Pandemic Darlings The pandemic economy, in original documents
Home Court filings In re: Bank of America California Unemployment Benefits Litigation — S.D. Cal., No. 21-md-02992 Memo of Points and Authorities — Bofa Ca Unemployment (Dkt. 566.1)

Court filing

Memo of Points and Authorities — Bofa Ca Unemployment (Dkt. 566.1)

Record facts

CourtU.S. District Court for the Southern District of California
Filed2025-10-17

U.S. District Court for the Southern District of California · No. 3:21-md-02992-GPC-MSB · Doc. 566-1 · 2025-10-17 · Docket on CourtListener

Summary

A memorandum of points and authorities filed October 17, 2025 as Document 566-1 in In re: Bank of America California Unemployment Benefits Litigation, Case No. 21-MD-02992-GPC-MSB, in the U.S. District Court for the Southern District of California. Filed by counsel for defendant Bank of America, N.A., it supports the bank's motion to exclude the opinions of a labor economist proffered by the plaintiffs as a damages expert. It argues that the report fails Daubert and Fed. R. Evid. 702 because it uses a 15.9% credit card interest rate as a conservative lower bound rather than evidence of actual damages. It recounts that on June 16, 2025 the court certified five classes of California unemployment insurance benefits recipients, for conduct running from September 28, 2020 to June 8, 2021. The memorandum is 27 pages and notes a hearing date of April 17, 2026.

Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used

Full text

BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE 
CASE NO. 21-MD-02992-GPC-MSB 
 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
JAMES W. MCGARRY (pro hac vice) 
JMcGarry@goodwinlaw.com 
GOODWIN PROCTER LLP 
100 Northern Avenue 
Boston, MA 02210 
Tel. +1 617 570 1000 
Fax: +1 617 523 1231 
SABRINA M. ROSE-SMITH (pro hac vice) 
SRoseSmith@goodwinlaw.com 
MATTHEW L. RIFFEE (pro hac vice) 
MRiffee@goodwinlaw.com 
GOODWIN PROCTER LLP 
1900 N Street, NW 
Washington, DC 20036 
Tel.: +1 202 346 4000 
Fax: +1 202 346 4444 
Attorneys for Defendant 
BANK OF AMERICA, N.A. 
[ADDITIONAL COUNSEL LISTED IN SIGNATURE BLOCK] 
 
UNITED STATES DISTRICT COURT 
FOR THE SOUTHERN DISTRICT OF CALIFORNIA 
SAN DIEGO DIVISION 
IN RE: BANK OF AMERICA 
CALIFORNIA UNEMPLOYMENT 
BENEFITS LITIGATION 
Case No. 21-MD-02992-GPC-MSB 
DEFENDANT BANK OF 
AMERICA, N.A.’S 
MEMORANDUM OF POINTS 
AND AUTHORITIES IN SUPPORT 
OF MOTION TO EXCLUDE 
PURPORTED EXPERT OPINIONS 
OF DAVID I. LEVINE 
Date: 
April 17, 2026 
Time: 
1:30 p.m. 
Dept: 
12A – 12th Floor 
Judge: 
Hon. Gonzalo P. Curiel 
 
ORAL ARGUMENT REQUESTED 
 
FILED PROVISIONALLY UNDER SEAL 
PURSUANT TO STIPULATED PROTECTIVE 
ORDER 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33472 
Page 1 of 27

 
i 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
TABLE OF CONTENTS 
Page 
INTRODUCTION ………………………………………………………………… 1 
BACKGROUND ........................................................................................................ 2 
I. 
Plaintiffs’ Claims and Damages Theories .......................................................... 2 
II. 
The Levine Report .............................................................................................. 4 
III. 
The McCrary Rebuttal ...................................................................................... 7 
STANDARD OF LAW .............................................................................................. 8 
ARGUMENT .............................................................................................................. 9 
I. 
Daubert Requires A Reliable Analysis, Not A “Conservative” Guess 
Untethered To Evidence. ................................................................................ 10 
II. 
Levine Offers No Evidence Of Damages, Individually Or Classwide. ........... 14 
CONCLUSION ......................................................................................................... 20 
 
 
 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33473 
Page 2 of 27

 
ii 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
TABLE OF AUTHORITIES 
 
Page(s) 
Cases 
Ayers v. Robinson, 
887 F. Supp. 1049 (N.D. Ill. 1995) ............................................................... 10, 11 
In re Blackbaud, Inc. Cust. Data Breach Litig., 
2024 WL 2155221 (D.S.C. May 14, 2024) ........................................................ 17 
Brighton Collectibles, Inc. v. RK Tex. Leather Mfg., 
923 F. Supp. 2d 1245 (S.D. Cal. 2013) (Curiel, J.) .......................................... 8, 9 
Comcast Corp. v. Behrend, 
569 U.S. 27 (2013) ......................................................................................... 9, 14  
Daubert v. Merrell Dow Pharms., Inc., 
509 U.S. 579 (1993) ................................................................................. 8, 17, 19 
Daubert v. Merrell Dow Pharms, Inc. (Daubert II), 
43 F.3d 1311 (9th Cir. 1995) ................................................................................ 9 
Domingo v. T.K., 
289 F.3d 600 (9th Cir. 2002) ................................................................................ 9 
United States ex rel. Fitzer v. Allergan, Inc., 
2024 WL 1156310 (D. Md. Mar. 18, 2024) ....................................................... 10 
Gen. Elec. Co. v. Joiner, 
522 U.S. 136 (1997) ........................................................................................... 12 
Guardant Health, Inc. v. Found. Med., Inc., 
2020 WL 2461551 (D. Del. May 7, 2020) ......................................................... 10 
Gutierrez v. Girardi, 
194 Cal. App. 4th 925 (2011) ............................................................................. 14 
Kewazinga Corp. v. Google LLC, 
2024 WL 4894840 (S.D.N.Y. Oct. 17, 2024) .................................................... 17 
Kolcraft Enters. v. Chicco USA, Inc., 
2018 WL 10772693 (N.D. Ill. July 16, 2018) .................................................... 12 
Metaswitch Networks Ltd. v. Genband US LLC, 
2016 WL 874775 (E.D. Tex. Mar. 7, 2016) ....................................................... 17 
Munoz v. JLO Automotive, Inc., 
2020 WL 6607789 (D. Conn. Nov. 12, 2020) .................................................... 19 
Obesity Rsch. Inst., LLC v. Fiber Rsch. Int’l LLC, 
165 F. Supp. 3d 937 (S.D. Cal. 2016) ................................................................ 14 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33474 
Page 3 of 27

