Racial Disparities in Access to Small Business Credit: Evidence from the Paycheck Protection Program (CESifo Working Paper No. 9345)
Summary
CESifo Working Paper No. 9345, Racial Disparities in Access to Small Business Credit: Evidence from the Paycheck Protection Program, by Sabrina T. Howell, Theresa Kuchler, David Snitkof, Johannes Stroebel and Jun Wong, dated October 2021. The abstract states that the authors explore sources of racial disparities in small business lending by studying the $806 billion Paycheck Protection Program (PPP). It states that, after controlling for a firm's zip code, industry, loan size, PPP approval date and other characteristics, Black-owned businesses were 12.1 percentage points (70% of the mean) more likely to obtain their PPP loan from a fintech lender than a traditional bank. It also states that among conventional lenders smaller banks were much less likely to lend to Black-owned firms, while the Top-4 banks exhibited little to no disparity after including controls.
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
- Issuer
- Source documents
- Document type
- Racial Disparities In Access To Small Business Credit Evidence From The Paycheck Protectio
Cite this document
“Racial Disparities in Access to Small Business Credit: Evidence from the Paycheck Protection Program (CESifo Working Paper No. 9345).”
Original at doi.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/racial-disparities-in-access-to-small-business-credit-evidence-from-the-paycheck-protection-program/
, SHA-256 196fb0bb0654f71ccc077c38944c565eed42a0d205713a8f77c44e0fe7af6752.Full text
Howell, Sabrina T.; Kuchler, Theresa; Snitkof, David; Stroebel, Johannes; Wong, Jun
Working Paper
Racial Disparities in Access to Small Business Credit:
Evidence from the Paycheck Protection Program
CESifo Working Paper, No. 9345
Provided in Cooperation with:
Ifo Institute – Leibniz Institute for Economic Research at the University of Munich
Suggested Citation: Howell, Sabrina T.; Kuchler, Theresa; Snitkof, David; Stroebel, Johannes; Wong,
Jun (2021) : Racial Disparities in Access to Small Business Credit: Evidence from the Paycheck
Protection Program, CESifo Working Paper, No. 9345, Center for Economic Studies and ifo Institute
(CESifo), Munich
This Version is available at:
https://hdl.handle.net/10419/248890
Standard-Nutzungsbedingungen: Terms of use:
Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Documents in EconStor may be saved and copied for your personal
Zwecken und zum Privatgebrauch gespeichert und kopiert werden. and scholarly purposes.
Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle You are not to copy documents for public or commercial purposes, to
Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich exhibit the documents publicly, to make them publicly available on the
machen, vertreiben oder anderweitig nutzen. internet, or to distribute or otherwise use the documents in public.
Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen If the documents have been made available under an Open Content
(insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, Licence (especially Creative Commons Licences), you may exercise
gelten abweichend von diesen Nutzungsbedingungen die in der dort further usage rights as specified in the indicated licence.
genannten Lizenz gewährten Nutzungsrechte.
9345
2021
October 2021
Racial Disparities in Access to
Small Business Credit:
Evidence from the Paycheck
Protection Program
Sabrina T. Howell, Theresa Kuchler, David Snitkof, Johannes Stroebel,
Jun Wong
Impressum:
CESifo Working Papers
ISSN 2364-1428 (electronic version)
Publisher and distributor: Munich Society for the Promotion of Economic Research - CESifo
GmbH
The international platform of Ludwigs-Maximilians University’s Center for Economic Studies
and the ifo Institute
Poschingerstr. 5, 81679 Munich, Germany
Telephone +49 (0)89 2180-2740, Telefax +49 (0)89 2180-17845, email office@cesifo.de
Editor: Clemens Fuest
https://www.cesifo.org/en/wp
An electronic version of the paper may be downloaded
· from the SSRN website: www.SSRN.com
· from the RePEc website: www.RePEc.org
· from the CESifo website: https://www.cesifo.org/en/wp
CESifo Working Paper No. 9345
Racial Disparities in Access to Small
Business Credit:
Evidence from the Paycheck Protection Program
Abstract
We explore the sources of racial disparities in small business lending by studying the $806 billion
Paycheck Protection Program (PPP), which was designed to support small business jobs during
the COVID-19 pandemic. PPP loans were administered by private lenders but federally
guaranteed, largely eliminating unobservable credit risk as a factor in explaining differential
lending by race. We document that even after controlling for a firm’s zip code, industry, loan size,
PPP approval date, and other characteristics, Black-owned businesses were 12.1 percentage points
(70% of the mean) more likely to obtain their PPP loan from a fintech lender than a traditional
bank. Among conventional lenders, smaller banks were much less likely to lend to Black-owned
firms, while the Top-4 banks exhibited little to no disparity after including controls. We use novel
data to show that the disparity is not primarily explained by differences in pre-existing bank or
credit relationships, firm financial positions, fintech affinity, or borrower application behavior. In
contrast, we document that Black-owned businesses’ higher rate of borrowing from fintechs
compared to smaller banks is particularly large in places with high racial animus, pointing to a
potential role for discrimination in explaining some of the racial disparities in small business
lending. We find evidence that when small banks automate their lending processes, and thus
reduce human involvement in the loan origination process, their rate of PPP lending to Black-
owned businesses increases, with larger effects in places with more racial animus.
JEL-Codes: G210, G230, G280, G410, J150.
Sabrina T. Howell Theresa Kuchler
NYU Stern / USA NYU Stern / USA
sabrina.howell@nyu.edu tkuchler@stern.nyu.edu
David Snitkof Johannes Stroebel
Ocrolus, New York / USA NYU Stern / USA
johannes.stroebel@nyu.edu
Jun Wong
University of Chicago / IL / USA
junwong@uchicago.edu
October 5, 2021
This paper subsumes results in an earlier note from December 2020 titled “Which lenders had the
highest minority share among their Payment Protection Program (PPP) loans?” We are grateful
to Georgij Alekseev, Sara Gong, and Cangyuan Li for superb research assistance, and to Enigma,
Middesk, Lendio, Biz2Credit, and Ocrolus for sharing part of their data. In particular, we thank
Scott Monaco at Ocrolus for his insights and analytical work. We are also grateful for help from
Sriya Anbil, Rohit Arora, Brock Blake, Katherine Chandler, Christine Dobridge, Kyle Mack,
Karen Mills, David Musto, Anand Narayan, Hicham Oudghiri, Madeline Ross, Kurt Ruppel, and
Sam Taussig. We thank seminar participants at NYU and the University of Georgia for their
helpful comments. Howell’s work on this project is funded by the Kauffman Foundation.
In recent years, policy makers in the U.S. and elsewhere have become increasingly concerned about
differences across racial groups in access to financial services such as consumer and small business credit
(Crutsinger, 2021; Abrams, 2021; Crowell, 2021). Understanding the determinants of observed racial
differences in credit access has been challenging due to the difficulty of disentangling the role of credit
risk from a possible independent role of race through channels such as preference-based discrimination.
In this paper, we study how and why lenders differed in their propensity to extend Paycheck Protection
Program (PPP) loans to small businesses owned by members of various racial and ethnic groups. The
PPP was established as part of the CARES Act, passed in March 2020, and was designed to help small
businesses struggling during the COVID-19 pandemic. With a volume of $806 billion, it is one of the
largest single public finance programs in U.S. history.
The design of the PPP makes it an attractive setting to study racial disparities in access to small
business credit. The Small Business Administration (SBA) did not issue specific guidance on loan
distribution, leaving the private financial institutions administering the loans to independently determine
which businesses to serve. As a result, institutional factors that determine banks’ regular lending
decisions may have also affected PPP loan originations. Importantly for our research design, PPP loans
were fully guaranteed by the federal government, and the loan amount was based solely on payroll,
largely eliminating credit risk and concerns about selection on other dimensions, such as loan size, from
banks’ lending decisions. When combined with information on other important determinants of lending
decisions, such as firm-level banking relationships and firm cash flows, the PPP data can therefore shed
light on the role of race and ethnicity in the lending decisions of different private lenders.
We work with public administrative data from the SBA on 11.8 million loans made between April 3,
2020 and May 31, 2021. In our analysis, we restrict the sample to “first draw” loans (some firms were
eligible for two loans) and to loans made before February 24, 2021, when the program’s rules changed
to explicitly prioritize lending to small firms and minorities. Within this sample of 5.7 million loans, we
build on a well-established literature to predict a business owner’s race and ethnicity using information
such as the owner’s name and location (Imai and Khanna, 2016; Humphries et al., 2019; Tzioumis,
2018). We gather owner names from business registrations in collaboration with the data analytics firm
Middesk. To improve the prediction, we train a random forest model on the subset of PPP loans in which
the borrower reports their race. Our model performs well at predicting self-reported race in a hold-out
sample of PPP borrowers not used to train the model. This assigned race should be interpreted as a signal
that is highly correlated with self-reported race and contains important socioeconomic content, consistent
with studies showing discrimination against job applicants with “African American-sounding” names
(Bertrand and Mullainathan, 2004; Milkman et al., 2012; Bartoš et al., 2016).1 Using this approach, we
assign race and ethnicity to the business owner for 4.2 million PPP loans.
We document substantial variation across different types of financial institutions in the propensity to
1
As we discuss in more detail below, in those instances in which the self-reported race differs from the predicted race, it is
plausible that many loan officers, who usually do not have access to borrowers’ true race but observe the inputs to the algorithm,
would treat borrowers in a way that is more aligned with the signal than with self-reported race. For example, a number of
borrowers with the last name “Huang” self-report to be Black but are assigned “Asian” by the algorithm.
1
extend PPP loans to Black-owned businesses. Within the analysis sample, about 8.6% of all loans and
2.9% of loans to employer businesses went to Black-owned businesses. Traditional banks made between
3.3% (among small banks) and 7.8% (Wells Fargo) of their PPP loans to Black-owned businesses. On the
other hand, fintech lenders made 26.5% of their PPP loans to Black-owned businesses. Overall, fintech
lenders were responsible for 53.6% of PPP loans to Black-owned businesses, while only accounting
for 17.4% of all PPP loans in the analysis sample. There are also differences across lenders in the
propensity to lend to White-, Asian- and Hispanic-owned firms. However, the difference in lending
between conventional lenders and fintech lenders is most striking for Black-owned businesses, and this
disparity is the focus of our paper.
We first consider important dimensions along which Black-owned businesses differ from other
firms. For example, their average PPP loan amount of $24,315 is less than half of that for Asian- and
Hispanic-owned businesses, and one-quarter of that for White-owned businesses. Since PPP lenders
were compensated with 5% of the loan value for loans under $350,000, the presence of fixed costs or
capacity constraints creates an incentive to prioritize the largest loans. If fintech lenders have smaller
fixed costs or more capacity, this could explain their disproportionate lending to smaller, Black-owned
firms. We therefore add granular controls for loan size. We also control for loan approval date and
borrower characteristics such as zip code, industry, business organizational form, and employer status.2
Overall, about two-thirds of the unconditional difference in Black-owned businesses’ propensity to get
fintech PPP loans relative to loans from other lenders is explained by differences in these characteristics.
However, even with these controls, which hold the profitability of lending and the availability of bank
branches fixed, we find that Black-owned businesses are 12.1 percentage points more likely to obtain their
PPP loans from a fintech firm than from traditional lenders relative to the mean probability of obtaining
a fintech PPP loan of 17.4%. Much of this effect is substitution away from non-Top-4 banks—especially
small- and medium-sized banks—towards fintech lenders by Black-owned businesses.3 For Top-4 banks,
we do not find a large difference in the propensity to extend PPP loans to Black-owned businesses after
including loan-level and firm-level controls. In this paper, we analyze what might explain the remaining
large disparity between Black-owned firms and other businesses in the propensity of getting a PPP loan
from a fintech firm rather than a smaller conventional lender.
First, observers of the PPP program have noted that banks tended to prioritize their own clients’
PPP loan applications, which may have distorted allocations away from the government’s intended “first
come, first serve” approach (Cowley, 2021; Li and Strahan, 2020). If banks indeed prioritized PPP
loan applications from their own clients, and if Black-owned businesses did not bank with active PPP
lenders (in particular among the small banks), this could explain some of the observed differences in their
propensity to eventually borrow from other lenders such as fintech firms. To assess whether this channel
2
Adding zip code fixed effects suggests that traditional lenders are less likely than fintechs to make loans to businesses—
even White-owned ones—in predominantly minority communities. While our paper focuses on explaining the disparity in
lending patterns conditional on observable characteristics such as location, understanding why different lenders have different
propensities to extend loans to any firms located in high-minority neighborhoods is an important question for future research.
3
When we consider lending to the other races and ethnicities in our fully controlled models, we find that small- and medium-
sized banks in particular were more likely to lend to White-owned businesses.
2
can explain the observed disparity, we match the PPP data to novel bank statement data from Ocrolus,
a firm that digitizes and analyzes financial documents for financial institutions. These bank statements,
which are available for about 170,000 PPP borrowers in our analysis sample, include information on
bank and credit relationships as well as cash flows. Within this matched sample—which selects on
having a checking account and a past fintech loan application—Black ownership is associated with a
5.5 percentage point higher probability of obtaining a PPP loan through a fintech lender, conditional on
controls. This disparity is unaffected by adding fixed effects for the identity of the bank where a borrower
has their primary business checking account. In other words, even though we find that banks generally
served their own clients at higher rates, this fact does not explain the higher rate of fintech PPP loans for
Black-owned businesses.
The bank-statement data also allows us to further explore why Black-owned businesses were more
likely to borrow from fintechs rather than smaller banks. In analyses that restrict the sample to firms
with checking accounts at different types of banks, we identify two channels through which Black-
owned businesses shifted to borrowing from fintechs rather than smaller banks. First, Black-owned firms
banking outside the top-4 banks were less likely to get their PPP loans from their checking account banks.
Second, among those firms that got PPP loans outside their checking account banks, Black-owned firms
were less likely to obtain loans from smaller banks, and more likely to obtain them from fintech lenders
instead. Quantitatively, this second channel, which captures racial differences in the rates of establishing
new credit relationships with different types of lenders, explains the majority of the observed disparity.
We next explore whether borrowers’ financial conditions can explain the relatively higher
probability of Black-owned businesses of obtaining PPP loans from fintech firms instead of smaller
banks. Even though PPP loans were fully guaranteed by the federal government, some traditional
lenders may have preferred lending to better-performing businesses that would be more attractive future
clients. In addition, loan officers at these banks, who are used to screening for creditworthiness, may
have preferred businesses with better financial positions as they do in their usual course of business,
even if a borrower’s financial performance would not directly affect the profitability of a given PPP
loan. We assess this possibility using monthly data from Enigma on firms’ credit and debit card
revenues, focusing on the period immediately prior to loan approval. In the matched sample of 820,000
firms, adding fixed effects for percentiles of card revenue during the PPP loan approval month and the
two previous months has no effect on the observed propensity of Black-owned firms to receive their PPP
loans from fintech lenders. Controlling for bank statement cash flows observed in the Ocrolus data also
does not affect the disparity. While both the card revenue and bank statement samples are not entirely
representative—in particular, they skew towards larger or more sophisticated firms—the fact that the
disparity persists in these samples is evidence of its importance across a wide range of small businesses.
A third possible explanation for the racial disparities in lending outcomes is firm application behavior
rather than lender decision-making. For example, it could be that Black-owned businesses prefer fintech
lenders and thus applied to these lenders at higher rates. This is unlikely given that we also observe a
disparity within a sample of firms with prior fintech loan applications which holds fixed a certain degree
3
of fintech affinity. Nonetheless, to further assess this possibility, we obtain PPP application data from
the marketplace lending platform Lendio. Lendio routed some PPP applications to conventional lenders,
some to fintech lenders, and some to both (routing was random conditional on loan size, geography, and
capacity criteria set by the lender partners). We find that even within a sample of about 46,000 PPP loan
applications that Lendio sent only to conventional lenders, Black-owned businesses were 3.9 percentage
points more likely to end up with a fintech PPP loan. This offers evidence consistent with the hypothesis
that conventional lenders more often reject applications from Black-owned businesses, pushing these
firms to subsequently apply to fintechs.
We also consider whether systematic differences between fintech lenders’ and conventional lenders’
compliance standards may have contributed to the observed differential probability of lending to
Black-owned businesses. Griffin et al. (2021) suggest that fintechs had different Know-Your-Customer
and Anti-Money Laundering (KYC/AML) policies.4 However, this is unlikely to be the explanation for
the differential lending across racial groups that we observe. Recall that conditional on observable firm
characteristics, the substitution towards fintech comes from relatively smaller banks, with no apparent
lending disparity at the Top-4 banks. It is not plausible that the Top-4 banks—which are the most
intensively regulated—have lower compliance standards than smaller banks.
The final mechanism that we consider is the role of human decision-making, which could enable
preference-based discrimination against Black borrowers. To explore this possibility, we examine how
the propensity of Black-owned businesses to obtain fintech loans varies with measures of racial
discrimination. Using six measures of racial animus—including racially biased Google searches,
implicit and explicit bias tests, and measures of local housing segregation—we find that the tendency of
Black-owned businesses to borrow from fintech lenders instead of smaller conventional lenders is
consistently higher in areas with more racial animus, even after controlling for firm and loan
characteristics. These results suggest that preference-based discrimination might explain at least some
of the substitution of Black-owned businesses towards fintech and away from smaller banks.
Automation of the loan application and approval process could help account for the absence of
disparate treatment by race at the Top-4 banks (after including controls) and the high rate of lending to
Black-owned firms at fintechs. Fintech lenders almost fully automate their underwriting processes,
leaving little scope for preference-based discrimination to affect the approval decision. In contrast,
conventional lenders traditionally rely more on human involvement and personal relationships between
loan officers and clients (Petersen and Rajan, 1994; Berger and Black, 2011). However, there is sizable
variation in the degree of automation across conventional lenders, with larger banks—those where we
see little difference in the conditional probability of lending to Black-owned businesses—being more
automated than smaller banks.5 Consistent with an important role of automated lending systems in
4
PPP lenders may have been responsible for loans shown to be fraudulent and were accountable for some but not all of the
compliance requirements that apply to normal small business lending (regulatory guidance on these requirements was at times
confusing and unclear, which may have led some lenders to err on the side of caution).
5
For example, one news article notes that “Large banks have avidly adopted robotic process automation...It’s tougher for
smaller banks to follow suit” (Crosman, 2020). As a second example, a 2018 Fannie Mae survey found that 76% of large banks
but only 47% of small banks were familiar with artificial intelligence or machine learning technology.
4
explaining our findings, we show that when small banks automated their loan origination procedures
during the PPP period, their rate of lending to Black-owned businesses increased relative to other small
banks that did not automate their processes. A differences-in-differences analysis shows that
automation more than doubled small banks’ propensity to lend to Black-owned businesses, with larger
effects in locations with higher racial animus. These findings support the conjecture that automation
might help reduce racial discrimination in the loan origination process.
Despite our findings suggesting an important role of taste-based discrimination in explaining some
of the observed patterns, there may be other remaining explanations for the residual racial gap in the
probability of borrowing from different lenders. Having said this, it is unclear how a mechanism besides
discrimination would account for the larger racial gap in areas with higher racial animus. In addition, our
controls for firm size, location, industry performance, and pre-existing banking and credit relationships
account for most possible explanations. It is also important to highlight that many of the variables we
control for in our regressions—such as the location of bank branches or the distribution of which firms
have checking accounts at banks—may themselves be the result of historical patterns of discrimination.
In other words, it is possible that the substantial differences in our controlled models represent a lower
bound estimate of the overall effect of discrimination on lending patterns.
Several contemporaneous papers offer results consistent with ours. Erel and Liebersohn (2020)
show that fintech lenders tended to extend more PPP loans in areas with higher minority population
shares. Fairlie and Fossen (2021) also use geographic variation to find that total PPP loan flows were
negatively correlated with the minority share of the population. Relative to these papers, we
demonstrate that even within a given geography, fintech lenders disproportionately lent to Black-owned
firms, showing that bank branch location cannot fully explain the observed patterns. In addition, we
make progress on identifying the mechanisms, ruling out that the patterns are largely explained by
differential banking relationships, financial performance, or application behavior, and finding evidence
consistent with preference-based discrimination among smaller banks without automated lending
processes. Chernenko and Scharfstein (2021) study PPP lending in a sample of nearly 11,000
restaurants in Florida, for which they can link owner names to voter registration records. They show
that minority-owned businesses are more likely to get non-bank PPP loans and conclude that racial bias
seems to explain the lending differences. Our paper benefits from vastly larger and richer data that is
more representative of PPP borrowers across geographies and industries. In addition, our use of
checking account data, credit card data, and loan application data allows us to directly assess a variety
of possible explanations in addition to discrimination. Finally, our analysis of cross-sectional and
time-series variation in the lending behaviors of different conventional lenders allows us to show that
the degree of automation in the lending process is a key factor in explaining the observed patterns.
More broadly, we contribute to the understanding of racial disparities in access to financial services.
This literature has mainly studied residential mortgage and consumer credit markets (Tootell, 1996;
Bayer et al., 2018; Dobbie et al., 2020; Bhutta and Hizmo, 2021; Ambrose et al., 2020; Giacoletti et al.,
2021; Begley and Purnanandam, 2021; Blattner and Nelson, 2021). While there is extensive work on
5
bias against Black people across a wide variety of settings (e.g., Arnold et al., 2018; Bertrand and
Mullainathan, 2004; Knowles et al., 2001; Anwar and Fang, 2006; Price and Wolfers, 2010), there is
less work on discrimination against Black business owners. Blanchflower et al. (2003) find racial
differences in access to small business credit, while Robb and Robinson (2018) do not find such
differences. Other work on the role of race in small business lending includes Fairlie and Robb (2007),
Asiedu et al. (2012), Bellucci et al. (2013), and Fairlie et al. (2020).
Our work also relates to a literature studying how the COVID-19 pandemic and associated policy
responses affected small businesses (Alekseev et al., 2020; Bartik et al., 2020a; Fairlie, 2020; Greenwood
et al., 2020; Kim et al., 2020). Much of the work on the PPP focuses on the effectiveness of the program
in mitigating job loss (Hubbard and Strain, 2020; Faulkender et al., 2020; Granja et al., 2020; Autor
et al., 2020; Bartik et al., 2020b; Barraza et al., 2020; Bartlett and Morse, 2020). Other researchers
have examined whether firm size or pre-existing banking relationships can explain access to PPP loans
(Humphries et al., 2020; Li and Strahan, 2020). Finally, we add to a literature that assesses the role of
fintech firms in the financial system (Seru, 2019; Philippon, 2019; Federal Reserve, 2020; Ranson, 2020;
Gopal and Schnabl, 2020). Most directly related is a literature that has explored the role of fintech lenders
in extending credit to traditionally underserved minorities (Tang, 2019; Balyuk et al., 2020; Fuster et al.,
2019; Buchak et al., 2018; Berg et al., 2020; Bartlett et al., 2021; D’Acunto et al., 2020).
1 The Paycheck Protection Program: Setting and Data
The Paycheck Protection Program (PPP) was first established as part of the Coronavirus Aid, Relief and
Economic Security Act (“CARES Act”), passed on March 27, 2020. The PPP provided 100%
government-guaranteed uncollateralized loans to firms which certified that their businesses were
“substantially affected by COVID-19.” For businesses using the loans to make eligible payments, such
as covering payroll, the loans would eventually be forgiven. The loan amount was fixed at 2.5 times
monthly pre-COVID payroll. The Small Business Administration (SBA) approved lenders and
individual loans, though this primarily involved a duplication check to avoid granting multiple loans to
a single entity. While Section 1102 of the CARES Act specifies that the program should prioritize
“small business concerns owned and controlled by socially and economically disadvantaged
individuals,” private lenders, which can face capacity constraints, were fully responsible for targeting
the funds and determining which PPP applications to prioritize. Media reports early in the life of the
PPP program raised concerns that banks were turning away large numbers of PPP applications from
minority-owned businesses (Simon and Rudegeair, 2020; Zhou, 2020; Beer, 2020).
The SBA compensated lenders for originating and servicing PPP loans according to the following
graduated, upfront fee system, creating incentives for lenders to prioritize larger loans:
• 5% for loans of not more than $350,000;
• 3% percent for loans of more than $350,000 and less than $2,000,000; and
• 1% percent for loans of at least $2,000,000.
6
The initial CARES Act authorized $349 billion in loan guarantees for the PPP, and issuance began on
April 3, 2020. Demand for PPP loans vastly exceeded expectations, and funding for the initial program
ran out on April 16, 2020. Congress approved a second PPP tranche of $310 billion on April 24, 2020,
and its distribution began on April 27, 2020. A third tranche of $284.5 billion was approved on December
27, 2020. In this round, firms were eligible to receive a “second draw” loan if they met certain conditions.
By the time the program closed permanently at the end of May 2021, 11.5 million loans, administered
by 5,467 lenders and totaling more than $800 billion, had been approved. As of June 2021, 64% of all
loans made in 2020 had been forgiven.
Lenders participated voluntarily in the PPP, and they entered and left the program over time. Fintechs
tended to enter somewhat later for several reasons. Some required special approval because they were
not regulated insured depository institutions or pre-approved SBA lenders. Others did not have large
enough balance sheets to originate many PPP loans and needed to wait for the PPP Liquidity Facility
established by the Federal Reserve to come online, which only occurred several weeks after the program
began. This facility enabled banks, and later fintechs, to post PPP loans as collateral for new funds with
which to originate loans. Another group of fintechs partnered with originating charter banks, such as
Celtic, in order to participate. In our analysis below, we control for the week of PPP loan approval to
ensure that our results are not affected by the dynamics of lender participation.
