Professional Documents
Culture Documents
12306
Journal of Accounting Research
Vol. No. xxxx 2020
Printed in U.S.A.
ABSTRACT
This study uses a comprehensive panel of tax returns to examine the finan-
cial reporting choices of medium-to-large private U.S. firms, a setting that
controls over $9 trillion in capital, vastly outnumbers public U.S. firms across
all industries, yet has no financial reporting mandates. We find that nearly
two-thirds of these firms do not produce audited GAAP financial statements.
∗ Questrom School of Business, Boston University and Norwegian Center for Taxation ;
† University of Chicago Booth School of Business.
Accepted by Mark Lang. We thank Mary Barth, Philip Berger, Michael Donohoe, Merle
Erickson, Christian Leuz, Nemit Shroff, Doug Skinner, Jennifer Tucker, Ross Watts, Joe
Weber, Teri Yohn, and participants at the 2017 SEC-NYU Dialogue on Securities Market
Regulation Conference, University of Amsterdam, Chicago, Illinois at Urbana-Champaign,
Illinois at Chicago, Lancaster, Florida, Michigan, Michigan State, Missouri, Penn State, Vrije
Universiteit, and Yale, the FASB/PCC Summer 2014 meeting, the 2014 University of Min-
nesota Empirical Conference, the 2014 American Accounting Association Annual Meeting,
and the 2014 FARS conference for comments. This paper was previously titled, “Financial
Reporting Choices of U.S. Private Firms: Large-Sample Analysis of GAAP and Audit Use” and
“Which Private Firms Follow GAAP and Why?” Lisowsky was approved access to tax return
information through a contractual agreement with the Internal Revenue Service subject to a
nondisclosure agreement. All statistics are presented in the aggregate in accordance with IRS
disclosure rules. Minnis is a member of the Private Company Council (PCC). Any opinions
are those of the authors and do not necessarily reflect the views of the IRS or the PCC. Minnis
gratefully acknowledges support from the ARAMARK Faculty Research Fund at the University
of Chicago Booth School of Business. An online appendix to this paper can be downloaded at
http://research.chicagobooth.edu/arc/journal-of-accounting-research/online-supplements.
1
C University of Chicago on behalf of the Accounting Research Center, 2020
2 P. LISOWSKY AND M. MINNIS
1. Introduction
A significant challenge in accounting research is observing the costs and
benefits of firms’ financial reporting choices. At the most basic level, U.S.
firms are required by the Securities and Exchange Commission (SEC) to
produce audited financial statements in accordance with U.S. GAAP as a
condition for raising publicly traded equity or debt. Moreover, many coun-
tries outside the United States require firms—both public and private—to
prepare audited financial statements in accordance with accounting stan-
dards, such as International Financial Reporting Standards (IFRS). As such,
it is difficult for researchers to study why firms produce audited GAAP fi-
nancial statements because regulation typically masks firms’ endogenous
choices to do so. However, one setting—U.S. firms without publicly traded
securities—has essentially no reporting mandates and thus serves as a fruit-
ful setting to better understand which firms produce audited GAAP finan-
cial statements (Allee and Yohn [2009]). We use a comprehensive panel
data set of confidential business tax returns from the Internal Revenue Ser-
vice (IRS) to understand the propensity of audited GAAP financial state-
ment production of medium-to-large private U.S. firms and to examine the
factors associated with these choices.
Understanding firms’ decisions to produce audited GAAP financial state-
ments is important for several reasons. First, because of the lack of a reg-
ulatory mandate, private firms will only choose to produce audited GAAP
statements if the benefits exceed the costs. This allows us to examine char-
acteristics of firms that presumably derive the most benefits from producing
audited financial statements—and as importantly, can point to mechanisms
that substitute for (and thus reduce the benefit of) such statements. For
example, academics have claimed it should be taken as “given that lenders
[are] the main users of private company financial statements” (Bradshaw
et al. [2014, p. 183]; emphasis added), or that GAAP accounting is less
useful to firms with more intangible, knowledge-based assets (Lev and Gu
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 3
[2016]). We assess these claims by examining the features of firms that en-
dogenously choose to produce audited GAAP statements.
Second, the Financial Accounting Foundation (FAF) recently launched
the Private Company Council (PCC) under the premise that GAAP was not
serving the needs of private firms. In doing so, however, little empirical
analysis has been conducted to understand which firms actually produce
GAAP statements. A baseline understanding of the characteristics of which
firms do so can inform standard setters as GAAP is adjusted to be less costly
and more beneficial to private companies.
Third, beyond understanding the factors related to accounting use, an
important development in the U.S. economy over the last 20 years has been
the steady growth of private firms. Estimates suggest that the number of
private firms increased by more than 10% from 1996 to 2016, whereas the
number of public firms decreased by almost 50% during this same period
(Doidge et al. [2017], Wursthorn and Zuckerman [2018]; U.S. Census Bu-
reau).1 Although there are many causes for this trend, our general under-
standing of this growing part of the economy—including financial report-
ing decisions—is lacking, especially outside of small U.S. businesses. Our
study therefore provides not only a baseline understanding of the use of ac-
counting in medium-to-large private firms in particular, but also enhances
our understanding of the nature of this setting overall.
We use confidential tax return data on medium-to-large private U.S. firms
in our analysis. Since 2008, the IRS has required firms with at least $10
million in assets to report on Schedule M-3 the set of accounting stan-
dards they follow for financial reporting, and whether their financial state-
ments are audited. We find that although the firms in our sample are not
small—each has at least $10 million in assets—they are numerous with
about 70,000 each year. Using conservative assumptions, we estimate that in
2010, nonfinancial private U.S. firms outnumber public firms in Compustat
by eighteen-to-one; this ratio remains significant at three-to-one even after
conditioning on revenues of at least $100 million.2 Moreover, private firms
are more numerous than public firms across all industries, again, even after
conditioning on only the larger firms in our sample. Collectively, the firms
in our sample deploy over $9 trillion in capital. As a result, the data illustrate
1 We define a private firm as one that has neither a publicly traded security nor a require-
ment to file reports with a financial market regulator, such as the SEC (Minnis and Shroff
[2017]). The number of public companies is sourced from CRSP, while the number of private
companies is the number of firms with at least 20 employees (following Doidge et al. [2017])
as reported by the Census Bureau’s Business Dynamics Statistics public data set through 2014.
2 These figures exclude firms from the Financial, Insurance, and Real Estate industries and
other firms with data issues. We describe our sample and data validation tests in section3 and
the online appendix. The ratios of private to public firms are based on 2010 data. The number
of public firms is sourced from Compustat and includes firms with at least $10 million in assets;
this figure was 3,885 in 2010. We have 70,425 firms in the IRS data in 2010. We stress that the
IRS data do not provide a comprehensive view of all private firms, but only those with at least
$10 million of assets.
4 P. LISOWSKY AND M. MINNIS
that the term “private company” is far from synonymous with “small busi-
ness.”3 In short, larger private firms control significant sums of U.S. capital
and far outnumber public firms; because they do not publicly report their
financial statements, we describe them as the “silent majority.”
We begin our analysis of financial reporting practices by tabulating high-
level statistics. We find that 37% of medium-to-large private firms choose to
produce audited GAAP financial statements. More interestingly, even after
conditioning on firms having both ownership dispersion and external debt,
the audited GAAP rate increases only marginally to 40%. That is, we find
many firms with multiple owners and millions of dollars of debt do not
produce audited GAAP statements, while many smaller, single-owner firms
without debt do. These statistics reflect a vast and heterogeneous market
for financial reporting in which firms weigh the costs and benefits of ac-
counting choices in the absence of regulatory mandates. In short, the pro-
duction of audited GAAP financial statements is not a necessary condition
for external financing.
