Nonprofit Organizations’ Financial Obligations and the Paycheck Protection Program — Management Science (2023)
Summary
A Management Science 69(7) article, publication date 2023-07-01, DOI 10.1287/mnsc.2023.4804, Nonprofit Organizations' Financial Obligations and the Paycheck Protection Program, by Daniel G. Neely, Gregory D. Saxton and Paul A. Wong. The abstract states that the authors examine nonprofit organizations' involvement in the Paycheck Protection Program and associations between pre-pandemic financial obligations and PPP participation and loan characteristics. It reports that nonprofits participated at a lower rate than other small business industries, that those with greater financial obligations were more likely to participate, and that approximately 11 percent of nonprofits failed to obtain loan forgiveness.
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
- Nonprofit Organizations Financial Obligations And The Paycheck Protection Program
- Case
- Nonprofit Organizations Financial Obligations And The Paycheck Protection Program
Cite this document
“Nonprofit Organizations’ Financial Obligations and the Paycheck Protection Program — Management Science (2023).”
Original at doi.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/nonprofit-organizations-financial-obligations-and-the-paycheck-protection-program/
, SHA-256 5ac0dc62ef820fc6401a7f26353ef9190e691861dbf913649a03d77f1ce69b27.Full text
UC Davis
UC Davis Previously Published Works
Title
Nonprofit Organizations’ Financial Obligations and the Paycheck Protection Program
Permalink
https://escholarship.org/uc/item/6sj1w6sf
Journal
Management Science, 69(7)
ISSN
0025-1909
Authors
Neely, Daniel G
Saxton, Gregory D
Wong, Paul A
Publication Date
2023-07-01
DOI
10.1287/mnsc.2023.4804
Supplemental Material
https://escholarship.org/uc/item/6sj1w6sf#supplemental
Copyright Information
This work is made available under the terms of a Creative Commons Attribution License,
availalbe at https://creativecommons.org/licenses/by/4.0/
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
Nonprofit Organizations’ Financial Obligations and the Paycheck Protection Program
(Forthcoming in Management Science)
Daniel G. Neely
University of Wisconsin – Milwaukee
neely@uwm.edu
Gregory D. Saxton
York University
gsaxton@yorku.ca
Paul A. Wong
University of California, Davis
panwong@ucdavis.edu
Acknowledgements: We would like to thank Steven Balsam, Brad Barber, Sudipta Basu, Brian
McAllister, Hollis Skaife, seminar participants at Temple University, the University of
California, Davis and the University of Wisconsin-Milwaukee, as well as conference attendees at
the 2022 American Accounting Association’s Government and Nonprofit Section Midyear
Meeting and the 2022 Haskayne and Fox Accounting Conference for their helpful comments and
assistance.
Nonprofit Organizations’ Financial Obligations and the Paycheck Protection Program
Abstract: We examine nonprofit organizations’ involvement in the Paycheck Protection Program (PPP).
The PPP provided participants with forgivable loans to pay employee salaries, increasing participants’
financial flexibility during the pandemic. We examine the associations between nonprofits’ pre-pandemic
financial obligations (e.g., long-term debt and donor-restricted net assets) and PPP participation and
participants’ loan characteristics. First, we find nonprofit organizations participated at a lower rate than
other small business industries and that nonprofits with greater financial obligations were more likely to
participate in the program. Second, we find financial obligations were positively associated with the loan
amount received as a percentage of total payroll costs. Lastly, while approximately 11 percent of
nonprofits failed to obtain loan forgiveness, we find nonprofits with restricted net assets were more likely
to have their loans forgiven. Our results suggest nonprofits with greater debt and donor obligations used
the PPP to increase their financial flexibility.
Keywords: COVID-19; Paycheck Protection Program; nonprofit organizations; financial obligations
2
1. Introduction
The United States federal government created the Paycheck Protection Program (“the program”
or “PPP”) in response to the economic effects of the Coronavirus (“COVID-19”) pandemic. The
program’s objective was to provide small business entities1 with resources to support company payroll
costs and maintain employment, and it funded approximately 11.8 million loans worth nearly $800
billion. These loans were issued at a one percent interest rate but were forgivable and converted into
grants if loan proceeds were primarily used for payroll costs and employees were retained.
Our study focuses on the nonprofit sector – an important component of the U.S. economy,
representing 5.7 percent of GDP, employing 6.4 percent of the workforce, and contributing $1.4 trillion to
the American economy in the first quarter of 2022 (Independent Sector 2022). Moreover, nonprofits
provide valuable services during times of crisis (Maher et al. 2020). Yet, few PPP-related studies and
media articles focus on the nonprofit sector. The existing literature generally surveys nonprofits and asks
questions related to the pandemic and PPP participation. These accounts noted that the program enabled
nonprofits to continue operations (Lyons 2021); however, some found the program’s eligibility criteria,
application process, and loan forgiveness provisions lacked clarity (Finchum-Mason et al. 2020).
Combined with nonprofits’ relative aversion to debt financing (Calabrese 2011; Garcia-Rodriguez et al.
2022), it is thus perhaps not surprising that disproportionately fewer nonprofits participated in the loan
program than for-profit entities (Schweitzer and Borawski 2021). Collectively, the evidence points to
nonprofits’ unique involvement with the PPP.
To help explain nonprofits’ engagement with the PPP, we focus on the role of financial
obligations – specifically long-term debt obligations and donor-restricted net assets – and examine
whether these obligations are associated with program participation and the loan amounts received and
ultimately forgiven. Nonprofits provide a unique setting to examine financial obligations. Some
nonprofits receive donations restricted for a time and/or purpose (e.g., endowment contributions or
1
The Small Business Administration (SBA) noted that entities potentially eligible for PPP funds included for-profit
businesses and nonprofit organizations (U.S. Small Business Administration 2021).
3
donations for specific organization expenses).2 These restrictions imply a future obligation imposed by
donors, constraining future resource usage. Thus, the nonprofit setting allows us to examine resource
obligations from both external financing and donor restrictions.
We posit that the forgivable PPP loans provided financial flexibility and reduced the constraints
of existing financial obligations to participants by temporarily funding payroll costs and allowing
participants to maintain employment and/or utilize resources intended for payroll on other activities.
Further, financial flexibility is a determinant of increased corporate investment (Arslan-Ayaydin et al.
2014); nonprofits with greater financial flexibility, or fewer obligations, have more resources to invest in
their programs. We argue those with existing debt and restricted net assets have greater financial
obligations than other nonprofits, a higher need for financial flexibility, and, therefore, an increased
likelihood to participate in the program, greater loan amount requests, and a higher likelihood of
obtaining loan forgiveness.
Our sample contains 103,886 PPP-eligible nonprofits, including 39,055 program participants. We
identify eligibility based on their number of employees and payroll expenses, inclusive of compensation,
benefits, and retirement costs, from Form 990 disclosures (i.e., less than or equal to 500 employees and
positive payroll expenses). Then, we identify PPP-participating nonprofits using the Small Business
Administration (SBA) PPP loan data. We use the SBA PPP loan data to obtain information about the
participants’ loan characteristics, including loan amount received, the intended use of loan proceeds by
allowable costs, and the loan amount forgiven. We gather financial statement data from the Form 990 and
use Part X to identify nonprofits with debt and restricted net assets.
Descriptively, the participation rate of PPP-eligible nonprofits in our sample was approximately
38 percent, considerably lower than the 70 percent participation rate of small businesses documented in
Schweitzer and Borawski (2021); and the average nonprofit participant received loans equal to 22.3
percent of annual payroll costs and approximately 89 percent of participants obtained loan forgiveness.