 
iii 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Opperman v. Path, Inc., 
2016 WL 3844326 (N.D. Cal. July 15, 2016) .............................................. 13, 17 
Orshan v. Apple Inc., 
2024 WL 4353034 (N.D. Cal. Sept. 30, 2024) ................................................... 10 
Sabicer v. Form Motor Co., 
362 F. Supp. 3d 837 (C.D. Cal. 2019) ................................................................ 14 
Stokes v. John Deere Seeding Grp., 
2014 WL 675820 (C.D. Ill. Feb. 21, 2014) ........................................................ 10 
Stragent, LLC v. Intel Corp., 
2014 WL 1389304 (E.D. Tex. Mar. 6, 2014) ..................................................... 10 
Treviso v. Nat’l Football Museum, Inc., 
2018 WL 4608197 (N.D. Ohio Sept. 25, 2018) ................................................. 15 
Unwired Planet, LLC v. Apple Inc., 
2017 WL 589195 (N.D. Cal. Feb. 14, 2017) ...................................................... 10 
Van v. LLR, Inc., 
962 F.3d 1160 (9th Cir. 2020) ............................................................................ 20 
Statutes 
15 U.S.C. § 1693m .............................................................................................. 9, 14 
Rules 
FED. R. EVID. 702 ................................................................................................. 8, 17 
 
 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33475 
Page 4 of 27

 
1 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Plaintiffs cannot survive summary judgment without carrying the burden of 
producing evidence of actual damages for themselves and their classes. Labor 
economist David I. Levine is one of the purported “experts” on whom they rely for 
this purpose. His report does not so much perform any expert analysis, but rather 
merely serves to put the imprimatur of a credentialed expert on a simplistic 
proposition that Plaintiffs could just as easily have argued without him, and in fact 
already did. Plaintiffs and their classes consist of debit cardholders who temporarily 
lost access to a portion of their benefits payments when Bank of America (BANA) 
denied their disputes of allegedly unauthorized charges. Those charges have all since 
been credited back to them (
), so their claims for actual damages rest on 
claims of being injured by the delay. 
Levine’s role in this is to assert that the class’s damages can be measured by 
the cost of borrowing money on a credit card at a 15.9% annual interest rate to make 
up for the temporarily unavailable funds. He does not, and cannot, contend that 
Plaintiffs or any class members actually borrowed money on credit cards and paid 
15.9% interest. Rather, he claims (more accurately, presumes) that they suffered 
other damages based on other expenses and inconveniences, and thus the 15.9% 
interest rate does not measure their actual damages, but rather serves as a 
“conservative” lower bound on an actual-damage total that for “most” class members 
he believes is much higher. In other words, Levine has no idea the actual damages 
suffered by the Plaintiffs or by any class member—he is just in search of some 
number, any number, comfortably beneath it, to avoid overshooting it. 
As a purported damages model, this fails the Daubert criteria for relevance and 
reliability. It is, in fact, indistinguishable from a purported damages model this Court 
already rejected. Plaintiffs relied on another of their experts at the class certification 
stage to propose a damages methodology based on the assumptions that (i) class 
members “would likely have needed alternative funds to mitigate the inability to 
access their funds” and (ii) that “the most likely source of credit for these consumers 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33476 
Page 5 of 27

 
2 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
was utilization of credit cards.” ECF 494 at 87­88. This Court determined that these 
“assumptions about class members . . . may not be true classwide” as the purported 
expert “has not provided any evidence in support,” and thus rejected the method for 
failure to “measure[] damages across the entire class.” Id. The same applies to 
Levine’s opinions, which rest on the same unsupported assumptions. 
 Moreover, Plaintiffs’ burden is to evidence their actual damages. A method 
that does not even attempt to calculate those damages is, by its own terms, irrelevant 
for that purpose. Calling it a “conservative lower bound” does not change that: 
precedent cases repeatedly affirm that the expert’s duty is to produce an analysis 
that’s reliable and accurate, not “conservative.” Levine’s concession that his analysis 
does not even apply on a classwide basis, but only to “most” of the class—a claim 
itself made without factual foundation, and without considering records on even a 
single member of the class to validate the assumption—compounds the defects. In 
fact, Levine’s analysis does not even apply to most of the class representatives, 
 
, and he 
admits that his method cannot be used to measure the damages of any person 
individually. But Plaintiffs carry the burden of producing evidence of their damages 
individually, and to show that the same evidence applies classwide—not merely to a 
tiny portion of the class. As Levine’s analysis concededly does neither, it should be 
excluded from the record. 
BACKGROUND 
I. 
Plaintiffs’ Claims and Damages Theories 
On June 16, 2025, this Court certified five classes of California unemployment 
insurance (UI) benefits recipients alleging that they contacted BANA claiming 
unauthorized use of their benefits prepaid debit card, but had those claims denied 
because BANA’s records showed that the disputed transactions were made in-person 
at ATMs, requiring the cardholder’s physical card and a claim that their private PIN 
passcode was somehow compromised. ECF 494 at 96-97. The relevant time period 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33477 
Page 6 of 27

 
3 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
runs about seven-and-a-half months, from September 28, 2020, when BANA 
implemented the challenged fraud-detection process (CFF-1), to June 8, 2021, when 
BANA ceased using CFF-1 to deny claims. See id.  
Separately, in July 2022, BANA entered into a settlement agreement with the 
Office of the Comptroller of the Currency (OCC) and the Consumer Financial 
Protection Bureau (CFPB) that included a framework providing full compensation 
for cardholders whose claims might have been inaccurately decisioned by CFF-1. See 
generally HX1 28; HX 29. 
 
 See HX 30 at 
77:7-17 (testifying that BANA 
 
 
). Following that 
settlement agreement, 
 
 
 
 HX 31 at 4 n.16; ECF 350-8 ¶¶ 9-14. In addition to paying 
cardholders the full value of their disputed claims, 
 
 
 
 HX 31 at 5-13. BANA used 
 
 even though it 
 
 and 
determined such a rate to be 
 
 Id. at 7, 9, 13. The Remediation Plan did not attempt to 
 
 id., at 12, but cardholders were 
 
 Id. 
 