We use agency and information theories as frameworks to better un-
derstand the heterogeneity in financial reporting practices of private U.S.
firms. These frameworks posit that firms maximize value and produce au-
dited GAAP financial statements only when the benefits of doing so exceed
the costs (Watts [1977], Holthausen and Leftwich [1983], Watts and Zim-
merman [1986]). A primary benefit of high quality–verified financial state-
ments is to reduce agency costs and information asymmetry arising from
contracting arrangements between firms and outside parties, such as capi-
tal providers and suppliers (e.g., Armstrong, Guay, and Weber [2010]). As
such, we empirically model private firms’ financial statement production as
a function of proxies for these costs.
The panel structure of the IRS data allows us to examine variation in
financial reporting in both the cross-section and time series. In the cross-
section, we find that firms produce audited GAAP financial statements con-
sistent with theory: larger firms with more ownership dispersion and debt
are significantly more likely to produce audited GAAP financial statements.
But we also uncover other significant dimensions: firms with growth oppor-
tunities, trade credit, and those investing in intangibles (both book and
R&D based) are also more likely to have audited GAAP statements. By
contrast, firms that are older (i.e., potentially reflecting relationships or
reputation-based interaction with capital providers), with greater “inside
3 We note that our descriptive facts offer a contrasting view of private firms relative to prior
research focusing on small businesses. For example, an unfortunately titled paper, “What do
Private Firms Look Like?” (Asker, Farre-Mensa, and Ljungqvist [2011]) has been cited as pro-
viding descriptive statistics of what private firms look like. However, as the authors of that paper
discuss, their sample of firms excludes larger private firms. This caveat is missed in the litera-
ture, which cites that paper as describing the population of private firms (see, e.g., Bradshaw
et al. [2014, p. 180]).
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 5
debt” (i.e., debt provided by owners), and with more physical assets (e.g.,
land) are less likely to produce audited GAAP statements.
In the time series, we find that firms raising equity or increasing own-
ership dispersion are more likely to begin producing audited GAAP state-
ments. These results are particularly pronounced in young, high-growth
firms raising outside equity. Interestingly, we find that firms increasing their
level of outside debt, or moving from no debt to some positive level of debt,
do not exhibit any higher or lower likelihood of initiating audited GAAP
statements, regardless of firm age. However, we are cautious about making
strong inferences based on a lack of results related to changes in debt be-
cause our data lack granular detail on specific debt contracts where results
may manifest.
In addition to considering which factors are associated with audited
GAAP financial reporting, the data reveal several notable quantitative facts.
First, although size is the most important determinant, we find that the
rate of audited GAAP production does not exceed 50% until the sample in-
cludes firms with revenues around $100 million. Thus, a substantial portion
of middle market firms does not produce audited GAAP statements. Sec-
ond, although the data reveal that debt is strongly positively related to the
production of audited GAAP statements, we also find that audited GAAP
statements are far from necessary to attract even substantive amounts of
debt. The propensity to produce audited GAAP statements of firms with
zero debt is 38%; conditioning on at least $5 million of debt, this propen-
sity only increases to 49%.4 These findings reinforce that alternative mech-
anisms, such as relationships and collateral, likely play an important role
to accessing debt capital, even in the setting of medium-to-large firms, and
suggest relaxing the assertions in Bradshaw et al. [2014] that lenders are
the main users of financial statements. Finally, the findings highlight the
overlooked importance of equity transactions and trade credit. For exam-
ple, we find that firms adding new equity investors initiate audited GAAP
statements 65% of the time compared to 44% for firms without equity trans-
actions. Given the massive and increasing prevalence of private equity in-
vesting, the role of financial statement choices in equity capital allocation
is an important area for future research to consider.
In sum, our paper sheds light on the prevalence and determinants of fi-
nancial reporting in medium-to-large private U.S. firms. The key findings
are threefold. First, the findings suggest that the debt contracting focus
in the literature is too narrow as other characteristics, such as equity capital
and trade credit, exhibit significant explanatory power. Second, firm youth,
4 We selected $5 million in debt because Slee [2011] suggests this level of debt as a common
threshold for borrowers to provide audited financial statements to banks. Anecdotally, we have
experienced multiple conversations in the course of writing this paper suggesting individuals’
expectations are that firms have to provide audited GAAP statements to their banks if firms
are borrowing more than $5 million in debt. These assumptions are not consistent with the
data.
6 P. LISOWSKY AND M. MINNIS
growth, and R&D are positively associated with audited GAAP reporting,
suggesting important monitoring roles of financial reporting. Third, de-
spite the number of characteristics we examine, many firms violate stan-
dard explanations for financial reporting choices. For example, some firms
with both ownership dispersion and high levels of debt do not produce au-
dited GAAP statements, leaving substantial unexplained heterogeneity to
explore in future research.
Our paper makes several contributions to the accounting literature. First,
we extend prior research investigating the uses and effects of financial
reporting in private firms, generally (e.g., Blackwell, Noland, and Win-
ters [1998], Allee and Yohn [2009], Minnis [2011], Dedman, Kausar, and
Lennox [2014], Kausar, Shroff, and White [2016], Minnis and Sutherland
[2017]), by focusing on larger private U.S. firms, specifically. Our extension
is important because these firms have greater complexity and agency con-
cerns than small U.S. businesses (Cendrowski et al. [2012]) and, as our
results suggest, the findings from the small business literature may not be
generalizable to the larger firms (e.g., Allee and Yohn [2009]). Further-
more, conditional on firms with at least $10 million in assets, our study cov-
ers the population of tax filing firms in the United States, avoiding sample
selection concerns (Minnis [2011], Badertscher et al. [2018]) or interpre-
tation issues in surveys (e.g., Allee and Yohn [2009]). The panel structure
of the IRS data also enables us to examine how firm characteristics vary with
changes in accounting choices. In all, we significantly broaden our scope
and knowledge about the market for private firm financial reporting.
Second, our results suggest that an almost exclusive focus in the litera-
ture on how debt explains audited GAAP use needs to be broadened (e.g.,
see references and discussion in Armstrong, Guay, and Weber [2010]). We
find that equity and trade credit-based explanations are powerful, yet un-
derexplored factors (Costello [2013]). Moreover, our results reveal that the
use of audited GAAP statements is far from a corner solution, as thousands
of firms do not fit standard “expectations.” Our results on equity and age,
as well as trade credit, suggest that the literature should at least reassess
the conventional view taking debt as the primary (or at least only) driver of
financial reporting choices in this setting.
Third, because our setting is similar in economic magnitude to public
firms, both cumulatively and at the firm level, it is interesting to contrast
our findings to those in research on public firms. Specifically, research on
public firms presents evidence that financial statements are less relevant for
knowledge- and intangibles-based, growth opportunity firms (e.g., Lev and
Gu [2016], Govindarajan, Rajgopal, and Srivastava [2018]). However, our
findings suggest that audited GAAP financial statements play a particularly
important role when monitoring and information channels evident in pub-
lic markets do not exist (e.g., Grossman [1976]). In firms with less physi-
cal assets, greater growth opportunities, and more intangible assets (such
as human capital, software, and R&D), our results suggest highly verified
financial reports become more informative mechanisms for monitoring and,
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 7
5 An audit is the highest level of assurance that an independent accountant can provide.
Other report types include reviews (which provide negative assurance) and compilations
(which provide no assurance regarding the financial report). The tax returns we analyze do
not report whether the financial statements have been reviewed or compiled.
6 See Benston [1985], Botosan et al. [2006], Kothari, Ramanna, and Skinner [2010], and
acting parties with the firm who may demand high quality financial reports are employees,
customers, and governments. Unfortunately, our data do not provide adequate proxies to ex-
amine the interactions with these parties, so we leave such a focus to future research.
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 9
8 Thus, that study suggests that only about 5% (i.e., 27% of 20%) of small firms have audits.
Unfortunately, the SSBF survey question does not ask what type of financial statements the
firms actually produce or whether they are audited; instead it asks what records the survey re-
spondent used to answer the survey questions. Specifically, the survey asks the respondent, “Do
you have records available to help you answer questions about the firm’s income, expenses,
and balance sheet, such as tax records, statements, worksheets, or any other records?” The
follow-up question is “What records are you using?” It is unclear whether the question refers
to the records being used to answer the survey, or the records being used in the business itself.
