Reply to Response to Motion re 566 Motion to Exclude the Purported Expert — In re BofA Unemployment Litigation (Dkt. 681)

Record facts

U.S. District Court for the Southern District of California · No. 3:21-md-02992-GPC-MSB · Doc. 681 · 2026-04-17 · Docket on CourtListener

Full text

REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
REPLY BRIEF IN SUPPORT OF 
DEFENDANT BANK OF AMERICA, 
N.A.¶S MOTION TO EXCLUDE 
PURPORTED EXPERT OPINIONS OF 
DAVID I. LEVINE (ECF 566) 
Date: 
April 17, 2026 
Time: 
1:30 p.m. 
Ctrm: 
12A ± 12th Floor 
Judge: 
Hon. Gonzalo P. Curiel 
FILED PROVISIONALLY UNDER SEAL 
PURSUANT TO STIPULATED PROTECTIVE 
ORDER 
 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58986     Page
1 of 16

i 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
In re Apple iPhone Antitrust Litig., 
2022 WL 1284104 (N.D. Cal. Mar. 29, 2022) ................................................. 3, 4 
Briseno v. ConAgra Foods, Inc., 
844 F.3d 1121 (9th Cir. 2017) .............................................................................. 9 
Comcast Corp. v. Behrend, 
569 U.S. 27 (2013) ......................................................................................... 2, 10 
Daubert v. Merrell Dow Pharms., Inc.,, 
509 U.S., 579, 591 (1993) .................................................................................... 5 
In re Flash Memory Antitrust Litig., 
2010 WL 2332081 (N.D. Cal. June 9, 2010) ....................................................... 4 
Gen. Elec. Co. v. Joiner, 
522 U.S. 136 (1997) ............................................................................................. 7 
Hartley v. Dillard¶s, Inc., 
310 F.3d 1054 (8th Cir. 2002) ........................................................................ 9, 10 
In re Lidoderm Antitrust Litig., 
2017 WL 679367 (N.D. Cal. Feb. 21, 2017) .................................................... 8, 9 
Medlock v. Taco Bell Corp., 
2015 WL 10791410 (E.D. Cal. Dec. 11, 2015) .................................................... 4 
Opperman v. Path, Inc., 
2016 WL 3844326 (N.D. Cal. July 15, 2016) ...................................................... 8 
In re Optical Disk Drive Antitrust Litig., 
303 F.R.D. 311 (N.D. Cal. 2014) ................................................................. 3, 4, 8 
In re Processed Egg Prods. Antitrust Litig., 
312 F.R.D. 124 (E.D. Pa. 2015) ........................................................................... 5 
Reed v. Advocate Health Care, 
268 F.R.D. 573 (N.D. Ill. 2009) ........................................................................... 5 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58987     Page
2 of 16

ii 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
Utne v. Home Depot USA, Inc., 
2022 WL 16857061 (N.D. Cal. Nov. 10, 2022) ................................................... 4 
Van v. LLR, Inc., 
962 F.3d 1160 (9th Cir. 2020) .............................................................................. 8 
Vaquero v. Ashley Furniture Indus., Inc., 
824 F.3d 1150 (9th Cir. 2016) ............................................................................ 10 
Statutes 
28 U.S.C. § 1332(d)(6) .............................................................................................. 3 
 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58988     Page
3 of 16

1 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
BANA moved to exclude David Levine¶s opinion on the damages supposedly 
suffered by the class for a very fundamental reason: Levine does not actually have an 
opinion on the damages (if any) suffered by the class. See ECF 566-1 (Mot.).1 What 
he has is a shot in the dark. He proposes to award the class damages based on the 
amount it would, hypothetically, have cost them to incur credit-card debt at a 15.9% 
interest rate²
 
, but because he imagines they were harmed by other things he did not (and 
could not) measure. So, as between credit-card interest rates he can measure and 
damages he cannot, Levine offers an opinion on the former. But that opinion does 
not even pretend to be a reliable way of measuring the latter.  
Levine concedes he considered no data or information on actual class member 
behavior or actual class member harms. He does not claim that his damages 
calculations apply to every class member, or to any class member²just some portion 
he imagines (without any empirical support) to constitute a ³majority.´ But he 
acknowledges myriad circumstances which could place some unknown and 
unknowable portion of the class outside that ³majority.´ And even within the 
imagined majority, Levine does not claim that his damages calculation is accurate or 
reliable. To the contrary, he proffers its very unreliability as a feature rather than a 
bug, because his guess is ³conservative.´ As Daubert requires a reliable method, not 
a conservative one, his unreliable opinions should be stricken from the record. 
ARGUMENT 
Plaintiffs¶ opposition brief undercuts their own purported expert right from 
their statement of the generic legal standard. They say, ³An expert¶s testimony should 
be admitted where Plaintiffs¶ damages model for assessing consequential 
 
