Impact of Race,Ethnicity,and Gender on the SBA Paycheck Protection Program (PPP) Loan Amounts

Abstract

The Paycheck Protection Program helped to preserve employment relationships during the sudden shutdown of economic activity due to the COVID-19 pandemic. This paper analyzes whether small business owners’ race, ethnicity, or gender played a role in the PPP loan amount received. In 2020 and 2021, non-White-, Hispanic-, and female-owned small businesses received smaller PPP loans than their business counterparts of the same size. Larger companies displayed increased discrepancies in loan amounts. From 2020 to 2021, disparities among non-White business owners decreased; however, female and Hispanic owners continued to receive less in PPP loans than male and non-Hispanic owners. Lee bounds estimates show that female-owned businesses in rural counties received smaller PPP loans per employee than female-owned businesses in urban counties. Structural interviews with PPP loan recipients in Northeast Ohio showed that businesses receiving smaller loan amounts had more difficulties and less knowledge about the loan application process compared to larger loans recipients.

Keywords

Minority-owned business Hispanic-owned business female-owned business economic inclusion Paycheck Protection Program (PPP)

The small business crisis has fallen disproportionally on minority- and female-owned businesses. From February 2020 to April 2020, the COVID-19 pandemic caused 41% of Black-owned businesses and 25% of female-owned firms in the United States to close (Fairlie, 2020). Small companies are important employers and contributors to the economy. In 2019, 1 million minority-owned small businesses in the United States employed almost 10 million workers and generated more than $1.5 trillion in economic output. Female-owned businesses represented nearly 300,000 of all minority-owned small businesses, employing 2.4 million workers (U.S. Census Bureau, 2019). Hispanic-owned businesses employed nearly 3 million people nationally. Minority- and female-owned small businesses tend to be smaller in size. Based on the U.S. Census Annual Business Survey, over half of them had one to four employees and nearly 80% had nine employees or less (see online Appendix Table A1).

This paper analyzes whether small business owners’ race, ethnicity, or gender played a role in the loan amount granted through the Paycheck Protection Program (PPP). PPP loans are guaranteed, and are potentially forgivable loans offered by the U.S. Small Business Administration (SBA) to small businesses affected by the pandemic to keep their workers on the payroll. With almost $800 billion approved for more than 11 million loans passing through 5,469 financial institutions, the PPP has been one of the largest economic stimulus programs in U.S. history.¹ In 2020, PPP provided 4,716,034 loans to small businesses who employed almost 60 million U.S. workers. In 2021, SBA distributed 6,154,040 PPP loans to 28 million U.S. workers. Minority- and female-owned businesses may have been more vulnerable to the pandemic, yet studies have shown that they face challenges when it comes to credit access. Our findings highlight that, from 2020 to 2021, disparities among non-White owners decreased; however, female and Hispanic owners continued to receive less in PPP loans than male and non-Hispanic owners. Policy makers can use our results to design policies that better target minority- and female-owned small businesses.

Literature Review and Contribution

Our contribution to the literature is fourfold. First, our paper adds to the literature on small-business credit access, a crucial element to their growth. The use of commercial financing reduces the exit rates of new firms (Yunwei & Minniti, 2015). Lack of access to credit is a leading cause of small business closure (Carter & Van Auken, 2006). For most small firms, survival is reliant on capital access (Neely & Van Auken, 2012). Given the inherent risk of starting a small business, banks and other funding sources tend to be wary of providing capital (Galli-Debicella, 2020). The SBA arose to fill this gap and provide greater funding sources to those unable to access credit (Craig et al., 2009). Research has shown that businesses that obtain SBA loans have a greater survival rate than those that do not (Galli-Debicella, 2020). PPP loans were instrumental in providing fast cash relief for businesses and funds to keep their employees on the payroll.

Second, we add to the literature on access to credit by minority-, Hispanic-, and female-owned small businesses. The Equal Credit Opportunity Act (ECOA) requires that equally qualified borrowers be able to access credit regardless of their race, ethnicity, gender, and other protected classes. Extensive extant literature unpacks the role of discrimination in credit access. Becker (1971) hypothesized that individuals who tend to discriminate behave as though they are willing to pay something, either directly or in the form of a reduced income, to be able to discriminate. This theory asserts that individuals discriminate because they prefer to avoid members of another race or gender. In the context of credit markets, a financial institution that would normally loan funds at rate r will require instead a rate of $r (1 + d)$ , with d being the discrimination coefficient. Phelps (1972) considered the case in which lenders use a demographic group as a proxy for the economically important factors when deciding to whom they will issue a loan.

According to the racial formation theory, racial inequalities stem from society's construction of race, which expands from the institutional level to day-to-day social interactions (Omi & Winant, 2014). Everybody learns some combination some version of the rules of racial classification and their own racial identity. Race continues to signify differences in distribution of economic, political, and cultural resources. State actions in the past and present have treated people in very different ways according to their race. Dominant groups (e.g., governments) have the power to reinforce or alleviate inequalities (Omi & Winant, 2014). Using the experience of five Latino-owned coffee shops in Los Angeles, Santellano (2021) argued that the effects of institutional racism surfaced in the pandemic and the PPP. Business owners who had established relationships with lenders benefitted over those who did not (Santellano, 2021). However, a previous relationship with a lender may be challenging for businesses with limited access to capital. The roots of limited access to capital and to credit may be derived from previous historic governmental programs, such as redlining.

The National Survey of Small Business Finances (NSSBF), collected by the Federal Reserve from 1987 to 2003, has been used by many researchers as a common source to investigate some of the factors influencing differentials in the credit market experiences of small businesses across different demographic groups. Findings from the 2003, 1998, and 1993 NSSBF revealed that women and minority borrowers experience greater difficulty securing loans than White male borrowers. Using the 1993 NSSBF ( $n =$ 3,400), Cavalluzzo and Cavalluzzo (1998) observed substantial ceteris paribus differences in denial rates between businesses owned by Black and White male owners. They concluded that there is insufficient evidence to eliminate discrimination as a potential explanation for some observed differences. Using data from the 1993 and 1998 NSSBF ( $n =$ 3,561), Blanchflower et al. (2003) found that Black-owned businesses were about twice as likely to be denied credit, even after controlling for observed differences in creditworthiness and other factors. Notably, the magnitude of the Black-White differential in small-business loan approval rates is considerably larger than in mortgage discrimination. According to Blanchard et al. (2008), differences in loan approval rates are driven by lenders’ stereotypes about the ability of Black and Hispanic business owners to succeed under some circumstances. Using data from the 1998 and 2003 NSSBF ( $n =$ 1,256), Asiedu et al. (2012) found that both Black and Hispanic business owners faced discrimination in obtaining loan renewals in 2003. Among large-scale, well-established small businesses, loan denial rates were 31.5% for minority-owned businesses and 12.3% for nonminority-owned businesses (Bates & Robb, 2013).

In the Small Business Credit Survey conducted by the Federal Reserve in 2016 ( $n =$ 7,916), Black-owned businesses applied for new funding at a rate 10% higher than White-owned businesses, but their approval rates were 19% lower. When looking only at firms approved for partial financing, 40% of minority-owned firms with good personal and/or business credit scores received the full amount requested, compared to 68% of nonminority-owned firms in good standing (U.S. Federal Reserve Banks, 2017). Similar trends continued in 2019 (U.S. Federal Reserve Banks, 2019).

Lack of access to credit is a large driver of the high failure rate of Hispanic-owned businesses (Canedo et al., 2014). Hispanic-owned businesses have shown rapid growth since 2000 but have quicker firm closure rates than those owned by nonminorities (Williams et al., 2012). According to the Survey of Business Owners, Hispanic women-owned businesses failed at a 47% rate (Mora & Dávila, 2014). Blanchard et al. (2005) found that Hispanic business owners were more likely than White male business owners to be denied credit and, when granted credit, paid a higher interest rate.

Credit constraints are a bigger issue for female-owned businesses (Carter & Allen, 1997; Giglio, 2020). Female-owned businesses are concentrated disproportionately in crowded subsections of lower-order services (e.g., retail, leisure, and hospitality) and are relatively scarce in value-added and capital-intensive sectors. This lessens the availability of credit. Asymmetric information for female-owned businesses causes credit to be more difficult to obtain or unfairly priced (Scalera & Zazzaro, 2001). In the 2016 Small Business Credit Survey, female owners applied for business loans at around the same rate as male owners but were much less likely to obtain financing (47% success compared to 61%). Fewer female-owned businesses were granted all the financing they requested in comparison to male-owned businesses, and more female-owned businesses did not receive any financial assistance (U.S. Federal Reserve Banks, 2017).

Third, we contribute to the literature examining geographic impacts on small business financing (secondary hypotheses). Using 2004 to 2011 Kauffman Firm Survey data ( $n =$ 657), Bates and Robb (2016) investigated loan accessibility among urban small businesses. Black-owned firms, other factors being equal, were associated with smaller loan amounts. Using the Community Reinvestment Act (CRA) small business lending data, Immergluck (2002) measured small business lending flows to neighborhoods in the Philadelphia metropolitan area ( $n =$ 1,197). Tracts with a large Black population received fewer loans after accounting for firm density, firm size, industrial mix, neighborhood income, and the credit quality of local firms.

Distribution of loans to small businesses in rural areas was among the three criteria to measure the impact of the changes introduced by the SBA to the PPP in 2021 (SBA, 2021). Rural America has become more racially and ethnically diverse over the last decade. Over one-fifth of rural America is non-White according to the 2020 U.S. Census population data (Rowlands & Love, 2021). Rural small businesses often lack adequate access to capital and broadband connectivity and are largely dependent upon the industries that were most immediately vulnerable to the pandemic effects (Love & Powe, 2020). Despite the importance of rural communities, policy narratives about rural America frequently neglect the experiences of Black and Hispanic populations (Ajilore & Willingham, 2019).