A&Y report that of the 4,004 firms responding, 790 responded as using financial statements,
but 1,682 responded as having used memory. Therefore, it is uncertain whether the findings
in A&Y refer to the actual production of financial statements, or simply the use of financial
statements in answering the SSBF survey.
9 Although studying small, entrepreneurial firms is certainly important for a variety of rea-
sons, our point is simply that “private firms” include not only small businesses, but also large,
12 P. LISOWSKY AND M. MINNIS
3. Data
We use panel data from the population of private U.S. firms with at least
$10 million in assets. These data are sourced from Schedule M-3 of the
U.S. federal income tax returns of Subchapter C corporations (Form 1120),
Subchapter S corporations (Form 1120S), and partnerships (Form 1065).
Since 2004, Schedule M-3 requires companies with assets of $10 million
or more to disclose to the IRS whether the financial statements have been
audited by an independent accountant, and since 2008, the IRS further
requires firms to report on Schedule M-3 which set of accounting standards
they use for financial reporting. We use these Schedule M-3 disclosures to
identify whether a firm undergoes a financial statement audit and what set
of accounting standards it follows for financial reporting.10
For the purposes of this study, the IRS generously provided access to one
of the authors all Schedules M-3 for entities filing Forms 1120, 1120S, and
1065 for fiscal years 2008–2010 at the consolidated U.S. parent level. By def-
inition, because filers of Schedule M-3 report assets of $10 million or more,
our sample begins with medium-to-large firms in the population; there are
no IRS disclosure requirements on financial statement audits or GAAP
usage for firms with assets less than $10 million. Despite this truncation,
table 1 reports that the initial data set provides 630,076 nonpublicly
traded firm-year observations across all three entity types.11 We remove
complex organizations with significant agency concerns and external capital demands that
are typically omitted from previously used data sets. Prior research also investigates two other
settings. First, non-U.S. settings used to investigate the accounting practices of private firms
include Dedman, Kausar, and Lennox [2014], Hope, Thomas, and Vyas [2011], Lennox and
Pittman [2011], and Kausar, Shroff, and White [2016]. However, they are limited to firms with
differing mandates regarding accounting and disclosure (for both public and private firms)
relative to the United States. Thus, the cost–benefit trade-off of financial reporting decisions
is likely different between the U.S. and non-U.S. settings, so the results may not be compara-
ble. Second, in a recent working paper, Gaver, Mason, and Utke [2019] examine the financial
reporting of private investment funds and find at most 53% of those funds prepare audited
GAAP statements. That study adds to research on financial reporting outside of public firms
and the debt contracting perspective by examining a specific equity fund setting.
10 We take several steps to ensure the internal and external validity of the data; see online
appendix A. In particular, we find that the summary statistics of audited GAAP production of
the firms in our tax return sample are quite similar to two other independently collected data
sets, reassuring our approach (see the appendix). Also see the online appendix for additional
details about Schedule M-3, including the 2010 version of the schedule.
11 Specifically, we omit any firm answering “Yes” to having filed with the SEC or having a
(1) financial firms (NAICS code 52) because their accounting and audit
choices are affected by regulation and (2) real estate and management
firms (NAICS codes 53 and 55) because there are many instances of firms
with significant assets, but no revenues or expenses, suggesting these are
nonoperating holding companies. We also drop firms with foreign owner-
ship greater than 25% because these entities may face alternative regulatory
regimes or may be subsidiaries of publicly traded foreign firms. This step
leaves a sample of 216,898 observations.12
The sample of 216,898 observations includes firms that file their tax re-
turns electronically (e-filers; n = 91,410) or on paper (paper filers; n =
125,488). One constraint of the IRS data is that the Schedule M-3 disclo-
sures on audit and GAAP use are available to us only for e-filers and not for
paper filers. We therefore use the data in two different ways. For any statis-
tic that considers the population, we use both paper and e-filers because
we have data on firm-level characteristics for both. However, for all anal-
yses that examine the relation between firm characteristics and account-
ing choices, we only use the 91,410 e-filing observations for which actual
12 By only using consolidated parent tax returns, and by eliminating financial, real estate,
and management firms, as well as all firms with substantial foreign ownership, we substan-
tially reduce the risk of double-counting entities and assets, for example, a private equity firm
(which would be considered a financial firm) owning a fund that invests in a private company.
Moreover, the author with confidential access did two hand-checks of the data: (1) he identi-
fied a sample of private companies from other data and confirmed they exist in the tax return
sample and (2) he examined a random sample of firms in the tax return data set to ensure
they are operating private companies. In addition, the list of company names was perused
to identify any systematic inclusion of shell companies (i.e., with unusual names) or other
oddities with the data. He found no evidence of inappropriate firms remaining in our sample.
14 P. LISOWSKY AND M. MINNIS
4. Analysis
4.1 DESCRIPTIVE ANALYSIS OF PRIVATE FIRMS
As the comprehensive data on medium-to-large private U.S. firms we use
in our study are rare, we begin our analysis by providing descriptive statistics
of the firms in table 2. All variables are defined in the appendix. The first six
columns report statistics for the entire sample (i.e., including paper filers),
and the last two columns report means for the paper and e-filing samples,
respectively. We winsorize continuous variables at the top and bottom 1%.13
Firms in our sample are substantially larger than prior studies of private
firms and control vastly more resources. The mean (median) firm has $53.7
($24.3) million in revenue and $57.7 ($19.4) million in total assets.14 For
2010, we find that the (unwinsorized) cumulative total assets controlled
by the private firms in our sample exceed $9 trillion (untabulated). The
statistics emphasize that these firms are much different from the private
firms examined in prior studies, for example, average total assets are less
than $0.4 million in Allee and Yohn [2009] and $6.5 million in Minnis
[2011].
13 Per IRS disclosure rules, we are not permitted to report the descriptive statistics at any
item is most similar to Compustat’s “sale” variable, plus interest, rents, royalties, and gains.
TABLE 2
Descriptive Statistics
Population Paper Only E-File Only
n = 216,898 n = 125,488 n = 91,410
T A B L E 2—Continued
Population Paper Only E-File Only
n = 216,898 n = 125,488 n = 91,410
FIG. 1.—Number of private and public firms and percentage of private firms with audited
GAAP financial statements by Compustat sales size quintile for the year 2010. This chart re-
ports the distribution of private firms (paper and e-file) partitioned by Compustat sales rev-
enue quintiles for the year 2010. The bar charts report the number of firms in each size
quintile and the line chart reports the percentage of private firms in that quintile that pro-
duce audited GAAP financial statements. The quintiles are based on the Compustat variable
“sale” for all nonfinancial U.S. firms in Compustat. The sizes of the private firms are based on
the variable Revenue as defined in the appendix. The financial reporting data (i.e., % Audited
GAAP) for the e-filing firms are as reported on Schedule M-3; estimates are used for the pa-
per filing firms as described in online appendix B. In comparing tax return data to Compustat
data, we note two important caveats. First, “Revenues” per the tax return are not an exact com-
parison to the “sale” variable in Compustat. Various differences in definitions, consolidation
rules, foreign versus domestic revenues all need to be considered. Second, Compustat does
not cover all firms that have SEC registration requirements—that is, it is not a comprehensive
sample of publicly traded firms. This chart is meant to demonstrate broad relative magnitudes
of the two settings rather than precise estimates.
Also in contrast to prior studies of private firms, the firms in our sam-
ple are much more similar to public U.S. firms. To illustrate this compari-
son, figure 1 plots the number of nonfinancial firms in both the IRS data
(paper and e-filers) and Compustat for the year 2010 by Compustat size
quintile (based on the variable “sale”). Although Compustat data are not
directly comparable to tax return data because of consolidation and other
accounting differences, this chart is simply intended to give a sense of the
relative magnitude of the two settings. As the chart indicates, there are sub-
stantially more private firms in the bottom three quintiles (total of 69,563)
compared to Compustat (total of 2,331), and the number of private firms
is similar in the fourth quintile (777 public firms versus 682 private firms).