2
Donor-restricted contributions are recorded as increases to restricted net assets (equity). Upon fulfillment of the
donor restriction, the net assets are reclassified from restricted net assets to unrestricted net assets.
4
We find financial obligations are positively associated with participation. Specifically, nonprofits with
long-term debt instruments (secured, unsecured, and/or tax-exempt debt) and restricted net assets were
more likely to obtain PPP loans. We find consistent results using an alternative measure of the magnitude
of financial obligations.3 We also examine the relationship with participants’ loan characteristics and find
financial obligations are positively associated with the loan amount as a percentage of total payroll costs
and that restricted net assets and the magnitude of financial obligations are positively associated with loan
forgiveness.
This study makes several contributions. First, we contribute to the growing PPP literature. Prior
research examines the barriers to participation (e.g., Humphries et al. 2020; Birdthistle and Silver 2021;
Beaton et al. 2020), employment outcomes (e.g., Autor et al. 2022; Bartik et al. 2021; Bartlett and Morse
2021; Chetty et al. 2020; Cole 2022; Doniger and Kay 2021; Granja et al. 2022; Humphries et al. 2020),
and the role of loan originators (e.g., Ballew et al. 2021; Erel and Liebersohn 2022; Griffin et al. 2023).
While prior studies focus on the program’s efficacy in maintaining employment levels at for-profit
businesses, our study provides evidence of the nonprofit sector’s involvement in the PPP by examining
the financial characteristics of PPP-eligible nonprofits.
We further contribute through our testing of organizational-level factors. The nascent PPP
literature provides anecdotal and survey evidence of program barriers, aggregate data on the socio-
economic correlates and economic effects of loan disbursements, analyses of lender data, and theoretical
modeling of optimal loan allocations. However, given data constraints on for-profit, private entities, prior
literature has struggled to quantitatively examine the organizational-level drivers of program
participation, characteristics, and outcomes. Leveraging the nonprofit IRS tax form and its filing
requirements allows us to operationalize organizational-level characteristics for, importantly, both
participating and non-participating PPP-eligible nonprofits.
3
Calculated using the sum of total restricted net assets and secured, unsecured, and tax-exempt debt scaled by total
assets.
5
Additionally, we contribute to the literature by exploring the factors associated with obtaining a
new source of funding during periods of considerable economic uncertainty. Previous studies suggest that
the appropriateness of a new revenue source is based in part on the types of services provided by the
nonprofit (Young 2007). At an aggregate level, studies have found that nonprofits that concentrate on
fewer revenue sources generally experience revenue growth (Foster and Fine 2007; Chikoto and Neely
2014). However, this may come at the price of less financial stability (Carroll and Stater 2009). Our study
isolates the determinants of obtaining PPP loans, thereby furthering our understanding of the decision to
obtain a new, temporary revenue source.
At a practical level, our study contributes to the understanding of government loan programs and
the characteristics of participating businesses. The findings are particularly relevant for designing
business-focused economic relief programs and maximizing their intended societal benefit during
economic downturns. Considering nonprofit organizations’ key pandemic role and their relatively low
rate of program participation, nonprofits may have been overlooked as a key participant, thus reducing the
program’s potential impact.
2. Paycheck Protection Program
In response to the economic disruption caused by COVID 19, the United States federal
government enacted the Coronavirus Aid, Relief, and Economic Security (CARES) Act (U.S. Department
of the Treasury 2021). The CARES Act includes the creation of the PPP to provide economic relief to
businesses through forgivable loans designed to incentivize employee retention during the pandemic (U.S.
Small Business Administration 2021). The general purpose of PPP loans was to finance payroll cost,
including employee salaries and benefits. Entities with 500 or fewer employees (including affiliated entity
employees) were eligible to apply for loans up to approximately 25 percent4 of eligible annual payroll
costs, with annual payroll costs of $100,000 per employee allowable (U.S. Small Business Administration
2021). Loan proceeds could not exceed $10 million.
4
Or two and a half months of the organization’s prior-year average monthly payroll costs.
6
Loans were forgiven if employees were retained and loan proceeds used on allowable costs
within 24 weeks of receipt (U.S. Small Business Administration 2021).5 The allowable costs for
forgiveness included payroll, mortgage interest, utilities, rent, and healthcare costs; however, at least 60
percent of proceeds had to be spent on payroll to obtain loan forgiveness. Any unforgiven loans incurred
an interest rate of one percent annually, with loan payments deferred for ten months and payable over five
years (U.S. Small Business Administration 2021).6
3. Data and Methodology
3.1 Sample Description
We obtain PPP loan data from the SBA website as of March 2022, which includes the name,
address, business sector, and demographics of the borrower, the loan approval amount and date, the loan
originator, and the loan amount forgiven. PPP data also includes loan proceed amounts intended for
payroll, interest, utilities, rent, and other allowable costs under the forgiveness provision. We restrict our
sample of PPP loans to those approved within 2020 to ensure all observations were initial loan
distributions.7
We leverage the mandatory nonprofit disclosures in the IRS Form 990 in our analyses. We obtain
the electronically filed Form 990s received by the IRS from 2010 to 2021, which are available on the
Amazon Web Services and IRS servers. We employ a multi-stage, supervised fuzzy matching algorithm
(Bai et al. 2018) based on the borrower’s name, zip code, and address to link nonprofits in the PPP dataset
to nonprofits in the Form 990 data. In a first stage, we identify nonprofits that match perfectly by name
and ZIP code. We then implement the fuzzy matching algorithm to match inexact participants, such as for
5
The number of full-time employees could not decrease and wages must not decrease by more than 25% for any
employee making less than $100,000 in 2019.
6
The Paycheck Protection Flexibility Act of 2020, enacted on June 5, 2020, made several modifications to the PPP,
including loosening stipulations and requirements for loan eligibility, repayment, and forgiveness. Specific changes
included 1) extending the maturity date after the borrower applies for loan forgiveness from 2 years to 5 years; 2)
extending the covered period from 8 weeks to 24 weeks from the loan origination date; 3) extending the date at
which employees must be rehired to avoid reduction of loan forgiveness from June 30th, 2020 to December 31,
2020; 4) reducing the amount of the PPP loan that must be spent on payroll to be eligible for forgiveness from at
least 75 percent to at least 60 percent; and 5) extending the loan deferral period (Congress.Gov 2021).
7
The Consolidated Appropriations Act of 2021 allowed for some 2020 participants to obtain a second loan in 2021.
7
variations of the same nonprofit name (e.g., “Enterprise Community Partners, Inc.,” “Enterprise
Community Partners,” or “Enterprise Commty Prtners”). While extensive manual verification is required
to ensure accuracy, supervised fuzzy matching allows us to successfully match such cases and thus obtain
a larger and more robust dataset (Bai et al. 2018).
Table 1, Panel A describes our sample selection. First, we retain the most recent filing for each
nonprofit prior to the CARES Act enactment, but we require filings with fiscal year ends in 2018 or later
to have financial data immediately prior to the program start date. We retain filings reporting 500 or fewer
employees and positive compensation expense to identify nonprofits eligible to participate. We also limit
our sample to nonprofits reporting positive assets and non-negative interest-bearing debt, salary and total
expenses, investible assets, and total liabilities to omit filings with possible Form 990 reporting errors
(Bowman et al. 2012), and we omit observations with missing data from our analyses. Our final sample
includes 39,055 PPP participants, which collectively received PPP loan approval of approximately $14.58
billion, and 64,831 non-participants. Panel B provides the number of PPP participants by approval month.
