Did the PPP and EIDL Program Get Disbursed to Minority Communities — NBER Working Paper 28321 (January 2021)

Summary

NBER Working Paper 28321 by Robert W. Fairlie and Frank Fossen, dated January 2021, asking whether the $660 billion Paycheck Protection Program and $220 billion Economic Injury Disaster Loan program were disbursed to minority communities in the early stages of COVID-19. The abstract states that, from data on the universe of loans and administrative data on employer firms, the authors generally find a slightly positive relationship between PPP loan receipt per business and the minority share of the population or businesses. It states that funds flowed to minority communities later than to communities with lower minority shares, and that PPP loan amounts are negatively related to the minority share.

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: Did The Ppp And Eidl Program Get Disbursed To Minority Communities Ear
Case: Did The Ppp And Eidl Program Get Disbursed To Minority Communities Ear

Cite this document

“Did the PPP and EIDL Program Get Disbursed to Minority Communities — NBER Working Paper 28321 (January 2021).” Original at www.nber.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/did-the-ppp-and-eidl-program-get-disbursed-to-minority-communities-ear/, SHA-256 76f10656e0e11c4a7a0f2302a5985435611f10ef2b66a591dcd88f5880fc11b2.

Full text

NBER WORKING PAPER SERIES

DID THE $660 BILLION PAYCHECK PROTECTION PROGRAM AND $220 BILLION
ECONOMIC INJURY DISASTER LOAN PROGRAM GET DISBURSED TO MINORITY
COMMUNITIES IN THE EARLY STAGES OF COVID-19?

Robert W. Fairlie
Frank Fossen

Working Paper 28321
http://www.nber.org/papers/w28321

NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
January 2021

We would like to thank Karen Mills and participants at the NBER Entrepreneurship Workshop,
the PPIC California labor market workshop, and the Kauffman Foundation Entrepreneurship
Issue Forum for comments and suggestions. The research project has also benefited from
numerous conversations with the press and policymakers. Special thanks go to Zach Mider, Jason
Grotto and Cedric Sam for discussions of the data and analysis, research assistance by Ian Allen
and Ege Can, and thanks go to participants in the Axios/Google Small Business Matters Expert
Panel and JPMorgan Chase Institute Data Dialogue meetings for their comments and questions.
The views expressed herein are those of the authors and do not necessarily reflect the views of the
National Bureau of Economic Research.

NBER working papers are circulated for discussion and comment purposes. They have not been
peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies
official NBER publications.

© 2021 by Robert W. Fairlie and Frank Fossen. All rights reserved. Short sections of text, not to
exceed two paragraphs, may be quoted without explicit permission provided that full credit,
including © notice, is given to the source.
Did the $660 Billion Paycheck Protection Program and $220 Billion Economic Injury Disaster
Loan Program Get Disbursed to Minority Communities in the Early Stages of COVID-19?
Robert W. Fairlie and Frank Fossen
NBER Working Paper No. 28321
January 2021
JEL No. I28,J15,L26

ABSTRACT

Social distancing restrictions and health- and economic-driven demand shifts from COVID-19
shut down many small businesses with especially negative impacts on minority owners. Is there
evidence that the unprecedented federal government response to help small businesses – the $659
billion Paycheck Protection Program (PPP) and the related $220 billion COVID-19 Economic
Injury Disaster Loans (EIDL) – which had a stated goal of helping disadvantaged groups, was
disbursed evenly to minority communities? In this descriptive research note, we provide the first
detailed analysis of how the PPP and EIDL funds were disbursed across minority communities in
the country. From our analysis of data on the universe of loans from these programs and
administrative data on employer firms, we generally find a slightly positive relationship between
PPP loan receipt per business and the minority share of the population or businesses, although
funds flowed to minority communities later than to communities with lower minority shares. PPP
loan amounts, however, are negatively related to the minority share of the population. The EIDL
program, in contrast, both in numbers and amounts, was distributed positively to minority
communities.

Robert W. Fairlie
Department of Economics
Engineering 2 Building
University of California at Santa Cruz
Santa Cruz, CA 95064
and NBER
rfairlie@ucsc.edu