Finally, we contribute to the fast-evolving literature on unequal access to the PPP program. Liu and Parilla (2020) broke down how small businesses in majority-White neighborhoods received PPP loans faster than small businesses in majority-Black and majority-Latino or Hispanic neighborhoods. Fairlie and Fossen (2021) also found that PPP funds flowed to minority communities later than to communities with lower minority shares. A Survey of Small Business Finances of over 9,000 employer firms found that Black-owned businesses were only half as likely as White-owned businesses to receive the entire PPP funding amount they requested, and they were nearly five times as likely to receive none (U.S. Federal Reserve Banks, 2021). According to Atkins et al. (2022a), Black-owned businesses received loans that were approximately 50% lower than observationally similar White-owned businesses. Erel and Liebersohn (2020) found that fintech was disproportionately used in ZIP codes with a larger minority share of the population.

The Paycheck Protection Program and Data Description

The PPP was authorized by the federal government's passage of the Coronavirus Aid, Relief, and Economic Security (CARES) Act in late March 2020. These loans were designed to directly incentivize small businesses, self-employed workers, sole proprietors, certain nonprofit organizations, and tribal businesses impacted by COVID-19 to keep their workers on the payroll through the pandemic. Congress built on the principles of the SBA's existing 7(a) loan guarantee program to distribute loans through certified lenders (banks, credit unions, CDFIs, and, eventually, financial technology [fintech] companies and nonbank lenders). SBA removed the majority of the 7(a) program's rules, requiring no fees, no credit scores, and no collateral from applicants.

To be eligible for the PPP in 2020, an applicant must have had 500 or fewer employees per location. On June 26, 2020, SBA provided an official guidance on how to calculate the maximum amount of a PPP loan for different types of businesses (SBA, 2020, June 26). The PPP loan amount was generally equal to 2.5 times average monthly payroll costs, subtracting any amounts paid to any individual employee in excess of $100,000 and any amounts paid to any employee whose principal place of residence is outside the United States. In addition, the loans could also be used to pay for mortgage interest, rent, utilities, worker protection costs related to COVID-19, and uninsured property damage costs caused by looting or vandalism during 2020. The SBA issues loan forgiveness if at least 60% of the proceeds are spent on payroll costs. If not forgiven, PPP loans have an interest rate of 1%.

This analysis employs the national database of PPP loan recipients released by the SBA on June 30, 2021. In 2020, during the first draw of PPP loans (approvals from April 3, 2000-August 8, 2000), the average size of a PPP loan was $113,754. Original PPP funds were depleted during the first round (April 3, 2020-April 16, 2020). The SBA started to accept applications again on April 27, 2020, with extra funds from the Paycheck Protection Program and Health Care Enhancement Act.

The second draw in 2021 (approvals from January 11, 2021-May 31, 2021) was available only to businesses with 300 or fewer employees that had sustained a 25% reduction in gross revenues between comparable quarters in 2019 and 2020. The maximum loan amount was kept at $10 million in 2020 and at $2 million in 2021. As a result, the average size of the PPP loan in 2021 was smaller than in 2020, $43,487.

To ensure all program rules were followed, all loans went through an automated review. SBA manually reviewed all loans of $2 million or more. In addition, any loan could be selected for a manual review (SBA, 2020). PPP loan applications had to be submitted in English to the approved PPP lender. SBA provided documents in 17 different languages to explain eligibility requirements, helped fill out applications, and answered frequently asked questions.² The SBA encouraged PPP loan recipients to support the American economy by asking them to purchase American-made equipment and products to the extent feasible.³ Demographic information for borrowers on the PPP loan application was requested on a voluntary basis.

SBA data in both years include 2,670,324 PPP loan recipients with known business owners’ race (Table 1). Based on this self-reported demographic information, minority (or non-White) business owners received 1,219,747 loans totaling $41.1 billion.⁴ The average PPP loan received by minority business owners was $67,811 in 2020 and $25,857 in 2021, almost half of what business owners with unanswered race and White business owners received.

Table 1.

Business Owner Characteristics of PPP Loans Recipients in 2020 and 2021.

Business owner characteristics	Number of loans		Mean		Std dev		Sum ($millions)
Business owner characteristics	2020	2021	2020	2021	2020	2021	2020	2021
Business owner race
Minority-owned	227,235	992,512	$67,811	$25,857	$220,041	$69,369	$15,400	$25,700
White-owned	605,698	844,879	$117,111	$50,447	$343,318	$151,925	$70,900	$42,600
Unanswered	3,883,111	4,316,649	$115,919	$46,178	$395,004	$154,429	$450,000	$199,000
Business owner ethnicity
Hispanic-owned	112,495	234,793	$84,180	$37,486	$245,538	$105,618	$9,470	$8,800
Non-Hispanic	962,203	1,820,701	$110,742	$44,691	$317,456	$132,116	$107,000	$81,400
Unanswered	3,641,346	4,098,546	$115,464	$43,296	$400,785	$150,791	$420,000	$177,000
Business owner gender
Female-owned	406,022	1,073,008	$77,786	$29,834	$233,573	$90,229	$31,600	$32,000
Male-owned	1,110,539	1,659,758	$127,416	$53,251	$361,673	$150,507	$142,000	$88,400
Unanswered	3,199,478	3,421,274	$113,576	$43,032	$403,387	$153,783	$363,000	$147,000
PPP loans (Total)	4,716,044	6,154,040	$113,754	$43,487	$382,164	$143,996	$536,000	$268,000

Source: SBA PPP Data released on June 30, 2021. All monetary values are in 2022 dollars.

SBA data include 3,130,192 PPP loan recipients that indicated their ethnicity (Table 1). Hispanic business owners received 112,495 loans during 2020 and 234,793 in 2021. The average loan amount received by Hispanic-owned businesses was $84,180 in 2020 and $37,486 in 2021. The difference between the average PPP amounts received by non-Hispanic and Hispanic small business owners reduced from 2020 to 2021, indicating a potential decrease in disparities in 2021. The average PPP loan for businesses with unanswered ethnicity was above the average PPP loan amount awarded to Hispanic-owned businesses.

SBA data include 4,249,327 PPP loan recipients that volunteered information about their gender (Table 1). A total of 406,022 PPP loans were awarded to female-owned businesses in 2020 and a little over 1 million loans in 2021. The average loan awarded to female business owners was $77,786 in 2020 and $29,834 in 2021. Although the average amount in 2021 is lower than that of 2020, the discrepancy with the average PPP loans awarded to male-owned and unanswered gender small businesses was smaller in 2021.

Differences in PPP amounts across minority-, Hispanic-, and female-owned small businesses may be partially explained by the different sizes of these businesses compared to White-, non-Hispanic-, and male-owned businesses (see online Appendix Table A2). PPP loan data included information on the number of jobs by each borrower, which we used as a proxy for the number of employees. On average, minority-owned businesses had fewer employees than White-owned businesses in both years. Female business owners also employed fewer people than male business owners. Our main regression includes interactions of the variables of interest (business owner race, ethnicity, gender) with three categories of company size: (1) less than 4 employees, (2) 5 to 19 employees, and (3) 20 to 499 employees.

To compare access to PPP loans in rural areas to urban areas, we used the rural or urban indicator included in SBA data released in June 2021. Small businesses located in urban counties received 80% of the PPP loans. The median PPP loan received by a small business in rural areas was smaller than in urban areas in both years.

The SBA also provided the names of the financial institutions (but no other identifiers) that facilitated the loan applications and distributions. PPP loans were allocated through eligible financial institutions. These eligible institutions included regional and community banks, credit unions, Farm Credit Associations, Community Development Financial Institutions (CDFIs), and Minority Depository Institutions, or any other lender approved by the SBA and enrolled in the program. To ensure the funds reached a lot of areas underserved by traditional banks, online lenders (e.g., fintechs) were made eligible to issue PPP loans on April 14, 2020.

We supplemented information on lender characteristics from various sources. Information on commercial banks was collected from the Federal Deposit Insurance Corporation call reports. Similarly, data on farm credit lenders were collected from call reports made available by the Farm Credit Association. Credit union information was collected from call reports from the National Credit Union Administration. Further, characteristics of nontraditional lenders (e.g., nonbanks, fintechs, CDFIs, nonprofits) were gathered from Mergent Online Non-Profit Explorer, 990 IRS forms, and company websites.⁵

Summary statistics for each variable are shown in Table 2. On average, PPP lenders had $10 billion in assets. Commercial banks distributed 70% of the PPP loans. Nontraditional lenders distributed 27%. The top lender was JPMorgan Chase Bank.⁶ On average, loans through commercial banks were larger than those made through nontraditional lenders.

Table 2.

Summary Statistics for the Variables Used in the Macro-Level Analysis for 2020 and 2021.