Most participants in our sample obtained loan approval in April 2020 (81 percent), the initial month of the
PPP loan distributions. Panel C provides the number of loans by loan amount. Most loans in our sample
are valued at less than $150,000, and about 4,200 loans were more than $1 million.
[TABLE 1]
3.2 Methodology
We use the following logistical regression to estimate the likelihood of PPP participation:
𝑃𝑃𝑃_𝐿𝑂𝐴𝑁 = 𝐿𝑇_𝐷𝐸𝐵𝑇 + 𝑅𝐸𝑆𝑇𝑅𝐼𝐶𝑇𝐸𝐷 + 𝐶𝑂𝑁𝑇𝑅𝑂𝐿𝑆 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑎𝑛𝑑 𝑆𝑡𝑎𝑡𝑒 𝐹𝑖𝑥𝑒𝑑 𝐸𝑓𝑓𝑒𝑐𝑡𝑠 (1)
The dependent variable, PPP_LOAN, identifies participating nonprofits and is equal to one if the
nonprofit received a PPP loan and zero otherwise. Our variables of interest relate to financial obligations.
LT_DEBT identifies nonprofits with tax-exempt, secured, or unsecured debt (Form 990, Part X, lines 20,
23, and 24, respectively), and RESTRICTED identifies nonprofits reporting restricted net assets resulting
from donor-restricted contributions (Part X, line 28). We also modify Equation (1) by replacing these two
8
indicator variables with a single measure of the magnitude of resource obligations, with OBLIGATIONS
equal to the natural log of one plus the sum of restricted net assets and tax-exempt, secured, and
unsecured debt scaled by total assets. In both specifications, we expect nonprofits with greater obligations
to be more likely to participate in the program.
We then estimate Equation (2) to examine the loan characteristics of PPP participants:
𝐿𝑂𝐴𝑁_𝐶𝐻𝐴𝑅𝐴𝐶𝑇𝐸𝑅𝐼𝑆𝑇𝐼𝐶𝑆 = 𝐿𝑇_𝐷𝐸𝐵𝑇 + 𝑅𝐸𝑆𝑇𝑅𝐼𝐶𝑇𝐸𝐷 + 𝐶𝑂𝑁𝑇𝑅𝑂𝐿𝑆 +
𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑎𝑛𝑑 𝑆𝑡𝑎𝑡𝑒 𝐹𝑖𝑥𝑒𝑑 𝐸𝑓𝑓𝑒𝑐𝑡𝑠 (2)
We examine two loan characteristics: loan amount and loan forgiveness. The dependent variable
LOAN_AMOUNT measures the approved loan amount relative to total payroll costs (including wages,
benefits, and retirement costs reported in Part I, Line 15). We expect nonprofits with greater obligations
to seek greater loan amounts to increase organizational capital and financial flexibility, and we predict
positive coefficients on our variables of interest. FORGIVEN identifies participants with fully forgiven
loans. On the one hand, nonprofits with long-term debt could use the program to refinance other debt
instruments and incur a one percent interest rate, resulting in a negative coefficient on LT_DEBT and
OBLIGATIONS. On the other hand, forgiven loans provide capital and additional financial flexibility that
does not require repayment, in which case we expect positive coefficients on our variables of interest.
We control for financial and entity-level characteristics common in the nonprofit accounting
literature (Boland et al. 2020; Chikoto and Neely 2014; Duguay et al. 2020), including governance,
financial resources, revenue characteristics, age, size, number of employees, location (rural or
metropolitan), business structure, political activities (see supplemental appendix for details), and industry8
and state fixed effects. Building on prior studies stressing the unique role of fintech institutions in
originating PPP loans (Erel and Liebersohm 2022; Griffin et al. 2023), we also include in Equation (2) a
measure identifying nonprofits whose loans originate with fintech lenders.
8
We use the National Taxonomy of Exempt Entities major groups to identify organization industries.
9
To account for differences in nonprofits in our sample with and without external obligations, we
entropy balance the regression covariates using a dichotomous variable equal to one if LT_DEBT or
RESTRICTED is equal to one. The covariates are balanced across groups on all three moments
(Hainmueller 2012). We generate entropy balanced weights for each subsample examined.
4. Results
4.1 Summary statistics
Table 2 contains summary statistics. In Panel A, approximately 37.6 percent of observations
participated in the program, which is less than the participation rate of other small business industries
(Schweitzer and Borawski 2021). Additionally, 29.1 percent of observations report long-term debt, 27.9
percent report restricted net assets, and the magnitude of obligations equals 14.7 percent of total assets.
Panel B reports loan characteristics summary statistics. On average, participating nonprofits received
loans equal to 22.3 percent of annual salary expense, and approximately 89 percent received full loan
forgiveness. Additionally, we find some variation in LOAN_AMOUNT, consistent with Schweitzer and
Borawski (2021), with an interquartile range from approximately 19.7 to 24.0. Panel C reports means and
proportions tests partitioning our full sample by participation. A greater percentage of PPP participants
report long-term debt compared to non-participants (33.6 percent versus 26.3 percent) and restricted net
assets (34.8 percent versus 23.8 percent), and the magnitude of obligations, as a percentage of total assets,
is greater for participants than non-participants (OBLIGATIONS equal to 0.154 and 0.143, respectively),
providing descriptive evidence that PPP participants have greater financial obligations.
[TABLE 2]
4.2 Multivariate Results
We present Equation (1) estimates in Table 3. In Column (1), we find the odds of nonprofits with
long-term debt and restricted net assets participating in the program are approximately 21 and 13 percent
greater than other eligible nonprofits, respectively.9 Further, the odds of participating increase by about
9
The odds are determined by subtracting one from the odds ratio, or ecoefficient – 1.
10
two percent with a one standard deviation increase in OBLIGATIONS (Column 2).10 Our results are
consistent with our expectations that nonprofits with financial obligations were more likely to participate.
[TABLE 3]
Next, we examine the loan characteristics of participating nonprofits and present the estimates of
Equation (2) in Table 4. First, we estimate LOAN_AMOUNT in Columns (1) and (2). As a percentage of
total payroll costs, loan amounts were approximately 0.538 and 0.683 percent greater for nonprofits with
long-term debt and restricted net assets, respectively, than other participants.11 Further, loan values
increased by approximately 0.283 percent of total payroll costs with a standard deviation increase in
OBLIGATIONS.12 While the allowable loan amount is determined by a formula, our results suggest there
was variation in the loan amount obtained relative to payroll costs, and those with secured debt and higher
borrowing costs obtained loans of greater values.13
[TABLE 4]
Our LOAN_AMOUNT variable includes total payroll costs, but the program restricts applicable
payroll costs to the first $100,000 per employee. There is therefore a possibility that LOAN_AMOUNT
could be inappropriately measured for nonprofits with employees earning more than $100,000 annually.
As robustness, we use information from Form 990, Part VII to identify nonprofits with any employee
earning more than $100,000 in the tax year and re-run the regressions in Columns (1) and (2) excluding
these nonprofits.14 In these untabulated tests we find no significant association with long-term debt and
loan amount using this sample; however, restricted net assets is positively associated with loan amounts.
10
The untabulated standard deviation of OBLIGATIONS in logged form for the full sample is 0.194; therefore, the
economic magnitude is calculated as e(0.273 x coefficient) – 1.
11
The average annual payroll for participants in our sample is approximately $1.75 million; therefore, the average
organization with secured debt received approximately $7,900 more than other participants.