Frank Fossen
Department of Economics
University of Nevada, Reno
1664 N. Virginia Street
Reno, NV 89557-0030
ffossen@unr.edu
1. Introduction
The widespread closing of stores and businesses in the United States and around the world due to
the coronavirus is unprecedented. Stores, factories and many other businesses have closed by
policy mandate, downward demand shifts, health concerns, or other factors. The number of
working business owners in the United States plummeted from 15.0 million in February 2020 to
11.7 million in April 2020 and has only partially rebounded since then (Fairlie 2020).1 The impacts
have also been disproportionately felt by race: business owner activity fell in the early-stages of
the pandemic by 41 percent among African-Americans and 32 percent among Latinx compared
with 17 percent among whites.2
Given the impact of the pandemic the federal government provided a response of larger
magnitude than ever seen before in terms of providing financial assistance to small businesses.
The largest program providing funds to small businesses is the $650 billion Paycheck Protection
Program (PPP).3 The Small Business Administration (SBA) administered program provides loans
to small businesses through banks, credit unions, and other financial institutions with the stated
goal of keeping small businesses open and retaining employees on the payroll. Loan amounts were
generally equal to 2.5 months of average payroll costs, and can be forgiven if the business retains
its employees. The program started providing loans on April 3, 2020, which was after most states
imposed social distancing restrictions in response to COVID-19, and ran through August 8, 2020
providing more than 5 million total loans.4 The $220 billion Economic Injury Disaster Loan
Program (EIDL) program, which is also administered by the SBA, is designed to provide either
loans or advances to small businesses that are losing revenues and sales due to COVID-19. Nearly
3.6 EIDL loans for $200 billion and nearly 5.8 million EIDL advances for $20 billion have been
provided to small businesses. EIDL loans can be used to cover up to 6 months of a wide array of

1
These findings prompted a large policy response, for example, new U.S. Senate bills (U.S. Senate 2020) and
California State bills (Newsom 2020) to provide assistance to minority and small businesses during the pandemic.
2
Cash balances and revenues of small businesses also fell substantially in the early stages of the pandemic (Farrell,
Wheat and Mac 2020).
3
The Coronavirus Aid, Relief, and Economic Security (CARES) Act also provided stimulus payments to
households and expanded unemployment insurance benefits to households. See Bhutta et al. (2020) for an analysis
of whether this cash assistance will help families cover expenses over a six month period of lost income (i.e.
April through September 2020), relative to a counterfactual where families would have had to rely solely on their
own liquid savings and standard UI benefits (e.g., benefits available in the absence of CARES).
4
On March 11, 2020, the World Health Organization (WHO) declared COVID-19 a pandemic. On March 16, the
San Francisco Bay Area imposed the first shelter-in-place restrictions in the country followed by the State of
California on March 19. New York State followed the next day. By early April most states imposed social
distancing restrictions.

1
working capital and normal operating expenses, such as continuation of health care benefits, rent,
utilities, and fixed debt payments. EIDL advances are grants and do not have to be repaid, but are
for smaller amounts ($1,000 per employee up to a maximum of $10,000). EIDL advances are
subtracted from the forgiveness amount of their PPP loan if they are received in addition to PPP
loans.
One of the stated goals in the CARES Act which included the PPP and EIDL programs
was to prioritize serving “underserved markets” and businesses owned by “socially and
economically disadvantaged individuals” (U.S. Congress 2020). Did the PPP and EIDL programs,
which provided 15 million loans or advances worth more than $850 billion to small businesses,
get disbursed to minority communities benefitting the businesses and employees in those
communities? Given the larger negative effects of COVID-19 on business inactivity among
minority businesses (Fairlie 2020) targeting these relief funds to minority communities might be
especially important.
In this descriptive research note we provide the first detailed analysis of how the PPP and
EIDL funds were disbursed across minority communities in the United States. Using
administrative data on the universe of PPP loans, EIDL loans, and EIDL advances, we explore
whether loans and advances were evenly distributed or not. We find that minority communities
received a large share of PPP loans. We generally find a slightly positive relationship between PPP
loan receipt per business and the minority share of the population. There is some evidence,
however, that the first round of funds was disproportionately disbursed to non-minority
communities and the second round of funds was disproportionately disbursed to minority
communities. When we focus on the minority share of employer businesses in an area we find
similar results. Focusing on PPP loan amount per employee, however, we find a negative
relationship with minority share of the population. In contrast, EIDL loans and advances, in both
number and amounts, were provided positively to minority communities. We find a strong positive
relationship in the receipt of these loans and advances by minority share of the population.
These results build on the findings from a few related working papers on the PPP program.
Granja et al. (2020) find that the first round of PPP funds flowed to areas more adversely affected
by the economic effects of the pandemic, but that the early PPP did not have a substantial effect
on local economic outcomes. Neilson et al. (2020) report based on survey data that the smallest
businesses were less aware of the PPP, less likely to apply, and conditional on applications filed

2
later, faced longer processing times, and were denied more often. Bartik et al. (2020) using firm-
level data and an instrumental variables approach find that PPP loans led to an increase in a
business’ expected survival, and a positive but imprecise effect on employment.5 Focusing on race,
Lederer et al. (2020) conducted matched-pair audit testing of financial institutions in Washington,
D.C. for PPP loans and find disparities between black and white testers in encouragement in
applying for a loan, products offered, and information provided by the bank representative.
Additionally, Erel and Liebersohn (2020) find that FinTech is disproportionately used to disburse
PPP funds in high minority share ZIP codes. This paper builds on the previous research by
providing the first comprehensive analysis of the relative disbursement of PPP and EIDL small
business funds to minority communities, and the first study, to our knowledge, of the EIDL
program. The findings are potentially important for future targeting and oversight of government
aid to preserve minority businesses and the jobs they create.6

