Court filing

DX 14.F.1 - Expert Report of Chloe N. East — In re Bank of America California Unemployment Benefits Litigation (Dkt. 594-1, S.D. Cal. No. 3:21-md-02992)

Filed October 23, 2025 in In re Bank of America California Unemployment Benefits Litigation; one of 1415 filings from this case.

Record facts

Court	U.S. District Court for the Southern District of California
Filed	2025-10-23

U.S. District Court for the Southern District of California · No. 3:21-md-02992-GPC-MSB · Doc. 594-1 · 2025-10-23 · Docket on CourtListener

Full text

DX 14.F.1
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43215
Page 1 of 27

Expert Report of Chloe N. East

UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF CALIFORNIA
SAN DIEGO DIVISION

In re Bank of America California
Unemployment Benefits Litigation,

This Document Relates to All Actions

Case No. 3-21-md-02992-GPC-MSB

EXPERT REPORT OF
CHLOE N. EAST

March 4, 2025

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43216
Page 2 of 27

Expert Report of Chloe N. East

Table of Contents
I.
Introduction ............................................................................................................................ 1
A. Qualifications..................................................................................................................... 1
B.
Summary of Assignment .................................................................................................. 1
C. Summary of Opinions ....................................................................................................... 2
II. Importance of Unemployment Insurance (“UI”) Benefits and Impacts of Denying
Beneficiaries Access to Their UI Funds ....................................................................................... 3
A. Importance of UI Benefits ................................................................................................ 3
B.
Impacts of Denying Beneficiaries Access to Their UI Funds ........................................ 9
III.
Appropriate Measures of Damages for Class Members Who Were Deprived of Access
to UI Funds by the Bank’s Actions and Inactions ..................................................................... 11
A. Costs of Borrowing ......................................................................................................... 11
B.
Costs of Cutting Spending .............................................................................................. 12
C. Conclusion ....................................................................................................................... 15
IV.
Appendices ........................................................................................................................ 16
Appendix A: CV ...................................................................................................................... 16
Appendix B: Materials Considered ....................................................................................... 23
Case Materials ....................................................................................................................... 23
Research ................................................................................................................................ 23
Data ....................................................................................................................................... 24

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43217
Page 3 of 27

Expert Report of Chloe N. East

I.
Introduction
A. Qualifications
1. I am an Associate Professor of Economics at the University of Colorado Denver. I
received a B.S. in Economics from the University of Texas at Austin in 2011, an M.A. in
Economics from the University of California, Davis in 2012, and a Ph.D. in Economics from the
University of California, Davis in 2016.
2. I founded and run the Equitable Policy Research Lab, which disseminates evidence-based
recommendations to advocates and policy-makers and offers training opportunities to individuals
from under-represented groups in economics. I am a Research Associate at the Institute for
Research on Poverty at the University of Wisconsin, a Non-Resident Fellow at the Brookings
Institution, a Faculty Research Fellow at NBER, and a Research Fellow at IZA.
3. My research has been published in top economics journals such as the American
Economic Review, the Journal of Labor Economics, the Journal of Public Economics, and the
Journal of Human Resources. My research has also been featured in major media news outlets
such as CNN, NPR, the New York Times, the Washington Post and others. I teach courses to
undergraduate and graduate students that cover U.S. public programs, causal inference and data
analysis tools.
4. I am a nationally recognized expert on U.S. safety net and social insurance programs
including Unemployment Insurance (“UI”). I have done research on UI since 2012 and have
published multiple papers studying the effects of UI as a cushion for people after they lose their
job. My CV is attached as Appendix A.
B. Summary of Assignment
5. Plaintiff's counsel has retained me in the case of In re Bank of America California
Unemployment Benefits Litigation, Case No. 3-21-md-02992-GPC-MSB. I was requested to
provide expert opinions on: 1) the vulnerability of UI recipients and the importance of UI
benefits in mitigating the harms that would otherwise be suffered by those who lose their jobs
through no fault of their own, 2) the impacts of denying UI beneficiaries access to their UI funds,
and 3) an appropriate compound interest rate that can be used under a common methodology to
calculate the consequential damages, or lost value to class members of their UI benefits, resulting
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43218
Page 4 of 27

Expert Report of Chloe N. East

from their inability to access their UI benefits at the time they were due as a result of the Bank’s
challenged policies and practices.
6. I am being compensated at a rate of $400 per hour, plus expenses, for my services.
7. In preparing this report, I have reviewed and analyzed several reports and declarations
from Plaintiffs’ expert Greg Regan and from Bank of America’s expert Dr. Victor Stango. I have
also reviewed the labor economics literature on UI and on borrowing while unemployed. I have
relied on both in forming my opinions as laid out in this report. The sources I relied on are
contained in footnotes and additional sources listed in Appendix B.
C. Summary of Opinions
8. UI Benefits Provide a Critical Safety Net for Individuals Experiencing Involuntary Job
Loss
a. UI recipients are a highly vulnerable and financially precarious group. UI is a
critical safety net program provided by the government to ensure that individuals
who lose their job can afford food, health care, transportation, housing, and other
life necessities.
9. The Typical UI Recipient Does Not Have Enough Savings to Survive Abruptly Losing
Access to UI Benefits Without Incurring Costs Elsewhere
a. In my expert opinion, the evidence is clear that the typical UI recipient does not
have enough savings to handle abruptly losing access to UI benefits, without
either increasing their borrowing, cutting their consumption, or both.
b. In my opinion, while there may be a few households with enough savings to cover
this temporary loss of UI benefits, these households are outliers and not
representative of the experience of the median UI recipient.
10. The Credit Card Interest Rate is a Minimum Estimate of the Costs Faced by UI
Recipients Who Lost Access to Their Benefits
a. Given the low levels of savings among UI recipients, many UI recipients are
dependent on their UI benefits to pay for basic life necessities; thus, the value of
receiving UI benefits when they are due is far greater than the dollar value of
those benefits if paid at some later point in time.
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43219
Page 5 of 27

