GERARD E. LYNCH, Circuit Judge:
In the course of defending against claims that it sold counterfeit products, defendant-appellant ePRO E-Commerce Limited (“ePRO”) engaged in persistent discovery misconduct: it failed to timely disclose the majority of the responsive documents in its possession, restricted a discovery vendor’s access, to its electronic data, and failed to impose an adequate litigation hold even after the court directed it to do so, which omission allowed custodians of relevant electronic data to delete thousands of documents and significant quantities of data, sometimes permanently. As a result, the United States District Court for the Southern District of New York (Vernon S. Broderick, /,) concluded that ePRO had willfully engaged in spoliation. It accordingly granted in substantial part plaintiff-appellee Klipsch Group, Inc.’s (“Klipsch”) motion for discovery sanctions, including a $2.7 million monetary sanction to compensate Klipsch for its corrective discovery efforts and a corresponding asset restraint in that amount, permissive and mandatory jury instructions, and an additional $2.3 million bond to preserve Klipsch’s ability to recover damages and fees at the end of the case. éPRO now brings this interlocutory appeal, raising various challenges to the evidentiary rulings and factual findings undergirding those sanctions. It also contends that the resulting sanctions are impermissibly punitive, primarily because they are disproportionate to the likely’value of the case. Klipsch, on cross-appeal,' argues that the district court erred by failing to infer that ePRO destroyed relevant sales data from the fact that it failed • to retain- backup copies of its live sales database."
We find no error in the district court’s factual findings, and wé conclude that the monetary sanctions it awarded properly compensated Klipsch for the’’ corrective discovery efforts it undertook with court permission in response to ePRO’s misconduct. In particular, we emphasize that discovery sanctions should be commensurate with the costs unnecessarily created by the sanctionable behavior. A monetary sanction in the amount of the cost of discovery efforts that appeared to be reasonable to undertake
ex ante
does not become- imper-missibly punitive simply because those efforts did not ultimately uncover more significant spoliation and fraud, or increase the likely damages in the underlying case. The district court’s orders imposing sanc
tions are accordingly AFFIRMED in all respects.
BACKGROUND
The facts presented below are drawn principally from the district court’s findings of fact.
I. Initial Steps in the Litigation
In August 2012, Klipsch, a manufacturer of sound equipment including headphones, sued • DealExtreme.com, a subsidiary of ePRO, a Chinese corporation, alleging that it was selling counterfeit' Klipsch headphones. ePRO does not dispute that some infringing sales occurred; however, throughout the proceedings, the parties have insisted on vastly different estimates of the extent of such sales. Klipsch alleged that ePRO sold at least $5 million in counterfeit. or otherwise infringing Klipsch products. ePRO, on the other hand, has consistently presented evidence that the sales of the relevant products amounted to less than $8,000 worldwide. The district court initially found ePRO’s evidence persuasive and did not substantially revise its view of the case’s value even after Klipsch’s investigators uncovered two other counterfeit Klipsch products being sold on ePRO’s sites a few months into the litigation.
As the case proceeded, however, ePRO’s failure to comply with its discovery obligations began to cast doubt on the reliability of its representations. By March 2013, as Klipsch was preparing to take depositions of ePRO’s employees in Hong Kong, ePRO had produced fewer than 500 documents. ePRO insisted that it did not possess any original sales documents, and instead turned over spreadsheets created specifically for this litigation that purported to list all relevant sales. In support of its contention that those sales records were complete, ePRO points out that each of the undercover purchases made by Klipsch’s investigators was accounted for therein, even though ePRO would have had no
a priori
means of identifying those transactions.
During a deposition of ePRO’s CEO, Daniel Chow, it became clear that ePRO had not placed a litigation hold on a substantial portion of its electronic data, including any emails or faxes. Around the same time, ePRO also admitted that it did possess transactional sales documents that should have been disclosed. In order to remedy those problems, ePRO agreed to retain a discovery vendor, FTI Consulting, to conduct a keyword search of its electronic documents. FTI’s search resulted in the production of an additional 40,000 documents, including 1,236 original sales documents. Some of the documents that were produced contradicted Chow’s testimony, suggesting that ePRO had misidentified the suppliers of the counterfeit products. And although the new production was voluminous, there were also indications that ePRO had artificially limited FTI’s investigation into its electronic records. In light of the new discovery, the magistrate judge supervising discovery (Michael H. Doling-er,
M.J.)
deemed it necessary for Klipsch to re-depose ePRO’s employees.