These four quintiles represent firms with revenues up to $2.3 billion each. It
is not until the largest quintile (i.e., firms each with revenues greater than
18 P. LISOWSKY AND M. MINNIS
15 If we condition that the private firm must be a C corporation, which is the most common
form of publicly traded firm, our distribution is as follows: 10,732 in Q1, 4,825 in Q2, 985
in Q3, 243 in Q4, and 66 in Q5, for a total of 16,851 private C corporations. Note that this
number far surpasses total public firms in Compustat, as well as includes some very large
private firms with revenues over $2.3 billion. Interestingly, more than half of the private firms
in all quintiles—even the largest two quintiles—are “pass through” entities (i.e., partnerships
and S corporations), not C corporations, suggesting that viewing pass through entities as only
“small businesses” is inaccurate.
16 “Taxable income” is referred to as “Ordinary Business Income” for pass-through entities.
We use the term “taxable income” for both types of entities for simplicity.
17 We use logged transformations in our regression models when measuring revenues
(LOG SIZE), external debt (LOG DEBT), internal debt (LOG DEBT FROM SH), and trade
credit (LOG ACCT PAY).
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 19
We caution that, unlike the income statement variables, which follow tax
rules for all firms regardless of the financial accounting basis they choose,
the balance sheet variables are reported using the book basis of account-
ing and could be mechanically related to firms’ accounting choices. For
example, if a firm uses the tax basis of accounting, assets such as equip-
ment (inventory) may be understated (overstated) relative to a firm that
uses GAAP. This is a primary reason why we use revenues rather than total
assets as our primary measure of size—revenues (as well as profit margin
and growth) are measured on the tax basis for all firms. As we discuss in the
next section, we also address this issue by using different specifications that
exclude balance sheet variables or substitute income statement variables for
balance sheet variables.
The tax returns also provide measures for ownership, age, and organi-
zational form. The IRS data contain the number of firm owners, but firms
are only required to report this field if the number of owners is 100 or
less. Thus, the IRS reporting requirement limits our ability to identify the
number of owners if they exceed 100 as only a few firms voluntarily dis-
close this number; otherwise the disclosure is left blank.18 Therefore, we
handle the NUMBER OF OWNERS variable three ways: (1) we set missing
values to equal 101 owners; (2) we exclude firms reporting more than
100 owners or missing values; and (3) we use indicator variables for vari-
ous ownership levels, including if it exceeds 100 or is missing. The own-
ership variable is right-skewed, as the mean NUMBER OF OWNERS is 14.5
including all available observations (after setting the missing values to 101)
and 6.9 after excluding firms with more than 100 owners or missing val-
ues. The variables reveal that 40% of our sample has just one or two
owners (OWNER EQ 1 and OWNER EQ 2), 27% have three to five owners
(OWNER EQ 3to5), and 8% have more than 100 owners or report missing
values (OWNER EQ 101orMore).19
18 The term “Owners” collectively indicates shareholders (for corporations), partners (for
partnerships), or members (for limited liability companies). Note that the data set has even
more right tail skewness than reported in the descriptive statistics because, as with all our
continuous variables, we winsorize the number of owners—when data are available—at the
99th percentile. In fact, the data contain 233 observations that (voluntarily) report more than
500 owners (the 99th percentile is 176 owners). These firms are professional services part-
nerships, which are exempt from SEC rules that require public financial reporting for firms
with 500 or more owners. Overall, when we use continuous owner data, we log the data to
create LOG NUM OWNER. We tabulate robustness tests in the online appendix and find that
the specification of the number of owners does not affect our inferences.
19 Unfortunately, the IRS data do not allow us to identify owners, so we cannot study owner-
ship variation in greater detail. In particular, given the recent increase in private equity (PE)
financing, one might be interested in the role of PE firms as owners. PitchBook, a relatively
comprehensive data set for North American PE activity (see Brown et al. [2015.]) reports that
about 5,900 companies were PE-backed in 2017 (PitchBook [2017]). It further estimates that
the number of PE-backed firms grew by 5% per year since 2009, suggesting that about 4,000
companies were PE-sponsored in 2009—and this number does not filter real estate or financial
firms as we have done in our study. PitchBook also reports that much of the PE deal activity
20 P. LISOWSKY AND M. MINNIS
In terms of age, only corporations (Forms 1120 and 1120S) report the
year of incorporation. The mean (median) corporation is 28 (24) years
old (AGE). We also report the share of our sample firms that are C corpora-
tions, the dominant form of organization for public firms. We find that only
25% of our sample firms are C corporations. The remaining firms are split
between S corporations (Form 1120S) or partnerships and limited liability
companies (Form 1065).
In table 2, we also report the population estimate of audited GAAP finan-
cial statement production. However, recall from section 3 that we have all
variables for the paper filing firms except their financial reporting choices.
Therefore, we estimate the population rate of audited GAAP production
using e-filing firms as counterfactual estimates for paper filing firms (see
the online appendix for details). As table 2 reports, we estimate that 37%
of the population of medium-to-large private U.S. firms produces audited
GAAP financial statements (GAAP AUDIT). This joint decision breaks down
to 79% of firms using GAAP (GAAP) and 38% having their financial state-
ments audited (AUDIT).20 Given the differences in size, ownership disper-
sion, and levels of debt between paper filers and e-filers, it is not surprising
that the estimated rate of audited GAAP production is lower for paper filers
(32%) compared to the actual rate for e-filers (44%). Figure 1 plots the per-
centage of private firms that produce audited GAAP financial statements in
each Compustat size quintile. Although this shows a strong positive relation
between firm size and GAAP audit use, it also highlights that even many
large firms do not produce audited GAAP financial statements.
As a final descriptive analysis, we present the distribution of (e-filing)
firm-years and propensity to produce audited GAAP statements by industry.
Table 3 sorts industries from highest to lowest rates of audited GAAP pro-
duction. Columns 1 and 2 report the number of firm-years and percentage
of sample, respectively.21 Column 3 reports the percentage of firm-years
producing audited GAAP statements. Information has the highest rate at
62%, followed closely by Utilities, Manufacturing, Services, and Healthcare.
These results suggest that as the U.S. economy continues its evolution to-
ward industries with substantial human and intangible capital (i.e., infor-
mation and services industries), audited GAAP reporting could be an im-
portant tool to verify, value, and monitor these firms. Consistent with this
occurs in the sub-$100 million market, suggesting that the activity is not concentrated solely
among the largest firms in our sample. Therefore, a likely upper bound on the share of PE-
owned firms in our population does not exceed 5.7% (or 4,000/70,000 private firms in 2009,
per our table 3).
20 We tabulate all accounting choices separately for e-filing firms in the online appendix. Of
the 16% of e-filing firms not claiming GAAP as their set of accounting standards, 12.5% use
tax basis, while 0.2% use IFRS, 0.6% use statutory, and 2.9% use an alternative set of standards
(e.g., cash basis). Note that the audit rate is slightly higher than the joint GAAP-and-audit rate
because a very small portion of firms have their non-GAAP financial statements audited.
21 The online appendix compares the distribution of firm-years across industries to other
view, table 3 also reveals that industries with significant tangible assets (e.g.,
Mining and Agriculture) and cash-based businesses (e.g., Food Service and
Retail Trade) have the lowest rates of audited GAAP statements. We note
that within Agriculture—the industry with the lowest audited GAAP rate of
25%—almost half of the firms have less than $5 million in revenues, yet
have at least $10 million in assets, reinforcing the notion that agriculture
not only includes many small operations but that this is also an asset inten-
sive industry.
The main takeaways from the descriptive analysis of the economics of
large private firms are as follows: They are far more numerous than public
companies, operate across all major industries, cumulatively deploy trillions
of dollars of capital, and have significant financial reporting heterogeneity.
