The opinion of the court was delivered by: John D. Bates United States District Judge
Presently before the Court are plaintiffs' motions for final approval of the settlement and for approval of attorney fees, reimbursement for costs and expenses, and incentive awards to the representative plaintiffs. For the reasons explained below, the Court concludes that final certification of the class and final approval of the settlement are warranted, and grants the petition for awards, fees, and costs.
I. PROCEEDINGS IN THIS CASE
Much of the procedural history and background of this case has been discussed in Trombley v. National City Bank, 759 F. Supp. 2d 20 (D.D.C. 2010), where the Court granted preliminary class certification and preliminary approval of the settlement agreement. Plaintiffs filed their class action complaint on February 17, 2010, alleging that defendant National City Bank*fn1 engaged in unlawful and deceptive practices by improperly charging its customers overdraft fees for insufficient funds on debit card transactions in violation of various state and federal laws. Compl. ¶¶ 2, 13-14. Specifically, plaintiffs allege that National City Bank "reorder[ed] electronic debit transactions from the highest dollar amount to lowest dollar amount so as to deplete the customer's available funds as quickly as possible while maximizing the number of overdraft fees collected." Compl. ¶ 2. Plaintiffs also allege that National City Bank provided false and misleading account balance information and failed properly to disclose its overdraft policies. Id.
In May 2010, the Court received notification from the U.S. Judicial Panel on Multidistrict Litigation ("MDL Panel") that an order had been entered, directing that this case be conditionally transferred to the Southern District of Florida to Multidistrict Litigation Proceeding No. 2036, In re Checking Account Overdraft Litigation ("MDL No. 2036"),which has been assigned to U.S. District Judge King. See Docket Entry No. 3. On July 28, 2010, plaintiffs and National City Bank (collectively, "the parties" or "the settling parties") entered into a settlement agreement and moved for preliminary approval in this Court. Soon thereafter, the parties informed the MDL Panel that they had reached a settlement, and the MDL Panel vacated its order to transfer the action and returned the case back to this Court to evaluate the proposed settlement. The MDL Panel's ruling did not bar the future transfer of the case to MDL No. 2036 if this Court did not approve the proposed settlement or fully resolve the case. See Order Vacating Conditional Transfer Order at 1 n.1, Docket Entry No. 14 (Aug. 9, 2010).
Robert Matos, the named plaintiff in Matos v. National City Bank, No. 10-cv-21771 (S.D. Fla. June 1, 2010), an overlapping class action that became part of MDL No. 2036, objected to the proposed settlement. Matos is represented by plaintiffs' lead counsel and the plaintiffs' executive committee ("PEC") in the multi-district litigation before Judge King. The Court held the preliminary fairness hearing on November 18, 2010 and heard arguments from the settling parties and objections from the PEC on behalf of Matos. On December 17, 2010, the Court held a brief telephone conference with counsel for plaintiffs and National City Bank to discuss several concerns with the notice and release provisions of the settlement agreement. Following that discussion, the parties filed a renewed motion for preliminary approval, a Revised Settlement Agreement, and a revised proposed order. See Docket Entry No. 34 (Dec. 22, 2010). After considering the submissions and arguments from the parties and the objectors, along with the Revised Settlement Agreement, the Court granted preliminary approval of the settlement and the notice plan, certified the settlement class, appointed Tycko & Zavareei LLP as class counsel and Ramona Trombley, Jeff Doehner, and Brian Wells as representative plaintiffs, and scheduled the final fairness hearing. See generally Order, Docket Entry No. 38 (Jan. 11, 2011); Trombley, 759 F. Supp. 2d at 30. The final fairness hearing was held on July 14, 2011, at which time the Court heard argument from the parties and from two of the objectors. After the hearing, the Court requested supplemental briefing from the parties and the objectors, along with a final status report, all of which were duly submitted.
II. TERMS OF THE SETTLEMENT
The terms of the settlement, which are laid out in the Revised Settlement Agreement, and the subsequent developments to those terms are briefly discussed below. Pursuant to the Revised Settlement Agreement, the Settlement Class was defined as follows:
All persons who hold or ever held a National City Account*fn2
who at any time during the Class Period incurred at least one Overdraft Fee*fn3
associated with at least one National City Debit Card
Transaction*fn4 that was not previously reversed,
refunded, or returned to the Settlement Class Member by
Excluded from the Settlement Class are National City Bank, any parent, subsidiary, affiliate or sister company of National City Bank, and all officers or directors of National City Bank, or any parent, subsidiary, affiliate or sister company at any time during the Class Period, and the legal representatives, heirs, successors, and assigns of any of the foregoing. The Court presiding over any motion to approve the Settlement Agreement is excluded from the Settlement Class. Also excluded from the Settlement Class is any person who timely submits a valid request to be excluded from this Settlement.