Expert Report of Chloe N. East

b. Most UI recipients who lose or are denied access to their UI benefits will need to
secure an alternative source of replacement funds and thus will turn to borrowing.
One of the most common methods of borrowing for UI recipients is credit cards.
c. Those who cannot borrow on credit cards will borrow using even more expensive
methods, will cut consumption, or both.
d. Therefore, UI recipients who were denied access to their UI benefits as a result of
the Bank’s challenged policies and practices were denied the full value of their UI
benefits, even if access to their benefits was later restored. The cost to UI
recipients who lost access to their UI benefits is at least as high as the interest
rates faced when borrowing on a credit card.
e. The average credit card interest rate is a conservative estimate of the costs faced
by the class members who were denied access to their UI benefits when due and
thus denied the full value of those benefits, since UI recipients almost all have
relatively low credit scores and so are likely to face high interest rates when
borrowing.
f. In my opinion, the average credit card interest rate of 20.8% is an appropriate
figure to use to calculate the cost to class members resulting from denial of access
to their principal claim amounts and frozen account balances.1
II.
Importance of UI Benefits and Impacts of Denying Beneficiaries Access to Their
UI Funds
A. Importance of UI Benefits
11. Many people experience involuntary job loss through no fault of their own, and UI
provides a critical safety net for those individuals and their families.
12. UI was especially important during the COVID-19 pandemic. During the first few
months of the pandemic, an estimated 22 million people (13% of the workforce) in the United
States lost their jobs.2

1 Plaintiffs’ counsel has informed me that another expert will be proposing a methodology for calculating
consequential damages that advocates for use of a compound interest rate.
2 Jeffry Bartash, “The U.S. has only regained 42% of the 22 million jobs lost in the pandemic. Here’s where they
are,” MarketWatch (August 7, 2020) (“The economy shed a record 22.2 million jobs in March and April after large
swaths of the U.S. were shut down to curb the spread of Covid-19, according [sic] the Labor Department’s survey of
business establishments.”).
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43220
Page 6 of 27

Expert Report of Chloe N. East

13. For decades, economists have been very clear about why the U.S. should have publicly
provided unemployment insurance: to help people maintain their spending at the same amount it
was pre-job-loss in the event they lose their job.3 In other words, UI is provided by the
government to ensure that individuals who lose their job can afford food, health care,
transportation, and can continue to stay current on bills, rent, and/or their mortgage. Specifically,
economists assert that if people were able to use their own personal savings to maintain spending
at the same amount it was pre-job-loss, then publicly provided UI benefits would not be
necessary.4 However, that is not the case. Economists have shown for decades that low levels of
private savings among people who lose their jobs demonstrates that those individuals cannot use
private savings to cover their essential expenses during unemployment.5
14. An alternative method, beyond savings, that households can use to maintain their
spending when a principal income-earner loses their job, is to borrow against current assets or
future earnings. If the credit market is complete and perfect, borrowing might be a satisfactory
alternative to publicly provided unemployment insurance. For the market to be complete and
perfect, though, several conditions must be satisfied: 1) the lender must be able to observe each
borrower’s probability of repaying their loan; 2) every potential borrower must have access to
borrowing, and the price of borrowing they face (interest rates) must reflect their likelihood of
paying back the loan; and 3) the lenders must make no profits. In reality, it is impossible to
perfectly observe each potential borrower’s likelihood of paying back the loan and lenders may
assume that only those who have lower ability to pay back loans will apply for loans, so lenders
charge higher prices and deny some loan applications to account for this (and do make profits).6
Because of these economic realities, economists conclude the credit market is not complete and

3 Jonathan Gruber, “The Consumption Smoothing Benefits of Unemployment Insurance,” 87 The American
Economic Review 192, 192 (1997); Raj Chetty, “Moral Hazard versus Liquidity and Optimal Unemployment
Insurance,” 116 Journal of Political Economy 173, 174 (2008); Kory Kroft and Matthew J. Notowidigdo, "Should
Unemployment Insurance Vary With the Unemployment Rate Theory and Evidence,” 83 The Review of Economic
Studies 1092, 1094 (2016).
4 Gruber, supra note 3 at 192.
5 Id.; Chetty, supra note 3 at 174; Jesse Rothstein and Robert G. Valletta, “Scraping by: Income and Program
Participation After the Loss of Extended Unemployment Benefits,” 36 Journal of Policy Analysis and Management
880, 884 (2017).
6 Stephen G. Cecchetti and Kermit L. Schoenholtz, "Adverse Selection: A Primer,” Money and Banking 1, 3 (2017)
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43221
Page 7 of 27