At Chow’s second deposition in November 2013, it became clear that, despite the magistrate judge’s clear directive, ePRO still had not imposed an adequate litigation hold. It is not contested that ePRO’s counsel had advised compliance and warned that noncompliance would bring consequences.
II. The Motion for Discovery Sanctions
In December 2013, Klipsch moved for discovery sanctions, asserting that as a result of ePRO’s failure to initiate a proper litigation hold or to promptly disclose documents, large quantities of relevant documents that would have reflected a larger volume of infringing sales had been lost. The magistrate judge agreed in large'part, observing that ePRO’s “trail of false and misleading representations” throughout the discovery process had created “evident uncertainty about the plausibility, as well as the accuracy, of the defendants’ current factual assertions about the scope of infringing sales and resultant profits.” Joint App. at 1184. But although the magistrate judge determined that ePRO had “substantially failed to meet [its] obligation to preserve, search for and timely produce documents,”
id.
at 1182, he declined to impose sanctions against ePRO at that time. Instead, he authorized Klipsch to undertake an independent forensic examination of ePRO’s computer systems. The magistrate judge instructed that Klipsch would pay for the examination in the first instance, and could apply for the apportionment or reallocation of those costs after obtaining the results.
Klipsch hired iDiscovery Services (“iDS”), led by Daniel Regard, to conduct its investigation. Regard’s initial report determined that ePRO employees who were custodians of responsive information had deleted files, emails, and other potentially relevant data. With respect to the actual sales data and documents, which the parties refer to as “Structured ESI,”
Regard concluded that ePRO’s live sales databases showed evidence of editing and omissions and therefore offered unreliable evidence of historical sales data. Because ePRO had not preserved the backup copies of its databases that it routinely generated for disaster recovery purposes, Regard concluded that no reliable record of historical data was available.
With respect to the information stored on employees’ computers and email accounts, which the parties refer to as “Unstructured ESI,” Regard found that the custodians had engaged in various forms of spoliation, including manually deleting thousands of files and emails, using data-wiping software shortly before iDS’s forensic examination was scheduled to begin, and updating their operating systems during the litigation period, which had the result of clearing out data regarding their program usage. Regard and his team were also denied access to several custodians’ email and private messenger accounts, despite ePRO’s admission that those accounts were sometimes used for business purposes.
After the search was completed,1 Klipsch filed an
ex parte
motion along with Regard’s initial report, asking the court to increase the hold on ePRO’s assets and enter a default judgment in Klipsch’s favor. Klipsch’s briefing suggested that Regard’s report showed evidence.that ePRO had engaged in massive amounts of data-deletion and tampering. The district court increased the hold on ePRO’s assets from $20,000 to $5 million and ordered ePRO to show cause why a default judgment should not be entered against it.
In opposition, ePRO submitted expert reports from Michael Jelen, an expert in Structured ESI forensics, and Erik Ham-merquist, an expert in Unstructured ESI. Klipsch then submitted an additional declaration from Regard that modified some
of his conclusions in response to ePRO’s experts’ critiques. The district court held a four-day evidentiary hearing- on Klipsch’s motion in January 2015, following which the parties submitted proposed findings of fact and conclusions of law as well as further declarations from Regard and Ham-merquist. The district court considered the final Regard report but excluded Hammer-quist’s proposed sur-rebuttal. ,
III. The District Court’s Sanctions Rulings
In " November 2015, the district court issued an order granting in part and denying in part Klipsch’s motion. In summary, it found that Klipsch failed to demonstrate that ePRO had destroyed or tampered with, its sales documents. The court accordingly denied Klipsch’s motion for default judgment and reduced the asset restraining order to $25,000, reflecting its view of the likely valuation of actual damages in the case.
The court concluded that Klipsch had shown that ePRO had willfully spoliated relevant Unstructured ESI. Although many' of the improperly deleted documents were recovered and no smoking gun was found among them, the district court emphasized that spoliation had resulted in some permanently unrecoverable files and data, and inferred from ePRO’s willful misconduct that the missing documents were relevant and that their absence prejudiced Klipsch1. As a result, it imposed the following sanctions: (1) an instruction requiring the jury to find that ePRO had destroyed relevant Unstructured ESI after, its duty to preserve those documents had been triggered; (2) an instruction permitting the jury to presume that the destroyed evi-' dence would, have- been favorable to Klipsch; and (3) reasonable costs and fees associated with the discovery motion, beginning with Klipsch’s second round of depositions in Hong Kong.