 See ECF 350-9 ¶¶ 14-15. 
 
1 Exhibits to the Declaration of Lindsay E. Hoyle shall be referred to as “HX.” 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33478 
Page 7 of 27

 
4 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Plaintiffs assert that 
 
 
 But the Court determined that “the 
Remediation Plan does not support a damages model” that carries Plaintiffs’ burden 
because it 
 
 ECF 494 at 87. In particular, the Court rejected a proposed 
method of basing damage calculations on assumptions that class members “would 
likely have needed alternative funds” 
 and 
would “most likely” have used credit card borrowing to obtain them, on the ground 
that Plaintiffs lacked “evidence showing that these assumptions are true as to most 
or even any of the EDD cardholders.” Id. at 87-88. 
II. 
The Levine Report 
David I. Levine claims expertise “in the field of labor economics” and teaches 
at the Haas School of Business. HX 35 ¶¶ 1, 7. Plaintiffs proffer him as a damages 
expert, primarily to furnish a “methodology for determining the value of the lost 
opportunity costs to the class members whose access to UI benefits was delayed or 
denied by the Bank’s challenged policies and practices”—which he considers “an 
appropriate measure of damages.” Id. ¶¶ 5, 16. Just like another of Plaintiffs experts, 
Greg J. Regan, the method Levine proposes is to use “the credit card interest rate” as 
“a conservative measure of the average opportunity cost faced by members of the 
impacted classes.” Id. ¶ 44. He “estimate[s] that class members paid an effective 
interest rate of at least 15.9%” and describes this 15.9% figure as “a conservative 
lower bound on the average opportunity cost funds [sic] for UI recipients in 
California during the COVID pandemic.” Id. ¶ 45. 
 
 But 
the theory is that consumers temporarily lacking access to funds might need 
“alternative sources of funds” during that time period, and borrowing funds on a 
credit card is one such alternative source. Id. ¶ 13. 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33479 
Page 8 of 27

 
5 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Despite using “the credit card interest rate” as the basis for his “measure of 
damages” (id. ¶¶ 16, 44), Levine does not claim 
 
 
 He claims this is irrelevant, because “the alternative sources of funds 
available to most class members are higher cost than the cost of credit card 
borrowing.” Id. ¶ 13. For example, he considers “borrowing from friends or family” 
to be one of the alternatives that is higher cost than the cost of credit card borrowing, 
because it might come with a “nonfinancial burden in terms of status or reputation.” 
HX 36 at 55:23-57:13. Thus, he assumes that “most” class members had to choose 
between credit card borrowing at an average 15.9% interest rate and an alternative 
whose “nonfinancial” costs were higher. Id.  
But in presuming this true for “most” class members, Levine does not contend 
it to be true for all class members, or even for any identifiable class member—indeed, 
he has no idea the number of class members for whom it is true. See id. at 57:24-
58:1. He did not review any information about any of the class members, nor even 
the class representatives. See, e.g., id. at 30:4-24, 39:2-18. He admits that “not 
everybody” would need to borrow money if faced with a temporary loss of access to 
UI benefits, that “some” people might simply access their own savings, and when 
questioned about how many class members did one or the other, admitted that “I 
don’t have any data.” Id. at 69:9-70:8; see also id. at 71:13-16 (“Q. Right. Do you 
have any data on the proportion of class members who chose to pay on credit cards 
in lieu of accessing liquid funds? A. No.”).  
That is why he couches his opinions in terms of proposing a damage measure 
in the form of “a conservative lower bound that would apply to the vast majority of 
the class members.” Id. at 32:19-24. But he acknowledges multiple factors that could 
make his methodology inapplicable to some unspecified “minority” of class 
members. Asked what factors “would cause somebody to belong to that small 
minority” for whom his damages methodology is inapplicable, he cited: 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33480 
Page 9 of 27

 
6 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Not have any liquidity constraints, not have concerns about precautionary 
savings, not have a high subjective discount rate, not be overly concerned 
about the duration of the delay, not be facing uncertainty in the pandemic 
about income or expenses or correlated shocks that might increase the need 
for precautionary savings. . . . Not have transaction costs be a large portion of 
the opportunity cost of not having access to funds, not have credit card debt 
that they could be repaying or any other high-interest rate debt they could be 
repaying, and anything I’m forgetting. . . . [N]ot be reducing consumption in 
things that are hard to substitute across time, not be relying on sources of 
credit with high nonmonetary costs in terms of reputation or status or social 
obligations. . . . I’m pretty sure I’ll want to add to this list as the day proceeds. 
Id. at 47:10-48:15; see also id. at 100:18-101:1 (similar). In brief, Levine’s 
“conservative lower bound that would apply to the vast majority of class members” 
does not actually measure the damages experienced by any class members, just a 
claimed “lower bound” on their alleged damages, and does not claim that even the 
“lower bound” applies to everyone, just to the “majority.” But whether it actually 
applies to the majority is just an assumption, not based on any review of any records 
of their behavior or experiences. 
Apart from his opinions on the damages allegedly suffered from temporarily 
losing access to a portion of their benefits, Levine also proposes a “methodology for 
calculating the value of the time lost by class members” who dialed BANA’s 
customer-service number and claim to have spent “excess” time on hold before 
reaching a representative. HX 35 ¶ 5. He does not offer any opinion on how much 
hold time is “excess[ive]” and is “unsure how [Plaintiffs] [a]re computing that.” 
HX 36 at 142:16-143:2. But he proposes to attach an economic value to whatever 
number Plaintiffs compute. His proposal is to value the time at “the minimum wage,” 
but again only proffers this as “a conservative estimate” and claims the actual value 
of the lost time is even higher. HX 35 ¶ 53. Most studies, he claims, “have found that 
the typical value of time is close to the median wage,” not the minimum wage. Id. 
¶ 48. But this “typical value” is really just an average value, based on some people 
valuing their time above the median and others below the median, because people 
tend to value their time in line with their own respective earnings potential. See 
HX 36 at 146:7-148:5. Thus, his minimum-wage measurement figure again is not 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33481 
Page 10 of 27

 
7 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
proffered as a damages measure applicable to all class members—only Levine’s idea 
of a “lower bound” applicable to the class on “average.” HX 35 ¶ 55. 
III. 
The McCrary Rebuttal 
Economist Justin McCrary addressed Levine’s conclusions and methods in a 
rebuttal report. He considered Levine’s claim of a 15.9% opportunity cost of 
temporarily lost funds based on the credit card interest rate “unsupported and 
inflated” for multiple reasons. HX 34 ¶ 17. First, there was “evidence to suggest 
proposed class members would have had access to alternative sources of capital, and 
no evidence they uniformly or typically turned to credit cards.” Id. ¶ 73. These 
alternative sources of capital “would have no or lower cost of borrowing than the 
average credit card interest rate,” but Levine did not consider them in his analysis. 
Id. The class representatives themselves were demonstrative. “Of the nine class 
representatives, . . . 
 