Variables	Description	Mean	Std dev	Observations
PPP Loans	Loan amount awarded through the PPP program	$73,973	$276,253	10,870,084
PPP per employee	PPP loan per employee / job	$11,855	$9,904	10,870,084
Owner Race
White-owned	White-owned businesses take the value of 1 and zero otherwise (base-case)	0.13	0.34	10,870,084
Minority-owned	Minority-owned businesses take the value of 1 and zero otherwise	0.11	0.32	10,870,084
Unanswered owner race	Owner race characteristics unknown take the value of 1 and zero otherwise	0.75	0.43	10,870,084
Owner Gender
Male-owned	Male-owned businesses take the value of 1 and zero otherwise (base-case)	0.25	0.44	10,870,084
Female-owned	Female-owned businesses take the value of 1 and zero otherwise	0.14	0.34	10,870,084
Unanswered owner gender	Owner gender characteristics unknown take the value of 1 and zero otherwise	0.61	0.49	10,870,084
Owner Ethnicity
Not Hispanic-owned	Not Hispanic owned businesses take the value of 1 and zero otherwise (base-case)	0.26	0.44	10,870,084
Hispanic-owned	Hispanic-owned businesses take the value of 1 and zero otherwise	0.03	0.18	10,870,084
Unanswered owner ethnicity	Owner ethnicity unknown take the value of 1 and zero otherwise	0.71	0.45	10,870,084
Rural county	Dummy that takes on the value of 1 if county is rural and zero otherwise	0.21	0.41	10,870,084
Loan demand	Total amount of CRA loans made to small businesses and farms in 2019	$8,889,054	$16.9 million	10,870,084
LMI	Low- to moderate-income counties take the value of 1 and zero otherwise	0.71	0.45	10,870,080
Lender assets	Total lender assets (in $ million)	$104,000	$483,000	10,866,467
Lenders
Credit union	Credit union lender takes the value of 1 and zero otherwise (base-case)	0.03	0.17	10,866,467
Commercial bank	Commercial bank lender takes the value of 1 and zero otherwise	0.70	0.46	10,866,467
Farm Credit Association	Farm Credit Association lender takes the value of 1 and zero otherwise	0.004	0.06	10,866,467
Nontraditional lender	Nontraditional (including nonbanks, Fintechs, CDFI) lender takes the value of 1 and zero otherwise	0.269	0.44	10,866,467
Second PPP round (Apr 27-Aug 8 2020)	Dummy that takes on the value of 1 if application is from the second round (April 27-August 28) and zero otherwise	0.68	0.47	4,716,040
Days to approval	Number of days between April 3, 2020, and the loan approval	54.62	39.79	10,870,084
Days in lockdown	Total days of mandated state or county lockdown	6.62	5.60	4,716,040
Nonprofit	Business is a nonprofit takes the value of 1 and zero otherwise	0.02	0.15	10,870,084
Company Size
Less than four employees	Companies with less than four employees take the value of 1 and zero otherwise (base-case)	0.71	0.46	10,870,084
5 to 19 employees	Companies with 5 to 19 employees take the value of 1 and zero otherwise	0.21	0.41	10,870,084
20 to 499 employees	Companies with 20 to 499 employees take the value of 1 and zero otherwise	0.08	0.28	10,870,084
Unanswered	Company size is unknown take the value of 1 and zero otherwise	0.0004	0.02	10,870,084

The number of days to approval represents the number of days from April 3 (when the program started) to the date of approval. As such, applications made later would have a greater number of days to approval. The number of mandated lockdown days was calculated using the dates from Wu et al. (2020). Small business loan demand was approximated by the amount in loans granted through the Credit Reinvestment Act (CRA).⁷ Under the CRA, commercial banks are required to collect and report on loans made to small businesses (Bostic & Lee, 2017). Data on the CRA were collected from the Federal Financial Institutions Examination Council. We aggregated the loan amounts awarded to small businesses and farms during the years of 2019 and 2020.

Hypotheses

This study investigates the role of business owners’ race, ethnicity, or gender in the PPP loan amount awarded. Using the ordinary least squares (OLS) estimation, we test three primary hypotheses: 1) minority-owned small businesses received a lower loan amount than White-owned small businesses, 2) Hispanic-owned small businesses received a lower loan amount than non-Hispanic small businesses, and 3) female-owned small businesses received a lower loan amount than male-owned small businesses. Then, we run a Lee's (2009) treatment-effect bound estimator to test three secondary hypotheses: 1) minority-owned small businesses located in rural counties received a lower loan amount per employee than those in urban counties, 2) Hispanic-owned businesses located in rural counties received a lower loan amount per employee than those in urban counties, and 3) female-owned businesses in rural counties received a lower loan amount per employee than those in urban counties.

Methodology

Our analysis has both macro and micro levels. At the macro level, we investigate whether business owner race, ethnicity, and gender affected the loan amount granted. We control for lender type, lender size (in assets), and business location characteristics. Our macro-level analysis strategy has two parts. First, we estimate a linear regression; then, we use a Lee (2009) bounds estimator to identify the bounds of the average treatment effects from owner race, ethnicity, and gender on the loan amount awarded. At the micro level, we conducted structural interviews with PPP recipients in Northeast Ohio. This region was chosen because COVID-19 heavily impacted its small businesses, and 40% of all layoffs and unemployment claims in Ohio have been attributed to it (Demko et al., 2020; Lendel, 2020). The micro-level analysis allowed for a qualitative and in-depth analysis of the PPP recipients’ experience with the PPP loans.

Macro-Level Analysis

Models 1 and 2—Testing Primary Hypotheses

The following model is used to test the three primary hypotheses:

\begin{aligned} \log (P P P l o a n a m o u n t)_{i} = & α + β_{0} X_{i} + β_{1} M_{i} + β_{2} H_{i} + β_{3} G_{i} \\ + β_{4} M_{i} * Y_{i} + β_{5} H_{i} * Y_{i} + β_{6} G_{i} \\ * Y_{i} + δ Z_{i} + θ W_{i} + τ_{i} + μ_{i} + l_{i} + ε_{i} \end{aligned}

(1)

where

X_{i}

is a matrix of small business and PPP characteristics, such as company size, nonprofit status, and days to loan approval.

M_{i}

H_{i}

and

G_{i}

are the treatment variables.

M_{i}

is a categorical variable based on the race of the business owner information provided: White-owned, minority-owned (or non-White), and unanswered business owner race.

H_{i}

is a categorical variable based on the ethnicity of the business owner: non-Hispanic business owners, Hispanic business owners, and unanswered business owner ethnicity.

G_{i}

is a categorical variable for the gender of the business owner: male-owned, female-owned, and unanswered business owner gender.

Y_{i}

is a categorical variable representing company size measured by the number of employees per PPP loan recipient interacted with race, ethnicity, and gender of the business owner. When not interacted, company size is a vector of

X_{i} . Z_{i}

is a matrix of lender characteristics, such as lender type and total lender assets.

Model (1) also includes a matrix of county-level characteristics: $W_{i}$ , such as rural status, loan demand, low- to moderate-income (LMI) status, and days in lockdown. We use the information on CRA loans to measure small business borrowing demand before COVID-19. CRA has an important role in lending to small businesses (Ding et al., 2018), especially those in low-income areas where businesses may be less resilient to shocks, such as the closure and reduction in demand brought on by the pandemic. Additionally, $τ_{i}$ controls for whether the loan was granted in the first round (April 3, 2020–April 16, 2020) or the second round (April 27, 2020–August 8, 2020). $μ_{i}$ is a control for state fixed effects and $l_{i}$ controls for the different industry sectors (21 different sectors in total). $μ_{i}$ and $l_{i}$ control for unobserved factors in a given state and industry sector, which may have affected the PPP loan amount granted.

We also run a regression where the dependent variable is the approved loan amount divided by the number of employees:

\log (P P P l o a n a m o u n t p e r e m p l o y e e)_{i} = α + β_{0} X_{i} + β_{1} M_{i} + β_{2} H_{i} + β_{3} G_{i} + δ Z_{i} + θ W_{i} + τ_{i} + μ_{i} + l_{i} + ε_{i}

(2)

In Model (2), matrix

X_{i}

includes nonprofit status and days to loan approval.

In our study, not all PPP recipients reported their race, ethnicity, and/or gender. The concurrent literature addressed the data incompleteness by measuring minority share of the population by race at the ZIP code level for employer businesses and at the county level for nonemployer businesses (Fairlie & Fossen, 2022), employing a Mills inverse ratio (Atkins et al., 2022a), using the cuisine type of the restaurants as a proxy for the race and ethnicity information of the owner (Fei & Yang, 2021), using a machine learning approach (Howell et al., 2021), and utilizing voter registration data (Chernenko & Scharfstein, 2022).

To test the primary hypothesis, we use a linear regression model with controls for the unanswered demographic information in the form of dummy variables (Unanswered Owner Race, Unanswered Owner Gender, Unanswered Owner Ethnicity), which allows us to conduct the analysis on the full set of PPP loans. For example, when analyzing PPP loan amounts based on race, we treat Unanswered Owner Race as a separate category and compare loan sizes between White and non-White business owners, and business owners with unanswered race.

Lee Bounds Estimator—Testing the Secondary Hypotheses

We use the Lee bounds estimator to account for selection into self-reporting of race, ethnicity, and/or gender when testing the secondary hypotheses. The Lee bounds estimator corrects for sample selection bias using a semiparametric approach, which requires less assumptions (Lee, 2009; Tauchmann, 2014).⁸ We analyze how being in rural areas versus being in urban areas affects PPP amount per employee.

The procedure first identifies the excess number of small businesses induced to be selected because of the treatment. In this case the treatment is whether they are in rural or urban areas and provided their race, gender, or ethnicity. It then trims the upper and lower tails of the outcome (e.g., PPP loan amount per employee) distribution by this number, yielding scenario bounds. For example, for the secondary hypothesis that female small business owners in rural counties received less than female small business owners in urban counties, the selection includes those that answered female and the treatment is their location (rural or urban).

The shares of observations with observed outcomes in the treatment group, $q_{T}$ , and its control group, $q_{c}$ , can be written as (Tauchmann, 2014):

q_{T} \equiv \frac{\sum_{i} I (T_{i} = 1, S_{i} = 1)}{\sum_{i} I (T_{i} = 1)} and

(3)

q_{C} \equiv \frac{\sum_{i} I (T_{i} = 0, S_{i} = 1)}{\sum_{i} I (T_{i} = 0)}

(4)

where

I (.)

denotes the indicator function;

Y_{i}

is the outcome (i.e., PPP loan amount per employee);

T_{i}

is the binary treatment indicator (i.e., if PPP loan recipient is located in a rural county);

S_{i}

is the binary selection indicator (i.e., if PPP loan recipient self-identified their race, gender, or ethnicity). When the treatment group is larger than the control group (

q_{C} < q_{T}

q = \frac{(q_{T} - q_{C})}{q_{T}}

and

1 - q

marks the quantiles where the control outcome distribution is trimmed (Tauchmann, 2014). This procedure excludes extreme values. As such, the marginal values of the outcome that are considered in the trimmed means are (

y_{q}^{T}

y_{1 - q}^{T}

y_{q}^{T} = G_{Y | T = 1, S = 1}^{- 1} (q)

(5)

y_{1 - q}^{T} = G_{Y | T = 1, S = 1}^{- 1} (1 - q)

(6)

where

G_{Y}^{- 1}

denotes the inverse distribution of the outcome function. We now have all elements to calculate estimates for the upper and lower bounds (Tauchmann, 2014):

{\hat{θ}}^{u p p e r} = \frac{\sum_{i} 1 (T_{i} = 1, S_{i} = 1, Y_{i} \geq y_{q}^{T}) Y_{i}}{\sum_{i} 1 (T_{i} = 1, S_{i} = 1, Y_{i} \geq y_{q}^{T})} - \frac{\sum_{i} 1 (T_{i} = 0, S_{i} = 1) Y_{i}}{\sum_{i} 1 (T_{i} = 0, S_{i} = 1)}

(7)

{\hat{θ}}^{l o w e r} = \frac{\sum_{i} 1 (T_{i} = 1, S_{i} = 1, Y_{i} \geq y_{1 - q}^{T}) Y_{i}}{\sum_{i} 1 (T_{i} = 1, S_{i} = 1, Y_{i} \geq y_{1 - q}^{T})} - \frac{\sum_{i} 1 (T_{i} = 0, S_{i} = 1) Y_{i}}{\sum_{i} 1 (T_{i} = 0, S_{i} = 1)}

(8)

Information on how the analytical standard errors are calculated can be found in Lee (2009). The Lee bounds estimator allows us to identify whether the effect is different from zero. In cases where the bounds do not include zero, we can conclude whether the treatment on treated was positive or negative, which allows us to test the secondary hypotheses.