Both parties moved for reconsideration. Other than a small revision to one aspect of its Unstructured ESI findings, however, the district court’s ruling remained unchanged. After a careful examination of Klipseh’s fee submissions, the court awarded Klipsch a total of $2.68 million as compensation for the additional discovery ■ efforts occasioned by ePRO’s misconduct.
The court rejected ePRO’s arguments that the amount was excessive either as compared to the small amount of actual compensatory-damages likely at issue in the case, or as a result of Klipsch’s failure to uncover evidence of any additional infringing sales or spoliation of sales records.
The court deemed ePRO to be a dissipation risk in light of its persistent failure to comply with court orders or discovery protocols. It accordingly imposed a restraint on ePRO’s assets in the amount of the monetary. sanctions. It also imposed an additional $2.3 million restraint, which amount it determined would be appropriate to secure Klipsch’s likely recovery of treble damages and attorney’s fees at the conclusion of the case. At ePRO’s request, the court permitted the parties to determine whether the resulting $5 million in total restraints would be held in a restraint, a separate bond, or some combination thereof.
The parties now appeal those rulings. ■
DISCUSSION
As a preliminary matter, we note that the restraint on ePRO’s assets is a form of injunctive relief that provides the basis for our jurisdiction to hear this inter
locutory appeal.
28 U.S.C. § 1292
(a)(1). Because the remainder of the parties’ arguments concern matters that are inextricably bound up with the validity of that injunction, we also have jurisdiction to resolve those disputes.
See Amador v. Andrews,
655 F.3d 89
, 95 (2d Cir. 2011).
We review each of the issues raised in these cross-appeals for abuse of discretion.
See West v. Goodyear Tire & Rubber Co.,
167 F.3d 776
, 779 (2d Cir. 1999) (discovery sanctions);
United States v. Fabian,
312 F.3d 550
, 557 (2d Cir. 2002), abrogated on other grounds by
United States v. Parkes,
497 F.3d 220
(2d Cir. 2007) (evidentiary determinations);
EEOC v. KarenKim, Inc.,
698 F.3d 92
, 99 (2d Cir. 2012) (grant of injunctive relief). “A district court abuses its discretion if it (1) bases its decision on an error of law or uses the wrong legal standard; (2) bases its decision on a clearly erroneous factual finding; or (3) reaches a conclusion that, though not necessarily the product of a legal error or a clearly erroneous factual finding, cannot be located within the range of permissible decisions.”
KarenKim,
698 F.3d at 99-100
(internal quotation marks omitted).
■ We determine that the district court did not err in its careful factual findings, nor did it base any aspect of its decision on an incorrect legal standard. The primary engine of ePRO’s appeal appears to be the fact that the district court imposed $2.7 million in sanctions in a case that it agreed will likely result in about $20,000 in damages. We hold that the district court’s award properly reflects the additional costs ePRO imposed on its opponent by refusing to comply with its discovery obligations. Because we conclude that the remainder of the parties’ arguments are also without merit, we affirm the district court’s rulings.
I. Evidentiary and Factual Arguments
The parties raise an assortment of evi-dentiary and factual challenges to the district court’s rulings. We assume the parties’ familiarity with the issues and we commend the district court for its cogent explanation of the highly technical factual details underpinning these debates.
A. Evidence Considered by the District Court
ePRO argues that the district court abused its discretion by admitting Regard’s supplemental reports and declarations, and by relying on any of his reports for the truth of the matter asserted therein. None of those evidentiary challenges has merit.
First, ePRO argues that some of Klipsch’s evidence should not have .been considered because it was. not timely submitted. The materials ePRO contends should have been excluded, however, were provided to ePRO before the January 2015 hearing. Although it is unclear from the record before us whether ePRO has adequately preserved its objection to the admission of that evidence, the district court was aware of ePRO’s concerns about the timing of the disclosures, and,. indeed, granted ePRO’s motion to strike (other) evidence that was submitted only days before the hearing. We see no abuse of discretion in-those rulings.
Next, ePRO argues that the district court abused its discretion when it considered Klipsch’s post-hearing supplemental expert report, but rejected ePRO’s response. It is not an abuse of discretion to give one party the last word, particularly after the court has already permitted the parties to exchange multiple expert reports and to cross-examine witnesses during a multi-day evidentiary hearing. In any event, striking the portions of the district
court’s factual findings that relied solely on Regard’s final report would have no impact on the validity of its overall ruling.