We now examine more rigorously how this financial reporting heterogene-
ity relates to firm characteristics and capital structure.
only consider e-filing firms for which we can link actual (rather than esti-
mated) financial reporting choices to firm-level data.
We begin by using two-way sorts to examine three constructs most com-
monly viewed as associated with financial statement production—firm size
(revenues), separation of ownership and control (number of owners), and
external debt. Table 4, panel A, partitions firm-years into size and own-
ership dispersion cells, and panel B partitions firm-years into size and
debt cells. Each cell reports the number of firm-years and proportion
of those firm-years with audited GAAP financial statements. We find that
the rate of audited GAAP financial statement production is clearly in-
creasing in all three variables, both unconditionally (examining the out-
side rows and columns) and conditionally on the other variable (exam-
ining the interior rows and columns). As expected, the upper left cells
have the lowest audited GAAP rates while the bottom right cells have the
highest.
Two additional insights emerge. First, although the relation between au-
dited GAAP rates and each of the variables is generally monotonic, it is
not linear. The relation between audited GAAP rates and ownership dis-
persion is the most apparent in firms with between 1 and 10 owners, and
the relation flattens out considerably after firms reach about 20 owners.
Increases in ownership dispersion on the intensive margin after this point
seem to have little relation with audited GAAP rates. Debt has a similarly
monotonic yet concave relation, with an inflection point near $25 million.
Data in table 4 and figure 1 (which reports the proportion of firms pro-
ducing audited GAAP statements by Compustat size quintile) show that
size has a concave relation that begins to flatten near $200 million in
revenues.
Second, the results from table 4 do not support corner solutions or text-
book “rules of thumb” about thresholds regarding the production of au-
dited GAAP financial statements.22 Although size, ownership dispersion,
and debt have strongly positive relations with audited GAAP use, many
firms do not meet standard expectations. For instance, focusing on firm-
years with both more than one owner and nonzero external debt results in a
conditional propensity to produce audited GAAP statements of only 40%,
that is, less than half the firms with ownership dispersion and outside debt.
Focusing specifically on debt, the propensity to produce audited GAAP
statements of firms with zero debt is 38% (as seen at the bottom of the
first column); after conditioning on firms with at least $5 million of debt,
this propensity increases to only 49% (i.e., again less than half the sample).
This difference suggests that many firms with relatively high levels of exter-
nal debt do not produce audited GAAP statements while many firms with
no debt do produce audited GAAP statements. The former result is consis-
tent with the idea that lenders to private companies—even large ones—rely
22 For example, in the textbook Private Capital Markets, Slee [2011] suggests $5 million in
debt as a common threshold for borrowers to provide audited financial statements to banks.
TABLE 4
Two-Way Sorts of Audited GAAP Financial Statement Rates by Size, Ownership Dispersion, and Debt
Panel A: Size and ownership dispersion partitions
Number of Owners
1 2–5 6–10 11–15 16–20 21–25 26–30 31–35 36–40 41–45 46–50 51–100 >100 Total
Total revenue ($ millions) <20 4,538 14,675 3,452 1,265 669 439 362 296 211 183 177 725 2,724 29,716
25% 19% 27% 34% 36% 40% 46% 38% 45% 46% 43% 51% 50% 27%
20–50 5,469 12,742 3,370 1,191 589 385 272 209 135 124 128 470 1,697 26,781
42% 37% 48% 53% 57% 56% 55% 59% 59% 57% 61% 69% 64% 44%
50–100 3,405 8,275 2,218 792 409 296 235 179 126 72 78 419 1,205 17,709
46% 46% 59% 67% 68% 73% 66% 73% 67% 54% 64% 60% 67% 52%
100–500 2,466 5,711 1,972 842 559 278 217 186 167 128 110 463 1,666 14,765
61% 61% 71% 81% 77% 82% 85% 83% 84% 80% 87% 77% 66% 67%
>500 337 584 263 124 124 59 61 46 38 33 31 119 620 2,439
69% 70% 84% 84% 85% 93% 85% 96% 89% 91% 81% 86% 60% 73%
Total 16,215 41,987 11,275 4,214 2,350 1,457 1,147 916 677 540 524 2,196 7,912 91,410
42% 36% 49% 56% 59% 61% 62% 62% 64% 61% 62% 64% 60% 44%
(Continued)
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES
23
24
T A B L E 4—Continued
Panel B: Size and debt partitions
0 0–5 5–10 10–15 15–20 20–25 25–30 30–35 35–40 40–45 45–50 50–100 >100 Total
Total revenue ($ millions) <20 10,493 7,206 4,849 3,014 1,503 786 486 322 214 155 128 361 199 29,716
25% 32% 27% 21% 23% 25% 28% 25% 29% 34% 38% 34% 39% 27%
P. LISOWSKY AND M. MINNIS
20–50 6,047 10,094 5,591 2,253 907 480 316 206 166 120 76 356 169 26,781
41% 43% 39% 46% 59% 58% 68% 69% 66% 68% 74% 71% 49% 44%
50–100 3,678 4,959 3,290 2,258 1,112 611 407 242 170 143 114 451 274 17,709
49% 53% 47% 44% 52% 63% 64% 71% 71% 75% 82% 78% 67% 52%
100–500 3,122 2,714 1,513 1,192 1,038 815 630 450 385 290 262 1,168 1,186 14,765
58% 63% 64% 67% 66% 67% 71% 70% 74% 77% 79% 80% 79% 67%
>500 505 204 105 75 78 66 50 53 40 46 39 281 897 2,439
55% 77% 79% 79% 67% 83% 80% 68% 83% 67% 72% 81% 79% 73%
Total 23,845 25,177 15,348 8,792 4,638 2,758 1,889 1,273 975 754 619 2,617 2,725 91,410
38% 44% 40% 40% 48% 53% 58% 59% 63% 66% 70% 72% 73% 44%
This table partitions the full sample of e-filing firms (2008–2010) by the number of owners and sale revenues (panel A) and amount of debt and sales revenue (panel B). Each
cell reports the number of firms in each portfolio and the percentage of those firms that produce audited GAAP financial statements.
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 25
23 Our data do not allow us to measure two prominent forms of equity and debt financing:
employee stock option plans and operating leases, respectively. The former is important be-
cause the Private Company Council has found that it is not uncommon for private companies
to issue stock-based compensation and yet struggle with the accounting for this form of com-
pensation (see https://www.fasb.org/pcc for documents related to this topic). The latter is
important because lessors typically have stronger repossession rights and therefore potentially
rely less on financial reporting mechanisms (see Sutherland [2018] for a discussion).
26 P. LISOWSKY AND M. MINNIS
TABLE 5
Firm Characteristics of Audited GAAP Financial Statement Production: Firm Level
Dependent Variable: GAAP AUDIT
24 In untabulated tests, we further investigate the interaction between land and debt by
interacting LOG DEBT with LAND TO ASSETS to the specification of table5, column 1. We
find a significantly negative coefficient on the interaction confirming the results in Minnis
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 27
and Sutherland [2017] that the presence of real estate mitigates the usefulness of financial
reporting in debt contracting. We also interact LOG DEBT with PE TO ASSETS and find a
negative, but statistically insignificant, coefficient.
25 At this point, it is worth contrasting in more detail our findings with prior research, in
particular Allee and Yohn’s [2009] (A&Y) table 6, which is the most similar analysis to our
table 5. A&Y examine the sophistication of accounting with two different dependent variables
in their table 6: a count variable ranging from 0 to 3, representing company prepared, com-
piled, reviewed, and audited financial statements, respectively, and an indicator variable for
whether or not the firm reports using an accrual basis of accounting. The first item to note
from A&Y is the lack of significant results for most variables. For example, firm size (total
assets)—the most significant variable in our study and others (e.g., Minnis [2011])—is not sig-
nificantly related to whether the firm has an audit or follows accrual accounting. In addition,
variables indicating new equity, the extent of leverage, the amount of sales growth are also in-
significant in A&Y. Ownership dispersion and trade credit is only significant in their accruals
regression. In contrast to the results in our table 5, firm age in A&Y is positively associated with
accounting sophistication. Ultimately, the lack of significance and mixed results of the A&Y
28 P. LISOWSKY AND M. MINNIS
table is likely driven in part by their small sample size (790 observations), noisy data points
generated by survey data, and the very different (i.e., small business) setting.