1.1 PPP Data
We obtain information on all PPP loans as of August 15, 2021 directly from the SBA. These data were
released following a court order and include the business name and address for all PPP loan recipients,
as well as information about the business type, loan size, self-reported number of jobs saved, the loan
originator, and the loan servicer. Subsequently, we term the loan originator the “PPP lender.” To construct
our dataset, we retain only a firm’s first loan and drop apparent duplicate loans, so that each firm appears
once. Specifically, we begin with a raw dataset from the SBA, which has 11.8 million loan observations.
Of these, 2.9 million are tagged as second draws (where a firm legally obtained a second PPP loan). After
dropping these, we are left with 8.8 million observations.
We also drop loans made after February 23, 2021. We do this because the Biden Administration
made drastic changes to the PPP, the first of which took effect on February 24. These changes included
first restricting loans to small firms with less than 20 employees, and then permitting only Community
Development Financial Institutions (CDFIs) to use PPP funds. Since our goal is to understand lending
behavior in a more representative population of both lenders and borrowers, we drop loans made after
February 23, 2021. This leaves us with 5.7 million PPP loans. However, our results are very similar
when we use the full time period, as well as when we include second draw loans.
The focus of our analysis is to understand differences by lender type in originating PPP loans to
minority-owned businesses. We classify all PPP lenders into the following mutually exclusive groups:
1. The Top-4 banks by assets (JP Morgan Chase, Bank of America, Wells Fargo, and Citibank);
7
2. Large banks: Banks with more than $100 billion in assets, excluding the Top-4;
3. Medium-sized banks: Banks with more than $2.2 billion in assets (but below $100 billion);
4. Small banks: Banks with less than $2.2 billion in assets;6
5. Credit unions: Based on the lender name (i.e., having ”credit union” or ”CU” at end of the name);
6. Community Development Financial Institutions (CDFIs) and nonprofits;7
7. Minority Depository Institutions (MDIs): As classified by the FDIC.
8. Fintech lenders: All lenders officially designated as such by the SBA. We further include online
lenders who originate primarily for or via fintech partners or platforms, online lenders founded
since 2005, and online lenders that received venture capital (VC) investment.8
This classification yields 11 mutually exclusive lender types (note each of the Top-4 banks is considered
as its own category). Table 1 shows how PPP loan origination varied across these lender types for
the full sample (Panel A) and the first-draw sample (Panel B). Focusing on Panel B, traditional banks
originated about 75% of PPP loans, with non-Top-4 banks responsible for 59% of all loans. Fintech
lenders originated 16.2% of all PPP loans, while credit unions and MDIs each originated about 4%, and
CDFIs 2%. Fintech and non-bank lenders made substantially smaller loans. The average (median) PPP
loan amount for fintech lenders is $31,921 ($15,500) compared to, for example, $87,182 ($20,833) for
small banks. The PPP loans originated by credit unions and Wells Fargo were the next-smallest in both
average and median. In Panel A of Appendix Table A.2, we show similar summary statistics for the
subset of loans for which we can predict borrower race.
1.2 Identifying Borrower Race and Ethnicity
A key element of our analysis is identifying the race and ethnicity of the owners of the firms participating
in the PPP. The SBA data contains details on owner race for a small and selected subset of PPP borrowers
who chose to self-report this information in their loan application, and for which the lender also chose
to report this information to the SBA. To identify a signal for race and ethnicity for a larger set of PPP
6
We include the roughly 6,000 loans by Business Development Corporations (BDCs) in the Small Bank category, since
these loans behave similarly in terms of the variables we study as the Small Bank loans.
7
We pool nonprofits with CDFIs because nonprofits issued very few loans (fewer than 15,000) and their loans have similar
characteristics as those of the CDFIs.
8
Appendix Table A.1 lists all lenders we classify as “fintech.” In some cases, the originator listed in the table made loans
primarily through fintech partners. For example, all of PayPal’s fintech loans were originated by WebBank, and Square’s loans
by Celtic Bank. The only fintech lender with a branch is Cross River. However, Cross River originated an overwhelming
quantity of loans for fintech partners such as Kabbage, they were founded in 2008, and received VC funding, so we consider
Cross River a fintech lender for our purposes. In the data, we do not observe loan referrals from traditional banks to other
lenders, so loans referred to fintech by other lenders would be classified as fintech loans. We also do not observe back-end
processors that do not show up as lenders or servicers, including Finastra, Ocrolus, and Customers Bank (which played this role
for other lenders even as it was also processing its own PPP loans). So some loans processed by fintech firms but originated by
other lenders would be classified according to their ultimate lender.
8
borrowers, we build on a well-established literature to predict race from a business owner’s name, and
the firm’s location, industry, and employer status.
We first identify a borrower firm’s individual owner or most senior executive. Our primary source
for this information is data on current firm officers as of July 2021 drawn from Secretary of State
registrations, provided to us by the analytics firm Middesk. The owner is identified as the first individual
listed as owner or principal under “business contacts” in Secretary of State filings. For non-employer
firms such as sole proprietorships, we also use the fact that the “business name” reported in the PPP
data usually corresponds to the owner’s name. Finally, we obtained applicant names for a sample of
PPP applicants from Lendio. We combine these data with public SBA information on the businesses’
address, industry, and employer status. Utilizing this data, we use a machine learning approach to
estimate the conditional probability of a business owner being Asian, Black, Hispanic, or White.
Our process consists of two steps. First, we follow the methodology in Imai and Khanna (2016) by
combining the Census list of last names (Word et al., 2008) with the census tracts of business locations
to estimate the conditional probability that an individual belongs to a certain racial group given their
last name and location.9 The second step combines the resulting Bayesian posterior probability with the
racial distribution of common first names and industries by employer status as features in a random forest
model with 1,000 trees (see Tzioumis, 2018; United States Census Bureau, 2012). We train and validate
the random forest model on the 809,119 PPP loans with self-reported race, a successfully geolocated
address, an identified owner name and information on a firm industry and employer status. The model
generates the probability that a borrower belongs to a certain race given first and last name, location,
industry, and employer status. We identify the borrower to be of the race with the largest probability
across the set of racial groups R. In a robustness test, we restrict the sample to borrowers who have an
assigned probability of more than 90% for their most likely races.
In total, we can predict the race for 4.18 million unique PPP borrowers. For the remaining firms,
we do not observe owner name or the geolocation fails. We also exclude about 30,000 loans for which
the algorithm predicts the owner race to be “Other.” We show in Panel A of Appendix Table A.3 that
the model correctly predicts the vast majority of the self-reported sample (note that, for most of our
analysis, we do not actually use the self-reported race in our analysis, but instead the prediction, so that
we consistently capture the signal sent by the name). For example, of those we predict to be Black, 93%
self-identify as Black (9.1/9.6 ≈ 0.93). To assess the out-of-sample quality of the prediction, which is
the most relevant for assessing our ability to predict the race and ethnicity of individuals who did not
self-identify, we randomly set aside a “hold-out” sub-sample of borrowers who self-identify race but
9
Suppose we denote the last name and census tract of an individual i as Li and Ti . The unobservable race or ethnicity
of the individual is denoted as Ri , and R ∈ {Asian, Black, Hispanic, W hite} is the available set of all such groups. We
would like to estimate P r(Ri = r|Li = l, Ti = t). The Census list of last names provides the racial and ethnic distribution
of 151,671 last names, P r(Ri = r|Li = l), which makes up 90% of the population in the 2000 Census. We obtain the racial
and ethnic distribution of each census tract from the American Community Survey, which gives P r(Ri = r|Ti = t) and the
population share of each census tract P r(Ti = t). Assuming that the location and last name of an individual are independent
conditional on race and ethnicity, Bayes’ rule implies P r(Ri = r|Li = l, Ti = t) = P 0 P r(T i =t|Ri =r)P r(Ri =r|Li =l)
0 0 .
r ∈R P r(Ti =t|Ri =r )P r(Ri =r |Li =l)
Here, P r(Ti = t|Ri = r) can once again be decomposed using Bayes’ Rule, allowing for a probabilistic prediction of an
individual’s race and ethnicity.
9
are not included in the training of the random forest model. Panel B of Appendix Table A.3 contains
the confusion matrix of our race prediction within the hold-out sub-sample. It shows that 75% of those
business owners that we predict to be Black self-identified as Black (5.7/7.6 ≈ 0.75).
We show the probability distribution for each race in Appendix Figure A.1. For example, Panel A
contains the set of borrowers whose predicted race is Black (the algorithm assigns Black to have the
highest probability across the race and ethnicity options). The graph shows the distribution of the
probability of being Black among these observations. Panel D of Appendix Table A.3 summarizes these
distributions for each race; among people predicted to be Black, the mean chance of being Black
according to the algorithm is 76%, with a median of 80%. In some of our tests, we use this probability
(instead of just a race dummy) to provide further evidence for our interpretations of our findings.
It is worth noting that, since loan officers often only have access to borrowers’ names and locations
(rather than true borrower race), they may respond to the race or ethnicity most associated with a given
name, rather than to the borrowers’ actual race or ethnicity. For example, two of the prediction
algorithm’s “errors” are individuals whose last names are Huang and Rodriguez, who identify as Black
but are predicted to be Asian and Hispanic, respectively. It is plausible that loan officers observing only
applicants’ names might also infer an incorrect race for these borrowers, and our algorithmically
assigned race may correspond more closely to the race inferred by a loan officer. Such behavior would
be highly consistent with findings from audit studies such as Bertrand and Mullainathan (2004), who
document discrimination against job applicants with “African American-sounding” names.
We call the sample for which we can identify race the “PPP Analysis Sample.” Panel A of Table
2 shows that, within this sample, 8.6% of business owners are Black, 7.5% are Hispanic, 8.9% are
Asian, and 75.0% are White. The distributions of originating lender and firm characteristics such as
loan amount are similar across the full sample (Table 1, Panel B) and the analysis sample (Table 2). For
example, the average PPP loan amount is $93,784 in the full sample and $93,666 in the analysis sample.
Appendix Table A.4 confirms that the distributions of firms’ business types, locations, and industries are
also similar in the full and analysis samples.10 This highlights that the sample for which we can predict
race is broadly representative of the overall PPP population, which, in turn, is relatively representative of
privately-owned U.S. businesses on industry and geography (see SBA May 2021 Program Report).
The racial composition of our PPP analysis sample is also comparable to that of the population of
small business owners in the United States. Appendix Table A.5 repeats Panel A of Table 2 for employers
and non-employers, respectively. For example, we predict that 2.9% of employer businesses in the PPP
analysis sample are Black-owned, compared to 2.1% of the population of comparable small business
owners in the 2012 U.S. Census Bureau Small Business Owners survey (United States Census Bureau,
2012). Among non-employer firms, 18.7% are Black-owned in our sample vs. 11.2% in the population.
10
We consolidate business types to seven categories from the 19 organizational forms in the SBA data: corporations, limited
liability corporations (LLC), non-profits, self-employed, sole proprietorships, subchapter S corporations, and other. We assign
cooperatives and other non-profit organizations under the non-profit umbrella. Independent contractors and self-employed
individuals are classified as self-employed, and limited liability partnerships are considered as LLCs. “Other” includes any
business types with less than 100,000 observations (such as employee stock ownership plans, joint ventures, etc.)
10
1.3 Bank Statement Data
To help distinguish between various explanations for differential lending of banks across races, we
acquire data from Ocrolus on borrower firms’ bank statements through July 2021. Ocrolus digitizes
documents for fintech companies, including bank statements that these lenders use in the underwriting
process, and thus has a large repository of business checking account statements. We match around
216,000 unique PPP borrowers in our analysis sample to Ocrolus’ database using information on the
business name and address. If several bank statements are available for a firm (the average firm has
three bank statements, mostly from 2019 and 2020), we focus on the most recent statement prior to the
issuance of the PPP loan.
Using information from the bank statements, we determine borrowers’ pre-existing banking and
credit relationships. We define a firm’s checking account bank as the bank that issued the statement.11
Panel C of Table 1 reports statistics on the whole bank statement-matched sample, organized by PPP
lender as in the previous panels. (We repeat Panel C for the subset of loans for which we can predict
borrower race in Panel B of Appendix Table A.2.) Within this sample, 28.5% of businesses obtained
their PPP loans from their checking account banks. There is sizable heterogeneity across banks. About
two-thirds of all PPP loans originated by Bank of America, JP Morgan Chase, and Wells Fargo went to
checking account clients of those banks. For other large banks, this number is 50%, and for medium and
small banks, it is 39.8% and 23.8%, respectively. Additional variables from these banks’ statements are
reported in Panel B of Table 3. For example, we find that almost half of all PPP borrowers have their
checking accounts at a Top-4 bank, while 16% have checking accounts at other large banks.
The bank statement sample and the full analysis sample are similar on many important dimensions,
such as loan amount, which is $80,987 in the bank statement sample and $93,666 in the analysis
sample. Bank statement-matched borrowers are more likely to be minority-owned. The main dimension
of selection is that firms matched to bank statement data have a much higher rate of fintech PPP
loans—36.3%, compared to 17.4% in the analysis sample. This reflects the fact that Ocrolus processes
loan applications for many fintech clients, and thus selects a sample of applicants with substantial
fintech affinity and experience.
We use text descriptions of transactions in the bank statements to identify credit relationships.
Specifically, we use the existence of a transaction to or from a lender to suggest a credit relationship of
some sort with this lender (e.g., a loan, credit line, or credit card payment). Among all borrowers, about
14.2% had a credit relationship with a fintech firm, while 80.0% had a credit relationship with a
traditional bank (Panel B of Table 3). Note that these credit relationships also include business credit
cards, and are thus much broader than other sources of data, such as UCC filings for secured debt. The
share of firms with access to external financing in the Ocrolus sample is naturally higher than in the
population of small businesses as reported by Alekseev et al. (2020), since Ocrolus usually only obtains
the bank statements for firms actively seeking external credit. There are no large differences by PPP
11
When a firm has statements from multiple banks, we identify the primary account as the one with the highest balance. Only
1% of firms have statements from multiple banks.
11
lender type in the propensity of firms to have prior credit relationships with a fintech or a traditional
lender (Table 1, Panel C). In some specifications below, we focus on the sample of firms with a history
of obtaining fintech credit, allowing us to rule out that any differential lending by fintech firms to
minority-owned businesses in this sample primarily reflects a higher familiarity or experience of
minority-owned businesses with fintech lenders in general.
We also use the bank statement data to calculate a firm’s monthly cash inflows and outflows as a
measure of firm financial performance. Panel B of Table 3 shows that mean net monthly cash inflow
across all firms is $9,124, while it is $6,332 among Black-owned businesses.
1.4 Card Revenue Data
To assess whether the real-time financial performance of small businesses helps explain our results, we
also gathered data from Enigma on monthly credit and debit card revenues. Enigma is a data analytics
company serving enterprise customers. Through a partnership with Verisk, a data warehouse that banks
employ to enable cross-issuer fraud checks, Enigma accesses real-time credit and debit card transactions
covering more than 60 banks, including all the major issuers. Their data include at least 60% of all U.S.
debit and credit card transactions.
About one million PPP borrowers were successfully merged between Enigma’s merchant identity
platform and the PPP loan data. Enigma provided monthly revenue data for these firms, which amounts
to over 70 million observations. For 813,812 of these firms, we observe revenue in the approval month
or the two months before (Enigma does not report these numbers if there were too few transactions in
a given month). We calculate average revenue across these months. Summary statistics are shown in
Panel D of Table 2 and in Panel D of Table 3. Notably, the Enigma-matched firms tend to be larger,
with a mean loan amount of $141,529 (compared to $93,666 in the main analysis sample). This reflects
Enigma being more likely to establish a merchant identity for firms that appear more frequently in their
card transaction data. Consistent with our previous measures for firm size—PPP loan amount and bank
statement cash inflows—the average card revenue for Black-owned firms is about half of that for the
other groups, at $23,169, compared to $42,557 for Hispanic- and Asian-owned firms, and $58,355 for
White-owned firms.
1.5 Other Data
Beyond the datasets described here, we use several further sources of data—including data on loan
applications through Lendio, and information on when small banks automated their lending procedures—
to explore the mechanism for our results. These data are described in tandem with the analyses in
Sections 3 and 4.
12
2 Who Lent to Minority Borrowers, and Why?
We begin this section by exploring whether certain types of lenders were more likely to extend PPP loans
to borrowers of a particular race or ethnicity. We then ask whether any disparities are explained by three
types of baseline characteristics that represent obvious hypotheses for lender differentiation: borrower
size, industry, and location. Finally, we examine the role of pre-existing banking and credit relationships,
firm financial conditions, and borrower application behavior in explaining any residual differences.
We first illustrate the main descriptive statistics. Panel A of Figure 1 shows the share of PPP loans
originated by each lender type made to Black-owned businesses (see also Panel A of Table 2). Only
3.3% of PPP loans originated by small banks went to Black-owned businesses. At large banks,
including the Top-4, Black-owned firms represent between 4% and 8% of loan recipients. At the top
end of the spectrum, CDFIs made 10.6% and fintech lenders made 26.5% of their loans to Black-owned
firms. CDFIs’ large share of lending to Black-owned businesses is consistent with their mission to
provide financial services to economically disadvantaged individuals within underserved communities.
The reasons for the large share of lending to Black-owned businesses by fintech lenders are not
immediately obvious as they do not, in general, have a mission to serve a particular racial or ethnic
group more than others. Appendix Table A.1 shows that while there is some variation across fintech
lenders in the share of PPP loans to Black-owned firms, this disparity is not driven by a few lenders.
Panel B of Figure 1 shows the share of all loans to Black-owned businesses by lender type (see also
Panel A of Table 3). Fintech lenders were responsible for 53.6% of PPP loans to Black-owned businesses
in our sample. We find similar results using only PPP loans with self-reported data on owner race (see
Appendix Figure A.2). We present the same figures for the other three racial and ethnic groups in our
data in Appendix Figures A.3-A.5. Key summary statistics for all the racial and ethnic groups are in
Table 3. Black-owned businesses exhibit by far the most striking disparity across lender types, especially
their propensity to obtain PPP loans from fintech firms instead of other lenders. The primary objective
of the rest of the paper is to assess a variety of plausible mechanisms for the observed disparities across
lender types in lending to Black-owned businesses.
2.1 Observable Loan and Borrower Characteristics
While the higher share of PPP loans to Black-owned businesses by fintech lenders is striking, it may
reflect the unique characteristics of those businesses. Black-owned firms receive the smallest PPP loans,
with a mean amount of $24,315, compared to about $54,000 for Hispanic- and Asian-owned firms,
and $110,317 for White-owned firms (Panel A of Table 3). Similarly, while 63.4% of all businesses
obtaining PPP loans are employer businesses, only 21.0% of Black-owned businesses obtaining PPP
loans are employers. These characteristics, in particular the differences by loan size, could explain some
of the differential lending rates. As Section 1 explains, lenders were compensated for originating PPP
loans with a fixed fraction of the loan amount. In the presence of fixed costs or capacity constraints, this
would incentivize lenders to first process the largest loans, which were disproportionately given to non-
13
minority-owned businesses. If capacity constraints or fixed costs were smallest among fintech lenders
(and largest among small banks), this could explain some of the differential lending rates across bank
types to Black-owned businesses.
In the next step, we explore the propensity for Black-owned businesses to obtain their PPP loans
from fintech or other lenders in a multivariate regression framework represented by Equation 1:
1(BankTypei ) = β 1(BlackOwnedi ) + Xi δ + εi . (1)
The dependent variable, BankT ypei , is an indicator for whether a PPP borrower gets their loan from a
certain type of lender. The key explanatory variable is an indicator for whether a firm is Black-owned, as
defined in the previous section. For example, when BankT ypei = F intechi , the coefficient β measures
the higher propensity (in percentage points) for Black-owned businesses of obtaining a PPP loan from a
fintech lender. The vector Xi captures control variables that vary across specifications.
Unconditionally (i.e., with no controls), we find that Black-owned businesses were 39.7 percentage
points more likely to obtain their PPP loan from a fintech lender (Column 1 of Panel A of Table 4).
In column 2, we include fixed effects for each percentile for the loan size distribution. Despite Black-
owned PPP borrowers having much smaller loans on average, controlling for loan amount explains only
a modest part of the disparity. Conditional on loan size, Black-owned businesses remain 31.6 percentage
points more likely to borrow from a fintech lender.
As many traditional lenders focus their operations on certain parts of the country and the Black
population is not evenly distributed (see Appendix Figure A.6), we also consider the possible role of
firm location in explaining the observed lending patterns.12 In column 3, we include firm zip code fixed
effects, and in column 4, we include firm census tract fixed effects. Including controls for the location
of the business has a sizable effect on the R2 of the regression. It also reduces the excess probability
that Black-owned businesses obtain PPP loans from fintech lenders to 23.2 percentage points (in the
model with zip code fixed effects), consistent with traditional lenders lending less in locations with more
minority-owned businesses (see Erel and Liebersohn, 2020). Indeed, in Appendix Table A.6, we verify
that all businesses located in areas with high minority ownership—even businesses in those areas that
are owned by White individuals—were somewhat more likely to obtain their PPP loans through fintech
lenders. An important question for future research is thus why traditional lenders are less likely to serve
people (of any race) in high-minority neighborhoods. This finding could, for example, be explained if
conventional lenders chose to locate fewer bank branches in high-minority neighborhoods, leading them
to process fewer PPP applications from local businesses in these areas. Overall, however, it appears that
lenders’ geographic focus can explain some, but not all (and not even the majority) of the racial gap in
the propensity to borrow from different types of lenders.
In the next step, we explore the role of the timing of the PPP application. Initially, only a few fintech
lenders were approved to participate in the program, with the rest entering in the subsequent months. If
12
While differently-sized loans have differential profitability to lenders, other characteristics, such as firm location, should
not affect the profitability of extending the federally-guaranteed PPP loans.
14
Black-owned businesses were more likely to obtain their PPP loans later in the program, then this may
have led to a natural relationship between race and the propensity to obtain PPP loans through a fintech
lender. Indeed, Figure 2 shows how the share of PPP loans made to Black-owned businesses (Panel A)
and the share of PPP loans made by fintech lenders (Panel B) both increased over time (see also Appendix
Table A.7). These patterns could either reflect a coincidence, whereby Black-owned businesses may have
only learned about and applied for the program in later periods when more fintech lenders were included,
or a causal relationship, whereby the absence of fintech lenders in the early stages of the program are part
of the reason why Black applicants struggled to obtain PPP loans during that period. After additionally
including “week of approval” fixed effects, the estimated β-coefficient remains at an economically large
17.5 percentage points.13 Therefore, a coincidence in timing cannot explain the majority of the disparity.
In columns 6 to 8 of Panel A of Table 4, we explore the role of industry, business type, and employer
status.14 Importantly, column 7 includes zip-by-industry interacted fixed effects, in case lenders perceive
certain industries in certain areas differently. In our most richly controlled model, the unexplained excess
share of loans to Black-owned businesses by fintech lenders is 12.1 percentage points (column 8). This
represents 70% of the mean chance of a fintech loan (throughout the paper, we report the mean of the
dependent variable towards the bottom of the table).15
In sum, fintech lenders made a substantially larger share of their loans to Black-owned businesses
than traditional lenders did, a disparity that does not appear with similar magnitude for other racial or
ethnic groups. Controlling for a rich set of observable loan and borrower characteristics can jointly
explain 69% of the higher chance that Black-owned businesses borrow from fintech lenders (.69 =
.397−.121
.397 ). Since many of the firm characteristics are correlated, it is challenging to attribute a relative
importance to each of these factors, and changing the order in which we include these controls in Table 4
changes their relative contribution to reducing the estimate of β. The key point for our analysis, however,
is that 31% of the gap remains unexplained, even in models that hold the profitability and timing of
lending and the location and characteristics of the firm fixed. Exploring possible explanations for this
remaining gap is the central contribution of our paper.
Across-Bank Heterogeneity. Given that Black-owned businesses are more likely to get their PPP loans
from fintech lenders, which types of lenders does this substitution come from? We explore this question
13
In Appendix Table A.8, we estimate Equation 1 separately for loans approved in each of the four phases of the PPP
program. This is also shown in Appendix Figure A.7. The main takeaway from Table 4 is consistent across rounds: fintech
lenders made a much larger share of their loans to Black-owned businesses compared to traditional lenders.
14
Borrower industry is captured with NAICS 3-digit industry fixed effects. Examples of industries in this classification
scheme are “Health and Personal Care Stores,” “Truck Transportation,” and “Food Services and Drinking Places.” Table 3
Panel A shows that business type and industry distribution differ by owner race. For instance, Black-owned businesses are
substantially more likely to be self-employed or sole proprietorships and less likely to be corporations or LLCs.
15
Appendix Table A.9 repeats this exercise but restricts to the sample of employer firms only. Although the magnitude of the
coefficients decreases somewhat, they remain robust and are larger relative to the mean chances of a fintech loan. With our full
set of controls in column 8, Black-owned employer firms are 7.5 percentage points more likely to get a fintech loan, which is
77% of the mean. We repeat these models using self-reported race in Panel A of Appendix Table A.12. The sample is restricted
to the subset of roughly one million loans for which race/ethnicity is reported in the SBA data. In this smaller and selected
sample, the results are quite similar to the main findings, though the magnitudes are somewhat larger.
15
in Panel B of Table 4, where we replace BankT ypei in Equation 1 with an indicator for obtaining
the PPP loan from a Top-4 bank (Bank of America, Wells Fargo, Citi, JP Morgan), a large bank, or a
small/medium bank. Unconditionally, Black-owned businesses were less likely to get their loans from
all of these types of banks, consistent with the findings in Table 2. However, after including controls,
this relationship is near-zero for the Top-4 banks, and the majority of the 12.1 percentage point fintech
disparity in column 8 of Panel A of Table 4 is accounted for by lower rates of small/medium bank PPP
lending to Black-owned businesses. Unconditionally, Black-owned businesses were 30 percentage points
less likely to get their PPP loans from a small bank (column 5); in the model with full controls (column
6), this gaps remains an economically large 8.1 percentage points.16,17 The results using self-reported
race in Appendix Table A.12 Panel B are similar.