Finally, ePRO argues that Regard’s reports should not have been relied on for the truth of the matters asserted therein because the district court allegedly sustained ePRO’s objection to the admission of his initial report for that purpose. But ePRO did not object to the acceptance of Regard as an expert, and his reports simply provided the factual details supporting his expert opinions. Thus, we conclude that challenge, too, lacks merit.
B. Spoliation of Unstructured ESI
ePRO argues that the district court erred when it found that ePRO had willfully spoliated Unstructured ESI. The district court found evidence of five methods of spoliation: (1) 4,596 responsive files or emails were manually deleted, although all of these documents were eventually recovered;
(2) seven employees used data-wiping programs shortly before the forensic examination began, resulting in 31 permanently unrecoverable files; (3) eighteen employees ran operating systems upgrades during the litigation hold period, which resulted in the loss of their program usage data; (4) ePRO failed to provide access to the email accounts of seven employees who had worked on its email direct marketing homepage but were no longer employed with the company, and deleted or failed to provide access to the email accounts of twelve other custodians of discoverable data; and (5) 32 out of 36 identified custodians, including ePRO’s CFO and the employee coordinating Regard’s access to ePRO’s data for his examination, refused to permit access to their accounts on a private messaging system, which, although primarily used for personal communication, were listed on their corporate email signatures.
The party seeking discovery sanctions on the basis of spoliation must show by a preponderance of the evidence: “(1) that the party having control over the evidence had an obligation to preserve it at the time it was destroyed; (2) that the records were destroyed with a culpable state of mind; and (3) that the destroyed evidence was relevant to the party’s claim or defense such that a reasonable trier of fact could find that it would support that claim or defense.”
Chin v. Port Auth. of N.Y. & N.J.,
685 F.3d 135
, 162 (2d Cir. 2012) (internal quotation marks omitted).
The district court did not clearly err in determining that each of those elements had been met here. ePRO’s arguments to the contrary with respect to the manually deleted emails and the operating system reinstallations fail to recognize that the district court did not find that each instance of spoliation was equally probative of willfulness, nor that every method resulted in the permanent destruction of presumptively relevant data. Instead, the district court inferred relevance and prejudice only from the unrecoverable data, and viewed the various means of deleting data as cumulative proof that ePRO’s spoliation was willful.
ePRO also spends significant effort trying to undermine the district court’s factu
al findings regarding exactly which employee used which data-wiping program at which time. But ePRO does not contest that the district court properly made the following findings: multiple custodians of discoverable data had data-wiping software on their devices during the litigation hold period; at least some of those programs were last used at a time very close to the initiation of Klipsch’s forensic examination; and the use of those programs resulted in the permanent deletion of 31 files. Those facts are sufficient to support the district court’s finding of willful spoliation, and ePRO’s expert’s opinion that some of the specific data had been commingled and therefore could not reliably be attributed to any particular individual does not change that conclusion. Moreover, ePRO’s evidence on commingling was contested and to some extent incomplete because ePRO’s expert testified that commingling was only one possible explanation for the duplicative data. Accordingly, the district court did not clearly err when it declined to adopt ePRO’s view of the evidence.
Finally, ePRO contends that its inability to provide access to certain email and messaging accounts should not have been construed as evidence of willful spoliation because those accounts were primarily for private use. But the district court determined that ePRO was required to provide access to all of those accounts because they were also sanctioned for business use. As the district court noted, it was undisputed that ePRO’s employees included their private messaging addresses in their work email signatures, installed the messaging software on their work computers, and sometimes used the specified private email accounts for business reasons. That ePRO did not have a software usage policy in place requiring its employees to segregate personal and business accounts or to otherwise ensure that professional communications sent through personal accounts could be preserved by the company for litigation purposes was the company’s own error. ePRO cannot use that oversight as an excuse to avoid discovery, nor can it complain because the resulting and wholly foreseeable deletion of material that could well have contained relevant evidence gave rise to sanctions.
Thus, we find no clear error in the district court’s factual findings regarding the spoliation of Unstructured ESI.