1 BANA has also moved to exclude the opinions of Plaintiffs¶ other purported 
damages experts, Chloe East and Greg Regan. See ECF 565, 567. All three opinions 
suffer from the same basic defect: offering opinions based on interest rates not 
³supported . . . with evidence showing that the[ir] assumptions are true as to most or 
even any´ class members. ECF 494 at 88. BANA thus incorporates the arguments 
made in its other Motions and Replies as equally applicable here. 
-
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58989     Page
4 of 16

2 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
damages . . . is attributable to their theory of liability and can be calculated for the 
class as a whole.´ ECF 617 (Opp.) at 6 (citing Comcast Corp. v. Behrend, 569 U.S. 
27, 35 (2013)). But Levine already admitted that his damages model does not apply 
to ³the class as a whole´²only to his imagined ³majority.´ See HX 36 39:2-6. And 
Plaintiffs¶ citation to Comcast is also self-defeating. Plaintiffs argue that ³[c]ourts 
routinely reject attempts to exclude expert testimony on the ground that the testimony 
is µspeculative,¶´ (Opp. 11), yet Comcast expressly rejects Plaintiffs¶ logic: 
The Court of Appeals simply concluded that respondents ³provided a method 
to measure and quantify damages on a class-wide basis,´ finding it 
unnecessary to decide ³whether the methodology was a just and reasonable 
inference or speculative.´ Under that logic, at the class-certification stage any 
method of measurement is acceptable so long as it can be applied classwide, 
no matter how arbitrary the measurements may be. Such a proposition would 
reduce Rule 23(b)(3)¶s predominance requirement to a nullity. 
569 U.S. at 35-36 (citation and brackets omitted). 
That describes the problem here. There is no disagreement that Levine¶s 
damages formula ³can be applied classwide.´ Id. (emphasis added). After all, it is 
barely even a formula at all²he just proposes taking every class member¶s claim 
amount and multiplying it by 15.9%. So, the arithmetic ³can´ of course be done. The 
dispute here is not about whether multiplication tables exist, but whether this two-
factor multiplication problem actually corresponds to reality²whether class 
members actually suffered damages equal to the products of Levine¶s equations. 
Plaintiffs do not even try to argue that they have, and instead argue that they 
should not be obliged to make that showing at all. They claim it suffices for Levine 
to estimate the ³average´ damages suffered across the class rather than the ³amount 
of damages suffered by each individual class member,´ because it is ³permissible´ 
to base a damages model on ³classwide (or µaggregate¶) approximations´ and 
³average[s].´ Opp. 4, 12, 15. This argument is faulty on multiple fronts. As a 
threshold matter, Plaintiffs cannot defend Levine¶s method as a mere exercise in 
averaging or aggregating, because that is not responsive to the objection that his 
figures are untethered to any harms actually suffered. Separately, aggregate damages 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58990     Page
5 of 16

3 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
calculations do not carry the Comcast burden of producing a reliable classwide 
damages methodology where the harms suffered (if at all) by individual class 
members vary and no method is proposed to account for those variations. 
An ³aggregate´ damages figure is, by definition, merely a sum of individual 
damages figures. See, e.g., 28 U.S.C. § 1332(d)(6) (amount in controversy ³[i]n any 
class action´ is the ³aggregate[] of ³the claims of the individual class members´). If 
the individual damages figures that comprise the aggregate are not reliable, then the 
aggregate is not reliable either. So, for Levine¶s opinion to clear the Daubert 
threshold on the ground that it measures aggregate damages, Plaintiffs must first 
show that the figures being aggregated are themselves reliable.  
As already shown, however, Levine, by his own admission, has no opinion on 
that question. He does not know if any of the class representatives suffered damages 
in the amounts yielded by his method. See Mot. 5, 14-15 (citing testimony). He does 
not know how many (if any) class members suffered damages in the amounts yielded 
by his methods. See id. He does not believe his methods are even capable of 
measuring the damages experienced by any individual class member. See id. 14-15. 
This is fatal. See, e.g., In re Optical Disk Drive Antitrust Litig., 303 F.R.D. 311, 321 
(N.D. Cal. 2014) (ruling that ³plaintiffs fail to show the expert reports answer the 
critical questions´ in ³calculating damages´ because the expert proposed to calculate 
damages ³in the aggregate´ without ³attempt[ing] to show that all or nearly all [class 
members] were [damaged] in that amount, or in any amount at all´); In re Apple 
iPhone Antitrust Litig., 2022 WL 1284104, *6-7 (N.D. Cal. Mar. 29, 2022) (striking 
damages methodology where model did not reliably measure individual damages of 
class members and let to ³absurd results´). 
Levine¶s opinion is being proffered to carry Plaintiffs¶ Comcast burden with 
opinions on the cost of credit-card borrowing, but this Court already held that this 
method ³does not support a damages model that satisfies Comcast´ because it ³is 
based on assumptions rather than evidence establishing each cardholder¶s 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58991     Page
6 of 16