2. Data
Partly in response to a Freedom of Information Act (FOIA) request and law suit threat by the
media, the SBA released complete loan-level microdata for the PPP and EIDL programs. The PPP
data cover the universe of loans provided through the program, which was from April 3, 2020 to
August 8, 2020. The PPP is divided into two rounds. The first round ran from April 3 to April 16
and consisted of $342 billion across about 1.6 million loans. The second round ran from April 27
to August 8 and included more than 3.5 million loans with a total value of roughly $180 billion.
In total there are 5.2 million loans and $522 billion.
The loan microdata include information on the amount of the loan for loans under $150,000
For larger loans, only ranges are reported ($150,000-350,000, $350,000-1 million, $1-2 million,
$2-5 million, and $5-10 million). Geographical information down to the zip code is provided in
the smaller loan data, whereas exact address and even the name of the business is included in the
larger loan data. The data also include information on industry, business type, jobs retained self-

5
Alstadsæter et al. (2020) use Norwegian administrative data to simulate the effects of Norwegian government
support and the U.S. PPP program to help businesses during the COVID-19 pandemic and find that these policies
supporting payroll can be partly effective.
6
A large literatures explores the causes and consequences of disparities in ownership and success of minority-
owned businesses. For broader discussions and reviews of these literature, see, for example, Davila and Mora
(2013); Fairlie and Robb (2008); Kerr and Kerr (2020); Parker (2018). See Fairlie (2020) and Stanford Latino
Entrepreneurship Initiative (2020) for evidence of COVID-19 impacts on minority-owned businesses.

3
reported by the business, and name of the lender.7 Information on the race, gender and veteran
status of the owner are incomplete. The application did not ask for demographic information on
the owners (see U.S. SBA 2020 for application form) and relied on banks to report the information.
The result is that only 10 percent of loans provide race information, and these are heavily
concentrated among a few banks.
The SBA also released loan and advance data from the EIDL program. The EIDL program
data are separated into the loans and advances. There were 3.6 million loans and 5.8 million
advances administered through the program. As of 9/14/20 $190 billion of the allocated $200
billion in loans have been handed out to small businesses. All of the $20 billion for EIDL advances
has been provided to small businesses.
To normalize the number of PPP or EIDL loans by zip code we calculate loans per
employer business. We use data from County Business Patterns (CBP) on business establishments
with employees. The data are provided by the U.S. Census Bureau at the zip code level as well as
other geographical levels. The CBP data on employer establishments do not include counts of
farms and nonprofits. We acquire farm data by zip code from the U.S. Department of Agriculture’s
National Agricultural Statistics Service (NASS). From the PPP loan data, we exclude nonprofit
businesses, which represent 3.5 percent of loans, businesses with a nonclassifiable industry (1.7
percent of the loans), self-employed persons (4.5 percent), and independent contractors (3.0
percent).
To normalize loan amounts we calculate average loan amounts per business employee in
each zip code. CBP data also includes employment levels for employer business establishments
down to the zip code level. The normalization adjusts for loan amount differences due to
differences in employment size by location, which is the general basis for loan amounts. Because
only ranges are reported for larger loans in the PPP administrative data we use the midpoint of
each range for each loan (e.g. we use $250,000 for a recorded range value of $150,000-350,000).
Using alternative assumptions such as 1/3 the range instead of the midpoint does not change the
relationship by minority share. EIDL loan and advance amounts are complete.

7
The top banks providing PPP loans were Bank of America (7%), JPMorgan Chase (5%), Cross River Bank (4%),
Kabbage (4%), and Wells Fargo (4%).

4
We compare these measures of loan receipt per employer business establishment and loan
amounts per employee to data from the Census of Population on the minority share of the
population across communities. We measure minority share of the population primarily by zip
code but also by county. In addition to analyzing the relationship between PPP and EIDL loan
receipt per business by the minority share of the population we examine the relationship by
minority share of employer businesses. We use data from the Annual Business Survey (ABS) on
employer businesses at the county level to calculate the minority share of employer businesses in
each location. Data from the ABS are not available at the zip code level.
Table 1 provides mean values (weighed by population and unweighted) for the main
variables of interest. Across zip codes, the average number of PPP loans per employer
establishment is 0.489. The average loan amount per employee (unconditional on receiving a loan)
is $4,404. EIDL loan receipt and amounts are lower. EIDL advances went out to more firms but
the amounts were much smaller than other funds. The minority share of the population across zip
codes has a mean of 0.389 and the minority share of employer businesses has a lower mean of
0.180 reflecting substantially lower business ownership rates among minorities.

Table 1: Descriptive Statistics

Weighted Unweighted
Mean Mean N
PPP loans per employer establishment 0.489 0.347 31,952
PPP average loan amount per employee $4,404 $4,892 30,356
EIDL loans per employer establishment 0.256 0.154 31,952
EIDL average loan amount per employee $1,515 $2,488 30,356
EIDL advances per employer establishment 0.577 0.335 31,952
EIDL average advance amount per employee $192 $262 30,356
Minority share of the population 0.389 0.229 32,670
Minority share of the population (county) 0.389 0.235 3,142
Minority share of employer businesses (county) 0.180 0.116 1,031
Notes: The statistics are at the zip code level if not otherwise indicated. The weighted means are
weighted by population. Areas with unobserved minority shares are excluded. The PPP average
loan amount per employee excludes loans to agriculture due to a lack of data on farm employees.