 
” Id. ¶ 75. 
 
 
 
 Id. Of the remaining class representatives, 
 
 
 
 Id. ¶¶ 75-76. Among the larger group of named Plaintiffs not 
proffered as class representatives, 48 provided interrogatory responses on their 
claimed damages, and “
.” Id. 
¶ 77. 
Furthermore, Levine based his conclusion that “most” class members would 
have needed to resort to credit card borrowing on the premise that “UI benefits are 
typically used for essential expenses.” Id. ¶ 73; HX 35 ¶ 12. But Levine’s claims 
about the “typical” uses of UI benefits were based on data related to equally typical 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33482 
Page 11 of 27

 
8 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
UI programs predating the pandemic—not the pandemic-era benefits at issue here. 
See HX 34 ¶¶ 50-58. The distinction is significant because during the pandemic, the 
“pandemic-related stimulus resulted in many households increasing their savings”—
in contrast to the usual profile of a UI benefits recipient—and because many of the 
“essential expenses” that recipients might otherwise have needed to pay were no 
longer essential: for example, borrowers could benefit from a year-long forbearance 
on making mortgage-loan payments, three-and-a-half years of forbearance on 
student-loan payments, and similar forbearance on auto loans. Id. ¶¶ 36, 38, 80 
(emphasis added). For all these reasons, far from its being the case that “most UI 
recipients turn to credit cards” when their benefits lapse (HX 35 ¶ 1), the pandemic 
era saw a decrease in credit card borrowing as consumers used the various forms of 
available assistance to pay down credit card debt. HX 34 ¶¶ 57, 67. Thus, “far from 
being conservative, the credit card borrowing rate . . . represents an inflated cost of 
the temporary inability to access a portion of UI benefits for the proposed class 
members.” Id. ¶ 103. 
STANDARD OF LAW 
Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharms., Inc., 509 
U.S. 579 (1993), “impose[] a special ‘gatekeeping obligation’ on trial judges” 
presented with expert testimony. Brighton Collectibles, Inc. v. RK Tex. Leather Mfg., 
923 F. Supp. 2d 1245, 1253 (S.D. Cal. 2013) (Curiel, J.). Under Rule 702, a witness 
proffered as an expert must satisfy the following requirements: 
(a) the expert’s scientific, technical, or other specialized knowledge will help 
the trier of fact to understand the evidence or to determine a fact in issue; (b) 
the testimony is based on sufficient facts or data; (c) the testimony is the 
product of reliable principles and methods; and (d) the expert has reliably 
applied the principles and methods to the facts of the case. 
Under Daubert, district courts must “carefully apply[] Federal Rule of Evidence 702 
to ensure that specialized and technical evidence is ‘not only relevant, but reliable.’” 
Brighton Collectibles, 923 F. Supp. 2d at 1253. The reliability standard tests: 
(1) whether the scientific theory or technique can (and has been) be tested; (2) 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33483 
Page 12 of 27

 
9 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
whether the theory or technique has been subjected to peer review and 
publication; (3) whether there is a known or potential error rate; and (4) 
whether the theory or technique is generally accepted in the relevant scientific 
community. 
Domingo v. T.K., 289 F.3d 600, 605 (9th Cir. 2002). In addition, “[o]ne very 
significant fact to be considered is whether the experts are proposing to testify about 
matters growing naturally and directly out of research they have conducted 
independent of the litigation, or whether they have developed their opinions 
expressly for purposes of testifying.” Daubert v. Merrell Dow Pharms, Inc. (Daubert 
II), 43 F.3d 1311, 1317 (9th Cir. 1995). 
“[T]he burden of proving the expert’s testimony satisfies Rule 702” and the 
Daubert standard rests on the Plaintiffs, as “[t]he proponent of the evidence.” 
Brighton Collectibles, 923 F. Supp. 2d at 1253. 
ARGUMENT 
Plaintiffs’ claims require them to furnish evidence of “actual damage sustained 
by [Plaintiffs] as a result of” the defendant’s conduct. 15 U.S.C. § 1693m(a)(1). And 
Because Plaintiffs assert their claims on behalf of a class, the method they use to 
evidence actual damages must measure them “across the entire class.” Comcast Corp. 
v. Behrend, 569 U.S. 27, 35 (2013); accord ECF 494 at 87­88. Levine’s proposed 
method is not admissible for either purpose. It does not even pretend to measure the 
“actual damage sustained” by any Plaintiff individually or by the “entire class” of 
Plaintiffs. It leaves both questions consigned to the realm of the unknown (and 
unknowable), and instead simply opines that whatever the correct measure of 
damages is for any Plaintiff or for the class as a whole, it can be assumed to be higher 
than his proposed “conservative lower bound.” This fails the Daubert standard on 
multiple grounds. It is not a reliable method for measuring actual damages (since it 
does not even pretend to do so) and is not based on any evidence or data. It is therefore 
not relevant to the Court or to a factfinder in assessing whether Plaintiffs have the 
evidence of actual damages that they need. 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33484 
Page 13 of 27

