Accepted Manuscript

Title: An Empirical Analysis of the Changes in Tax Audit

Focus on International Transfer Pricing

Author: K. Hung Chan Agnes W.Y. Lo Phyllis Mo

PII: S1061-9518(15)00002-6
DOI: http://dx.doi.org/doi:10.1016/j.intaccaudtax.2014.12.001
Reference: ACCAUD 193

To appear in: Journal of International Accounting, Auditing and Taxation

Received date: 2-10-2013

Revised date: 15-12-2014
Accepted date: 18-12-2014

Please cite this article as: CHAN, K. H., LO, A. W. Y., and MO, P.,An
Empirical Analysis of the Changes in Tax Audit Focus on International
Transfer Pricing, Journal of the Chinese Institute of Chemical Engineers (2015),
http://dx.doi.org/10.1016/j.intaccaudtax.2014.12.001

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
*Title Page (including author details)

An Empirical Analysis of the Changes in Tax Audit Focus

on International Transfer Pricing

by

t
ip
cr
K. Hung CHAN and Agnes W.Y. LO
Department of Accountancy

us
Lingnan University
Tuen Mun, Hong Kong

and
an
Phyllis MO
M
Department of Accountancy
City University of Hong Kong
Kowloon Tong, Hong Kong
e d
pt

April 2014
ce
Ac

Correspondences concerning this paper should be sent to:

Professor Agnes Lo
Department of Accountancy
Lingnan University
Tuen Mun, Hong Kong

Tel. No.: (852)2616-8163

Fax No. : (852) 2466-4751
E-mail: wylo@ln.edu.hk

Page 1 of 32
*BLINDED Manuscript (no author details)

1
2 An Empirical Analysis of the Changes in Tax Audit Focus
3
4 on International Transfer Pricing
5
6
7 Abstract
8
9 International transfer pricing is an important financial management mechanism allowing
10
11

t
12 multinational corporations to maneuver funds internationally. The manipulation of reported

ip
13
14 profits often triggers investigations from tax authorities. With the increased globalization of
15

cr
16
17
economies and changes in the business environment, tax authorities in many countries have
18

us
19 refined their enforcement of transfer pricing regulations. In this study, we use archival data in
20
21 China to examine whether tax authorities have changed their focus on auditing multinational
22
23
24
25
26
an
companies over the past two decades. Our results indicate that Chinese tax authorities have

significantly reduced their focus on auditing wholly foreign-owned enterprises, and placed
27
M
28
29 more focus on Western multinationals and larger companies in the late 2000s as compared
30
31 with tax audits in the early 1990s. Tax audits in the late 2000s also focus on export-oriented
32
d

33
34 and loss firms. The findings show that changes in the business environment, regulations and
te

35
36 the audit expertise of tax officials can lead to a shift in the focus of tax audits of international
37
p

38
transfer pricing.
39
ce

40
41
42
43 Keywords: foreign investment enterprises; international transfer pricing; tax audit focus
44
Ac

45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64 Page 2 of 32
65
 1
2 An Empirical Analysis of the Changes in Tax Audit Focus
3
4 on International Transfer Pricing
5
6
7
8
9 1. Introduction
10
11

t
12 This study provides empirical insights into transfer pricing audit focus in China. We

ip
13
14 examine firm characteristics that affect the probability of firms being selected for transfer
15

cr
16
17
pricing audit and investigate whether there is any longitudinal change in tax authorities audit
18

us
19 selection focus over the past two decades.
20
21 International transfer pricing has long been regarded as an important financial
22
23
24
25
26
an
management mechanism for multinational corporations (MNCs) to maneuver funds

internationally and to choose the countries in which profits are reported. MNCs frequently
27
M
28
29 use transfer pricing as a means of reducing global income tax payments. Since 1995, biennial
30
31 surveys conducted by Ernst & Young have found that transfer pricing continues to be rated
32
d

33
34 by management of MNCs as the most important international tax issue (Ernst & Young,
te

35
36 2012). According to the Ernst & Young survey, the risk of transfer pricing audits is
37
p

38
increasing and MNCs expect to devote more resources to transfer pricing compliance,
39
ce

40
41 particularly in China, the U.S. and India. According to a Deloitte webcast survey in 2012, 35
42
43 percent of MNCs reported that they had a Chinese affiliate that had undergone a transfer
44
Ac

45
46 pricing audit within the past three years (Bell, 2012).
47
48 Tax authorities consider transfer pricing the most common form of tax avoidance by
49
50
51 MNCs. This form of tax avoidance costs China an annual RMB 30 billion (US$4.7 billion)
52
53 loss in tax revenue and Chinese tax authorities have stepped up measures to combat such tax
54
55
56
avoidance (Pappas, 2012). China is one of the particularly dangerous zones for MNCs
57
58 because Chinese tax authorities very likely impose penalties after an audit. According to
59
60
61
62
63
64 1
Page 3 of 32
65
 1
2 Ernst & Young (2012), more than three-quarters of firms in China with transfer pricing audit
3
4 adjustments had to pay penalties, and the average penalty rate was 25% of the additional tax.
5
6
7 Given these financial implications, it is important that MNCs understand how Chinese tax
8
9 authorities select transfer pricing audit targets and how this selection changes over time.
10
11

t
12 China has systematically liberalized its foreign direct investment regime since 1979. For

ip
13
14 example, in the mid-1990s, the Chinese government extended some of the special provisions
15

cr
16
17
used to attract foreign direct investment in special economic zones to other locations across
18

us
19 the country. After Chinas admission to the WTO in 2001, many sectors, such as power
20
21 generation, transportation, port development, oil exploration and exploitation, which were
22
23
24
25
26
an
previously restricted, are now open to foreign investors. As a result, according to Zhang