5
3. Results
3.1 Regional Patterns in PPP Loans and EIDL
PPP loans were spread across the country and not limited to a few regions. Figure 1, Panel A,
provides a state heat map of PPP loan receipt per employer business establishment. A few states
had levels of above 0.55 loans per employer business and a few states had levels between 0.27 and
0.34 loans per employer business. States on the East Coast tended to have higher rates of loan
receipt per business, and states in the Midwest tended to have lower rates of loan receipt per
business. EIDL loan receipt per business (Panel B) also was generally spread across the country.
The patterns are somewhat stronger regionally, however, with the coasts having higher levels of
loan receipt per business than the middle of the country. EIDL advances (Panel C) show a
somewhat similar pattern across states. The main takeaway from these figures, however, is that
PPP loan, EIDL loan and EIDL advances receipt per business was spread across the country and
not limited to only a few states or regions.

Figure 1: Distribution of loan receipts per employer establishment across states
Panel A: PPP loans

6
Panel B: EIDL loans

Panel C: EIDL advances

3.2 PPP Loan Receipt Patterns by Minority Communities
We turn to analyzing how PPP loan receipt was distributed across minority communities. Figure
2 displays PPP loan receipt per employer establishment by minority share of the population at the

7
zip code level. Panel A shows the relationship weighted by the total population and Panel B shows
the relationship without population weights. The figure also includes plotted quadratic regression
lines to help show the relationship. Before discussing the results, two important points are noted.
First, we do not report confidence intervals (i.e. “whiskers”) because we use the universe of PPP
loans and administrative data on employer firms based on the Census Business Register. Second,
we focus on the raw relationship between PPP loan receipt and minority share of the population
without controlling for other factors because we are trying to capture the influences of these
neighborhood characteristics. For example, if minority communities have higher poverty rates and
that is correlated with receipt of PPP loans then we want to include that in our measurement. Even
if the driver of loan receipt is income it is reflected in race and that is what we are trying to capture.

Figure 2: PPP loans per employer establishment by minority share
Panel A: Weighted by population Panel B: Unweighted

0.60
0.44

0.55
0.39

0.50
0.34
0.45

0.29
0.40

0.24
0.35

0.30 0.19
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean number of PPP loans per employer establishment in zip codes by minority share of
the population. Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic
regressions at the zip code level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard
deviation.

The relationship appears to be mostly flat between loan receipt and minority population
share. Both weighted and unweighted figures show a slight positive relationship between loan
receipt per business and the minority share of the population across zip codes in the United States.
Most of the averages by minority share fall within the range of half a standard deviation from the

8
median, as indicated by the double arrow on the Y-axis. Using the weighted figure, moving from
the 25th percentile minority share of the population (16 percent minority) to the 75th percentile
minority share of the population (59 percent minority) loan receipt only increases from 0.49 to
0.52 PPP loans per employer business establishment.
The PPP disbursed funds in two rounds with adjustments and awareness about the program
changing between the two. The first round was April 3 to April 16, 2020 and consisted of 1.6
million loans. The second round ran from April 27 to August 8 and consisted of 3.6 million loans.
Figure 3 displays the first round relationship, and Figure 4 displays the second round relationship.
Different patterns are revealed when separating by rounds. In the first round, loan receipt went
disproportionately to non-minority communities. The figure shows a stronger negative
relationship, with a decline of 0.05 loans per business between the 25th and 75th percentiles in
minority shares. The second round of funding, however, showed the opposite pattern. In this case,
there is an unequivocal positive relationship between loan receipt and minority population share.
Moving from the first quartile to the third quartile in minority share is associated with an increase
of 0.08 PPP loans per firm.

Figure 3: PPP loans per employer establishment by minority share in the 1st round
Panel A: Weighted by population Panel B: Unweighted

0.18
0.18
0.16

0.16 0.14

0.12
0.14

0.1
0.12
0.08

0.1
0.06

0.08 0.04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean number of PPP loans per employer establishment in zip codes by minority share of
the population in the first round of the PPP program (April 3-April 16, 2020). Panel A uses population weights, Panel
B is unweighted. The dashed lines are from quadratic regressions at the zip code level. For perspective, the double
arrow on the Y-axis indicates the median ± ½ standard deviation.

9
Figure 4: PPP loans per employer establishment by minority share in the 2nd round
Panel A: Weighted by population Panel B: Unweighted

0.52 0.38

0.47
0.33

0.42
0.28
0.37
0.23
0.32
0.18
0.27

0.22 0.13

0.17 0.08
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean number of PPP loans per employer establishment in zip codes by minority share of
the population in the second round of the PPP program (April 27-August 8, 2020). Panel A uses population weights,
Panel B is unweighted. The dashed lines are from quadratic regressions at the zip code level. For perspective, the
double arrow on the Y-axis indicates the median ± ½ standard deviation.