 
10 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
I. 
Daubert Requires A Reliable Analysis, Not A “Conservative” Guess 
Untethered To Evidence. 
“Daubert asks whether expert opinions are reliable and relevant, not whether 
they are conservative.” Orshan v. Apple Inc., 2024 WL 4353034, *3 (N.D. Cal. Sept. 
30, 2024). “If a damage model could survive Daubert by simply underestimating true 
damages, an expert could avoid having a court exclude her opinions by picking an 
arbitrary damage figure that is comfortably below any reasonable amount of true 
damages even though such an opinion would be plainly unreliable.” Id. But that is all 
Levine has done, and Plaintiffs cannot “sneak” his model “past Daubert” in that 
manner. Id. 
The case law is unequivocal about this. “Just because an approach is 
conservative does not mean it is reliable.” United States ex rel. Fitzer v. Allergan, 
Inc., 2024 WL 1156310, *5 (D. Md. Mar. 18, 2024); accord, e.g., Unwired Planet, 
LLC v. Apple Inc., 2017 WL 589195, *2 (N.D. Cal. Feb. 14, 2017) (“Nor does the 
fact that the . . . model yields relatively conservative results offer assurance that those 
figures are useful or reliable. . . . Instead, by shrouding the results in an air of 
legitimacy, these factors raise the risk of juror confusion, making it all the more 
essential that the Court exercise its gatekeeping function until admission is 
adequately supported.”).2 As one court reasoned:  
Note, for example, how quick [the purported expert] is to exploit the method’s 
malleability by suggesting that he could have picked a higher number . . . and 
that by opting for the lower figure he is somehow rendering a “conservative” 
opinion. Maybe so, but a conservative opinion in that sense does not equate 
to a scientific one. Someone who states on the basis of a dull pain in his right 
knee that he thinks it is going to rain less than .1 inch expresses a conservative, 
but surely an unscientific, opinion. 
Ayers v. Robinson, 887 F. Supp. 1049, 1060-61 (N.D. Ill. 1995).3  
 
2 See also Guardant Health, Inc. v. Found. Med., Inc., 2020 WL 2461551, *18 (D. 
Del. May 7, 2020) (“merely labelling a value ‘conservative’ is no substitute for a 
showing that there is an evidentiary foundation for the particular percentage 
selected”); Stokes v. John Deere Seeding Grp., 2014 WL 675820, *4 (C.D. Ill. Feb. 
21, 2014) (“the fact that an opinion is conservative does not make it scientific”). 
3 See also Stragent, LLC v. Intel Corp., 2014 WL 1389304, *4 (E.D. Tex. Mar. 6, 
2014) (rejecting “attempts to justify [expert’s] estimate on the ground that it is ‘quite 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33485 
Page 14 of 27

 
11 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Levine’s reference to the credit card interest rate is perhaps not quite as 
unscientific as the proverbial dull pain in the knee, but it is not as far off as he thinks. 
In both situations, the dispositive fact is that the premise for the opinion has no 
analytical connection to the conclusion. Levine is not claiming that every class 
member (or any class member) was damaged in the amount of the credit card interest 
rate because they paid credit card interest. Rather, his claim is that their damages are 
based on other factors (and most likely higher), but since he doesn’t know what they 
are, he will use the credit card interest rate instead of actually trying to measure them 
or risk overestimating them. Nothing about their alleged or actual damages has 
anything to do with the credit card interest rate—it has no connection at all to any 
class member’s alleged or actual damages apart from Levine’s confidence that 
whatever those damages are, they are probably higher.  
Demonstrating this at his deposition, Levine speculated about a consumer who 
“loses access to $1,000 for 3 days,” which “means the damages are on the order of 
$1.50 using my method.” HX 36 at 50:3-14. The $1.50 is supposed to represent “the 
transaction cost of dealing with why our debit card isn’t working as expected,” but 
Levine simultaneously contended that the actual cost “is more than $1.50 for the vast 
majority of the class.” Id. Levine then repeatedly dodged questions asking him to 
describe “the connection” between the actual costs “and the credit card interest rate,” 
but ultimately admitted there was no such connection at all beyond his assertion that 
the actual costs are “at least” that much. Id. at 50:25-53:8.  
Q. So this person’s damages are not based on the credit card rate, you’re just 
certain that whatever damages those are, they are more than the credit card 
rate? 
A. For the vast majority of people, the transaction cost would be more than 
what would be implied by the credit card interest rate, yes. 
*     *     * 
Q. Do the damages they suffered have anything to do with the credit card 
 
conservatively low,’ because “[d]espite [plaintiff’s] assertions, a ‘conservative 
opinion in that sense does not equate to a scientific one’”) (quoting Ayers, 887 F. 
Supp. at 1060)); ECF 494 at 87 (method “based on assumptions rather than evidence” 
that “could fall short of compensating all cardholders” is not “a damages model that 
satisfies Comcast”). 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33486 
Page 15 of 27

 
12 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
interest rate other than being more than it? 
A. In this hypothetical, I’m assuming that their only cost are these transaction 
costs. They could also be—they might have other costs and then some of those 
would be tied to the credit card interest rate. But I’m saying, in this simplest 
example where they have no other cost, the credit card interest rate remains a 
conservative lower bound. 
 
Id. (objections omitted). The fatal problem is that calling the interest rate a “lower 
bound” does not establish any connection between that rate and Plaintiffs’ damages. 
If the class were limited to people who actually paid interest on loans, it would be 
fair to assume that the lower bound on what they paid is the lowest interest rate 
available. But in a class not limited to people who paid interest on loans, selecting an 
interest rate as the lower bound of damages is arbitrary and disconnected from their 
theory of recovery. See, e.g., Kolcraft Enters. v. Chicco USA, Inc., 2018 WL 
10772693, *4 (N.D. Ill. July 16, 2018) (“the arbitrary baseline taints the entire 
damages calculation”). 
In other words, the interest rate “is connected to existing data only by the ipse 
dixit of the expert.” Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997). It serves as 
the “lower bound” of damages only because Levine says so. But its relevance is 
actually even more attenuated than that. Levine doesn’t even offer his say-so that the 
interest rate is connected to Plaintiffs’ actual damages. He expressly concedes 
otherwise, that their actual damages are based on other costs unaccounted for in his 
model, and that the interest rate has nothing to do with those costs: 
Q. . . . Where their only costs are the transaction costs, their costs are greater 
than the credit card interest rate; is that your opinion? 
A. Yes, for these small or short claims. . . . [T]he transaction costs are going 
to be higher. . . . So the point is, even for small or short claims, the credit card 
interest rate, even if they have no need to borrow, remains a very conservative 
lower bound. 
Q. Can the credit card interest rate be used to calculate their transaction costs? 
A. I’m using it as a conservative lower bound. Does that answer your 
question? 
Q. Not quite. So I understand your opinion that the interest rate is a 
conservative lower bound, meaning the transaction costs are higher than the 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33487 
Page 16 of 27

 
13 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
figure produced by the interest rate. Am I stating that accurately? 
A. Yes. 
Q. Can the interest rate be used to figure out what the transaction costs are? 
A. That’s not part of how I calculated the transaction costs, using the credit 
card interest rate. 
 
