example, if a consumer contacted BBW about a Vanilla Bean lotion and informed us they did not like the smell, and thought the lotion was too thin, that would register as two separate entries in the system.
13. It took two people nearly 10 days to search for and compile results of all consumer claims related to the Eucalyptus Spearmint, 3-wick candle alone. Another 10-12 hours on these Eucalyptus Spearmint results is anticipated to filter the results for accuracy and responsiveness to plaintiff's request. We are in the process of producing those results to counsel for production in this case.
14. Accordingly, the process to respond to specific requests for each any every candle system sold by BBW would be extraordinarily cumbersome, time consuming and burdensome.
The Hearing
I will point out, Judge, something that may have gone unnoticed here, but there was an affidavit from a Ms. Tabitha Vaughn, in-house at L Brands, I think. And it's attached to the Joint Statement. And she talks about, "Oh, this is a great burden on, on this company to have to go through and — `cause the way we've set up our system it's difficult to find claims relating to this product and to this particular candle. And then when we get that particular candle we have to, in many instances, manually read all of the entries to see if it involves unexpected flaring or explosions."
Plaintiff's Counsel Argument, ECF No. 49 at 9.
MR. OSTERMAN: Sure. And that's what we tried to address in the Tabitha Vaughn affidavit. What she's saying is, "For us to check our claim system," okay, "for us to, to check our claim system, we are looking for, first of all, what product. We can't search by eucalyptus mint because we sell — we're a fragrance company. We're, we're a company that sells fragrant products. So we sell air fresheners in cars. We sell lotions, creams, lot of things that have the word eucalyptus mint. We've got to look for eucalyptus mint candle. And sometimes that's checked in a — that is something that — that the — that is supposed to be checked, but it's not always and faithfully entered that way. So the only way we can be sure of what product we're looking for when we first narrow it to eucalyptus mint, that can take the claims database for, for a company that sells thousands of different products a year and
* * *
So every, everything that comes up as eucalyptus mint candle they're then going into and they have to read because they can't search for, they can't do string searches. They can search for the word, as I understand it, they can do one-word searches, but we get things like "flame" or "fire." For every candle claim, or a high percentage of them, is going to be talking about, "I didn't like how it burned. It produced soot. The flame went out. The flame was too high. The flame was too low. The flame seemed off center. It was up against the glass." Those are all kinds of things that get, that get entered in and somebody has to manually go through and look for something that's consistent with what they're describing. That's what Tabitha Vaughn explained to us and what we were trying to put into the affidavit to her, that it is a labor intensive, time-intensive process to search for all that.
Defense Counsel Argument, ECF No. 49 at 19-20.
The Order
Despite not specifically referencing the Vaughn declaration, the undersigned was persuaded by that declaration and the Smith declaration that permitting a search of all candles for flashover events was not proportional to the needs of the case in terms of burden—especially because of the necessary manual search. As set forth in the first discovery order:
. . . and Defendant has shown that to respond with regard to all of the various iterations of candle it markets would be a daunting exercise, the court will, at this point limit the scope of production to that involving all triple wick candles regardless of fragrance as the similarities between this formulation of candle would appear to the court to be most relevant to Plaintiff's ultimate burden of proof.1 Defendant need not, therefore, produce documents regarding other candle configurations at this point. [Fn 2]
ECF 40 at 4
Footnote 2:
Defendant protested at hearing that even this limited discovery is too much. Defendant contends that although its product files, or complaint files, are somewhat computerized, it is not possible to perform a Boolean search (key word search) on these files. If this is so— time to retain an IT expert who can perform this task if a manual search is too burdensome. Also, for the first time, after exclaiming in its part of the Joint Statement that the candle market is comprised of a great many types of candles, and this is why the burdensome, non-proportional discovery should be denied, Defendant's counsel at hearing asserted that three wick candles comprise the vast, vast majority of its candles sold, and that the court's proffered limitation is unworkable. This different assertion is too little-too late, unverified, and seemingly doubtful.
Post-Order Discovery
Plaintiff's counsel, focusing on candle mishaps, engaged in post-order discovery to ascertain whether there was indeed a computerized method to discover flashover incidents. And indeed there was.
• Customer comments are assigned a "reason code" in BBW's customer database system (Exhibit 2, Tabatha Vaughan Dep. Tr. 21:13-24:13,31:7-20, 42:6-9, 44:15-46:2 (Mar. 14, 2018));
• BBW has at least two reason codes for sudden and unexpected flaring. The first is the code for a candle fire (CPR CD FRE). The second is the code for a candle flashover (CPRU CN FLS) (id. at 29:2-8, 31:7-13);
• By running a search for candle fires and candle flashovers, all noncandle products would be eliminated (id. at 32:23-33:11, 75:24-76:6, 77:4-79:15);
• BBW could identify all candle fires and candle flashovers involving the Eucalyptus Spearmint candle by applying only two filters: reason code and product SKU (id.);
• BBW could identify all candle fires and candle flashovers of its three wick candles using the same two filters (id.);
• It did not take two people "nearly 10 days" to search for consumer claims regarding the Eucalyptus Spearmint, three-wick candles. Instead, those two employees only worked one-to-two hours per day in each of those 10 days and devoted the rest of their time to their regular work duties (id. at 95:13-25.)