The heterogeneity across different conventional banks helps assess the potential role of compliance
with Know-Your-Customer and Anti-Money-Laundering (KYC/AML) regulations in explaining the
differential lending to Black-owned businesses. The Top-4 banks, as international institutions with
charters in all or most U.S. states, are the most tightly regulated banks (Congress, 2009). Consequently,
the fact that, conditional on controls, the most-regulated and least-regulated entities in our sample—the
Top-4 banks and the fintech lenders—had the highest probability of granting PPP loans to Black-owned
businesses makes differential compliance standards an unlikely explanation.18
We are also interested in understanding whether small banks’ low rates of lending to Black-owned
businesses reflect higher rates of lending to all other races or to one race in particular. Panel C of Table 4
includes indicators for the three other races and ethnicities as explanatory variables. Since the four racial
and ethnic groups span the data, the coefficients should be interpreted as the probability of obtaining
a PPP loan from a particular lender type relative to the probability for Black-owned businesses. After
including controls, there is zero difference between Black- and White-owned businesses in the chances
of getting a PPP loan from a Top-4 bank (column 4), while Asian- and Hispanic-owned businesses were
marginally more likely to get their PPP loan from a Top-4 bank. In contrast, on a percentage point
basis, small- and medium-sized banks lent particularly more to White-owned businesses (columns 7-8).
For example, unconditionally, White-owned businesses are 35 percentage points more likely than Black-
owned businesses to get their PPP loans from smaller banks (column 7). With controls, this gap remains
an economically large 10 percentage points, and we continue to find similar results using self-reported
race (Appendix Table A.12 Panel C) or excluding fintech loans (Appendix Table A.10 Panel B).
16
Note, however, that the 2.5 percentage points effect identified for non-Top-4 large banks corresponds to a larger share of
their unconditional mean probability of originating any loan, at 26%, than the 8.1 percentage points effect at medium and small
banks, at 17% of the mean.
17
We visualize the across-lender patterns by comparing all 11 types of lending institutions simultaneously in Figure 3. Here,
we show the degree to which the lender types were statistically different from one another in their propensity to lend to each
of the four racial and ethnic groups, conditional on our rich array of controls. The fraction of fintechs’ loans to Black-owned
businesses was over five percentage points higher than the fraction for other lender types. MDIs made a disproportionate share
of their loans to Asian-owned businesses. Note that the reversal for MDIs in Hispanic loans relative to the summary statistics
reflects the location control; in particular, a very large MDI in Puerto Rico.
18
To underline this cross-bank heterogeneity independent of fintech loans, Appendix Table A.10 Panel A repeats Panel B of
Table 4 but excludes fintech loans. The results show that relative to all remaining lenders, Top-4 banks are significantly more
likely to lend to Black-owned businesses (columns 3-4) while smaller banks are much less likely to do so (columns 7-8).
16
Strength of Algorithm’s Race Signal. In our baseline analysis, we assign every individual the race
with the highest probability from our machine learning algorithm. We next explore whether, among
individuals predicted to be Black, the strength of the race signal correlates with variation in PPP lender
type. Specifically, we divide all individuals predicted to be Black into five quintiles based on the
algorithm’s predicted probability that they are Black (see Section 1.2 for the full distribution of this
probability). Individuals in the 5th quintile deliver the strongest signal that they are Black to lenders
given their name and location.19
Table 5 shows that the magnitude of our main results increases monotonically with the signal
strength. As in Panels A and B of Table 4, the omitted group is all borrowers not predicted to be Black.
Panel A shows that across the quintiles of increasing probability that the business owner is Black, the
chance of having a fintech loan increases. In the fully controlled model in column 8, Black-owned
businesses in the first quintile are only 4.3 percentage points more likely than other groups to have a
fintech loan. In the third quintile, they are 12.7 percentage points more likely, and in the fifth quintile,
they are 24.7 percentage points more likely. Similarly, Panel B explores the probability of obtaining
loans from conventional lenders, again with the most striking differences among small- and
medium-sized banks (columns 5-6). Conditional on controls, borrowers in the first quintile are 3.1
percentage points less likely to get a small/medium bank loan, while those in the fifth quintile are 15.2
percentage points less likely to do so. In sum, these results help to validate the algorithm and point to a
potentially important role for the degree of “Blackness” as suggested by the audit literature (Bertrand
and Mullainathan, 2004).
2.2 Borrower Bank Relationships and Financial Situation
In this section, we explore the importance of two channels that may help explain why the propensity
of Black-owned businesses to obtain PPP loans varies across lender types: differences in pre-existing
banking and credit relationships, and differences in firm financial health.
Pre-existing Banking Relationships. A number of observers of the PPP program have highlighted
that many banks tended to first serve their own clients’ PPP loan applications, which may have distorted
allocations away from the government’s intended “first come, first serve” approach (Cowley, 2021;
Rosenberg and Myers, 2020; Li and Strahan, 2020).20 If banks indeed prioritized administering PPP
loan applications from their own clients, and if Black-owned businesses did not bank with active PPP
19
In the self-reported data, individuals in the 3rd, 4th, and 5th quintiles of probability Black have more than a 99% chance
of self-identifying as Black (recall that these quintiles are calculated within the population for whom Black has the highest
probability across the races). The 1st quintile individuals have a 78% chance of being Black, and the 2nd quintile individuals
have a 97% chance.
20
There are a number of explanations for prioritizing PPP applications from existing customers. First, it may have been
cheaper to process these applications. Second, this might be optimal for banks if receiving a PPP loan increases the chances that
a borrower repays existing loans, including possible loans to the PPP lender (Granja et al., 2020). Third, even in the absence of
an existing credit relationship between banks and their clients, banks might prioritize existing clients if they perceived a positive
net present value from the relationships, and receiving a PPP loan would improve the chances of those clients surviving.
17
lenders, this could explain some of the observed differences in their propensity to eventually borrow
from other lenders such as fintech firms.
To quantify the effect of pre-existing banking relationships, we turn to the sample of PPP borrowers
matched to bank statement data. Indeed, we find that conventional banks’ PPP clients were also often
their business checking account clients. Panel C of Table 1 shows that 28.5% of PPP borrowers had a
checking account at their PPP lender; of the PPP loans originated by large banks and the Top-4 banks,
more than 50% went to existing checking account customers. For fintech lenders, which do not usually
offer checking accounts, this number was essentially zero.21
Although conventional lenders served their own clients at higher rates, we show in Table 6 that this
fact does not explain the higher rate of fintech PPP loans for Black-owned businesses. First, we estimate
the fully controlled model from Table 4, Panel A, column 8, in the bank statement-matched sample.
Here, Black-owned businesses are 5.5 percentage points more likely to obtain their PPP loan from a
fintech lender (column 1). Selection into having a checking account could help explain the smaller effect
size; that is, the larger gap in the full sample could reflect Black-owned businesses being less likely to
have any banking relationships, and therefore less likely to have a “house bank” prioritizing their PPP
loan application. However, we do not find this explanation compelling. This story would suggest that
Black-owned firms should be a smaller share of the bank statement-matched sample than the full sample,
because the former obviously requires the firm to have a checking account. As shown in Panels A and
B of Table 2, this is not the case; in fact, Black-owned firms even represent a somewhat larger share of
the bank statement-matched sample. Instead, we believe that the somewhat smaller magnitude of the
baseline effect in the bank statement-matched sample likely reflects a higher rate of clients with fintech
affinity, which would raise the share of fintech PPP loans among borrowers of all races and ethnicities.
In Column 2 of Table 6, we add fixed effects for the identity of the bank where the firm has a
checking account. In this model, we are comparing, for example, the origination of PPP loans to Black-
owned firms and other firms with a checking account at JP Morgan Chase. The inclusion of these fixed
effects has essentially no effect on the estimated probability of Black-owned businesses to obtain their
PPP loans from fintech firms. Therefore, the observed racial difference in this probability is not driven
by Black-owned firms holding their checking accounts at banks that were less active as PPP lenders.
We next assess whether, after controlling for observable characteristics, there are racial differences in
the propensity of a firm to obtain its PPP loan from its checking account bank. On average, Black-owned
businesses were between 1.1 and 1.7 percentage points less likely to obtain their PPP loans from their
checking account banks (Table 6, Panel B, columns 7-8).22 In Table 7, we split the sample of checking
21
Table 3 Panel B shows that within the bank statement-matched sample, Black-owned businesses are the most likely, at
17.2%, to have their checking account with a lender that is not a traditional bank. Black-owned businesses have similar or
slightly higher chances of having a credit relationship with another non-fintech lender in this sample, though this could reflect
business credit cards which represent a more arms-length relationship than traditional small business loans.
22
Anecdotes from the popular press offer examples of Black-owned businesses failing to obtain PPP loans through their
checking account banks. For example, the Associated Press interviewed Lisa Marsh, who is Black and the owner of MsPsGFree,
a Chicago-based gluten-free baking business (Rosenberg and Myers, 2020): “Lisa Marsh tried in vain to get banks to process
her application. She first applied in June but she couldn’t get answers on her status from her bank, a subsidiary of a big national
bank. She also got nowhere with smaller community banks... [Marsh] finally applied through an online lender in late July and
18
account holders by the identity of the checking account bank. Among firms with checking accounts at
Top-4 banks (Panel A), we find no differential probability of Black-owned businesses to obtain their PPP
loans from their checking account bank. The negative coefficient in the full sample stems from large
banks (Panel B), with a smaller relationship for smaller banks (Panel C).
In columns 2-5 of Table 7, we show that Black-owned firms’ differential chance of getting a fintech
loan versus a non-Top-4 bank loan is similarly large regardless of where they have their checking account,
and exists even for borrowers with checking accounts at Top-4 banks. This difference is largely explained
by variation in the identity of the PPP lender among those firms that do not obtain their PPP loan from
their checking account bank.
Overall, these findings highlight two channels that made it more likely for Black-owned businesses
to obtain their PPP loans from fintech lenders. First, Black-owned firms with checking accounts at non-
Top-4 banks were somewhat less likely to obtain their PPP loans from their checking account bank.
Second, among firms that obtained PPP loans outside their checking account banks (which includes both
Black-owned and non-Black-owned firms), Black-owned firms were much less likely to obtain loans
from non-Top-4 banks, and much more likely to obtain them from fintech lenders. Quantitatively, this
second channel, which captures racial differences in the rates of establishing new banking relationships
with different types of lenders, explains the majority of the observed disparity.
Pre-existing Credit Relationships. We next explore the possible role of prior credit relationships in
explaining the observed differences in PPP lending outcomes. Specifically, in column 3 of Panel A of
Table 6, we include indicators for whether a PPP borrower has credit relationships with any fintech and
conventional lenders. Unsurprisingly, a prior credit relationship with a fintech lender is associated with a
significantly higher chance of obtaining a PPP loan from a fintech lender. At the same time, columns 2,
4, and 6 of Panel B show that such a prior relationship reduces the likelihood of getting a loan from all
types of traditional banks. Similarly, having previously received credit from a non-fintech lender reduces
the likelihood of getting a fintech PPP loan and increases the probability of a non-fintech PPP loan.
The preferential treatment of firms with prior credit relationships, however, does not account for the
disproportionate lending to Black-owned businesses by fintech lenders in the PPP program: Black-owned
businesses are 5.6 percentage points more likely to get their PPP loan from a fintech lender compared
to other PPP borrowers, even after conditioning on the identity of the checking account bank and the
presence of credit relationships with both fintech and non-fintech lenders (column 3 of Table 6). This is
largely because, within the bank statement-matched sample, there are no substantial differences across
races in having various credit relationships (see Table 3, Panel B).
got her loan a few days before the PPP ended. “I was very frustrated and almost gave up,” she says.” In a similar story, the New
York Times described Black auto dealership owner Jenell Ross who, “sought a Paycheck Protection Program loan, [but] her
longtime bank told her to look elsewhere” (Cowley, 2021). In the absence of formal statistical analysis like the one we provide,
it is obviously impossible to say whether such challenges were disproportionately encountered by Black-owned businesses.
19
Firm Financial Health. We next explore whether the differential financial performance of minority-
owned firms can explain the remaining observed racial disparities in PPP lending. Recall that PPP
lenders were not responsible for any loan losses; therefore, creditworthiness should be irrelevant to the
one-shot decision of making a PPP loan. However, lenders may nonetheless have been more likely to
lend to businesses in better financial positions, for example, because they might represent more attractive
future customers. There could also be some stickiness in the behavior of loan officers who are used
to screening for creditworthiness, causing them to screen on financial signals as they do in their usual
course of business.
Panel B of Table 3 shows that gross and net inflows at Black-owned businesses are, at the median,
less than half of their values for the other racial groups. To assess whether these differences affect our
results after conditioning on firm size, we first add granular controls for a firm’s gross and net cash
inflows from the most recent bank statement in column 4 of Panel A of Table 6. Black-owned businesses
remain 5.5 percentage points more likely to receive a fintech loan, suggesting that the cash-flow situation
of borrowers does not explain the observed racial differences in the identity of the PPP lender.23
We also assess whether differences in contemporaneous revenue explain the observed disparity. If
lenders treated Black-owned firms differently because they were doing relatively poorly during the
pandemic, controlling for revenue during this period should eliminate the disparity. Our data on credit
and debit card spending from Enigma allows us to observe the firm over the course of the COVID-19
economic crisis and PPP loan application period. We use these data to control for revenue around the
time of application. We show results using the card revenue-matched sample in Table 8. We continue to
include all controls available in the PPP data.
Given the sample differences highlighted in Section 1.4, we first re-estimate the main model to
establish a baseline. Column 1 of Table 8 shows that, in this sample, Black-owned firms are 1.6
percentage points more likely to get a fintech PPP loan—this effect corresponds to about 17% of the
mean probability. In this sample, Black-owned firms are slightly more likely to get their PPP loans from
Top-4 banks (columns 3-4), and substantially less likely to get PPP loans from small- and medium-sized
banks (columns 7-8). Importantly, adding fixed effects for 100 equal-sized groups of credit card revenue
has no effect on these racial disparities, consistent with our findings when we controlled for cash flows
observed in the bank statement data. This pattern continues to hold when we measure revenue only in
the approval month (Appendix Table A.14).24 Hence, while the Enigma sample generally includes
larger and more consumer-oriented firms among which Black-owned businesses appear to be at less of a
disadvantage, we find no evidence that real-time revenue differences help explain the main findings.
23
We repeat Table 6 with indicators for whether a business is Asian-, Hispanic-, and White-owned in Appendix Table A.11.
We continue to find similar results. For example, these regressions show that, after controlling for bank and credit relationships
as well as cash flows, small banks are significantly more likely than other non-fintech lenders to serve White-owned businesses.
24
In Panel B of Table 8, we restrict the sample to firms that appear especially harmed by the COVID-19 economic crisis.
These are firms for which we observe monthly revenue in February 2020 but not in the approval month. Note that we do not
observe revenue if there are fewer than 30 transactions. Therefore, these “struggling” firms have either no activity or limited
activity relative to February. We repeat our main models from Table 4 within this sample of “struggling” firms. In the model
with full controls, and among this sample of struggling firms, Black-owned firms are 2.1 percentage points more likely to get
their PPP loan from a fintech firm.
20
3 Borrower Application Behavior
Our primary data includes only originated loans and does not contain information on loan applications.
This leaves open the possibility that the differential rate at which Black-owned businesses end up
borrowing from different types of lenders is the result of their application behavior. In this section, we
therefore explore whether the observed equilibrium lending outcomes might in part be explained by
Black-owned businesses being more likely to apply to fintech lenders than to non-Top-4 banks.
3.1 Differential Fintech Affinity.
One reason Black-owned businesses may have been more likely to apply to fintech firms is that they
may have been more tech-savvy or have had higher fintech affinity.25 To test this hypothesis, in Panel
D of Table 7, we condition on firms in the bank statement-matched data that we observe having a pre-
existing credit relationship with fintech firms. Within this sample of firms, all of which have shown
a certain degree of past fintech affinity, we continue to find an economically important substitution of
PPP borrowing of Black-owned businesses from small and medium banks towards fintech lenders. This
finding reduces the likelihood that the results are driven by an increased fintech affinity of Black-owned
businesses.
3.2 Analysis within PPP Application Data from Lendio.
For some borrowers who applied for a PPP loan through Lendio, a marketplace platform for small
business loans, we have information on the set of lenders to which Lendio forwarded the application.
Lendio is a platform through which businesses could submit PPP applications, which were then
forwarded to around 300 partner lenders. Some of these Lendio partner lenders were fintech firms, and
some were conventional banks. Our conversations with Lendio executives, including CEO Brock Blake,
suggest that the routing of the applications to lenders was random conditional on loan size, geography,
and capacity criteria set by the lender partners. Banks then had the opportunity to follow up with the
applicant firms to complete the application (firms also retained the option to apply for a PPP loan
through other lenders). In these data, we can assess whether Black-owned businesses remain more
likely to ultimately get a fintech loan conditional on the types of banks Lendio forwards these loans to.
We obtained data on all PPP loan applications through Lendio up until November 2020. The data
include applications from about 267,000 firms, about 176,000 of which we can link to final PPP loans
in our data. The remaining applicants either never got a PPP loan, or we were unable to match them
to firms in the main dataset. In our analysis, we focus on the sample that ultimately received PPP loans
since this is the sample for which we can generate our name-based race proxy. We divide the lenders who
25
A different possibility within the scope of application behavior is that because Black communities were disproportionally
affected by COVID-19, and because banks are more associated with in-person interaction than fintechs, Black business owners
were more likely to apply to fintechs in order to avoid in-person interaction. In unreported analysis, we find that COVID-19
prevalence cannot explain the main finding. Even in areas with low COVID-19 cases and deaths, Black-owned businesses were
substantially more likely to receive fintech PPP loans.
21
ultimately originated the loans into two groups: fintech and conventional lenders (where conventional
includes all other lender types). The main sample characteristics according to this split are summarized
in Panel C of Table 2. About half the PPP loans to firms that applied through Lendio were originated
by fintech lenders, a result that is unsurprising given that Lendio is an online platform. The share of
Black borrowers across conventional and fintech lenders, as well as the average difference in loan size,
however, is quite similar to the main samples. Panel C of Table 3 shows that the average application was
routed to 1.5 lenders, of which 0.9 were fintechs and 0.6 were conventional lenders.
In Table 9, we again present results from Equation 1. As before, we begin by establishing a baseline
by replicating the result from column 8 of Panel A of Table 4 within the Lendio sample. Column 1
of Table 9 shows that, conditional on our controls, Black-owned businesses in this sample were 2.9
percentage points more likely to obtain their PPP loan from a fintech lender. The somewhat smaller
disparity compared to our baseline estimate likely reflects the composition of lenders that Lendio partners
with. To show this, column 2 explores a sample of PPP loans originated by Lendio partners, but to firms
that did not apply through Lendio (thereby removing any effects that might occur due to firm selection
into the Lendio data). The coefficient of 2.6 percentage points is very similar to the previous column,
indicating that the smaller baseline disparity likely reflects lender composition.
Next, in column 3, we add interacted fixed effects for the number of lenders to which Lendio routed
the application in both the fintech and conventional categories. The disparity remains at 2.8 percentage
points, suggesting that differential application behavior by Black-owned businesses has at most a small
role in explaining the lending patterns by race documented in this paper. Instead, this result points to
lender decision-making playing a role. In Column 4, we restrict the sample to loan applications that were
sent to at least one conventional lender, while Column 5 focuses on applications that were sent only to
conventional lenders. Both specifications control for the number of lenders the application was sent to in
each category. These models continue to find that Black-owned businesses were more likely to ultimately
have a fintech loan. For example, the estimate in column 5 suggests that Black-owned businesses are 3.9
percentage points (36% of the mean) more likely to get a fintech loan than non-Black-owned businesses,
even when Lendio only sent their application to conventional lenders. These Black-owned firms obtained
fintech PPP loans outside of Lendio.
Overall, our results in this section suggest that racial differences in the probability of applying to
conventional lenders versus fintech lenders cannot explain our findings and that, instead, different
behavior across lender types is likely to explain why Black-owned businesses receive fintech loans at
substantially higher rates.
4 Discrimination: The Role of Racial Animus and Automation
In this final section, we explore whether preference-based discrimination may contribute to explaining
some of the remaining disparity between fintechs and smaller conventional lenders in their propensity to
provide PPP loans to Black-owned businesses. This channel would involve potential biases among loan
22
officers during the manual review and processing of PPP applications. Such manual involvement of loan
officers was relatively common among conventional lenders, especially among smaller and medium-
sized banks. For example, one industry article profiled the approach of The Piedmont Bank in Georgia,
a small SBA-preferred lender:
“While The Piedmont Bank considered some automated, online solutions, they ultimately
decide to process the applications manually...Everyone who works there is preparing to put
in long hours and a lot of elbow grease. They know they’re going up against big banks and
their automated systems.” (Smith, 2020)
As a second example, an industry article mentions that:
“In the initial round of the Paycheck Protection Program, First Bank in Hamilton, N.J.,
leaned on its bankers rather than technology to help small businesses stay afloat. That
manual labor “ironically turned out to be a good thing, because we had people helping
small businesses through the process, and they had a number and name to talk to,” said
Patrick Ryan, president and CEO of the $2.3 billion-asset bank.” (Cross, 2021)26
When reviewing PPP loan applications, loan officers may become aware of applicant race through a
number of channels. One is visually through manual review of driver’s licenses, which were required
(in color) for all PPP applicants. A second channel is through information such as the applicant’s name,
which we have shown to be highly predictive of race.27
If preference-based discrimination contributed to the observed higher probability of otherwise similar
Black-owned businesses obtaining a PPP loan through a fintech lender rather than a non-Top-4 bank, we
would expect this gap to be larger in regions with higher racial animus. We next explore this hypothesis
and find that Black-owned businesses were indeed more likely to get their PPP loans from fintech lenders
(and less likely to get them from small- or medium-sized banks) in regions with higher racial animus. We
also study the lending behavior of small banks that automated their lending processes during our sample
period, and explore whether following the automation, there were significant changes in their lending to
Black-owned businesses.
4.1 Racial Animus Data.
We collect six measures of racial animus at the local level. The first measure comes from
Stephens-Davidowitz (2013) and is calculated at the level of the designated media market. It measures
the percentage of an area’s Google searches that contain racially charged words. The second measure
26
As a third example, RCB Bank of Oklahoma and Kansas did not even allow borrowers to apply online for PPP loan
forgiveness, but instead stated that “A Loan Officer will contact you to discuss your forgiveness eligibility and provide you
with the appropriate loan application via DocuSign” (see RCB Bank Example). As a fourth example, Skip advised customers
that “While many lenders spent the time to automate processes and increase their throughput, many have not and may be doing
manual review” (see Skip Example).
27
Most Americans can infer race for a large fraction of names, perhaps not with the accuracy of our algorithm, but well
enough to lead to systematic bias (Bertrand and Mullainathan, 2004; Milkman et al., 2012; Bartoš et al., 2016).
23
follows Bursztyn et al. (2021) and is based on how favorably White respondents rate Black Americans
as a group in the Nationscape survey; individual responses are aggregated up to the congressional
district level (Tausanovitch and Vavreck, 2020). The third measure of racial animus is based on the
Implicit Association Test (IAT), which measures implicit bias against Black individuals. The fourth
measure is based on a survey question that explicitly asks individuals who just took the IAT for their
feelings towards Black Americans. These IAT-based measures are aggregated up to the county-level
(Xu et al., 2014). The last two measures of racial animus are based on the extent of local residential
segregation (Massey and Denton, 1988). The first of these, the dissimilarity index, measures how
similar the distribution of White and Black residents are across city tracts. The second is the isolation
index, which measures the probability of a Black resident sharing the same city tract with another Black
resident. The segregation measures are available at the metro/micropolitan statistical area (MSA) level.
Appendix A describes these measures of racial animus in more detail, and examines their geographic
variation and the degree to which they are correlated with one another. Importantly, Appendix Figure
A.9 shows that the places where racial animus is high differ substantially across our measures,
indicating that they offer somewhat independent signals of animus.
4.2 Differential Effects by Racial Animus.
Table 10 estimates whether, for a Black-owned firm, the probability of obtaining a PPP loan from
different lenders varies with the degree of racial animus in the firm’s location. In Panels A, B, and C,
the dependent variables are indicators for obtaining a PPP loan from a fintech, a Top-4 bank, and a
non-Top-4 bank, respectively. In each panel, column 1 includes the same controls as the specification in
column 8 of Panel A, Table 4. In columns 2-7, we interact our indicator for Black-owned businesses
with each of the proxies for racial animus. The location fixed effects absorb any direct effect of racial
animus on the probability of borrowing from fintech lenders that is constant across all borrowers. Each
measure of racial animus is standardized to have a mean of zero and a standard deviation of one, so the
coefficients can be interpreted as the effect, in percentage points, of a one standard deviation increase in
the racial animus measure.
In Panel A of Table 10, we consider the effects of increased racial animus on the probability of
a Black-owned firm obtaining a PPP loan from a fintech lender. We find a robust positive interaction
between the various racial animus measures and Black ownership. The coefficient magnitudes vary
from 0.4 to 2.9 percentage points. This implies that, relative to the mean chances of a fintech loan of
17.4%, a one standard deviation increase in racial animus is associated with a 2.3% to 17% increase
in the probability that a Black-owned business obtains their PPP loan from a fintech lender. With the
implicit bias (IAT) measure—which is probably the most widely used in the academic literature—the
coefficient estimate of 1.3 percentage points implies a 7.5% increase. In sum, while the magnitude of the
relationship varies somewhat across measures, we find robust evidence that in areas with higher racial
animus, Black-owned businesses are more likely to obtain their PPP loans from fintech lenders.
We repeat this exercise for the Top-4 banks in Panel B. Here we see near-zero coefficients on the
24
interaction, which flip signs across the animus measures. This is consistent with our previous analysis
finding no racial differences in the conditional probability of obtaining a PPP loan from a Top-4 bank.