Results of Macro-Level Analysis

Table 3 presents the results for 2020, 2021, and both years (2020 and 2021) from Model (1) and Table 4 from Model (2). The models control for exogenous factors such as loan-, lender-, county-, and state-specific factors that may explain the differences in the PPP loan amounts awarded. Other factors for which models’ control are business owner race, ethnicity, and gender. Standard errors are clustered at the county level. We use state and industry fixed effects.

Table 3.

PPP Loan Amount Results from Model 1 (Continues on Next Page).

	Log PPP loans	Log PPP loans	Log PPP loans	Log PPP loans
	2020	2020	2021	2020/2021
Minority-owned	0.042**	0.046**	0.207***	0.060***
	(0.0136)	(0.014)	(0.0090)	(0.0174)
Minority-owned 5 to 19 Employees	−0.233***	−0.278***	−0.437***	−0.296***
	(0.0172)	(0.0189)	(0.0136)	(0.0221)
Minority-owned 20 to 499 Employees	−0.370***	−0.306***	−0.542***	−0.330***
	(0.0225)	(0.0224)	(0.0184)	(0.0241)
Female-owned	−0.169***	−0.172***	−0.132***	−0.175***
	(0.00616)	(0.00617)	(0.0085)	(0.0078)
Female-owned 5 to 19 Employees	0.001	−0.003	−0.071***	−0.008
	(0.0052)	(0.0078)	(0.0086)	(0.0072)
Female-owned 20 to 499 Employees	−0.052***	−0.069***	−0.090***	−0.002
	(0.0075)	(0.0128)	(0.0097)	(0.0084)
Hispanic-owned	−0.053***	−0.060***	−0.053*	−0.069***
	(0.0095)	(0.0098)	(0.0322)	(0.0146)
Hispanic-owned 5 to 19 Employees	−0.021	−0.057***	−0.108***	−0.078***
	(0.0175)	(0.0167)	(0.0170)	(0.0172)
Hispanic-owned 20 to 499 Employees	−0.083***	−0.067***	−0.098***	−0.063**
	(0.0202)	(0.0179)	(0.0283)	(0.0196)
Rural county	−0.183***	−0.168***	−0.137***	−0.166***
	(0.0077)	(0.0073)	(0.0087)	(0.0075)
Log loan demand	−0.005	−0.003	0.001	0.002
	(0.0035)	(0.0034)	(0.0035)	(0.0031)
Low- to moderate-income	0.015*	0.015	−0.024***	0.016*
	(0.0075)	(0.0076)	(0.0091)	(0.0074)
Log lender assets	0.007***	−0.003		−0.002
	(0.0014)	(0.0016)		(0.0013)
Commercial bank	0.208***	0.170***	0.156***	0.173***
	(0.0100)	(0.0099)	(0.0082)	(0.0074)
Farm Credit Association	0.463***	0.401***	0.383***	0.390***
	(0.0150)	(0.0144)	(0.0142)	(0.0138)
Nontraditional lender	0.145***	0.173***	0.170***	0.101***
	(0.0135)	(0.0132)	(0.0130)	(0.0133)
Second PPP round	−0.344***
	(0.0066)
Days to approval	−0.003***	−0.006***	−0.001	−0.004***
	(0.0002)	(0.0002)	(0.000069)	(0.000099)
Lockdown days	0.034
	(0.0214)
Nonprofit	0.216***	0.180***	0.168***	0.195***
	(0.0111)	(0.0119)	(0.0124)	(0.0105)
Unanswered owner race	0.005	0.009	0.119***	0.059***
	(0.0062)	(0.0064)	(0.0089)	(0.0055)
Unanswered owner gender	−0.116***	−0.122***	−0.115***	−0.154***
	(0.0041)	(0.0043)	(0.0044)	(0.0041)
Unanswered owner Hispanic	−0.015***	−0.018***	−0.040***	−0.011***
	(0.0038)	(0.0038)	(0.0040)	(0.0032)
5 to 19 Employees	1.388***	1.448***	1.656***	1.525***
	(0.0052)	(0.0062)	(0.0098)	(0.0066)
20 to 499 Employees	2.793***	2.571***	3.096***	2.671***
	(0.0114)	(0.0102)	(0.0144)	(0.0102)
Unanswered company size	4.745***	3.584***	5.343***	3.679***
	(0.0927)	(0.0176)	(0.1967)	(0.0148)
Minority-owned company size unanswered	−0.831***	−0.595***	−0.720	−0.544***
	(0.1882)	(0.0379)	(0.5246)	(0.0372)
Unanswered race-owned 5 to 19 employees	−0.039***	−0.043***	−0.097***	−0.078***
	(0.0056)	(0.0057)	(0.0078)	(0.0059)
Unanswered race-owned 20 to 499 employees	−0.056***	−0.046***	−0.173***	−0.105***
	(0.0079)	(0.0073)	(0.0105)	(0.0061)
Unanswered race-owned company size unanswered	0.034	−0.048***	−0.561	−0.130***
	(0.0929)	(0.0113)	(0.3567)	(0.0096)
Female-owned company size unanswered	−0.694***	−0.069***	−0.267	−0.0351**
	(0.1507)	(0.0128)	(0.5617)	(0.01297)
Unanswered gender-owned 5 to 19 employees	0.0522***	0.042***	0.005	0.061***
	(0.0038)	(0.0039)	(0.0061)	(0.0039)
Unanswered gender-owned 20 to 499 employees	0.150***	0.092***	0.148***	0.123***
	(0.0082)	(0.0060)	(0.0010)	(0.00597)
Unanswered gender-owned company size unanswered	−0.102	0.138***	0.429*	0.193***
	(0.0848)	(0.0111)	(0.2485)	(0.0104)
Hispanic-owned company size unanswered	−0.470*	−0.229***	−0.516	−0.171***
	(0.2059)	(0.0298)	(0.8783)	(0.02940)
Unanswered Hispanic-owned 5 to 19 employees	0.022***	0.011*	0.011**	−0.003
	(0.0045)	(0.0044)	(0.0052)	(0.0040)
Unanswered Hispanic-owned 20 to 499 employees	0.111***	0.049***	0.88***	0.043***
	(0.0070)	(0.0062)	(0.0078)	(0.0054)
Unanswered Hispanic-owned company size unanswered	0.149	0.122***	0.251	0.115***
	(0.1089)	(0.0101)	(0.4128)	(0.0099)
Constant	9.516***	9.741***	9.285***	9.524***
	(0.1530)	(0.0623)	(0.0553)	(0.0539)
State fixed effects	Yes	Yes	Yes	Yes
Industry fixed effects	Yes	Yes	Yes	Yes
Year fixed effects				Yes
Observations	4,667,834	4,663,570	6,114,938	8,134,794
R-squared	0.66	0.67	0.52	0.65
Clusters (county level)	3,138	3,138	3,125	3,125

Standard errors in parentheses. Not all 2021 lender assets were available for nontraditional lenders reason why it is not controlled for in 2021.

Source: Authors’ calculations based on SBA, FDIC, FCA, NCUA, FFIEC, Mergent Online, 990 Internal Revenue Service files or internal audits, Wu et al. (2020).

* p < 0.1, ** p < 0.05, *** p < 0.01.

Table 4.

PPP Loan Amount per Employee Results from Model 2.

	Log PPP per employee	Log PPP per employee	Log PPP per employee
	2020	2021	2020/2021
Minority-owned	−0.107***	0.152***	−0.035**
	(0.0092)	(0.0110)	(0.0173)
Female-owned	−0.121***	−0.088***	−0.117***
	(0.0044)	(0.0077)	(0.0066)
Hispanic-owned	−0.079***	−0.039	−0.072***
	(0.0060)	(0.0341)	(0.0110)
Rural county	−0.065***	−0.018**	−0.029***
	(0.0065)	(0.0077)	(0.0056)
Log loan demand	0.009	0.018***	0.012**
	(0.0054)	(0.0070)	(0.0046)
Low- to moderate-income	0.061***	−0.022**	0.040***
	(0.0064)	(0.0111)	(0.0073)
Log lender assets	−0.012***		−0.017***
	(0.0013)		(0.0013)
Commercial bank	0.021***	0.080***	0.067***
	(0.0074)	(0.0081)	(0.0055)
Farm Credit Association	0.260**	0.284***	0.189***
	(0.0160)	(0.0116)	(0.0107)
Nontraditional lender	0.068***	0.288***	0.086***
	(0.0111)	(0.0122)	(0.0109)
Second PPP round	−0.203***
	(0.0065)
Days to approval	0.0003	0.002***	−0.0001
	(0.0002)	(0.00007)	(0.00011)
Days in lockdown	0.050
	(0.0379)
Nonprofit	−0.031***	−0.159***	−0.048***
	(0.0087)	(0.0130)	(0.0094)
Unanswered owner race	−0.040***	0.041***	−0.021***
	(0.0053)	(0.0077)	(0.0047)
Unanswered owner gender	−0.029***	−0.044***	−0.038***
	(0.0030)	(0.0035)	(0.0033)
Unanswered owner Hispanic	0.007***	−0.021***	−0.001
	(0.0020)	(0.0032)	(0.0019)
Constant	8.801***	8.648***	9.003***
	(0.2658)	(0.1071)	(0.0742)
Year fixed effects			Yes
Industry	Yes	Yes	Yes
State	Yes	Yes	Yes
County	Yes	Yes	Yes
Observations	4,667,658	6,114,923	8,139,340
R-squared	0.14	0.12	0.12
Cluster (county)	3,129	3,110	3,132

Standard errors in parentheses. Not all 2021 lender assets were available for nontraditional lenders’ reason why it is not controlled for in 2021.