4 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
experience,´ and the assumption that ³impacted cardholders would likely have 
needed´ to ³utiliz[e] credit cards´ was not ³supported [] with evidence showing that 
these assumptions are true as to most or even any of the EDD cardholders[].´ ECF 
494 at 87-88. Levine, likewise, does not supply evidence that this assumption is ³true 
as to most or even any´ class members. He does not base his proposal to measure 
damages by credit-card interest rates on any belief (much less data) 
 
. Rather, his expressly stated position is 
that he does not know if they did, but he believes they were damaged in some other 
way that he cannot measure. See Mot. 11-12 (quoting testimony). 
Unable to deny this, Plaintiffs maintain that ³Levine¶s testimony need not be 
based on each class member¶s individual financial circumstances to be admissible.´ 
Opp. 11. But Plaintiffs do need to demonstrate that Levine¶s assumptions are valid 
on the individual level in order to proffer them as valid in the aggregate. See, e.g., In 
re Optical Disk Drive, 303 F.R.D. at 321; In re Apple iPhone, 2022 WL 1284104, 
*16 (³plaintiffs may rely on aggregate damage estimates, but must also establish that 
there is a method, common across the class, for arriving at individual damages´) 
(internal quotation omitted); Utne v. Home Depot USA, Inc., 2022 WL 16857061, *5, 
*7 (N.D. Cal. Nov. 10, 2022) (granting motion to strike ³aggregate´ damage 
calculations because ³neither model here purports to derive a total award from a sum 
of individual class members¶ awards´ and ³[w]ere the jury to return an aggregate 
award, it would µnot be possible to know which [class members] are entitled to share 
in the award,¶ or how much each should receive´); Medlock v. Taco Bell Corp., 2015 
WL 10791410, *6 (E.D. Cal. Dec. 11, 2015) (rejecting damages methodology based 
on ³aggregate [] figures . . . instead of the actual rates´ applicable to the class 
members); In re Flash Memory Antitrust Litig., 2010 WL 2332081, *10 (N.D. Cal. 
June 9, 2010) (rejecting methodology that ³look[s] only at an average price trend,´ 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58992     Page
7 of 16

5 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
ignoring ³individual variances´).2
Recognizing these problems, a few pages later in the same brief, Plaintiffs 
insist that ³Levine has explicitly reckoned with potential variations in the time value 
of money to the particular population at issue.´ Opp. 15 (emphasis added). The 
citation Plaintiffs supply in support of this proposition is to Paragraphs 9 to 11 of 
Levine¶s Report. The Court can scour them in search of any ³reckon[ing] with 
potential variations´ among class members and find no trace of it: 
9. Using an aggregate measure of harm is common practice in the field of 
economics, and aggregates are an appropriate way to represent classwide 
harm. 
10. The harm suffered by class members who were deprived of their UI 
benefits cannot be represented only by the principal amount of the benefits 
they were denied. UI recipients who were denied their benefits also suffered 
harm caused by the delay itself, which can be described as the ³opportunity 
cost´ faced by individuals in the absence of UI. 
11. In my opinion, a compound interest rate is an appropriate way of 
measuring this opportunity cost. Specifically, a compound interest rate, 
applied to the principal amount of delayed funds and the length of time those 
funds were inaccessible, is an appropriate way to measure the harm caused by 
the Bank¶s policies and practices. 
HX 35 ¶ 9-11. The ³potential´ for ³variations´ is not even mentioned. Levine is very 
expressly attempting to do the opposite: far from ³reckon[ing] with potential 
variations,´ he is trying to proffer a justification for avoiding that reckoning.  
But the proffered justification does not help Plaintiffs. It may in fact be 
³common practice in the field of economics´ to measure harms in the ³aggregate.´ 
But that says nothing about whether the measure is a reliable one. And it likewise 
says nothing about its ³fit´ to the case, since ³scientific validity for one purpose is 
not necessarily scientific validity for other, unrelated purposes.´ Daubert v. Merrell 
Dow Pharms., Inc., 509 U.S., 579, 591 (1993). Even if a competent economist 
proffers a reliable measure of harm on an aggregate basis, that does not mean it 
 