C. Lack of Spoliation of Structured ESI
In what has been styled a cross-appeal but might be better understood as advocating an additional or alternative ground on which to affirm the sanctions, Klipsch argues that the district court erred when it declined to find that ePRO had willfully spoliated direct evidence of sales, which the parties refer to as “Structured ESI.” Klipsch concedes that its examination did not uncover any direct evidence that ePRO’s sales data had actually been tampered with, nor that ePRO was otherwise concealing a greater volume of infringing sales. Instead, Klipsch argues that ePRO’s
failure to preserve static backup copies of its sales databases deprived Klipsch of. a crucial means of verifying the historical accuracy of its live sales data.
ePRO maintains a' live sales database that extracts and cross-references data from separate modules tracking inventory, orders, and shipments. Regard observed that the sales records in the live database were consistent with what had previously been produced, but because he found evh dencé that the database had been subject to editing," he nevertheless concluded that the sales records were unreliable. Jelen, ePRO’s Structured ESI expert, agreed that it was possible to edit the live database and noted that some such edits would be expected in the normal course of business; however, he opined that it would be extremely difficult to completely remove evidence of a transaction from all three modules, and noted there was no evidence of such an effort nor of any inconsistency across the modules.
In addition to the live database, ePRO’s policies also require it to generate backup copies of the sales database for disaster recovery purposes on a regular basis as well as every time the system undergoes a specified “major event.” Thpse copies presumably would not have been subject to, editing in the normal course, and therefore could have been used as references against which to check the accuracy of the live data. Klipsch asserts that ePRO had an obligation to preserve the backup copies of its sales databases as unique sources of historical sales data,
But even where a party has shown that its opponent had an obligation to preserve certain evidence and willfully failed to do so, sanctions, need not automatically follow; instead, we have simply held that such conduct
may
provide sufficient circumstantial evidence that relevant and prejudicial data was lost.
See, e.g., Fujitsu Ltd. v. Fed. Exp. Corp.,
247 F.3d 423
, 436 (2d Cir. 2001) (“Once a court has- concluded that a party was under an obligation to preserve the evidence that it destroyed, it must then consider whether the evidence was intentionally destroyed, and the likely contents of that evidence. The determination of an appropriate sanction for spoliation, if any, is confined to the discretion of the trial judge[,]”) (internal citations omitted). That' is to say, the district court retains its1 discretion to impose discovery sanctions—or not—so long as that determination is not based on a legal or factual error.
See
id.
We conclude that the district court did not abuse its discretion here when it chose not to infer that the lost backups would have included information that was relevant or that its unavailability prejudiced Klipsch. Jelen’s testimony provided adequate support for a determination that' the otherwise undetéctable destruction of sales data was unlikely, and, as the district court pointed out, even Klipsch’s own expert was unwilling to affirmatively endorse the theory that such conduct had actually occurred.
In sum, we find no clear error with respect to the district court’s factual findings, nor any, abuse of discretion with respect to its evidentiary rulings.
II. Amount of Monetary Sanction
ePRO argues that the monetary sanctions imposed against it are so out of proportion to the value of the evidence uncovered by Klipsch’s efforts or to the likely ultimate value of the case as to be impermissibly punitive and a violation of due process. That position, although superficially sympathetic given the amount of the sanction, overlooks the fact that ePRO caused Klipsch to accrue those costs by failing to comply with its discovery obligations. Such compliance is not optional or negotiable; rather, the integrity of our civil
litigation process requires that the parties before us, although adversarial to one another, carry out their duties to maintain and disclose the relevant information in their possession in good faith.
The extremely broad discovery permitted by the Federal Rules depends on the parties’ voluntary participation. The system functions because, in the vast majority of cases, we can rely on each side to preserve evidence and to disclose relevant information when asked (and sometimes even before then) without being forced to proceed at the point of a court order.
See Cine Forty-Second St. Theatre Corp. v. Allied Artists Pictures Corp.,
602 F.2d 1062
, 1068 (2d Cir. 1979) (observing that “embroil[ing] trial judges in day-to-day supervision of discovery” is “a result directly contrary to the overall scheme of the federal discovery rules”). The courts are ill-equipped to address parties that do not voluntarily comply: we do not have our own investigatory powers, and even, if we did, the spoliation of evidence would frequently be extremely difficult for any outsider to detect.
Moreover, noncompliance vastly increases the cost of litigation by drawing out deadlines and necessitating motion practice. But “[a]n undertaking on the scale of the large contemporary suit brooks none of the dilation, posturing, and harassment once expected in litigation.”
Id.
at 1067-68
. Accordingly, we have held that discovery sanctions are proper even against parties who have belatedly complied with their obligations, because an alternative rule “would encourage dilatory tactics, and compliance with discovery orders would come only when the backs of counsel and the litigants were against the wall.”