Source: Authors’ calculations based on SBA, FDIC, FCA, NCUA, FFIEC, Mergent Online, 990 Internal Revenue Service files or internal audits, Wu et al. (2020).

*p < 0.1, **p < 0.05, *** p < 0.01.

Goodness-of-fit is indicated by R-squared, which ranges from 0.52 to 0.67 for Model (1) and 0.12 to 0.14 for Model (2). We will discuss the results for both models separately. We also will discuss the results from the Lee bounds estimations that test the secondary hypotheses (Tables 5–7).

Table 5.

Lee Bounds Estimates for Female Business Owners in Rural Areas versus in Urban Areas in 2020 and 2021.

	PPP loan amount per employee
	2020	2021	2020	2021
Female
Lower	−$3,897***	−$6,390***	−$3,919***	−$5,673***
	($31.5)	($32.3)	($31.6)	($33.2)
Upper	−$932***	$348***	−$948***	$460***
	($36.1)	($33.0)	($36.0)	($33.2)
Covariate	No	No	LMI	LMI
Observations	4,716,009	6,154,040	4,716,009	6,154,040
Number of cells			2	2
Selected observations	405,987	1,073,008	405,987	1,073,008
Overall trimming proportion	0.16	0.27	0.16	0.27
Treatment	Rural	Rural	Rural	Rural
Selection	Female	Female	Female	Female
95% confidence interval	[-$3,900, -$870]	[-$6,400, $402]	[-$4,000, -$890]	[-$5,700, $515]

Notes: Standard errors in parentheses. *** p < 0.01.

All monetary values are in 2022 dollars.

The covariate used was the Low- and Moderate-Income Counties (LMI) identifier.

Table 6.

Lee Bounds Estimates for Minority Business Owners in Rural Areas versus Urban Areas in 2020 and 2021.

	PPP loan amount per employee
	2020	2021	2020	2021
Minority
Lower	−$8,366***	−$6,804***	−$8,135***	−$6,570***
	($200.4)	($22.9)	($200.3)	($24.6)
Upper	$4,466***	$6,618***	$4,389***	$5,775***
	($188.5)	($37.9)	($188.5)	($34.6)
Covariate	No	No	LMI	LMI
Observations	4,716,042	6,154,040	4,716,042	6,154,040
Number of cells			2	2
Selected observations	227,232	992,512	227,232	992,512
Overall trimming proportion	0.60	0.55	0.60	0.55
Treatment	Rural	Rural	Rural	Rural
Selection	Non-White	Non-White	Non-White	Non-White
95% confidence interval	[-$8,700, $4,800]	[-$6,842, $6,700]	[-$8,500, $4,700]	[-$6,600, $5,800]

Notes: Standard errors in parentheses. *** p < 0.01.

All monetary values are in 2022 dollars.

The covariate used was the Low- and Moderate-Income Counties (LMI) identifier.

Table 7.

Lee Bounds Estimates for Hispanic Business Owners in Rural Areas versus Urban Areas in 2020 and 2021.

	PPP loan amount per employee
	2020	2021	2020	2021
Hispanic
Lower	−$3,700***	−$9,388***	−$3,582***	−$9,087***
	($94.5)	($68.2)	($95.4)	($69.6)
Upper	$1,186***	$3,137***	$872***	$2,944***
	($88.1)	($58.7)	($93.1)	($60.9)
Covariate	No	No	LMI	LMI
Observations	4,716,042	6,154,040	4,716,042	6,154,040
Number of cells			2	2
Selected observations	112,493	234,793	112,493	234,793
Overall trimming proportion	0.23	0.47	0.23	0.47
Treatment	Rural	Rural	Rural	Rural
Selection	Hispanic	Hispanic	Hispanic	Hispanic
95% confidence interval	[-$3,900, $1,300]	[-$9,500, $3,200]	[-$3,700, $1,000]	[-$9,200, $3,000]

Notes: Standard errors in parentheses. *** p < 0.01

All monetary values are in 2022 dollars.

The covariate used was the Low- and Moderate-Income Counties (LMI) identifier.

PPP Loan Amount Results from Model 1

Our quantitative results point to discrepancies between the PPP amounts received by minority-owned businesses versus White-owned businesses. Similar discrepancies for 2020 have been found in Atkins et al. (2022a). Minority business owners with four employees or less received higher loan amounts, on average, than White-owned businesses: 4.6% ($5,233) higher in 2020, 21% ($9,132) higher in 2021, and 6% ($4,438) higher over both years. However, as the company size increases, minority-owned businesses receive, on average, less in PPP loans. Minority-owned businesses with 5 to 19 employees received a 23% smaller PPP loan than their White-owned business counterparts in 2020 and 2021. In dollar amounts this represented $26,391 less in 2020 and $10,002 less in 2021. Over both years they received 24% ($17,458) less than White-owned businesses with 5 to 19 employees.⁹ Minority-owned businesses with 20 to 499 employees received a 26% ($29,576) smaller PPP loan than their White-owned business counterparts in 2020, 34% ($14,568) less in 2021, and 27% ($19,973) less over both years.

Similar to the case of the owner's race, the owner's ethnicity played an important role in the amount of PPP loans awarded in 2020, 2021, and cumulatively over both years. Hispanic-owned businesses with four employees or less received 6% ($6,825) smaller PPP loans than non-Hispanic-owned businesses of the same size in 2020, 5% ($2,174) less in 2021, and 7% ($5,178) less over both years. For small businesses with 5 to 19 employees, Hispanic owners in 2020 received 12% ($13,309) less than non-Hispanic business owners. In 2021 the difference was 16% ($7,001) less and 15% ($10,874) less over both years. Larger companies’ sizes displayed similar disparities in the PPP loan amounts awarded. Hispanic-owned businesses with 20 to 499 employees received 13% ($14,447) smaller PPP loans than their non-Hispanic business counterparts in 2020, 15% ($6,567) less in 2021, and 13% ($9,764) less over both years.

The small business owner's gender also played a role in the PPP loan amount awarded. On average, female-owned businesses with four employees or less received 17% ($19,338) smaller PPP loans than male-owned businesses of the same size in 2020, 13% ($5,653) less in 2021, and 17.5% ($13,315) less over both years. The difference received when comparing male- and female-owned companies with 5 to 19 employees is only statistically significant in 2021, where female-owned businesses received 20% ($8,828) less. When comparing female- and male-owned companies of 20 to 499 employees, female-owned companies received 24% ($27,415) less in 2020, 22% ($9,654) less in 2021, and 18% ($13,093) less over both years.

In studying racial disparities in the PPP, Chernenko and Scharfstein (2022) found that the location of the business can explain up to 5% of these differences. In our case we find that, on average, small businesses in rural counties received over both years 17% smaller PPP loans than those in urban counties. Atkins et al. (2022a) also found that businesses in rural ZIP codes received less PPP loan dollars in 2020. In 2021, the discrepancy between the size of PPP loans in rural and urban counties decreased. Love and Powe (2020) showed that rural small businesses require coordinated relief to weather the crisis due to barriers in capital access and broadband connectivity. There is a need to connect underbanked small business owners in rural areas with capital and capacity-building resources.

Counties with higher median income levels are positively related to the loan amount awarded in 2020, but negatively related in 2021 and when both years are considered (Table 3). The 2020 result aligns with Schweitzer and Borawski (2021) findings showing that PPP loans had a broad reach to low- and middle-income communities served by the CRA loans. In our case, however, we do not find loan demand, as proxied by CRA lending, to have a statistically significant effect.

Results also show that larger lenders, those with a higher value of assets, distributed larger amounts of PPP loans in 2020. When we consider both years, the effect is statistically insignificant. Higher PPP loan amount approvals are associated with loan applications through commercial banks and farm credit institutions when compared to those provided by credit unions. This finding corresponds with our interview findings, that larger banks sought larger companies and thus received loan applications for larger dollar amounts. In 2020, loans approved in the second round (April 27, 2020-August 8, 2020) were 34% smaller than those in the first round (April 3, 2020-April 16, 2020). This result may be associated with the fact that a greater number of smaller businesses applied later in the program, while larger companies applied early on.

Taken together, our results show discrepancies in the PPP loan amounts awarded to business owners that are female, non-White, and/or Hispanic. However, there are indications that the size of the discrepancy, in monetary terms, decreased from 2020 to 2021. In Model (1), firm size categories are used as exogenous variables; however, there may be concerns over potential endogeneity. Therefore, in Model (2), instead of using firm size as exogenous, we transform the dependent variable to PPP loan per employee.

PPP Loan Amount per Employee Results from Model 2

In Model (2), the dependent variable is the approved loan amount divided by the number of employees (Table 4). Most coefficients of interest remain with similar marginal effect signs as in the results from Model (1). It is important to highlight the increased role of nontraditional lenders in comparison to the other lenders from 2020 to 2021. Fei and Yang (2021) and Howell et al. (2021) also found that minority-owned firms were more likely than other firms to get their PPP loan through fintech lenders than through traditional lenders. Table 4 shows that counties with a high loan demand received slightly higher PPP loans per employee in 2021. Rural small businesses received smaller PPP loans per employee in both years, with a lower magnitude in 2021.