26 We also re-estimate table 5 using a logit model and find similar results. As an exploratory
analysis, we also re-estimate the regression for each industry and tabulate the results in the
online appendix. The results are generally consistent across industry (e.g., the coefficients on
size, ownership dispersion, and debt are almost always of the same sign and statistical signifi-
cance).
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 29
across the variables and include all three-digit NAICS industries with at least
three public firms.27
We first note that, in contrast to the firm-level results of table 5, the
industry-level results in column 1 of table 6 reveal little association be-
tween industry-level profitability and the use of audited GAAP financial
statements.28 However, we find that industries with high levels of profitabil-
ity dispersion (ROA DISP) have higher rates of audited GAAP use. We infer
from the dispersion results that audited GAAP statements are more useful
when earnings are less certain.
Next, we see no relation between industry-level leverage and private
firms’ audited GAAP production. However, this result is likely driven by
omitted variable issues, as industries with more leverage typically have
more collateralizable assets. Note that this result highlights a central point,
though: in debt contracting for private firms, alternative mechanisms can
substitute for financial statements, diminishing their benefit. In fact, the co-
efficient on industry-level PPE TO ASSETS is significantly negative, which
suggests that the use of audited GAAP statements is muted in the pres-
ence of significant tangible assets that are more easily verified upon visual
inspection.
Finally, we find evidence that industries with high levels of booked
intangible assets (INTAN TO ASSETS), unrecognized intangibles
(R&D TO SALES), and growth opportunities (GROWTH) have signif-
icantly higher rates of audited GAAP statements. MTB is positive, but not
significant at standard levels. These findings are particularly intriguing
given the repeated calls that GAAP’s treatment of intangible assets does
not meet a cost–benefit threshold (FAF [2011]) and that the way GAAP
treats research and knowledge-based firms is not informative to financial
statement users (Lev and Gu [2016]).29 We speculate that growth op-
portunity firms benefit from audited GAAP statements in two ways: (1)
they anticipate raising external capital in the future, and audited GAAP
statements facilitate this process; and (2) financial statement auditors can
27 Specifically, to transform the variable into decile ranks, we create deciles for each variable,
subtract 1 from the decile rank value and divide the result by 9. As such, the variable values
range from 0 to 1.
28 Interestingly, the linear results mask a nonlinear relation. In untabulated results, when
we include a second-order ROA term, the first-order coefficient is significantly negative and
the second-order coefficient is significantly positive. Differentiating with respect to ROA, we
find a minimum point outside of the ROA distribution, such that the relation between indus-
try profitability and audited GAAP statements is monotonically negative through the sample
distribution of ROA, with the relation becoming weaker with profitability.
29 An important caveat—and a detail that future research can consider—is that we are un-
able to measure qualified audit opinions. We suspect that intangible assets are common causes
of qualified opinions. Moreover, firms engaging in activities creating booked intangibles (e.g.,
acquisitions) are likely to benefit from audited GAAP statements for reasons other than their
intangible assets. Our results simply find that booked intangibles are positively rather than
negatively related to audited GAAP statements. We also caution against drawing causal infer-
ences.
TABLE 6
Firm Characteristics of Audited GAAP Financial Statement Production: Industry Level
30
∗∗
GROWTH 0.087 0.054
(2.11) (1.48)
∗∗
R&D TO SALES 0.140∗∗∗ 0.098
(4.60) (2.62)
MTB 0.060 −0.006
(1.36) (−0.11)
∗∗ ∗∗
CONSTANT 0.010 −0.045 0.009 0.033 −0.066∗∗∗ −0.051∗ −0.074∗∗∗ −0.038 −0.119
(0.43) (−2.16) (0.36) (1.61) (−2.79) (−1.96) (−3.65) (−1.56) (−2.49)
Adjusted R2 −0.1% 4.1% −0.3% 5.9% 12.9% 6.4% 23.1% 2.2% 23.0%
Observations 64 64 64 64 64 64 64 64 64
This table analyzes industry-level rates of audited GAAP financial statement production using a linear probability model regression for the 64 nonfinancial three-digit NAICS
industries with at least three firms with sufficient data to calculate all variables. The dependent variable is the estimated fixed effect coefficient for each industry from the firm-level
regression results in table 5, column 1. All independent variables are industry-based measures for publicly held firms sourced from Compustat or CRSP. ROA is net income scaled
by total assets for the average firm in the industry. ROA DISP is the within-industry interquartile range of ROA. LEV is the sum of total short- and long-term debt scaled by total
assets for the average firm in the industry. PPE TO ASSETS is the percentage of total assets composed of net property, plant, and equipment for the average firm in the industry.
INTAN TO ASSETS is intangible assets divided by total assets for the average firm in the industry. GROWTH is the percentage revenue growth for the average firm in the industry.
R&D TO SALES is research & development scaled by total sales for the average firm in the industry. MTB is the market value of assets (market value of equity plus book value of total
debt) divided by the book value of total assets for the average firm in the industry. To facilitate comparability across the coefficients, all variables are placed into deciles and scaled
between [0, 1]. Heteroskedasticity robust t-statistics reported below the coefficient estimates. Continuous variables are winsorized at the 1st and 99th percentile levels. ∗∗∗ , ∗∗ , and ∗
denote significance at the 1%, 5%, and 10% levels, respectively.
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 31
30 Note that our proxies identify both types of intangibles identified by Skinner [2011]:
(1) intellectual capital and lack of physical substance, and (2) identifiable and recognized.
Although our finding that firms in industries with more intangibles are more likely to have
audited GAAP statements suggests that the current approach to accounting for intangibles is
net cost-beneficial for these firms, we are cautious in making this inference. Nevertheless, our
results suggest that the private firm U.S. setting could be a fruitful setting to investigate the
costs and benefits of intangible accounting.
31 For example, INTAN TO ASSETS, GROWTH, and R&D TO SALES are all positively corre-
lated and each of the three variables remains significant in column 9 if the other two are not
included.
32 Although the frequency of beginning audited GAAP statements may seem low, note that
this is for only one year and firms typically do not change this decision frequently. If this rate
is cumulated over a number of years, a substantial portion of the firms would experience a
change. See the online appendix for a Markov transition matrix of changes in both type of
standards and whether the financial reports are audited. We also repeat our analysis for firms
ceasing the production of audited GAAP financial statements. Results are reported in the
online appendix; inferences mirror those reported here.
32 P. LISOWSKY AND M. MINNIS
TABLE 7
Firm Characteristics of Audited GAAP Financial Statement Production: Firm Level Changes
Dependent Variable: BEGIN GAAP AUDIT
statements, while those with more land are less likely. Moreover, rein-
forcing one of the main findings of this paper, we continue to find ev-
idence that high-growth, intangibles-based firms are more likely to view
audited GAAP statements as cost-beneficial. In addition, we find that au-
dited GAAP rates are increasing in property and equipment relative to
total assets, but decreasing in the ratio of property and equipment to
Wages.
In column 1, we also assess the importance of financing-based reasons for
preparing audited GAAP statements by including three indicator variables
measuring changes in equity and debt financing. OWNER INC (DEBT INC)
equals 1 if the number of shareholders (sum of short and long-term ex-
ternal debt) increases from 2008 to 2010, and 0 otherwise. Because equity
transactions can occur among owners (e.g., one owner could sell to one or
more other owners) or between owners and the firm (e.g., the firm could
raise equity by selling to new owners), we also include RAISE EQUITY, a
variable equal to 1 if the amount of capital stock plus additional paid in
capital (as measured on Schedule L) increases between 2008 and 2010,
and 0 otherwise. The results in table 7, column 1, suggest that increases
in owners and equity are positively related to the decision to begin pro-
ducing audited GAAP statements, while increases in debt are not. The co-
efficients on both OWNER INC and RAISE EQUITY are significantly posi-
tive (and economically significant relative to the unconditional mean of
5%), whereas the coefficient on DEBT INC is small and statistically in-
significant. We obtain an almost identical estimate using an indicator
34 P. LISOWSKY AND M. MINNIS
variable for NEW DEBT, equal to 1 if the firm moves from 0 to positive debt
(untabulated).