2 Accord, e.g., Reed v. Advocate Health Care, 268 F.R.D. 573, 591 (N.D. Ill. 2009) 
(collecting cases rejecting ³reliance on averages´ absent evidence ³all members of 
the proposed class suffered´ the same harms); In re Processed Egg Prods. Antitrust 
Litig., 312 F.R.D. 124, 159 (E.D. Pa. 2015) (³The case law understandably allows 
for averages and aggregations, but only if the court is convinced that the averages 
and aggregations are not masking individualized issues.´). 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58993     Page
8 of 16

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 P ROCTER LLP 
ATmR?\'£\S ATLA\\' 
suffices to carry Plaintiffs' Comcast burden. 
But Levine's proposed "aggregate" rests on nothing reliable. The above-cited 
reference to the pablum in Paragraphs 9 to 11 of his report is not the only effort by 
Plaintiffs to prop up its reliability that falls flat on its own terms. For example, 
Plaintiffs assert that "Levine analyzed substantial data in formulating his expert 
opinion on an appropriate interest rate that reflects the 'time value' of the UI benefits 
that were delayed or denied." Opp. 7. But in fact, Levine analyzed almost no such 
data-as Plaintiffs admit later on in their brief, when they are defending Levine 
against the charge of failing to consider "sufficient facts or data," and instead of 
describing the "facts or data" he considered, they merely argue that "experts may rely 
on 'knowledge and experiences that are not necessarily in the record.' ... This is 
exactly what Dr. Levine has done." Id. 17. When Plaintiffs refer to Levine 
considering "data," all they really mean is that he considered general interest-rate 
data. He did not consider any data, documents, or information from the class 
representatives or the class members to see whether that interest-rate data was 
applicable to them in any way. See Mot. 5, 7, 17-19. 
The result is that his opinions on the likelihood class members turned to credit-
card borrowing are entirely without empirical foundation. Plaintiffs quote Levine's 
opinion "that it is unlikely that 'the typical class member . . . could fall back on 
savings to cover expenses the whole time their UI benefits were denied."' Opp. 4. 
But Levine considered no data or information pertinent to how likely this is: 
. 
• 
• 
• 
• 
, ~~-
havenot 
) 
expert's 
stated beli_ef t a~ they are not like y to be "typical" of the class. See, e.g., 
HX 36 95.25-97.20.J 
• He considered no data on how long the "whole time" actually was in which 
th~ payments were delayed, other than r~lyi11:g on how "Plamtiffs' counsel 
month" masks the act that nearly 30% were not, and were paid within 30 
6 
REPLY ISO BANA's MOT. TO EXCLUDE LEVINE 
CASE No. 3:21-MD-02992-GPC-MSB 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58994     Page
9 of 16