In contrast, we estimate a robust negative interaction term for non-Top-4 banks in Panel C, suggesting
that higher racial animus makes Black-owned firms less likely to receive their PPP loans from those non-
Top-4 banks. The coefficient using the implicit bias (IAT) measure of -2.4 percentage points implies that
a one-standard-deviation higher racial animus score is associated with a 4.2% decrease in the probability
of Black-owned businesses obtaining their PPP loans from non-Top-4 lenders. In Appendix Table A.15,
we repeat these three panels within the bank statement-matched sample and find similar results, even
after controlling for the identity of the checking account bank and firm financial performance.
4.3 Automation as a Mechanism.
A number of our previous results suggest that the degree of automation in the lending process may help
explain some of the disparities in the propensity to serve Black-owned businesses. First, fintechs, which
have automated processes at the core of their business models, made a disproportionately large share of
their loans to Black-owned businesses. Second, among conventional lenders, smaller banks account for
most of the lower rate of PPP lending to Black-owned businesses, particularly in models that control for
observable borrower characteristics. In contrast, after including controls, we find no such differences in
the probability of borrowing from Top-4 banks. While we do not have systematic data on the degree
of automation by bank, automation is widely believed to increase in bank size and to be particularly
widespread at the very largest banks. Survey and anecdotal evidence support this narrative. One industry
article notes that “Large banks have avidly adopted robotic process automation . . . It’s tougher for smaller
banks to follow suit” (Crosman, 2020). A 2018 Fannie Mae survey found that 76% of large banks but
only 47% of small banks were familiar with artificial intelligence or machine learning technology. At
smaller banks, humans—with all their biases—generally play a larger role in the loan origination process.
As a colorful example, Cross (2021) explains:
“In community banking, when you’re closing a loan, you’re probably closing it with a lady
or gent you went to high school with, maybe on the hood of a Cadillac at a Friday night
football game or Sunday after church. Those things are nice, but they don’t scale.”
In this last section, we further study the effect of automation by exploiting the fact that, during the
PPP loan period, a number of small banks automated their loan origination processes, many of them by
collaborating with fintech software providers such as Numerated and Biz2Credit. For example, one bank
official attested that: “Compared to 10 days of manual lending, with every bank resource that we had, in
terms of volume of new loans generated, we were able to do it in 2 days with [Numerated].” As a second
example, Cross (2021) explains that:
“When HV Bancorp in Doylestown, Pennsylvania, first went live with the Paycheck
Protection Program last April, “we just had bodies in front of keyboards using the Small
Business Administration’s E-Tran system and entering applications,” said Hugh Connelly,
25
chief lending officer in the business banking division of Huntingdon Valley Bank...The
urgency of the Paycheck Protection Program propelled community banks to find a speedier
way to disburse loans to small businesses than relying on phone and email. Many turned to
software to originate loans, automate the underwriting process, collect documents and
transmit the information to the SBA’s processing system.”
Our formal analysis of the effects of automation on lending patterns proceeds in two steps. We first
identify a set of small banks that automated their lending processes during the PPP period. We then
show that their rate of lending to Black-owned businesses increased disproportionately following the
automation compared with other small banks that did not automate their lending processes.
Data. We gather data from the fintech firm Biz2Credit on small banks that—motivated by the influx
of PPP loan applications—hired Biz2Credit to automate lending during the second round of the PPP.
Biz2Credit offers a white label SaaS product called Biz2X that banks can license to outsource and
automate their loan processing and underwriting. Once a bank automates using Biz2X, loan application
materials are automatically redirected from the bank’s website to Biz2Credit. Biz2Credit then conducts
fraud checks and ensures compliance with PPP eligibility rules, makes a decision, and forwards the
required materials back to the bank to originate the loan.28 Biz2Credit provided us with the launch date
of their service for their clients during the PPP. We matched these clients to our list of banks in the SBA
data. We also manually searched newspaper articles to identify additional automating small banks. The
banks that are not Biz2Credit clients automated via a range of other fintech service providers, including
Customers Bancorp, Numerated, and Fountainhead. For some of these manually identified banks, we
only have a rough date of automation, potentially creating some noise in our estimation.
Appendix Table A.16 shows summary statistics for the automating banks as well as for the control
group of all other small banks.29 After removing Biz2Credit clients that we do not classify as small banks
or that have automation dates after our analysis period, we are left with 20 small banks that automated
during the sample period; those banks account for about 75,000 PPP loans in our analysis sample, or
3.8% of all PPP loans originated by small banks. Among automating banks, about half of their total
loans occur after automation. Automating banks are somewhat larger than the average small bank.
The Effect of Automation. Appendix Table A.16 shows that, on average, automating banks’ share of
loans to Black-owned businesses increased after automation, from about 4.1% to about 11.2%. However,
this trend could, at least in part, reflect the overall increase in loans to Black-owned businesses over time.
To identify a treatment effect of automation, we therefore use a standard differences-in-differences model
to evaluate the effect of automation on treated banks relative to other small banks that never automated:
28
For non-PPP marketplace loans, there is a more complicated underwriting algorithm in which Biz2Credit applies additional
data and a more complicated AI-based algorithm to make determinations.
29
It is possible that some of the small banks in the control group also automated at some point before, during, or after the
PPP, and we were unable to identify this via our Biz2Credit data and manual search efforts. Any misclassification of banks into
the control group will likely bias us against finding an effect.
26
1(BlackOwnedi,b,t ) = αb + αt + β 1(PostAutob,t ) + Xi δ + εibt . (2)
We estimate Equation 2 only within the sample of small banks. The dependent variable is an indicator
whether PPP loan i made by bank b at time t is to a Black-owned business. Here, 1(PostAutob,t ) is an
indicator for bank b having automated—i.e., having started service with Biz2Credit or another white label
fintech—as of date t. αb is a fixed effect for the bank, allowing us to remove the effects of any baseline
differences in lending to Black-owned businesses between automating and non-automating banks; αt is a
fixed effect for the week of loan approval, removing the effects of any general time trends in the share of
loans by small banks to Black-owned businesses. Xi represents other loan-level controls. The coefficient
of interest β represents the effect of automation.
We report results from estimating Equation 2 in Table 11. In column 1 of Panel A, we report the
results of the baseline differences-in-differences model controlling only for bank and time fixed effects.
The coefficient implies that the share of loans to Black-owned businesses increased by six percentage
points after automation, relative to a pre-automation share of 4.1%. In column 2, we add the same set
as of controls as in Table 4, Panel A, column 8. The coefficient remains an economically large and
statistically significant 4.3 percentage points: automation more than doubled small banks’ propensity to
lend to Black-owned businesses, even conditional on other firm and loan characteristics.30 In columns
3-5 of Table 11 Panel A, we consider how lending to the other races or ethnicities was affected. We see
a small increase in the rate of lending to Hispanic- and Asian-owned firms, suggesting some additional
benefit of automation for other minorities. As in our previous analyses of small bank lending, it appears
that most of the increase in lending to Black-owned firms came at the expense of White-owned firms.
One possible concern with the empirical design in Equation 2 is that the automating banks may have
already been increasing their share of loans to Black-owned businesses prior to the automation event.
Indeed, a bank’s decision to automate could have been driven by a desire to serve more Black-owned
businesses. To assess this concern, we use the following dynamic differences-in-differences specification
to examine whether, in the period prior to automation, we see a differential trend in the share of loans
extended by automating small banks to Black-owned businesses:
n
1(BlackOwnedi ) = βt 1{τ = n} + αb + αt + Xi δ + εibt .
X
(3)
−n
In Equation 3, τ denotes the event month, so that τ = 0 for loans made in the month the bank automated,
τ = 1 for the first month after the bank automated, and so on. Loans with τ ≤ −1 were made prior
to the time when an automating bank automated. The coefficients are relative to the omitted period,
τ = −1, which represents the month prior to automation. The model also includes fixed effects for bank
and origination period, as well as other controls defined previously.
30
In our baseline analysis, we continue to cluster standard errors by borrower zip code. In Appendix Table A.17, we present
the main models with alternative standard errors. We cluster standard errors by bank in columns 1-2, and double cluster standard
errors by bank and week approved in columns 4-6. The statistical significance of our results is robust to both approaches.
27
In Figure 4, we plot the βt coefficients from the dynamic differences-in-differences model in
Equation 3. Panel A shows estimates with only bank and time fixed effects (as in Table 11, Panel A,
column 1). Panel B adds the additional vector of controls (as in Table 11, Panel A, column 2). We do
not report more than three months before automation because the automation dates (usually in late
Spring and late Fall after a period in which the PPP program was in active) mean that we rarely observe
PPP loans four months prior to an automation event. The graphs indicate no pre-trend before
automation and a clear upwards trend afterwards. In order to eliminate potential compositional effects,
we also conduct a robustness check at the weekly level that restricts the sample to a balanced panel of
11 automating banks that have observations at least six weeks on both sides of the automation date. The
results, in Appendix Figure A.8, also indicate no pre-trends and a clear discontinuity in the weeks
following automation, even though for some of the banks our data do not include the exact week of
automation. Overall, the figures demonstrate an immediate and persistent discontinuity around the
automation event, consistent with a causal effect of automation on the lending behavior of small banks.
In the last part of this section, we examine whether the effect of automation on the share of loans
to Black-owned businesses is larger in areas with higher racial animus. In Panel B of Table 11, we
interact our automation dummy with our six indicators for racial animus. We see positive and significant
interactions for four of the six measures, with the remaining two positive but insignificant. For example,
the interaction coefficient of 0.009 for the implicit bias test implies that, relative to the mean share of
loans to Black-owned businesses in this small bank sample, automation increased this share by 24% more
in areas with one standard deviation higher racial animus. The fact that automation has larger effects on
the share of loans to Black-owned businesses in areas with higher racial animus provides additional
evidence that one of the mechanisms through which automation increases those loans is by reducing the
effect of preference-based discrimination.
5 Conclusion
The original legislation authorizing the PPP included an explicit mandate to prioritize socioeconomically
disadvantaged businesses. Yet, in practice, many conventional bank lenders did not serve Black-owned
businesses in proportion to their share in the PPP borrower population. Instead, it was fintech firms that
made a disproportionate share of loans to Black-owned businesses, accounting for over half of the PPP
loans to Black-owned businesses. Among conventional lenders, small banks had a particularly low rate
of lending to Black-owned businesses. Why would this have occurred given that PPP loans were 100%
guaranteed by the federal government? This question is the focus of our paper.
We show that a rich array of basic borrower characteristics—including location, loan amount, loan
approval date, industry, and business form—can explain about two-thirds of the disparity between
fintechs and other lenders. However, even after these controls, Black-owned businesses remain about 12
percentage points more likely than other firms to get their PPP loan from a fintech lender. Much of this
substitution towards fintech lenders comes from a lower propensity of Black-owned businesses to get
28
loans from smaller banks. We show that differential pre-existing bank relationships, real-time revenue,
fintech affinity, and firm application behavior cannot fully explain this gap.
We find suggestive evidence consistent with a role for preference-based discrimination à la Becker
(1957) in explaining some of the lower rates of lending to Black-owned businesses among smaller
conventional lenders. Specifically, we show that the observed differences are larger in areas with higher
racial animus. It is important to note that many of the variables we condition on in this analysis—such
as bank branch location or the distribution of firms with checking accounts at banks—may be the result
of historical patterns of discrimination. It is therefore possible that the substantial differences in our
controlled models represent a lower bound for the overall effect of discrimination on lending patterns.
We also present evidence that automation may help explain why disparate treatment is concentrated
among smaller banks. We show that after small banks automated their lending processes, their rate of
lending to Black-owned businesses increased substantially. This contributes to the ongoing conversation
about the equity effects using machine learning algorithms in lending (Fuster et al., 2020; Berg et al.,
2020; Blattner et al., 2021; Bartlett et al., 2021). While concerns that algorithms may discriminate
because they are trained on biased data are warranted, our results suggest that there may be benefits from
automation. Specifically, by eliminating manual review conducted by biased humans, automation could
reduce the incidence of taste-based discrimination.
29
References
Abrams, Amanda, “To Help Black Developers, Programs Start With Access to Capital,” The New York
Times, 2021, March 21.
Alekseev, G., S. Amer, M. Gopal, T. Kuchler, J. W. Schneider, J. Stroebel, and N. Wernerfelt, “The
Effects of COVID-19 on US Small Businesses: Evidence from Owners, Managers, and Employees,”
Technical Report, NBER Working Paper No. w27833 2020.
Ambrose, Brent W, James Conklin, and Luis A Lopez, “Does Borrower and Broker Race Affect the
Cost of Mortgage Credit?,” Review of Financial Studies (Forthcoming), 2020.
Anwar, Shamena and Hanming Fang, “An Alternative Test of Racial Prejudice in Motor Vehicle
Searches: Theory and Evidence,” American Economic Review, 2006, 96 (1), 127–151.
Arnold, David, Will Dobbie, and Crystal S Yang, “Racial bias in bail decisions,” The Quarterly
Journal of Economics, 2018, 133 (4), 1885–1932.
Asiedu, Elizabeth, James A Freeman, and Akwasi Nti-Addae, “Access to credit by small businesses:
How relevant are race, ethnicity, and gender?,” American Economic Review, 2012, 102 (3), 532–37.
Autor, David, David Cho, Leland D. Crane, and Mita Goldar, “An Evaluation of the Paycheck
Protection Program Using Administrative Payroll Microdata,” Technical Report, National Bureau of
Economic Research 2020.
Balyuk, Tetyana, Allen N Berger, and John Hackney, “What is fueling fintech lending? the role of
banking market structure,” The Role of Banking Market Structure (June 23, 2020), 2020.
Barraza, Santiago, Martı́n Rossi, and Timothy J Yeager, “The Short-Term Effect of the Paycheck
Protection Program on Unemployment,” Available at SSRN 3667431, 2020.
Bartik, Alexander W, Marianne Bertrand, Zoe Cullen, Edward L Glaeser, Michael Luca, and
Christopher Stanton, “The impact of COVID-19 on small business outcomes and expectations,”
Proceedings of the National Academy of Sciences, 2020, 117 (30), 17656–17666.
, Zoe B Cullen, Edward L Glaeser, Michael Luca, Christopher T Stanton, and Adi Sunderam,
“The targeting and impact of Paycheck Protection Program loans to small businesses,” Technical
Report, National Bureau of Economic Research 2020.
Bartlett, Robert, Adair Morse, Richard Stanton, and Nancy Wallace, “Consumer-lending
discrimination in the FinTech era,” Journal of Financial Economics, 2021.
Bartlett, Robert P and Adair Morse, “Small Business Survival Capabilities and Policy Effectiveness:
Evidence from Oakland,” Available at SSRN 3660991, 2020.
Bartoš, Vojtěch, Michal Bauer, Julie Chytilová, and Filip Matějka, “Attention discrimination: Theory
and field experiments with monitoring information acquisition,” American Economic Review, 2016,
106 (6), 1437–75.
Bayer, Patrick, Fernando Ferreira, and Stephen L Ross, “What Drives Racial and Ethnic Differences
in High-Cost Mortgages? The Role of High-Risk Lenders,” The Review of Financial Studies, 2018, 31
(1), 175–205.
Becker, Gary S., “economics of discrimination,” 1957.
Beer, Tommy, “Minority-Owned Small Businesses Struggle To Gain Equal Access To PPP Loan
Money,” 2020. [Forbes Online; accessed 09-Dec-2020].
Begley, Taylor A and Amiyatosh Purnanandam, “Color and credit: Race, regulation, and the quality
of financial services,” Journal of Financial Economics, 2021, 141 (1), 48–65.
Bellucci, Andrea, Alexander Borisov, and Alberto Zazzaro, “Do banks price discriminate spatially?
Evidence from small business lending in local credit markets,” Journal of Banking & Finance, 2013,
37 (11), 4183–4197.
Berg, Tobias, Valentin Burg, Ana Gombović, and Manju Puri, “On the rise of fintechs: Credit scoring
30
using digital footprints,” The Review of Financial Studies, 2020, 33 (7), 2845–2897.
Berger, Allen N and Lamont K Black, “Bank size, lending technologies, and small business finance,”
Journal of Banking & Finance, 2011, 35 (3), 724–735.
Bertrand, Marianne and Sendhil Mullainathan, “Are Emily and Greg more employable than Lakisha
and Jamal? A field experiment on labor market discrimination,” American Economic Review, 2004,
94 (4), 991–1013.
Bhutta, Neil and Aurel Hizmo, “Do minorities pay more for mortgages?,” The Review of Financial
Studies, 2021, 34.
Blanchflower, David G, Phillip B Levine, and David J Zimmerman, “Discrimination in the small-
business credit market,” Review of Economics and Statistics, 2003, 85 (4), 930–943.
Blattner, Laura and Scott Nelson, “How Costly is Noise? Data and Disparities in Consumer Credit,”
arXiv preprint arXiv:2105.07554, 2021.
, , and Jann Spiess, “Unpacking the Black Box: Regulating Algorithmic Decisions,” 2021.
Buchak, Greg, Gregor Matvos, Tomasz Piskorski, and Amit Seru, “Fintech, regulatory arbitrage, and
the rise of shadow banks,” Journal of Financial Economics, 2018, 130 (3), 453–483.
Bursztyn, Leonardo, Thomas Chaney, Tarek Alexander Hassan, and Aakaash Rao, “The Immigrant
Next Door: Exposure, Prejudice, and Altruism,” Technical Report, National Bureau of Economic
Research 2021.
Chernenko, Sergey and David S. Scharfstein, “Racial Disparities in the Paycheck Protection
Program,” Technical Report, Available at SSRN: https://ssrn.com/abstract=3907575 or
http://dx.doi.org/10.2139/ssrn.3907575 2021.
Congress, Too Big to Fail?: The Role of Antitrust Law in Government-funded Consolidation in the
Banking Industry: Hearing Before the Subcommittee on Courts and Competition Policy of the
Committee on the Judiciary, House of Representatives, One Hundred Eleventh Congress, First Session,
March 17, 2009, US Government Printing Office, 2009.
Cowley, Stacy, “Minority Entrepreneurs Struggled to Get Small-Business Relief Loans,” The New York
Times, 2021, April 4.
Crosman, Penny, “Community Banks Get on Board With Bots,” American Banker, 2020.
Cross, Miriam, “Small Banks Count on PPP Tech Advances to Speed Traditional Lending,” American
Banker, 2021.
Crowell, Charlene, “Black America Needs a ‘New Normal’: Equitable Credit Access to Build Wealth,”
The Philadelphia Tribune, 2021, March 30.
Crutsinger, Martin, “Administration pledges more support for minority businesses,” AP News, 2021,
February 5.
D’Acunto, Francesco, Pulak Ghosh, Rajiv Jain, and Alberto G Rossi, “How Costly are Cultural
Biases?,” Available at SSRN 3736117, 2020.
Dobbie, Will, Andres Liberman, Daniel Paravisini, and Vikram S Pathania, “Measuring bias in
consumer lending,” Review of Economic Studies, 2020.
Erel, Isil and Jack Liebersohn, “Does FinTech Substitute for Banks? Evidence from the Paycheck
Protection Program,” Technical Report, National Bureau of Economic Research 2020.
Fairlie, R., “The impact of COVID-19 on small business owners: Evidence from the first three months
after widespread social-distancing restrictions,” Journal of Economics & Management Strategy, 2020,
29, 727–740.
Fairlie, Robert W, Alicia Robb, and David T Robinson, “Black and White: Access to Capital Among
Minority-Owned Startups,” Technical Report, National Bureau of Economic Research 2020.
and Alicia M Robb, “Why are black-owned businesses less successful than white-owned businesses?
31
The role of families, inheritances, and business human capital,” Journal of Labor Economics, 2007,
25 (2), 289–323.
and Frank Fossen, “Did the 660BillionP aycheckP rotectionP rogramand220 Billion Economic
Injury Disaster Loan Program Get Disbursed to Minority Communities in the Early Stages of COVID-
19?,” Technical Report, National Bureau of Economic Research 2021.
Faulkender, Michael W, Robert Jackman, and Stephen Miran, “The Job Preservation Effects of
Paycheck Protection Program Loans,” Available at SSRN 3767509, 2020.
Federal Reserve, “Federal Reserve expands access to its Paycheck Protection Program Liquidity Facility
(PPPLF) to additional lenders, and expands the collateral that can be pledged,” Press Release, 2020,
April 30.
Fuster, Andreas, Matthew Plosser, Philipp Schnabl, and James Vickery, “The role of technology in
mortgage lending,” The Review of Financial Studies, 2019, 32 (5), 1854–1899.
, Paul Goldsmith-Pinkham, Tarun Ramadorai, and Ansgar Walther, “Predictably unequal? the
effects of machine learning on credit markets,” The Effects of Machine Learning on Credit Markets
(October 1, 2020), 2020.
Giacoletti, Marco, Rawley Heimer, and Edison G Yu, “Using High-Frequency Evaluations to Estimate
Discrimination: Evidence from Mortgage Loan Officers,” Available at SSRN 3795547, 2021.
Gopal, Manasa and Philipp Schnabl, “The rise of finance companies and FinTech lenders in small
business lending,” NYU Stern School of Business, 2020.
Granja, J., C. Makridis, C. Yannelis, and E. Zwick, “Did the Paycheck Protection Program Hit the
Target?,” Technical Report, NBER Working Paper No. w27095. 2020.
Greenwood, Robin M, Benjamin Charles Iverson, and David Thesmar, “Sizing up corporate
restructuring in the covid crisis,” NBER Working Paper, 2020, (w28104).
Griffin, John M, Samuel Kruger, and Prateek Mahajan, “Did FinTech Lenders Facilitate PPP
Fraud?,” Available at SSRN 3906395, 2021.
Hubbard, R Glenn and Michael R Strain, “Has the Paycheck Protection Program Succeeded?,”
Technical Report, National Bureau of Economic Research 2020.
Humphries, J. E., N. Mader, D. Tannenbaum, and W van Dijk, “Does eviction cause poverty? quasi-
experimental evidence from Cook County, IL.,” Technical Report, NBER Working Paper No. w26139
2019.
Humphries, John Eric, Christopher A Neilson, and Gabriel Ulyssea, “Information frictions and
access to the Paycheck Protection Program,” Journal of public economics, 2020, 190, 104244.
Imai, K. and K. Khanna, “Improving ecological inference by predicting individual ethnicity from voter
registration records,” Political Analysis, 2016, pp. 263–272.
Kim, Olivia S, Jonathan A Parker, and Antoinette Schoar, “Revenue Collapses and the Consumption
of Small Business Owners in the Early Stages of the COVID-19 Pandemic,” Technical Report,
National Bureau of Economic Research 2020.
Knowles, John, Nicola Persico, and Petra Todd, “Racial Bias in Motor Vehicle Searches: Theory and
Evidence,” Journal of Political Economy, 2001, 109 (1), 203–229.
Li, Lei and Philip Strahan, “Who supplies ppp loans (and does it matter)? banks, relationships and the
covid crisis,” Technical Report, National Bureau of Economic Research 2020.
Lopez, German, “For years, this popular test measured anyone’s racial bias. But it might not work after
all.,” Vox, 2017.
Massey, Douglas S and Nancy A Denton, “The dimensions of residential segregation,” Social forces,
1988, 67 (2), 281–315.
Milkman, Katherine L, Modupe Akinola, and Dolly Chugh, “Temporal distance and discrimination:
32
An audit study in academia,” Psychological science, 2012, 23 (7), 710–717.
Petersen, Mitchell A and Raghuram G Rajan, “The benefits of lending relationships: Evidence from
small business data,” The journal of finance, 1994, 49 (1), 3–37.
Philippon, Thomas, “On fintech and financial inclusion,” Technical Report, National Bureau of
Economic Research 2019.
Price, Joseph and Justin Wolfers, “Racial discrimination among NBA referees,” The Quarterly journal
of economics, 2010, 125 (4), 1859–1887.
Ranson, Diana, “The Fed Needs to Step Up to Help Main Street,” Inc.com, 2020, August 28.
Robb, Alicia and David T Robinson, “Testing for Racial Bias in Business Credit Scores,” Small
Business Economics, 2018, 50 (3), 429–443.
Rosenberg, Joyce M. and Justin Myers, “Minority-owned companies waited months for loans, data
shows,” The Associated Press, 2020, December 31.
Seru, Amit, “Regulating banks in the era of fintech shadow banks,” The Andrew Crocket Memorial
Lecture, Bank for International Settlements, Basel, 2019, 30.
Simon, Ruth and Peter Rudegeair, “In Race for Small-Business Loans, Winning Hinged on Where
Firms Bank,” The Wall Street Journal, 2020, April 20.
Smith, David, “The Ballad of the Small Banker: An SBA Lender’s Experience with PPP Loans,” 2020.
[FICO Blog Online; May 07 2020].
Stephens-Davidowitz, Seth I, “The cost of racial animus on a black presidential candidate: Using google
search data to find what surveys miss,” Available at SSRN 2238851, 2013.
Tang, Huan, “Peer-to-peer lenders versus banks: substitutes or complements?,” The Review of Financial
Studies, 2019, 32 (5), 1900–1938.
Tausanovitch, Chris and Lynn Vavreck, “Democracy Fund + UCLA Nationscape, Oct 10–17, 2019
(version 20200814),” Technical Report 2020.
Tootell, Geoffrey MB, “Redlining in Boston: Do mortgage lenders discriminate against
neighborhoods?,” The Quarterly Journal of Economics, 1996, 111 (4), 1049–1079.
Tzioumis, Konstantinos, “Demographic aspects of first names,” Scientific data, 2018, 5, 180025.
United States Census Bureau, “Survey of Business Owners (SBO),” 2012.
Word, David L., Charles D. Coleman, Robert Nunziata, and Robert Kominski, “Demographic
aspects of surnames from Census 2000,” 2008.
