Strategic Subsidiary Disclosure

• Author: SCOTT D. DYRENG,JEFFREY L. HOOPES,PATRICK LANGETIEG,JARON H. WILDE
• Date: 01 June 2020
• DOI: http://doi.org/10.1111/1475-679X.12308
• Published date: 01 June 2020

DOI: 10.1111/1475-679X.12308

Journal of Accounting Research

Vol. 58 No. 3 June 2020

Printed in U.S.A.

Strategic Subsidiary Disclosure

SCOTT D. DYRENG,

∗JEFFREY L. HOOPES,

†

PATRICK LANGETIEG,

‡AND JARON H. WILDE

§

Received 16 March 2018; accepted 2 April 2020

ABSTRACT

Although subsidiary disclosures in firms’ filings with the Securities and Ex-

changes Commission (SEC; Exhibit 21) represent the most granular required

public disclosure of a firm’s geographic footprint, little is understood about

the quality of the disclosure, and anecdotal evidence suggests firms may not

fully comply with the disclosure requirements. We use data provided by multi-

national firms to the Internal Revenue Service regarding their foreign sub-

sidiary locations to explore the accuracy of public subsidiary disclosures on

∗Fuqua School of Business, Duke University; †Kenan-Flagler Business School, University

of North Carolina at Chapel Hill; ‡Internal Revenue Service; §Tippie College of Business,

University of Iowa.

Accepted by Douglas Skinner. All data work for this project involving confidential tax-

payer information was done at IRS facilities, on IRS computers, by IRS employees, and at no

time were confidential taxpayer data ever outside of the IRS computing environment. Pro-

fessor Wilde is an IRS employee under an agreement made possible by the Intragovernmen-

tal Personnel Act of 1970 (5 U.S.C. 3371–3376). We thank John Guyton and Lisa Rupert of

the IRS for help and guidance with this project and we thank Barry Johnson, Alicia Miller,

and Michael Weber for facilitating this project through the Joint Statistical Research Program

of the Statistics of Income Division of the IRS. We also thank Ted Christensen, Matt Cob-

abe, Ed deHaan (discussant), Jimmy Downes, Rich Frankel, Michelle Hanlon, Xiumin Martin,

Jon Medrano, Michele Mullaney, Tom Omer, Nemit Shroff, and workshop participants at the

Twelfth Annual Bauer Accounting Research Symposium, MIT, 2018 Colorado Summer Ac-

counting Research Conference, Louisiana State University Regional Workshop Series, Univer-

sity of Illinois-Urbana Champaign, University of Mannheim, University of Nebraska-Lincoln,

University of Oregon, Northwestern, Penn State University, Virginia Tech University, the Se-

curities and Exchange Commission, U.S. Department of the Treasury – Office of Tax Analysis,

Washington State University, and Washington University in St. Louis for their helpful com-

ments. Jaron Wilde gratefully acknowledges support from Thomas and Margaret Kloet and

from the Palmer Fellowship. The views expressed here are ours alone, and do not reflect the

views of the Internal Revenue Service.

643

CUniversity of Chicago on behalf of the Accounting Research Center, 2020

644 S.D.DYRENG,J.L.HOOPES,P.LANGETIEG,AND J.H.WILDE

Exhibit 21 of Form 10-K per SEC rules. The overall incidence of nondisclo-

sure is low, suggesting that most firms comply with Exhibit 21 disclosure rules,

and that for most applications, Exhibit 21 disclosures provide a reasonable

proxy for locations of significant subsidiaries. Nevertheless, there is some

evidence of nondisclosure, particularly when subsidiaries are in tax havens,

when the firm is more highly scrutinized in the media, or when the firm has

other characteristics consistent with low-quality disclosures such as SEC com-

ment letters.

JEL codes: H25; H26; M41; M48

Keywords: financial disclosure; corporate tax; reputational costs of tax

planning

1. Introduction

Subsidiary disclosures in firm filings with the Securities and Exchanges

Commission (SEC) reflect the most granular required public disclosure

of a firm’s geographic footprint. However, we have limited understanding

about the quality of these disclosures, and anecdotal evidence suggests

firms may not fully comply with the disclosure requirements (e.g., Gram-

lich and Whitaker-Poe [2013]). In this study, we undertake the first

comprehensive study of information contained in the SEC’s required list

of “significant” subsidiaries, reported in Exhibit 21 of the Form 10-K, by

comparing locations reported in Exhibit 21 with those from tax filings

made with the Internal Revenue Service (IRS).1We engage in this exercise

for two reasons. First, researchers use Exhibit 21 as a source for information

related to the scope of a firm’s geographic footprint, including as a proxy

for various tax planning strategies requiring legal operations in tax haven

countries (e.g., Law and Mills [2014], Higgins, Omer, and Phillips [2015],

Law and Mills [2017]).2Second, investors use Exhibit 21 to identify “sys-

temic risk, firm interconnectivity” and to “understand complex structures

employed by some firms” (SEC Investor Advisory Committee [2016]),

and reporters and nonprofit organizations use Exhibit 21 disclosures to

scrutinize companies with operations in tax haven countries (e.g., Phillips

et al. [2016]).3Despite this widespread use, surprisingly little is understood

1Firms are required to disclose both foreign and domestic significant subsidiaries on Ex-

hibit 21. We focus only on the disclosure of foreign subsidiaries. We explain the details of

Exhibit 21 reporting requirements in section 2, and in the appendix.