7 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
days of the dispute²a variation Levine does not account for in any way. 
See ECF 589-1 (MSJ) at 23 n.4. 
He considered no data on any ³borrowing´ activity by class members and 
no data on any ³other costs´ class members occurred. His opinion on how 
³likely´ class members were to incur these costs is a classic opinion 
³connected to existing data only by the ipse dixit of the expert.´ Gen. Elec. 
Co. v. Joiner, 522 U.S. 136, 146 (1997). 
He considered no data on the amount of each class member¶s claim. As 
economist Victor Stango pointed out (and Plaintiffs do not dispute), in 
assessing whether a consumer would need to borrow funds to cover a 
temporarily disputed claim amount, it is highly relevant whether the claim 
amount is a small one or a large one. See HX 40 ¶ 52 (It is also relevant 
how much they had in savings or other accessible liquid funds, (see id. 
¶ 48), on which Levine also considered no data.). 
Thus, it is unsurprising that while Plaintiffs assert that Levine considered 
³information concerning this case and the class,´ (Opp. 7), they cannot actually 
identify any information ³concerning . . . the class´ that Levine considered. 
Plaintiffs¶ Opposition also relies on several misstatements (and extreme 
exaggerations) as to the nature of BANA¶s challenge. They cite cases for the 
proposition that it is permissible to base a damages methodology on ³approximated 
value[s].´ Opp. 13 n.4. But nothing in BANA¶s Motion challenged the use of 
approximations. Rather, BANA challenged the assumption that Levine¶s 15.9% 
interest rate (or median-wage figure) is a reliable and supported approximation of 
anything. Mot. 1, 5-7. Plaintiffs also accuse BANA of ³argu[ing] that Dr. Levine 
does not precisely measure each class members¶ individual damages,´ then set up a 
straw man in attributing to BANA an ³insistence that the only permissible means of 
calculating damages requires Plaintiffs to identify with complete certainty the exact 
quantum of each individual¶s damages.´ Opp. 14, 16. But BANA has not faulted 
Levine for a failing to ³precisely measure´ individual damages, nor has it suggested 
³the exact quantum´ must be established ³with complete certainty.´ Rather, BANA¶s 
challenge is quite nearly the opposite: that Levine¶s method does involve attaching a 
precise value to every class member¶s individual damages, but does so without 
³evidence showing [his] assumptions are true as to most or even any of the EDD 
cardholders´ (ECF 494 at 88)²and without considering the varying circumstances 
• 
• 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58995     Page
10 of 16

8 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 leave him admitting that his method would ³overcompensate some class 
members while undercompensating others.´ Opperman v. Path, Inc., 2016 WL 
3844326, *14 (N.D. Cal. July 15, 2016); Mot. 16-17. This is not a criticism that 
Levine¶s figures merely lack ³complete certainty.´ Opp. 16. It is a criticism that they 
lack evidentiary foundation altogether. 
Plaintiffs, therefore, fall back on arguing that Levine¶s method does not need 
any evidentiary foundation. They cite Van v. LLR, Inc., 962 F.3d 1160 (9th Cir. 
2020), for its proposition that ³[i]nterest is simply a way of measuring [the] injury, 
not the injury itself.´ Opp. 12 n.3 (quoting Van, 962 F.3d at 1165). In other words, 
Plaintiffs are arguing that they do not have to evidence any connection between the 
credit-card interest rate and any class member injuries, because they are not claiming 
credit-card interest as ³the injury,´ only using the interest rate as a proxy for some 
other unspecified injury. But that merely begs the question whether it is a reliable 
measure of that other injury²and Levine makes no claim that it is (indeed, he 
expressly disclaims any connection between the two). See Mot. 11-12; In re Optical 
Disk, 303 F.R.D. at 321 (rejecting damages analysis because it ³assumes the very 
proposition that the [plaintiffs] are now offering it . . . to show´). Nothing in Van 
supports the notion that damages for lost ³time value of money´ can be measured by 
the cost of borrowing on credit cards. See Mot. 20. This Court has already held that 
damages based on the cost of borrowing on credit cards cannot be awarded without 
evidence that the assumptions about ³utilization of credit cards . . . are true.´ ECF 
494 at 88. Plaintiffs¶ concession that they cannot evidence this is dispositive. 
Notably, Levine is not actually proposing an ³aggregate´ damages 
methodology in the sense contemplated by the case law. An example of an admissible 
aggregate damages methodology appears in In re Lidoderm Antitrust Litig., 2017 WL 
679367 (N.D. Cal. Feb. 21, 2017), an antitrust case alleging ³inflated costs for [a] 
brand name and generic´ drug. Id. *1. The court there approved an expert¶s proposal 
to ³take[] an aggregate approach to damages´ ³based on aggregated purchases²
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58996     Page
11 of 16