Id.
at 1068
;
see also S. New Eng. Tel. Co. v. Glob, NAPs Inc.,
624 F.3d 123
, 149 (2d Cir. 2010). When, as a result of an opponent’s persistently uncooperative behavior, it appears reasonable
ex ante
to conduct expensive corrective discovery efforts, we see no reason why the party required to undertake those efforts should not be compensated simply because it eventually turned out that the obstructive conduct had hidden nothing of real value to the case. Those costs must be placed on the uncooperative opponent in order to deter recalcitrant parties from the cavalier destruction or concealment of materials that the law requires them to retain and disclose.
When we apply those principles to the case at hand, it is clear that the district court did not abuse its discretion by imposing monetary sanctions calculated to make Klipsch whole for the extra cost and efforts, it reasonably undertook in response to ePRO’s recalcitrance. Even accepting the district court’s conclusion that Klipsch’s forensic review failed to uncover proof that evidence of additional infringing sales had been destroyed, neither Klipsch nor the magistrate judge was in any position to know that such would prove to be the case before Klipsch undertook its rer view. On the contrary, ePRO’s persistent refusal to comply with the discovery process provided ample grounds to suspect that the degree of its obstructive conduct would be commensurate with the value of the evidence it was hiding from its adversary and the court.
ePRO’s principal objections to that conclusion fall into two buckets: first, it invokes Rule 37 of the Federal Rules of Civil Procedure to suggest that the district court has rewarded Klipsch for engaging in unnecessary and excessive discovery; and second, it argues that the sanctions are disproportionate either to the degree of success Klipsch achieved on its motion or to the ultimate value of the case. We find both arguments unconvincing.
A. Rule 37’s Limitations on Unnecessary Discovery Sanctions
¿PRO challenges the fee amount as failing to adhere to the principles provided by Rule 37(e) of the Federal Rules of Civil Procedure, which governs when a court may impose sanctions for the spoliation of electronically stored information. The district court did not, in fact, rely on Rule 37 as the grounds for the monetary sanctions; instead, the sanctions were imposed under the court’s inherent power to manage its own affairs.
Nevertheless, ePRO argues that the monetary sanctions were inappropriate under either source of authority. It argues that the monetary sanctions imposed here improperly reward Klipsch for excessive discovery efforts, suggesting Klipsch was not justified in taking such expensive measures to obtain the documents' to which it was entitled because it should have known that the ultimate recovery on the merits would remain small in any event. Alternatively, ePRO argues that Klipsch should not be compensated for these efforts because it would have undertaken them regardless of ePRO’s conduct. Neither of those arguments finds support in the record before us.
First, we note that there is no special rule requiring parties to suffer an opponent’s open and notorious discovery misconduct in small value cases. Moreover, it is important to recognize the context in which Klipsch undertook its forensic examination, of ePRO’s files. At that point, the magistrate judge supervising discovery observed that ePRO’s evasive conduct had created “evident uncertainty” about the accuracy of its representations regarding the likely value of the case. Joint App. at 1184. Klipsch had every reason to suspect that the documents its search would uncover were likely to be even more damaging than the ones ePRO had already grudgingly disclosed, which suggested that ePRO had misled Klipsch and the court about the provenance of the infringing products and the nature of its business.
Second, the history of the case makes clear that the sanctions and fees awarded in this case were carefully limited to costs Klipsch incurred in direct response to ePRO’s misconduct. Klipsch obtained approval from the magistrate judge prior to each of its substantive efforts, and in each case, that approval was given only after ePRO had already squandered an opportunity to correct its own errors. For example, ePRO’s failure to implement a litigation hold was first discovered in March 2013, during Klipsch’s first round of depositions with ePRO employees, but ePRO was not sanctioned at that time, nor was Klipsch given carte blanche to explore ePRO’s files. Instead, ePRO was permitted to hire its own discovery expert to correct the error, which resulted in the production of substantial additional discovery. Klipsch then spent approximately $550,000 on a second round of depositions occasioned by that late production. It is evident that the district court did not detect any abusive conduct on the part of Klipsch, such as the piling on of discovery demands and investigatory initiatives in order to burden its
adversary with wasteful expenses, motions practice, and sanctions. ePRO does not appear to contest the reasonableness of permitting Klipseh to take those remedial depositions, nor can it plausibly assert that Klipseh would have insisted on doing so even if ePRO’s initial production had béen complete or timely. And only in March 2014, after ePRO had repeatedly shown itself to be an untrustworthy participant in the discovery process, did the magistrate judge determine that Klipseh was “fully justified” in seeking to undertake an independent forensic examination. Joint App. at 1187.