12
One standard deviation of OBLIGATIONS for participating nonprofits is 0.183.
13
A colleague with first-hand experience applying for a PPP loan noted the terms of forgiveness were not clearly
defined at the time they applied for the loan. Further, he noted there was some discretion regarding the inputs into
the model, specifically stating “we were conservative as far as excluding funds paid to owners. We also were
conservative with some of the amounts for wages and distributions because we were trying to ensure we didn’t
overstate [our payroll costs].”
14
Part VII requires the disclosure of management compensation and compensation for the five highest-compensated,
non-management employees above $100,000.
11
Further, the magnitude of resource obligations is positively associated with loan amounts, suggesting
some association between financial obligations and the loan principal amount.
In Columns (3) and (4), we examine the likelihood of loan forgiveness among participants. We
find no association between having long-term debt and loan forgiveness; however, nonprofits with
restricted net assets were approximately 12 percent more likely to obtain loan forgiveness than other
nonprofits. Additionally, the magnitude of resource obligations is positively associated with the
likelihood of forgiveness.
5. Additional Analyses
5.1 Nonprofits with For-Profit Characteristics
Our focus is on tax-exempt entities with different operating incentives than for-profit businesses
(Eldenburg and Krishnan 2008). As a result, our inferences may not extrapolate to the broader population
of PPP-eligible businesses. We explore our findings’ generalizability to for-profit entities by identifying a
subsample of nonprofits that exhibit “for-profit” characteristics by earning a majority of their revenue
from non-donative sources (MAJORITY_DONATIONS = 0), such as program service revenue and
investment income. We estimate our regressions using this subsample and report results in Table 5. Our
results for participation and loan amounts are consistent with our main findings, suggesting some ability
to extrapolate our findings to for-profit entities. However, we find no significance in our forgiveness test.
[Table 5]
5.2 April 2020 PPP Loans
As shown in Table 1 Panel B, approximately 80 percent of participants in our sample received
their loans in April of 2020. However, as previously stated in footnote 6, Congress altered the program
provisions in June of 2020, which included changes to loan forgiveness provision, loan duration, and the
covered period to utilize loan proceeds on allowable costs. Therefore, the factors associated with
participation may have changed after the enactment of these modifications. To account for these
programmatic changes, we modify Equation (1) to estimate the likelihood of loan approval in April 2020,
and estimate Equation (2) using participants with April 2020 loan approval. We report the results in Table
12
6. Our results are generally consistent with our main findings; however, OBLIGATIONS is not associated
with loan forgiveness for this subsample.
[Table 6]
5.3 Lending Relationships
The SBA utilized the financial industry to originate loans; therefore, those with pre-existing
lending relationships may have had an advantage in applying for and obtaining loans. We omit nonprofits
with long-term debt and re-estimate the likelihood of participation. In these untabulated tests, we find
nonprofits with restricted net assets were still more likely to participate in the program and obtained larger
loans as a percentage of payroll costs.
5.4 Loan Proceeds for Interest Costs
The program’s loan forgiveness provision required participants to allocate at least 60 percent of
proceeds on payroll costs, but it also allowed participants to use proceeds on other costs including
mortgage interest costs. In effect, entities could lower their cost of debt by allocating up to 40 percent of
loan proceeds to interest costs and receive loan forgiveness. We examine the intended loan proceeds use
toward mortgage interest costs15 and isolate participating nonprofits with secured debt. We estimate the
amount of loan proceeds intended for interest conditional on the amount of interest expense and the value
of secured debt and report the results in Table 7. We find interest expense and secured debt are positively
associated with the loan proceeds intended for interest, which suggests some nonprofits used the PPP to
effectively reduce their cost of debt.
[Table 7]
15
The SBA discloses the loan proceeds intended for each allowable cost rather than actual proceed usage. While the
intended and actual use may differ, we believe the risk of understatement is low for the interest-related fields. The
primary PPP purpose was to support payroll costs, given that nonprofits were required to use a majority of PPP
proceeds on payroll costs to obtain loan forgiveness. As a result, participating entities likely overstated their
intentions to use funds on payroll and understated their intentions for other allowable costs. This likely
understatement of interest-related intended costs biases against our findings.
13
6. Conclusion
This study examines characteristics of nonprofit organizations participating in the PPP and the
characteristics of loans they obtained. The PPP issued loans to entities with the intent to incentivize
employee retention. These loans were forgivable if the funds were primarily used on payroll costs and
borrowers retained their employees, and participation provided additional capital and financial flexibility
to participants. We explore whether existing nonprofit financial obligations are associated with the
likelihood of participation, the requested loan amounts, and the likelihood of loan forgiveness.
We leverage the nonprofit setting to obtain mandatory financial and employment disclosures and
identify nonprofits eligible to participate in the program. Using a sample of PPP-eligible nonprofit
organizations, we find nonprofits with greater resource obligations were more likely to participate and
obtain a larger loan as a percentage of total payroll costs. Additionally, we find some evidence that
resource obligations are positively associated with the likelihood of loan forgiveness. Collectively, these
results provide additional insights on the program and the financial characteristics of participants.
14
References:
Arslan-Ayaydin, Ö, Florackis C, Ozkan A (2014) Financial flexibility, corporate investment and
performance: Evidence from financial crises. Rev Quant Finan Acc 42: 211–250
Autor D, Cho D, Crane LD, Goldar M, Lutz B, Montes J, Peterman WB, Ratner D, Villar D, Yildirmaz
A. (2022) An evaluation of the Paycheck Protection Program using administrative payroll
microdata. J. Public Econom. 211:104664.
Bai L, Koveos P, Liu M (2018) Applying an ontology-augmenting XBRL model to accounting
information system for business integration. Asia-Pac. J. Accounting Econom. 25(1–2): 75–97.
Ballew HB, Nicoletti A, Stuber SB (2021) The effect of the Paycheck Protection Program and financial
reporting standards on bank risk-taking. Management Sci. 68(3):2363-2371.
Bartik AW, Cullen ZB, Glaeser EL, Luca M, Stanton CT, Sunderam A (2021) The targeting and impact
of Paycheck Protection Program loans to small businesses. NBER Working Paper No. 27623,
National Bureau of Economic Research, Cambridge, MA.
Bartlett III RP, Morse A (2021) Small-business survival capabilities and fiscal programs: Evidence from
Oakland. J. Financial Quant. Anal. 56(7):2500-2544.
Beaton E, Colchin E, Ma Y (2020) Ohio nonprofit COVID-19 survey. Report, Ohio Nonprofit COVID-19
Survey Project, Columbus, OH.
Birdthistle WA, Silver J (2021) Funding crises: An empirical study of the Paycheck Protection
Program. Buffalo Law Rev. 69(5):1541-1620.
Boland CM, Harris EE, Petrovits CM, Yetman MH (2020) Controlling for corporate governance in
nonprofit research. J. Governmental Nonprofit Accounting 9(1): 1–44.
Bowman W, Tuckman HP, Young DR (2012) Issues in nonprofit finance research: Surplus, endowment,
and endowment portfolios. Nonprofit Voluntary Sector Q. 41 (4): 560–579.
Calabrese, TD (2011) Testing competing capital structure theories of nonprofit organizations. Public
Budgeting & Finance 31(3): 119-143.
Carroll DA, Stater KJ (2009) Revenue diversification in nonprofit organizations: Does it lead to financial
stability? J. Public Adm. Res. Theory 19(4):947–966.
Chetty R, Friedman JN, Hendren N, Stepner M (2020). The economic impacts of COVID-19: Evidence
from a new public database built using private sector data. NBER Working Paper No. 27431,
National Bureau of Economic Research, Cambridge, MA.