9 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
combining brand only, generic and brand, and then generic only purchases to create 
aggregating purchasing figures²and then estimating . . . a µbut-for¶ price´ based on 
the aggregate data. Id. *10. The key fact, as that passage indicates, is that the 
aggregate damages model was still based on actual data about individual class 
members¶ purchases, fed into an aggregate on which further analysis was performed. 
Given this, the court (appropriately) questioned whether individual damages could 
be reliably deduced from the aggregate, and was satisfied that they could be. See id. 
*11 (concluding ³damages can be apportioned . . . using[] a pro rata formula based 
on each [class member¶s] purchase[s]´). Here, Levine¶s proposal to measure the 
aggregate cost of credit-card borrowing is not based on any data about class 
members¶ actual borrowing, so there is no data that could be used to deduce a 
distribution rooted in any class member¶s actual harms (if any). This is no small 
distinction. It is the difference between aggregate damages and random numbers. 
Plaintiffs contend there is ³longstanding caselaw holding that expert 
opinions²particularly the opinions of economics experts²are admissible even 
when based on µgeneralizations¶ rather than . . . [the] damages suffered by each 
individual class member.´ Opp. 12. Nothing in the cases they cite endorses the notion 
that Plaintiffs are entitled to rely on generalizations with no evidence that those 
generalizations ³are true as to most or even any of the EDD cardholders,´ as this 
Court (correctly) required. ECF 494 at 88. The only issue presented in Briseno v. 
ConAgra Foods, Inc., 844 F.3d 1121 (9th Cir. 2017), was ³whether, to obtain class 
certification under Federal Rule of Civil Procedure 23, class representatives must 
demonstrate that there is an µadministratively feasible¶ means of identifying absent 
class members.´ Id. at 1123. That issue is not presented in this case, and it has nothing 
to do with the use of ³generalizations´ in measuring damages on any basis.  
The Eighth Circuit¶s ruling in Hartley v. Dillard¶s, Inc., 310 F.3d 1054 (8th 
Cir. 2002), is even further afield. Proffering it to support the notion that Levine need 
not account for ³damages suffered by . . . individual class member[s]´ (Opp. 12) 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58997     Page
12 of 16

10
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
raises serious doubts about whether Plaintiffs have even read the case. It was not a 
class action at all, but an individual employment discrimination case. The expert in 
question was not opining on damages; rather, the expert opined on whether the 
plaintiff¶s job termination could be ³justif[ied]´ by ³declining profits´ at the mall 
where she worked, or whether the claim of declining profits was just a pretext for 
discrimination. Hartly, 310 F.3d at 1058, 1060. Plaintiffs cite it for its ruling that it 
was appropriate to consider the expert¶s opinion that ³that the financial problems of 
the McCain Mall store were consistent with what was happening to department stores 
in malls around the country.´ Id. at 1060. Plaintiffs inexplicably mischaracterize this 
as holding that it is permissible to base a damages methodology on ³generalizations 
about national trends,´ (Opp. 13), but the expert was not proffering a damages 
methodology at all. The case says nothing whatsoever about the circumstances in 
which damages might be measured in an aggregate or ³general´ fashion. 
The Ninth Circuit has ³interpreted Comcast to mean that plaintiffs must be 
able to show that their damages stemmed from the defendant¶s actions that created 
the legal liability.´ Vaquero v. Ashley Furniture Indus., Inc., 824 F.3d 1150, 1154 
(9th Cir. 2016) (internal quotation marks and citation omitted); accord Comcast, 569 
U.S. at 35 (³a model purporting to serve as evidence of damages in [a] class action 
must measure only those damages attributable to th[e] theory´ of liability). Plaintiffs¶ 
liability case seeks compensation for ³the loss of use of money.´ ECF 494 at 86. 
Levine proposes to measure damages for that alleged lost use of money by credit-
card interest rates, but he does not, and cannot, claim that the payment of credit-card 
interest is something that ³stemmed from the defendant¶s actions.´ Vaquero, 824 
F.3d at 1154. There is thus a fatal mismatch between the theory of liability and the 
damages model Levine offers. It should be stricken as unreliable and irrelevant. 
CONCLUSION 
For these reasons and those set forth in BANA¶s Motion and evident from the 
record, BANA Levine¶s report and opinions should be stricken and excluded. 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58998     Page
13 of 16

11
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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:   February 20, 2026 
Respectfully submitted, 
 
 
 
 
By: 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 Street 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 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.58999     Page
14 of 16

12
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 
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 
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 681     Filed 02/20/26     PageID.59000     Page
15 of 16

13 
REPLY ISO BANA¶S MOT. TO EXCLUDE LEVINE 
CASE NO. 3:21-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 February 20, 2026.  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. 
 
 
Dated:   February 20, 2026 
 
s/ Matthew L. Riffee 
 
 
MATTHEW L. RIFFEE 
 
Case 3:21-md-02992-GPC-MSB     Document 681     Filed 02/20/26     PageID.59001     Page
16 of 16

File and source

File

gov.uscourts.casd.709615.681.0.pdf

Size

1,724,217 bytes

SHA-256

3aea88403599435f1bec410e6f87540b572226db12e87e8423552fe377aec7b7

Our copy

gov.uscourts.casd.709615.681.0.pdf

Original

PACER (login required)