Revised Settlement Agreement ("Rev. Settlement Agreement") ¶ 7. The "Class Period" was defined as July 1, 2004 through August 15, 2010. Id. ¶ 1(d). Moreover, customers who originally opened accounts at PNC Bank or PNC Bank's predecessor banks other than National City were not included in the settlement. Id. ¶ 4.
The agreement provided for a $12,000,000 settlement fund, inclusive of all attorneys' fees, costs, expenses, and incentive payments to representative plaintiffs. National City Bank also agreed to provide an additional $500,000 for notice and claims administration costs.On March 14, 2011, the parties advised the Court that National City had agreed to increase its contribution by an additional $1 million for notice and claims administration costs, see Docket Entry Nos. 39-40, followed by an agreement by the parties that National City would pay an additional $300,000 towards notice and claims administration costs. See Ex. C to Final Status Report (Aug. 31, 2011). Any costs in excess of the $1.8 million would be deducted henceforth from the settlement fund. See Rev. Settlement Agreement ¶ 9-10. As a result, the ultimate amount to be paid by National City for the settlement of the claims is $13.8 million. As part of the settlement, National City also agreed not to object to incentive awards in the amount of $5,000 to each of the representative plaintiffs, or an award of attorneys fees of not more than 25% of the settlement fund. Id. ¶¶ 10, 12.
Pursuant to the Revised Settlement Agreement, class members would receive a maximum reimbursement of $36 for each eligible overdraft charge incurred during two calendar months during the class period of July 1, 2004 to August 15, 2010. Claims would be paid from the moneys remaining in the settlement fund after attorneys' fees, costs, and other expenses were paid. Rev. Settlement Agreement ¶ 31. If any funds remained after accounting for all the fees, costs, and claims, the leftover amount would then be distributed to claimants on a pro rata basis. After that, any amount remaining would be paid out through a cy pres distribution to a mutually-agreed upon organization subject to approval by the Court. Id. ¶ 31(b). At the final fairness hearing, class counsel stated that because of the number of claimants and the number of overdrafts per claimant, it was anticipated that there would be no cy pres fund. Final Fairness Hr'g Tr. ("Tr.") 9:7-10.
With respect to notifying class members about the settlement, National City provided the Notice Administrator, Hilsoft Notifications ("Hilsoft"), with the names and last known addresses of National City customers who had incurred an overdraft fee. Individual notices were sent out by first-class mail to potential class members. Rev. Settlement Agreement ¶¶ 15, 17-18; Affidavit of Cameron R. Azari ("Azari Aff.") ¶¶ 25-29 (May 26, 2011). In addition to individualized notices, other forms of notice were also employed. Published notice appeared once on a weekday in the highest circulation newspaper in twenty two designated market areas, which comprised 92% of National City branch locations. Rev. Settlement Agreement ¶ 19(a); Azari Aff. ¶¶ 30-33. A press release was issued to 4,490 major press outlets. Rev. Settlement Agreement ¶ 19(b); Azari Aff. ¶ 34. Electronic forms of notice were also provided, including the creation of a website where class members could obtain information about the settlement and documents such as claim forms, and file claims online. PNC's website also posted a notice targeted to former National City customers with online accounts, directing these customers to the aforementioned website. Rev. Settlement Agreement ¶ 19(c)-(d); Azari Aff. ¶¶36-37.
To receive a refund under the settlement agreement, a class member was required to submit a claim form declaring under penalty of perjury that she or he incurred at least one overdraft fee associated with that member's National City debit card transaction. On the form, the class member could claim compensation for each incurred and eligible overdraft fee during any two calendar months within the Class Period. The two months need not be consecutive. The class member could opt to have the Claim Administrator, Epiq Class Action and Claims Solutions ("Epiq"), determine the two months during the class period in which the member incurred the greatest number of overdraft fees ("Option 1") using data provided by National City. Alternatively, the class member could choose the two months ("Option 2"). Id. ¶¶ 26-27.