Plaintiff's Motion for Sanctions, ECF No. 57 at 3.
Defendant proffered its version of the important facts from the Vaughn deposition:
Q. How long did it take for [the head of IT] to gather all the entries relating to candle fires or flashovers?
A. It took an estimated 8 to 10 hours. Originally, we thought it would be 10 to 12, but it was a little less than that.
Q. Do you know the process that he used to pull all of the entries that had reason codes CDFRE or CDFLSH?
A. Yes. I know it's a very tedious process.
Q. And can you describe the process to me?
A. So we're able to go into E Power Center and set up a filter with columns based on what type of information we're requesting. So we would request the two reason codes that we've shared here, and then we have to create more columns for the dates received. So whatever dates the reports are requested from.
Q. Is that, the date —
MR. OSTERMAN: Were you done with your answer?
THE WITNESS: No.
MR. STANNER: That's okay.
BY MR. STANNER: The date that you just talked about, that date is in the system, right?
Vaughan Dep. Tr. at 32:23-33:19.
[Defendant's Argument] The premise of Plaintiff's sanction motion is that Ms. Vaughan and BBW could have been more efficient. See Dkt. No. 57 at 5 ¶ 11 ("There was no need to manually search the narrative sections."). But Plaintiff's contention completely ignores Ms. Vaughan's testimony:
So part of the reason it could also take longer for this is if a customer emails us regarding a eucalyptus spearmint candle, lotion, doesn't matter what, and we don't speak to that individual directly but we just have their complaint, we use a reason code that would not categorize under the candle reason codes. And they don't contact us after that. That could produce a whole list of uncategorized reason codes. So that's part of it, too. So I want to be clear. While we search by those two candle reason codes, we could have it where they've emailed us in, and that would go under our basic reason code. Until we actually speak to them and update to candle, it could fall under that category as well. So that's more of time, too. It's very time consuming to run a filter on all of those things and weed out the candles. That's another reason we had to go that route which is more time consuming. So I want to make sure that I'm sharing that, because it's a big part of what we did.
Vaughan Dep. Tr. at 87:23-88:21.
Ms. Vaughan further testified that the process was time consuming and tedious because she also needed to search the case notes to eliminate materials that were not relevant:
Q. You described at the beginning of Paragraph 6 [of your affidavit], you say, unfortunately, the database does not allow for Boolean searches to be conducted, and therefore filters were applied for the product SKU number. That's a true statement?
A. Yes.
Q. Okay. You then go on to describe how, because of the limitations in the system, you were required to review the case text and narrative form for each of the entries that came back on your initial search, correct?
A. Yes.
Q. And what things were you looking for in the case text in order to provide information responsive to what you were being asked to provide?
A. Looking for key words. We wanted to make sure we were identifying 3-wick candles. So if we did a search by the two reason codes, that could display case texts for all candles, votives, single-wicks, two-wicks. So we had to search that case text to eliminate anything that would not be relevant to the report that was being requested.
Q. Okay. And so you were reviewing that to make sure — you said you applied, the first filter you applied, was the SKU for eucalyptus spearmint, because that's what you understood was the product at issue in the case, right?
A. Yes.
Q. And [Plaintiff's counsel] is making the point that you could have applied as a first filter a candle fire code, and that would have brought back just candles and not all of the other products that eucalyptus spearmint is in. You understand that's the point he's making?
A. Yes, I understand that. In looking back, I could have done that. But at the end of the day we still had to look through the case text to get a true reflection of what we were trying to bring forward.
Q. Okay. And were you trying to deceive anybody with the affidavit that was submitted?
A. No.
* * *
Q. So no matter which filter you applied first, you would have still had to manually review the case text in order to satisfy yourself that you were being inclusive and responsive to the request, correct?
A. Yes. It is a Court order. We want to make sure that we're providing all accurate information.
Q. And you were — in many instances the SKU is not known. So you may know it's a candle related claim.
A. Mm-hmm.
Q. But you have no idea if it was related to eucalyptus spearmint or if it was a 3-wick candle, correct?
A. Correct.
Q. So the only way to determine if it was eucalyptus spearmint and a 3-wick candle, if there's no SKU number in the database, was for you to do a manual review of the case text?
A. Correct.
Q. And that's what you were trying to describe in the affidavit as being time consuming, and I think you used the word today tedious?
A. Yes, it is tedious. And to even go more up on that is if one of our associates doesn't state it it's a 3-wick. So let's say Mary says 3-wick and Amy next to her says candle, it's going to display different results, because I'm searching by 3-wick and I have to search by a regular candle, and then it could be a votive, single-wick, and then we have to go in and determine.
Q. Okay. And you estimate that you've given us today is that that took you between 10 and 20 hours to comply with the original effort of the search limited to eucalyptus spearmint, correct?