Because the costs for which Klipseh is being compensated were reasonably incurred in direct response to ePRO’s misconduct, we cannot conclude that the district court abused its discretion by requiring ePRO to pay monetary sanctions in that amount.
B. Proportionality
'
Even assuming that Klipseh reasonably undertook these efforts, ePRO contends that the monetary sanctions should nevertheless be reduced on the basis of a principle of proportionality putatively borrowed from fee-shifting cases in the civil rights context.
ePRO first argues that Klipseh should not be completely compensated for its efforts because it did not achieve complete success on its motion. The district court did, however, grant Klipseh substantial relief: it upheld Klipsch’s contention that ePRO willfully failed to fulfill its discovery obligations after several court warnings, and it determined that ePRO’s spoliation of its Unstructured ESI was sufficiently serious to warrant adverse jury instructions.
See Metrokane, Inc. v. Built NY,
Inc., No. 06 CIV. 14447(LAK)(MHD),
2009 WL 637111
, at *4 n.3 (S.D.N.Y. Mar. 6, 2009) (determining a discovery sanctions motion to be substantially successful where “the court upheld the movant’s contention that Metrokane had failed to fulfill its discovery obligations in the very manner suggested by Built N.Y. and awarded a full range of appropriate relief’). That the district court chose not to view Klipsch’s failure to demonstrate willful spoliation of Structured ESI as a lack of success sufficient to justify a proportional reduction of the monetary sanctions was a decision soundly within its discretion.
ePRO also takes various tacks to assert that the sanction should be reduced in light of the small amount of money likely at issue on the merits. But it has not identified any authority limiting a district court’s discretion to award a compensatory discovery sanction on the basis of the ultimate damages award. Instead, because discovery sanctions are typically decided independently from the ultimate outcome of the case, it appears that courts routinely award such sanctions without any discussion of the ultimate merits recovery.
See, e.g., E. I. DuPont de Nemours & Co. v.
Kolon Indus., Inc.,
No. 3:09-CV-058,
2013 WL 458532
, at *3 & n.2 (E.D. Va. Feb. 6, 2013) (observing that there was no need, to consider the amount in controversy as a factor in setting a proper discovery sanction);
Qualcomm Inc. v. Broadcom Corp.,
No. 05-CV-1958-B (BLM),
2008 WL 66932
, at *17 (S.D. Cal. Jan. 7, 2008), vacated in part on other grounds, No. 05-CV-1958-RMB (BLM),
2008 WL 638108
(S.D. Cal. Mar. 5, 2008) (making no reference to the ultimate recovery in the case). And in the present, circumstance, the discovery motion at issue was decided before the case has concluded; accordingly, tethering monetary sanctions to the ultimate amount in controversy would restrict the court’s discretion to a number that remains speculative and indeterminate.
It is hardly clear, in any event, that the civil rights cases provide an appropriate analogy for discovery sanctions.
Congress has provided for fee-shifting in that context in order to provide an incentive to qualified counsel to undertake representation of impecunious plaintiffs with meritorious cases.
See Perdue v. Kenny A. ex rel. Winn,
559 U.S. 542
, 552,
130 S.Ct. 1662
,
176 L.Ed.2d 494
(2010). But that provision was also intended to avoid “windfalls to attorneys” that would result'from-granting them fees- no sane, paying client would expend in pursuit of the same relief.
Id.
Accordingly, courts have sometimes held that claims seeking large fees for minuscule success should be reduced or denied.
See, e.g., Farrar v. Hobby,
506 U.S. 103
, 115,
113 S.Ct. 566
,
121 L.Ed.2d 494
(1992) (observing that a nomiñal damages award usually will not support any accompanying attorney’s fee award);
Husain v. Springer,
579 Fed.Appx. 3
, 5 (2d Cir. 2014) (reducing fee award in light of the plaintiffs’ limited success and recovery).
Discovery sanctions are different. As we have emphasized above, a party that disregards its obligations may create a reasonable suspicion that further investigation is warranted, and thereby imposes costs on its adversary that would never have been incurred had the party complied with its obligations in the first instance. In that situation, the offended adversary’s counsel is not being rewarded for its success in the litigation; rather, the adversary is simply being compensated for costs it should not have had to bear.
Moreover, to whatever extent that civil rights fee-shifting cases might provide a useful comparison, they. d° not provide unequivocal support for ePRO’s proportionality arguments. Because success in some meritorious civil rights cases will reasonably require attorney’s fees in excess of the plaintiffs ultimate recovery, we have stated in that context that '“[a] presumptively correct ‘lodestar’ figure should not be reduced simply because a plaintiff recovered a low damage award.”