Chikoto GL, Neely DG (2014) Building nonprofit financial capacity: The impact of revenue
concentration and overhead costs. Nonprofit Voluntary Sector Q. 43(3):570-588.
Cole A (2022) The impact of the Paycheck Protection Program on small businesses: Evidence from
administrative payroll data. Working Paper, MIT, Cambridge.
15
Congress.Gov (2021) H.R. 7010 - Paycheck Protection Program Flexibility Act of 2020. Retrieved
January 19, https://www.congress.gov/bill/116th-congress/house-bill/7010
Doniger CL, Kay BS (2021) Ten days late and billions of dollars short: The employment effects of delays
in Paycheck Protection Program financing. Report, Federal Reserve, Washington, DC
Duguay R, Minnis M, Sutherland A (2020) Regulatory spillovers in common audit markets. Management
Sci. 66(8):3389-3411.
Eldenburg LG, Krishnan R (2008) The influence of ownership on accounting information expenditures.
Contemporary Accounting Rev. 25(3):739-772.
Erel I, Liebersohn J (2022) Can fintech reduce disparities in access to finance? Evidence from the
Paycheck Protection Program. J. Financial Econom. 146(1):90-118.
Finchum-Mason E, Husted K, Gugerty MMK, Barnhart E (2020) Local impacts of a global crisis: How
Washington state nonprofits are responding to COVID-19. Nancy Bell Evans Center on
Nonprofits & Philanthropy.
Foster W, Fine G (2007). How nonprofits get really big. Stanford Social Innovation Rev. 5(2), 46-55.
Garcia-Rodriguez I, Romero-Merino ME, Santamaria-Mariscal M (2022) Capital structure and debt
maturity in nonprofit organizations. Nonprofit Voluntary Sector Q. 51(6): 1451–1474.
Granja J, Makridis C, Yannelis C, Zwick E (2022) Did the Paycheck Protection Program hit the target? J.
Financial Econom. 145(3):725-761.
Griffin JM, Kruger S, Mahajan P (2023) Did fintech lenders facilitate PPP fraud? J. Finance.
Forthcoming.
Hainmueller, J (2012) Entropy balancing for casual effects: A multivariate reweighting method to
produce balanced samples in observational studies. Political Analysis 20(1):25-46.
Humphries JE, Neilson C, Ulyssea G (2020) Information frictions and access to the Paycheck Protection
Program. J. Public Econom. 190:104244.
Independent Sector (2022) Health of the U.S. nonprofit sector: A quarterly review. Retrieved November
9, 2022, https://independentsector.org/wp-content/uploads/2022/07/sector-health-q1-2022_v3.pdf
Lyons C (2021) “A blessing”: How PPP kept Greater Lansing nonprofits afloat. Lansing State Journal
(June 24, 2021). Retrieved July 27, 2021, from
https://www.lansingstatejournal.com/story/news/2021/06/24/paycheck-protection-program-loans-
greater-lansing-nonprofit-organizations/5318179001/.
Maher CS, Hoang T, Hindery A (2020) Fiscal responses to COVID-19: Evidence from local governments
and nonprofits. Public Admin. Rev. 80(4):644-650.
Schweitzer ME, Borawski G (2021) Which industries received PPP loans? Federal Reserve Bank of
Cleveland, Economic Commentary, 2021-08, DOI: 10.26509/frbc-ec-202108
16
U.S. Department of the Treasury (2021) About the CARES Act and the Consolidated Appropriations Act.
Retrieved January 19 2022, https://home.treasury.gov/policy-issues/coronavirus/about-the-cares-
act
U.S. Small Business Administration (2021) Paycheck Protection Program. Retrieved January 19, 2022,
https://www.sba.gov/funding-programs/loans/covid-19-relief-options/paycheck-protection-
program
Young DR (2007) Toward a normative theory of nonprofit finance. Young D, ed. Financing Nonprofits:
Putting Theory into Practice (AltaMira Press, Lanham, MD), 339-372.
17
Appendix. Variable Definitions
Variable Definition
Dependent Variables
PPP_LOAN Equal to one if the nonprofit received a PPP loan, zero otherwise.
LOAN_AMOUNT Equal to the approved loan amount scaled by total salary expense, multiplied by 100.
FORGIVEN Equal to one if the nonprofit’s PPP loan was fully forgiven, zero otherwise.
PPP_LOAN_APRIL Equal to one if the nonprofit received a PPP loan in April 2020, zero otherwise.
AMOUNT_INT Equal to the natural log of one plus the loan amount intended for mortgage interest.
Variables of Interest
LT_DEBT Equal to one if the nonprofit reported ending secured mortgages, unsecured notes payable
or tax-exempt bonds, zero otherwise.
RESTRICTED Equal to one if the nonprofit reported ending restricted net assets, zero otherwise.
OBLIGATIONS Equal to the natural log of one plus the sum of total restricted net assets, secured mortgage,
unsecured notes payable and tax-exempt bonds scaled by total assets.
INT_EXP Equal to the natural log of one plus total interest expense.
LN_SECURED_DEBT Equal to the natural log of one plus total ending secured debt.
Control variables
COST_OF_DEBT Equal to total interest expense divided by average tax-exempt bond, secured mortgages,
and unsecured debt liabilities multiplied by 100. COST_OF_DEBT is equal to zero if
LT_DEBT is equal to zero.
GOV_INDEX Equal to the sum of 5 indicator variables representing: majority independent board
members, CEO compensation approval, no delegation of management duties, financial
statement audit, and 990 form availability.
INV_ASSETS Equal to the natural log of one plus total ending cash, savings, and investments scaled by
total expenses.
FINTECH Equal to one if the loan originator is a fintech institution as defined by Erel and Liebersohn
(2020), zero otherwise.
BOARD_SIZE Equal to the natural log of one plus the number of board members.
MAJORITY_DONATIONS Equal to one if total public contributions exceed 50 percent of total revenue, zero
otherwise.
INVESTMENTS Equal to one if the nonprofit reported investment assets, zero otherwise.
GRANTS Equal to one if the nonprofit received grant revenue, zero otherwise.
UBI Equal to one if the nonprofit reported unrelated business income, zero otherwise.
AGE Equal to the natural log of one plus the nonprofit’s age, where age is equal to the fiscal year
of the 990 filing less the nonprofit’s formation year.
VOLUNTEERS Equal to the natural log of one plus the nonprofit’s volunteers.
ASSETS Equal to the natural log of one plus the nonprofit’s total assets.
EMPLOYEES Equal to the natural log of one plus the nonprofit’s employee count.
RURAL Equal to one if the nonprofit is in a rural or micropolitan statistical area, zero otherwise.
RELATED_PARTIES Equal to one if the nonprofit reported transactions with related parties in Form 990,
Schedule R, zero otherwise.
GROUP_RETURN Equal to one if the nonprofit filed a group return, zero otherwise.
LOBBYING Equal to one if the nonprofit engaged in lobbying activities, zero otherwise.
POLITICAL_CAMPAIGN Equal to one if the nonprofit engaged in political campaign activities, zero otherwise.