In column 2 of table 7, we focus only on corporations to use proxies
for inside debt and age. Because we noted in the previous analysis that
the relation between firm age and production of audited GAAP statements
mutes after approximately four years and because we also conduct two-way
sorts based on age in subsequent tests, we dichotomize age as an indica-
tor variable YOUNG equal to 1 if the firm is less than four years old, and 0
otherwise. Consistent with our cross-sectional tests, we find that firms with
more inside debt are less likely to begin producing audited GAAP state-
ments (though changes in inside debt do not coincide with changes in fi-
nancial reporting), while younger firms are more likely to begin producing
audited GAAP statements.
In column 3, we further explore how a firm’s reputation (i.e., age) and
potential for having built relationships could mediate the usefulness of
financial statement production in the presence of raising debt and eq-
uity. Results show that young firms that are adding owners (YOUNG ×
OWNER INC) or raising equity (YOUNG × RAISE EQUITY) are substantially
more likely to initiate audited GAAP statement production. Furthermore,
conditioning on age has no such relation with respect to increases in out-
side or inside debt (i.e., the coefficients on both YOUNG × DEBT INC and
YOUNG × DEBT FROM SH INC are insignificant).33
We further illustrate the economic significance of the interactions be-
tween age and capital raising using contingency tables. In table 8, we parti-
tion the sample from column 3 of table 7 along two dimensions: firm age
and either ownership increases (panel A) or debt increases (panel B).34
Overall, three insights emerge from this analysis. First, audited GAAP state-
ments are not necessary for expanding ownership or attracting external
debt—that is, all cells report values less than 100%. Second, noting that the
right-hand column of cells of the contingency tables in both panels have
the larger values, increases in ownership dispersion and increasing debt
are associated with initiating audited GAAP statements—but, noting that
the lower right hand cell values are substantially larger, this association is
much stronger for young firms. Third, although increasing debt is associ-
ated with increased audited GAAP statement production, this difference is
generally less than half the value of increasing ownership. Overall, these
33 We conduct several additional analyses to further investigate the relation with debt
changes. First, we consider whether investigating the extensive margin matters. We condition
the sample only on firms with zero debt in 2008 and create an indicator variable for those
with at least some amount of debt in 2010. We continue to find no significant results on the
indicator (or with an interaction with firm age). Second, we re-estimate table 7, column 2, but
interact age with DEBT CHANGE (continuous amount of change in debt). Again, we do not
find that conditioning on age produces significant results between debt changes and audited
GAAP changes.
34 See the online appendix for additional analyses partitioning on age and raising new
equity.
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 35
TABLE 8
2 × 2 Contingency Tables Conditional on Age, Ownership, and Debt
Panel A: Partitioning by firm age and change in
Panel B: Partitioning by firm age and change in debt
ownership
Ownership Dispersion Debt
No change/ No change/
decrease Increase Diff decrease Increase Diff
3.5% 5.1% 1.5% 3.4% 4.4% 1%*
Old Old
4,118 514 3,346 1,286
Age
Age
12.0% 33.3% 21.3%** 13.0% 17.0% 4.1%
Young Young
266 30 208 88
Diff 8.5%*** 28.3%*** 29.8%*** Diff 9.6%*** 12.6%*** 13.6%***
This table reports the propensity for corporations to initiate audited GAAP financial statements condi-
tional on the age of the firm and increase in either ownership dispersion or the level of debt. The sample
is restricted to the balanced panel of corporations without audited GAAP financial statements in 2008. The
cells of the contingency tables report the percentage of firms that begin producing audited GAAP financial
statements in 2009. Young firms are those less than four years old. Ownership dispersion is defined as the
number of shareholders of the corporation. Debt is defined as the sum of external short-term and long-term
debt per Schedule L. Changes in ownership dispersion and debt are measured from the year 2008 to the
year 2009. The model specifications exclude firms with number of owners >100. ∗∗∗ , ∗∗ , and ∗ denote sig-
nificance at the 1%, 5%, and 10% levels, respectively, using robust standard errors clustered by three-digit
NAICS industry code.
results along with those of table 7 suggest that young firms bringing in new
owners—a setting with substantially less research than debt explanations—
have a particularly high propensity to begin producing audited GAAP
statements.
APPENDIX
Variable Definitions
IRS Form Source
A P P E N D I X—Continued
IRS Form Source
A P P E N D I X—Continued
IRS Form Source
REFERENCES
ABDEL-KHALIK, R. “Why Do Private Companies Demand Auditing? A Case for Organizational
Loss of Control.” Journal of Accounting, Auditing, and Finance 8 (1993): 31–52.
ALLEE, K., AND T. YOHN. “The Demand for Financial Statements in an Unregulated Environ-
ment: An Examination of the Production and Use of Financial Statements by Privately Held
Small Businesses.” Accounting Review 84 (2009): 1–25.
ANANTHARAMAN, D.; V. FANG; AND G. GONG. “Inside Debt and the Design of Corporate Debt
Contracts.” Management Science 60 (2013): 1260–80.
40 P. LISOWSKY AND M. MINNIS
ARMSTRONG, C.; W. GUAY; AND J. WEBER. “The Role of Information and Financial Reporting
in Corporate Governance and Debt Contracting.” Journal of Accounting and Economics 50
(2010): 179–234.
ASKER, J.; J. FARRE-MENSA; AND A. LJUNGQVIST. “What Do Private Firms Look Like?” 2011.
Available at SSRN: https://ssrn.com/abstract=1659926.
AZAR, J.; M. SCHMALZ; AND I. TECU. “Anticompetitive Effects of Common Ownership.” Journal
of Finance 73 (2018): 1513–65.
BARRIOS, J.; P. LISOWSKY; AND M. MINNIS. “Measurement Matters: Financial Reporting and
Productivity.” Working paper, University of Chicago and Boston University, 2019.
BARTON, J., AND G. WAYMIRE. “Investor Protection Under Unregulated Financial Reporting.”
Journal of Accounting and Economics 38 (2004): 65–116.
BADERTSCHER, B. A.; J. KIM; W. KINNEY; AND E. OWENS. “Verification Services and Fi-
nancial Reporting Quality: Assessing the Potential of Review Procedures” (July 26,
2018). Simon Business School Working Paper No. FR 17–17. Available at SSRN:
https://ssrn.com/abstract=3016422 or https://doi.org/10.2139/ssrn.3016422
BENSTON, G. “The Market for Public Accounting Services: Demand, Supply and Regulation.”
Journal of Accounting and Public Policy 4 (1985): 33–79.
BERGER, P.; M. MINNIS; AND A. SUTHERLAND. “Commercial Lending Concentration and Bank
Expertise: Evidence from Borrower Financial Statements.” Journal of Accounting and Eco-
nomics 64 (2017): 253–77.
BLACKWELL, D.; T. NOLAND; AND D. WINTERS. “The Value of Auditor Assurance: Evidence from
Loan Pricing.” Journal of Accounting Research 36 (1998): 57–70.
BOTOSAN, C.; H. ASHBAUGH-SKAIFE; A. BEATTY; P. DAVIS-FRIDAY; P. HOPKINS; K. NELSON; K.
RAMESH; R. UHL; M. VENKATACHALAM; AND G. VRANA. “Financial Accounting and Reporting
Standards for Private Entities.” Accounting Horizons 20 (2006): 179–94.