2Numerous studies use Exhibit 21 data. See, for example, (Dyreng and Lindsey [2009],

Lisowsky [2010], Dyreng, Hanlon, and Maydew [2012], Dyreng, Lindsey, and Thornock

[2013], Black, Dikolli, and Dyreng [2014], Law and Mills [2014], Dyreng et al. [2015], Han-

lon, Lester, and Verdi [2015], Demere, Donohoe, and Lisowsky [2016], Dyreng and Markle

[2016], Akamah, Hope, and Thomas [2017], Bozanic et al. [2017], Chow, Hoopes, and May-

dew [2017], De Simone, Mills, and Stomberg [2017], Dyreng et al. [2017], Heckemeyer,Olligs,

and Overesch [2017], Law and Mills [2017]).

3See the letter from the SEC Investor Advisory Committee, https://www.sec.gov/spotlight/

investor-advisory-committee-2012/iac-approved-letter-reg-sk-comment-letter-062016.pdf.

STRATEGIC SUBSIDIARY DISCLOSURE 645

about the information contained in Exhibit 21. Are firms complying with

the disclosure requirements as set forth by the SEC? Do firms strategically

decide which subsidiaries to disclose and which to omit? What factors drive

the decision to disclose versus omit a significant subsidiary from Exhibit 21?

The answers to these questions are not clear. On the one hand, the infor-

mation contained in Exhibit 21 is sparse, and unlikely to provide taxing au-

thorities or other regulators with new information (Donohoe, McGill, and

Outslay [2012]). Furthermore, Gallemore, Maydew, and Thornock [2014]

find very little evidence of negative reputational consequences to tax avoid-

ance and Chen, Schuchard, and Stomberg [2019] find that firms do not

change their tax avoidance activities in response to media scrutiny. If the

information in Exhibit 21 is sufficiently benign so as not to alarm taxing au-

thorities or other regulators, or if the value to investors is sufficiently large,

then it is likely firms would comply with SEC requirements.

On the other hand, even though Chen, Schuchard, and Stomberg [2019]

find little evidence that firms change their tax avoidance following nega-

tive media coverage, indeed, the information firms disclose on Exhibit 21

to the public at large could provide material for reputation-damaging arti-

cles in the popular press and protests by concerned citizens. For example,

Hanlon and Slemrod [2009] find evidence of (weak) negative market reac-

tions when firms are accused of tax sheltering in the press. Similarly, Austin

and Wilson [2017] document that firms with valuable consumer reputa-

tions have a lower probability of engaging in tax shelters. Anecdotally, the

BBC has reported that Starbucks, Google, and Amazon have all been tar-

gets of public tax shaming episodes (see Barford and Holt [2013] and De

Simone et al. [2016] for examples of tax shaming). If these costs are large

enough, it is possible some firms will strategically omit disclosure of some

subsidiaries.

Overall, our evidence suggests that most firms comply with the Exhibit

21 disclosure requirements. Specifically, we examine whether the disclo-

sures on Exhibit 21 are an accurate representation of the firm’s significant

subsidiaries. We examine the information contained in Exhibit 21 to deter-

mine whether firms disclose information as required by the SEC, and if not,

why not. To do this, we compare foreign subsidiary information contained

in Exhibit 21 with a comprehensive data set of U.S. multinational firms’

foreign subsidiaries obtained from IRS tax filings for the years 2006–15.4

In contrast to potential omissions on firms’ Exhibit 21, which carry trivial

4We use data from the universe of Form 5471, “Information Return of U.S. Persons With

Respect to Certain Foreign Corporations” filed both electronically and on paper with the

IRS. Form 5471 is only required for subsidiaries organized as corporations, but Exhibit 21 is

required for all subsidiaries regardless of legal form (e.g., LLC, LP), which many public firms

have and often use as special purpose entities (Feng, Gramlich, and Gupta [2009]). Thus, all

subsidiaries of sufficient size found in the IRS data should be disclosed on Exhibit 21, but

some noncorporate subsidiaries disclosed on Exhibit 21 will not be present in the IRS data

we use. To this end, we note that our estimates represent a lower bound of nondisclosure in

Exhibit 21. We also note that Form 8865, “Return of U.S. Persons with Respect to Certain