In terms of the different rounds of the PPP, the first $349 billion was exhausted after just
two weeks of being available. Given unmet need by small businesses for assistance, Congress
approved an additional $310 billion. The change in the slope of the relationship between the two
rounds might be caused by a few factors. First, applying for PPP loans early on favored having
long established relationships with banks which minority businesses were less likely to have (Mills
2020). Second, much of the early money flowed through smaller community banks which were
often in rural areas because these banks were nimbler at accessing the aid (Bloomberg 2020). In
the second round larger banks with more urban and racially diverse customer bases caught up.
Third, minority-owned businesses tend to be smaller than non-minority-owned businesses (Census
2016; Fairlie and Robb 2008), and smaller businesses typically took longer to complete required
paperwork because they often did not have in-house accountants, legal help, or other support.
Finally, FinTech and other online lenders were brought in and approved by the SBA, and these
lenders were often active in minority areas (Liu and Parilla 2020). It is unclear how costly the
delay was in receiving loans to minority businesses and communities.

10
We turn to analyzing the relationship between PPP loans per business and the minority
share of businesses in the community. To measure the minority share of businesses we use data
from the Annual Business Survey (ABS) on employer businesses at the county level.8 Data are not
available at the zip code level.9 Figure 5 displays the relationship. The unweighted numbers do not
indicate a clear pattern and are mostly consistent with a flat relationship. The weighted numbers
by population size indicate a slight positive relationship. The relationship is not strong however.
For example, moving from the 25th percentile of counties in the minority share of businesses (9
percent minority share) to the 75th percentile (25 percent) is associated with an increase of 0.02
PPP loans per employer business.

Figure 5: PPP loans per employer establishment by minority share of businesses
Panel A: Weighted by population Panel B: Unweighted

0.63
0.56

0.61
0.54

0.59
0.52

0.57
0.5

0.55
0.48

0.53 0.46

0.51 0.44

Notes: The charts show the mean number of PPP loans per employer establishment in counties by minority share of
businesses. Loans to agricultural businesses are excluded. Panel A uses population weights, Panel B is unweighted.
The dashed lines are from quadratic regressions at the county level. For perspective, the double arrow on the Y-axis
indicates the median ± ½ standard deviation.

8
We exclude PPP loans to agricultural businesses here due to a lack of data on the minority status of farmers.
9
We also examine the relationship between PPP loans per employer business by minority share of the population at
the county level. The results are similar to those at the zip code level.

11
Figures 6 and 7 display the relationship between loan receipt and minority business share
for the first and second rounds, respectively. Similar to our findings using the minority share of
the population, again we find that in the first round there appears to be a negative relationship
between loan receipt and minority business share, and in the second round the relationship switches
to being positive.

Figure 6: PPP loans per employer establishment by minority share of businesses in the 1st round
Panel A: Weighted by population Panel B: Unweighted

0.28
0.27
0.26
0.25
0.24
0.23
0.22
0.21
0.2
0.19
0.18
0.17

0.15 0.16

0.13 0.14

0.11 0.12

0.09 0.1

Notes: The charts show the mean number of PPP loans per employer establishment in counties by minority share of
businesses in the first round of the PPP program (April 3-April 16, 2020). Loans to agricultural businesses are
excluded. Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic regressions
at the county level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard deviation.

12
Figure 7: PPP loans per employer establishment by minority share of businesses in the 2nd round
Panel A: Weighted by population Panel B: Unweighted

0.46
0.41

0.41
0.36

0.36 0.31

0.31 0.26

0.26 0.21

Notes: The charts show the mean number of PPP loans per employer establishment in counties by minority share of
businesses in the second round of the PPP program (April 27-August 8, 2020). Loans to agricultural businesses are
excluded. Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic regressions
at the county level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard deviation.

PPP Loan Amounts
The disbursement of PPP funds across communities by minority share might differ when measured
by loan amounts instead of number of loans. Figure 8 displays average loan amounts per business
employee by minority share in the population at the zip code level. We standardize the Y-axis by
reporting the range of ± ½ standard deviations around the median loan size (cutting off at zero in
case of the unweighted chart). We find a slight downward relationship with minority share.
Moving from the 25th to the 75th percentile in minority share is associated with a decrease from
$4652 to $4204 in average loan amount per employee.

13
Figure 8: PPP loan amounts per employee by minority share
Panel A: Weighted by population Panel B: Unweighted

10000
5200

8000
4700

6000
4200

4000
3700

3200 2000

2700 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean amounts of PPP loans per employee in zip codes by minority share of the population.
Loans to agricultural businesses are excluded. Loans reported as a range are approximated by using the interval
midpoint. Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic regressions
at the zip code level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard deviation.