18
Table 1
Observation Detail
Panel A: Sample
Most recent Form 990 filing from each nonprofit prior to the CARES Act enactment
but after January 1, 2018 215,119
Omit observations:
More than 500 employees or zero reported employees (91,415)
Non-positive total salary expense, total expenses, investible assets, total assets,
long-term debt, and total liabilities (1,885)
Received PPP loan in 2021 (12,816)
Missing regression variables (5,117)
103,886
PPP loan recipients 39,055
Non-PPP loan recipients 64,831
Panel B: PPP Loan Recipient by Approval Month
Month in 2020 Number of Loans Approved
April 31,003
May 6,202
June 1,130
July 437
August 283
Total 39,055
Panel C: PPP Loan Amount
Loan Amount Number of Loans
($0, $10,000) 1,277
[$10,000, $50,000) 10,544
[$50,000, $150,000) 10,492
[$150,000, $500,000) 8,941
[$500,000, $1,000,000) 3,599
[$1,000,000, $2,000,000) 2,757
[$2,000,000, $10,000,000] 1,445
19
Table 2
Summary Statistics
Panel A: Full Sample (N = 103,886)
Variable Mean Std. Dev. 25% Median 75%
PPP_LOAN 0.376 0.484 0.000 0.000 1.000
LT_DEBT 0.291 0.454 0.000 0.000 1.000
RESTRICTED 0.279 0.449 0.000 0.000 1.000
OBLIGATIONS 0.147 0.273 0.000 0.000 0.186
COST_OF_DEBT 1.422 3.311 0.000 0.000 0.110
GOV_INDEX 4.079 0.973 3.000 4.000 5.000
INV_ASSETS 0.792 2.035 0.000 0.163 0.644
BOARD_SIZE 11.875 39.898 6.000 9.000 14.000
MAJORITY_DONATIONS 0.322 0.467 0.000 0.000 1.000
INVESTMENTS 0.330 0.470 0.000 0.000 1.000
GRANTS 0.392 0.488 0.000 0.000 1.000
UBI 0.070 0.255 0.000 0.000 0.000
AGE 31.202 24.043 14.000 26.000 42.000
VOLUNTEERS 162.373 504.377 0.000 16.000 94.000
ASSETS (in millions) 6.466 17.300 0.272 1.030 4.120
EMPLOYEES 49.438 84.962 5.000 14.000 49.000
RURAL 0.146 0.353 0.000 0.000 0.000
RELATED_PARTIES 0.183 0.387 0.000 0.000 0.000
GROUP_RETURN 0.001 0.027 0.000 0.000 0.000
LOBBYING 0.065 0.246 0.000 0.000 0.000
POLITICAL_CAMPAIGN 0.001 0.030 0.000 0.000 0.000
Variables are defined in the Appendix. Summary statistics for continuous variables are unlogged.
20
Table 2 (continued)
Panel B: Loan Characteristics Sample (N = 39,055)
Variable Mean Std. Dev. 25% Median 75%
LOAN_AMOUNT 22.304 7.338 19.699 21.626 24.035
FORGIVEN 0.889 0.315 1.000 1.000 1.000
LT_DEBT 0.337 0.473 0.000 0.000 1.000
RESTRICTED 0.348 0.476 0.000 0.000 1.000
OBLIGATIONS 0.154 0.249 0.000 0.014 0.227
COST_OF_DEBT 1.598 3.327 0.000 0.000 2.436
GOV_INDEX 4.289 0.901 4.000 5.000 5.000
INV_ASSETS 0.774 1.789 0.003 0.249 0.722
FINTECH 0.061 0.239 0.000 0.000 0.000
BOARD_SIZE 13.145 53.960 7.000 11.000 15.000
MAJORITY_DONATIONS 0.322 0.467 0.000 0.000 1.000
INVESTMENTS 0.397 0.489 0.000 0.000 1.000
GRANTS 0.434 0.496 0.000 0.000 1.000
UBI 0.083 0.275 0.000 0.000 0.000
AGE 34.398 25.463 16.000 29.000 45.000
VOLUNTEERS 210.156 579.468 0.000 25.000 130.000
ASSETS (in millions) 7.300 17.000 0.432 1.520 5.725
EMPLOYEES 61.404 93.128 7.000 22.000 69.000
RURAL 0.153 0.360 0.000 0.000 0.000
RELATED_PARTIES 0.173 0.378 0.000 0.000 0.000
GROUP_RETURN 0.000 0.022 0.000 0.000 0.000
LOBBYING 0.079 0.270 0.000 0.000 0.000
POLITICAL_CAMPAIGN 0.001 0.031 0.000 0.000 0.000
Variables are defined in the Appendix. Summary statistics for continuous variables are unlogged.
21
Table 2 (continued)
Panel C: Means and Proportions Tests
Subsample
PPP_LOAN
=1 =0
Variable (n = 39,055) (n = 64,831) Difference
LOAN_AMOUNT 22.304 --
FORGIVEN 0.889 --
LT_DEBT 0.336 0.263 0.073***
RESTRICTED 0.348 0.238 0.110***
OBLIGATIONS 0.154 0.143 0.011***
COST_OF_DEBT 1.598 1.316 0.282***
GOV_INDEX 4.289 3.953 0.336***
INV_ASSETS 1.025 1.253 -0.228***
FINTECH 0.061 --
BOARD_SIZE 13.145 11.110 2.035***
MAJORITY_DONATIONS 0.322 0.323 -0.001
INVESTMENTS 0.397 0.289 0.108***
GRANTS 0.434 0.367 0.067***
UBI 0.083 0.062 0.021***
AGE 34.398 29.277 5.121***
VOLUNTEERS 210.156 133.588 76.568***
ASSETS (in millions) 7.261 5.987 1.274***
EMPLOYEES 61.404 42.230 19.174***
RURAL 0.153 0.142 0.011***
RELATED_PARTIES 0.173 0.189 -0.016***
GROUP_RETURN 0.000 0.001 0.000**
LOBBYING 0.079 0.056 0.023***
POLITICAL_CAMPAIGN 0.001 0.001 0.000
Variables are defined in the Appendix. Summary statistics for continuous variables are unlogged.
***
indicates statistical significance at the 0.01 level.
22
Table 3
PPP Loan Participation
(1) (2)
Variable DV = PPP_LOAN
LT_DEBT 0.189***
(0.000)
RESTRICTED 0.123***
(0.000)
OBLIGATIONS 0.111***
(0.007)
COST_OF_DEBT -0.004 0.008***
(0.239) (0.002)
GOV_INDEX 0.187*** 0.188***
(0.000) (0.000)
INV_ASSETS -0.053** -0.056**
(0.030) (0.021)
BOARD_SIZE 0.239*** 0.242***
(0.000) (0.000)
MAJORITY_DONATIONS 0.221*** 0.218***
(0.000) (0.000)
INVESTMENTS 0.263*** 0.266***
(0.000) (0.000)
GRANTS -0.015 -0.014
(0.448) (0.494)
UBI -0.033 -0.029
(0.395) (0.447)
AGE 0.065*** 0.067***
(0.000) (0.000)
VOLUNTEERS 0.030*** 0.031***
(0.000) (0.000)
ASSETS -0.055*** -0.051***
(0.000) (0.000)
EMPLOYEES 0.189*** 0.189***
(0.000) (0.000)
RURAL 0.044 0.046
(0.115) (0.103)
RELATED_PARTIES -0.387*** -0.388***
(0.000) (0.000)
GROUP_RETURN -0.916** -0.910**
(0.012) (0.013)
LOBBYING 0.022 0.019
(0.558) (0.614)
POLITICAL_CAMPAIGN 0.097 0.104
(0.730) (0.708)
Industry and State Fixed Effects Yes Yes
Constant -2.006*** -2.030***
(0.000) (0.000)
Pseudo R2 0.047 0.046
Observations 103,886 103,886
Table 3 estimates the likelihood of participating in the program using an entropy balanced logistic regression.