HX 36 at 53:21-55:7. In sum, Levine admits that the correct measure of Plaintiffs’ 
damages would be based on “transaction costs” and unspecified other costs they may 
have incurred, but does not base his damages measure on those costs. Instead, he 
bases his damages measure on costs they did not incur, only because he is confident 
that it makes his measure “conservative.” But that does not make it correct or reliable. 
This is fatal. 
Levine’s other damages measure, using the minimum wage as a basis for 
attaching a value to the alleged “lost time” of the customer-service class, suffers from 
the same fundamental defects. He does not assert that any member of this class could 
or would have collected compensation for their time at the minimum-wage rate if 
they were not spending that time telephoning Bank of America’s customer service. 
Rather, he admits that every class member values their time differently, but he will 
use the minimum wage “[t]o err on the side of a conservative estimate.” HX 35 ¶ 53. 
But to err on the side of a conservative estimate is still to err. That is especially so 
when the erroneous estimate has no factual basis beyond “the unsurprising 
conclusion that people do not like waiting on hold for customer service.” Id. ¶ 51. 
That conclusion may perhaps be unsurprising, but it is also unscientific as a basis for 
calculating damages. A dislike of waiting on hold is not a justification for attaching 
a wholly arbitrary value to the time, much less for affording that arbitrary value the 
imprimatur of an “expert.” See, e.g., Opperman v. Path, Inc., 2016 WL 3844326, *14 
(N.D. Cal. July 15, 2016) (“No damages number arising from this model will apply 
to all class members,” “because consumers do not have identical preferences” and 
“each class member will place a very different value on” their alleged injury). 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33488 
Page 17 of 27

 
14 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
II. 
Levine Offers No Evidence Of Damages, Individually Or Classwide. 
Plaintiffs argued at the class certification stage that they “need only ‘establish 
that damages are capable of measurement on a classwide basis,’” and that 
“[u]ncertainty regarding class members’ damages does not prevent certification of a 
class as long as a valid method has been proposed for calculating those damages.” 
ECF 324-1 at 37-38 (emphasis in original); see generally Comcast, 569 U.S. 27. But 
this case is no longer at the certification stage. Plaintiffs’ burden at the summary 
judgment stage is not merely to show that actual damages are “capable” of classwide 
measurement, but to produce actual evidence of those damages. See 15 U.S.C. 
§ 1693m(a)(1)4. Since Levine’s report furnishes no such evidence, it is not relevant 
to the Court or to a factfinder, and does not clear the Daubert hurdle. 
Levine admits that his method cannot be used to calculate any Plaintiff’s 
damages—or any class member’s damages—individually. Or, more pointedly, he 
eventually admitted as much after multiple efforts to dodge the question: 
Q. Can your method be used to measure the individual harms suffered by any 
individual class member? 
A. I was asked to create a method to get a conservative lower bound for 
estimating the aggregate harm. I wasn’t asked to— 
Q. . . . So can your method be used to assess the harm experienced by any 
individual class member? 
A. I wasn’t asked to do that, so I don’t . . . 
Q. Do you have an opinion on it?  
A. Ask the question again. 
Q. Can your method be used to assess the harm experienced by any individual 
class member? 
A. I’m hesitant to answer because it’s an ill-posed question. My method was 
 
4 See also Sabicer v. Form Motor Co., 362 F. Supp. 3d 837, 840 (C.D. Cal. 2019) 
(negligence claim requires damages); Obesity Rsch. Inst., LLC v. Fiber Rsch. Int’l 
LLC, 165 F. Supp. 3d 937, 947 (S.D. Cal. 2016) (plaintiff must “establish a loss or 
deprivation of money or property sufficient to qualify as injury in fact” under Unfair 
Competition Law); Gutierrez v. Girardi, 194 Cal. App. 4th 925, 932 (2011) 
(fiduciary duty requires showing of “damage proximately caused by the breach) 
(internal quotations omitted).  
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33489 
Page 18 of 27

 
15 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
trying to create a conservative lower bound for the class and then use the 
individualized information on claimed amounts and duration. It makes it very 
hard to say how . . . it was not designed for any other purpose besides that. . . . 
Q. Can your method be used to assess the harms experienced by any member 
of the class? 
A. I wasn’t designing the method with that intent, and I don’t have an expert 
opinion on that. I just—you guys are welcome to hire me to answer that 
question, but I just don’t have an opinion that I can defend right now. 
HX 36 at 73:22-75:25 (objections omitted). The exchange continued further in that 
vein, and BANA’s counsel then asked Levine, “[F]or any individual class member, 
if we wanted to know what their damages were, what would we do?” Id. at 77:18-
78:3. Levine had no idea. He responded that “that’s a really different question” from 
the one on which he opined. Id.; see also id. at 79:11-18. He ultimately admitted that 
he considers his method appropriate for assessing the aggregate damages for a class 
of thousands, but not for smaller groups of people or for any one person. See id. at 
80:24-81:16. Having obtained class certification on the premise that “individualized 
damages issues do not alone defeat certification” (ECF 324-1 at 37), Plaintiffs cannot 
now escape the implication that “each individual class member will need to submit 
proof of their damages.” Treviso v. Nat’l Football Museum, Inc., 2018 WL 4608197, 
*7 (N.D. Ohio Sept. 25, 2018) (ruling proposed expert methodology unacceptable for 
failure to account for costs incurred by individual class members, “all of which are 
damages Plaintiff alleges the class is entitled to recover”). The conceded inability of 
Levine’s method to serve that purpose renders it irrelevant and inadmissible.  
Likewise, Levine expressly disclaimed any notion that his method could be 
used to prove the damages incurred by the class representatives. He considered “the 
named plaintiffs . . . a nonrandom sample that was small,” and so “didn’t look at [] 
any evidence about them when preparing [his] report.” HX 36 at 30:13-18; see also 
id. at 92:14-20 (“I had not looked at the interrogatories of the named plaintiffs 
because they’re a nonrandom sample and a small sample.”). He had not even 
reviewed Plaintiffs’ complaint. Id. at 84:13-17. But BANA’s counsel presented 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33490 
Page 19 of 27