In 2020, minority-owned businesses received, on average, 11% less per employee than their White-owned business counterparts ($1,065 less). In 2021, however, minority-owned businesses received 15% more than White-owned businesses ($2,024 more per employee). Over the course of both years, the discrepancy was 3.5% smaller. Our findings reflect reports that in 2021 PPP loans reached more minority-owned businesses (Fairlie & Fossen, 2022).

Fairlie and Fossen (2022) focused on the raw relationship between PPP loan receipts and minority share of the population without controlling for other factors in 2020 and 2021. They emphasized that more research is needed on the dynamic of PPP receipt over the rounds of funding. Our results show that female and Hispanic business owners did not experience greater access to PPP loans in 2021 as did non-Whites. The PPP evolved as time went on; however, the decline in discrepancies between PPP loan amounts per employee among minority- and White-owned businesses is not as large for Hispanic- and female-owned businesses. Hispanic-owned businesses received 8% ($786) less per employee than non-Hispanic business owners in 2020, 4% less ($519) in 2021, and 7% less ($854) over both years. Female-owned businesses received 12% less ($1,204) per employee compared to male-owned businesses in 2020, 9% less ($1,172) in 2021, and 12% less ($1,387) over both years. In monetary terms, Hispanic-owned businesses and female-owned businesses received smaller PPP loans per employee in 2021 than their business counterparts. As a next step we analyze whether there were discrepancies between the PPP loans per employee awarded to minority, female, and Hispanic business owners according to their location using Lee bounds estimation.

Lee Bounds Results for Secondary Hypotheses

In Table 5, the selection is whether a PPP recipient reported their gender as female; the treatment group is female-owned small businesses located in rural counties and the control group is female-owned small businesses located in urban counties. Using Lee bounds, we find that female-owned businesses in rural counties received a lower PPP loan amount per employee in 2020 than urban counties, between $948 and $3,919 less (Table 5). In 2021, the confidence interval for the treatment effect overlaps the value of zero, meaning that the difference in the PPP amount per employee between female-owned businesses in rural counties and female-owned businesses in urban counties was not statistically different from zero.

Using results in Tables 6 and 7, we reject the hypotheses that: (1) minority-owned small businesses located in rural counties received a lower loan amount per employee than minority-owned small businesses in urban counties and (2) Hispanic-owned small businesses located in rural counties received a lower loan amount per employee than Hispanic-owned small businesses in urban counties in both years of PPP distribution. These results indicate that overall location did not cause further discrepancies in PPP loans per employee within minority and Hispanic owners. Smaller disparities in 2021 may be associated with the entry of nontraditional lenders. Atkins et al. (2022b) reported smaller differences in PPP loans between White and Black business owners as bank competition increases and fintech lenders are allowed to receive PPP loan applications.

Robustness Checks

We run a series of robustness checks to control for unobserved effects due to the month of PPP loan approval, ZIP code level effects, state effects, and the interaction of these. To control for the average demographic characteristics of business owners in a certain location, we run county and ZIP code fixed effects (see online Appendix Table A3).¹⁰ We also control for underlying factors that may be affecting the entire population in a state at a given time by interacting state and month. The statistical significance of the main coefficients, their magnitude, and signs are robust to these variations. We also run regressions with less exogenous variables and with year, state, industry, and county fixed effects (see online Appendix Table A4). The coefficients are larger than when other controls are added, leading us to think that they are biased because of the omitted variables. A model with more variables has a better fit and can control for possible sources of discrepancies in loan amounts not fully captured by only using the fixed effects.

Data Limitations

As mentioned in the data section, answering the race, ethnicity, or gender portions on the PPP loan application was optional. As such, substantial portions were left unanswered. Instead of removing the Unanswered observations, we control for them. However, we are unable to provide insights into the Unanswered data because we do not observe them. Such limitation could mean that our results may be under or overestimating the true values. A reason why we may be observing differences in the PPP loan amounts may be due to other factors, such as owners’ lack of knowledge about their eligibility. We attempted to identify these factors in the interviews we conducted. Findings from the interviews are discussed in the qualitative results section of the paper.

Micro-Level Analysis—Structural Interviews

We performed 19 structural interviews with a variety of small businesses in Northeast Ohio (NEO) to provide a qualitative assessment of businesses’ experiences applying for and receiving PPP loan funds in 2020. The sample consisted of businesses in NEO, an 18-county area containing approximately 36% of Ohio's population, employment, and gross domestic product (GDP). Poverty and low income are persistent issues in NEO that contribute to economic distress, as the region lost high-paying manufacturing jobs and low-wage service-sector jobs took their stead. Adding to this, the pandemic forced many businesses to close. In 2020, nearly 40% of previously employed NEO residents filed for unemployment (Figure 1). The pandemic has heavily affected small businesses that supported nearly 80% of all jobs in NEO (Demko et al., 2021).

Figure 1

Unemployment Filings in Northeast Ohio in 2020.

At the time the interviews were conducted, SBA data did not provide business names for loans of $150,000 or less. We reached this group through local communities’ Facebook pages, personal contacts, and publicly available USA Spending data on PPP. Companies that received more than $150,000 in loan funding were identified using SBA data for business names and locations and the Mergent Intellect Database for corresponding contact information. The response rate was the same for each outreach format.

From March through May 2021, a total of 316 e-mails were sent inviting businesses to participate in the interview, with a follow-up reminder about 1 week after the first e-mail. There was a 6% positive response rate from the e-mails. Respondents self-selected to participate and did not receive any direct benefits from the interview. The interviews were held using Zoom or telephone and ranged from 15 to 50 min, averaging around 30 min in length.

Online Appendix B lists the 18 questions asked of each participant (Approved IRB-FY2021-147). By design, the interview questions prompted a discussion about small businesses’ access to PPP loans. The first question asks if a business had any of the following certifications: minority-owned, female-owned, or veteran-owned. The rest of the interview questions focus on the pandemic impact on business and experience with the PPP loan. This allowed us to analyze responses related to the borrowers’ demographic characteristics and allowed the respondents to provide answers about their experience with the program without any influence. Gender and racial discrimination are sensitive topics. Mentioning discrimination in any form could provide context for the answers that is not related to credit access. To ensure unbiased responses, the interview invitations did not disclose the primary and secondary hypotheses of this research.

Small businesses in the sample represented various industries of the economy such as professional services, wholesale trade, transportation, information services, construction, administrative services, and manufacturing. In terms of legal form, participants included 12 LLCs, four S Corporations, two C Corporations, and one nonprofit. The largest business had 240 employees and two participants were sole proprietors. The PPP loan amounts that respondents received ranged from about $20,000 to over $2 million. Ten businesses received $150,000 or less in PPP loan funding, whereas nine businesses received over $150,000. Only 10 PPP recipients included payroll in their loan application, while nine added rent, utilities, and personal protective equipment expenses. The sample consisted of 11 female-owned and 8 male-owned businesses, 3 of which are minority-owned and 16 are White-owned. Themes emerged based on the size of the PPP loan received rather than industry, business owner characteristics, or the like. The interview findings are thus grouped according to loan size received.

Interview Findings

In general, the companies interviewed did not report any issues with PPP as it related to gender or race. In the sample of nine large PPP loan recipients, six businesses had an official minority- or women-owned business certification. In the sample of 10 small PPP loan recipients, three were women-owned and one was women- and minority-owned. However, none of the four had formal certification. These findings suggest that large PPP loan recipients are more likely to have an official minority- or women-owned certification–possibly because of a greater awareness of its potential benefits. Two of the businesses that received larger PPP loan amounts specified that certification helps them at the regional level, for example, with bidding on certain jobs.

It is important to highlight that the sample is not representative of all experiences but provides a case study of small businesses in Northeast Ohio. In diving deeper into the responses, strong differences emerged between businesses that received $150,000 or less in PPP funds and those that received more than $150,000. In general, businesses that received smaller loan amounts reported more difficulty with the loan application process, filed more applications, and pursued more unique funding sources such as fintech lenders.

Some universal experiences emerged across most of the businesses interviewed. Except for one interviewee, businesses received the full loan amount requested. Only two remembered that the application asked loan recipients to buy American-made products when possible. Almost every respondent expressed frustration with the application process and confusion concerning its demands. At the same time, most interviewees indicated that the PPP loans helped them to retain employees and all respondents conveyed appreciation for this form of government assistance.

Despite receiving the PPP loan, a wide variety of strategies were implemented through the pandemic beyond layoffs, including partial and complete furloughs, reduced hours, and reduced salaries and wages for both employees and management. Several business owners did not take a salary for at least one quarter of 2020, whereas some larger businesses instituted management salary cuts of up to 25%. One business discussed its strategy of approaching employees later in their career and asking them to take early retirement; approximately 20 of its 240 employees were either terminated or took the early retirement option. Other businesses also had employees who retired due to the increased health risks posed by the pandemic.

Interview Findings for Small PPP Loans Recipients ($150,000 or Less)

This group included 10 small businesses that received $150,000 or less in PPP loan funding. The smaller the loan request, the more businesses pursued unique lenders for PPP loans. These businesses repeatedly reported difficulties in applying for the program using commercial banks, even those with whom they had extensive working relationships. This resulted in these businesses having to submit multiple applications with multiple banks, and for several of them to ultimately pursue fintech lenders like Lendio, Ready Capital, and Paypal. Respondents cited the size of their business as the reason for these difficulties. Interviewees in this category included many smaller businesses, some of whom pointed to challenges obtaining attention or responses from lenders due to the program's incentivization structure, which favored larger businesses because banks were paid a percentage of the disbursed loan amounts.

In addition to difficulties finding a lender, many recipients of smaller PPP loans communicated a lack of transparency about PPP amount limits and permitted uses. For example, a few respondents did not realize that rent, mortgage, and utility payments could be included and even expressed their desire that these expenses had been permitted to be part of the calculations. This lack of clarity caused business owners to miss out on the opportunity to request higher PPP loan amounts. One business owner, in particular, stated that while the PPP loan was very helpful, it was not enough and that they wished they had been able to include rent and utilities in the loan amount.