BRADSHAW, M.; D. BENS; C.A. FROST; E. GORDON; S. MCVAY; G. MILLER; R. PFEIFFER; M. PLUM-
LEE; C. SHAKESPEARE; W. THOMAS; AND F. WONG. “Financial Reporting Policy Committee of
the American Accounting Association’s Financial Accounting and Reporting Section: Ac-
counting Standard Setting for Private Companies.” Accounting Horizons 28 (2014): 175–92.
BROWN, G.; R. HARRIS; T. JENKINSON; S. KAPLAN; AND D. ROBINSON. “What Do Differ-
ent Commercial Data Sets Tell Us About Private Equity Performance?” Working paper,
2015.https://ssrn.com/abstract=2706556.
CASSAR, G.; C. ITTNER; AND K. CAVALUZZO. “Alternative Information Sources and Information
Asymmetry Reduction: Evidence from Small Business Debt.” Journal of Accounting and Eco-
nomics 59 (2015): 242–63.
CENDROWKSI, H.; L. PETRO; J. MARTIN; AND A. WADECKI. Private Equity: History, Governance, and
Operations, Second edition. Hoboken, NJ: John Wiley & Sons, 2012.
CHOW, C. “The Demand for External Auditing: Size, Debt and Ownership Influences.” Ac-
counting Review 57 (1982): 272–91.
COPPENS, L., AND E. PEEK. “An Analysis of Earnings Management by European Private Firms.”
Journal of International Accounting, Auditing and Taxation 14 (2005): 1–17.
COSTELLO, A. “Mitigating Incentive Conflicts in Inter-firm Relationships: Evidence from Long-
Term Supply Contracts.” Journal of Accounting and Economics 56 (2013): 19–39.
DEDMAN, E.; A. KAUSAR; AND C. LENNOX. “The Demand for Audit in Private Firms: Recent
Large-Sample Evidence from the UK.” European Accounting Review 23 (2014): 1–23.
DIAMOND, D. “Monitoring and Reputation: The Choice Between Bank Loans and Directly
Placed Debt.” Journal of Political Economy 99 (1991): 689–721.
DOIDGE, C.; G. ANDREW KAROLYI; AND RENÉ M. STULZ. “The U.S. Listing Gap.” Journal of Finan-
cial Economics 123 (2017): 464–87.
EDMANS, A., AND Q. LIU. “Inside Debt.” Review of Finance 15 (2011): 75–102.
FINANCIAL ACCOUNTING FOUNDATION (FAF). Blue-Ribbon Panel on Standard Setting for Private
Companies Report to the Board of Trustees of the Financial Accounting Foundation. Stamford, CT:
FAF, 2011.
FINANCIAL ACCOUNTING FOUNDATION (FAF). “Financial Accounting Foundation Establishes
New Council to Improve Standard Setting for Private Companies.” 2012. Available at
PRIVATE U.S. FIRMS AND FINANCIAL REPORTING CHOICES 41
http://www.accountingfoundation.org/cs/ContentServer?c=FAFContent C&pagename=
Foundation/FAFContent C/FAFNewsPage&cid=1176160049265.
GAVER, J.; P. MASON; AND S. UTKE. “Financial Reporting Choices of Private Funds.” Working
paper, 2019. Available at SSRN: https://ssrn.com/abstract=3092331.
GE, W., AND S. MCVAY. “The Disclosure of Material Weaknesses in Internal Control after the
Sarbanes-Oxley Act.” Accounting Horizons 19 (2005): 137–58.
GOVINDARAJAN, V.; S. RAJGOPAL; AND A. SRIVASTAVA. “Why Financial Statements
Don’t Work for Digital Companies.” 2018. Harvard Business Review. Available at
https://hbr.org/2018/02/why-financial-statements-dont-work-for-digital-companies.
GROSSMAN, S. “On the Efficiency of Competitive Stock Markets Where Trades have Diverse
Information.” Journal of Finance 31 (1976): 573–85.
HOLTHAUSEN, R. W., AND R. W. LEFTWICH. “The Economic Consequences of Accounting
Choice Implications of Costly Contracting and Monitoring.” Journal of Accounting and Eco-
nomics 5 (1983): 77–117
HOPE, O. K.; W. THOMAS; AND D. VYAS. “Financial Credibility, Ownership, and Financing Con-
straints in Private Firms.” Journal of International Business Studies 42 (2011): 935–57.
JENSEN, M., AND W. MECKLING. “Theory of the Firm: Managerial Behavior, Agency Costs and
Ownership Structure.” Journal of Financial Economics 3 (1976): 305–60.
KAUSAR, A.; N. SHROFF; AND H. WHITE. “Real Effects of Audit Choice.” Journal of Accounting and
Economics 62 (2016): 157–81.
KOTHARI, S. P.; K. RAMANNA; AND D. SKINNER. “Implications for GAAP from an Analysis of
Positive Research in Accounting.” Journal of Accounting and Economics 50 (2010): 246–86.
LEFTWICH, R. “Discussion of: “Investor Protection under Unregulated Financial Reporting”
(by Jan Barton and Gregory Waymire).” Journal of Accounting and Economics 38 (2004): 117–
28.
LENNOX, C., AND J. PITTMAN. “Voluntary Audits Versus Mandatory Audits.” Accounting Review
86 (2011): 1655–78.
LEV, B., AND F. GU. “The End of Accounting and the Path Forward for In-
vestors and Managers.” 2016. Available at https://onlinelibrary.wiley.com/doi/
book/10.1002/9781119270041
LISOWSKY, P.; M. MINNIS; AND A. SUTHERLAND. “Economic Growth and Financial Statement
Verification.” Journal of Accounting Research 55 (2017): 745–94.
MINNIS, M. “The Value of Financial Statement Verification in Debt Financing: Evidence from
Private U.S. Firms.” Journal of Accounting Research 49 (2011): 457–506.
MINNIS, M., AND N. SHROFF. “Why Regulate Private Firm Disclosure and Auditing?” Accounting
and Business Research 47 (2017): 473–502.
MINNIS, M., AND A. SUTHERLAND. “Financial Statements as Monitoring Mechanisms: Evidence
from Small U.S. Commercial Loans.” Journal of Accounting Research 55 (2017): 197–233.
PETERSEN, M., AND R. RAJAN. “The Benefits of Lending Relationships: Evidence from Small
Business Data.” Journal of Finance 49 (1994): 3–37.
PITCHBOOK. “U.S. PE Middle Market Report 2017 Annual.” 2017. Available at
www.pitchbook.com.
SKINNER, D. “Accounting for Intangibles—A Critical Review of Policy Recommendations.” Ac-
counting and Business Research 38 (2011): 191–204.
SLEE, R. Private Capital Markets, Second edition. Hoboken, NJ: John Wiley and Sons, 2011.
SUNDARAM, R., AND D. YERMACK. “Pay Me Later: Inside Debt and Its Role in Managerial Com-
pensation.” Journal of Finance 62 (2007): 1551–88.
SUTHERLAND, A. “Does Credit Reporting Lead to a Decline in Relationship Lending? Evidence
from Information Sharing Technology.” Journal of Accounting and Economics 66 (2018): 123–
41.
WATTS, R. “Corporate Financial Statements, A Product of the Market and Political Processes.”
Australian Journal of Management 2 (1977): 53–75.
WATTS, R., AND J. ZIMMERMAN. “Towards a Positive Theory of the Determination of Accounting
Standards.” Accounting Review 53 (1978): 112–34.
42 P. LISOWSKY AND M. MINNIS
WATTS, R., AND J. ZIMMERMAN. “Agency Problems, Auditing, and the Theory of the Firm: Some
Evidence.” Journal of Law and Economics 26 (1983): 613–33.
WATTS, R. L., AND J. L. ZIMMERMAN. Positive Accounting Theory. Englewood Cliffs, N.J.: Prentice-
Hall, 1986.
WURSTHORN, M., AND G. ZUCKERMAN. “Fewer Listed Companies: Is That Good or
Bad for Stock Markets?” The Wall Street Journal, January 4, 2018. https://www.wsj.
com/articles/fewer-listed-companies-is-that-good-or-bad-for-stock-markets-1515100040.