As of September 1, 2011, approximately 187,679 "timely claims" have
been received and verified, with a remaining number of 10,511
additional claims that are undergoing further analysis. Final Status
Report at 1-2. The majority of the claimants (183,339) selected Option
1, allowing Epiq to determine the number of eligible overdrafts. An
estimated 4,340 claimants
chose Option 2, in which claimants themselves selected the two months
for reimbursement. Approximately seventy-six requests for opt-outs
have been received. Final Status Report at 2; Ex. B. Through July 31,
2011, Hilsoft has incurred approximately $192,620.00 in fees and
expenses, see Declaration of Robert Oseas ¶ 6 (Sept. 2, 2011); Ex. A
to Final Status Report (Hilsoft Statement of Activity),*fn5
and Epiq has incurred $1,841,902 in fees and expenses
associated with the settlement. Combined, they together estimate that
total notice and claims costs will ultimately amount to $2.3 million,
with $1.8 million of that amount to be paid by National City, and the
remaining amount to be paid out of the settlement fund. Final Status
Report at 3. In other words, the parties expect that approximately
$500,000 would be taken out of the settlement fund proceeds to be paid
towards notice and claims administration costs. After deductions for
certain fees, costs, and expenses, including attorneys' fees, class
counsel estimate that approximately $8,407,000 will be available for
distribution to claimants. Id. at 4.
The Revised Settlement Agreement also contained a release, discharging
the "Released Parties *fn6 . . . from any and all
rights, claims, liabilities . . . that Releasing Parties *fn7
ever had . . . or may have in the future, that result from .
. . or relate to in any way to the conduct, omissions, duties or
matters alleged in the Complaint and in the MDL Proceedings (insofar
as [they] relate to the putative National City national class and state subclass . .
.)" Revised Settlement Agreement ¶ 37. This release extended to and
included any claims relating to (1) failures to adequately and fully
disclose the procedure for authorizing and approving overdrafts
arising out of debit card transactions, or the manner in which items
were posted to accounts, (2) failures to warn when debit card
transactions would cause an account to be withdrawn, (3) authorizing
and approving debit card transactions when they resulted in
overdrafts, (4) the use of posting orders, and (5) violations of the
Electronic Funds Transfer Act and Regulation E. See id. ¶ 37. After
entering into the Revised Settlement Agreement, PNC agreed to shorten
the release period from August 15, 2010 to June 21, 2010, the date in
which PNC's integration of National City accounts was completed.See
Settling Parties Supp. Mem. ("Supp. Mem.") at 14, Docket Entry No. 52
(Aug. 29, 2011).
The notice provided to class members stated the following regarding the release of claims:
Unless you exclude yourself, you give up the right to sue the Defendant for all of the claims that the proposed settlement resolves. You must exclude yourself from this Settlement Class to start your own lawsuit or be part of any different lawsuit relating to the claims in this case. For example, unless you exclude yourself, you will give up your right to be included in the proposed National City national class and state subclass in the litigation currently pending in the U.S. District Court of the Southern District of Florida as part of the In re Checking Account Overdraft Litigation, MDL No. 2036, or any other litigation.
See Exs. 3-5 to Mot. for Prelim. Approval of Rev. Settlement Agreement, Docket Entry No. 34-1.
III. ASSESSMENT OF DAMAGES
Edgeworth Economics ("Edgeworth") was retained by plaintiffs to quantify the injury to class members arising from the practice of resequencing electronic debt transactions, and prepared a report authored by two economists, Jesse David and Kevin W. Christensen ("Edgeworth Report"). To prepare the report, Edgeworth relied on discussions with and information provided by National City, documents produced through informal and formal discovery, and public information. Edgeworth Report ¶ 8.
Edgeworth estimated damages for the class period in four steps. First, evaluating sample data taken during September 2010, Edgeworth estimated the monetary difference between the overdraft fees actually charged on the electronic debit transactions and what would have been charged had National City's resequencing practice not been in place. From this amount, Edgeworth calculated the average excess overdraft fee charged per each overdraft fee that was actually charged ("Average Excess Overdraft Fee"). Id. ¶ 18A.*fn8 Edgeworth engaged in two "but for" scenarios to determine the range of damages. In the first scenario, where the transactions would have been arranged from low-to-high, Edgeworth calculated an Average Excess Overdraft Fee amount of $5.44 for each overdraft fee actually charged. Id. ¶ 24. Posting the transactions in simulated chronological order resulted in an Average Excess Overdraft Fee amount of $2.15.*fn9
Second, Edgeworth calculated the number of relevant overdraft fees incurred from August 2005 to August 2010*fn10 using information provided by National City. Id. ¶ 18B. Third, Edgeworth then applied the Average Excess Overdraft Fee to the number of overdraft fees charged during the class period to obtain the total monetary amount of excess overdraft fees, adjusting for fees that had been reversed or not collected by National City. Id. ¶ 18C. Fourth and finally, Edgeworth made downward adjustments to the amount of projected damages to account for reversed and/or non-collectible overdraft fees and "chronic" overdrafters. Id. ¶ 18.