3.3 Economic Injury Disaster Loan (EIDL) Programs
Although the PPP program has received a lot of attention, the federal government also approved
the $220 billion EIDL program, which also provides aid to small businesses during COVID-19,
but has received much less attention. There are two programs, EIDL loans and EIDL advances.
EIDL loans are not forgivable and must be paid back in full. EIDL advances are grants and do not
have to be repaid, but are for smaller amounts ($1,000 per employee up to $10,000 total).
Figure 9 displays EIDL loan receipt per employer establishment by minority share of the
population across zip codes. The relationship between loan receipt and minority population share
shows a clear upward pattern. If we move from the lowest quartile minority share (16 percent) to
the highest quartile minority share (59 percent) loan receipt increases from 0.20 to 0.31 EIDL loans
per employer business establishment.

14
Figure 9: EIDL loans per employer establishment by minority share
Panel A: Weighted by population Panel B: Unweighted

0.48 0.32

0.43
0.27

0.38
0.22
0.33
0.17
0.28
0.12
0.23

0.18 0.07

0.13 0.02
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean number of EIDL loans per employer establishment in zip codes by minority share
of the population. Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic
regressions at the zip code level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard
deviation.

Figure 10 displays EIDL advance receipt per employer establishment by minority share of
the population in zip codes. The relationship between advance receipt and minority population
share shows a similarly strong upward pattern. Movement from the lowest quartile to the highest
quartile minority share loan receipt increases from 0.42 to 0.69 EIDL advances per employer
business establishment.

15
Figure 10: EIDL advances per employer establishment by minority share
Panel A: Weighted by population Panel B: Unweighted

1.41
0.74

1.21 0.64

1.01 0.54

0.44
0.81
0.34
0.61
0.24

0.41
0.14

0.21 0.04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean number of EIDL advances per employer establishment in zip codes by minority
share of the population. Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic
regressions at the zip code level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard
deviation.

EIDL Loan Amounts
Figure 11 displays EIDL loan amounts per employee by minority share of the zip code. Similar to
the number of loans we find a positive relationship between loan amounts and minority share of
the population based on the weighted chart. An increase in EIDL loans per employee from $1404
to $1624 is associated with the interquartile range in minority share across zip codes. Figure 12
displays EIDL advances per employee by minority share. We also find a positive relationship for
EIDL advances increasing from a weighted average of $148 to $198 per employee when moving
from the 25th to the 75th percentile in minority share.

16
Figure 11: EIDL loan amounts per employee by minority share
Panel A: Weighted by population Panel B: Unweighted

2000 4000
1800
3500
1600
3000
1400
2500
1200
2000
1000
1500
800

600 1000

400 500

200 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean EIDL loan amounts per employee in zip codes by minority share of the population.
Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic regressions at the zip
code level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard deviation.

Figure 12: EIDL advance amounts per employee by minority share
Panel A: Weighted by population Panel B: Unweighted

600 450

400
500
350
400 300

250
300
200

200 150

100
100
50

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Notes: The charts show the mean EIDL advance amounts per employee in zip codes by minority share of the
population. Panel A uses population weights, Panel B is unweighted. The dashed lines are from quadratic regressions
at the zip code level. For perspective, the double arrow on the Y-axis indicates the median ± ½ standard deviation.

17
4. Conclusions
Given the shutdown of the economy to slow down the spread of the novel coronavirus Congress
agreed to a massive level of expenditures in the 2020 CARES Act to help small businesses stay
open and retain employees. Two components directly providing loan and grant assistance to small
businesses, the PPP and EIDL programs, provided a total of nearly 15 million separate loans or
advances, and a staggering level of expenditures of roughly $850 billion. The total number and
amount of support for small businesses in the United States is unprecedented. Given that the
programs were to help disadvantaged businesses (U.S. Congress 2020) we provide the first study
of whether loans and advances from these programs were indeed distributed positively to minority
communities.
Using administrative data on the universe of PPP loans, EIDL loans, and EIDL advances,
we explore how loans and advances were distributed. We find that funding from these programs
both flowed to minority communities and away from minority communities. Focusing first on PPP
loans, we generally find a slightly positive relationship between PPP loan receipt per business and
the minority share of the population. There is some evidence, however, that the first round of funds
was disproportionately disbursed to non-minority communities and the second round of funds was
disproportionately disbursed to minority communities. When we focus on the minority share of
employer businesses in an area we find similar results: slightly positive relationship but differential
relationships by disbursement rounds. Focusing on PPP loan amount per employee we find a
negative relationship with minority share of the population. EIDL loans and advances, in both
number and amounts, were provided positively to minority communities. We find a strong positive
relationship in the receipt of these loans and advances by the minority share of the population.
Although analyzing patterns of PPP and EIDL funding receipt across minority
communities by using the universe of loan-level data across minority communities is important,
the loan-level data are limited by not having information on loan receipt by race and ethnicity. To
be sure, there is some information in the PPP loan data, but only 10 percent of loans include race
and ethnicity (and in a non-representative way by lender), and none of the loans in the EIDL data
provide information on race and ethnicity. There is always the possibility that minority businesses
did not evenly receive loans in geographical areas even with high minority shares of the population
or high minority shares of businesses. The federal government has been criticized heavily for not