Interview Findings for Large PPP Loans Recipients (Over $150,000)

Nine of the businesses we interviewed received loan amounts over $150,000, with the highest loan being over $2 million. These businesses varied widely in size, from employee counts of 10 to 240. All businesses received funding through a commercial bank; none had to pursue an alternative funding source. Nevertheless, many of these respondents felt frustrated with their bank, the application process, and the need to submit multiple applications. One business aptly explained the lack of communication: “The funny thing is, it seems the rules were so vague that the amount of money you were eligible for depended on what person you got on the phone with and what their interpretation and understanding of the rules were.” However, overall, their experience was more streamlined when compared with the smaller loan recipients. As one business owner stated, “It became just a matter of where we fell in line.”

These businesses also reported getting approved for funds more quickly—within a few days to 3 weeks. One business owner shared that they were “pretty much instantly approved; it only took a day or two both times.” Another echoed this, stating, “Approval was pretty immediate.” On the other hand, businesses receiving smaller loan amounts reported that approval and receipt of funds took from 10 days to 6 weeks. Larger businesses did report spending several more hours on their applications, given the complexities of having a larger business with greater employee counts. Some businesses reported spending “up to 20 or 30 hours” on their application, with others saying it took several days to gather the requisite information. On the other hand, smaller businesses said the application took as few as 8 minutes to complete, echoing the expectations set by SBA.¹¹ Some businesses stated that lenders could have been incentivized to work with businesses of all sizes.

Our qualitative results showed that businesses received unequal treatment from PPP lenders based on their size, not the demographics of their ownership. Information frictions in determining access to PPP resources have been more binding for smaller than for larger businesses (Humphries et al., 2020). These findings have important implications for minorities because previous literature found that minority-owned firms are substantially smaller than White-owned firms (Fairlie & Robb, 2004; Perry & Romer, 2020). Similarly, when comparing male and female business owners, research has found that women operate smaller businesses in terms of annual turnover and employment size (Carter & Shaw, 2006). Literature emphasizes an overall hesitancy by female and minority business owners to ask for credit. This hesitancy is linked to a lack of experience, as well as previous negative experiences with lenders.

Policy Implications and Conclusions

Of all small businesses, minority-owned have proven to be most vulnerable. Even before the pandemic, these businesses showed signs of limited financial health in terms of profitability and credit scores, and they are more likely to be concentrated in industries most immediately affected by the pandemic (Dua et al., 2020). Policy makers need to account for the preexisting inequalities in the current systems that administer relief programs. Otherwise, the enacted policies and programs will likely reproduce those same inequities (Atkins et al., 2022a). Greater access to more affordable loans may facilitate economic recovery, enable minority- and female-owned firms to improve their credit scores, and grow business. In addition, expanded access to borrowing opportunities may encourage women to enter into small business ownership at higher rates, which is important because the pandemic has hit women harder than men (Stevenson, 2020). In February 2020, women's labor force participation rate was 57%, the lowest it has been since 1988 (Ewing-Nelson, 2021).

Our quantitative results confirmed hypotheses about discrepancies between the PPP loan amounts received by minority-, Hispanic-, and female-owned businesses during 2020. The discrepancy in PPP loans to women- and minority-owned businesses may have stemmed from a lack of access and knowledge about the program itself. From 2020 to 2021, disparities among non-White owners decreased, however female- and Hispanic- owners continued to receive less in PPP loans than male and non-Hispanic owners. Less disparity in PPP loans in 2021 is likely due to increased information transparency and the inclusion of nontraditional lenders. Nevertheless, female and Hispanic owners did not experience the greater access to PPP loans that non-Whites experienced.

As such, female- and minority-owned businesses tend to fall under the case of businesses that may have applied for smaller PPP loans. Our qualitative findings suggest that smaller businesses faced difficulties in identifying lenders as well as knowing whether they were eligible for the program and the amount they could request. Similar to Humphries et al. (2020), we found that the smallest businesses had the least awareness of government assistance programs, including PPP, relative to larger ﬁrms. Small loan recipients in our sample did not have formal women- or minority-owned business certification. Interview respondents confirmed that certifying a business is time consuming. However, having a certain type of business certification helps get business contracts. Thus, it is important to increase awareness of smaller women- and minority-owned businesses about the benefits of official certifications; namely, capacity-related trainings, mentor programs, networking activities, access to consulting services, and business financing assistance (Schirmer, 2013). Minority- and female-owned small businesses may have benefited from free technical assistance with the PPP application, as they also tend to have fewer resources to pay business consultant fees. The effective policy requires that its rules and regulations be clear and transparent and that they reach eligible applicants through various business networks and associations. For example, the smaller PPP recipients we interviewed mentioned using local Facebook groups and Twitter to navigate the application process and crowdsource finding lenders.

Similar to Bartik et al. (2020), our results identify the important role of borrower-lender relationships when applying for the PPP loan. Although PPP loans were intended for small businesses of all sizes, responses from the interviews suggest that larger businesses were preferred by larger lenders, such as commercial banks. Nonbank lenders were often sought as the lenders of last resort. Businesses with loans of $150,000 or less also recounted having to submit multiple applications with multiple banks, with many of them ultimately having success at smaller, local banks rather than larger multistate banks. This means that policy makers may want to pay close attention to bank mergers and the exits of local banks from underserved communities. Policy makers may also want to add further incentives to lenders that reach out to smaller businesses of diverse ownership and size, such as banks that have high participation in CRA lending.

If the goal of a policy is to target ethnic disparities and support inclusion in federal aid for entrepreneurs, then it cannot be one size fits all (Ajilore & Willingham, 2019; Carpenter & Loveridge, 2018; Hamilton, 2020). Chetty et al. (2020) raised an important concern that the loans were taken by firms that did not intend to lay off many employees to begin with. Santellano (2021) documented challenges of Hispanic business owners in accessing PPP funds during the early stages of the program, July 2020. We performed our interviews from March through May 2021, focusing on experience with the PPP loans in 2020. More qualitative analysis on small businesses’ access to PPP loans in 2021 and their experience with loan forgiveness may provide further insights.

Footnotes

Acknowledgments

We would like to thank Isabella McKnight, Georgina Figueroa, Anne Boyd, and David H. Nason for their excellent research assistance. We would also like to thank Dr. Iryna Lendel, Dr. Merissa Piazza, Matt Ellerbrock, Professor Ted Jaenicke, Professor Doug Wrenn, Professor Chyi Lyi (Kathleen) Liang, participants of the Northeastern Agricultural and Resource Economics Association (NAREA) informal meeting on March 26, 2021, Professor Solomiya Shpak, and participants of the Kyiv School of Economics Academic Seminar on March 11, 2021, for their helpful comments. All errors are our own.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was prepared, in part, using federal funds under award ED16CHI3030036 from the U.S. Economic Development Administration, U.S. Department of Commerce. The statements, findings, conclusions, and recommendations are those of the author(s) and do not necessarily reflect the views of the U.S. Economic Development Administration or the U.S. Department of Commerce.

ORCID iD

Iryna Demko

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biography

Dr. Iryna Demko is a lead data analyst at Progressive Insurance. This work was performed while Iryna was a research associate in the Center for Economic Development at Cleveland State University. Dr. Demko testified to the U.S. Congress at a hearing about the PPP with a summary of this research. She has her Ph.D. degree in agricultural, environmental, and regional economics from The Pennsylvania State University.

Ana Claudia Sant’Anna is an assistant professor in the Division of Resource Economics and Management at West Virginia University. Her research interests focus on strategic decision making within the fields of agricultural production, agribusiness, and ag finance. Ana Claudia earned her Ph.D. from the Department of Agricultural Economics at Kansas State University.

References

Ajilore

Willingham

(2019, July). Redefining rural America [Report]. Center for American Progress. https://www.americanprogress.org/article/redefining-rural-america/.

Asiedu

Freeman

Nti-Addae

(2012). Access to credit by small businesses: How relevant are race, ethnicity, and gender? The American Economic Review, 102(3), 532–537. https://doi.org/10.1257/aer.102.3.532

Atkins

Cook

Seamans

(2022b). Fintechs and racial discrimination in lending. https://ssrn.com/abstract=4034821.

Atkins

Cook

Seamans

(2022a). Discrimination in lending? Evidence from the Paycheck Protection Program. Small Business Economics, 58(2), 843–865. https://doi.org/10.1007/s11187-021-00533-1

Bartik

A. W.

Cullen

Z. B.

Glaeser

E. L.

Luca

Stanton

C. T.

Sunderam

(2020). The targeting and impact of Paycheck Protection Program loans to small businesses (No. w27623). National Bureau of Economic Research.

Bates

Robb

(2013). Greater access to capital is needed to unleash the local economic development potential of minority-owned businesses. Economic Development Quarterly, 27(3), 250–259. https://doi.org/10.1177/0891242413477188

Bates

Robb

(2016). Impacts of owner race and geographic context on access to small business financing. Economic Development Quarterly, 30(2), 159–170. https://doi.org/10.1177/0891242415620484

Becker

(1971). The Economics of Discrimination (2nd ed). University of Chicago Press.

Blanchard

Zhao

Yinger

(2005). Do credit market barriers exist for minority and woman entrepreneurs? Center for Policy Research Paper No. 90, Syracuse University Maxwell School of Citizenship and Public Affairs.

10.

Blanchard

Zhao

Yinger

(2008). Do lenders discriminate against minority and woman entrepreneurs? Journal of Urban Economics, 63(2), 467–497. https://doi.org/10.1016/j.jue.2007.03.001

11.

Blanchflower

Levine

Zimmerman

(2003). Discrimination in the small-business credit market. Review of Economics and Statistics, 85(4), 930–943. https://doi.org/10.1162/003465303772815835

12.

Bostic

R. W.

Lee

(2017). Small business lending under the Community Reinvestment Act. Cityscape (Washington, D C ), 19(2), 63–84. https://www.huduser.gov/portal/periodicals/cityscpe/vol19num2/ch6.pdf

13.

Canedo

Stone

Black

Lukaszewski

(2014). Individual factors affecting entrepreneurship in hispanics. Journal of Managerial Psychology, 29(6), 755–772. https://doi.org/10.1108/JMP-11-2012-0333

14.