 
16 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Levine with the class representatives’ allegations and discovery responses at his 
deposition, and asked him what information he considered relevant to forming an 
opinion on their damages. Levine stated at the outset that his method was not 
“designed” for that purpose and could not be used to determine whether their 
damages were above or below the credit card interest rate. E.g., id. at 86:2-24, 
102:22-24. 
Nonetheless, Levine had no trouble opining that at least some class 
representatives had experienced an array of pecuniary and nonpecuniary damages 
based on their allegations and discovery responses. What he could not do was 
articulate any connection between those damages and the credit card interest rate. 
Levine opined that class member Stephanie Moore suffered consequential damages 
in the form of “
 
 
.” Id. at 95:25-97:20.5 But his 
method “wasn’t designed to value each of these,” and he was confident Moore 
“suffer[ed] more harm” than his method measured. Id. As for class representative 
Kuang Ting Chong, who “
 
 and thus did not claim any of the sort of 
consequential damages claimed by Moore, Levine determined that Chong did not 
furnish “enough information here to judge” whether or not his damages were above 
or below the figure Levine’s method would produce based on the credit card interest 
rate. Id. at 102:5-24. 
The conceded inapplicability of Levine’s method to the actual class 
representatives—and to some other unspecified proportion of the class—is highly 
significant to assessing the relevance and reliability of his analysis under the Daubert 
standard. A damages method that underestimates damages for “most” class members 
 
5 See also HX 37 at No. 14. 
6 See HX 38 at No. 14.  
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33491 
Page 20 of 27

 
17 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
and overestimates damages for the rest is of no use to the Court. See, e.g., Opperman, 
2016 WL 3844326, *14 (“It may be that the average damages that [a proffered expert 
economist]’s model would predict will be very close to the damages actually suffered 
by every class member, but there is no way of knowing this. It is equally or more 
likely that his model would overcompensate some class members while 
undercompensating others.”); ECF 494 at 87. The prospect of overcompensation here 
is not merely “likely,” but certain. Levine has specifically described the 
characteristics of class members he admits his method would overcompensate. See 
HX 36 at 47:10-48:15, see also id. at 100:18-101:1. He also described the 
characteristics of class members he believes his method would undercompensate. 
E.g., id. at 95:25-97:20 (discussing Moore); ECF 494 at 87 (rejecting method that 
“could fall short of compensating all cardholders” as unsupported by “evidence 
establishing each cardholder’s experience”). 
Further, a “key question” under Daubert is whether the expert’s method “can 
be (and has been) tested.” 509 U.S. at 593. A method that does not produce reliable 
and accurate conclusions as applied to the class representatives or some other sample 
of the class ipso facto cannot serve as a reliable and accurate method as applied to 
the class as a whole. See, e.g., In re Blackbaud, Inc. Cust. Data Breach Litig., 2024 
WL 2155221, *10, 13-14 (D.S.C. May 14, 2024) (excluding expert who “tested his 
method on three named plaintiffs,” failed to “indicate whether or how he verified the 
accuracy of the [method] for those three individuals,” and “chose not to conduct any 
testing beyond [the] three named plaintiffs”). 
Daubert also requires that expert opinion be “based on sufficient facts or data.” 
FED. R. EVID. 702(b); see, e.g., Kewazinga Corp. v. Google LLC, 2024 WL 4894840, 
*4 (S.D.N.Y. Oct. 17, 2024) (striking damages computation as not based on 
“sufficient facts or data,” because “a very high-level view” “does not mean that those 
numbers are sufficiently reliable to develop a ratio to compute damages”); 
Metaswitch Networks Ltd. v. Genband US LLC, 2016 WL 874775, *3 (E.D. Tex. 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33492 
Page 21 of 27

 
18 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Mar. 7, 2016) (excluding part of expert’s report as “not supported by sufficient ‘facts 
or data’” where expert “admit[ted] in his report that ‘there is no information available 
with regard to’” the subject matter, and ruling that “[a]n absence of information is 
not a license to speculate”). Levine plainly fails to base his method on “sufficient 
facts or data,” given that the facts or data he considered relevant to the harms 
allegedly incurred by individual Plaintiffs like Moore are not accounted for in his 
method, and given his admission that he did not even have enough facts or data to 
form an opinion on the harms allegedly incurred by individual Plaintiffs like Chong. 
Plaintiffs’ complaint itself is also replete with claims of injury by the named Plaintiffs 
that Levine makes no effort to account for with his method. E.g., ECF 304 ¶ 519 
(claiming consequential harms from missing rent payments, eviction, car 
repossession, and skipped cable and electric bills).  
Applying Levine’s method to the entire class, as he urges, merely exacerbates 
its unreliability. As already shown, Levine admits his method does not apply on a 
classwide basis—only to a claimed “majority” of the class, although Levine has no 
facts or data to back up that claim, or to assess how large that “majority” is even if 
the claim is true. But he sets forth multiple factors that would make a class member 
a member of the “minority” whom he admits his damages method would over-
compensate. These include class members who: 
• “did not face any liquidity constraints”; 
• “did not have [a] high subjective discount rate”; 
• “w[ere] not reducing their spending on hard-to-shift items”; 
• “w[ere] not foregoing repaying a credit card”; 
• had “precautionary savings that permitted them to self-insure against 
shocks to their income or expenses”; 
• did “[n]ot have transaction costs be a large portion of the opportunity cost 
of not having access to funds”; 
• were not “relying on sources of credit with high nonmonetary costs in 
terms of reputation or status or social obligation.” 
HX 36 at 47:10-48:15; see also id. at 100:18-101:1. These are not fanciful 
hypotheticals: as noted, multiple named Plaintiffs confirmed 
 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33493 
Page 22 of 27