18
collecting this information and plans on collecting demographic information when processing
forgiveness on the PPP loans. Future research needs to address this critical question.
Another criticism of the programs is that there was no collection of information on
applications for loans that were denied. There is no way to gauge demand and unmet need for these
loans by minority businesses and in minority communities. Although there is currently no
information by race, the Census Bureau’s Small Business Pulse Survey indicates that by early
August most businesses in their survey who asked for PPP or EIDL funds reported receiving them
(U.S. Census Bureau 2020). But, this is an important concern. There might exist large disparities
by race, and there is a major difference in potential policy response between whether minority
businesses needed these loans but faced barriers (e.g. lack of established bank relationships, lack
of information about loans, digital divide, or discrimination) or if they did not need loans or needed
smaller loans. Another concern is that many minority businesses did not have employees and the
programs were primarily focused on serving employer businesses. Finally, many minority
businesses might have been reluctant to apply for PPP loans because of uncertainty over future
revenues due to entering the pandemic in a weakened position (Mills and Battisto 2020).
The findings presented in this research note have implications for trends in broader
inequality. Minority-owned businesses are important for local job creation (as minority owners
disproportionately hire minority workers), economic advancement, and longer-term wealth
inequality (Boston 1999, 2006; Stoll et al. 2001; Bradford 2003, 2014; Fairlie and Robb 2008).
With major losses in business activity among minority businesses in the early stages of the
pandemic (Fairlie 2020) minority business owners have already lost substantial amounts of income
from their businesses. If the pandemic continues over a long period of time the long-term economic
consequences on minority businesses could be severe. Many minority business owners will not
have the resources to weather prolonged closures, reduced demand from health concerns, and a
more comprehensive recession. Just prior to the pandemic when small business owners were asked
what actions they would take if faced with a two-month revenue loss roughly half said they would
use their own funds and 17 percent said they would close or sell the business (Mills et al. 2020).
But, the latest Census data indicate that the median level of wealth among black families is $13,000
and Latinx families is $20,000 compared with $139,000 among white families possibly making it
difficult to use their own funds for an extended period of time (U.S. Census Bureau 2015).

19
The government just passed a new $892 billion COVID relief package, and private
foundations and companies are promising to help. Can these programs help small businesses
survive the setbacks and shutdowns due to the coronavirus pandemic, or will more assistance be
needed? Furthermore, will an added shift in consumer behavior away from large online retailers
towards small businesses be needed? In light of the Black Lives Matter movement there has been
an unprecedented push to support black-owned stores around the country, and states have
promoted shopping local (e.g. California’s #ShopSafeShopLocal). In the end, getting the virus in
check and restoring customer, owner and employee confidence in health risks is likely the first real
step to a full recovery for small businesses.