Xu, Kaiyuan, Brian Nosek, and Anthony Greenwald, “Psychology data from the race implicit
association test on the project implicit demo website,” Journal of Open Psychology Data, 2014, 2
(1).
Zhou, Li, “The Paycheck Protection Program failed many Black-owned businesses,” 2020. [Vox Online;
accessed 09-Dec-2020].
33
Figure 1: Black-Owned Business PPP Lending by Institution Type
(A) Share of PPP Loans to Black-Owned Businesses by Institution Type
Small Banks
Minority Dep Inst
Citibank
Medium Banks
JPMC Bank
Large Banks
Bank of America
Wells Fargo Bank
Credit Union
CDFI/Nonprofit
Fintech
0 5 10 15 20 25
Share of All Loans Made by Lender Type to Black-Owned Businesses
(B) Share of PPP Lender Institution Type among Black-Owned Businesses
Small Banks
Minority Dep Inst
Citibank
Medium Banks
JPMC Bank
Large Banks
Bank of America
Wells Fargo Bank
Credit Union
CDFI/Nonprofit
Fintech
0 20 40 60
Share Loans to Black-Owned Businesses Made by Lender Type
Note: This figure shows the shares of PPP loans made by originating lender type to Black-owned businesses. Panel A shows
the Black share of PPP loans made by originating lender type (P (Black-owned|Originating Lender Type)). Panel B shows the
shares of PPP loans from originating lender type made to Black-owned businesses (P (Originating Lender Type|Black-owned)).
34
Share of Fintech Loans in Week Share of Loans to Black Borrowers in Week
0 20 40 60 80 0 10 20 30
4/01/20 4/01/20
4/08/20 4/08/20
4/15/20 4/15/20
4/22/20 4/22/20
4/29/20 4/29/20
5/06/20 5/06/20
5/13/20 5/13/20
5/20/20 5/20/20
5/27/20 5/27/20
Note: This figure shows shares of PPP loans made to Black-owned businesses (Panel A) and made by fintech lenders (Panel B)
6/03/20 6/03/20
6/10/20 6/10/20
6/17/20 6/17/20
35 6/24/20 6/24/20
7/01/20 7/01/20
Week of
Week of
7/08/20 7/08/20
7/15/20 7/15/20
7/22/20 7/22/20
7/29/20 7/29/20
(B) Fintech PPP Lending by Week
8/05/20 8/05/20
Break in PPP Break in PPP
1/08/21 1/08/21
(A) Black-Owned Businesses PPP Lending by Week
1/15/21 1/15/21
1/22/21 1/22/21
1/29/21 1/29/21
Figure 2: Black-Owned Businesses PPP Lending by Week
2/05/21 2/05/21
2/12/21 2/12/21
2/19/21 2/19/21
by week of loan approval. The dashed red line denotes a hiatus in the PPP program from August 2020 to January 2021.
Figure 3: Conditional Share of PPP Loans to Each Race by Institution Type
(A) Black-Owned Businesses (B) White-Owned Businesses
Minority Dep Inst Minority Dep Inst
Small Banks Fintech
Medium Banks Bank of America
Large Banks Citibank
CDFI/Nonprofit JPMC Bank
JPMC Bank Wells Fargo Bank
Bank of America CDFI/Nonprofit
Credit Union Credit Union
Citibank Large Banks
Wells Fargo Bank Medium Banks
Fintech Small Banks
.03 .04 .05 .06 .07 .08 .09 .1 .11 .76 .78 .8 .82 .84 .86 .88 .9
Share Black-Owned Share White-Owned
(C) Hispanic-Owned Businesses (D) Asian-Owned Businesses
Minority Dep Inst Credit Union
Small Banks Wells Fargo Bank
Medium Banks Medium Banks
Large Banks Small Banks
Credit Union Large Banks
JPMC Bank Fintech
Citibank CDFI/Nonprofit
Bank of America JPMC Bank
Wells Fargo Bank Citibank
CDFI/Nonprofit Bank of America
Fintech Minority Dep Inst
.015 .02 .025 .03 .035 .04 .045 .05 .055 0 .05 .1 .15 .2
Share Hispanic/Latino-Owned Share Asian-Owned
Note: This figure shows shares of PPP loans made to businesses predicted to be Black-owned by originating lender type. Each
graph presents coefficients from variants of the following regression: Black-ownedi = βLender Typei + γXi + i , where Xi
is a vector of fixed effects for borrower zip code, loan amount percentile (in 100 bins), approval week, 3-digit NAICS industry,
business type, and employer status. Standard errors are clustered by zip code. In each panel, we change the dependent variable
to be an indicator for whether a borrower is a Black- (Panel A), White- (Panel B), Hispanic- (Panel C), or Asian- (Panel D)
owned business.
36
Figure 4: Share of Loans to Black-Owned Businesses Before and After Small Bank Automation
(A) Simple Differences-in-Differences (B) With Additional Controls
Note: This figure reports dynamic differences-in-differences estimates at the monthly level, using Equation 3. Period 0
following the dashed vertical line correspond to the automation month. Panel A includes fixed effects for the bank and week
of loan approval. Panel B adds the vector of controls included in Table 11, Panel A, column 1. We do not include more than
three months before automation because the automation dates (in late Spring and late Fall after a large gap in the PPP program)
mean that we observe essentially no loans made four month prior to . Standard errors are clustered by zip code. The grey bars
represent 95% confidence intervals.
37
Table 1: Summary Statistics by Lender Type
Panel A: All PPP Loans
PPP Loan Amount ($)
Number Number Share Total
Lenders Loans Loans Amt ($ bn) Mean P10 P50 P90
All 4,889 11,768,450 100% 798.7 67,867 4,199 20,678 124,500
Bank of America 1 491,032 4.2% 34.4 70,080 4,820 20,833 127,855
Citibank 1 47,767 0.4% 4.8 100,359 8,200 29,300 188,900
JPMC Bank 1 438,571 3.7% 41.6 94,745 6,542 27,500 181,897
Wells Fargo Bank 1 280,693 2.4% 13.8 49,179 4,181 18,777 97,042
Large Banks 17 898,175 7.6% 108.8 121,127 6,000 28,133 230,895
Medium Banks 377 2,139,089 18.2% 271.3 126,807 6,080 30,405 266,000
Small Banks 3,196 2,385,728 20.3% 179.0 75,024 4,130 20,832 151,100
Credit Union 946 378,673 3.2% 15.7 41,479 3,602 15,882 82,422
CDFI/Nonprofit 187 1,516,834 12.9% 34.4 22,711 4,400 20,000 20,833
Minority Dep Inst 133 424,976 3.6% 29.3 68,924 3,823 20,053 131,015
Fintech 29 2,766,912 23.5% 65.6 23,726 3,083 17,208 29,166
Panel B: All First Draw PPP Loans Before Feb 24th
PPP Loan Amount ($)
Number Number Share Total
Lenders Loans Loans Amt ($ bn) Mean P10 P50 P90
All 4,864 5,692,097 100% 533.8 93,784 4,500 20,833 176,100
Bank of America 1 346,477 6.1% 25.6 73,863 4,375 20,833 130,610
Citibank 1 31,341 0.6% 3.4 108,697 8,200 29,800 199,300
JPMC Bank 1 290,131 5.1% 29.7 102,374 6,570 27,597 190,560
Wells Fargo Bank 1 202,027 3.5% 10.5 52,113 4,078 18,520 100,000
Large Banks 17 590,817 10.4% 81.8 138,458 6,200 30,625 258,331
Medium Banks 377 1,405,565 24.7% 197.7 140,632 6,200 33,295 288,874
Small Banks 3,191 1,360,381 23.9% 118.6 87,182 4,524 20,833 173,375
Credit Union 936 227,205 4.0% 10.3 45,402 3,541 15,932 88,900
CDFI/Nonprofit 178 112,320 2.0% 8.2 72,687 3,989 20,800 144,422
Minority Dep Inst 133 201,952 3.5% 18.6 91,885 4,200 20,833 173,500
Fintech 28 923,881 16.2% 29.5 31,921 2,937 15,500 55,782
Panel C: Bank and Credit Relationships Sample (Ocrolus)
SME Has
Checking
Credit With Any: Monthly Net Cash Inflow ($)
N Acct with
PPP Lender Fintech Non-Fintech Mean P10 P50 P90
All 216,240 28.5% 14.2% 79.8% 9,124 -36,671 1,374 62,879
Bank of America 15,171 63.9% 15.6% 86.6% 8,755 -45,847 1,433 70,230
Citibank 1,360 53.5% 12.1% 66.2% 10,282 -45,101 1,144 75,981
JPMC Bank 12,616 70.2% 17.5% 89.9% 14,478 -66,644 4,537 111,547
Wells Fargo Bank 8,584 73.7% 13.2% 88.0% 14,451 -34,225 4,069 74,553
Large Banks 21,381 50.0% 14.6% 82.5% 12,188 -42,302 2,584 77,336
Medium Banks 37,338 39.8% 14.5% 75.5% 9,558 -50,997 1,581 76,855
Small Banks 27,580 23.8% 14.2% 75.1% 7,757 -50,065 1,330 71,616
Credit Union 7,255 35.7% 13.0% 71.6% 6,292 -27,811 749 47,421
CDFI/Nonprofit 3,865 8.2% 13.2% 78.0% 7,674 -36,688 1,205 56,613
Minority Dep Inst 4,954 21.6% 13.8% 75.4% 5,410 -52,939 1,117 64,189
Fintech 76,136 0.0% 13.3% 80.3% 7,697 -20,136 946 41,984
Note: This table reports summary statistics about PPP loans by originating lender type, where each PPP lender is assigned to
a single type. The data in Panel A include all PPP loans, including “second draw” loans, which is a firm’s second PPP loan
(accounting for about 2.8 million loans). Panel B repeats the statistics in our starting sample, which is composed of first draw
PPP loans between April 3, 2020 and February 23, 2021. All subsequent statistics and analysis are drawn from subsamples of
the data included in Panel B. Panel C reports statistics about banking and credit relationships as well as financial performance
from those borrowers in the sample of bank statements. The first column, ”SME has Checking Account with PPP Lender,”
means that the borrower’s business checking account bank is the same institution that originated their PPP loan. The remaining
variables are derived from transactions on the borrowers’ most recent monthly bank statement. In this table, we include all
loans, regardless of whether race is populated. Appendix Table A.2 repeats Panels B and C for the subset with predicted race.
38
Table 2: Sample Characteristics
Panel A: Analysis Sample
N = 4,183,623
Share Mean Loan Share Share Share Share
Loans Amt ($) Asian Borr Black Borr Hisp Borr White Borr
All 100% 93,666 8.9% 8.6% 7.5% 75.0%
Bank of America 6.4% 75,094 17.9% 6.9% 12.3% 62.9%
Citibank 0.5% 108,719 17.6% 3.9% 10.9% 67.6%
JPMC Bank 5.3% 101,318 14.7% 4.5% 9.6% 71.2%
Wells Fargo Bank 3.7% 52,984 11.9% 7.8% 12.3% 68.0%
Large Banks 9.4% 146,505 7.7% 5.3% 5.5% 81.4%
Medium Banks 24.3% 142,532 6.7% 4.1% 5.3% 84.0%
Small Banks 23.9% 87,164 4.3% 3.3% 3.2% 89.3%
Credit Union 4.1% 45,412 5.2% 9.7% 6.9% 78.1%
CDFI/Nonprofit 2.0% 71,564 7.9% 10.6% 7.1% 74.4%
Minority Dep Inst 3.0% 98,406 25.7% 3.9% 13.6% 56.9%
Fintech 17.4% 31,228 11.2% 26.5% 13.2% 49.2%
Panel B: Bank and Credit Relationships Sample (Ocrolus)
N = 168,360
Share Mean Loan Share Share Share Share
Loans Amt ($) Asian Borr Black Borr Hisp Borr White Borr
All 100% 80,897 10.2% 15.3% 11.0% 63.5%
Bank of America 7.1% 67,838 14.9% 9.7% 15.6% 59.7%
Citibank 0.6% 109,947 18.0% 4.3% 14.0% 63.6%
JPMC Bank 5.9% 87,618 13.0% 6.4% 12.7% 67.9%
Wells Fargo Bank 4.0% 53,761 11.4% 9.7% 14.9% 64.0%
Large Banks 9.0% 104,023 9.2% 8.1% 8.1% 74.6%
Medium Banks 16.9% 133,265 7.6% 5.9% 8.3% 78.2%
Small Banks 12.6% 118,637 6.9% 6.1% 6.9% 80.1%
Credit Union 3.4% 56,194 6.7% 14.2% 9.6% 69.5%
CDFI/Nonprofit 1.8% 66,732 8.9% 17.8% 11.9% 61.4%
Minority Dep Inst 2.4% 114,058 34.2% 5.4% 13.3% 47.0%
Fintech 36.3% 42,379 10.0% 28.7% 12.7% 48.6%
Panel C: Application Sample (Lendio)
N = 175,660
Share Mean Loan Share Share Share Share
Loans Amt ($) Asian Borr Black Borr Hisp Borr White Borr
All 100% 65,184 10.8% 11.3% 9.5% 68.4%
Conventional 47.7% 88,688 10.1% 6.9% 8.9% 74.0%
Fintech 52.3% 43,765 11.5% 15.3% 10.0% 63.3%
39
Table 2: Sample Characteristics (Continued)
Panel D: Card Revenue Sample (Enigma)
N = 813,812
Share Mean Loan Share Share Share Share
Loans Amt ($) Asian Borr Black Borr Hisp Borr White Borr
All 100% 141,529 15.7% 2.8% 6.6% 74.9%
Bank of America 7.5% 93,136 29.3% 4.1% 11.5% 55.1%
Citibank 0.5% 137,962 24.7% 2.5% 11.0% 61.8%
JPMC Bank 5.5% 140,825 21.6% 2.5% 9.4% 66.4%
Wells Fargo Bank 3.8% 68,501 23.2% 4.5% 12.0% 60.3%
Large Banks 11.8% 193,926 11.8% 2.9% 5.3% 79.9%
Medium Banks 29.1% 185,730 10.6% 2.2% 5.2% 82.0%
Small Banks 23.0% 135,672 9.3% 1.7% 3.9% 85.1%
Credit Union 3.4% 80,015 10.8% 3.8% 6.0% 79.4%
CDFI/Nonprofit 2.0% 105,914 13.7% 3.9% 6.6% 75.8%
Minority Dep Inst 3.6% 136,611 42.2% 2.2% 8.9% 46.7%
Fintech 9.7% 57,138 26.0% 5.5% 10.3% 58.2%
Note: This table shows loan characteristics and borrower race and ethnic breakdown by lender type. Panel A includes the
analysis sample (all loans for which we can predict race). Panel B restricts to the bank statement-matched sample, which
includes borrowers for whom we observe a bank statement prior to the PPP loan approval date. Panel C restricts to the Lendio-
matched sample. We limit the Lendio sample to borrowers whose loan approval date is after their Lendio application date and
are sent to at least one lender by Lendio. Panel D restricts to the Enigma-matched sample. We limit the Enigma sample to
borrowers for whom we observe card revenue prior to loan approval.
40
Table 3: Summary Statistics by Predicted Race
Panel A: Analysis Sample
All Asian-Owned Black-Owned Hispanic-Owned White-Owned
Loan Amount
Mean Loan Amount ($) 93,666 53,492 24,315 54,287 110,317
Median Loan Amount ($) 20,833 20,071 14,886 18,218 $23,700
Share Total Loans Made by Bank Types
Bank of America 6.4% 12.8% 5.1% 10.5% 5.3%
Citibank 0.5% 1.1% 0.3% 0.8% 0.5%
JPMC Bank 5.3% 8.7% 2.8% 6.8% 5.0%
Wells Fargo Bank 3.7% 4.9% 3.3% 6.0% 3.3%
Large Banks 9.4% 8.1% 5.8% 6.9% 10.2%
Medium Banks 24.3% 18.3% 11.5% 17.1% 27.3%
Small Banks 23.9% 11.4% 9.1% 10.2% 28.4%
Credit Union 4.1% 2.4% 4.7% 3.8% 4.3%
CDFI/Nonprofit 2.0% 1.7% 2.4% 1.9% 2.0%
Minority Dep Inst 3.0% 8.8% 1.4% 5.5% 2.3%
Fintech 17.4% 21.7% 53.6% 30.5% 11.4%
Business Types
Corporation 27.8% 40.3% 11.2% 28.7% 28.1%
LLC 26.8% 24.4% 19.9% 23.9% 28.1%
Nonprofit 2.8% 1.0% 2.1% 1.5% 3.2%
Other 0.5% 0.4% 0.5% 0.4% 0.5%
Self Employed 8.6% 6.9% 23.6% 13.9% 6.6%
Sole Proprietorship 20.4% 14.9% 38.4% 21.5% 18.8%
Subchapter S Corporation 13.1% 12.0% 4.3% 10.1% 14.6%
Employer Institution 63.4% 73.8% 21.0% 57.8% 67.6%
Industries (3 Digit NAICS)
Professional/Technical Services 12.7% 7.8% 10.6% 10.8% 13.7%
Ambulatory Health Care Services 7.5% 10.3% 8.2% 6.3% 7.2%
Food and Drinking Services 5.9% 14.7% 3.5% 8.7% 4.8%
Personal and Laundry Services 6.0% 12.0% 15.6% 7.5% 4.0%
Specialty Trade Contractors 5.4% 0.9% 2.2% 6.3% 6.2%
Other 62.6% 54.3% 60.0% 60.4% 64.1%
Observations 4,183,623 372,993 359,366 313,389 3,137,875
41
Table 3: Summary Statistics by Predicted Race Continued
Panel B: Bank and Credit Relationships Sample (Ocrolus)
All Asian-Owned Black-Owned Hispanic-Owned White-Owned
Banking Relationships
Share Checking Acct from Top 4 Banks 47.8% 57.5% 47.1% 61.9% 44.1%
Share Checking Acct from Large Banks 15.8% 13.6% 19.1% 11.6% 16.0%
Share Checking Acct from Small/Medium Banks 27.5% 20.3% 16.5% 19.9% 32.7%
Share Checking Acct from Others 8.9% 8.7% 17.2% 6.7% 7.2%
Credit Relationships
Share Credit Relationship with Fintech 14.2% 13.2% 7.9% 12.5% 16.2%
Share Credit Relationship with Non-Fintech 80.0% 78.9% 81.4% 85.8% 78.9%
Share Credit with Checking Account Bank 17.6% 18.4% 14.4% 19.5% 17.9%
Cash In/Outflows
Mean Monthly Cash Inflow ($) 231,390 223,615 81,706 201,382 273,948
Median Monthly Cash Inflow ($) 65,232 70,141 11,636 58,644 87,949
Mean Monthly Cash Outflow ($) 218,738 213,055 72,270 189,775 260,004
Median Monthly Cash Outflow ($) 60,890 67,033 8,960 52,466 82,450
Mean Monthly Net Cash Inflow ($) 9,016 7,426 6,332 8,681 9,977
Median Monthly Net Cash Inflow ($) 1,332 1,232 703 1,698 1,676
Observations 168,360 17,166 25,782 18,531 106,881
Panel C: Application Sample (Lendio)
All Asian-Owned Black-Owned Hispanic-Owned White-Owned
Number of Lenders Sent
All 1.480 1.564 1.464 1.498 1.468
Fintech 0.886 1.024 0.995 0.870 0.852
Conventional 0.594 0.540 0.469 0.628 0.616
Share of Lenders Sent
Fintech & Conventional 14.7% 16.8% 15.8% 15.4% 14.2%
Fintech Only 55.5% 62.0% 64.4% 53.7% 53.4%
Conventional Only 29.8% 21.2% 19.8% 30.9% 32.4%
Observations 175,660 16,711 17,442 14,644 105,584
Panel D: Card Revenue Sample (Enigma)
All Asian-Owned Black-Owned Hispanic-Owned White-Owned
Card Revenue
Mean (Approval & 2 Prev Mos) 53,968 43,422 23,169 42,557 58,355
Median (Approval & 2 Prev Mos) 24,993 21,320 10,273 21,118 27,178
Observations 813,812 127,957 23,082 53,622 609,151
Struggling (Rev in 2/20, not at Appr) 20.8% 24.0% 24.4% 20.5% 20.0%
Observations 970,270 142,680 30,960 64,077 732,553
Note: This table reports summary statistics by race/ethnicity for our four main samples. Panel A contains loan and firm
characteristics for the full analysis sample. Panel B summarizes information from bank statement-matched data. Panel C
contains statistics about the data on PPP applications from the Lendio platform, restricted to applications linked to a PPP loan
in the SBA data. “Conventional” includes all non-fintech lenders. Panel D summarizes the credit and debit card data. We define
firms as “struggling” if they had revenue in February, 2020 but not in the month of approval.
42
Table 4: Business Owner Race and PPP Lender Type
Panel A: Fintech PPP Loan
Dependent Variable: 1(Fintech)
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black-Owned) 0.397∗∗∗ 0.316∗∗∗ 0.232∗∗∗ 0.269∗∗∗ 0.175∗∗∗ 0.155∗∗∗ 0.148∗∗∗ 0.121∗∗∗
(0.005) (0.004) (0.002) (0.003) (0.002) (0.002) (0.002) (0.002)
Loan Amount FE No Yes Yes Yes Yes Yes Yes Yes
Zip Code FE No No Yes No Yes Yes No Yes
Census Tract FE No No No Yes No No No No
Approval Week FE No No No No Yes Yes Yes Yes
Industry FE No No No No No Yes No Yes
Zip-by-Industry FE No No No No No No Yes No
Business Type FE No No No No No No No Yes
Employer Status FE No No No No No No No Yes
Dep Var Mean 0.174 0.174 0.174 0.174 0.174 0.174 0.174 0.174
R2 0.086 0.138 0.218 0.188 0.291 0.313 0.390 0.356
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
Panel B: Bank PPP Loan
Dependent Variable: 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
(1) (2) (3) (4) (5) (6)
1(Black-Owned) -0.049∗∗∗ -0.008∗∗∗ -0.039∗∗∗ -0.025∗∗∗ -0.301∗∗∗ -0.081∗∗∗
(0.002) (0.001) (0.001) (0.001) (0.003) (0.001)
Loan Amount FE No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes
Business Type FE No Yes No Yes No Yes
Employer Status FE No Yes No Yes No Yes
Dep Var Mean 0.159 0.159 0.094 0.094 0.482 0.482
R2 0.001 0.317 0.001 0.131 0.029 0.376
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
43
Panel C: Other Races
Dependent Variable: 1(Fintech) 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
(1) (2) (3) (4) (5) (6) (7) (8)
1(Asian-Owned) -0.319∗∗∗ -0.121∗∗∗ 0.161∗∗∗ 0.030∗∗∗ 0.023∗∗∗ 0.015∗∗∗ 0.091∗∗∗ 0.032∗∗∗
(0.005) (0.002) (0.003) (0.001) (0.001) (0.001) (0.006) (0.002)
1(Hispanic-Owned) -0.231∗∗∗ -0.083∗∗∗ 0.127∗∗∗ 0.018∗∗∗ 0.011∗∗∗ 0.022∗∗∗ 0.066∗∗∗ 0.037∗∗∗
(0.006) (0.002) (0.003) (0.001) (0.001) (0.001) (0.004) (0.002)
1(White-Owned) -0.422∗∗∗ -0.129∗∗∗ 0.028∗∗∗ 0.001 0.044∗∗∗ 0.028∗∗∗ 0.350∗∗∗ 0.100∗∗∗
(0.005) (0.002) (0.002) (0.001) (0.001) (0.001) (0.003) (0.001)
Loan Amount FE No Yes No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes No Yes
Business Type FE No Yes No Yes No Yes No Yes
Employer Status FE No Yes No Yes No Yes No Yes
44 Dep Var Mean 0.174 0.174 0.159 0.159 0.094 0.094 0.482 0.482
R2 0.108 0.357 0.016 0.317 0.003 0.132 0.068 0.378
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
Note: This table reports estimates of Equation 1. The dependent variable in Panel A is an indicator for whether the originating lender is fintech. Panel B repeats the specifications
in columns 1 and 8 for indicators for whether the originating lender is a Top-4 bank (columns 1–2), large bank (columns 3–4), and small/medium-sized bank (columns 5–6).
Panel C repeats Panel B, but adds two columns for fintech loans and considers the other three races/ethnicities. Here, Black-owned businesses represent the single omitted group,
so the coefficients should be interpreted relative to them. Control variables all pertain to the borrower firm and their particular PPP loan. Loan Amount FE are 100 indicator
variables for each percentile of the loan size distribution. Zip Code and Census Tract FE are indicators for each zip code and census tract. Approval Week FE are indicators for
the week in which the PPP loan was approved by SBA. Industry FE are 104 indicators for NAICS 3-digit classifications that appear in the data. Business type FE are 7 indicators
for the firm’s business type (see Table 3). Employer status is an indicator for whether the firm has at least one employee. Standard errors are clustered by borrower zip code.