Edgeworth estimated that a total of 59.7 million overdraft fees were charged by National City during the class period. Because National City did not keep data distinguishing overdraft fees incurred from electronic debit transactions from other types of transactions, the Edgeworth Report relied on an FDIC study of overdraft programs, which provided an industry-wide estimate that 48.8% of overdraft fees are caused by electronic debit transactions. Id. ¶¶ 26-27. Edgeworth also adjusted this amount downward assuming a 26.52% rate of reversed and/or uncollectible fees, which was the median uncollectible and reversed rate used by an expert in Gutierrez v. Wells Fargo Bank, 730 F. Supp. 2d 1080 (N.D. Cal. 2010), concluding that an estimated total number of overdraft fees of 29.2 million during the class period were attributable to electronic debit transactions. Id. ¶ 27. Multiplying that amount by the Average Excess Overdraft Fee, Edgworth concluded that based on the first scenario of low to high reordering, total damages would have amounted to $116.5 million and, applying the two months with the greatest number of overdraft fees to account for "frequent" or "chronic" overdrafters, total damages (i.e., excess overdraft fees) would amount to $50.8 million. See id. ¶¶ 28-29. Edgeworth concluded if the transactions had been chronologically ordered (the second scenario), the damages would amount to $46 million and, applying the two months with the greatest number of overdraft fees, total damages (i.e., excess overdraft fees) would amount to$20 million. Id. ¶¶ 28-29. Plaintiffs assert that the estimated damages arrived at in the second scenario, representing chronological posting, is more likely a proper representation of damages than the first scenario, where damages would have been based on a requirement that banks post transactions from low to high. See Mot. for Approval at 22. In summary, the range of damages spanned from $116.5 million to $46 million, and, discounting for chronic overdrafters, from $50.8 million to $20 million. Hence, plaintiffs assert that the $12 million settlement fund would be in the range of 10% to 26% of the two damages calculations, and, taking the two months in which a class member incurred the greatest number of overdrafts, 24% to 60% of the possible recovery. See id. at 20.
A. Impact of Regulation E
Edgeworth, relying on representations made by National City, explained that a data sample containing account-level transaction data contemporaneous with the class period "was not feasible to obtain" because of the way in which the data had been stored and archived when National City merged with PNC. Id. ¶ 11 n.4. Instead, Edgeworth extracted a sampling of this information during September 2010, which was taken from PNC account holders who had previously been National City account holders ("sample data"). The sample data consisted of approximately 1.3 million transactions and appeared to include all transactions on the day an overdraft fee had been incurred, as well as all transactions from the previous day. Id. ¶ 11. From this information, Edgeworth was able to determine the balance at the start of the day, the transactions that caused the overdraft fee, and the overdraft fees assessed to compare the total overdraft fees charged to those that would have been charged but for the high to low reordering of transactions. Id. ¶ 11.
The sample data was taken during a period of time after the Federal Reserve had made recent amendments to 12 C.F.R. § 205 ("Regulation E"),*fn11 which imposed certain disclosure and affirmative opt-in requirements on banks before enrolling customers in overdraft protection services. See Gutierrez, 730 F. Supp. 2d at 1135-36 (describing relevant changes to Regulation E). Edgeworth assumed for the purposes of its report that "the average characteristics and behavior of customers are the same before and after changes to Regulation E" and accordingly calculated excess overdraft fees using the sample data as "representative of what would have occurred during the Class Period. " Edgeworth Report ¶ 14. However, Edgeworth also acknowledged that "it is possible that the average characteristics and behavior of customers with overdraft protection . . . are somewhat different from the average characteristics and behavior of customers with such protection prior to Regulation E." Id.¶ 15.