Cowan v. Prudential Ins. Co. of Am.,
935 F.2d 522
, 526 (2d Cir. 1991). And we have recognized that the conduct of the opposing party
commonly justifies the award of .such “disproportionate” fees. In
Kassim v. City of Schenectady,
415 F.3d 246
(2d Cir. 2005), we explained:
[I]n litigating a matter, an attorney is in part reacting to forces beyond the attorney’s control, particularly the conduct of opposing counsel and of the court. If the attorney is compelled to defend against frivolous motions and to make motions to compel compliance with routine discovery demands, or to respond to unreasonable demands of the court for briefing or for wasteful, time-consuming court appearances, the hours required to litigate even a simple matter can expand enormously. It is therefore difficult to generalize about the appropriate size of the fee in relation to the amount in controversy.
Id.
at 252
;
see also Serin v. N. Leasing Sys., Inc.,
501 Fed.Appx. 39
, 41 (2d Cir. 2012) (same).
In sum, we see nothing in ePRO’s proportionality arguments compelling us to conclude that the district court abused its discretion by awarding full compensation for efforts that were
ex ante
a reasonable response to ePRO’s own evasive conduct.
The proportionality that -matters here is that the amount of the sanctions was plainly proportionate—indeed, it was exactly equivalent—to the costs ePRO inflicted on Klipsch in its reasonable efforts to remedy ePRO’s misconduct.
III. Amount of Asset Restraints
The district court concluded that ePRO, a foreign company with few ties to the United States, had shown itself to be a dissipation risk by repeatedly failing to comply with court orders or its own attorneys’ instructions. It therefore imposed .two mechanisms to ensure Klipsch’s full recovery: first, it restrained $2.7 million of ePRO’s assets to cover the amount of the monetary sanctions; and second, it required ePRO to post. a. bond for $2.3 million, a sum representing what the court deemed to be Klipsch’s likely recovery on the merits as well as its remaining attorney’s fees, which would; be recoverable under the Lanham Act. At ePRO’s request, however, it permitted ePRO to hold the full $5 million in either an asset restraint, a bond, or some combination thereof.
'1 ePRO now contends that the district court’s rulings are contradictory because the court determined that the infringing sales were unlikely to exceed $8,000, but nevertheless imposed a $5 million asset restraint. That argument simply ignores the district court’s explanation that the total sum is.intended to cover the monetary sanctions, and. remaining attorney’s fees in the case- in addition, to the merits award, and, as the court pointed out, there is ample precedent for .requiring the posting of a-bond in these circumstances.
See Johnson v. Kassovitz,
No.. 97 CIV. .5789 DLC,
1998 WL 655534
, at *1 (S.D.N.Y. Sept. -24, 1998) (in the context of the Lan-ham Act, observing that, security of attorney’s fees may be included in a bond for costs under S.D.N.Y. Local Rule 54.2) (collecting cases).
Nothing that we say in this opinion ■should be taken as condoning excessive and disproportionate discovery demands, countenancing the tactical use of discovery sanction motions to inflict gratuitous costs on adversaries, or derogating from the re
sponsibility of district courts to ensure •that litigation proceeds in a responsible and cost-efficient manner.
See
Fed. R. Civ. P. 1 (directing that the Rules of Civil Procedure “should be construed, administered, and employed by the court and the parties to secure the just, speedy, and
inexpensive
determination of every action”) (emphasis added); Fed. R. Civ. P. 26(b)(1) (scope of discovery should be “proportional to the needs of the case, considering
[inter
alia] the amount in controversy”). If it turns out, as the district court has estimated, that the amount of actual damages in this case is modest in relation to the costs spent on the litigation, that would be a highly regrettable outcome.
But the question before the district court, and before us, is which party should be held responsible for those costs. ePRO does not ever contend that Klipsch’s initial discovery demands were unreasonable or disproportionate to the merits- of the case. Nor does it seriously argue'that the magistrate judge erred in allowing Klipsch to take the steps it took to remedy ePRO’s refusal to comply with those demands. The district court reasonably concluded, after a full and fair hearing, that it was ePRO’s noncompliance with its legal obligations that occasioned the excessive costs in this case, and we find no reason why ePRO should not therefore be required to pay them.
CONCLUSION
We have considered ePRO’s remaining arguments and found them unpersuasive. For the reasons stated above, the judgment of the district court is AFFIRMED.