∗∗∗ ∗∗ ∗
, , indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
Table 5: Signal Strength of Black Business Ownership and PPP Lender Type
Panel A: Fintech PPP Loan
Dependent Variable: 1(Fintech)
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black Owned—First Quintile) 0.188∗∗∗ 0.140∗∗∗ 0.087∗∗∗ 0.110∗∗∗ 0.064∗∗∗ 0.056∗∗∗ 0.057∗∗∗ 0.043∗∗∗
(0.003) (0.003) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
1(Black Owned—Second Quintile) 0.252∗∗∗ 0.196∗∗∗ 0.149∗∗∗ 0.169∗∗∗ 0.111∗∗∗ 0.100∗∗∗ 0.099∗∗∗ 0.080∗∗∗
(0.003) (0.003) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
1(Black Owned—Third Quintile) 0.359∗∗∗ 0.290∗∗∗ 0.233∗∗∗ 0.257∗∗∗ 0.176∗∗∗ 0.158∗∗∗ 0.155∗∗∗ 0.127∗∗∗
(0.003) (0.003) (0.002) (0.003) (0.002) (0.002) (0.002) (0.002)
1(Black Owned—Fourth Quintile) 0.521∗∗∗ 0.429∗∗∗ 0.358∗∗∗ 0.389∗∗∗ 0.275∗∗∗ 0.245∗∗∗ 0.239∗∗∗ 0.193∗∗∗
(0.004) (0.004) (0.003) (0.003) (0.002) (0.002) (0.003) (0.002)
1(Black Owned—Fifth Quintile) 0.663∗∗∗ 0.551∗∗∗ 0.456∗∗∗ 0.500∗∗∗ 0.358∗∗∗ 0.312∗∗∗ 0.299∗∗∗ 0.247∗∗∗
(0.006) (0.005) (0.003) (0.004) (0.003) (0.003) (0.003) (0.003)
Loan Amount FE No Yes Yes Yes Yes Yes Yes Yes
Zip Code FE No No Yes No Yes Yes No Yes
Census Tract FE No No No Yes No No No No
Approval Week FE No No No No Yes Yes Yes Yes
Industry FE No No No No No Yes No Yes
Zip-by-Industry FE No No No No No No Yes No
Business Type FE No No No No No No No Yes
Employer Status FE No No No No No No No Yes
Dep Var Mean 0.174 0.174 0.174 0.174 0.174 0.174 0.174 0.174
R2 0.104 0.151 0.228 0.199 0.297 0.318 0.393 0.359
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
Panel B: Bank PPP Loan
Dependent Variable: 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
(1) (2) (3) (4) (5) (6)
1(First Quantile) 0.018∗∗∗ -0.002 -0.012∗∗∗ -0.007∗∗∗ -0.198∗∗∗ -0.031∗∗∗
(0.002) (0.001) (0.001) (0.001) (0.003) (0.002)
1(Second Quantile) -0.014∗∗∗ -0.007∗∗∗ -0.019∗∗∗ -0.015∗∗∗ -0.227∗∗∗ -0.059∗∗∗
(0.002) (0.001) (0.001) (0.001) (0.003) (0.002)
1(Third Quantile) -0.038∗∗∗ -0.009∗∗∗ -0.035∗∗∗ -0.026∗∗∗ -0.286∗∗∗ -0.086∗∗∗
(0.002) (0.001) (0.001) (0.001) (0.003) (0.002)
1(Fourth Quantile) -0.086∗∗∗ -0.013∗∗∗ -0.056∗∗∗ -0.041∗∗∗ -0.365∗∗∗ -0.123∗∗∗
(0.002) (0.001) (0.001) (0.001) (0.003) (0.002)
1(Fifth Quantile) -0.126∗∗∗ -0.012∗∗∗ -0.072∗∗∗ -0.058∗∗∗ -0.432∗∗∗ -0.152∗∗∗
(0.002) (0.002) (0.001) (0.002) (0.004) (0.003)
Loan Amount FE No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes
Business Type FE No Yes No Yes No Yes
Employer Status FE No Yes No Yes No Yes
Dep Var Mean 0.159 0.159 0.094 0.094 0.482 0.482
R2 0.003 0.317 0.002 0.132 0.031 0.377
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
Note: This table reports estimates of Equation 1. The independent variables are quintiles of the probability that the business is
Black-owned, within the subset of individuals predicted to be Black by our algorithm. The algorithm predicts an individual to be
Black if that is the highest probability race/ethnicity. In the regression models, the omitted group is all borrowers predicted not
Black (as in Table 4 Panels A-B). Dependent variables and controls are as described for Table 4. Standard errors are clustered
by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
45
Table 6: Black Business Ownership and PPP Lender Type with Bank and Credit Relationship Controls
Panel A: Fintech PPP Loan
Dependent variable: 1(Fintech)
(1) (2) (3) (4)
1(Black-Owned) 0.055∗∗∗ 0.055∗∗∗ 0.056∗∗∗ 0.055∗∗∗
(0.004) (0.004) (0.004) (0.004)
1(Credit from Fintech) 0.075∗∗∗ 0.078∗∗∗
(0.003) (0.003)
1(Credit from Conv.) -0.012∗∗∗ -0.011∗∗∗
(0.003) (0.003)
Loan Amount FE Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes
Months Since Statement FE No Yes Yes Yes
Checking Acct Bank FE No Yes Yes Yes
Monthly Cash Inflow FE No No No Yes
Monthly Net Cash Inflow FE No No No Yes
Dep Var Mean 0.363 0.363 0.363 0.363
Observations 168,360 168,360 168,360 168,360
Panel B: Bank PPP Loan
Dependent Variable:
1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank) 1(PPP Lender is)
Checking Acct Bank
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black-Owned) 0.005 -0.000 -0.017∗∗∗ -0.019∗∗∗ -0.047∗∗∗ -0.036∗∗∗ -0.017∗∗∗ -0.011∗∗∗
(0.003) (0.003) (0.003) (0.002) (0.004) (0.003) (0.004) (0.004)
1(Credit from Fintech) -0.025∗∗∗ -0.011∗∗∗ -0.034∗∗∗ -0.040∗∗∗
(0.003) (0.002) (0.003) (0.004)
1(Credit from Conv.) 0.001 0.005∗∗ -0.002 0.011∗∗∗
(0.002) (0.002) (0.003) (0.003)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes Yes
Months Since Statement FE No Yes No Yes No Yes No Yes
Checking Acct Bank FE No Yes No Yes No Yes No Yes
Monthly Cash Inflow FE No Yes No Yes No Yes No Yes
Monthly Net Cash Inflow FE No Yes No Yes No Yes No Yes
Bank Statement Sample Latest Latest Latest Latest Latest Latest Latest Latest
Dep Var Mean 0.177 0.177 0.090 0.090 0.295 0.295 0.274 0.274
Observations 168,360 168,360 168,360 168,360 168,360 168,360 168,360 168,360
Note: This table reports estimates of Equation 1, focusing on the role of bank and credit relationships. The sample is restricted
to bank statement-matched data. We include only information from a firm’s latest statement prior to the loan approval. The
dependent variable in Panel A is an indicator for whether a PPP loan is originated by a fintech lender. The dependent variables
in Panel B are indicators for whether the originating lender is a Top-4 bank (columns 1–2), a large bank (columns 3–4), a
small/medium-sized bank (columns 5–6), or the borrower’s checking account bank (columns 7–8). We report coefficients on
indicators for whether the borrower has previous credit relationships with fintech and non-fintech lenders. Checking Acct Bank
FE are indicators for the bank where the borrower has its main business checking account, so that we compare borrowers who
bank with the same institution. Monthly Net Cash Inflow FE and Monthly Cash Inflow FE are each a set of 100 percentile
indicators for monthly net cash inflow and total cash inflow, respectively. Other controls are as described in Table 4. Standard
errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
46
Table 7: Black Business Ownership and PPP Lender Type by Checking Account Bank Type
Lender is
Dep Var:
Checking Acct
1(Fintech) Banks:
Bank
1(Top 4) 1(Large) 1(Small/Medium)
(1) (2) (3) (4) (5)
Panel A: Sample of Borrowers with Checking Accounts at Top 4 Banks
1(Black-Owned) -0.006 0.051∗∗∗ -0.007 -0.010∗∗∗ -0.034∗∗∗
(0.005) (0.006) (0.005) (0.003) (0.004)
Observations 80,560 80,560 80,560 80,560 80,560
Dep Var Mean 0.248 0.402 0.310 0.048 0.183
Panel B: Sample of Borrowers with Checking Accounts at Non-Top 4 Large Banks
1(Black-Owned) -0.024∗∗ 0.044∗∗∗ 0.009 -0.033∗∗∗ -0.036∗∗∗
(0.010) (0.011) (0.006) (0.010) (0.009)
Observations 26,539 26,539 26,539 26,539 26,539
Dep Var Mean 0.261 0.364 0.062 0.309 0.210
Panel C: Sample of Borrowers with Checking Accounts at Small/Medium-Sized Banks
1(Black-Owned) -0.015 0.056∗∗∗ -0.001 -0.008 -0.045∗∗∗
(0.010) (0.010) (0.005) (0.005) (0.010)
Observations 46,352 46,352 46,352 46,352 46,352
Dep Var Mean 0.355 0.256 0.051 0.053 0.577
Panel D: Sample of Borrowers with Fintech Credit Relationship
1(Black-Owned) 0.019 0.032∗∗ 0.016 -0.009 -0.037∗∗∗
(0.014) (0.014) (0.012) (0.009) (0.011)
Observations 23,890 23,890 23,890 23,890 23,890
Dep Var Mean 0.298 0.339 0.194 0.093 0.302
Business Type FE Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes
Loan Amount FE Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes
Checking Acct Bank FE Yes Yes Yes Yes Yes
Months Since Statement FE Yes Yes Yes Yes Yes
Monthly Cash Inflow FE Yes Yes Yes Yes Yes
Monthly Net Cash Inflow FE Yes Yes Yes Yes Yes
Credit Rel. Controls Yes Yes Yes Yes Yes
Note: This table reports estimates of a modified Equation 1, focusing on various samples of firms with different checking
account bank and credit relationships. Panels A, B, and C limit the sample to PPP borrowers with checking accounts at Top-
4 banks, non-Top-4 large banks, and small/medium banks, respectively. Panel D limits the sample to PPP borrowers who
have previous credit relationship with fintech lenders. Across all panels, the dependent variable in column 1 is an indicator
for whether a PPP loan is originated by the borrower’s checking account bank. The dependent variables in columns 2–5 are
indicators for whether a PPP loan is originated by a fintech lender, Top-4 bank, non-Top-4 large bank, and small/medium bank,
respectively. Controls are as described in Tables 4 and 6. Standard errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate
statistical significance at the 1%, 5%, and 10% levels, respectively.
47
Table 8: Black Business Ownership and PPP Lender Type with Revenue Controls
Panel A: Controls for Average Revenue in Approval and 2 Previous Months
Banks:
Dependent variable: 1(Fintech) 1(Top 4) 1(Large) 1(Small/Med)
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black-Owned) 0.016∗∗∗ 0.016∗∗∗ 0.008∗∗∗ 0.008∗∗∗ -0.002 -0.003 -0.019∗∗∗ -0.018∗∗∗
(0.003) (0.003) (0.003) (0.003) (0.002) (0.002) (0.003) (0.003)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes Yes
Card Revenue FE No Yes No Yes No Yes No Yes
Dep Var Mean 0.097 0.097 0.174 0.174 0.118 0.118 0.521 0.521
Observations 813,812 813,812 813,812 813,812 813,812 813,812 813,812 813,812
Panel B: Sample Restricted to Firms Struggling during Approval Month
Banks:
Dependent variable: 1(Fintech)
1(Top 4) 1(Large) 1(Small/Med)
(1) (2) (3) (4)
1(Black-Owned) 0.021∗∗∗ 0.006 -0.011∗∗∗ -0.019∗∗∗
(0.005) (0.005) (0.004) (0.006)
Loan Amount FE Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes
Dep Var Mean 0.105 0.170 0.119 0.512
R2 0.205 0.343 0.186 0.367
Observations 203,357 203,357 203,357 203,357
Note: This table reports estimates of a modified Equation 1. In Panel A, we add controls for firm revenue from credit and debit
card transactions during and prior to the PPP loan approval month. The sample is restricted to those matched to Enigma data on
credit and debit card transactions. We take the mean card revenue over the loan approval month and previous two months, then
construct card revenue FE as a set of 100 indicators for each percentile of average monthly card revenue. In Panel B, we restrict
the sample to firms that appear especially harmed by the COVID-19 economic crisis. These are firms for which we observe
monthly revenue in February 2020 but not in the approval month. Note that we do not observe revenue if there are fewer than 30
transactions. Therefore, these “struggling” firms have either no activity or limited activity relative to February. Other controls
are as described in Table 4. Standard errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the
1%, 5%, and 10% levels, respectively.
48
Table 9: Black Business Ownership and Fintech PPP Loans using Loan Applications via Lendio
Dependent Variable: 1(Fintech PPP)
(1) (2) (3) (4) (5)
1(Black-owned) 0.029∗∗∗ 0.026∗∗∗ 0.028∗∗∗ 0.048∗∗∗ 0.039∗∗∗
(0.004) (0.001) (0.003) (0.006) (0.008)
Loan Amount FE Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes
Application Week FE No N/A Yes Yes Yes
# Fintech × # Non-Fintech Sent FE No N/A Yes Yes Yes
Lenders Sent All N/A All Any Conv. Only Conv.
Sample Applied via Lendio Banks, Not Applied via Applied via Applied via
Lendio Applied via Lendio Lendio Lendio
Lendio
Dep Var Mean 0.516 0.771 0.516 0.240 0.108
Observations 196,294 692,926 196,294 68,566 45,761
Note: This table reports estimates of a modified Equation 1, focusing on whether Black-owned firms that applied to
conventional lenders via the Lendio platform were nonetheless more likely to end up with a fintech PPP loan. The sample
is restricted to Lendio applications linked to a PPP loan in the SBA data. The dependent variable is an indicator for whether
a PPP loan is originated by a fintech lender. Column 1 includes all merged Lendio applications. Column 2 assesses whether
the coefficient magnitude in column 1 relative to Table 4 column 8 reflects Lendio partner lender composition. It uses the
sample of non-Lendio PPP loans that were originated by the subset of lenders who appear in the Lendio data. Column 3 repeats
column 1 but adds a set of indicators controlling for the number of lenders that an application was sent to in both the fintech
and non-fintech (“conventional”) categories and indicators for the week of Lendio application. Column 4 restricts the sample to
borrowers who were sent to at least one conventional lender. Column 5 restricts the sample to borrowers who were sent only to
conventional lenders; that is, borrowers who were sent to both fintech and conventional lenders are excluded from the sample.
Other controls are as described in Table 4. Standard errors are clustered by zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at
the 1%, 5%, and 10% levels, respectively.
49
Table 10: Black Business Ownership and Lender Identity: The Effect of Racial Animus
Panel A: Fintech PPP Loans as Dependent Variable
(1) (2) (3) (4) (5) (6) (7)
1(Black-Owned) 0.121∗∗∗ 0.124∗∗∗ 0.121∗∗∗ 0.120∗∗∗ 0.121∗∗∗ 0.117∗∗∗ 0.107∗∗∗
(0.002) (0.001) (0.002) (0.001) (0.001) (0.001) (0.001)
1(Black-Owned) × Animus 0.004∗∗ 0.014∗∗∗ 0.013∗∗∗ 0.011∗∗∗ 0.016∗∗∗ 0.029∗∗∗
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Racial Animus Measure Stephens- IAT (Implicit) IAT (Explicit) Segregation Segregation
Nationscape
Davidowitz (Dissimilarity) (Isolation)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes
Dep Var Mean 0.174 0.174 0.174 0.174 0.174 0.174 0.174
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
Panel B: Top-4 Bank PPP Loans as Dependent Variable
(1) (2) (3) (4) (5) (6) (7)
1(Black-Owned) -0.008∗∗∗ -0.009∗∗∗ -0.008∗∗∗ -0.009∗∗∗ -0.008∗∗∗ -0.008∗∗∗ -0.009∗∗∗
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
1(Black-Owned) × Animus 0.010∗∗∗ -0.001 0.008∗∗∗ 0.004∗∗∗ -0.004∗∗∗ 0.004∗∗∗
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Racial Animus Measure Stephens- IAT (Implicit) IAT (Explicit) Segregation Segregation
Nationscape
Davidowitz (Dissimilarity) (Isolation)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes
Dep Var Mean 0.159 0.159 0.159 0.159 0.159 0.159 0.159
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
Panel C: Non-Top 4 Bank PPP Loans as Dependent Variable
(1) (2) (3) (4) (5) (6) (7)
1(Black-Owned) -0.106∗∗∗ -0.107∗∗∗ -0.105∗∗∗ -0.103∗∗∗ -0.105∗∗∗ -0.103∗∗∗ -0.093∗∗∗
(0.002) (0.001) (0.001) (0.001) (0.001) (0.002) (0.001)
1(Black-Owned) × Animus -0.013∗∗∗ -0.013∗∗∗ -0.024∗∗∗ -0.017∗∗∗ -0.012∗∗∗ -0.029∗∗∗
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Racial Animus Measure Stephens- IAT (Implicit) IAT (Explicit) Segregation Segregation
Nationscape
Davidowitz (Dissimilarity) (Isolation)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes
Dep Var Mean 0.576 0.576 0.576 0.576 0.576 0.576 0.576
Observations 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623 4,183,623
Note: This table reports estimates of a modified Equation 1, focusing on the interaction between the indicator for Black-owned
business and a standardized measure of racial animus in the borrower location. The dependent variable differs across the three
panels: In Panel A, it is an indicator for a fintech PPP loan, in Panel B, it is an indicator for a Top-4 bank PPP loan, and in Panel
C, it is an indicator for a non-Top 4 bank PPP loan. In each panel, column 1 includes the same controls as the specification in
Table 4 Panel A column 8. The racial animus measures are as follows: column 2 uses the number of racially charged searches
in a designated media market (DMA); column 3 uses responses to the question on favorability toward Black people in the
Nationscape survey aggregated to the congressional district level; columns 4–5 use the implicit and explicit score from the
Implicit Association Test (IAT) aggregated to the county level; columns 6–7 use the dissimilarity and isolation index at the
metropolitan statistical area (MSA) level. All racial animus measures are standardized at their respective levels of geography,
weighted by the number of PPP loans. Controls are as described in Tables 4 and 6. Standard errors are clustered by zipcode.
∗∗∗ ∗∗ ∗
, , indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
50
Table 11: Effect of Automation during PPP on Lending to Black-Owned Small Businesses
Panel A: Loan Share by Race and Ethnicity by Bank Automation
Dependent Variable: 1(Black-Owned) 1(Owned by:)
Hispanic Asian White
(1) (2) (3) (4) (5)
1(After Automation) 0.060∗∗∗ 0.043∗∗∗ 0.008∗∗∗ 0.009∗∗∗ -0.060∗∗∗
(0.003) (0.003) (0.002) (0.003) (0.004)
Bank FE Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes
Loan Amount FE No Yes Yes Yes Yes
Zip Code FE No Yes Yes Yes Yes
Industry FE No Yes Yes Yes Yes
Business Type FE No Yes Yes Yes Yes
Employer Status FE No Yes Yes Yes Yes
Dep Var Mean 0.037 0.037 0.043 0.055 0.865
Observations 2,024,674 2,024,674 2,024,674 2,024,674 2,024,674
Panel B: Loan Share by Race and Ethnicity by Bank Automation and Racial Animus
Dependent variable: 1(Black-Owned)
(1) (2) (3) (4) (5) (6)
1(After Automation) 0.045∗∗∗
0.042∗∗∗
0.041 ∗∗∗
0.042 ∗∗∗
0.037 ∗∗∗
0.041∗∗∗
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
1(After Automation) × 0.004∗ 0.005 0.009∗∗∗ 0.008∗∗∗ 0.025∗∗∗ 0.001
Racial Animus (0.002) (0.003) (0.003) (0.002) (0.002) (0.002)
Racial Animus Measure Stephens- Nationscape IAT (Implicit) IAT (Explicit) Segregation Segregation
Davidowitz (Isolation) (Dissimilarity)
Bank FE Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes
Loan Amount FE Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes
Dep Var Mean 0.037 0.037 0.037 0.037 0.037 0.037
Observations 2,024,674 2,024,674 2,024,674 2,024,674 2,024,674 2,024,674
Note: This table reports estimates of Equation 2, estimated on the sample of PPP loans extended by small banks. Columns 1 and
2 of Panel A shows the effect of automation on the probability that a loan is extended to a Black-owned businesses. Columns 3-
5 consider effects on lending to Hispanic-, Asian-, and White-Owned businesses, respectively, using the fully controlled model
from column 2. Panel B interacts the automation indicator with our measures of local racial animus (from Table 10), continuing
to use the fully controlled model from Panel A, column 2. All racial animus measures are standardized at their respective levels
of geography, weighted by the number of PPP loans. Controls are as described in Tables 4 and 6. Standard errors are clustered
by zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
51
Appendix
(For Online Publication)
sss
Figure A.1: Race Probability Distributions
(A) Black-Owned Businesses (B) White-Owned Businesses
5
6
4
3 4
Density Density
2
2
1
0 0
.2 .4 .6 .8 1 .2 .4 .6 .8 1
Probability Probability
(C) Hispanic-Owned Businesses (D) Asian-Owned Businesses
4 5
4
3
3
Density
2 Density
2
1
1
0 0
.2 .4 .6 .8 1 .2 .4 .6 .8 1
Probability Probability
Note: This figure shows the probability distribution for each race generated by our algorithm. Specifically, each graph contains
the sample of borrowers predicted by the algorithm to be the particular race, which means that the race has the highest
probability. For example, Panel A contains the subset of borrowers whose highest probability race is Black. The graph shows
the algorithm’s predicted chance that they are Black.
Internet Appendix 1
Figure A.2: Black-Owned Business PPP Lending by Institution Type (Self-Reported)
Note: This figure shows shares of PPP loans made to businesses that self-identify as Black-owned by lender type. The sample
limits to 1,098,682 loans to businesses for which the data includes self-reported race.
Internet Appendix 2
Figure A.3: Hispanic-Owned Business PPP Lending by Institution Type
(A) Share of PPP Loans to Hispanic-Owned Business by Institution Type
(B) Share of PPP Lender Institution Type among Hispanic-Owned Business
Note: This figure shows the shares of PPP loans made by originating lender type to Hispanic-owned businesses. Panel
A shows the Hispanic share of PPP loans made by originating lender type (P (Hispanic-owned|Originating Lender Type)).
Panel B shows the shares of PPP loans from originating lender type made to Black-owned businesses
(P (Originating Lender Type|Hispanic-owned)).
Internet Appendix 3
Figure A.4: Asian-Owned Business PPP Lending by Institution Type
(A) Share of PPP Loans to Asian-Owned Business by Institution Type
(B) Share of PPP Lender Institution Type among Asian-Owned Business
Note: This figure shows the shares of PPP loans made by originating lender type to Asian-owned businesses. Panel A shows
the Asian share of PPP loans made by originating lender type (P (Asian-owned|Originating Lender Type)). Panel B shows the
shares of PPP loans from originating lender type made to Asian-owned businesses (P (Originating Lender Type|Asian-owned)).
Internet Appendix 4
Figure A.5: White-Owned Business PPP Lending by Institution Type
(A) Share of PPP Loans to White-Owned Business by Institution Type
(B) Share of PPP Lender Institution Type among White-Owned Business
Note: This figure shows the shares of PPP loans made by originating lender type to White-owned businesses. Panel
A shows the White share of PPP loans made by originating lender type (P (White-owned|Originating Lender Type)).
Panel B shows the shares of PPP loans from originating lender type made to White-owned businesses
(P (Originating Lender Type|White-owned)).
Internet Appendix 5
Figure A.6: Geographic Distribution of PPP Loans to Black-Owned Businesses
(A) PPP Borrowers (Analysis Sample)
(B) Black Share of Population
Note: Panel A of this figure shows the geographic distribution of PPP loans to Black-Owned businesses. Panel B shows the
geographic distribution of the Black population, measured as the share of Black people in the county. These data are from the
2019 U.S. Census Bureau ACS. Internet Appendix 6
Figure A.7: Black-Owned Business PPP Lending by Institution Type and Round
(A) Round 1 (4/3/2020–4/16/2020) (B) Round 2 Early (4/27/2020–5/13/2020)
JPMC Bank
Citibank
Small Banks
Minority Dep Inst
Medium Banks
Large Banks
Bank of America
Wells Fargo Bank
Credit Union
CDFI/Nonprofit
Fintech
0 5 10 15
Share of All Loans Made by Lender Type to Black-Owned Businesses
(C) Round 2 Late (5/14/2020–8/9/2020) (D) Round 3 (1/12/2021–2/23/2021)
Citibank Small Banks
Small Banks Minority Dep Inst
JPMC Bank Citibank
Minority Dep Inst Medium Banks
Medium Banks JPMC Bank
Wells Fargo Bank Bank of America
Large Banks Large Banks
Bank of America Wells Fargo Bank
Credit Union Credit Union
CDFI/Nonprofit CDFI/Nonprofit
Fintech Fintech
0 10 20 30 0 10 20 30 40 50
Share of All Loans Made by Lender Type to Black-Owned Businesses Share of All Loans Made by Lender Type to Black-Owned Businesses
Note: This figure shows the shares of PPP loans made by originating lender type to Black-owned businesses by PPP round.
Panel A limits the sample to Round 1 PPP approvals. Panel B limits the sample to Round 2 early approvals. Panel C limits the
sample to Round 2 late approvals. We distinguish Round 2 late from early (where early is intended to represent the initial rush)
by defining early as ending on the last day on which there were at least 30,000 loans issued. The results are not sensitive to
using an alternative threshold. Panel D limits the sample to Round 3 approvals.
Internet Appendix 7
Figure A.8: Weekly Balanced Sample Event Study fo the Share of Loans to Black-Owned
Businesses Before and After Small Bank Automation
Note: This figure reports dynamic differences-in-differences estimates, using Equation 3. We employ a weekly balanced
panel. Here, we only include the 11 automating banks that have observations at least six weeks on both sides of the automation
date. We include bank and time fixed effects (as in Table 11 Panel A column 1). Standard errors are clustered by zip code. The
grey bars represent 95% confidence intervals.