Variables are defined in the Appendix. *** indicates statistical significance at the 0.01 level. Two-tailed p-values
are in parentheses.
23
Table 4
Loan Characteristics
(1) (2) (3) (4)
Variable DV = LOAN_AMOUNT DV = FORGIVEN
LT_DEBT 0.538*** -0.017
(0.000) (0.749)
RESTRICTED 0.683*** 0.113**
(0.000) (0.010)
OBLIGATIONS 1.545*** 0.296***
(0.000) (0.010)
COST_OF_DEBT -0.001 0.023 0.004 0.000
(0.953) (0.103) (0.559) (0.950)
GOV_INDEX -0.349*** -0.321*** -0.028 -0.020
(0.000) (0.000) (0.362) (0.509)
INV_ASSETS 0.756*** 0.790*** 0.047 0.062
(0.000) (0.000) (0.528) (0.403)
FINTECH -0.377 -0.366 -0.318*** -0.315***
(0.122) (0.133) (0.001) (0.001)
BOARD_SIZE -0.131 -0.086 -0.027 -0.015
(0.363) (0.546) (0.634) (0.784)
MAJORITY_DONATIONS -0.042 -0.054 0.122** 0.124**
(0.741) (0.671) (0.041) (0.036)
INVESTMENTS -0.395*** -0.362*** -0.024 -0.017
(0.002) (0.005) (0.693) (0.780)
GRANTS -0.137 -0.125 -0.180*** -0.178***
(0.237) (0.280) (0.000) (0.001)
UBI 0.118 0.116 0.066 0.063
(0.574) (0.583) (0.504) (0.525)
AGE -0.438*** -0.417*** 0.026 0.029
(0.000) (0.000) (0.479) (0.428)
VOLUNTEERS -0.021 -0.012 0.034*** 0.036***
(0.468) (0.673) (0.002) (0.001)
ASSETS -0.079 -0.076 0.022 0.017
(0.251) (0.256) (0.423) (0.516)
EMPLOYEES -0.095 -0.090 -0.062* -0.060*
(0.215) (0.239) (0.055) (0.064)
RURAL -0.575*** -0.575*** -0.000 -0.003
(0.000) (0.000) (0.996) (0.969)
RELATED_PARTIES 1.195*** 1.181*** 0.040 0.038
(0.000) (0.000) (0.566) (0.581)
GROUP_RETURN 1.431 1.414 0.217 0.211
(0.479) (0.467) (0.758) (0.765)
LOBBYING -0.371* -0.384* 0.030 0.032
(0.066) (0.058) (0.758) (0.744)
POLITICAL_CAMPAIGN -0.392 -0.417 1.822* 1.803*
(0.580) (0.559) (0.077) (0.080)
Industry and State Fixed Effects Yes Yes Yes Yes
Constant 26.249*** 26.048*** 2.194*** 2.172***
(0.000) (0.000) (0.000) (0.000)
R2/Pseudo R2 0.034 0.032 0.029 0.029
Observations 39,055 39,055 39,055 39,055
Table 4, Columns (1) and (2) estimate the PPP participants’ loan amount. Columns (3) and (4) estimate the likelihood
of PPP participant loan forgiveness. Variables are defined in the Appendix. ***, **, and * indicate statistical significance
at the 0.01, 0.05, and 0.10 levels, respectively. Pseudo R2 is provided for the logistical regressions. Two-tailed p-
values are in parentheses.
24
Table 5
Nonprofits with For-Profit Characteristics
(1) (2) (3) (4) (5) (6)
Variable DV = PPP_LOAN DV = LOAN_AMOUNT DV = FORGIVEN
LT_DEBT 0.211*** 0.555*** -0.019
(0.000) (0.000) (0.764)
RESTRICTED 0.139*** 0.466*** 0.047
(0.000) (0.000) (0.377)
OBLIGATIONS 0.184*** 0.989*** 0.211
(0.000) (0.003) (0.135)
COST_OF_DEBT -0.003 0.010*** -0.002 0.026 0.005 0.001
(0.388) (0.002) (0.901) (0.102) (0.513) (0.826)
GOV_INDEX 0.187*** 0.190*** -0.098 -0.082 -0.014 -0.009
(0.000) (0.000) (0.233) (0.313) (0.720) (0.811)
INV_ASSETS -0.009 -0.010 1.088*** 1.104*** 0.116 0.127
(0.768) (0.757) (0.000) (0.000) (0.253) (0.212)
FINTECH -0.325 -0.314 -0.247** -0.244**
(0.332) (0.350) (0.034) (0.036)
BOARD_SIZE 0.207*** 0.213*** -0.096 -0.063 -0.098 -0.091
(0.000) (0.000) (0.599) (0.727) (0.167) (0.193)
INVESTMENTS 0.330*** 0.334*** -0.466*** -0.446*** 0.011 0.014
(0.000) (0.000) (0.006) (0.009) (0.883) (0.845)
GRANTS -0.084*** -0.082*** -0.223 -0.213 -0.211*** -0.210***
(0.001) (0.001) (0.127) (0.147) (0.001) (0.001)
UBI -0.048 -0.045 0.011 0.018 0.129 0.127
(0.282) (0.319) (0.966) (0.941) (0.286) (0.294)
AGE 0.101*** 0.104*** -0.448*** -0.435*** 0.061 0.063
(0.000) (0.000) (0.000) (0.000) (0.182) (0.167)
VOLUNTEERS 0.028*** 0.029*** -0.024 -0.019 0.036** 0.037**
(0.000) (0.000) (0.520) (0.610) (0.016) (0.013)
ASSETS -0.073*** -0.069*** -0.116 -0.109 0.013 0.010
(0.000) (0.000) (0.142) (0.162) (0.730) (0.794)
EMPLOYEES 0.190*** 0.191*** -0.025 -0.022 -0.056 -0.054
(0.000) (0.000) (0.783) (0.814) (0.191) (0.208)
RURAL 0.057* 0.059* -0.454** -0.450** 0.022 0.022
(0.095) (0.085) (0.011) (0.012) (0.804) (0.809)
RELATED_PARTIES -0.390*** -0.392*** 1.279*** 1.266*** 0.053 0.052
(0.000) (0.000) (0.000) (0.000) (0.521) (0.526)
GROUP_RETURN -0.708* -0.701* 1.435 1.417 0.293 0.284
(0.064) (0.068) (0.535) (0.529) (0.726) (0.734)
LOBBYING 0.045 0.043 -0.664** -0.673** -0.059 -0.057
(0.349) (0.372) (0.020) (0.019) (0.627) (0.641)
POLITICAL_CAMPAIGN 0.251 0.255 1.372** 1.343** 1.522 1.506
(0.476) (0.466) (0.023) (0.028) (0.147) (0.152)
Industry and State Fixed
Yes Yes Yes Yes Yes Yes
Effects
Constant -1.854*** -1.905*** 25.249*** 25.079*** 2.107*** 2.094***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
R2/Pseudo R2 0.050 0.048 0.038 0.036 0.030 0.030
Observations 70,403 70,403 26,465 26,465 26,465 26,465
Table 5 estimates Equations (1) and (2) for a sample of observations earning most of their revenue from non-donation
sources. Variables are defined in the Appendix. ***, **, and * indicate statistical significance at the 0.01, 0.05, and 0.10
levels, respectively. Pseudo R2 is provided for logistical regressions. Two-tailed p-values are in parentheses.