At the final fairness hearing, the Court expressed concern with the Edgeworth Report's reliance on post-Regulation E data in quantifying the damages that occurred as a result of National City's resequencing policy. Plaintiffs conceded that the effect of Regulation E on the incurrence of overdraft fees was unknown, but nevertheless urged this Court to assume that Regulation E created no impact on the estimated valuation of damages. Tr. 15:4-16:9. The Court directed the parties to submit supplemental memoranda to further substantiate this claim, and to explain why the use of post-Regulation E data by Edgeworth was appropriate in this circumstance, where the damages were incurred by the class prior to the enactment of Regulation E. See Order for Supp. Br. & Final Status Rpt. ¶ 3, Docket Entry No. 47 (July 27, 2011).
In response, Edgeworth claimed that post-Regulation E data was used in
only one aspect
of calculating damages -- to calculate the "Average Excess Overdraft
Fee." According to Edgeworth, data regarding the actual number of
overdrafts incurred by the class during the class period (in other
words, pre-Regulation E data) was used to determine the frequency or
volume of overdraft fees incurred by account holders on a per-day
basis. As such, Edgeworth's damages calculation was "based on the
actual number of relevant overdrafts that occurred prior to the
implementation of Regulation E." David Decl. ¶¶ 5, 10.*fn12
Edgeworth claims that it "did not simply estimate the total
amount of overall excess overdraft revenue (caused by high-to-low
reordering) from the post-Regulation E sample period and then use that
amount to extrapolate what might have happened in prior months.
Instead, [Edgeworth] applied the Average Excess Overdraft Fee Amount
to the actual number of overdraft fees incurred by class members prior
to Regulation E." Id. ¶ 11 (emphasis added). Edgeworth also maintains
that both before and after filing its report, it conducted research
regarding the effect of Regulation E on daily transactional overdraft
behavior (i.e. "the banking behavior of accountholders within the day"
or "the average composition of accountholders eligible to incur
overdraft fees") and, based on that research, found "no literature" to
indicate "that any such changes . . . did occur." Id. ¶ 14. In any
event, Edgeworth claims that even if customers' behavior had changed
due to Regulation E,
Edgeworth's "methodology already includes a certain tolerance for such variation and implicitly accounts for small variations that may occur through such changes." Id. ¶ 15. Hence, according to Edgeworth, its assessment already "accounts for any changes in the rate or volume of overdraft fees potentially caused by Regulation E." Id. ¶ 10.
B. Data Retrieval and Analysis of Class Period Data
At the final fairness hearing, the Court also expressed concern withrespect to the settling parties' generalized assertions that data retrieval and usage of National City Bank data from the class period would be unduly burdensome and costly. The parties were therefore directed "to provide specific information regarding the feasibility of obtaining and analyzing such data, including the estimated cost of obtaining such data as well as the estimated time frames required for providing such data . . . ." Order for Supp. Br. & Final Status Rpt. ¶ 2.
In response, the settling parties submitted a declaration from lead consultant Steven Visser, the Managing Director of Navigant Consulting ("Navigant"). Navigant has significant experience assisting banks in extracting customer transactional data from bank mainframe systems. See Supp. Mem. at 2; Declaration of Steven Visser ("Visser Decl.") ¶ 2 (Aug. 26, 2011). It estimates that the total labor costs associated with locating, extracting, and converting all known and relevant transactional data for National City account holders for the entire class period into a database format would be approximately $620,000 and take approximately six months to complete. Visser Decl. ¶ 16. Navigant also estimated that the additional non-labor costs to support the data retrieval, which include hardware, software, and miscellaneous data center costs, would be approximately $153,000, for a total cost of approximately $773,000. Id.
¶¶ 17-18 & 21. These cost estimates do not account for additional relevant reports with information about authorizations, holds on deposits, and pending debit transactions, which, if available, would cost an additional $387,000 to extract and convert. Id. ¶ 20. The estimated total cost for the data retrieval project was projected by Navigant to be approximately $1,160,000. Id.
¶ 21. In addition to Navigant's projections, Edgeworth estimates that it would cost an additional $175,000 to $300,000 to assess such data to perform a damages analysis for the class period for each customer. David Decl. ¶ 19. Taking Edgeworth's and Navigant's estimates together, in sum, the total estimated cost for retrieving and analyzing account level data for the class members during the class period is estimated to be around $1,335,000 to $1,460,000.
Class counsel have identified ten objections that have been received as of July 9, 2011. Three of these objections were timely filed with the Court and are summarized briefly below.*fn13
Those three objections are addressed in further detail in the Court's discussion of whether final class certification and ...