 
19 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
 
 (see, e.g., id. at 101:5-102:24), so the class 
of a hundred thousand people undoubtedly contains more. But Levine’s method 
would award damages to class members without considering any “facts or data” 
relevant to whether they are in the population to whom he admits his method does 
not apply. 
Lastly, Daubert asks whether a proffered expert’s methods enjoy “general 
acceptance” in the field. 509 U.S. at 594. “Widespread acceptance can be an 
important factor in ruling particular evidence admissible, and ‘a known technique 
which has been able to attract only minimal support within the community’ may 
properly be viewed with skepticism.” Id. Levine does not make any claims that using 
the credit card interest rate to calculate damages for a class of people who did not pay 
credit card interest enjoys any such general acceptance. The most he offers are his 
citations to academic studies for the proposition that “[w]hen UI payments disappear, 
most UI recipients turn to credit cards to cover those expenses.” HX 35 ¶ 12. That 
falls far short of indicating any general acceptance for his method here, where there 
is no evidence that most class members turned to credit cards and ample evidence of 
class members who did not. And, as McCrary pointed out in rebuttal, this case does 
not present a situation where “UI payments disappear[ed]”—the only claim is that 
“the proposed class members were temporarily unable to access a portion of their UI 
benefits.” HX 34 ¶ 48. Levine does not present any studies that measure the 
opportunity costs of that scenario, much less any studies that do so using the method 
he proposes. 
Nor is there any precedent case law ratifying this method of calculating 
damages for a temporary loss of access to funds. The cases do not permit plaintiffs 
to claim interest expenses as damages with no evidence of actually paying them. See, 
e.g., Munoz v. JLO Automotive, Inc., 2020 WL 6607789, *3 (D. Conn. Nov. 12, 2020) 
(interest charges not cognizable as “actual damages” with “no evidence in the record” 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33494 
Page 23 of 27

 
20 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
that plaintiff “actually incurred those interest charges”). To the contrary, in Van v. 
LLR, Inc., 962 F.3d 1160, 1161-65 (9th Cir. 2020), the Ninth Circuit assessed “time 
value of money” damages for a delayed reimbursement based on the 4.35% interest 
the plaintiff would have earned on the money, not the interest the plaintiff would 
have paid to borrow the money. See ECF 494 at 86­87. This Court found Van’s logic 
applicable here, but rejected the notion that the interest rate could be based on 
assumptions about “increased utilization of credit cards” where Plaintiffs had no 
“evidence showing these assumptions are true as to most or even any of the EDD 
cardholders’ experience.” Id. at 88. 
Levine’s method thus enjoys no general acceptance in the field of economic 
damage analysis, nor any general acceptance by the courts as a method of calculating 
actual damages as a matter of law. It is thus neither a reliable measure of Plaintiffs’ 
actual damages nor relevant to the Court in ruling on summary judgment, since it sets 
forth no material factual evidence about the actual damages suffered by any Plaintiff 
or the Plaintiff class as a whole. 
Defendant thus respectfully submits that Levine’s unreliable and irrelevant 
analysis should be stricken from the record. 
CONCLUSION 
For each and all of the foregoing reasons, Levine’s report and opinions should 
be excluded and stricken from the record. 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33495 
Page 24 of 27

 
21 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
Dated: October 17, 2025 
By: 
Respectfully submitted, 
 
/s/ Matthew L. Riffee 
 
MATTHEW L. RIFFEE (pro hac vice) 
MRiffee@goodwinlaw.com 
SABRINA M. ROSE-SMITH (pro hac vice)  
SRoseSmith@goodwinlaw.com 
KEITH LEVENBERG (pro hac vice) 
KLevenberg@goodwinlaw.com 
GOODWIN PROCTER LLP 
1900 N St. NW  
Washington, DC 20036  
Tel: +1 202 346 4000  
Fax: +1 202 346 4444 
JAMES W. MCGARRY (pro hac vice) 
JMcGarry@goodwinlaw.com 
GOODWIN PROCTER LLP 
100 Northern Avenue 
Boston, MA 02210 
Tel.: +1 617 570 1000 
Fax: + 1 617 523 1231 
LAURA G. BRYS (SBN 242100) 
LBrys@goodwinlaw.com 
GOODWIN PROCTER LLP 
601 S Figueroa St., Suite 4100 
Los Angeles, CA 90017 
Tel.: +1 213 426 2500 
Fax: +1 617 346 4444 
VALERIE A. HAGGANS (pro hac vice) 
VHaggans@goodwinlaw.com 
LINDSAY E. HOYLE (pro hac vice) 
LHoyle@goodwinlaw.com 
GOODWIN PROCTER LLP 
620 Eighth Avenue 
New York, NY 10018 
Tel: +1 212 813-8800 
Fax: +1 212 355-3333 
YVONNE W. CHAN (pro hac vice) 
YChan@jonesday.com 
JONES DAY 
100 High Street 
Boston, MA 02110 
Tel.: +1 617 960 3939 
Fax: +1 617 449 6999 
JANICE P. BROWN (SBN 114433) 
jbrown@myersnave.com 
MATTHEW B. NAZARETH (SBN 278405) 
mnazareth@myersnave.com 
MEYERS NAVE 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33496 
Page 25 of 27

 
22 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
600 B Street, Suite 1650 
San Diego, CA 92101 
Attorneys for Defendant 
BANK OF AMERICA, N.A. 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33497 
Page 26 of 27

 
23 
BANA’S MEM. ISO MOT. TO EXCLUDE LEVINE  
 
CASE NO. 21-MD-02992-GPC-MSB 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
GOODWIN PROCTER LLP 
ATTORNEYS AT LAW 
CERTIFICATE OF SERVICE 
I hereby certify that I electronically filed the foregoing with the clerk of the 
court for the United States District Court for the Southern District of California by 
using the CM/ECF system on October 17, 2025. I further certify that all participants 
in the case are registered CM/ECF users and that service will be accomplished by the 
CM/ECF system. I certify under penalty of perjury that the foregoing is true and 
correct. 
 
Executed: October 17, 2025  /s/ Matthew L. Riffee 
 
 
Case 3:21-md-02992-GPC-MSB     Document 566-1     Filed 10/17/25     PageID.33498 
Page 27 of 27

File and source

File
gov.uscourts.casd.709615.566.1.pdf
Size
388,801 bytes
SHA-256
e1942ad5f7899acc0f4d4270eb29ceacd5f85ad84b3e592f1abaf87100c49ff4
Our copy
gov.uscourts.casd.709615.566.1.pdf
Original
PACER (login required)
Back to top