Expert Report of Chloe N. East

perfect and that there are constraints on borrowing, with the result that publicly provided UI is
necessary to correct for this market failure.7
15. Economists who study the low savings levels of unemployed people and the importance
of UI in helping them make ends meet often focus on the experience of a representative person
within a large and representative data set. In general, empirical analysis in economics
emphasizes the importance of using large, representative samples, and recognizes that accurate
conclusions may be drawn from studying a sample similar to the population of interest. This
approach to economic data analysis has been the norm throughout my experience in academia—
especially when studying large populations and the impact of policies like UI.
16. In my experience, it is not appropriate to predict the effects on the mean or median
person based on results from extreme values. Drawing conclusions based on a small sample or a
selected sample, such as one drawn only from one set of extreme values, does not lead to
generalizable conclusions. Economists do the opposite—we look to aggregate data, including the
mean or median, and draw conclusions about the average experience. As such, the economics
literature I have reviewed in my time studying UI and analyzing why UI benefits are important,
frequently relies on the well-established fact that the median UI unemployed person who
receives UI benefits has low savings available to compensate for the loss of income. Economists
focus primarily on the median in conducting these analyses because the distribution of savings
among UI recipients, like the population in general, is skewed instead of evenly distributed,
which is explained in more detail below. Examples from the literature I have reviewed include:
“The median 25-64-year-old worker has gross financial assets equivalent to less than three
weeks of income, and the average unemployment spell for those becoming unemployed lasts
approximately 13.1 weeks.”8
“The median checking account balance in the JPMCI sample is $1250…”9

7 Gruber, supra note 3 at 195-97.
8 Gruber, supra note 3 at 192 n.2.
9 Peter Ganong and Pascal Noel, “Consumer Spending during Unemployment: Positive and Normative
Implications,” 109 American Economic Review 2383, 2392 (2019).
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43222
Page 8 of 27

Expert Report of Chloe N. East

“[M]edian liquid wealth net of unsecured debt is only $128, suggesting that many
unemployed individuals may not be in a position to smooth consumption while
unemployed.”10
“We list median values because the means are heavily influenced by high values in the long
tail of the distribution of household wealth,” and “[t]hus, median liquid financial wealth is
not far above zero and is only equal to about one-half of monthly household income.”11
Note that different papers report different nominal dollar values, depending on when each paper
was published, but even without adjusting them to all be in the same dollars, the findings are
strikingly similar in concluding that savings among the unemployed are low across all these
studies.
17. There are several potential measures of personal savings (also referred to as “liquidity”)
as highlighted above in the sampled papers. I focus in this report on one common measure that is
frequently used in the economics literature: the household’s net liquid wealth as a percentage of
their monthly income.12 This measure takes into account differences in household size,
household needs, and financial status. Net liquid wealth is the difference between liquid assets
and unsecured debt. Liquid assets include the value of jointly and individually owned checking
and savings accounts, money market accounts and funds, and certificates of deposit. Unsecured
debt includes the amount owed on credit cards and store bills, medical, and education debt.
18. I use the Survey of Income and Program Participation (SIPP) data to evaluate the savings
of UI recipients in California during 2020 and 2021.13 This is a common data set used in the
economics literature to understand the effects of job loss and the role of UI.14 With this data, I
document that UI recipients in California during the pandemic had low levels of savings and
were thus unable to maintain consumption at pre-job-loss levels (thus becoming unable to afford

10 Chetty, supra note 3 at 196.
11 Jesse Rothstein and Robert G. Valletta, “Scraping by: Income and Program Participation After the Loss of
Extended Unemployment Benefits,” 36 Journal of Policy Analysis and Management 880, 890-91 (2017).
12 The economics research that uses the same data I use here uses the same measure of liquidity constraint. Chetty,
supra note 3 at 195; Rothstein, supra note 11 at 889. Other economics research that does not have access to the
exact same information uses similar, but not exactly the same, measures of liquidity.
13 “Survey of Income and Program Participation,” U.S. Census Bureau, https://www.census.gov/programs-
surveys/sipp.html, accessed July 9, 2024.
14 See Chetty, supra note 3 at 196; Rothstein, supra note 11 at 885; Chloe N. East and David Simon, “The safety net
and job loss: How much insurance do public programs provide?,” 238 Journal of Public Economics 1, 3 (2024).
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43223
Page 9 of 27

Expert Report of Chloe N. East

to maintain spending on food, bills, housing, and other necessities) by relying on their own
savings. I measure household savings of UI recipients as the net wealth the household has in the
year before UI receipt as a percentage of the household’s monthly income, where household
monthly income is used to proxy for the household’s monthly expenditures. I measure all of this
before UI benefits are received, to avoid capturing the effect of UI itself. This is a common
approach in the literature to characterize the financial well-being of people who eventually
receive UI.15 I also adjust all dollar values in the SIPP to be in constant 2024 dollars. 44% of
those who receive UI had negative net wealth before they received UI—meaning that their
unsecured debt was more than they had in liquid savings. I recode these negative values to be 0
to be conservative. Finally, I “winsorize” the net wealth variables (a statistical process that
replaces the most extreme values in the data with representative values) to reduce the impact of
outliers on the conclusions.16
19. Within the sample of UI recipients who received benefits in California between 2020 and
2021 in the SIPP, the median UI recipient had 7.2% of monthly household income in net
household wealth. The average (mean) UI recipient had 219.7% of their monthly household
income in net household wealth. The difference between the median and mean illustrate that the
distribution is highly skewed and thus reliance on the mean values would put too much weight
on very few, extreme high values that are not representative of the typical UI recipient's wealth.
Indeed, in the literature, economists caution against reading too much into extreme and
uncommon high values of net wealth since they are not representative of most UI recipients, as
noted in the quotations above. The difference between the mean and median here also
demonstrate that most UI recipients had very little net wealth, and therefore savings and ability
to cover their expenses when they lose access to UI benefits, between 2020-2021.
20. To further understand the savings of UI recipients in California between 2020 and 2021, I
calculate that 67% of UI recipients have household net wealth under 100% of household monthly