20
References
Annette Alstadsaeter, Julie Brun Bjorkheim, Wojciech Kopczuk, and Andreas Okland (2020),
Norwegian and U.S. Policies Alleviate Business Vulnerability Due to the COVID-19 Shock
Equally Well, National Tax Journal, 73:3, pp. 805-828
Bartik, Alexander W., Zoe B. Cullen, Edward L. Glaeser, Michael Luca, Christopher T. Stanton,
and Adi Sunderam. 2020. “The targeting and impact of Paycheck Protection Program loans to
small businesses,” National Bureau of Economic Research WP No. w27623.
Bhutta, Neil, Jacqueline Blair, Lisa Dettling, and Kevin Moore (2020), COVID-19, the CARES
Act, and Families Financial Security, National Tax Journal, 73:3, pp. 645-672
Bloomberg. 2020. “White America Got a Head Start on Small-Business Virus Relief” by Jason
Grotto, Zachary R. Mider and Cedric Sam, July 30, 2020.
Bohn, Sarah, Marisol Cuellar Mejia, and Julien Lafortune. 2020. “The Economic Toll of
COVID-19 on Small Business,” Public Policy Institute of California.
Boston, Thomas D. 1999. “Generating Jobs Through African American Business Development,”
in J. Whitehead and C. Harris, eds. Readings in Black Political Economy. Dubuque: Kendall-
Hunt.
Boston, Thomas D. 2006. “The Role of Black-owned Businesses in Black Community
Development” ed. Paul Ong, Jobs and Economic Development in Minority Communities:
Realities, Challenges, and Innovation. Temple University Press
Bradford, William D. 2003. “The Wealth Dynamics of Entrepreneurship for Black and White
Families in the U.S.,” Review of Income and Wealth, 49(1): 89-116.
Bradford, William D. 2014. "The “Myth” that Black Entrepreneurship can Reduce the Gap in
Wealth between Black and White Families," Economic Development Quarterly 28.3: 254-269.
Centers for Disease Control and Prevention (CDC). 2020. “Identifying the Source of the
Outbreak,” https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/about-
epidemiology/identifying-source-outbreak.html
Centers for Disease Control and Prevention (CDC). 2020. “Trends in Number of COVID-19
Cases in the US Reported to CDC, by State/Territory,” https://www.cdc.gov/covid-data-
tracker/#trends
Dávila, Alberto, and Marie Mora. 2013. Hispanic Entrepreneurs in the 2000s: An Economic
Profile and Policy Implications. Stanford University Press, Stanford.
Desai, Sameeksha, and Jessica Looze. 2020. “Business Owner Perceptions of COVID-19 Effects
on the Business: Preliminary Findings,” Trends in Entrepreneurship, No. 10. Kauffman
Foundation.
Erel, Isil, and Jack Liebersohn. 2020. “Does FinTech Substitute for Banks? Evidence from the
Paycheck Protection Program.” NBER Working Paper No. 27659.
21
Farrell, Diana, Chris Wheat and Chi Mac. 2020. “Small Business Financial Outcomes during the
Onset of COVID-19,” JPMorgan Chase & Co. Institute Report.
Fairlie, Robert W. 2020. “The Impact of Covid-19 on Small Business Owners: Evidence of
Early-Stage Losses from the April 2020 Current Population Survey,” May 23, 2020. Stanford
University (SIEPR) Working Paper No. 20-022.
Fairlie, Robert W., and Frank M. Fossen. 2019. “Opportunity versus Necessity Entrepreneurship:
Two Components of Business Creation,” National Bureau of Economic Research, Working
Paper No. w26377.
Fairlie, Robert W., and Alicia M. Robb. 2008. Race and Entrepreneurial Success: Black-, Asian-
and White-Owned Businesses in the United States, Cambridge: MIT Press.
Granja, João, Christos Makridis, Constantine Yannelis, and Eric Zwick. 2020. “Did the Paycheck
Protection Program Hit the Target?” National Bureau of Economic Research, Working Paper No.
w27095.
Kerr, Sari Pekkala, and William Kerr. 2020. “Immigrant Entrepreneurship in America: Evidence
from the Survey of Business Owners 2007 & 2012,” Research Policy 49.3: 103918.
Lederer, Anneliese, Sara Oros, Sterling Bone, Glenn Christensen, and Jerome Williams. 2020.
“Lending Discrimination within the Paycheck Protection Program,” National Community
Reinvestment Coalition, Washington D.C.
Liu, Sifan, and Joseph Parilla. 2020. “New Data Shows Small Businesses in Communities of
Color Had Unequal Access to Federal COVID-19 Relief,” Brookings Institution Report.
Mills, Claire Kramer, and Jessica Battisto. 2020. “Double Jeopardy: COVID-19’S Concentrated
Health and Wealth Effects in Black Communities,” Federal Reserve Bank of New York.
Mills, Claire Kramer, Jessica Battisto, Mels de Zeeuw, Scott Lieberman, and Ann Marie
Wiersch. 2020. Small Business Credit Survey, Federal Reserve Banks.
Neilson, Christopher, John Eric Humphries, and Gabriel Ulyssea. 2020. “Information Frictions
and Access to the Paycheck Protection Program,” National Bureau of Economic Research,
Working Paper No. w27624, forthcoming Journal of Public Economics.
Newsom, Gavin. 2020. “Governor Newsom Signs Bills to Support Small Businesses Grappling
with Impact of COVID-19 Pandemic, Bolster Economic Recovery,” State of California, Office
of the Governor.
Parker, Simon C. 2018. The Economics of Entrepreneurship. Cambridge University Press
Stanford Latino Entrepreneurship Initiative. 2020. The impact of COVID-19 on Latino-owned
business. Stanford, CA: Stanford Latino Entrepreneurship Initiative.
Stoll, Michael A., Steven Raphael, and Harry J. Holzer. 2001. “Why Are Black Employers More
Likely than White Employers to Hire Blacks?” Institute for Research on Poverty, University of
Wisconsin, Madison.
22
U.S. Census Bureau. 2016. “Survey of Business Owners (SBO) - Survey Results: 2012”
U.S. Census Bureau. 2020. “Small Business Pulse Survey.” https://portal.census.gov/pulse/data/
U.S. Senate. 2020. “Cardin, Cantwell, Schumer, Booker, Cortez Masto & Harris Introduce
Legislation to Invest in Minority-Owned Businesses,” U.S. Senate Committee on Small Business
& Entrepreneurship.
U.S. Small Business Administration 2020. “Paycheck Protection Program Borrower Application
Form,” https://www.sba.gov/sites/default/files/2020-07/PPP-Borrower-Application-Form-
508.pdf
Wilmoth, Daniel. 2020. “Small Business Facts: Early Data Show Severe Disruptions,” U.S.
Small Business Administration.

File and source

File: did-the-ppp-and-eidl-program-get-disbursed-to-minority-communities-ear.pdf
Size: 788,181 bytes
SHA-256: 76f10656e0e11c4a7a0f2302a5985435611f10ef2b66a591dcd88f5880fc11b2
Our copy: did-the-ppp-and-eidl-program-get-disbursed-to-minority-communities-ear.pdf
Original: www.nber.org