A. Yes.
Q. And [the head of IT], you estimate, spent a similar amount of time, another 10 to 20 hours?
A. So myself and [the head of IT] together spent 10 to 20 hours
ECF No. 59 at 3-7
Ultimately, a nearly 600 page spread sheet (computer generated in the main) of 1,280 candle flashover incidents was given to plaintiff's counsel. The court is unsure whether this represents the number of 3 wick candle incidents or whether there are other one or two wick incidents as well.
Discussion
A. The Proper Basis for Sanctions
Plaintiff asserts several bases for the imposition of sanctions: Rule 11 (false certification in pleadings), the inherent power of the court to punish false statements, and ultimately, Fed. R. Civ. P 37 (without differentiation). ECF No. 57 at 1. Although plaintiff later discussed Rule 37(b) as opposed to (a), the thrust of the motion included all of Rule 37. In the reply brief, ECF No. 63 at 6, plaintiff made this clear.
However, by the very words of Fed. R. Civ. P. 11, this rule does not apply to discovery disputes. Rule 11(d).
The court need not endeavor to determine whether it should utilize its inherent authority to rectify a false statement before the court, as the "falsity" did not stem from a perjurious statement, but rather occurred as a result of an evasive or incomplete disclosure.
Rule 37(a) (4) expressly lists the evasive or incomplete disclosure as a basis for a motion to compel further discovery, i.e., such a response is treated as a "failure to disclose, answer or respond." Rule 37(a)(5) requires the imposition of sanctions unless the "failure to disclose etc." was substantially justified or the imposition of sanctions would be otherwise unjust. Subsection (C) of Rule 37(a)(5) allows a court to "apportion" sanctions if the entire discovery response was partially not substantially unjustified. While some cases have determined that the "losing party" has the burden of proving substantial justification, Easley v. U.S. Home Corp, 2012 WL 3245526 (D.Nev. 2012), the better formulation in this case where each party was partially successful on the issue upon which sanctions are sought is that the party against whom sanctions are sought has the burden to demonstrate substantial justification. Zig Zag Holdings LLC v. Hubbard, 2014 WL 3724800 *1 (N.D. Cal. 2014), citing Hyde and Drath v. Baker, 24 F.3d 1162, 1171 (9th Cir. 1994). This is especially true in situations where the basis for the sanctions motion is a failure to disclose.
B. Whether Sanctions Are Substantially Justified
Defendant, at the very least, and despite the request made for "all candle" flashover incidents, purposefully shot itself in the foot when it maintained its initial position that it would not produce any information except scattered incidents of public record, or otherwise information about only three wick Eucalyptus Spearmint candles. It then argued that producing information about even the very candle at issue, the three wick Eucalyptus Spearmint candles, was overly burdensome. As it turned out, and as demonstrated by the Vaughn declaration testimony, it was logistically more burdensome to maintain this position than it would have been simply to turn over the computer generated information of candle flashover incidents—something plaintiff had initially requested. Ultimate production of the flashover spreadsheet objectively demonstrates this fact. In essence, defendant created its own logistical burden by fashioning a self-created, manual search burden as a reason to preclude discovery into seemingly a substantial number of flashover incidents.2
As the undersigned discussed at hearing on the sanctions motion, he felt misled on the arguments defendant was making at the initial motion to compel with respect to the discovery related to flashover incidents. Combined with the "millions of candles distributed" testimony of Steven Smith, there was at the initial motion to compel process, a nearly complete emphasis on the need for a very burdensome manual search; there was no discussion of the computer generated reason codes which could have been implemented to greatly reduce this proffered burden. The conversation about having to manually winnow candles from other types of fragrance products, and even then manually determine whether the incident was one responsive to the flashover requests, and then manually limiting such search to three wick Eucalyptus Spearmint three wick candles was at best an obfuscation. It appears to have been born of a steadfast desire to keep from view in this litigation the number of flashover incidents in general.3
The undersigned considered and relied upon defendant's arguments on the initial motion to compel when it stated that defendant had "won the day" (at least in significant part), and limited discovery to three wick candles. If the information about reason codes had been imparted during the initial motion to compel proceedings, the undersigned might well not have made even that limitation with respect to the discovery requests for flashover incidents.4
Therefore, with respect to the important discovery about other candle flashover incidents, the undersigned does not find that defendant's positions in the discovery meet and confer, nor in its arguments in papers or at hearing, to be substantially justified.
However, given that defendant was substantially justified in resisting discovery requesting information in general, e.g., designs of all candles, or manufacturing processes for all candles, something which could not have been reduced to a simple "reason code" computer search, plaintiff should not recover sanctions for the entirety of the discovery motion process. The undersigned finds that a rough cut of 50% is equitable in apportioning the flashover incident discovery (certainly an important part of the discovery) which was impeded by defendant's conduct.
Conclusion
The undersigned finds that defendant's responses involving flashover incidents, as set forth previously, were not substantially justified.
Plaintiff shall, within ten days of the filed date of this order, submit a declaration detailing the time and expenses, attorney fees' hourly rate—all the items which would be considered in a lodestar analysis. The court will reduce any final figure by 50%.
Defendant may oppose the amounts sought within ten days of the filing of the declaration by plaintiff; plaintiff may file a short reply, if desired, within five days of the filing of the opposition.
IT IS SO ORDERED.