Carpenter

C. W.

Loveridge

(2018). Differences between Latino-owned businesses and White-, Black-, or Asian-owned businesses: Evidence from census microdata. Economic Development Quarterly, 32(3), 225–241. https://doi.org/10.1177/0891242418785466

15.

Carter

Allen

(1997). Size determinants of women-owned businesses: Choice or barriers to resources? Entrepreneurship & Regional Development, 9(3), 211–220. https://doi.org/10.1080/08985629700000012

16.

Carter

Van Auken

(2006). Small firm bankruptcy. Journal of Small Business Management, 44(4), 493–512. https://doi.org/10.1111/j.1540-627X.2006.00187.x

17.

Carter

Shaw

(2006). Women’s business ownership: Recent research and policy development. DTI Small Business Service Research Report.

18.

Cavalluzzo

(1998). Market structure and discrimination: The case of small businesses. Journal of Money, Credit, and Banking, 30(4), 771–792. https://doi.org/10.2307/2601128

19.

Chernenko

Scharfstein

D. S.

(2022). Racial disparities in the paycheck protection program (No. w29748). National Bureau of Economic Research.

20.

Chetty

Friedman

Hendren

Stepner

(2020). The economic impacts of COVID 19: Evidence from a new public database built from private sector data. Opportunity Insights.

21.

Correia

(2017). Linear models with high-dimensional fixed effects: An efficient and feasible estimator [Working Paper]. http://scorreia.com/research/hdfe.pdf.

22.

Craig

Jackson

Thomson

(2009). The economic impact of the Small Business Administration’s intervention in the small firm credit market: A review of the research literature. Journal of Small Business Management, 47(2), 221–231. https://doi.org/10.1111/j.1540-627X.2009.00269.x

23.

Demko

Lendel

Figueroa

(2021). Navigating government statistics of small businesses for pandemic assistance. Urban Publications. https://engagedscholarship.csuohio.edu/urban_facpub/1713

24.

Demko

Lendel

Piazza

(2020). From economic slowdown to recession. Urban Publications. https://engagedscholarship.csuohio.edu/urban_facpub/1660

25.

Ding

Lee

Bostic

(2018). Effects of the Community Reinvestment Act (CRA) on small business lending [Working Paper WP 18-27]. Federal Reserve Bank of Philadelphia.

26.

Dua

Mahajan

Millan

Stewart

(2020). COVID-19’s effect on minority-owned small businesses in the United States. McKinsey & Company.

27.

Erel

Liebersohn

(2020). Does Fintech substitute for banks? Evidence from the Paycheck Protection Program [Working Paper No. 27659. National Bureau of Economic Research.

28.

Ewing-Nelson

(2021). Another 275,000 women left the labor force in January. National Women’s Law Center.

29.

Fairlie

R. W.

(2020). The impact of COVID-19 on small business owners: Evidence of early-stage losses from the April 2020 current population survey [Working Paper No. w27309]. National Bureau of Economic Research.

30.

Fairlie

R. W.

Fossen

(2021). 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? [Working Paper No. w28321]. National Bureau of Economic Research.

31.

Fairlie

R. W.

Fossen

(2022). The 2021 Paycheck Protection Program reboot: Loan disbursement to employer and nonemployer businesses in minority communities [Working Paper No. w29732]. National Bureau of Economic Research.

32.

Fairlie

R. W.

Robb

A. M.

(2004). Why are Black-owned businesses less successful than White-owned businesses? The role of families, inheritance, and business human capital, [Discussion Paper No. 1292]. IZA.

33.

Fei

C. Y.

Yang

(2021). Fintech and racial barriers in small business lending. University of North Carolina at Chapel Hill Working Paper.

34.

Galli-Debicella

(2020). The efficacy of SBA loans on small firm survival rates. Journal of Small Business Strategy, 30(2), 26–34. https://libjournals.mtsu.edu/index.php/jsbs/article/view/1590

35.

Giglio

(2020). Access to credit and women entrepreneurs: A systematic literature review. European Research Studies, 23(4), 312–335. https://doi.org/10.35808/ersj/1686

36.

Hamilton

(2020). From survival to revival: How to help small businesses through the COVID-19 crisis. The Hamilton Project.

37.

Howell

S. T.

Kuchler

Snitkof

Stroebel

Wong

(2021). Lender automation and racial disparities in credit access [Working Paper No. w29364]. National Bureau of Economic Research.

38.

Humphries

Neilson

Ulyssea

(2020). Information frictions and access to the Paycheck Protection Program. Journal of Public Economics, 190(October), 1–8. https://doi.org/10.1016/j.jpubeco.2020.104244

39.

Immergluck

(2002). Redlining redux: Black neighborhoods, Black-owned firms, and the regulatory cold shoulder. Urban Affairs Review, 38(1), 22–44. https://doi.org/10.1177/107808702401097781

40.

Lee

D. S.

(2009). Training, wages, and sample selection: Estimating sharp bounds on treatment effects. The Review of Economic Studies, 76(3), 1071–1102. https://doi.org/10.1111/j.1467-937X.2009.00536.x

41.

Lendel

(2020). Covid-19 recession impacts on NEO [Slides deck]. Cleveland State University. 2020 Chicago Regional EDA University Center Showcase. https://urban.csuohio.edu/sites/default/files/2020_CSU_CED_UC_Showcase_Final.pdf.

42.

Liu

Parilla

(2020). New data shows small businesses in communities of color had unequal access to federal COVID-19 relief. Brookings Institution.

43.

Love

Powe

(2020, December 1). Rural small businesses need local solutions to survive. Brookings Institution.

44.

Mora

M. T.

Dávila

(2014). Gender and business outcomes of black and Hispanic new entrepreneurs in the United States. American Economic Review, 104(5), 245–249. https://doi.org/10.1257/aer.104.5.245

45.

Neely

Van Auken

(2012). An examination of small firm bootstrap financing and use of debt. Journal of Developmental Entrepreneurship, 17(1), 1–12. https://doi.org/10.1142/S1084946712500021

46.

Omi

Winant

(2014). Racial formation in the United States (3rd ed.). Routledge. https://doi.org/10.4324/9780203076804

47.

Perry

A.M.

Romer

(2020). To expand the economy, invest in Black business. Brookings Institute.

48.

Phelps

(1972). The statistical theory of racism and sexism. American Economic Review, 62(9), 659–661. https://www-jstor-org-443.web.bisu.edu.cn/stable/1806107

49.

Rowlands

D. W.

Love

(2021, September 28). Mapping rural America’s diversity and demographic change. Brookings Institution.

50.

Santellano

(2021). Compounded inequality: How the US Paycheck Protection Program is failing Los Angeles Latino small businesses. Ethnic and Racial Studies, 44(5), 794–805. https://doi.org/10.1080/01419870.2020.1844265

51.

Scalera

Zazzaro

(2001). Group reputation and persistent (or permanent) discrimination in credit markets. Journal of Multinational Financial Management, 11(4-5), 483–496. https://doi.org/10.1016/S1042-444X(01)00037-8

52.

Schirmer

M. L.

(2013). Minority business certification programs in a Mid-Atlantic state: Participants’ perceptions of capacity building. Wilmington University (Delaware).

53.

Schweitzer

Borawski

(2021, May 28). How well did PPP loans reach low- and moderate-income communities? [Number 2021-13]. Economic Commentary. Federal Reserve Bank of Cleveland, 2021(13), 1–6. https://doi.org/10.26509/frbc-ec-202113

54.

Small Business Administration. (2020, June 26). Paycheck Protection Program. How to calculate maximum loan amounts – by business type . https://www.sba.gov/sites/default/files/2020-12/How-to-Calculate-Loan-Amounts-508_6-26-20.pdf

55.

Small Business Administration (2021, March 9). Fact sheet: changes to Paycheck Protection Program by Biden-Harris administration increase equitable access to relief [Press release].

56.

Small Business Administration. (2020). Paycheck Protection Program (PPP) provided emergency COVID-19 relief to millions of small businesses and their workers. Myth vs Fact .

57.

Stevenson

(2020). The initial impact of COVID-19 on labor market outcomes across groups and the potential for permanent scarring. Hamilton Project Essay, 2020(16), 1–11. https://www.hamiltonproject.org/assets/files/Stevenson_LO_FINAL.pdf

58.

Tauchmann

(2014). Lee (2009) treatment-effect bounds for nonrandom sample selection. The Stata Journal, 14(4), 884–894. https://doi.org/10.1177/1536867X1401400411

59.

U.S. Census Bureau (2019). Annual Business Survey: Employment size of firm statistics for employer firms by sector, sex, ethnicity, race, and veteran status for the U.S., states, and metro areas: 2019. Table ID: AB1900CSA04.

60.

U.S. Federal Reserve Banks (2017). Small Business Credit Survey. Report on minority-owned firms. 10.55350/sbcs-20171109.

61.

U.S. Federal Reserve Banks (2019). Small Business Credit Survey. Report on minority-owned firms. 10.55350/sbcs-20191213.

62.

U.S. Federal Reserve Banks. (2021). Small Business Credit Survey. 2021 Report on firms owned by people of color. 10.55350/sbcs-20210415

63.

Williams

Ortiz-Walters

Gavino

(2012). Entrepreneurial education and Latina business owner preferences: Do gender and race matter? Journal of Women’s Entrepreneurship and Education, 1(2), 74–98. http://ebooks.ien.bg.ac.rs/498/1/12_jwe_1-2_5.pdf

64.

Wooldridge

J. M.

(2013). Introductory econometrics: A modern approach (5th ed). South Western Cengage Learning.

65.

Smith

Khurana

Siemaszko

DeJesus-Banos

(2020, March 25). Coronavirus lockdowns and stay-at-home orders across the U.S. NBC News. https://www.nbcnews.com/health/health-news/here-are-stay-home-orders-across-country-n1168736.

66.

Yunwei

Minniti

(2015). External financing and the survival of black-owned start-ups in the US. Eastern Economics Journal, 41(3), 387–410. https://doi.org/10.1057/eej.2014.23

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB