Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25

Contents lists available at SciVerse ScienceDirect

Journal of International Accounting,

Auditing and Taxation

The determinants of thinly capitalized tax avoidance

structures: Evidence from Australian firms
Grantley Taylor a,∗ , Grant Richardson b,1
a
School of Accounting, Curtin Business School, Curtin University, GPO Box U1987, Perth 6845, Western Australia, Australia
b
Discipline of Accounting and Information Systems, Business School, The University of Adelaide, 10 Pulteney Street, Adelaide 5005,
South Australia, Australia

a r t i c l e i n f o a b s t r a c t

Keywords: This paper examines the determinants of thinly capitalized structures of publicly-listed
Thin capitalization Australian firms. Based on a hand-collected sample of 203 publicly-listed Australian firms
Tax avoidance
over the 2006–2009 period (812 firm-years), our regression results indicate that the thin
Corporate governance
capitalization position of firms is significantly and positively associated with multination-
Australia
ality, tax haven utilization, withholding taxes and tax uncertainty. Multinationality and
the use of tax havens are, in particular, strongly associated with thin capitalization. Our
additional regression results provide evidence that shows that corporate governance mon-
itoring mechanisms relating to board of director independence, institutional ownership and
big-4 auditor utilization are significantly negatively associated with firms adopting thinly
capitalized tax avoidance structures.
© 2013 Elsevier Inc. All rights reserved.

1. Introduction

International corporate tax avoidance2 carried out by multinational firms represents a significant problem for many
western countries including Australia, Canada, the UK and the US (see, e.g., Benshalom, 2008; Christensen & Murphy, 2004).3
Large losses in tax revenue due to the tax deductibility of high levels of debt have encouraged many developed countries to
adopt thin capitalization rules to protect their domestic tax bases (Dahlby, 2008; Richardson, Hanlon, & Nethercott, 1998;
Smith, 1996). In fact, tax authorities such as the Australian Taxation Office (ATO) have long been concerned with the erosion
of tax revenue through the shifting of debt between related parties domiciled in variably taxed jurisdictions (ATO, 2002,
2004a, 2006). Specifically, a multinational firm has an incentive to finance its foreign direct investment with debt if the
host-country’s corporate tax rate is higher than that of the home country because interest payments are tax deductible
at a higher tax rate if the borrowing is carried out by a foreign subsidiary rather than by the parent firm (Dahlby, 2008).
Research by Dahlby (2008) shows that both internal (related-party) and external (third-party) debt of foreign subsidiaries
is responsive to tax rate differentials.

∗ Corresponding author. Tel.: +61 8 9266 3377; fax: +61 8 9266 7196.
E-mail addresses: [Link]@[Link] (G. Taylor), [Link]@[Link] (G. Richardson).
1
Tel.: +61 8 8313 0582; fax: +61 8 8223 4782.
2
Consistent with existing tax research (see, e.g., Chen et al., 2010; Frank et al., 2009; Lanis & Richardson, 2011), we define corporate tax avoidance as
the downward management of taxable income through tax planning activities. The practice encompasses tax planning activities that are legal, those that
may fall into a gray area and those that are illegal, and thus the term is broadly defined. Moreover, although we use the term “tax avoidance” throughout
the paper, it can be used interchangeably with “tax aggressiveness” and “tax management.”
3
Jog and Tang (2001) provide empirical evidence suggesting that the increase in the debt-to-asset ratio of US subsidiaries operating in Canada between
1984 and 1994 may have reduced the amount of corporate tax paid by all US foreign-controlled firms by at least US$1.3 billion.

1061-9518/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
[Link]
 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25 13

Thin capitalization refers to a corporate investment decision by the firm to finance business operations primarily with debt
capital rather than equity capital in its capital structure (Richardson et al., 1998; Taylor & Tower, 2009; Taylor, Tower, & Van
der Zahn, 2010). The excessive use of debt financing in the form of thinly capitalized structures by subsidiary firms located
in higher tax jurisdictions constitutes an important international corporate tax avoidance technique used by multinational
firms (Shackelford & Shevlin, 2001; Shackelford, Slemrod, & Sallee, 2007).4 In Australia, the thin capitalization rules are
aimed at preventing multinational taxpayers from allocating a disproportionate amount of debt in the capital structures of
their Australian operations by disallowing interest deductions for excessive debt financing (Richardson et al., 1998; Taylor
et al., 2010). These rules constitute anti-avoidance tax provisions that apply to firms whose assets are funded by a high level
of debt capital and a relatively low level of equity capital in their capital structures.5 Method statements in the Australian
tax legislation outline the process by which a firm can calculate the maximum amount of interest-bearing debt that can
generate interest deductions in a given fiscal year (referred to as the “maximum allowable debt”). A thinly capitalized firm
is a firm with a level of debt that exceeds 75% of its total debt plus equity (referred to as the “safe harbor limit”). Australia’s
tax provisions6 rely on accounting standards to determine what constitutes assets, liabilities and equity when computing a
firm’s thin capitalization position. An accurate and objective measure of a firm’s thin capitalization position and whether it
has breached the safe harbor limit can thus be determined by using financial statement data from the annual report.
This study investigates the determinants of thinly capitalized structures of publicly-listed Australian firms. Based on a
hand-collected sample of 203 publicly-listed Australian firms over the 2006–2009 period (812 firm-years), our regression
results show that firms’ thin capitalization positions are significantly and positively associated with multinationality, tax
haven utilization, withholding taxes and tax uncertainty. Our additional regression results provide evidence that shows that
corporate governance monitoring mechanisms relating to board of director independence, institutional ownership and big-4
auditor employment are significantly negatively associated with firms’ adopting thinly capitalized tax avoidance structures.
Our study makes several contributions to the literature. First, while there are several past studies that have examined the
link between taxes and debt sourcing and use, and the operation of the thin capitalization provisions in general, to the best
of our knowledge there is no empirical research that examines explicitly the determinants of thin capitalization. Previous
research relating to thin capitalization has examined this relationship in general terms from a tax compliance perspective
without quantitatively assessing firms’ thin capitalization positions. Strategic use of debt in a particular jurisdiction such as
Australia, which has a relatively high statutory corporate tax rate (i.e., 30%), can result in significant tax benefits. Thus, it
important to understand some of the potential underlying factors that may lead firms to include significantly more debt in
their capital structure compared to their industry peers. Our research therefore contributes to an understanding of the role
that taxes play in a firm’s capital management decisions. Second, this study utilizes a novel measure of thin capitalization
based on a method statement contained in the Australian tax legislation. This metric provides an objective means to ascertain
how firms employ tax deductible debt to reduce corporate taxes. In fact, use of the method statement in the Australian tax
legislation, together with reliance on the accounting definition of financial statement elements, permits replication of our
tax legislation-based thin capitalization measures in other jurisdictions, such as Canada, New Zealand and the UK. Third, this
study also provides some unique empirical evidence that shows that corporate governance monitoring mechanisms pertain-
ing to board of director independence, institutional ownership and big-4 auditor utilization are negatively associated with
firms adopting thinly capitalized structures. Finally, our findings provide policymakers and regulators with valuable infor-
mation about the most important determinants of thinly capitalized structures of firms, as well as the corporate governance
monitoring mechanisms that may assist in controlling them.
The rest of the paper is organized as follows. Section 2 provides a brief review of the literature on international corporate
tax avoidance. Section 3 considers the relevant theory in this area of research and develops hypotheses. Section 4 reports
the research design, and Section 5 summarizes and analyzes the empirical results. Finally, Section 6 concludes the paper.

2. Literature review

Research by Slemrod (2001) and Rego (2003) finds that highly leveraged firms in the US have lower effective tax rates
(ETRs) because of the use of debt deductions to reduce taxable income. Similarly, Beuselinck, Buysschaert, and Deloof (2005),
Graham and Tucker (2006) and Dyreng, Hanlon, and Maydew (2008) all find that successful long-run tax-avoidant firms in
the US have significantly higher leverage in their capital structures.
Mills and Newberry (2004) find the taxable income levels of US foreign controlled subsidiaries (FCC’s) vary significantly
in line with the worldwide tax incentives of their foreign parent firms. Foreign multinational firms with relatively low
foreign tax rates are found to have sourced more debt and reported less taxable income in their US FCCs than those with
relatively high average foreign tax rates. The authors provide evidence that suggests that FCCs source debt in the US, where

4
Multinational groups rely substantially on debt financing due to the tax deductibility of interest payments, economic factors that render it far easier,
more efficient and less costly for a firm to raise capital through the use of debt capital rather than equity capital, the maintenance of credit ratings, the
fungibility of debt in that borrowings can be readily transferred across jurisdictions and between group subsidiaries, and/or a mixture of these factors
(Benshalom, 2008; Smith, 1996).
5
The Australian thin capitalization rules are contained in Division 820 of the Income Tax Assessment Act 1997 (ITAA97).
6
Section 820-680 of ITAA97.
14 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25

greater tax deductions are available on the interest payable on that debt. Similar evidence of tax-motivated income shifting
is found by Newberry and Dhaliwal (2001) who show that US multinational firms’ source interest deductions in overseas tax
jurisdictions to shift income, and that tax incentives influence managerial decisions about the location of debt. Graham and
Harvey (2001) find that the interest deductibility of debt is a moderately important determinant of firms’ capital structure.
Desai, Foley, and Hines (2004) also find that a 10% increase in a host-country’s tax rate increases the foreign subsidiary’s
debt-to-asset ratio by 2.8%. According to Desai et al. (2004), internal borrowing is particularly sensitive to host-country tax
rates.
Mintz and Weichenrieder (2005) analyze German subsidiaries operating in 68 countries and find that a one percentage
point increase in a host country’s tax rate increases the debt-to-asset ratio of a wholly-owned foreign subsidiary by 0.30–0.57
percentage points. Beattie, Goodacre, and Thomson (2006) also note that UK firms’ choice of financing decisions is based
on target debt levels that are, in turn, related to the long-term survival of the firm, the maintenance of earnings and cash
flows, restrictive covenant conditions, interest rates and the tax deductibility of interest payments. Overall, Beattie et al.
(2006) provide evidence of the important role that taxes play in UK firms financing decisions. Finally, Huizinga, Laeven, and
Nicodeme (2008) present evidence that European firms’ indebtedness depends on domestic tax rates and tax rate differences
between the parent and subsidiary firms.
In summary, a firm’s international financing strategy has significant income tax implications. Because of Australia’s
geographic isolation and small capital market, many Australian firms regularly carry out debt financing internationally
(especially in the US) to fund their domestic business operations. Therefore, debt-related tax strategies play a significant
role in the financing decisions of Australian firms. In this paper, we argue that several variables play an important role in
determining the thin capitalization position of Australian firms, including multinationality, tax haven utilization, withholding
taxes and tax uncertainty. In the following section, we discuss the rationale and literature support for each of these variables,
and develop our hypotheses for empirical testing.

3. Theory and hypotheses development

3.1. Multinationality

Australia’s tax legislation is based on the general principle that resident taxpayers are subject to taxation on their gross
income from all sources either inside or outside Australia.7 Nevertheless, there are also specific tax provisions that cover
international corporate dealings. For Australian resident firms deriving foreign-sourced income, there are several tax impli-
cations that should be considered. For example, related firms can take advantage of their group operating structure and tax
rate differentials across jurisdictions to shift income between group members to minimize the overall tax burden of the
corporate group (Beuselinck et al., 2005).
Given that multinational firms normally apply efficient tax planning across group entities, it is expected that firms with
subsidiaries in the corporate group that derive income from foreign sources will engage in greater tax avoidance activities. In
fact, Rego (2003) finds that US multinational firms are more successful at avoiding corporate taxes than purely domestic firms.
Moreover, Hanlon, Mills, and Slemrod (2005) report that US FCCs have more than doubled the levels of tax non-compliance
relative to domestic controlled firms. Finally, Dyreng et al. (2008) show that US firms with greater international exposure
tend to have more opportunities available to engage in tax avoidance strategies. To test the impact of multinationality on
thin capitalization, we develop the following hypothesis:

H1. All else being equal, corporate multinationality is positively associated with thin capitalization.

3.2. Tax haven utilization

Tax avoidance also can be facilitated if a member of the corporate group is resident in a jurisdiction with tax haven
status that offers advantageous financial, legal and taxation treatment (ATO, 2004a; OECD, 2006).8 Tax havens may impose
either no or negligible amounts of corporate tax, have laws or administrative practices that prevent the effective exchange of
information, and lack transparency with regard to financial and tax arrangements (e.g., regulatory, legal and administrative
provisions and access to financial records). Tax havens can also promote tax avoidance by permitting the reallocation of
taxable income from high taxed jurisdictions to low taxed jurisdictions, and by reducing the amount of domestic tax levied
on foreign income (Hanlon & Slemrod, 2009; OECD, 2006).

7
See Sections 6-5 and 6-10 of ITAA97. Australian resident firms are taxed on their foreign-sourced income unless specifically exempted from doing so.
8
The OECD (2006) identifies three key factors that determine whether a jurisdiction constitutes a tax haven as follows: (1) it levies no or a nominal
amount of tax; (2) it lacks an effective information exchange mechanism; and (3) it is characterized by a lack of transparency. Tax havens differ in the extent
to which these particular characteristics apply. The OECD (2006) recognizes a total of 33 tax havens around the world, which are reported in Appendix A
of the paper.
 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25 15

According to Desai (2006) and Desai, Foley, and Hines (2006), it is possible for tax avoidant firms to incorporate controlled
entities in tax havens in order to significantly avoid domestic taxes.9 Firms can exploit the secrecy laws of tax havens in an
attempt to conceal income that may otherwise be subject to domestic taxation and take advantage of the lack of effective
information exchange and transparency in these havens (Desai, 2006; OECD, 2006). In fact, Harris, Morck, Slemrod, and
Yeung (1993) find US tax liabilities (as a percentage of US sales or assets) to be lower in their sample of US multinational
firms with a legal presence in a tax haven.10
Firms incorporated in a country with tax haven status may play an important role for the entire corporate group.
For instance, they may control the treasury, insurance, business and service functions for the corporate group or
assist in the tax-efficient transfer of funds between group members.11 Thus, efficient tax planning across group
entities involving tax haven-incorporated firms could have a significant impact on the transparency and accountabil-
ity of the corporate group (ATO, 2004a; Wilson, 2009). A lack of transparency and accountability at the corporate
group level may, in turn, influence the corporate mindset on such issues as the provision of information to tax
authorities.
Desai and Hines (2002) show that firms incorporated in tax havens are able to shift income from high tax jurisdic-
tions to low tax jurisdictions through the use of transfer pricing, inter-firm debt and the transfer of intangible assets. It
is expected that special purpose finance, investment or insurance incorporated subsidiaries in tax havens are likely to
assist in the avoidance of corporate taxes. In fact, a multinational enterprise could use a financing entity in a tax haven to
secure tax deductions for interest payments made on debt by its subsidiaries in high-tax countries (Richardson et al., 1998;
Slemrod & Wilson, 2009; Wilson, 2009). To test the impact of tax havens on thin capitalization, we develop the following
hypothesis:

H2. All else being equal, tax haven utilization is positively associated with thin capitalization.

3.3. Withholding taxes

Withholding tax is charged on Australian interest (10%), unfranked dividends (30%) and royalty income (30%) paid to
non-resident firms. In considering withholding taxes, it is necessary to determine whether an investment return constitutes
debt interest or equity interest.12 This categorization determines whether a return paid by a firm on a financing interest that
it has issued is deductible (i.e., treated as interest on debt) or frankable (i.e., treated as a dividend). If it is deemed to be a return
on debt interest, then it is subject to interest withholding tax. However, if it is deemed to be a return on an equity interest,
then it is subject to dividend withholding tax. Accordingly, this characterization determines whether interest withholding
tax or dividend withholding tax is payable on non-resident distributions or whether an instrument is considered to be debt
capital for thin capitalization purposes.
The treatment of a payment as debt interest or equity interest can vary across different tax jurisdictions depending on
the character of the payment, so it is possible that lower amounts of withholding tax could be paid. Intragroup transfers
of debt, dividends and royalties can be structured to reduce the amount of tax payable by characterizing a transaction as
an interest payment rather than as a dividend payment or royalty payment (Collins & Shackelford, 1998). Alternatively, the
payment could be characterized as a payment for services, thereby avoiding the payment of withholding taxes altogether.
If thin capitalization is integrated with a firm’s financing decisions in an attempt to avoid corporate taxes, then funds may
be borrowed in high tax jurisdictions to receive a tax deduction on interest payments in those jurisdictions. The borrowed
funds may then be transferred to a low taxed jurisdiction for corporate use. Nevertheless, a firm may borrow funds at a
favorable interest rate in a specific jurisdiction and then novate the loans to firms domiciled in a high taxed jurisdiction to
receive larger tax deductions on interest payments.
Notwithstanding the details of the financing arrangements undertaken by a firm, it is likely that the development of
a thinly capitalized structure will involve some form of fund transfers across variably taxed jurisdictions. Fund transfers
typically involve the payment of withholding taxes. Therefore, if withholding taxes apply, then there is a greater likelihood
that the firm could be employing a thinly capitalized structure to avoid corporate taxes. To test the impact of withholding
taxes on thin capitalization, we develop the following hypothesis:

H3. All else being equal, corporate withholding taxes are positively associated with thin capitalization.

9
However, it is possible that controlled entities incorporated in tax havens may also be established for legitimate business purposes and/or because they
represent the lower taxed location of several possible locations in which that business could be conducted.
10
Nevertheless, it should be noted that not all of the empirical research in this area shows that tax haven affiliates help firms to achieve lower corporate
taxes. In fact, Dyreng and Lindsey (2009) investigate the differences in the domestic, foreign and global tax rates that apply to US corporate groups with
and without tax haven incorporated entities. They find that while having at least one foreign operation located in a tax haven incrementally reduces the
global tax rate for US firms from 36% to 34.5% on average, many of the firms with tax haven links are also associated with high (US) domestic tax rates on
their foreign income.
11
Many transactions involving firms incorporated in tax havens may occur for legitimate business purposes (ATO, 2004a). However, information trans-
parency relating to these transactions could still be lacking.
12
The rules for defining what constitutes equity or debt are set-out in Division 974 of ITAA97.
16 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25

3.4. Tax uncertainty

Corporate taxes affect the financing choices, restructuring decisions, capital appropriations, compensation arrangements
and the day-to-day operational decisions of a firm (Chen, Chen, Cheng, & Shevlin, 2010; Desai & Dharmapala, 2006). Man-
agement may face significant uncertainty in deriving tax estimates based on differing interpretations of taxation law (Desai
& Dharmapala, 2006; Frischmann, Shevlin, & Wilson, 2008; Slemrod, 2004) and calculation of tax assets and tax liabilities. In
fact, a firm could face significant uncertainty in estimating tax assets (e.g., estimates of deferred tax assets based on meeting
the probability criteria that tax losses can be recouped at a future point in time) or tax liabilities (e.g., provisions for taxes
based on the global variability in corporate taxes payable).
In Australia, firms are required to disclose any tax-related contingent assets and liabilities in accordance with Australian
Accounting Standard AASB 137, Provisions, Contingent Liabilities and Contingent Assets. Contingent assets and liabilities may
arise because of unresolved tax disputes with the ATO. Australian firms recognize liabilities for anticipated tax audit issues
based on management’s estimate of whether additional taxes will be payable sometime in the future. This situation will
occur if the final tax outcome differs from current and deferred tax estimates in the present accounting period.
Tax uncertainty may be used by a firm’s management as a tool for camouflaging or masking tax avoidance activities
(Desai & Dharmapala, 2006), including the use of thinly capitalized structures to reduce the overall amount of corporate
tax payable.13 In fact, the derivation of a tax benefit through various tax avoidance arrangements requires complexity
and obfuscation to prevent detection, thereby shielding managerial opportunism with regard to corporate tax avoidance.
Obfuscation may be achieved through concealment of obligations, earnings manipulation, diversion of transactions, secrecy
and through the creation of an environment of uncertainty in deriving tax estimates (Desai & Dharmapala, 2007). To test
the impact of tax estimate uncertainties (“tax uncertainty”) on thin capitalization, we develop the following hypothesis:

H4. All else being equal, corporate tax uncertainty is positively associated with thin capitalization.

4. Research design

4.1. Sample selection and data source

We examine the determinants of thin capitalization over the 2006–2009 period. These four years were chosen because
they represent the most recent years for which financial statement data are currently available. Our initial sample comprised
the top 300 Australian Stock Exchange (ASX) listed firms as of December 31, 2009. This number was reduced to 203 after
excluding: financial firms (39); insurance firms (11); US firms subject to GAAP (16); property partnership or trust entities
(11); and firms that had not reported financial information across all four years because they were newly incorporated,
taken-over or merged with other firms during the study period (20).14 Finally, all data were hand-collected from the annual
reports of all sample firms.

4.2. Dependent variable

Under Australia’s thin capitalization provisions, accounting standards are used to determine what constitutes assets,
liabilities and equity.15 The thin capitalization provisions outline the process by which an entity can calculate the maximum
amount of interest-bearing debt (MAD) that can give rise to interest deductions in a fiscal year.16 A thinly capitalized
entity is an entity with a level of debt in its capital structure that exceeds 75% of its total debt plus equity. This metric is
known as the “safe harbor limit.” We calculate a firm’s thin capitalization position by utilizing the safe harbor test,17 which
involves the calculation of the safe harbor debt amount (SHDA) using the method statement outlined in Section 820-95 (for
outward investing general entities) of ITAA97. The SHDA is 75% of the average asset value of Australian operations net of

13
Moreover, a firm’s use of complex tax structures can produce sufficient obscurity to facilitate managerial self-dealing (see, e.g., Desai & Dharmapala,
2006; Dharmapala & Riedel, 2011).
14
Finance and insurance firms are subject to different thin capitalization rules compared to other firms and are also subject to unique regulatory require-
ments compared to other firms. Moreover, firms reporting under US GAAP and property partnership or trust entities are also excluded from the sample
because of differing reporting requirements.
15
Taxation Ruling TR 2002/20 Income tax: Thin Capitalization – Definition of assets and liabilities for the purposes of Division 820 states that: “It is also
appropriate to adopt the accounting meaning of the term ‘liabilities’ for the purposes of Division 820” (Paragraph 5). It continues: “In more specific terms,
the safe harbor debt amount calculations (and the worldwide gearing debt amount calculation which is based on the safe harbor debt amount) determine
the safe harbor debt/equity mix based on the entity’s assets and liabilities. The calculation is broadly based on the fundamental accounting equation of
assets = liabilities + equity.”
16
The fixed safe harbor gearing ratio is adopted as the first-tier test. Then, if exceeded, an arm’s length test or test based on the worldwide gearing limit
is applied.
17
Determination of the MAD is not made using the arm’s length test or worldwide gearing ratio, given that these methods are reliant on firm-specific
assumptions and factors.
 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25 17

non-interest-bearing liabilities and investments in associates.18 The method statement provided in Section 820-95 of ITAA97
(ATO, 2004b) is summarized in Appendix B of the paper.
Measures for each of the variables used to calculate the SHDA are computed as follows:

SHDA = (average total assets − non-IBL) × 75%, (1)

where non-IBL = the non-interest-bearing liabilities of the firm.

We also calculate a measure of average debt as follows:

Average debt = total interest − bearing liabilities (IBL) of the firm. (2)

Finally, a measure of MAD is computed as follows:

average debt
MAD ratio = . (3)
SHDA of the firm
The MAD ratio computed in accordance with Eq. (3) represents our first measure of thin capitalization (THINCAP1).
To improve the robustness of our empirical results, we compute a second measure of thin capitalization (THINCAP2)
which is based on a dummy variable of 1 if the firm has breached the MAD position (i.e., MAD > 1) in accordance with the
thin capitalization provisions, otherwise 0. A firm with a MAD ratio in excess of that based on Eq. (3) is considered potentially
non-compliant with the thin capitalization provisions because its average debt level exceeds its SHDA.

4.3. Independent variables

The independent variables in this study are represented by multinationality, OECD recognized tax haven, withholding
tax and tax uncertainty. Our proxy measure of multinationality (MULTI) is a dummy variable of 1 if the firm has at least one
subsidiary firm that is incorporated outside Australia, otherwise 0. The predicted sign for MULTI is positive. OECD recognized
tax haven (TAXHAV) is measured as a dummy variable of 1 if the firm has at least one subsidiary firm that is incorporated
in an OECD recognized tax haven, otherwise 0.19 The predicted sign for TAXHAV is positive. Withholding taxes (WTAX) is
measured as a dummy variable of 1 if the firm is subject to Australian withholding taxes, otherwise 0, and its predicted
sign is positive. Finally, our proxy measure for tax uncertainty (UNCERT) is measured as a dummy variable of 1 if the firm
has issued a statement noting that it has major difficulties in estimating its tax liability, otherwise 0. The predicted sign for
UNCERT is positive.

4.4. Control variables

We also include several control variables in our base regression model to control for other effects, including: foreign
exposure (FOR), firm size (SIZE), the market-to-book ratio (MKTBK), return on assets (ROA), capital intensity (CAPINT),
inventory intensity (INVINT), R&D intensity (RDINT), industry sector effects (INDSEC) and year effects (YEAR).
We include the extent of foreign exposure (FOR) as a control variable in our base regression model in keeping with
previous research which finds that FOR is positively associated with corporate tax avoidance (see, e.g., Mills & Newberry,
2004; Rego, 2003). FOR is measured in our study as the proportion of total foreign sourced income divided by total income
and its predicted sign is positive.
SIZE also is included in the base regression model given that Rego (2003) argues that larger firms can achieve economies
of scale via tax planning and have the incentives and resources readily available to them to reduce the amount of corporate
taxes payable. Hanlon et al. (2005) also find that larger firms tend to have greater tax deficiencies relative to their actual tax
liability. We measure SIZE as the natural log of total assets, and its predicted sign is positive.
MKTBK is included in our base regression model to control for growth (see, e.g., Adhikari, Derashid, & Zhang, 2006; Gupta
& Newberry, 1997; Lanis & Richardson, 2011). However, due to the inconsistent results obtained with MKTBK in previous
tax avoidance research (see, e.g., Adhikari et al., 2006; Gupta & Newberry, 1997), no sign prediction is made for this variable.
ROA, measured as pre-tax profit divided by total assets, is included in our base regression model to control for operat-
ing performance or profitability of the firm (Gupta & Newberry, 1997). No sign prediction is made for ROA owing to the

18
As an example, ABC Ltd. is a listed Australian firm with an average asset value of AUD$100 million. The average values of its relevant associate entity
debt, associate entity equity, controlled foreign debt, controlled foreign entity equity and non-debt liabilities are AUD$10 million, AUD$8 million, AUD$5
million, AUD$2 million and AUD$5 million, respectively. Deducting these amounts from the average asset value leaves AUD$70 million. Multiplying AUD$70
million by 3/4 results in AUD$52.5 million, the SHDA. The average debt amount (predominantly IBL) is compared to the SHDA. If the average debt amount is
greater than AUD$52.5 million, then debt deductions on the excess amount may be denied. The proxy measure of the SDHA and determination of whether
an entity can be regarded as thinly capitalized are based on the accounting definition of assets and liabilities (ATO, 2002, 2004b, 2006). Finally, the thin
capitalization tax provisions rely on the valuation rules in the accounting standards to provide the values of assets and non-debt liabilities (ATO, 2002).
19
Other methods of measuring THAV were included in our regression models, such as the proportion of tax haven incorporated subsidiaries. Our results
were not significant. In the case of the small-to-medium sized firms (e.g., firms with less than AUD$600 million total assets), tax avoidance may be assisted
greatly through the use of only one tax haven, especially if that tax haven provides a treasury function or is the financing entity for the group. Our sample
comprises approximately 34% of firms which could be regarded as small-to-medium in size.
18 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25

inconsistent findings in previous tax avoidance research for this variable (see, e.g., Adhikari et al., 2006; Gupta & Newberry,
1997).
CINT and INVINT are also included as control variables in our base regression model for high capital intensive or inventory
intensive firms, respectively (Stickney & McGee, 1982). CINT is positively associated with tax avoidance due to accelerated
depreciation charges based on asset lives (Stickney & McGee, 1982). Moreover, to the extent that INVINT is a substitute for
CINT, inventory intensive firms should be less tax avoidant than capital intensive firms, so INVINT is positively associated
with tax avoidance (Stickney & McGee, 1982). We measure CINT as net property, plant and equipment divided by total assets
and INVINT as inventory divided by total assets. We also include RDINT as a control variable in our base regression model as
previous research (e.g., Gupta & Newberry, 1997; Richardson & Lanis, 2007) finds that RDINT is positively associated with
tax avoidance due to R&D expenditure being tax-deductible. We measure RDINT as R&D expenditure divided by total assets,
and its predicted sign is positive.
We include nine INDSEC dummy variables based on the two-digit Global Industry Classification Standard (GICS) codes
as control variables in our base regression model because it is possible that thin capitalization fluctuates across industry
sectors (Dyreng et al., 2008). They are denoted by the energy, materials, industries, consumer discretionary, consumer
staples, health care, information technology, telecommunications and utilities20 industry sectors. No sign predictions are
made for the INDSEC dummies.
Finally, year (YEAR) dummy variables are also included in our base regression model to control for differences in tax
avoidance activities that could possibly exist across the 2006–2009 sample years.21 No sign predictions are made for the
YEAR dummies.

4.5. Base regression model

The base regression model employed to examine the major determinants of thinly capitalized tax avoidance structures
is as follows:

THINCAPit = ˛0 + ˇ1 MULTIit + ˇ2 TAXHAVit + ˇ3 WTAXit + ˇ4 UNCERTit + ˇ5 FORit + ˇ6 SIZEit + ˇ7 MKTBKit

+ ˇ8 ROAit + ˇ9 CINTit + ˇ10 INVINTit + ˇ11 RDINTit + ˇ12−19 INSECit + ˇ20−22 YEARit + εit (4)

where:

i = firms 1 through 203;

t = financial years 2006–2009;
THINCAP = thin capitalization proxy measures (THINCAP1 and THINCAP2);
MULTI = dummy variable of 1 if the firm has at least one subsidiary firm that is incorporated outside of Australia, otherwise
0;
TAXHAV = dummy variable of 1 if the firm has at least one subsidiary firm that is incorporated in an OECD recognized tax
haven, otherwise 0;
WTAX = dummy variable of 1 if the firm is subject to Australian withholding taxes, otherwise 0;
UNCERT = dummy variable of 1 if the firm has issued a statement noting that it has significant uncertainties in estimating
its tax liability, otherwise 0;
FOR = foreign income divided by total income;
SIZE = the natural logarithm of total assets;
MKTBK = market value of equity divided by book value of equity;
CINT = net property, plant and equipment divided by lagged total assets;
INVINT = inventory divided by lagged total assets;
RDINT = R&D expenditure divided by lagged total assets;
ROA = pre-tax income divided by total assets;
INDSEC = industry sector dummy variable of 1 if the firm is represented in the specific GICS category, otherwise 0;
YEAR = year dummy variable of 1 if the year falls within the specific year category, otherwise 0; and
ε = the error term.

5. Empirical results

5.1. Descriptive statistics

Table 1 reports the descriptive statistics for the dependent variable (THINCAP1 and THINCAP2), independent variables
(MUTLI, TAXHAV, WTAX and UNCERT) and control variables (FOR, SIZE, MKTBK, ROA, CINT, INVINT and RDINT). Table 1 shows

20
With utilities being the omitted industry sector in our regression analysis.
21
With the 2009 year being the omitted year in the regression analysis.
 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25 19

Table 1
Descriptive statistics.

Variable N Mean Std. Dev. Minimum Median Maximum

THINCAP1 812 .434 .376 0 .419 1.496

THINCAP2 812 .073 .261 0 0 1
MULTI 812 .764 .563 0 1 1
TAXHAV 812 .337 .473 0 0 1
WTAX 812 .510 .500 0 1 1
UNCERT 812 .376 .487 0 0 1
FOR 812 .256 .350 0 .070 .365
SIZE 812 20.089 2.800 15.639 20.267 25.281
MKTBK 812 3.771 9.560 −2.537 2.498 6.721
ROA 812 .056 .014 −.243 .066 .352
CINT 812 .784 .208 0 .416 .866
INVINT 812 .105 .289 0 .642 .740
RDINT 812 .006 .004 0 0 .043

Variable definitions: THINCAP1 = (MAD ratio – see Eq. (3)) the average debt divided by the firm’s safe harbor debt amount; THINCAP2 = a dummy variable
of 1 if the firm has breached the MAD position (i.e., MAD > 1) in accordance with the thin capitalization provisions, otherwise 0; MULTI = a dummy variable
of 1 if the firm has at least one subsidiary firm that is incorporated outside Australia, otherwise 0; TAXHAV = a dummy variable of 1 if the firm has at
least one subsidiary firm that is incorporated in an OECD (2006) recognized tax haven, otherwise 0; WTAX = a dummy variable of 1 if the firm is subject
to Australian withholding taxes, otherwise 0; UNCERT = a dummy variable of 1 if the firm has issued a statement noting that it has major difficulties in
estimating its tax liability, otherwise 0; FOR = total foreign income divided by total income; SIZE = the natural logarithm of total assets; MKTBK = market
value of equity divided by the book value of equity; ROA = pre-tax income divided by total assets; CINT = property, plant and equipment divided by total
assets; INVINT = inventory divided by total assets; and RDINT = research and development expenditure divided by total assets.

that THINCAP1 has a mean (median) of .434 (.419) and that THINCAP2 has a mean (median) of .073 (0). Hence, THINCAP2
provides evidence that shows that, on average, 7.3% of the firms in our sample have breached the MAD ratio according to
Australian tax legislation. Overall, an acceptable range of variation is observed for all variables, and there is a reasonable
level of consistency between the means and medians thus reflecting normality of distributions.

5.2. Correlation results

The Pearson pairwise correlation results are presented in Table 2. We find significant correlations between THINCAP1
and MULTI, TAXHAV, WTAX, UNCERT, SIZE, ROA, CINT, INVINT and RDINT (p < .10 or better, with predicted signs where
appropriate), and between THINCAP2 and TAXHAV, UNCERT and SIZE (p < .05 or better, with predicted signs). Table 2 also
shows that only moderate levels of collinearity exist between the explanatory variables used in this study. Finally, we also
computed variance inflation factors (VIFs) when estimating our base regression model to test for signs of multicollinearity
between the explanatory variables. Given that none of the VIFs exceeds five, we conclude that multicollinearity is not
problematic in our study (Hair, Black, Babin, Anderson, & Tatham, 2006).

5.3. Regression results

Table 3 reports the regression results for THINCAP1 (OLS regression analysis) and THINCAP2 (logistic regression analysis).
For the THINCAP1 regression model, Table 3 shows that the regression coefficients for MULTI and TAXHAV are positive and

Table 2
Pearson correlation results.

THINCAP1 THINCAP2 MULTI TAXHAV WTAX UNCERT FOREIGN SIZE MKTBK ROA CINT INVINT RDINT

MULTI .122*** .026 1

TAXHAV .158*** .081*** .276*** 1
WTAX .245*** .010 .361*** .516*** 1
UNCERT .090*** .067** .088*** .133*** .105*** 1
FOR .008 .024 .389*** .256*** .380*** .125*** 1
SIZE .171*** .108*** .253*** .226*** .284*** .037 .022 1
MKTBK .008 .017 .016 .035 .042 .027 .047 .046 1
ROA −.055* −.039 .101*** .069* .102 ***
−.023 −.005 .198*** .008 1
CINT .057* .003 −.041 .036 .037 .069** −.042 −.002 .033 .006 1
INVINT −.048* −.023 .−.003 .040 .003 .017 −.025 .045 .010 .098*** −.013 1
RDINT .086*** .032 .172*** .085** −.038 .027 .116*** −.175*** −.004 −.083** −.018 −.067* 1

Variable definitions: See Table 1.

N = 812 for all variables.
The p-values are one-tailed for the directional hypotheses and two-tailed otherwise.
*
Significance at the .10 level.
**
Significance at the .05 level.
***
Significance at the .01 level.
20 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25

Table 3
Base regression model results.

Variable OLS regression Logistic regression

a
Predicted sign THINCAP1 Predicted sign THINCAP2a

Intercept ? 3.613 (.13) ? 2.590 (2.85)***

MULTI + .197 (3.15)*** + .024 (3.02)***
TAXHAV + .169 (3.31)*** + .139 (2.76)***
WTAX + .167 (4.74)*** + .098 (.26)
UNCERT + .004 (1.55)* + .483 (1.60)*
FOR + .006 (.17) + .002 (.55)
SIZE + .367 (.99) + .032 (.99)
MKTBK ? −.001 (−.05) ? −.001 (−.22)
ROA ? −.078 (−.83) ? −.016 (−1.77)*
CINT + .358 (1.99)** + .051 (.45)
INVINT − −.474 (.18) − −.105 (−.46)
RDINT + .399 (1.58)* + .456 (3.34)***
INDSEC ? Yes ? Yes
YEAR ? Yes ? Yes

Adj. R2 (%) 18.42% Pseudo R2 (%) 10.89%

F-Value 9.98 Chi2 80.57
(Two-tailed p-value) (.001) (Two-tailed p-value) (.001)
N 812 N 812

Variable definitions: INDSEC = industry sector dummy variables; Year = year dummy variables; and see Table 1 for other variable definitions. The p-values
are one-tailed for the directional hypotheses and two-tailed otherwise.
a
Coefficient estimates with the t-statistics in parentheses. Standard errors are corrected using the White (1980) procedure.
*
Significance at the .10 level.
**
Significance at the .05 level.
***
Significance at the .01 level.

significantly associated with thin capitalization (p < .01), thereby providing support for H1 and H2. Firms with thinly capital-
ized structures tend to have at least one foreign incorporated subsidiary. We also find that firms with at least one subsidiary
firm incorporated in an OECD recognized tax haven are more likely to have thinly capitalized structures. Additionally, the
regression coefficient for WTAX is positive and significantly associated with thin capitalization (p < .01), thus providing sup-
port for H3. Development of thinly capitalized structures is typically associated with firms that are exposed to Australian
withholding taxes. The regression coefficient for UNCERT is positive and significantly associated with thin capitalization
(p < .10), thus providing some support for H4. It seems that development of thinly capitalized structures is associated with
firms’ disclosure of uncertainty in calculating its tax liabilities. Finally, for our control variables, the regression coefficient for
CINT is positive and significantly associated with thin capitalization (p < .05), as expected. Similarly, the regression coefficient
for RDINT is positive and significantly associated with thin capitalization (p < .10).
In terms of the THINCAP2 regression model, we observe that the regression coefficients for MULTI and TAXHAV are positive
and significantly associated with thin capitalization (p < .01), thus providing additional support for H1 and H2. Furthermore,
we find that the regression coefficient for UNCERT is positive and significantly associated with thin capitalization (p < .10),
which provides some further support for H4. Finally, the ROA and RDINT control variables are positive and significantly
associated with thin capitalization (p < .10 and p < .01, respectively), as expected.
In summary, our regression results show that thin capitalization is significantly and positively associated with multina-
tionality, tax haven utilization, withholding taxes and tax uncertainty. In fact, given the magnitude and significance levels
of the regression coefficients in our study, variables relating to multinationality and tax haven utilization are, in particular,
strongly positively associated with a firm’s thin capitalization position.

5.4. Robustness checks

We perform several robustness checks to assess the reliability of the regression results reported in Table 3. First, we
drop all of the control variables from the base regression model and obtain similar results for MULTI, TAXHAV, WTAX
and UNCERT. Second, we enter the control variables consecutively into the base regression model and our results remain
unchanged. Overall, the regression coefficients for MULTI, TAXHAV, WTAX and UNCERT are stable in the base regression
model. Finally, to deal with potential outlier problems, we re-estimate the base regression model after excluding several
outliers based on the method recommended by Neter, Wasserman, and Kutner (1996). The predicted signs and statistical
significance of the various regression coefficients are similar to those reported in Table 3.

5.5. Corporate governance monitoring mechanisms

The regression results reported in Table 3 show that thin capitalization is positively associated with multinationality,
tax haven utilization, withholding taxes and tax uncertainty. However, recent research provides evidence that suggests that
 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25 21

corporate governance monitoring mechanisms may assist in limiting the tax avoidance activities of firms. For example, Lanis
and Richardson (2011) find that a higher proportion of outside directors on the board of directors reduces the likelihood of tax
avoidance. We thus extend our base regression model in Eq. (4) to empirically test the impact of several corporate governance
monitoring mechanisms on thin capitalization, including: (1) board of director independence; (2) level of institutional
ownership; (3) existence of a big-4 external auditor; and (4) board of director statement that the firm’s risk management
system is operating effectively.
Our conjecture is that corporate governance monitoring mechanisms are associated with a reduction in tax avoidance
activities of firms. Emerging literature in the area of corporate governance and taxation by Desai and Dharmapala (2006,
2007) and Lanis and Richardson (2011) suggests that analyzing the link between corporate governance and tax avoidance
may offer some important insights into the real-world consequences of tax policies and the effectiveness of various corporate
governance mechanisms. An empirical test of this link is also important due to recent concerns raised by tax authorities (e.g.,
the ATO) that the tax avoidance activities of firms (e.g., the strategic use of debt along with tax havens and transfer pricing)
are contributing to the progressive erosion of government tax revenue as evidenced by the decline in corporate ETRs and
the increasing number of firms reporting zero tax liabilities (ATO, 2010).
The board of directors represents the highest-level internal control mechanism for monitoring the actions of top man-
agement in a firm (Fama & Jensen, 1983). Independent directors have incentives to perform their monitoring tasks properly,
rather than colluding with top managers to seize shareholder wealth, and thus the addition of independent directors to a
board increases its ability to monitor top management (Fama & Jensen, 1983). Research by Lanis and Richardson (2011)
shows that a higher proportion of independent directors on a board reduces the likelihood of tax avoidance. We measure
board of director independence (BODI) as the proportion of a firm’s board members who are independent directors. The
predicted sign for BODI is negative.
Institutional ownership can also serve as an important corporate governance monitoring mechanism (Beasley, 1996;
Bhojraj & Sengupta, 2003). The existence of institutional ownership adds to the incentive to monitor management because
institutional shareholders have greater power and influence over the board of directors and management than do smaller
shareholders (Jensen, 1993; Shleifer & Vishny, 1986). Institutional ownership (INST) is measured in our study as the per-
centage of ordinary share capital owned by the top five institutional shareholders of a firm. The predicted sign for INST is
negative.
Another important corporate governance monitoring mechanism is the utilization of a big-4 external audit firm (Beasley,
1996; Klein, 2002), which could have an impact on tax avoidance. The use of an external auditor from one of the big-4
audit firms may assist in reducing tax avoidance behavior via the enhanced monitoring of management and the provision
of higher quality audits. Some research (e.g., Matsumura & Tucker, 1992; Rezaee, 2005) finds a positive association between
big-4 auditor use and the probability of detecting financial statement fraud. Conversely, other research (e.g., Badertscher,
Katz, & Rego, 2009) finds that the implementation of successful tax avoidance strategies requires the services of professional
consultants, including those working in the big-4 audit firms. We measure big-4 auditor (AUD) as a dummy variable of 1
if the firm employs a big-4 auditor, otherwise 0. No sign prediction is made for AUD due to the inconsistent findings in
previous research.
The final corporate governance monitoring mechanism that we consider in our study is the existence of an effective
risk management policy as assessed by the board of directors (D’Ascenzo, 2010; OECD, 2009). The Corporate Governance
Council’s Principle 7 issued by the ASX states that a firm should establish a sound risk management policy that deals with risk
oversight, risk management, the firm’s risk profile and the degree of internal control (ASX, 2007).22 In fact, such a policy is
part of the board’s oversight role, and its responsibility to oversee the formation and implementation of a risk management
system in the firm (Erle, 2008). Risk management (RISKMGT) is measured in this study as a dummy variable of 1 if the firm
has a statement in its annual report indicating that its risk management system has been operating effectively and that the
board of directors attests to the system of internal controls and risk management policy that are in place, otherwise 0. The
predicted sign for RISKMGT is negative.
Our base regression model in Eq. (4) is thus extended to include the above-mentioned corporate governance monitoring
mechanism variables as follows:

THINCAPit = ˛0 + ˇ1 MULTIit + ˇ2 TAXHAVit + ˇ3 WTAXit + ˇ4 UNCERTit + ˇ5 BODIit + ˇ6 INSTit + ˇ7 AUDit

+ ˇ8 RISKMGTit + B9 FORit + ˇ10 SIZEit + ˇ11 MKTBKit + ˇ12 ROAit + ˇ13 CINTit + ˇ14 INVINTit

+ ˇ15 RDINTit + ˇ16−23 INSECit + ˇ24−26 YEARit + εit (5)

where:

BODI = the proportion of board members who are independent directors;

INST = the percentage of ordinary share capital owned by the top five institutional shareholders of a firm;
AUD = a dummy variable of 1 if the firm employs a big-4 auditor, otherwise 0; and

22
Of course, a firm’s risks include such tax-related risks as corporate tax avoidance (D’Ascenzo, 2010; Erle, 2008; OECD, 2009).
22 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25

Table 4
Extended regression model results – corporate governance monitoring variables.

Variable OLS regression Logistic regression

a
Predicted sign THINCAP1 Predicted sign THINCAP2a

Intercept ? 7.862 (.783) ? 5.186 (3.70)***

MULTI + .281 (5.18)*** + .021 (2.83)***
TAXHAV + .720 (2.34)*** + .290 (3.06)***
WTAX + .712 (4.40)*** + .259 (.64)
UNCERT + .986 (1.46)* + .505 (1.66)**
BODI − −.085 (−1.41)* − −.016 (−1.64)**
INST − −.128 (−1.65)** − −.303 (−2.88)***
AUD ? −.977 (−2.54)*** ? −.186 (−.37)
RISKMGT − −.131 (−.41) − −.233 (−.35)
FOR + .010 (.29) + .002 (.43)
SIZE + .046 (.79) + .032 (1.01)
MKTBK ? −.001 (−.06) ? −.001 (−.06)
ROA ? −.112 (−1.20) ? −.020 (−2.06)**
CINT + .780 (2.19)** + .096 (.73)
INVINT − −.123 (−.05) − −.095 (−.43)
RDINT + .350 (1.41)* + .151 (3.48)***
INDSEC ? Yes ? Yes
YEAR ? Yes ? Yes

Adj. R2 (%) 16.38% Pseudo R2 (%) 14.34%

F-Value 10.19 Chi2 85.12
(Two-tailed p-value) (.001) (Two-tailed p-value) (.001)
N 812 N 812

Variable definitions: BODI = the proportion of board members who are independent directors; INST = the percentage of ordinary share capital owned by a
firm’s top five institutional shareholders; AUD = a dummy variable of 1 if the firm employs a big-4 auditor, otherwise 0; and RISKMGT = a dummy variable
of 1 if the firm has a statement in its annual report indicating that its risk management system has been operating effectively and that its board of directors
attests to the system of internal controls and risk management policy that is has in place, otherwise 0; and see Tables 1 and 3 for other variable definitions.
The p-values are one-tailed for directional hypotheses and two-tailed otherwise.
a
Coefficient estimates with the t-statistics in parentheses. Standard errors are corrected using the White (1980) procedure.
*
Significance at the .10 level.
**
Significance at the .05 level.
***
Significance at the .01 level.

RISKMGT = a dummy variable of 1 if the firm has a statement in its annual report indicating that its risk management system
has been operating effectively and the board of directors attests to the system of internal controls and the risk management
policy that are in place, otherwise 0.

The extended regression model results of the impact of various corporate governance monitoring mechanisms on thin
capitalization are reported in Table 4. For the THINCAP1 regression model, we find that the regression coefficient for BODI
is negative and significantly associated with thin capitalization (p < .10). It seems that a higher proportion of independent
directors on a firm’s board reduces the likelihood of the firm approaching or exceeding its MAD position under the thin
capitalization provisions. Our results are consistent with Lanis and Richardson’s (2011) general findings on board composi-
tion and tax avoidance. We also observe that the regression coefficient for INST is negative and significantly associated with
thin capitalization (p < .05). It appears that the existence of institutional ownership adds to the incentive to monitor man-
agement as institutional shareholders have greater power and influence over the board of directors and management than
do smaller shareholders (Jensen, 1993; Shleifer & Vishny, 1986). Moreover, we find that AUD is negative and significantly
associated with thin capitalization (p < .01), which is consistent with previous research (e.g., Matsumura & Tucker, 1992;
Rezaee, 2005). Big-4 auditor utilization seems to reduce the probability of a firm approaching or exceeding its MAD position
under the thin capitalization provisions. However, the regression coefficient for RISKMGT is not significant. Moreover, in
terms of the regression coefficients for our other independent variables, MULTI, TAXHAV, WTAX and UNCERT are positive
and significantly associated with thin capitalization (p < .10 or better). Finally, regarding our control variables, the regression
coefficients for CINT and RDINT have a positive and significant association with thin capitalization (p < .10 or better).
For the THINCAP2 regression model, we once again find that the regression coefficients for BODI and INST are negative and
significantly associated with thin capitalization (p < .05 and p < .01, respectively). Nevertheless, the regression coefficients for
AUD and RISK are not significant. In terms of the regression coefficients for our other independent variables, MULTI, TAXHAV
and UNCERT are positive and significantly associated with thin capitalization (p < .05 or better), whereas the regression
coefficient for WTAX is insignificant. Finally, for our control variables, the regression coefficients for ROA and RDINT have a
negative/positive and significant association with thin capitalization (p < .05 or better).
Overall, the extended regression results provide some unique evidence that corporate governance monitoring mech-
anisms relating to board of director independence, institutional ownership and big-4 auditor employment reduces the
likelihood of firms approaching or exceeding their MAD positions under the thin capitalization tax legislation.
 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25 23

6. Conclusion

This study investigates the determinants of thinly capitalized structures of publicly-listed Australian firms. Based on the
magnitude and significance levels of the regression coefficients in our study, variables pertaining to multinationality and
tax haven utilization are, in particular, significantly and positively associated with firms’ thin capitalization position.23 We
also present supplementary evidence which shows that corporate governance monitoring mechanisms relating to board of
director independence, institutional ownership and big-4 auditor employment are significantly negatively associated with
firms’ thinly capitalization position.
This study extends the extant literature on corporate tax avoidance by considering the determinants of thinly capitalized
tax avoidance structures in empirical terms for possibly the first time. The use of the method statement in the Australian
tax legislation together with reliance on accounting definitions of financial statement elements permits duplication of this
tax legislation-based thin capitalization measure in other jurisdictions (e.g., Canada, New Zealand and the UK). Our findings
also provide policymakers and regulators with valuable information about the major determinants of thin capitalization
structures, as well as on the corporate governance monitoring mechanisms that can assist in controlling them in practice.
For instance, the ATO and the Treasury may review, enforce or modify the thin capitalization provisions if there is continued
evidence of erosion of tax revenue attributable to excessive interest deductions, particularly in light of widening budget
deficits. Furthermore, the absence of robust thin capitalization rules could increase the incentive of multinational firms
to shift debt and related deductions to higher taxed jurisdictions (such as Australia) providing them with a competitive
advantage over domestic firms (The Treasury, 2012).
This study is subject to several limitations. First, the sample is drawn from publicly-listed Australian firms. Because of data
unavailability, it was not possible to include unlisted firms in our sample. Second, given that tax return data are private, we
constructed our measures of thin capitalization based on financial statement data. Finally, our base regression model could
be incomplete. For example, the role of tax authorities could impact on the likelihood of firms developing thin capitalization
structures.
Future research could consider the thinly capitalized tax avoidance structures in other western countries, including
Canada, New Zealand and the UK that applies similar thin capitalization rules to those of Australia with the aim of preven-
ting multinational taxpayers from allocating a disproportionate amount of debt in the capital structures of their domestic
operations. We encourage further research in this area.

Appendix A. List of OECD (2006) recognized tax havens

The OECD’s (2006) 33 registered tax havens are Anguilla, Antigua and Bermuda, Bahamas, Bahrain, Bermuda, Belize,
British Virgin Islands, Cayman Islands, Cook Islands, Cyprus, Dominica, Gibraltar, Grenada, Guernsey, Isle of Man, Jersey,
Liberia, Malta, Marshall Islands, Mauritius, Montserrat, Nauru, Netherlands Antilles, New Caledonia, Panama, Samoa, San
Marino, Seychelles, St. Lucia, St. Kitts and Nevis, St. Vincent and the Grenadines, Turks and Caicos Islands and Vanuatu.

Appendix B. Calculation of a firm’s thinly capitalized position using the method statement in section 820-95 of
ITAA97

The method statement provided in the ITAA97 can be summarized as follows:

• Step 1: Work-out the average value, for the income year, of all assets of the entity.
• Step 1A: Reduce the result of Step 1 by the average value, for that year, of all excluded equity interests in the entity.
• Step 2: Reduce the result of Step 1A by the average value, for that year, of all associated entity debt of the entity.
• Step 3: Reduce the result of Step 2 by the average value, for that year, of all associate entity equity of the entity.
• Step 4: Reduce the result of Step 3 by the average value, for that year, of all controlled foreign entity debt of the entity.
• Step 5: Reduce the result of Step 4 by the average value, for that year, of all controlled foreign entity equity of the entity.
• Step 6: Reduce the result of Step 5 by the average value, for that year, of all of the non-debt liabilities of the entity. If the
result of this step is a negative amount, then it is taken to be nil.
• Step 7: Multiply the result of Step 6 by 3/4.
• Step 8: Add to the result of Step 7 the average value, for that year, of the entity’s associate entity excess amount. The result
of this step is the safe harbor debt amount.

23
Moreover, firms which have breached the MAD position in accordance with the thin capitalization provisions (i.e., MAD > 1), are potentially non-
compliant compliant with the thin capitalization provisions, and thus may indicate a concerted effort by management to use debt as a means to reduce
the amount of corporate taxes payable. While firms may be compliant with the thin capitalization provisions if their MAD position is less than one, firms
having a MAD position close to the upper limit of one may still reflect a tendency of these firms to use debt aggressively to assist in reducing the amount
of corporate taxes payable.
24 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25

References

Adhikari, A., Derashid, C., & Zhang, H. (2006). Public policy, political connections, and effective tax rates: Longitudinal evidence from Malaysia. Journal of
Accounting and Public Policy, 25, 574–595.
ASX. (2007). Corporate Governance Principals and Recommendations. Australian securities exchange, Sydney. Available from: [Link]
Australian Taxation Office (ATO). (2002). Taxation ruling TR 2002/20, Income tax: Thin capitalization – Definition of assets and liabilities for the purposes of
division 820. Canberra, ACT: ATO. Available at: [Link]
Australian Taxation Office (ATO). (2004a). Tax havens and tax administration, NAT 10567-01. 2004. Canberra, ACT: ATO. Available at: [Link]
Australian Taxation Office (ATO). (2004b). Thin capitalization: Guide, NAT 4461. Canberra, ACT: ATO. Available at: [Link]
Australian Taxation Office (ATO). (2002). Large business and tax compliance. Canberra, ACT: ATO. Available at: [Link]
Australian Taxation Office (ATO). (2010). Large business and tax compliance. Canberra, ACT: ATO. Available at: [Link]
Badertscher, B., Katz, S., & Rego, S. (2009). The impact of private equity ownership on corporate tax avoidance. Working Paper 10-004, Harvard Business School.
Beasley, M. S. (1996). An empirical analysis of the relation between the board of director composition and financial statement fraud. The Accounting Review,
71(4), 443–465.
Beattie, V., Goodacre, A., & Thomson, S. J. (2006). Corporate financing decisions: UK survey evidence, Journal of Business. Finance and Accounting, 33(9–10),
1402–1434.
Benshalom, I. (2008). The quest to tax interest income in a global economy: Stages in the development of international income taxation. Virginia Tax Review,
27(3), 631–708.
Beuselinck, C., Buysschaert, A., & Deloof, M. (2005). Business groups, taxes and earnings management. Paper prepared for the European Accounting Association
Congress, Gothenburg, Sweden, May 17–20, 2005.
Bhojraj, S., & Sengupta, P. (2003). Effect of corporate governance on bond ratings and yields: The role of institutional investors and outside directors. Journal
of Business, 76(3), 455–475.
Chen, S., Chen, X., Cheng, Q., & Shevlin, T. (2010). Are family firms more tax aggressive than non-family firms? Journal of Financial Economics, 95, 41–61.
Christensen, J., & Murphy, R. (2004). The social irresponsibility of corporate tax avoidance: Taking CSR to the bottom line. Development, 47(3), 37–44.
Collins, J. H., & Shackelford, D. A. (1998). Global organizations and taxes: An analysis of the dividend, interest, royalty, and management fee payments
between U.S. multinationals’ foreign affiliates. Journal of Accounting and Economics, 24, 151–173.
Dahlby, B. (2008). Taxation of Inbound direct investment: Economic principles and tax policy considerations. Research report prepared for the Advisory Panel
on Canada’s System of International Taxation.
D’Ascenzo, M. (2010). What’s tax got to do with it? Speech to the Australian Institute of Company Directors. Sydney, 16 February 2010. Available at:
[Link]
Desai. (2006). The demand for tax haven operations. Journal of Public Economics, 90, 513–531.
Desai, & Dharmapala, D. (2006). Corporate tax avoidance and high-powered incentives. Journal of Financial Economics, 79(1), 145–179.
Desai, & Dharmapala, D. (2007). Taxation and corporate governance: An economic approach. In Taxation and Corporate Governance Conference, Munich, 2007
Desai, M. A., Foley, C. F., & Hines, J. R. (2004). A multinational perspective on capital structure choice and internal capital markets. Journal of Finance, 59,
2451–2487.
Desai, M., Foley, C. F., & Hines, J. R. (2006). The demand for tax haven operations. Journal of Public Economics, 90, 513–531.
Desai, & Hines, J. R. (2002). Expectations and expatriations: Tracing the causes and consequences of corporate inversions. National Tax Journal, 55(3),
409–440.
Dharmapala, D., & Riedel, N. (2011). Earnings shocks and tax-motivated income-shifting: Evidence from European multinationals. Working Paper 1101, Oxford
University, Centre of Business Taxation.
Dyreng, S., Hanlon, M., & Maydew, E. (2008). Long-run corporate tax avoidance. The Accounting Review, 83(1), 61–82.
Dyreng, S., & Lindsey, B. P. (2009). Using financial accounting data to examine the effect of foreign operations located in tax havens and other countries on
U.S. multinational firms’ tax rates. Journal of Accounting Research, 47(5), 1283–1316.
Erle, B. (2008). Tax risk management and board responsibility. In W. Schön (Ed.), Tax and corporate governance (pp. 205–220). Berlin/Heidelberg: Springer-
Verlag.
Fama, E. F., & Jensen, M. C. (1983). Separation of ownership and control. Journal of Law and Economics, 26, 301–325.
Frank, M. M., Lynch, J. L., & Rego, S. O. (2009). Are financial and tax reporting aggressiveness reflective of broader corporate policies? The Accounting Review,
84(2), 467–496.
Frischmann, P. J., Shevlin, T., & Wilson, R. (2008). Economic consequences of increasing the conformity in accounting for uncertain tax benefits. Journal of
Accounting and Economics, 46(2–3), 261–278.
Graham, J. R., & Harvey, C. R. (2001). The theory and practice of corporate finance: Evidence from the field. Journal of Financial Economics, 60(2), 187–243.
Graham, J. R., & Tucker, A. (2006). Tax shelters and corporate debt policy. Journal of Financial Economics, 81, 563–594.
Gupta, S., & Newberry, K. (1997). Determinants of the variability on corporate effective tax Rates: Evidence from longitudinal data. Journal of Accounting
and Public Policy, 16(1), 1–34.
Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate data analysis (6th ed.). Upper Saddle River, NJ: Pearson Prentice-Hall.
Hanlon, M., Mills, L., & Slemrod, J. (2005). An empirical examination of corporate tax noncompliance. Working Paper 1025, University of Michigan, University
of Texas.
Hanlon, M., & Slemrod, J. (2009). What does tax aggressiveness signal? Evidence from stock price reactions to news about tax shelter involvement. Journal
of Public Economics, 93(1/2), 126–141.
Harris, D., Morck, R., Slemrod, J., & Yeung, B. (1993). Income shifting in U.S. multinational firms. In A. Giovannini, R. G. Hubbard, & J. Slemrod (Eds.), Studies
in international taxation (pp. 277–302). Chicago: University of Chicago Press.
Huizinga, H., Laeven, L., & Nicodeme, G. (2008). Capital structure and international debt shifting. Journal of Financial Economics, 88, 80–118.
Jensen, M. C. (1993). The modern industrial revolution, exit, and the failure of internal control systems. Journal of Finance, 48, 831–880.
Jog, V., & Tang, J. (2001). Tax reforms, debt shifting and tax revenues: Multinational corporations in Canada. International Tax and Public Finance, 8, 5–25.
Klein, A. (2002). Audit committee, board of director characteristics, and earnings management. Journal of Accounting and Economics, 33, 375–400.
Lanis, R., & Richardson, G. (2011). The effect of board of director composition on corporate tax aggressiveness. Journal of Accounting and Public Policy, 30(1),
50–70.
Matsumura, E. M., & Tucker, R. R. (1992). Fraud detection: A theoretical foundation. The Accounting Review, 67(4), 753–782.
Mills, L. F., & Newberry, K. J. (2004). Do foreign multinationals’ tax incentives influence their U.S. income reporting and debt policy? National Tax Journal,
57(1), 89–107.
Mintz, J., & Weichenrieder, A. (2005). Taxation and the financial structure of German outbound FDI. CESIFO Working Paper no. 1612.
Neter, J., Wasserman, W., & Kutner, M. (1996). Applied linear regression models (3rd ed.). Chicago: Irwin.
Newberry, K. J., & Dhaliwal, D. S. (2001). Cross-jurisdictional income shifting by U.S. multinationals: Evidence from international bond offerings. Journal of
Accounting Research, 39(3), 643–662.
OECD. (2006). The OECD project on harmful tax practices: 2006 update on progress in member countries. Available at:. [Link]
[Link]/dataoecd/1/17/[Link]
OECD. (2009). OECD forum on tax administration, information note: General administration. Paris: OECD.
Rego, S. O. (2003). Tax-avoidance activities of U.S. multinational firms. Contemporary Accounting Research, 20(4), 805–833.
Rezaee, Z. (2005). Causes, consequences, and deterrence of financial statement fraud. Critical Perspectives on Accounting, 16, 277–298.
 G. Taylor, G. Richardson / Journal of International Accounting, Auditing and Taxation 22 (2013) 12–25 25

Richardson, G., & Lanis, R. (2007). Determinants of the variability in corporate effective tax rates and tax reform: Evidence from Australia. Journal of
Accounting and Public Policy, 26(6), 689–704.
Richardson, G., Hanlon, D., & Nethercott, L. (1998). Thin capitalization: An Anglo-American comparison. The International Tax Journal, 24(2), 36–66.
Shackelford, D. A., & Shevlin, T. (2001). Empirical tax research in accounting. Journal of Accounting and Economics, 31(2), 321–387.
Shackelford, D. A., Slemrod, J., & Sallee, J. M. (2007). A unifying model of how the tax system and generally accepted accounting principles affect corporate
behavior. Working Paper, University of North Carolina and University of Michigan.
Shleifer, A., & Vishny, R. W. (1986). Large shareholders and corporate control. Journal of Political Economy, 95, 461–488.
Slemrod, J. (2001). A general model of the behavior response to taxation. International Tax and Public Finance, 8(2), 119–128.
Slemrod, J. (2004). The economics of corporate tax selfishness. National Tax Journal, 57, 877–899.
Slemrod, J., & Wilson, J. D. (2009). Tax competition with parasitic tax havens. Journal of Public Economics, 93, 1261–1270.
Smith, A. M. C. (1996). New Zealand’s thin capitalization rules. Canadian Tax Journal, 44(6), 1525–1551.
Stickney, C., & McGee, V. (1982). Effective corporate tax rates - the effect of size, capital intensity, leverage, andother factors. Journal of Accounting and Public
Policy, 1, 125–152.
Taylor, G., & Tower, G. (2009). Impact of adoption of IFRS on the thinly capitalised position of Australian companies. Journal of Tax Research, 7(1), 37–53.
The Treasury. (2012). Business tax working group: Final report on the tax treatment of losses. Canberra, Australia: The Treasury. Available at:
[Link]
Taylor, G., Tower, G., & Van der Zahn, M. (2010). The influence of international taxation structures on corporate financial disclosure patterns. Accounting
Forum, 35, 32–46.
Wilson, R. (2009). An examination of corporate tax shelter participants. The Accounting Review, 84(3), 969–999.