25
26
Table 6
April 2020 Loans
(1) (2) (3) (4) (5) (6) (7) (8)
Variable DV = PPP_LOAN_APRIL DV = LOAN_AMOUNT DV = FORGIVEN
LT_DEBT 0.277*** 0.421*** 0.486*** -0.036
(0.000) (0.000) (0.001) (0.532)
RESTRICTED 0.135*** 0.117*** 0.650*** 0.098*
(0.000) (0.001) (0.000) (0.054)
OBLIGATIONS 0.284*** 0.697*** 1.518*** 0.189
(0.000) (0.000) (0.000) (0.151)
COST_OF_DEBT -0.003 0.013*** 0.001 0.021*** -0.000 0.020 0.011 0.007
(0.385) (0.000) (0.866) (0.002) (0.983) (0.179) (0.136) (0.292)
GOV_INDEX 0.206*** 0.206*** 0.084*** 0.079*** -0.330*** -0.302*** -0.030 -0.022
(0.000) (0.000) (0.001) (0.001) (0.000) (0.001) (0.429) (0.549)
INV_ASSETS -0.049* -0.056** -0.026 -0.034 0.703*** 0.740*** 0.071 0.085
(0.056) (0.029) (0.652) (0.561) (0.001) (0.000) (0.403) (0.320)
FINTECH -0.353 -0.341 -0.258* -0.255*
(0.322) (0.339) (0.069) (0.072)
BOARD_SIZE 0.319*** 0.321*** 0.433*** 0.428*** -0.203 -0.158 -0.043 -0.032
(0.000) (0.000) (0.000) (0.000) (0.243) (0.360) (0.524) (0.629)
MAJORITY_DONATIONS 0.250*** 0.241*** 0.212*** 0.188*** -0.026 -0.035 0.145** 0.149**
(0.000) (0.000) (0.000) (0.000) (0.857) (0.810) (0.042) (0.036)
INVESTMENTS 0.234*** 0.238*** 0.000 0.004 -0.407*** -0.376** -0.097 -0.092
(0.000) (0.000) (0.996) (0.932) (0.006) (0.012) (0.151) (0.175)
GRANTS -0.021 -0.019 -0.047 -0.048 -0.126 -0.117 -0.180*** -0.178***
(0.339) (0.380) (0.293) (0.285) (0.359) (0.396) (0.003) (0.003)
UBI -0.036 -0.031 0.026 0.029 0.213 0.211 0.048 0.045
(0.366) (0.441) (0.768) (0.744) (0.378) (0.383) (0.685) (0.702)
AGE 0.066*** 0.071*** 0.022 0.030 -0.450*** -0.430*** 0.057 0.059
(0.000) (0.000) (0.462) (0.317) (0.000) (0.000) (0.184) (0.171)
VOLUNTEERS 0.038*** 0.039*** 0.047*** 0.048*** -0.023 -0.015 0.033** 0.034***
(0.000) (0.000) (0.000) (0.000) (0.491) (0.658) (0.013) (0.009)
ASSETS -0.035*** -0.028*** 0.048** 0.060*** -0.066 -0.066 0.016 0.012
(0.000) (0.002) (0.018) (0.003) (0.422) (0.414) (0.608) (0.712)
EMPLOYEES 0.218*** 0.219*** 0.230*** 0.231*** -0.134 -0.129 -0.072* -0.070*
(0.000) (0.000) (0.000) (0.000) (0.151) (0.167) (0.071) (0.078)
RURAL 0.127*** 0.132*** 0.400*** 0.410*** -0.604*** -0.604*** -0.053 -0.056
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.509) (0.486)
RELATED_PARTIES -0.336*** -0.341*** 0.019 0.008 1.267*** 1.253*** 0.041 0.041
(0.000) (0.000) (0.761) (0.896) (0.000) (0.000) (0.604) (0.606)
27
GROUP_RETURN -0.828** -0.821** 0.130 0.132 1.884 1.852 -0.080 -0.090
(0.034) (0.035) (0.861) (0.856) (0.345) (0.340) (0.909) (0.898)
LOBBYING 0.004 -0.004 -0.007 -0.027 -0.438* -0.445* 0.118 0.121
(0.918) (0.928) (0.937) (0.752) (0.061) (0.058) (0.305) (0.293)
POLITICAL_CAMPAIGN 0.198 0.212 0.162 0.181 0.011 -0.001 1.586 1.574
(0.491) (0.454) (0.760) (0.735) (0.987) (0.999) (0.128) (0.132)
Industry and State Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes
Constant -3.025*** -3.086*** -1.562*** -1.664*** 26.353*** 26.158*** 2.288*** 2.280***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
R2/Pseudo R2 0.064 0.062 0.091 0.089 0.036 0.035 0.029 0.029
Observations 103,886 103,886 39,055 39,055 31,003 31,003 31,003 31,003
Table 6, Columns (1) - (4) estimate the likelihood of PPP participation in April 2020 for all PPP-eligible nonprofits, shown in Columns (1) and
(2); and for all nonprofits obtaining loan approval in 2020, shown in Columns (3) and (4). Columns (5)-(8) estimate Equation (2) for observations
with April 2020 loan approvals. Variables are defined in the Appendix. ***, **, and * indicate statistical significance at the 0.01, 0.05, and 0.10
levels, respectively. Pseudo R2 is provided for logistical regressions. Two-tailed p-values are in parentheses.
28
Table 7
Intended Proceed Use – Mortgage Interest
(1) (2)
Variable DV = AMOUNT_INT
INT_EXP 0.031*** 0.020***
(0.000) (0.000)
LN_SECURED_DEBT 0.101***
(0.000)
GOV_INDEX -0.044 -0.037
(0.215) (0.299)
INV_ASSETS -0.192** -0.125
(0.013) (0.107)
FINTECH -0.246** -0.247**
(0.020) (0.019)
BOARD_SIZE -0.070 -0.054
(0.271) (0.393)
MAJORITY_DONATIONS 0.047 0.060
(0.475) (0.357)
INVESTMENTS -0.034 -0.020
(0.590) (0.749)
GRANTS 0.010 0.009
(0.849) (0.868)
UBI 0.106 0.089
(0.210) (0.292)
AGE 0.050 0.061
(0.225) (0.138)
VOLUNTEERS -0.004 -0.001
(0.747) (0.925)
ASSETS 0.090*** 0.005
(0.000) (0.864)
EMPLOYEES 0.063** 0.076***
(0.023) (0.007)
RURAL 0.224*** 0.232***
(0.004) (0.003)
RELATED_PARTIES -0.088 -0.093
(0.163) (0.142)
GROUP_RETURN -0.728*** -0.755***
(0.000) (0.000)
LOBBYING -0.012 -0.004
(0.904) (0.965)
POLITICAL_CAMPAIGN -0.053 -0.085
(0.946) (0.913)
Industry and State Fixed Effects Yes Yes
Constant -0.782* -0.880*
(0.085) (0.050)
R2/Pseudo R2 0.035 0.038
Observations 10,565 10,565
Table 7 estimates PPP participants’ intended loan proceed usage on mortgage interest for
a sample of nonprofits that have secured debt. Variables are defined in the Appendix. ***,
**
, and * indicate statistical significance at the 0.01, 0.05, and 0.10 levels, respectively.
Two-tailed p-values are in parentheses.
29
File and source
- File
- nonprofit-organizations-financial-obligations-and-the-paycheck-protection-program.pdf
- Size
- 555,000 bytes
- SHA-256
- 5ac0dc62ef820fc6401a7f26353ef9190e691861dbf913649a03d77f1ce69b27
- Original
- doi.org