Internet Appendix 8
Table A.1: List of fintechs
Lender Num Loans Median Loan Amt (Thou) Share Black Borrowers
Cross River Bank 174,436 18,540 0.250
Kabbage 137,687 10,416 0.275
Celtic Bank Corporation 125,601 9,569 0.183
Lendio 101,828 25,000 0.074
WebBank 64,522 13,750 0.098
Customers Bank 61,993 11,524 0.112
Readycap Lending 40,272 20,800 0.056
Itria Ventures 24,459 17,102 0.430
Intuit Financing 14,472 18,424 0.044
Newtek Small Business Finance 13,018 16,500 0.111
Fundbox 11,401 12,588 0.234
MBE Capital Partners 5,330 15,980 0.413
FC Marketplace 5,072 25,486 0.077
Harvest Small Business Finance 4,298 76,298 0.045
Fountainhead SBF 2,684 70,000 0.061
CRF Small Business Loan Company 2,137 20,800 0.175
Sunrise Banks National Association 1,749 23,900 0.094
Accion 1,298 13,211 0.054
Fund-Ex Solutions Group 1,172 75,000 0.044
The Bancorp Bank 1,159 68,700 0.020
Centerstone SBA Lending 980 63,850 0.012
Grow America Fund 593 29,517 0.202
Evolve Bank and Trust 517 25,000 0.128
NBKC Bank 303 14,600 0.079
immito 159 23,000 0.302
Loan Source 150 25,000 0.060
BayBank 145 21,526 0.014
VelocitySBA 81 77,200 0.012
All 797,516 15,777 0.184
Note: This table lists the firms we identify as fintech lenders. We report their number of loans, their median loan amount, and
the share of their PPP loans made to Black-owned businesses.
Internet Appendix 9
Table A.2: Summary Statistics by Lender Type for Sample with Predicted Race
Panel A: All First Draw PPP Loans Before Feb 24th with Race Prediction
PPP Loan Amount ($)
Number Number Share Total
Lenders Loans Loans Amt ($ bn) Mean P10 P50 P90
All 4,846 4,212,001 100% 394.4 93,642 4,465 20,833 176,900
Bank of America 1 267,283 6.3% 20.1 75,217 4,500 20,833 134,225
Citibank 1 22,973 0.5% 2.5 108,931 8,200 29,500 200,000
JPMC Bank 1 221,750 5.3% 22.5 101,379 6,465 27,442 188,998
Wells Fargo Bank 1 153,416 3.6% 8.1 53,110 4,160 18,825 102,794
Large Banks 17 411,150 9.8% 59.0 143,532 6,200 31,500 272,847
Medium Banks 377 1,022,007 24.3% 145.8 142,626 6,200 33,600 295,600
Small Banks 3,186 1,000,417 23.8% 87.2 87,183 4,509 20,833 174,500
Credit Union 926 173,966 4.1% 7.9 45,462 3,550 16,000 89,600
CDFI/Nonprofit 176 82,547 2.0% 5.9 71,615 3,906 20,800 145,833
Minority Dep Inst 132 127,778 3.0% 12.6 98,436 4,800 22,100 190,217
Fintech 28 728,714 17.3% 22.8 31,274 2,855 15,343 54,564
Panel B: Bank and Credit Relationships Sample with Race Prediction (Ocrolus)
SME Has
Checking
Credit With Any: Monthly Net Cash Inflow ($)
N Acct with
PPP Lender Fintech Non-Fintech Mean P10 P50 P90
All 169,818 27.7% 14.2% 80.1% 9,047 -36,399 1,340 62,189
Bank of America 12,036 64.1% 15.7% 86.9% 8,278 -47,275 1,479 69,427
Citibank 1,084 51.8% 12.0% 66.1% 9,706 -45,705 961 78,408
JPMC Bank 9,917 69.7% 17.7% 90.0% 15,144 -64,159 4,641 112,415
Wells Fargo Bank 6,805 73.5% 13.0% 87.8% 14,252 -35,662 3,949 78,582
Large Banks 16,032 49.4% 14.6% 82.5% 12,186 -42,680 2,551 77,493
Medium Banks 28,595 38.7% 14.5% 75.5% 9,445 -51,446 1,494 76,657
Small Banks 21,362 22.7% 14.4% 75.4% 7,655 -51,202 1,254 71,749
Credit Union 5,670 34.6% 13.1% 72.1% 6,821 -27,779 829 49,106
CDFI/Nonprofit 3,060 8.5% 13.7% 79.2% 8,311 -36,405 1,189 57,953
Minority Dep Inst 4,003 21.4% 14.2% 75.5% 5,395 -54,560 1,169 63,714
Fintech 61,254 0.0% 13.3% 80.7% 7,580 -19,331 932 40,861
Note: This table reports summary statistics about PPP loans by originating lender type, where each PPP loan is assigned
to a single type. We restrict the sample to loans for which we can predict the business owner’s race. In Panel B, the first
column, ”SME has Checking Account with PPP Lender” means that the borrower’s business checking account bank is the same
institution that originated their PPP loan. The remaining variables are derived from transactions on the borrowers’ most recent
monthly bank statement.
Internet Appendix 10
Table A.3: Race Prediction Out of Sample Confusion Matrix
Panel A: Prediction in Full Self-Identified Sample
Predicted
Observed Asian Black Hispanic Other White All
Asian 10.7% 0.0% 0.0% N/A 0.4% 11.2%
Black 0.0% 9.1% 0.0% N/A 0.33% 9.5%
Hispanic 0.0% 0.0% 9.7% N/A 0.6% 10.3%
Other N/A N/A N/A N/A N/A 2.7%
White 0.1% 0.1% 0.1% N/A 69.6% 69.8%
All 10.8% 9.6% 9.8% N/A 69.9% 100%
Panel B: Prediction in Hold-Out Self-Identified Sample
Predicted
Observed Asian Black Hispanic Other White All
Asian 7.8% 0.2% 0.2% 0.0% 2.2% 10.4%
Black 0.1% 5.7% 0.2% 0.0% 3% 9.0%
Hispanic 0.2% 0.3% 5.9% 0.0% 2.8% 9.2%
Other 0.2% 0.2% 0.1% 0.0% 1.8% 2.3%
White 1.5% 1.2% 1.5% 0.0% 64.9% 69.1%
All 9.8% 7.6% 7.9% 0.0% 74.7% 100%
Panel C: Summary Statistics of Race Probability Distributions
N Mean Median SD
Predicted Asian-Owned 372,993 0.801 0.846 0.166
Predicted Black-Owned 359,366 0.758 0.796 0.186
Predicted Hispanic-Owned 313,389 0.776 0.814 0.167
Predicted White-Owned 3,137,875 0.851 0.902 0.143
Note: Panel A of this table shows the race prediction of the random forest model for the full sample of self-reported individuals
with geolocated addresses, including those on which the model was trained. We do not retain the Other prediction or use it in
analysis. The sample is 809,119. Panel B of this table shows the out-of-sample validation of the random forest model. It is
restricted to the hold-out sample of self-reported borrowers that was not used to train the random forest model, and contains
234,632 observations. (The random forest model with 1,000 trees was trained on 574,487 observations with self-reported race.)
In both Panel A and B, the percents represent percent of total observations in the sample. Panel C contains summary statistics
about the probability distributions by predicted race/ethnicity.
Internet Appendix 11
Table A.4: Additional Sample Characteristics
Full PPP Sample PPP Analysis Sample Matched to Ocrolus
N = 5,663,719 N = 4,183,623 N = 168,360
Mean P50 Mean P50 Mean P50
Lender Type
Bank of America 6.1% 6.4% 7.1%
Citibank 0.6% 0.5% 0.6%
JPMC Bank 5.1% 5.3% 5.9%
Wells Fargo Bank 3.6% 3.7% 4.0%
Large Banks 10.1% 9.4% 9.0%
Medium Banks 24.8% 24.3% 16.9%
Small Banks 24.0% 23.9% 12.6%
Credit Union 4.0% 4.1% 3.4%
CDFI/Nonprofit 2.0% 2.0% 1.8%
Minority Dep Inst 3.6% 3.0% 2.4%
Fintech 16.3% 17.4% 36.3%
Borrower Firm Business Type
Corporation 27.0% 27.8% 31.6%
LLC 28.9% 26.8% 31.6%
Nonprofit 3.4% 2.8% 1.0%
Other 0.6% 0.5% 0.3%
SelfEmployed 7.8% 8.6% 8.5%
Sole Proprietorship 19.5% 20.4% 16.9%
Subchapter S Corporation 12.8% 13.1% 10.2%
Borrower Firm Characteristics
Employer Status 64.6% 63.4% 68.1%
Loan Amount ($) 93,802 20,833 93,666 20,833 80,897 21,121
Borrower Firm Industry
Professional/Technical Services 12.5% 12.7% 13.7%
Ambulatory Health Care Services 7.5% 7.5% 6.5%
Food and Drinking Services 5.9% 5.9% 6.5%
Personal and Laundry Services 5.8% 6.0% 6.2%
Specialty Trade Contractors 5.4% 5.4% 5.5%
Other 62.8% 62.6% 61.6%
Borrower Firm Census Divisions
East North Central 13.3% 14.8% 12.3%
East South Central 5.2% 5.2% 3.0%
Middle Atlantic 12.6% 8.9% 10.4%
Mountain 7.2% 7.3% 7.1%
New England 4.9% 5.4% 3.8%
Pacific 15.3% 16.7% 22.8%
South Atlantic 19.4% 20.6% 26.6%
West North Central 9.6% 9.3% 2.8%
West South Central 11.7% 11.7% 11.2%
Borrower Race/Ethnicity
Share Black-Owned 8.6% 15.3%
Share Hispanic/Latino-Owned 7.5% 11.0%
Share Asian-Owned 8.9% 10.2%
Share White-Owned 75.0% 63.5%
Note: This table shows additional summary statistics across three samples: the sample with all PPP loans, analysis sample for
which we have successfully predicted borrower race, and the bank statement-matched sample. We highlight the distribution of
the top-five 3-digit NAICS industries, and code the rest as “Other.”
Internet Appendix 12
Table A.5: Sample Characteristics for Employers and Non-employer Firms
Panel A: Analysis Sample - Employers
N = 2,695,027
Share Mean Loan Share Share Share Share
Loans Amt ($) Asian Borr Black Borr Hisp Borr White Borr
All 100% 140,318 10.4% 2.9% 6.9% 79.8%
Bank of America 7.6% 95,213 19.6% 4.2% 12.7% 63.5%
Citibank 0.6% 151,187 19.4% 2.5% 10.8% 67.3%
JPMC Bank 6.3% 129,937 15.7% 2.4% 9.8% 72.1%
Wells Fargo Bank 3.9% 73,101 13.9% 3.7% 13.0% 69.4%
Large Banks 11.4% 185,867 8.0% 2.9% 5.3% 83.7%
Medium Banks 28.5% 186,355 7.1% 1.8% 5.2% 85.9%
Small Banks 22.9% 134,582 5.7% 1.4% 3.5% 89.4%
Credit Union 3.6% 73,705 6.3% 3.9% 6.6% 83.2%
CDFI/Nonprofit 1.9% 109,627 8.7% 4.3% 6.1% 80.9%
Minority Dep Inst 3.2% 139,354 32.2% 1.9% 11.0% 54.9%
Fintech 10.0% 66,944 15.6% 8.2% 11.4% 64.8%
Panel B: Analysis Sample - Non-Employers
N = 1,547,701
Share Mean Loan Share Share Share Share
Loans Amt ($) Asian Borr Black Borr Hisp Borr White Borr
All 100% 11,825 6.4% 18.7% 8.7% 66.2%
Bank of America 4.3% 11,625 12.3% 15.3% 11.5% 60.9%
Citibank 0.5% 16,377 13.7% 7.1% 11.2% 68.0%
JPMC Bank 3.6% 12,322 11.6% 10.9% 9.4% 68.1%
Wells Fargo Bank 3.2% 11,220 7.8% 16.5% 11.0% 64.8%
Large Banks 5.9% 12,127 6.7% 13.5% 6.4% 73.4%
Medium Banks 16.7% 12,247 5.4% 10.9% 5.5% 78.1%
Small Banks 25.0% 12,006 2.0% 6.4% 2.7% 88.8%
Credit Union 5.0% 10,095 3.9% 17.2% 7.4% 71.5%
CDFI/Nonprofit 2.1% 10,832 6.6% 20.7% 9.0% 63.7%
Minority Dep Inst 2.7% 11,372 12.0% 8.2% 19.0% 60.9%
Fintech 31.0% 11,741 8.7% 36.6% 14.2% 40.5%
Note: This table shows loan characteristics and borrower race and ethnic breakdown by lender type for employer firms (Panel
A) and non-employer firms (Panel B).
Internet Appendix 13
Table A.6: Fintech PPP Loans and the Zip Code’s Black Share of Population
Dependent Variable: 1(Fintech)
Sample of Borrowers: All Black White
(1) (2) (3) (4)
1(Black-Owned) 0.159∗∗∗ 0.131∗∗∗
(0.002) (0.002)
zipshare black 0.125∗∗∗ 0.089∗∗∗ 0.078∗∗∗
(0.004) (0.004) (0.004)
Loan Amount FE Yes Yes Yes Yes
State FE Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes
Dep Var Mean 0.174 0.174 0.536 0.114
Observations 4,183,623 4,183,623 359,366 3,137,875
Note: This table reports estimates of a modified Equation 1, focusing on the role of the Black population in the borrower firm’s
neighborhood. Columns 2-4 include a continuous variable for the Black share of the population in a zip code. Columns 3-4
limit the sample to only Black-owned and White-owned businesses, respectively. We include state FE (indicators for each U.S.
state and territory). Other controls are as described in Table 4. Standard errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗
indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
Internet Appendix 14
Table A.7: PPP Round Characteristics
Round 1 Round 2 (Early) Round 2 (Late) Round 3
N = 1,176,645 N = 1,830,721 N = 744,242 N = 432,015
Mean P50 Mean P50 Mean P50 Mean P50
Lender Type
Bank of America 0.6% 12.3% 4.2% 0.6%
Citibank 0.4% 0.8% 0.3% 0.1%
JPMC Bank 1.7% 9.3% 3.2% 1.8%
Wells Fargo Bank 0.1% 6.3% 3.9% 1.7%
Large Banks 11.5% 10.2% 7.1% 4.2%
Medium Banks 37.9% 23.0% 13.1% 12.3%
Small Banks 33.9% 17.1% 12.7% 44.6%
Credit Union 3.4% 4.5% 4.9% 3.3%
CDFI/Nonprofit 2.2% 1.5% 2.4% 2.3%
Minority Dep Inst 4.2% 2.7% 1.9% 3.2%
Fintech 4.0% 12.2% 46.3% 25.8%
Borrower Firm Business Type
Corporation 35.7% 32.5% 16.8% 5.5
LLC 32.4% 29.0% 21.0% 12.1
Non-Profit 4.2% 2.7% 1.5% 0.2
Self-Employed 1.1% 5.5% 21.9% 19.5
Sole Proprietorship 7.6% 16.5% 29.4% 55.6
Subchapter S Corporation 18.3% 13.1% 9.2% 5.9
Other 0.6% 0.6% 0.2% 1.2
Borrower Firm Characteristics
Employer Status 88.7% 68.0% 38.8% 17.8%
Loan Amount ($) 200,471 58,600 68,394 20,833 29,148 13,344 21,010 15,650
Borrower Firm Industry
Professional/Technical Services 12.8% 13.6% 13.5% 7.2%
Ambulatory Health Care Services 8.5% 8.4% 6.5% 2.5%
Food and Drinking Services 7.9% 5.7% 4.9% 3.1%
Personal and Laundry Services 2.9% 5.3% 11.0% 8.4%
Specialty Trade Contractors 6.6% 5.3% 4.7% 3.5%
Other 61.2% 61.8% 59.5% 75.2%
Borrower Firm Census Divisions
East North Central 17.0% 13.3% 13.4% 17.6%
East South Central 6.5% 4.4% 4.3% 6.6%
Middle Atlantic 7.4% 9.7% 11.1% 5.9%
Mountain 8.2% 7.7% 6.0% 5.1%
New England 7.2% 5.4% 4.5% 2.5%
Pacific 11.7% 20.8% 19.1% 8.6%
South Atlantic 17.9% 21.9% 25.7% 14.1%
West North Central 10.8% 5.7% 4.9% 27.8%
West South Central 13.2% 11.2% 10.5% 11.7%
Borrower Race/Ethnicity
Share Black-Owned 2.1% 6.1% 19.9% 17.5%
Share Hispanic/Latino-Owned 3.7% 8.1% 12.4% 7.1%
Share Asian-Owned 5.9% 10.7% 11.3% 5.3%
Share White-Owned 88.3% 75.2% 56.4% 70.1%
Note: This table shows summary statistics of PPP borrowers across the various PPP rounds for the analysis sample.
Round 1: 4/3/2020–4/16/2020; Round 2 Early: 4/27/2020–5/13/2020; Round 2 Late: 5/14/2020–8/9/2020; Round 3:
1/12/2021–2/23/2021. We highlight the distribution of the top-five 3-digit NAICS industry, and code the rest as “Other.”
Internet Appendix 15
Table A.8: Black Business Ownership and PPP Lender Type by PPP Round
Dependent Variable: 1(Fintech)
PPP Round: 1 2 (Early) 2 (Late) 3
(1) (2) (3) (4)
1(Black-Owned) 0.025∗∗∗ 0.035∗∗∗ 0.087∗∗∗ 0.177∗∗∗
(0.002) (0.001) (0.002) (0.003)
Loan Amount FE Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes
Dep Var Mean 0.040 0.122 0.461 0.257
Observations 1,181,767 1,846,203 747,991 436,060
Note: This table reports estimates of Equation 1, specifically replicating the specification in Table 4 column 8 for each PPP
round. We distinguish Round 2 late from early (where early is intended to represent the initial rush) by defining early as ending
on the last day on which there were at least 30,000 loans issued. The results are not sensitive to using an alternative threshold.
Controls are as described in Table 4. Standard errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance
at the 1%, 5%, and 10% levels, respectively.
Internet Appendix 16
Table A.9: Black Business Ownership and PPP Lender Type – Employer Firms Only
Panel A: Fintech PPP Loan
Dependent Variable: 1(Fintech)
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black-Owned) 0.186∗∗∗ 0.165∗∗∗ 0.119∗∗∗ 0.142∗∗∗ 0.081∗∗∗ 0.080∗∗∗ 0.073∗∗∗ 0.075∗∗∗
(0.004) (0.004) (0.003) (0.003) (0.002) (0.002) (0.002) (0.002)
Loan Amount FE No Yes Yes Yes Yes Yes Yes Yes
Zip Code FE No No Yes No Yes Yes No Yes
Census Tract FE No No No Yes No No No No
Approval Week FE No No No No Yes Yes Yes Yes
Industry FE No No No No No Yes No Yes
Zip-by-Industry FE No No No No No No Yes No
Business Type FE No No No No No No No Yes
Dep Var Mean 0.097 0.097 0.097 0.097 0.097 0.097 0.097 0.097
R2 0.011 0.029 0.076 0.061 0.145 0.164 0.272 0.177
Observations 2,653,490 2,653,490 2,653,490 2,653,490 2,653,490 2,653,490 2,653,490 2,653,490
Panel B: Bank PPP Loan
Dependent Variable: 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
(1) (2) (3) (4) (5) (6)
1(Black-Owned) 0.036∗∗∗ -0.008∗∗∗ 0.001 -0.018∗∗∗ -0.235∗∗∗ -0.056∗∗∗
(0.003) (0.002) (0.002) (0.002) (0.004) (0.002)
Loan Amount FE No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes
Business Type FE No Yes No Yes No Yes
Dep Var Mean 0.184 0.184 0.114 0.114 0.517 0.517
R2 0.000 0.335 0.000 0.139 0.006 0.346
Observations 2,653,490 2,653,490 2,653,490 2,653,490 2,653,490 2,653,490
Note: This table reports estimates of Equation 1 on the sample consisting of only employer firms. The dependent variable in
Panel A is an indicator for whether the originating lender is fintech. Panel B repeats the specifications in columns 1 and 8 for
indicators for whether the originating lender is a Top-4 bank (columns 1–2), large bank (columns 3–4), and small/medium-sized
bank (columns 5-6). Control variables all pertain to the borrower firm and their particular PPP loan. Controls are as described
in Table 4. Standard errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10%
levels, respectively.
Internet Appendix 17
Table A.10: Business Owner Race and PPP Lender Type – Excluding Fintech Lenders from Sample
Panel A: Bank PPP Loans to Black-Owned Businesses
Dependent Variable: 1(PPP Lender is 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
Checking Acct Bank)
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black-Owned) -0.033∗∗∗ -0.013∗∗ 0.059∗∗∗ 0.010∗∗∗ 0.008∗∗∗ -0.019∗∗∗ -0.142∗∗∗ -0.022∗∗∗
(0.006) (0.007) (0.003) (0.001) (0.002) (0.001) (0.003) (0.002)
Loan Amount FE No Yes No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes No Yes
Business Type FE No Yes No Yes No Yes No Yes
Employer Status FE No Yes No Yes No Yes No Yes
Dep Var Mean 0.286 0.286 0.159 0.159 0.098 0.098 0.480 0.480
R2 0.000 0.202 0.001 0.388 0.000 0.149 0.004 0.330
Observations 108,564 108,564 3,483,287 3,483,287 3,483,287 3,483,287 3,483,287 3,483,287
Panel B: Bank PPP Loans to White-Owned Businesses
Dependent Variable: 1(PPP Lender is 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
Checking Acct Bank)
(1) (2) (3) (4) (5) (6) (7) (8)
1(White-Owned) 0.011∗∗∗ 0.010∗∗ -0.155∗∗∗ -0.023∗∗∗ -0.013∗∗∗ -0.017∗∗∗ 0.235∗∗∗ 0.073∗∗∗
(0.004) (0.004) (0.003) (0.001) (0.001) (0.001) (0.003) (0.001)
Loan Amount FE No Yes No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes No Yes
Business Type FE No Yes No Yes No Yes No Yes
Employer Status FE No Yes No Yes No Yes No Yes
Dep Var Mean 0.286 0.286 0.159 0.159 0.098 0.098 0.480 0.480
R2 0.000 0.202 0.025 0.388 0.000 0.150 0.036 0.333
Observations 108,564 108,564 3,483,287 3,483,287 3,483,287 3,483,287 3,483,287 3,483,287
Note: This table reports estimates of Equation 1 on the sample excluding PPP loans from fintech lenders. The independent
variable in Panel A is an indicator for whether the borrower is a Black-owned business. The independent variable in Panel
B is an indicator for whether the borrower is a White-owned business. In both panels, the dependent variables are indicators
for whether the originating lender is the borrower’s checking account bank (columns 1–2), Top-4 bank (columns 3–4), large
bank (columns 5–6), and small/medium-sized bank (columns 7–8). Control variables all pertain to the borrower firm and their
particular PPP loan. Loan Amount FE are 100 indicator variables for each percentile of the loan size distribution. Zip Code FE
are indicators for each zip code. Approval Week FE are indicators for the week in which the PPP loan was approved by SBA.
Industry FE are 104 indicators for NAICS 3-digit classifications that appear in the data. Business type FE are 7 indicators for
the firm’s business type (see Table 3). Employer status is an indicator for whether the firm has at least one employee. Standard
errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
Internet Appendix 18
Table A.11: Other Race Business Ownership & PPP Lender Type (Bank & Credit Relationship Controls)
Panel A: Fintech PPP Loan
Dependent variable: 1(Fintech)
(1) (2) (3) (4) (5)
1(Asian-Owned) -0.066∗∗∗ -0.070∗∗∗ -0.070∗∗∗ -0.068∗∗∗ -0.050∗∗∗
(0.005) (0.005) (0.005) (0.005) (0.007)
1(Hispanic-Owned) -0.041∗∗∗ -0.037∗∗∗ -0.038∗∗∗ -0.037∗∗∗ -0.021∗∗∗
(0.005) (0.005) (0.005) (0.005) (0.007)
1(White-Owned) -0.057∗∗∗ -0.058∗∗∗ -0.059∗∗∗ -0.057∗∗∗ -0.041∗∗∗
(0.004) (0.004) (0.004) (0.004) (0.006)
1(Credit from Fintech) 0.075∗∗∗ 0.078∗∗∗ 0.109∗∗∗
(0.003) (0.003) (0.005)
1(Credit from Conv.) -0.012∗∗∗ -0.011∗∗∗ -0.015∗∗∗
(0.003) (0.003) (0.004)
Loan Amount FE Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes
Months Since Statement FE No Yes Yes Yes Yes
Checking Acct Bank FE No Yes Yes Yes Yes
Monthly Cash Inflow FE No No No Yes Yes
Monthly Net Cash Inflow FE No No No Yes Yes
Bank Statement Sample Latest Latest Latest Latest Latest within
6 months
Dep Var Mean 0.363 0.363 0.363 0.363 0.424
Observations 168,360 168,360 168,360 168,360 91,870
Internet Appendix 19
Panel B: Bank PPP Loan
Dependent Variable: 1(PPP Lender is) 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
Checking Acct Bank
(1) (2) (3) (4) (5) (6) (7) (8)
1(Asian-Owned) -0.007 -0.010∗∗ 0.003 0.005 0.014∗∗∗ 0.018∗∗∗ 0.014∗∗∗ 0.013∗∗∗
(0.005) (0.005) (0.005) (0.004) (0.003) (0.003) (0.005) (0.005)
1(Hispanic-Owned) 0.028∗∗∗ 0.021∗∗∗ 0.008∗ 0.003 0.011∗∗∗ 0.015∗∗∗ 0.025∗∗∗ 0.019∗∗∗
(0.005) (0.005) (0.004) (0.004) (0.003) (0.003) (0.004) (0.004)
1(White-Owned) 0.020∗∗∗ 0.012∗∗∗ -0.010∗∗∗ -0.002 0.019∗∗∗ 0.020∗∗∗ 0.060∗∗∗ 0.046∗∗∗
(0.004) (0.004) (0.003) (0.003) (0.003) (0.003) (0.004) (0.004)
1(Credit from Fintech) -0.040∗∗∗ -0.025∗∗∗ -0.011∗∗∗ -0.035∗∗∗
(0.004) (0.003) (0.002) (0.003)
1(Credit from Conv.) 0.011∗∗∗ 0.001 0.005∗∗ -0.002
(0.003) (0.002) (0.002) (0.003)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes Yes
Months Since Statement FE No Yes No Yes No Yes No Yes
Checking Acct Bank FE No Yes No Yes No Yes No Yes
Monthly Cash Inflow FE No Yes No Yes No Yes No Yes
Monthly Net Cash Inflow FE No Yes No Yes No Yes No Yes
Bank Statement Sample Latest Latest Latest Latest Latest Latest Latest Latest
Dep Var Mean 0.274 0.274 0.177 0.177 0.090 0.090 0.295 0.295
Observations 168,360 168,360 168,360 168,360 168,360 168,360 168,360 168,360
Note: This table reports estimates of a modified Equation 1, focusing on the role of bank and credit relationships and using
indicators for the three other races/ethnicities. The sample is restricted to those matched to bank statement data. In all columns
except for Panel A Column 5, we include only information from a firm’s latest statement prior to the loan approval. Panel A
Column 5 includes only the latest statement if it is within six months of loan approval. The dependent variable in Panel A is an
indicator for whether a PPP loan is originated by a fintech lender. The dependent variables in Panel B are indicators for whether
the originating lender is the borrower’s checking account bank (columns 1–2), Top-4 bank (columns 3–4), large bank (columns
5–6), and small/medium-sized bank (columns 7–8). We report coefficients on indicators for whether the borrower has previous
credit relationships with fintech and non-fintech lenders. Controls are as described in Table 6. Standard errors are clustered by
borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
Internet Appendix 20
Table A.12: Self-Reported Business Owner Race and PPP Lender Type
Panel A: Fintech PPP Loan
Dependent Variable: 1(Fintech)
(1) (2) (3) (4) (5) (6) (7) (8)
1(SelfID Black-Owned) 0.507∗∗∗ 0.414∗∗∗ 0.314∗∗∗ 0.344∗∗∗ 0.222∗∗∗ 0.201∗∗∗ 0.192∗∗∗ 0.186∗∗∗
(0.006) (0.004) (0.003) (0.003) (0.002) (0.002) (0.003) (0.002)
Loan Amount FE No Yes Yes Yes Yes Yes Yes Yes
Zip Code FE No No Yes No Yes Yes No Yes
Census Tract FE No No No Yes No No No No
Approval Week FE No No No No Yes Yes Yes Yes
Industry FE No No No No No Yes No Yes
Zip-by-Industry FE No No No No No No Yes No
Business Type FE No No No No No No No Yes
Employer Status FE No No No No No No No Yes
Dep Var Mean 0.173 0.173 0.173 0.173 0.173 0.173 0.173 0.173
R2 0.181 0.239 0.335 0.294 0.417 0.437 0.556 0.459
Observations 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682
Panel B: Bank PPP Loan
Dependent Variable: 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
(1) (2) (3) (4) (5) (6)
1(SelfID Black-Owned) -0.041∗∗∗ -0.013∗∗∗ -0.043∗∗∗ -0.028∗∗∗ -0.380∗∗∗ -0.128∗∗∗
(0.002) (0.001) (0.001) (0.001) (0.004) (0.002)
Loan Amount FE No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes
Business Type FE No Yes No Yes No Yes
Employer Status FE No Yes No Yes No Yes
Dep Var Mean 0.136 0.136 0.090 0.090 0.485 0.485
R2 0.001 0.341 0.002 0.197 0.058 0.436
Observations 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682
Internet Appendix 21
Panel C: Other Races
Dependent Variable: 1(Fintech) 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
(1) (2) (3) (4) (5) (6) (7) (8)
1(SelfID Asian-Owned) -0.409∗∗∗ -0.160∗∗∗ 0.165∗∗∗ 0.039∗∗∗ 0.037∗∗∗ 0.015∗∗∗ 0.123∗∗∗ 0.041∗∗∗
(0.006) (0.002) (0.003) (0.002) (0.002) (0.002) (0.006) (0.003)
1(SelfID Hispanic-Owned) -0.349∗∗∗ -0.124∗∗∗ 0.087∗∗∗ 0.009∗∗∗ 0.023∗∗∗ 0.031∗∗∗ 0.112∗∗∗ 0.069∗∗∗
(0.007) (0.002) (0.003) (0.002) (0.002) (0.001) (0.005) (0.002)
1(SelfID White-Owned) -0.523∗∗∗ -0.189∗∗∗ 0.015∗∗∗ 0.009∗∗∗ 0.049∗∗∗ 0.028∗∗∗ 0.449∗∗∗ 0.149∗∗∗
(0.006) (0.002) (0.002) (0.001) (0.002) (0.001) (0.004) (0.002)
Loan Amount FE No Yes No Yes No Yes No Yes
Zip Code FE No Yes No Yes No Yes No Yes
Approval Week FE No Yes No Yes No Yes No Yes
Industry FE No Yes No Yes No Yes No Yes
Internet Appendix 22
Business Type FE No Yes No Yes No Yes No Yes
Employer Status FE No Yes No Yes No Yes No Yes
Dep Var Mean 0.173 0.173 0.136 0.136 0.090 0.090 0.485 0.485
R2 0.199 0.460 0.025 0.342 0.003 0.197 0.131 0.439
Observations 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682 1,098,682
Note: This table reports estimates of Equation 1. The independent variables are the self-reported race/ethnicity of the borrower, and the sample is restricted to the subset of
loans for which race/ethnicity is self-reported. Note that in the main analysis, we use only predicted race (self-reported race is used to train the random forest algorithm, but
does not replace the prediction for self-reported observations). The dependent variable in Panel A is an indicator for whether the originating lender is fintech. Panel B repeats
the specifications in columns 1 and 8 for indicators for whether the originating lender is the borrower’s checking account bank (columns 1–2), a Top-4 bank (columns 3–4), large
bank (columns 5–6), and small/medium-sized bank (columns 7–8). Panel C repeats Panel B, but adds two columns for fintech loans and considers the other three race/ethnicities.
Here, Black-owned businesses represent the single omitted group, so the coefficients should be interpreted relative to them. Control variables all pertain to the borrower firm
and their particular PPP loan. Loan Amount FE are 100 indicator variables for each percentile of the loan size distribution. Zip Code and Census Tract FE are indicators for
each zip code and census tract. Approval Week FE are indicators for the week in which the PPP loan was approved by SBA. Industry FE are 104 indicators for NAICS 3-digit
classifications that appear in the data. Business type FE are 7 indicators for the firm’s business type (see Table 3). Employer status is an indicator for whether the firm has at least
one employee. Standard errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
Table A.13: Business Owner Race and PPP Lender Type with Bank and Credit Relationship Controls
(Excl. Fintech Lenders from Sample)
Panel B: Bank PPP Loans to Black-Owned Businesses
Dependent Variable: 1(PPP Lender is) 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
Checking Acct Bank
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black-Owned) -0.010 0.000 0.008 0.005 -0.010∗∗ -0.015∗∗∗ -0.037∗∗∗ -0.024∗∗∗
(0.007) (0.007) (0.006) (0.006) (0.005) (0.005) (0.006) (0.006)
1(Credit from Fintech) -0.001 0.007∗ 0.000 -0.007∗
(0.005) (0.004) (0.003) (0.004)
1(Credit from Conv.) 0.008∗ -0.000 0.006∗ -0.012∗∗∗
(0.005) (0.003) (0.003) (0.004)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes Yes
Months Since Statement FE No Yes No Yes No Yes No Yes
Checking Acct Bank FE No Yes No Yes No Yes No Yes
Monthly Cash Inflow FE No Yes No Yes No Yes No Yes
Monthly Net Cash Inflow FE No Yes No Yes No Yes No Yes
Bank Statement Sample Latest Latest Latest Latest Latest Latest Latest Latest
Dep Var Mean 0.430 0.430 0.278 0.278 0.141 0.141 0.464 0.464
Observations 168,360 168,360 168,360 168,360 168,360 168,360 168,360 168,360
Panel B: Bank PPP Loans to White-Owned Businesses
Dependent Variable: 1(PPP Lender is) 1(Top 4 Bank) 1(Large Bank) 1(Small/Med Bank)
Checking Acct Bank
(1) (2) (3) (4) (5) (6) (7) (8)
1(Asian-Owned) -0.032∗∗∗ -0.037∗∗∗ -0.002 -0.005 0.005 0.012∗∗ -0.019∗∗ -0.020∗∗∗
(0.008) (0.008) (0.007) (0.007) (0.006) (0.006) (0.008) (0.008)
1(Hispanic-Owned) 0.037∗∗∗ 0.027∗∗∗ 0.016∗∗ 0.008 0.006 0.011∗ 0.013 0.010
(0.008) (0.008) (0.007) (0.007) (0.006) (0.006) (0.008) (0.007)
1(White-Owned) 0.013∗ 0.001 -0.013∗∗ -0.007 0.012∗∗ 0.016∗∗∗ 0.052∗∗∗ 0.035∗∗∗
(0.007) (0.007) (0.006) (0.006) (0.005) (0.005) (0.007) (0.006)
1(Credit from Fintech) -0.001 0.007∗ 0.000 -0.008∗
(0.005) (0.004) (0.003) (0.004)
1(Credit from Conv.) 0.007 -0.001 0.006∗ -0.012∗∗∗
(0.005) (0.003) (0.003) (0.004)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes Yes
Months Since Statement FE No Yes No Yes No Yes No Yes
Checking Acct Bank FE No Yes No Yes No Yes No Yes
Monthly Cash Inflow FE No Yes No Yes No Yes No Yes
Monthly Net Cash Inflow FE No Yes No Yes No Yes No Yes
Bank Statement Sample Latest Latest Latest Latest Latest Latest Latest Latest
Dep Var Mean 0.430 0.430 0.278 0.278 0.141 0.141 0.464 0.464
Observations 168,360 168,360 168,360 168,360 168,360 168,360 168,360 168,360
Note: This table reports estimates of Equation 1 on the Ocrolus bank statement matched sample excluding PPP loans from
fintech lenders. The independent variable in Panel A is an indicator for whether the borrower is a Black-owned business.
The independent variable in Panel B is an indicator for whether the borrower is a White-owned business. In both panels, the
dependent variables are indicators for whether the originating lender is the borrower’s checking account bank (columns 1–2),
Top-4 bank (columns 3–4), large bank (columns 5–6), and small/medium-sized bank (columns 7–8). Control variables all
pertain to the borrower firm and their particular PPP loan. Controls are as described in Table 6. Standard errors are clustered
by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels, respectively.
Internet Appendix 23
Table A.14: How Race Predicts Fintech with Card Revenue Controls within Approval Month
Banks:
Dependent variable: 1(Fintech) 1(Top 4) 1(Large) 1(Small/Med)
(1) (2) (3) (4) (5) (6) (7) (8)
1(Black-Owned) 0.024∗∗∗ 0.015∗∗∗ 0.013∗∗∗ 0.014∗∗∗ -0.011∗∗∗ -0.011∗∗∗ -0.012∗∗∗ -0.011∗∗∗
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.004) (0.004)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes Yes
Card Revenue At Approval FE No Yes No Yes No Yes No Yes
Dep Var Mean 0.097 0.097 0.179 0.179 0.122 0.122 0.515 0.515
Observations 582,099 582,099 582,099 582,099 582,099 582,099 582,099 582,099
Note: This table reports estimates of a modified Equation 1, focusing on the role of firm revenue during the PPP loan approval
month. The sample is restricted to those matched to Enigma data on credit and debit card transactions. Controls are as described
in Table 8. Standard errors are clustered by borrower zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10%
levels, respectively.
Internet Appendix 24
Table A.15: Black Business Ownership and Fintech PPP Loans Mediated by Racial Animus (Bank Statement-Matched Sample)
Panel A: Fintech PPP Loans as Dependent Variable
Dependent variable: 1(Fintech)
(1) (2) (3) (4) (5) (6) (7)
1(Black-Owned) 0.059∗∗∗ 0.060∗∗∗ 0.058∗∗∗ 0.057∗∗∗ 0.059∗∗∗ 0.058∗∗∗ 0.052∗∗∗
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
1(Black-Owned) × Racial Animus 0.013∗∗∗ 0.011∗ 0.022∗∗∗ 0.023∗∗∗ 0.004 0.016∗∗∗
(0.005) (0.005) (0.005) (0.005) (0.004) (0.004)
Racial Animus Measure Stephens- Segregation Segregation
Nationscape IAT (Implicit) IAT (Explicit)
Davidowitz (Dissimilarity) (Isolation)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes
Internet Appendix 25
Industry FE Yes Yes Yes Yes Yes Yes Yes
Business Type FE Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes
Checking Acct Bank FE Yes Yes Yes Yes Yes Yes Yes
Months Since Statement FE Yes Yes Yes Yes Yes Yes Yes
Monthly Cash Inflow FE Yes Yes Yes Yes Yes Yes Yes
Monthly Net Cash Inflow FE Yes Yes Yes Yes Yes Yes Yes
Credit Controls No Yes Yes Yes Yes Yes Yes
Dep Var Mean 0.361 0.361 0.361 0.361 0.361 0.361 0.361
Observations 169,818 169,818 169,818 169,818 169,818 169,818 169,818
Panel B: Top-4 Bank PPP Loans as Dependent Variable
Dependent variable: 1(Top 4)
(1) (2) (3) (4) (5) (6) (7)
1(Black-Owned) 0.001 0.000 0.001 0.000 0.001 0.001 -0.002
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
1(Black-Owned) × Racial Animus 0.012∗∗∗ 0.003 0.009∗∗ 0.007∗ 0.002 0.008∗∗
(0.003) (0.004) (0.004) (0.004) (0.003) (0.003)
Racial Animus Measure Stephens- IAT (Implicit) IAT (Explicit) Segregation Segregation
Nationscape
Davidowitz (Dissimilarity) (Isolation)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes
Internet Appendix 26
Business Type FE Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes
Checking Acct Bank FE Yes Yes Yes Yes Yes Yes Yes
Months Since Statement FE Yes Yes Yes Yes Yes Yes Yes
Monthly Cash Inflow FE Yes Yes Yes Yes Yes Yes Yes
Monthly Net Cash Inflow FE Yes Yes Yes Yes Yes Yes Yes
Credit Controls No Yes Yes Yes Yes Yes Yes
Dep Var Mean 0.176 0.176 0.176 0.176 0.176 0.176 0.176
Observations 169,818 169,818 169,818 169,818 169,818 169,818 169,818
Panel C: Non-Top 4 Bank PPP Loans as Dependent Variable
Dependent variable: 1(Non-Top 4 Bank)
(1) (2) (3) (4) (5) (6) (7)
1(Black-Owned) -0.061∗∗∗ -0.062∗∗∗ -0.061∗∗∗ -0.060∗∗∗ -0.061∗∗∗ -0.060∗∗∗ -0.050∗∗∗
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
1(Black-Owned) × Racial Animus -0.016∗∗∗ -0.014∗∗∗ -0.026∗∗∗ -0.027∗∗∗ -0.014∗∗∗ -0.029∗∗∗
(0.005) (0.005) (0.005) (0.005) (0.004) (0.004)
Racial Animus Measure Stephens- IAT (Implicit) IAT (Explicit) Segregation Segregation
Nationscape
Davidowitz (Dissimilarity) (Isolation)
Loan Amount FE Yes Yes Yes Yes Yes Yes Yes
Zip Code FE Yes Yes Yes Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes
Internet Appendix 27
Business Type FE Yes Yes Yes Yes Yes Yes Yes
Employer Status FE Yes Yes Yes Yes Yes Yes Yes
Checking Acct Bank FE Yes Yes Yes Yes Yes Yes Yes
Months Since Statement FE Yes Yes Yes Yes Yes Yes Yes
Monthly Cash Inflow FE Yes Yes Yes Yes Yes Yes Yes
Monthly Net Cash Inflow FE Yes Yes Yes Yes Yes Yes Yes
Credit Controls No Yes Yes Yes Yes Yes Yes
Dep Var Mean 0.389 0.389 0.389 0.389 0.389 0.389 0.389
Observations 169,818 169,818 169,818 169,818 169,818 169,818 169,818
Note: This table repeats Table 10 but within the bank statement-matched (Ocrolus) sample. It reports estimates of a modified Equation 1, focusing on the interaction between
the indicator for Black-owned business and a standardized measure of racial animus in the borrower location. The dependent variable differs across the three panels: In Panel
A, it is an indicator for a fintech PPP loan, in Panel B, it is an indicator for a Top-4 bank PPP loan, and in Panel C, it is an indicator for a non-Top 4 bank PPP loan. In each
panel, column 1 includes the same controls as the specification in Table 4 Panel A column 8. The racial animus measures are as follows: column 2 uses the number of racially
charged searches in a designated media market (DMA); column 3 uses responses to the question on favorability toward Black people in the Nationscape survey aggregated to
the congressional district level; columns 4–5 use the implicit and explicit score from the Implicit Association Test (IAT) aggregated to the county level; columns 6–7 use the
dissimilarity and isolation index at the metropolitan statistical area (MSA) level. All racial animus measures are standardized at their respective levels of geography, weighted by
the number of PPP loans. Controls are as described in Tables 4 and 6. Standard errors are clustered by zipcode. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10%
levels, respectively.
Table A.16: Summary Statistics on Automation during PPP among Small Banks
Automating Banks Never-Automated Banks p-value
N = 74,965 N = 1,949,709
Unique Banks 20 3,870
Share Loans After Automation 0.531
Assets (Million $) 28,408 11,129 0.000
Asian-Owned Share Before Automation 0.064 0.055 0.000
Asian-Owned Share After Automation 0.070 0.055 0.000
Black-Owned Share Before Automation 0.041 0.035 0.000
Black-Owned Share After Automation 0.112 0.035 0.000
Hispanic-Owned Share Before Automation 0.058 0.042 0.000
Hispanic-Owned Share After Automation 0.073 0.042 0.000
White-Owned Share Before Automation 0.837 0.868 0.000
White-Owned Share After Automation 0.746 0.868 0.000
Note: This table contains summary statistics of the sample used in the automation analysis. It compares small banks that
automated (left column) to those that did not (right column), and reports the p-value of a t-test comparing the difference of the
two means.
Internet Appendix 28
Table A.17: Automation during PPP and Lending to Black-Owned Small Businesses
Dependent Variable: 1(Black-Owned)
(1) (2) (3) (4)
1(After Automation) 0.060∗∗∗ 0.043 ∗∗∗
0.060∗∗∗ 0.043∗∗∗
(0.011) (0.006) (0.013) (0.007)
Level of Clustering Bank Bank Bank-Week Bank-Week
Bank FE Yes Yes Yes Yes
Approval Week FE Yes Yes Yes Yes
Loan Amount FE No Yes Yes No
Zip Code FE No Yes Yes No
Industry FE No Yes Yes No
Business Type FE No Yes Yes No
Employer Status FE No Yes Yes No
Dep Var Mean 0.037 0.037 0.037 0.037
Observations 2,024,674 2,024,674 2,024,674 2,024,674
Note: This table reports estimates of Equation 2, but adjusts the clustering of standard errors. It shows the effect of automation
on the chances a loan is to a Black-owned businesses. Standard errors are clustered by bank in columns 1-2, and double
clustered by bank and week approved in columns 3-4. ∗∗∗ ,∗∗ ,∗ indicate statistical significance at the 1%, 5%, and 10% levels,
respectively.
Internet Appendix 29
A Supplemental Details on Racial Animus Measures
This appendix describes the construction of our six measures of racial animus.
A.1 Stephens-Davidowitz (2013) Measure
This measure uses Google search data to proxy for social attitudes. Stephens-Davidowitz (2013) defines
racial animus using the search rate of a racially charged word in a designated media market (DMA) from
2004 to 2007. He defines the racially charged search rate in a DMA j as
h i
Google searches including the word “Word(s)”
Total Google searches j
Racially Charged Search Ratej = h i . (4)
Google Searches including the word “Word(s)”
Total Google searches max
Figure A.9 Panel A shows the geographical distribution of racially charged searches by DMA, and Figure
A.10 Panel A shows the distribution of racially charged searches by DMA.
A.2 Nationscape Survey
Nationscape is a large public opinion survey conducted in the lead-up to the 2020 elections (see
Tausanovitch and Vavreck, 2020). To capture racial bias, we follow Bursztyn et al. (2021) in using
responses to the question: ”Here are the names of some groups that are in the news from time to time.
How favorable is your impression of each group or haven’t you heard enough to say? — Blacks.” The
scale is from 1 to 4, with 1 being very favorable and 4 being very unfavorable. We keep only the White
respondents. Figure A.9 Panel B maps the geographical distribution of the average response by
congressional district, the finest geographic level available. We treat the response ”Haven’t heard
enough” as missing. Figure A.10, Panel B shows the distribution of individual responses to the question
by White respondents.
A.3 Project Implicit IAT Score
Project Implicit runs the Implicit Association Tests (IATs). In particular, we use data from the Race
IAT, which measures explicit and implicit bias against different races (Xu et al., 2014). We keep only
responses from White, non-Hispanic respondents.
Implicit bias is measured by asking respondents to first use two buttons (“E” or “I”) on their keyboard
to identify a series of faces that flash on the screen as Black or White and then a series of words that flash
on the screen as good or bad. In the following rounds, both faces and words will flash on the screen, but
the respondents will still be limited to “E” or “I” — only “E” could now mean “Black or good” while
“I” will mean “White or bad” in one round, and later be reversed so “E” means “Black or bad” and “I”
means “White or good” in the next round. The idea behind the IAT is that a slower reaction to selecting
“good” when “Black” is linked to it or “bad” when “White” is linked to it implies an implicit bias against
Black people or bias in favor of White people (see Lopez, 2017, for more details). The IAT score is
then calculated as the average difference in average speed per participant between each corresponding
“actual” and practice block scaled by the pooled standard deviation.
Figure A.9 Panel C shows the geographical distribution of average IAT scores by county. We omit
counties with fewer than 50 respondents. Figure A.10 Panel C shows the distribution of the IAT score at
the respondent level.
Internet Appendix 30
A.4 Project Implicit IAT Explicit Attitude
Following the IAT, respondents are asked to take a follow-up survey. We follow Bursztyn et al. (2021)
in using the first question in the survey to proxy for explicit attitudes: “Please rate how warm or cold
you feel toward the following groups (0 = coldest feelings, 5 = neutral, 10 = warmest feelings): African
Americans.” The responses are on a scale from 0 to 10.
Figures A.10 Panel D and A.9 Panel D show the response distribution and geographical distribution
of the explicit score, respectively. We omit counties with fewer than 50 respondents.
A.5 Dissimilarity Index
As an alternative to surveys to measuring racial bias, we also estimated residential segregation at the
metropolitan statistical area (MSA) level to proxy for racial bias. Among many other things, residential
segregation could represent a certain revealed preference of the residents.
The residential segregation literature proposes various different methods of measuring segregation.
Here, we consider the dissimilarity index, which measures how similar the distribution of Black residents
in a tract is relative to the distribution of White residents in the same tract (Massey and Denton, 1988):
N
1 X wi bi
DT = − (5)
2 WT BT
i=1
where bi and wi correspond to the share of Black and White people in tract i, respectively, and bT and wT
correspond to the same shares in MSA T . The higher the index, the more segregated an MSA is. Figure
A.10 Panel E shows the distribution of the dissimilarity index. Figure A.9 Panel E maps MSA-level
residential segregation across the U.S.
A.6 Isolation Index
We estimate an additional measure of residential segregation: the isolation index. Isolation measures
the extent to which Black people only interact with other Black people, instead of other White people
(Massey and Denton, 1988):
N
X bi bi
IT = × (6)
BT bi + wi
i=1
with variables defined as above. As Black residents are more isolated, the isolation index approaches one.
Note that the isolation index measures a different dimension of segregation compared to the dissimilarity
index. The dissimilarity index does not consider the relative size of the groups being compared. For
example, a particular MSA might be “even” according to the dissimilarity index, but if the population of
Black residents is much smaller than that of White residents, then the MSA will be high on the isolation
index. Figure A.10 Panel F shows the distribution of the isolation index. Figure A.9 Panel F maps the
geography of isolation.
Internet Appendix 31
Figure A.9: Geographical Distribution of Racial Animus
(A) Racially Charged Searches (B) Nationscape Survey
(C) IAT Implicit Scores (D) IAT Explicit Scores
(E) Dissimilarity Index (F) Isolation Index
Note: This figure shows the geographical distribution of each proxy of racial bias used in the analysis. Racially charged
searches (Panel A) are at the designated media market (DMA) level, plotted at the county level. IAT scores (Panels B and C)
are aggregated at the county level; counties with less than 50 respondents are coded as omitted. Nationscape surveys (Panel D)
are aggregated at the congressional district level. The dissimilarity and isolation index (Panels E and F) are calculated at the
metro/micropolitan statistical area (MSA) level.
Internet Appendix 32
Figure A.10: Distribution of Racial Animus
(A) Racially Charged Searches
(B) Nationscape Survey
(C) IAT Implicit Scores (D) IAT Explicit Scores
(E) Dissimilarity Index (F) Isolation Index
Note: This figure shows the distribution of each proxy of racial bias used in the analysis. Racially charged searches (Panel
A) are at the designated media market (DMA) level. IAT scores (Panels B and C) are at the respondent level. Nationscape
surveys (Panel D) are at the respondent level. The dissimilarity and isolation index (Panels E and F) are calculated at the
metro/micropolitan statistical area (MSA) level.
Internet Appendix 33
File and source
- File
- racial-disparities-in-access-to-small-business-credit-evidence-from-the-paycheck-protection-program.pdf
- Size
- 8,701,010 bytes
- SHA-256
- 196fb0bb0654f71ccc077c38944c565eed42a0d205713a8f77c44e0fe7af6752
- Original
- doi.org