15 Chetty, supra note 3 at 178; Rothstein, supra note 11 at 890.
16 Winsorizing is the process of assigning the most extreme observations with more representative values (this can
apply to low and high values). I winsorize at the 90th percentile, which means I assign the value at the 90th
percentile to all observations above the 90th percentile. The median net wealth is unchanged if I winsorize at the 10th
percentile or do not winsorize at all. Several papers studying the effect of UI, the authors note that they “winsorize”
the variables at a similarly high percentile in order for outliers to not have too much impact on their conclusions.
Ganong & Noel, supra note 9 at 2396; Peter Ganong, Fiona Greig, Pascal Noel, Daniel M. Sullivan, and Joseph
Vavra, “Spending and Job-Finding Impacts of Expanded Unemployment Benefits: Evidence from Administrative
Micro Data,” 114 American Economic Review 2898, Online Appendix A-4 (2024).
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43224
Page 10 of 27

Expert Report of Chloe N. East

income, and 79% have less than three months’ worth of income in savings, which is the
suggested amount households should have saved in the event of a job loss or other hardship.17
Taken together, this evidence points to the vast majority of UI recipients in California between
2020 and 2021 being unable to cover their expenses if they suddenly lose access to UI benefits.
21. Additionally, I have used data from the Federal Reserve Survey of Household Economics
and Decisionmaking (SHED) to calculate that 20% of UI recipients in the pandemic reported
being behind on educational expenses.18 Additionally, 21% reported that they expected to be
unable to pay or only make a partial payment that month on expenses such as rent or mortgage,
credit card bills, utility bills, phone or cable bills, car payments, or student loans. Another 22%
of UI recipients in 2020-2021 reported having unpaid medical debt. Thus, UI recipients are
already in a precarious financial situation even before losing UI benefits. Given the low levels of
savings and pre-existing debt among those receiving UI, when UI benefits are cut off from their
benefits, recipients have to either find some other way to finance their spending, like borrowing,
or cutting back on necessities, or some combination of the two.
22. The sample of UI recipients who received benefits in California between 2020 and 2021
had very similar net wealth as a percentage of their income as compared to a sample of UI
recipients across the country in this sample time period, and to a sample of UI recipients in the
entire country in 2019. The data shows that 33% of California UI recipients in 2020-2021, 29%
of UI recipients in the United States as a whole in 2020-2021, and 30% of UI recipients in the
United States as a whole in 2019, had a net worth of one month’s worth of their pre-job-loss
income. This suggests that low net wealth of UI recipients during the pandemic period was not
unusual. The nationwide and California-specific samples of UI recipients during the pandemic
were also similar in terms of their income levels, share of female head of households, household
size, and number of children (statistics produced using the SIPP). When using data from the
SHED, which does not identify state of residence, I instead use a national sample of UI

17 Vanguard, “Emergency Fund: What’s the right emergency fund amount?” (“To prepare for income shocks, many
experts suggest keeping enough money in your emergency fund to cover 3 to 6 months' worth of living expenses.”),
https://investor.vanguard.com/investor-resources-education/emergency-
fund#:~:text=To%20prepare%20for%20income%20shocks,%242%2C500%20to%20cover%20spending%20shocks.
18 “Survey of Household Economics and Decisionmaking,” Board of Governors of the Federal Reserve System,
https://www.federalreserve.gov/consumerscommunities/shed.htm, accessed September 26, 2024.
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43225
Page 11 of 27

Expert Report of Chloe N. East

recipients, which are plausibly generalizable to the population of UI recipients in California
during 2020-2021 because of the similarity of these samples.
23. Before the pandemic, California UI replaced about 50% of individual’s lost earnings.19
Due to temporary federal policies that made UI more generous, the replacement rate in several
months of the pandemic was higher, and for some it reached around 100% for a few months in
2020.20 For many who lost their jobs in the pre-pandemic period, UI was the only state or federal
government program that provided a meaningful amount of income support.21 During the
pandemic, those who got UI often received stimulus payments as well,22 but even with these
more generous UI payments and stimulus checks, household wealth remained low. I calculated
that even after receiving stimulus payments, most UI recipients in California still had less than
one month's worth of income in net household wealth (calculated in the SIPP). Thus, households
receiving UI in this period could not rely on stimulus checks or other government transfers to
cover their expenses without UI benefits.
B. Impacts of Denying Beneficiaries Access to Their UI Funds
24. Past research has shown the importance of UI in allowing households to cover essential
expenses in the face of unemployment.23 To conduct this research, social scientists took
advantage of changes in UI generosity to determine how much consumption and expenditures
changed when UI generosity changed. This research concluded that providing more generous UI
benefits leads to a smaller drop in consumption during unemployment. For example,
expenditures on groceries fall by 16% and expenditures on medical care fall by 14-15% when UI
benefits run out.24 Importantly, these expenditure changes follow a predictable reduction in UI

19 Alex Bell, Thomas J. Hedin, Geoffrey Schnorr, and Till von Wachter, “An Analysis of Unemployment Insurance
Claims in California During the COVID-19 Pandemic,” California Policy Lab 1, 6 (2020).
20 Peter Ganong, Pascal Noel, and Joseph Vavra, “US Unemployment Insurance Replacement Rates During the
Pandemic,” 191 Journal of Public Economics 1, 1 (2020).
21 East & Simon, supra note 14 at 7.
22 Michael Karpman and Gregory Acs, “Unemployment Insurance and Economic Impact Payments Associated with
Reduced Hardship Following CARES Act,” Urban Institute 1, 3 (2020).
23 Gruber, supra note 3 at 199; Chloe N. East and Elira Kuka, “Reexamining the consumption smoothing benefits of
Unemployment Insurance,” 132 Journal of Public Economics 32, 32 (2015); Peter Ganong and Pascal Noel,
“Consumer Spending during Unemployment: Positive and Normative Implications,” 109 American Economic
Review 2383, 2384 (2019).
24 Ganong & Noel, supra note 9 at 2399.
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43226
Page 12 of 27

Expert Report of Chloe N. East

benefits, and the drop in expenditures and consumption would likely be even larger in the face of
an unexpected reduction in UI.25 All of this points to UI being a crucial source of income during
unemployment that helps individuals pay for food, medical care, and other necessities. As
summarized by Gruber, “results [that show how consumption responds to UI generosity]
therefore decisively reject the notion that there are complete private consumption insurance
markets for unemployment spells,”26 meaning that without UI, households must turn to
expensive borrowing or else incur costs related to foregoing these necessities.
25. An economic study of the pandemic period also confirms that UI served the same
purpose in this period as well; when federal expansions to UI benefits expired in 2020 and 2021,
those who had previously received these expansions reduced their spending dramatically.27
26. Additionally, my own calculations using the Census Household Pulse data show that
about 69% of UI recipients in California in 2020 reported that they used the UI benefits to cover
their spending needs.28
27. Based on this research, it is my opinion that UI benefits provide a critical lifeline for
recipients, many of whom have little to no savings, and that delaying benefits deprives those
families of a critical income source. When UI benefits are delayed, it is my expert opinion that
those families are unlikely to have other sources of funds to pay their bills, buy food, and afford
medical care, so they are forced to incur debt or pay the high costs associated with cutting back
on those necessities.

25 Peter Ganong, Fiona Greig, Pascal Noel, Daniel M. Sullivan, and Joseph Vavra, “Spending and Job-Finding
Impacts of Expanded Unemployment Benefits: Evidence from Administrative Micro Data,” 114 American
Economic Review 2898, 2899 (2024).
26 Gruber, supra note 3 at 195.
27 Peter Ganong, Fiona Greig, Pascal Noel, Daniel M. Sullivan, and Joseph Vavra, “Spending and Job-Finding
Impacts of Expanded Unemployment Benefits: Evidence from Administrative Micro Data,” 114 American
Economic Review 2898, 2932 (2024).
28 “Household Pulse Survey Release,” U.S. Census Bureau, Phases 1-4.2 (April 23, 2020-September 16, 2024),
https://www.census.gov/programs-surveys/household-pulse-survey/data/datasets.html, accessed September 18,
2024.
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43227
Page 13 of 27

Expert Report of Chloe N. East

III.
Appropriate Measures of Damages for Class Members Who Were Deprived of
Access to UI Funds by the Bank’s Actions and Inactions
A. Costs of Borrowing
28. The value of UI benefits is not simply the dollar amount of the benefits, but the fact that
benefits are paid when a household is facing an unanticipated change in income. Thus, the
benefits are worth more than their dollar value because they provide insurance at a time the
affected households are particularly vulnerable. The adverse impacts of a sudden loss of access
to UI benefits cannot be fully rectified by later lump sum receipt of those lost benefits, because
to a UI recipient, the timing of the benefit payment matters. This is because households that
receive UI are already financially constrained, so the loss of UI benefits, even for a short period
of time, will cause further financial hardship that cannot easily be undone.
29. Based on my review of the literature, when UI benefits are abruptly cut off, households
will adjust their finances in a combination of two ways. First, households may turn to expensive
methods of borrowing, such as credit card utilization, in order to finance their expenses. Second,
households may cut back on these expenses and incur the significant costs associated with late
bill payments, food insecurity, etc. Households can also do a combination of these two activities.
I consider these possibilities, and the costs associated with each, next.
30. Many Americans are heavily reliant on credit cards to make ends meet29 and UI
recipients are no exception; 76% of UI recipients in California during the pandemic had
household credit card debt even before they received UI.30 Credit card borrowing is the most
common source of borrowing among people who receive UI, and UI recipients are more likely
than the general population to report that they pay off their credit card debt over time instead of
each month, meaning they accrue debt and interest costs.31 Furthermore, compared to all
households in 2020-2021, UI recipients were 52% more likely to be denied credit and 54% more

29 Khristopher J. Brooks, “Americans continue to rack up credit card debt, hitting a record $1.14 trillion,” CBS
News, August 6, 2023, https://www.cbsnews.com/news/credit-card-debt-total-us-2024/ (“Americans have
increasingly been relying on credit cards to make ends meet, with 6 in 10 adults, or 60%, using credit cards to buy
groceries in 2023, according to a May report by the Urban Institute.”).
30 Calculated using statistics on UI recipients from the SIPP.
31 Calculated using statistics from the Federal Reserve SHED data
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43228
Page 14 of 27

Expert Report of Chloe N. East

likely to be given less credit than they desired.32 Thus, people receiving UI may turn to even
more expensive forms of borrowing compared to credit cards, such as pawn shop loans or auto
title loans, to make ends meet. These alternative borrowing methods can have annual percentage
rates (APRs) up to 200% to 300%.33
31. Furthermore, unemployment causes an increase in default on debt and UI helps people
avoid defaulting on existing debt and ending up in a worse financial position in the longer
run.34 Defaulting on debt causes long-term harm because it leads to reduced credit scores, less
ability to borrow, and/or more expensive borrowing options in the future.35
B. Costs of Cutting Spending
32. Other households that are not able to turn to borrowing to finance their essential expenses
may be forced to cut back on necessities when they lose UI benefits. While some budget
tightening might be easy for households to do, two facts point to how costly this cutting back can
be. First, as I calculated in the SHED and described above, individuals who receive UI apply for
more credit but are denied, which indicates they would prefer not to cut consumption by as much
as they do. For example, 33% of those who receive UI and apply for more credit have their
request denied.36 According to economic theory, the value of UI benefits is higher the more
recipients must cut their consumption, so the cost of losing access to UI benefits will be higher if
people cannot borrow.37 Second, households cut consumption of important necessities, including
food and medical care, rather than just non-essential items, when they stop receiving UI or

32 Calculated using statistics from the Federal Reserve SHED data.
33 Bev O’Shea, “Should You Take a Pawnshop Loan?,” NerdWallet, March 1, 2024,
https://www.nerdwallet.com/article/loans/personal-loans/pawnshop-loans (“While payday loans and car title loans
can easily top 400% APR, pawnshop loans may be more affordable, with APRs around 200%.”); Marcie Geffner,
“Car title loans: 3 things to know before getting one,” CreditKarma, November 11, 2023,
https://www.creditkarma.com/personal-loans/i/car-title-
loans#:~:text=Car%20title%20loans%20have%20high%20fees%20and%20interest%20rates,-
With%20a%20car&text=This%20translates%20into%20an%20annual,total%20cost%20of%20the%20loan ("This
translates into an annual percentage rate, or APR, of more than 300%.”).
34 Niklas Flamang and Sreeraahul Kancherla, “Unemployment Insurance as a Financial Stabilizer: Evidence from
Large Benefit Expansions,” Working Paper 1, 15 (2023).
35 Ben Luthi, “What Happens if I Default on a Loan?,” Experian, January 22, 2024,
https://www.experian.com/blogs/ask-experian/what-does-it-mean-to-default-on-a-loan/.
36 Statistics calculated using the SHED.
37 Kory Kroft and Matthew J. Notowidigdo, "Should Unemployment Insurance Vary With the Unemployment Rate
Theory and Evidence,” 83 The Review of Economic Studies 1092, 1094 (2016).
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43229
Page 15 of 27

Expert Report of Chloe N. East

receive less generous UI. Recall that even while receiving UI, most households already face the
need to tighten their budgets and cut spending because their UI benefits generally do not make
up for their entire lost earnings. Thus, losing UI benefits when households already have to cut
their spending is even more costly. Taken together, the costs of cutting consumption after the
loss of UI benefits are likely at least as high as the costs of the preferred option of borrowing. It
is my opinion that the cost of borrowing on a credit card is likely an underestimate of the actual
costs to cutting consumption.
33. As an example of the importance UI plays in helping households afford necessities, and
the costs to households if UI benefits are taken away and they have to cut back, I focus on food
consumption. It is well documented that food purchases fall when people lose their jobs, and UI
helps to lessen this decline after a job loss.38 In fact, receiving UI was associated with a 35%
decrease in food insecurity and a 48% decrease in the likelihood of reducing meals because of
financial constraints in 2020.39 Food insecurity has many negative and well documented
downstream impacts that generate real costs to the impacted individuals. In the literature, food
insecurity is associated with greater cognitive problems, higher risks of birth defects, higher
probabilities of asthma, higher levels of chronic disease, stress and anxiety, and more.40
34. As a second example, I consider the fact that it is well documented that UI helps people
afford health care and health insurance coverage.41 The loss of UI benefits will cause people to
reduce health care spending, increase their risk of becoming uninsured, and can also lead to
increases in medical debt. Using SHED data, I find that 22% of UI recipients in 2020-2021
report having unpaid medical debt. Loss of UI benefits will only exacerbate the problems faced
by these individuals with medical debt. Having medical debt comes with its own set of issues;
according to the Kaiser Family Foundation, the majority of adults with health care debt have had

38 Gruber, supra note 3 at 195; East & Kuka, supra note 23 at 33; Ganong & Noel, supra note 9 at 2405.
39 Julia Raifman, Jacob Bor, and Atheendar Venkataramani, “Association Between Receipt of Unemployment
Insurance and Food Insecurity Among People Who Lost Employment During the COVID-19 Pandemic in the
United States,” 4 JAMA Netw Open 1, 9 (2021).
40 Craig Gundersen, Brent Kreider, and John Pepper, “The Economics of Food Insecurity in the United States,” 33
Applied Economic Perspectives and Policy 281, 289 (2011).
41 Elira Kuka, “Quantifying the Benefits of Social Insurance: Unemployment Insurance and Health,” 102 The
Review of Economics and Statistics 490, 490 (2020); Ganong & Noel, supra note 9 at 2399.
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43230
Page 16 of 27

Expert Report of Chloe N. East

to cut back on basic necessities or use up their limited savings to pay down their debt.42 Unpaid
medical debt could further be sent to collection agencies and negatively affect credit in this time
period. Additionally, health insurance is associated with improved health outcomes both in the
short and longer run.43 Worse health leads to an increase in costs of about $1,500 per year (in
2013 USD) due to both increases in out-of-pocket medical spending and decreases in earnings.44
35. Because decreasing consumption can lead to such negative impacts, I expect that these
costs will exceed the cost of credit card interest rates these same individuals might have incurred
if they had been able to borrow. The credit card interest rates thus serve as an appropriate
minimum bound for the costs of the class, whether they borrow or simply cut spending. This
estimate likely undervalues the cost of being denied access to UI funds, because some people in
the class likely had to borrow using even more expensive methods and others were not able to
borrow and thus had to incur the extreme costs and harmful outcomes related to cutting back on
food and medical care along with other necessities.
36. To understand the cost of credit card borrowing for UI recipients, I used statistics from
the SHED data. I calculate that the median UI recipient in 2020-2021 had a “Good” credit score.
This corresponds to scores of about 739 and below.45 Thus, the vast majority of UI recipients
would face a minimum APR of 20.8% on their credit cards. If anything, this is an underestimate
since the APR would be as high as 23.6% for some of the group.46 I therefore recommend using
a borrowing rate of 20.8% as an estimate of the cost to class members resulting from losing
access to UI benefits, even temporarily.

42 Lunna Lopes, Audrey Kearney, Alex Montero, Liz Hamel, and Mollyann Brodie, “Health Care Debt in The U.S.:
The Broad Consequences of Medical and Dental Bills,” Kaiser Family Foundation, June 16, 2022,
https://www.kff.org/report-section/kff-health-care-debt-survey-main-findings/ (“When asked about some specific
problems they or someone in their household may have experienced in the past five years as a result of their health
care debt, six in ten adults with health care debt say they cut back spending on food, clothing, and basic household
items (63%).”).
43 Sarah Miller, Norman Johnson, and Laura R. Wherry, “Medicaid and Mortality: New Evidence from Linked
Survey and Administrative Data,” 136 Quarterly Journal of Economics 1783, 1820 (2021).
44 Mariacristina De Nardi, Svetlana Pashchenko, and Ponpoje Porapakkarm, "The Lifetime Costs of Bad
Health,” The Review of Economic Studies 1, 3 (2024).
45 Beverly Harzog, “What is a Fair Credit Score?,” at “What is a Fair VantageScore?,” U.S. News & World Report,
July 1 2024, https://money.usnews.com/credit-cards/articles/what-is-considered-a-fair-credit-
score#:~:text=If%20you%20have%20a%20fair,t%20let%20that%20discourage%20you.
46 Consumer Financial Protection Bureau, “The Consumer Credit Card Market,” at 19, Table 1, and 46-50, Figure 3
September 2021, https://files.consumerfinance.gov/f/documents/cfpb_consumer-credit-card-market-report_2021.pdf
(showing credit score ranges and average APR data).
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43231
Page 17 of 27

Expert Report of Chloe N. East

C.
Conclusion
37. Based on my research and review of the economic literature on UI recipients in
California during the stated time period, it is my opinion that the classwide harm caused by
delayed UI benefits can be measured, in part, by using an interest rate that represents the costs
incurred by those class members when the critical lifeline of UI was abruptly taken away (in
addition to the dollar value of the benefits themselves). Each class member incurred a related
cost: some turned to credit card borrowing, others to higher cost forms of borrowing, and others
were forced to go without food, forego medical care, or fall behind on bills because no credit was
available to them. Using the credit card interest rate, as a result, is in my opinion a conservative
estimate of the aggregate costs faced by the class while waiting for their UI benefits.

~~~~~~~~~~~~~~~

Dated: March 4, 2025

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43232
Page 18 of 27

Expert Report of Chloe N. East

IV.
Appendices
Appendix A: CV

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43233
Page 19 of 27

Expert Report of Chloe N. East

ED
liforni
D • , Jun 201
L . E.oor omics,
r iv
B .. E or tni • , ,Jth
1
,
uivc ··ty of I .xAS,
0
tin, ,to 2 11
Ho
d
b
Eff ac
f the Pub·
il:n
·ki
m·
oun a.l of Public E rwt
10S
,

1 nin n
uen
of lnu ·g
ion En£ r m nt; Hou h
and Hig}
m 11 s • • ork (,
. \clasq
} Iz.A 01
·o. 1 02
Human R
60 /ssu 1 1 Ja
1. ulti-g ncrational lmpa t
to th
l Li!◄
E po ur
M die id a.r d tl
_
a.1 d L. Whci-ry ,
BER W tkh1 P.q.>ct
1.1S, 0 .1
Heal
1. Page,
n Eco o
{1}: 9 -
t Eft ts f lnuni.gration Enforccm nt (with A. Hit cs, P. L ck,
. V
q
2 , IZA OP ·o. 114
Jo in a.l of u
E rwttics 41U}
or tion
ll
onpt
t H pital and o al H altb D partm nt
>
ion H altb In
tm n
b
on ro
tal ?
it. 1 T.
<111-
t.h
. . L , ru d R. Lind 00th), Medtcal Ca
n
pplc a Da: . Adult F
d
tamp Eligibi.lit and H altl't
at Utili2ation
(wi h
. 1''ric
n), lZA OP 'o. 11 4.,:;,
john lnsti u -0
'o king P pc . o. 19- %,
tt
•
,i Jou, ud of Hea.ttJ, Ecot om·
6(.1) 2 -929.
Ho
do E
LU
Op
rtu
(, it
of
· 1' u
o,·
b Effect of }"'ood
l'tanght.g E ie-ibili
b
abor uppl
226.
n
hUdtns
1 H rria.t H out
, La
t
V
,a11
r
Equa.lily
from lnunigrant
7.
Hamon • s, 1, 0 -
amini.ng
nsumption
mootlting B n Jlt of Un mplo m nt l ur-
( i l K
.
td al of l'u fc Ecot om ·c , 1,'12, .1 -.SO.
PAPERS
Work: Evid n
from Quasi-Rand ml
k),
BER
'o king Pa er 313 7,
µ ,1
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43234
Page 20 of 27

Expert Report of Chloe N. East

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43235
Page 21 of 27

Expert Report of Chloe N. East

ELLO.
HIP
D '"'" ... , .. S"'--4
(am
nt to
D
db low)
l
. Bi I r)
0
IS. l'LdtlO l m'J}r
t
r
.
Bifou ·kj F
h .
Doct t l
•
ati
sea • .1
a.ul
rn
ffo.i
Davi
,rat t,
Davi
_01
_OF
01-
014
_014
_014
_011
l
I j
(1◄
0r
le Public ,.
2
H) (x2), 0
(. 2),
, 2 19, 2020, 2021);
2024-2025
2022-202,'J
Le 1.ard
l , 0
<.llif om,. , Da
,-,,
r-n·a, Da
Ir
nou j
{it cltut·ng
. B
A
0.
tti
•
kc
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43236
Page 22 of 27

Expert Report of Chloe N. East

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43237
Page 23 of 27

Expert Report of Chloe N. East

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43238
Page 24 of 27

Expert Report of Chloe N. East

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43239
Page 25 of 27

Expert Report of Chloe N. East
23
Appendix B: Materials Considered
Case Materials
Order Re Preliminary Injunction dated May 17, 2021
Preliminary Injunction dated June 2, 2021
CFPB Consent Order dated July 14, 2022
OCC Consent Order dated July 14, 2022
Expert Report of Greg Regan dated August 29, 2024
Expert Report of Victor Stango dated October 24, 2024
Expert Report of Greg Regan dated November 21, 2024
Research
Camille Landais and Johannes Spinnewijn, “The Value of Unemployment Insurance,” Review of
Economic Studies 88, no. 6, 2021, pp. 3041-3085
Federal Reserve, “How Insured Are Workers Against Unemployment? Unemployment Insurance
and the Distribution of Liquid Wealth,” André Victor, D. Luduvice, and Anaya Truss-Williams,
2024
Federal Reserve, “The Impact of Government Transfer Payment Frequency on Consumption:
Evidence from Delayed UI,” by Michael Gelmen, Zachary Orlando, and Dhiren Patki, 2024
Federal Reserve Bank of San Francisco, “Enhanced Unemployment Insurance Benefits in the
United States During COVID-19: Equity and Efficiency,” by Robert G. Valletta and Mary Yilma,
Working Paper 2024-15, 2024
Fei Man and Valerie Tarasuk, “Employment Insurance may mitigate impact of unemployment on
food security: Analysis on propensity-score matched sample from the Canadian Income Survey,”
Preventative Medicine 169, 2023
James X. Sullivan, “Borrowing During Unemployment: Unsecured Debt as a Safety Net,” Journal
of Human Resources 43, no. 2, 2008, 383-412
Jonathan Gruber, “The Wealth of the Unemployed,” Industrial and Labor Relations Review 55,
no. 1 2001, 79-94.
Judith Bartfeld and Fei Men, “Policy and economic factors that affect food security,” IRP Focus
36, no. 3, 2020, pp. 13-25
Marianne P. Bitler, Jonah B. Gelbach, and Hilary W. Hoynes, “What Mean Impacts Miss:
Distributional Effects of Welfare Reform Experiments,” American Economic Review 96, no. 4,
2006, pp. 988-1012
Mathematica, “A Longitudinal Survey of Unemployment Insurance Recipients in Two Regions in
California,” by Joanne Lee, Karen Needels, and Walter Nicholson, 2017
Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43240
Page 26 of 27

Expert Report of Chloe N. East

Office of Financial Research, U.S. Department of the Treasury, “Household Liquidity
Measurement: A New Approach,” by Dasol Kim and Nick Schwartz, no. 24-03, 2024

Peter Ganong and Pascal Noel, “Liquidity versus Wealth in Household Debt Obligations: Evidence
from Housing Policy in the Great Recession,” American Economic Review 110, no. 10, 2020, pp.
3100-3138

Peter Ganong, Pascal Noel, and Joseph Vavra, “US unemployment insurance replacement rates
during the pandemic,” Journal of Public Economics 191, 2020, 104273

Politico, “Unemployment assistance to millionaires soared during pandemic,” by Brian Faler,
November 2022.

PPIC, “Unemployment Benefits in the COVID-19 Pandemic,” by Sarah Bohn, Marisol Buellar
Mejia, and Julien Lafortune, April 2020.

Vox, “American layoffs and firings are at a 20-year low,” April 2019

Data

“Survey of Income and Program Participation,” U.S. Census Bureau, 2019-2023,
https://www.census.gov/programs-surveys/sipp.html, accessed July 9, 2024.
“Survey of Household Economics and Decisionmaking,” Federal Reserve, 2019-2023,
https://www.federalreserve.gov/consumerscommunities/shed.htm, accessed September 26, 2024
“Household
Pulse
Public
Use
File,”
U.S.
Census
Bureau,
2020-2024,
https://www.census.gov/programs-surveys/household-pulse-survey/data/datasets.html, accessed
September 18, 2024.

Case 3:21-md-02992-GPC-MSB Document 594-1 Filed 10/23/25 PageID.43241
Page 27 of 27

File and source

File: gov.uscourts.casd.709615.594.1.pdf
Size: 9,397,718 bytes
SHA-256: 4e24796465039b5159d90f681d8f83423545e23ff5c7fba2e4e8bd1591f5e4ec
Our copy: gov.uscourts.casd.709615.594.1.pdf
Original: PACER (login required)

DX 14.F.1 - Expert Report of Chloe N. East — In re Bank of America California Unemployment Benefits Litigation (Dkt. 594-1, S.D. Cal. No. 3:21-md-02992)

Record facts

Full text

File and source

Related records