(2006), there have been significant changes in the characteristics of foreign investment in
27
M
28
29 China since the early 1990s. Changes identified by Zhang (2006) include (1) investment by
30
31 large MNCs has grown rapidly, with over 80 percent of all Fortune 500 companies now
32
d

33
34 having established operations in China; (2) the entry mode of many MNCs has shifted from
te

35
36 joint ventures to wholly foreign owned projects; (3) while small-scale, labor-intensive, and
37
p

38
export-oriented foreign direct investment projects were dominant in the 1980s and the early
39
ce

40
41 1990s, large-scale, capital- and technology-intensive, and local market-oriented foreign direct
42
43 investments have increased substantially; and (4) while the majority of foreign direct
44
Ac

45
46 investment is still from Hong Kong and Taiwan, the proportion of foreign direct investment
47
48 from these sources has been reduced from about 80 percent to 60 percent in the past 15 years.
49
50
51 Chan and Chow (1997a) provided the first empirical study on transfer pricing audits in
52
53 China during 1992-93. Tax audit focus was measured by the difference between the audited
54
55
56
samples and their respective population distributions in the early 1990s. Chan and Chows
57
58 univariate tests showed that transfer pricing audits in China were confined mainly to
59
60
61
62
63
64 2
Page 4 of 32
65
 1
2 small-sized companies and transfer of tangible goods. Tax authorities also tended to focus on
3
4 auditing Hong Kong-Taiwan sourced companies and wholly foreign-owned companies
5
6
7 (WFOEs). Given deepening economic reforms, changes in the business environment and
8
9 regulations, and a greater ability of enforcing transfer pricing regulations by Chinese tax
10
11

t
12 authorities in the past two decades, we expect corresponding changes in tax authorities focus

ip
13
14 on auditing foreign investment enterprises (FIEs) for transfer pricing. In this paper, we update
15

cr
16
17
and extend Chan and Chow (1997a) by assessing audit focus in 2006-10 and comparing it
18

us
19 with the results from 1992-93 to identify any longitudinal changes. Our results show that
20
21 audit focus on Hong Kong-Taiwan sourced FIEs and small FIEs has been reduced
22
23
24
25
26
an
significantly, suggesting more audit efforts toward large Western FIEs (mainly U.S. and

Japanese FIEs) in the 2000s. Moreover, tax authorities no longer focus on auditing wholly
27
M
28
29 foreign-owned FIEs. Indeed, for foreign investors, the form of investment should no longer
30
31 be a concern. Instead, in the late 2000s, tax audits focus on export-oriented and loss firms.
32
d

33
34 This paper contributes to the literature on tax audits and planning and should enhance
te

35
36 MNCs awareness of the changing risk of international transfer pricing audits. Increasing
37
p

38
pressure on governments to raise revenue and the dedication of additional resources for
39
ce

40
41 transfer pricing enforcement lead to increasing efforts on tax audits often resulting in
42
43 additional tax assessment and penalties. In China, once a firm is selected for transfer pricing
44
Ac

45
46 audit, an adjustment of taxable income almost always results (Chan and Lo, 2005). Therefore,
47
48 greater knowledge about transfer pricing audit focus, should enable MNCs to tailor their
49
50
51 financial management, investment and transfer pricing strategies to minimize the probability
52
53 of being audited, reduce tax costs, and improve the quality and quantity of documentation to
54
55
56
support their related party dealings.
57
58 The remainder of this paper is organized as follows. The next section discusses
59
60
61
62
63
64 3
Page 5 of 32
65
 1
2 changes in the Chinese business environment related to transfer pricing in the past two
3
4 decades. The third section formulates the research hypotheses. The fourth section explains the
5
6
7 research design, and the fifth section provides empirical results. The last section concludes.
8
9
10
11

t
12 2. Changes in the Business Environment in China during 1991-2010

ip
13
14 2.1. The Growth of Foreign Investment and Trade in China
15

cr
16
17
China has experienced rapid growth in GDP and foreign direct investment in the past
18

us
19 two decades. The average annual GDP growth rate in China was about 9 percent for
20
21 2001-2010 (IMF, 2004; 2010). Foreign direct investment inflows to China reached US$105.7
22
23
24
25
26
an
billion in 2010, making China the largest recipient of foreign direct investment in the

developing world (UNCTAD, 2011). In 2010, China became the largest export nation and the
27
M
28
29 second largest import nation (behind the U.S.) in the world (IMF, 2011). FIEs in China play
30
31 an increasingly important role in its foreign trade. In 2010, total imports and exports by FIEs
32
d

33
34 accounted for 52.9 percent and 54.6 percent of Chinas total imports and exports, respectively
te

35
36 (MOC, 2012). Related-party transactions between FIEs in China and their overseas affiliated
37
p

38
companies have traditionally accounted for a large proportion of these import and export
39
ce

40
41 transactions. Chan and Chow (1997b) find that 88 percent of export-oriented FIEs in China
42
43 trade with their overseas related companies for more than 70 percent of their total imports
44
Ac

45
46 and exports. For non-export-oriented FIEs, 53 percent have more than half of their imports
47
48 from their related companies. Chan and Lo (2004) report that 80 percent of their sample FIEs
49
50
51 had inter-affiliate trade accounting for more than 75 percent of their total trade. In the 2000s,
52
53 there has been a significant increase in Chinas outward direct investment, often in
54
55
56
partnership with MNCs; the average growth rate for outward direct investment is 55 percent
57
58 for 2003-2009 (Wang and Wang, 2011). This outward direct investment should further
59
60
61
62
63
64 4
Page 6 of 32
65
 1
2 increase related party transactions between companies in China and their overseas affiliates.
3
4 Due to the significant amount of related-party transactions of FIEs, international transfer
5
6
7 pricing is a very important issue in China for both the Chinese government and the MNCs
8
9 invested there. Consequently, tax avoidance through transfer pricing is an important concern
10
11

t
12 for the Chinese government. Figure 1 shows a summary of selected statistics relating to the

ip
13
14 business environment in China for the period 1991-2010.
15

cr
16
17
[Insert Figure 1 here]
18

us
19
20 2.2. Currency Appreciation and Foreign Exchange Control
21
22
23
24
25
26
an
Exchange rate and foreign exchange control policies have significant impacts on

companies operation and transfer pricing decisions in China (Chan and Lo, 2004). In the
27
M
28
29 1990s, Chinas currency exchange rate was strictly controlled by the government. In 2005,
30
31 China started to adopt a managed floating exchange rate regime based on the market supply
32
d

33
34 and demand of a basket of currencies. The reform of the Renminbi (RMB) exchange rate
te

35
36 regime has progressed steadily, and the RMB appreciated about 24 percent between 2005 and
37
p

38
39
2010 (Bloomberg News, 2010). In 2010, the Peoples Bank of China deepened reform of the
ce

40
41 RMB exchange rate regime to further increase exchange rate flexibility. Appreciation of the
42
43 RMB has a significant impact on the profitability of FIEs. Particularly, profits of
44
Ac

45
46 export-oriented enterprises dropped significantly due to appreciation of the RMB.
47
48
49
50
China also had strict foreign exchange controls. The relevant regulations are
51
52 governed by the State Administration of Foreign Exchange (SAFE). FIEs capital injections,
53
54 cross-border trade and services transactions settled in foreign exchange, overseas financing
55
56
57 and profit repatriation are all subject to exchange controls. However, since joining the WTO
58
59
60
61
62
63
64 5
Page 7 of 32
65
 1
2 in 2001, China has liberalized its foreign exchange control policies. Currently, FIEs can buy
3
4 and sell foreign currencies for trading purposes (e.g., sales and purchase of goods and
5
6
7 services). The payment and receipt of foreign exchange accounts should be supported by
8
9 documents and settled with self-owned foreign currencies. FIEs can also buy foreign
10
11

t
12 currencies from designated financial institutions engaged in the settlement of foreign

ip
13
14 currency transactions or sell foreign currencies to those institutions. However, foreign
15

cr
16
17
exchange control on capital accounts (i.e., equity or securities investment, loans, derivative
18

us
19 deals, etc.) is still in force and will be gradually relaxed by the SAFE (Deloitte, 2011; SAFE
20
21 2012).
22
23
24
25
26
2.3. Changes in Transfer Pricing Regulations
an
27 The development of national legislation on transfer pricing in China started in 1991. The
M
28
29
30 first transfer pricing provision was set forth in Article 13 of the Income Tax Law of the
31
32 Peoples Republic of China for Enterprises with Foreign Investment and Foreign Enterprise.
d

33
34
te

35
Operational rules are provided under Articles 52 to 58 of the Detailed Rules and Regulations
36
37 on the Implementation of the Income Tax Law for Enterprises with Foreign Investment and
p

38
39 Foreign Enterprises, which define associated enterprises and explain the different categories
ce

40
41
42 of related-party transactions and types of transfer pricing methods that have been used. In
43
44 1992, the State Administration of Taxation (SAT) issued two circulars, Guoshuifa (1992) No.
Ac

45
46
47 237 and Guoshuifa (1992) No. 242. Guoshuifa (1992) No. 237 provides further details on the
48
49 definition of associated enterprises. It also specifies the reporting requirements for business
50
51
52 transactions between associated enterprises, clarifies how transfer pricing adjustments will be
53
54 made, sets out the time limits for transfer pricing investigations, and imposes reporting
55
56
requirements on every FIE. Guoshuifa (1992) No. 242 is a set of implementation rules and
57
58
59 procedures that serve as an internal operational manual for the Chinese tax authorities. It sets
60
61
62
63
64 6
Page 8 of 32
65
 1
2 out general procedures for dealing with transfer pricing audits, including selection and scope
3
4 of investigation, audit techniques and procedures, the scope of adjustments and adjustment
5
6
7 methods. It also explains collection and inquiry of price information, procedures and
8
9 authority for approving adjustments, appeals, procedures and methods to compile, report,
10
11

t
12 document and administer audit cases. This circular serves as the first specific guidelines to

ip
13
14 tax officials for conducting transfer pricing tax audits.
15

cr
16
17 After gaining transfer pricing audit experience and evaluating the need for more
18

us
19
20
comprehensive and updated transfer pricing regulations, in 1998 the SAT issued Guoshuifa
21
22 (1998) No. 59. This regulation sets criteria for selecting target FIEs for tax audits,
23
24
25
26
an
standardizes transfer pricing examination procedures and strengthens internal coordination

27 between local tax authorities and the SAT. Guoshuifa (1998) No. 59 also introduced advance
M
28
29 pricing agreements (APAs), although the explanations on APAs were very brief and the
30
31
32 circular was subsequently abolished in 2006.
d

33
34
te

35 In 2008, a new Enterprise Income Tax Law was issued unifying the income tax rates for
36
37 foreign and domestic enterprises in China (Deloitte, 2008). Under this new law, companies
p

38
39
are required to have contemporaneous transfer pricing documentation as part of an
ce

40
41
42 enterprises transfer pricing compliance requirement. Similar to Guoshuifa (1998), the new
43
44 law reiterates criteria for selecting audit targets based mainly on a firms financial condition
Ac

45
46
47 and nature of operations. To supplement the new Income Tax Law, Guoshuifa (2009) No. 2
48
49 was issued to provide specific guidance on contemporaneous documentation and disclosure
50
51
52 requirements, audit and adjustment procedures and APAs arrangements. Given these new
53
54 regulations, Chinas transfer pricing regime is now much closer to the standards of developed
55
56
57 countries (Deloitte, 2009). In addition, Chinese tax authorities are now fully capable to audit
58
59 FIEs of different industries and business operations.
60
61
62
63
64 7
Page 9 of 32
65
 1
2 The development of transfer pricing regulations also suggests that transfer pricing
3
4 audits in China have been improving. Since the issuance of Guoshuifa (1992) No. 242, the
5
6
7 Chinese government has continued to strengthen its transfer pricing legislative framework by
8
9 introducing new regulations and rules aiming at standardizing transfer pricing examination
10
11

t
12 procedures and strengthening internal coordination between local tax bureaus and the SAT.

ip
13
14 During the last decade, the Chinese government has increased efforts to build transfer pricing
15

cr
16
17
audit teams. The SAT organized training courses with the OECD to provide transfer pricing
18

us
19 lessons to officers of Chinese tax bureaus (SAT, 2009). Selected tax officers were also sent
20
21 overseas for training. Therefore, Chinese tax authorities are now better positioned to audit
22
23
24
25
26
an
larger firms than in the early 1990s. The SAT also promotes information sharing and

cooperation among tax bureaus for auditing large MNCs (Ernst & Young, 2011).
27
M
28
29
30
31
32
d

3. Hypothesis Development
33
34
te

35 Tax compliance is often modeled as a game between taxpayers and tax authorities. Tax
36
37 authorities seek to maximize collections of tax revenue net of audit costs (Beck and Jung,
p

38
39
1989; Sansing, 1993). Using analytical models, Rhoades (1999) shows that tax authorities
ce

40
41
42 will base their audit decisions on the likelihood of misstatement in tax returns and the audit
43
44 cost. In other words, the probability that a company will be selected for audit depends on the
Ac

45
46
47 perceived likelihood and amount of misstatements as well as the related tax audit cost.
48
49 Chan and Chow (1997a) analyze several firm characteristics of FIEs that were subject to
50
51
52 transfer pricing audits by Chinese tax authorities in the early 1990s. Their study uses
53
54 univariate tests to compare distributions of FIE audit sample data with population
55
56
57 characteristics. Specifically, their results indicate that tax authorities tend to focus on auditing
58
59 wholly foreign-owned companies, Hong Kong and Taiwan sourced FIEs and small or
60
61
62
63
64 8
Page 10 of 32
65
 1
2 medium-sized FIEs. To investigate whether there is any change in audit focus over the past
3
4 two decades along these three specific dimensions, we formulate three hypotheses.
5
6
7 3.1. Form of Investment
8
9 Joint venture is a common form of foreign investment in China because it is usually
10
11

t
12 easier for foreign companies to enter into a new market with the help of a local partner.

ip
13
14 Relationships with local partners are important for foreign investors of joint ventures (Chan
15

cr
16
17
and Lo, 2004). Local partners can act as a bridge between foreign investors and local
18

us
19 government, and can help establish effective contacts with local distributors, suppliers,
20
21 customers and workers (Ambler and Witzel, 2000). Chinese partners are equipped with local
22
23
24
25
26
an
market information and normally have a good relationship with the local government.

Up-to-date local market information can help the ventures penetrate into the local Chinese
27
M
28
29 market. A good relationship with the Chinese local government is also very helpful to smooth
30
31 business operations and to negotiate bureaucratic hurdles.
32
d

33
34 Another common form of FIEs in China are wholly foreign-owned enterprises (WFOEs).
te

35
36 However, China in the past had restricted the establishment of WFOEs, particularly in certain
37
p

38
sensitive industries. Although foreign partners have incentives to shift profits out of joint
39
ce

40
41 ventures to avoid profit sharing with local partners, Chan and Chow (1997a) find that
42
43 WFOEs are more likely to shift profits out of China through transfer pricing manipulations
44
Ac

45
46 than joint ventures. This is because the local partners in joint venture can monitor the FIEs
47
48 and such monitoring helps prevent the FIEs from shifting profits out of China. This was an
49
50
51 important reason that WFOEs were more likely than joint ventures to be audited in the early
52
53 1990s.
54
55
56
With rapid social and economic developments and Chinas admission to the WTO, the
57
58 legal system and business environment in the 2000s are more transparent and more
59
60
61
62
63
64 9
Page 11 of 32
65
 1
2 harmonized with international practices than in the 1990s. With a better understanding of the
3
4 business environment and practices in China, foreign investors reliance on the local partners
5
6
7 has gradually been reduced. In addition, with the liberalization of foreign direct investment
8
9 policy, the Chinese government has expanded the industry sectors that can be served by
10
11

t
12 WFOEs. As a result, more foreign investors have established companies in China in the form

ip
13
14 of WFOEs. The percentage of WFOEs increased from 20 percent in 1993 to about 50 percent
15

cr
16
17
in 2010. Since Chinese tax authorities are now more familiar with the way WFOEs operate,
18

us
19 suspicions about WFOEs shifting profits out of China have declined. Furthermore, the
20
21 appreciation of Chinese currency may have also reduced the incentive for WFOEs to shift
22
23
24
25
26
an
profits out of China. Thus, we expect that Chinese tax authorities no longer focus on selecting

WFOEs for audit in 2006-10 as they were in 1992-93. Accordingly, we formulate the
27
M
28
29 following hypothesis.
30
31 Hypothesis 1: Ceteris paribus, the tax audit focus on WFOEs has significantly
32
d

33
34 declined over the past two decades.
te

35
36 3.2. Source of Investment
37
p

38
Performance evaluation is an important consideration for setting transfer prices.
39
ce

40
41 Management is not likely to manipulate transfer prices if they adversely affect
42
43 performance-based compensation. Hong Kong and Taiwan sourced FIEs are often closely
44
Ac

45
46 held, owner-managed family businesses (Liu, 1999). The agency problems between owners
47
48 and management are reduced in this setting. Klassen (1997) finds that high inside ownership
49
50
51 concentration reduces the reliance on public financial reporting. Company information
52
53 including the setting of transfer prices and their effect on FIEs profits can be efficiently
54
55
56
communicated from the management to owners with few information asymmetry problems.
57
58 Therefore, Hong Kong and Taiwan sourced FIEs have lower financial reporting costs for
59
60
61
62
63
64 10
Page 12 of 32
65
 1
2 underreporting profits since they have higher inside ownership structure and can efficiently
3
4 inform shareholders about the companies value through channels other than financial
5
6
7 statements. Thus, the nontax costs of reporting lower book income through transfer pricing
8
9 manipulations are lower for Hong Kong and Taiwan sourced FIEs compared with other FIEs.
10
11

t
12 In addition, auditing Hong Kong and Taiwan sourced FIEs may be easier than auditing

ip
13
14
15 other sourced FIEs because of similarities in culture and language. Furthermore, the tax rates

cr
16
17 in Hong Kong and Taiwan are generally lower than in China, while the tax rates of most
18

us
19
20
Western countries, especially the U.S. and Japan, are higher than in China. Indeed, Hong
21
22 Kong and Taiwan FIEs constitute over 97% of our sample FIEs whose tax rate is below that
23
24
25
26
an
of China. This suggests greater incentive for FIEs to shift profits from China to Hong Kong

27 or Taiwan rather than to the U.S. or Japan. Chan and Chow (1997a) found that in the early
M
28
29 1990s, Hong Kong and Taiwan sourced FIEs were more likely to be audited than other FIEs.
30
31
32 More recently, there has been significant improvement in the corporate governance of
d

33
34
te

35 Hong Kong and Taiwan companies. These improvements coupled with improved
36
37 communications with tax authorities may have reduced the tax audit focus on Hong Kong and
p

38
39
Taiwan sourced FIEs in the 2000s. In addition, the relaxation of foreign exchange controls
ce

40
41
42 should reduce the need for these FIEs to transfer profits out of China. In the past, this was a
43
44 serious problem for Hong Kong-Taiwan sourced FIEs as they required frequent transfers of
Ac

45
46
47 funds with China. Finally, in recent years, due to economic and financial problems in the U.S.
48
49 and Europe, Hong Kong and Taiwan sourced FIEs have quickly increased their investments
50
51
52 to develop Chinas domestic markets. This further reduces their need to transfer funds out of
53
54 China. Consequently, we expect a reduction in tax audit focus on Hong Kong and Taiwan
55
56
57 sourced FIEs over time. The following hypothesis is formulated:
58
59
60
61
62
63
64 11
Page 13 of 32
65
 1
2 Hypothesis 2: Ceteris paribus, the tax audit focus on Hong Kong and Taiwan
3
4 sourced FIEs has significantly declined over the past two decades.
5
6
7 3.3. Size of Investment
8
9
10 Accounting researchers use firm size as a proxy for political sensitivity and
11

t
12 hypothesize that the larger the firm, the more likely the manager is to choose conservative

ip
13
14
15 accounting procedures (Watts and Zimmerman, 1990). For tax planning, larger firms may

cr
16
17 devote more resources and have more sophisticated tax departments that are able to engage in
18

us
19
20
complex tax planning including transfer pricing manipulations (Scholes et al., 1992; Yin and
21
22 Cheng, 2004). In the early 1990s, expertise and resources in China for auditing large MNCs
23
24
25
26
an
were very limited (SAT, 2004a), because such audits required a sophisticated audit team and

27 were very costly. Alternatively, small firms require less expertise and fewer resources to audit
M
28
29 whether a WFOE or joint venture. Thus, Chan and Chow (1997a) found that small and
30
31
32 medium FIEs were more likely to be audited compared to large ones in the early 1990s.
d

33
34
te

35 Since Chinese tax authorities have gained experience on transfer pricing audits and
36
37 guidelines for transfer pricing tax audits have been better developed, tax authorities should
p

38
39
now be more able to audit larger FIEs. Thus, we expect that the tax audit focus on small FIEs
ce

40
41
42 has been reduced significantly over the past two decades. The following hypothesis is
43
44 formulated:
Ac

45
46
47 Hypothesis 3: Ceteris paribus, the tax audit focus on small and medium FIEs has
48
49 significantly declined over the past two decades.
50
51
52
53
54
55 4. Data Collection and Model
56
57 4.1. Data Collection
58
59
60
61
62
63
64 12
Page 14 of 32
65
 1
2 We address our research questions using data from the tax bureaus in cities where
3
4 foreign enterprises are concentrated. The data were from randomly selected FIEs that were
5
6
7 audited and not audited on transfer pricing during the period 2006-2010. Data on firm
8
9 characteristics including the form of investment (i.e. wholly foreign-owned or joint venture),
10
11

t
12 size (measured by sales), source of investment (home country of the investor), profitability,

ip
13
14 percentage of export sales and industry were collected. We examined a total of 169 FIEs that
15

cr
16
17
were audited on transfer pricing and 169 FIEs that were not audited but had related-party
18

us
19 transactions in 2006-2010. Firms with advance pricing agreement (APA) were excluded from
20
21 our sample because the probability of APA firms being audited is very different from
22
23
24
25
26
non-APA firms. an
We also relied on the data from Chan and Chow (1997a) for companies

audited by the Chinese tax authorities in 1992-93. Their data comprises 81 transfer pricing
27
M
28
29 audit cases. They did not include cases with immaterial audit adjustments (mostly for small
30
31 FIEs) or cases where audit guidelines were not sufficiently followed. Following similar
32
d

33
34 procedures, the 2006-2010 audited sample is representative of serious and reliable transfer
te

35
36 pricing audit cases. Thus, these two sets of audited samples should be comparable. In order to
37
p

38
compare the audit focus on a difference-in-difference basis, we also used a sample of 81 FIEs
39
ce

40
41 that had related-party transactions but were not audited in 1992-1993. Details of these audited
42
43 and non-audited cases are discussed below in the Descriptive Statistics section.
44
Ac

45
46 4.2. Model
47
48 To examine whether there are significant changes in the audit focus between 1992-1993
49
50
51 and 2006-2010, we use the following logistic regression model (Chan et al., 2006a):
52
53 AUDIT = 0 +1 WFOE +2 SOURCE +3 SIZE +4 PERIOD +5PERIOD*WFOE
54
55
56
+6 PERIOD* SOURCE +7 PERIOD*SIZE +8 EXPORT + 9 INDUSTRY
57
58 + 10 ROS + 11 LOSS + (1)
59
60
61
62
63
64 13
Page 15 of 32
65
 1
2 The dependent variable of the regression, AUDIT, is a dichotomous variable indicating
3
4 whether the FIE was selected for transfer pricing audit by Chinese tax authorities. AUDIT
5
6
7 equals 1 if a FIE was audited by Chinese tax authority due to transfer pricing manipulation,
8
9 and 0 otherwise. The first test variable, WFOE equals 1 if a FIE is a wholly foreign-owned
10
11

t
12 enterprise, and 0 otherwise. SOURCE equals 1 if a FIE is sourced from Hong Kong or

ip
13
14 Taiwan, and 0 if the FIE is sourced from the U.S., Japan or other countries. We confirmed
15

cr
16
17
that all Hong Kong-Taiwan firms are genuine Hong Kong-Taiwan firms by tracing the
18

us
19 original nationality of their parent companies. SIZE equals 1 if a FIEs sales are greater than
20
21 one third of the sample firms for a given period (i.e. 1992-93 or 2006-10), and 0 otherwise.
22
23
24
25
26
an
PERIOD is a dummy variable which equals 1 if the audits were taken in 2006-10.

27 Chan and Chow (1997a) found that based on univariate analysis, Chinese tax authorities
M
28
29 focused on wholly foreign-owned (WFOE=1), Hong Kong-Taiwan sourced (SOURCE=1)
30
31
32 and small and medium-sized companies (SIZE) in 1992-93. Therefore, we expect the
d

33
34 coefficients 1 and 2 to be positive and 3 to be negative. The interaction variables,
te

35
36
37
PERIOD*WFOE, PERIOD*SOURCE, and PERIOD*SIZE, are used to test whether there is
p

38
39 any significant change in the factors affecting the FIEs being selected for audit. For example,
ce

40
41 if the coefficient on PERIOD*WFOE (5) is significantly negative, it will indicate that the
42
43
44 probability of wholly foreign-owned FIEs being selected for audit in 2006-10 is significantly
Ac

45
46 lower than that in 1992-93. According to our hypotheses, we expect 5 and 6 to be negative
47
48
49 and 7 to be positive. To test if WFOEs are still more likely to be selected for tax audit in
50
51 2006-10, we test the significance of the sum of the coefficients on WFOE and
52
53
54 PERIOD*WFOE (i.e., 1 + 5). Similar tests are performed for SOURCE and SIZE.
55
56
57 The control variable EXPORT controls the impact of FIE activity orientation on the
58
59 probability of being audited. Export-oriented FIEs (i.e., those FIEs that export 70 percent or
60
61
62
63
64 14
Page 16 of 32
65
 1
2 more of their products in a calendar year as specified by the Ministry of Commerce in China,
3
4 EXPORT=1) should have a higher probability of being audited because they are more likely
5
6
7 to shift profits out of China through transfer pricing according to Chan et al. (2006b).
8
9 Export-oriented firms also tend to have larger amount of related party transactions with
10
11

t
12 oversea entities than other firms (Chan and Lo, 2005).

ip
13
14 Prior studies on international transfer pricing in China have found that certain industrial
15

cr
16
17
sectors are more likely to undertake transfer pricing manipulations. Chan and Chow (1997b)
18

us
19 reveal that in the early 1990s, Chinese FIEs in audio and video equipment, garment and
20
21 textile, plastic products and chemicals industries tended to over-price their imports and
22
23
24
25
26
an
underprice their exports (i.e. not following the arms length principle) and thus reported

lower profit. To control for the potential effect of industries on the selection of transfer
27
M
28
29 pricing audits, we include INDUSTRY as a control variable for the regression. The dummy
30
31 variable equals 1 if a FIE is in the above-mentioned industries, and 0 for other industries.
32
d

33
34 We include a variable, ROS, to control for a firms profitability. ROS is defined as profit
te

35
36 after tax divided by sales. We also include a control variable, LOSS, to control for the effect
37
p

38
of the financial condition of the FIEs on the probability of being audited. LOSS equals 1 if
39
ce

40
41 profit before tax is less than 0, and 0 otherwise.1
42
43
44
Ac

45
46 5. Empirical Results
47
48 5.1. Descriptive Statistics
49
50
51 Table 1 provides descriptive statistics for the firm attributes of the 169 (81) FIEs which
52
53
1
54 We prepared a detailed correlation matrix of all regression variables to test for multicollinearity. Most of the
55 correlation coefficients between regression variables are significant at the 1 or 5 percent levels. Except the
56 correlation between ROS and LOSS (correlation coefficient =-0.644), the correlations between all other
57 independent variables are less than 0.30, indicating that multicollinearity is not a problem in our regression
58 analysis. Thus, we omit detailed presentation of the correlation matrix.
59
60
61
62
63
64 15
Page 17 of 32
65
 1
2 were selected for transfer pricing audits and 169 (81) FIEs which had related party
3
4 transactions but were not selected for transfer pricing audits in the 2000s (1990s). Table 1,
5
6
7 Column A, shows that 55.6 percent of the FIEs in the 2000s' audited sample are wholly
8
9 foreign-owned enterprises, 78.7 percent are sourced from Hong Kong or Taiwan, 27.2 percent
10
11

t
12 are large FIEs, 91.7 percent are export-oriented enterprises, 26.0 percent are in industries that

ip
13
14 have a higher risk of transfer pricing manipulation, and 31.9 percent have losses. Compared
15

cr
16
17
with the non-audited sample (Column B) which is in line with the respective national
18

us
19 distributions for FIEs in China (China Commerce Yearbook, 2011), the univariate tests show
20
21 that Hong Kong and Taiwan sourced FIEs, small FIEs, export-oriented FIEs, firms with lower
22
23
24
25
26 pricing audits in the 2000s.
an
profitability (ROS) and firms that have losses are more likely to be selected for transfer

27
M
28
29 Table 1, Column C, shows that 46.9 percent of the FIEs in the 1990s audited
30
31 sample are wholly foreign-owned enterprises, 92.6 percent are sourced from Hong Kong or
32
d

33
34 Taiwan, 11.1 percent are large FIEs, 90.1 percent are export-oriented enterprises, 34.6 percent
te

35
36 are in industries perceived to have a higher risk of transfer pricing manipulation, and 55.6
37
p

38
percent have losses. These audited sample distributions are comparable to those in the Chan
39
ce

40
41 and Chow (1997a) study, whereas the distributions of the non-audited sample are in line with
42
43 the respective national distributions for FIEs in China.2 Based on the results of univariate
44
Ac

45
46 tests for the comparison of distributions between the audited sample and non-audited sample,
47
48 we find that wholly foreign-owned enterprises, Hong Kong and Taiwan sourced FIEs,
49
50
51 export-oriented FIEs, small FIEs, and firms that have losses were more likely to be selected
52
53 for transfer pricing audits in 1990s. These results are consistent with our hypotheses and are
54
55
56
57 2
Chan and Chow (1997a) compare the audited sample with the national distribution rather than the non-audited
58 sample and they use registered capital as a measure of size. In addition, their study does not provide an analysis
59 about export-oriented enterprises and loss firms, nor does it deal with firms profitability.
60
61
62
63
64 16
Page 18 of 32
65
 1
2 similar to Chan and Chow (1997a)s findings. Thus, based on univariate tests, the results in
3
4 the 2000s are similar to those in the 1990s, except that, in the 2000s, wholly foreign-owned
5
6
7 enterprises are not significantly more likely to be audited than joint ventures.
8
9 [Insert Table 1 here]
10
11 5.2. Multivariate Analysis

t
12

ip
13
14 Table 2 reports logistic regression results for the likelihood that a FIE with different
15

cr
16 firm characteristics will be selected for transfer pricing audits by Chinese tax authorities.
17
18
Overall, the R-square of the model is 0.375, and the signs of the coefficients of the test

us
19
20
21 variables are all consistent with our hypotheses. 3 WFOE and SOURCE are positively
22
23
24
25
26
an
associated with AUDIT; whereas SIZE is negatively associated with AUDIT. Thus, wholly

foreign-owned enterprises, Hong Kong or Taiwan sourced FIEs, and small FIEs were more
27
M
28
29
likely to be selected for audit in the 1990s. In addition, the coefficients on PERIOD*WFOE
30
31 and PERIOD*SOURCE are negative and significant at 5 percent and 1 percent levels,
32
d

33 respectively; the coefficient on PERIOD*SIZE is positive and significant at the 1 percent

34
te

35
36 level. Thus, the audit focuses on wholly foreign-owned enterprises, Hong Kong or Taiwan
37
p

38 sourced FIEs, and small and medium FIEs have all been significantly reduced over the past
39
ce

40
41 two decades. These results in turn suggest that, compared with the 1990s, Chinese tax
42
43 authorities are now focusing more on larger Western (mainly U.S. and Japanese) FIEs.
44
Ac

45
46 [Insert Table 2 here]
47
48
49 We also find that the sum of coefficients of (i) WFOE and PERIOD*WFOE (0.016);
50
51
52
(ii) SOURCE and PERIOD*SOURCE (-0.462), and (iii) SIZE and PERIOD*SIZE (0.220)
53
54 are not significantly different from zero. That is, in the 2000s, tax authorities, instead of
55
56
57 3
We computed variance inflation factors (VIF) for our independent variables. The results show that the smallest
58 VIF is 1.047 and the largest VIF is 3.802. Therefore, multicollinearity is unlikely to be problematic in our
59 regression because all VIFs are lower than 10 (Gujarati, 1999; Green, 2008).
60
61
62
63
64 17
Page 19 of 32
65
 1
2 focusing on WFOEs, have a more balanced emphasis on auditing FIEs of different forms. Tax
3
4 audit efforts have also been shifted toward large Western FIEs to the extent that there is no
5
6
7 longer an audit focus on small and medium Hong Kong or Taiwan FIEs. This change can
8
9 significantly affect foreign investors in choosing their investment strategy.
10
11

t
12 In sum, the results of the logistic regression confirm that the form of investment,

ip
13
14
15 source of investment, and size of FIEs significantly affect the probability of being selected for

cr
16
17 transfer pricing audits in China in the 1990s but not in the 2000s. Our empirical evidence
18

us
19
20
shows that there are significant changes of audit focus between 1990s and 2000s.
21
22 Specifically, wholly foreign-owned FIEs were more likely to be audited in the early 1990s
23
24
25
26
an
than joint ventures. However, in the 2000s, the tax audit focus on this type of FIEs has been

27 reduced significantly and Chinese tax authorities no longer show any preference in audit
M
28
29 selection based on the form of investment. During the 1990s, Hong Kong and Taiwan sourced
30
31
32 FIEs and small FIEs were more likely to be audited. However, in the 2000s, they are no
d

33
34 longer more likely to be audited as Chinese tax audit expertise improves. Overall our
te

35
36
37
empirical results reveal that the audit focus on transfer pricing has gradually shifted toward
p

38
39 Western and larger FIEs.4 In a separate regression analysis based on 2000s data only, we find
ce

40
41 that export-oriented firms and loss firms are significantly more likely to be audited in the
42
43
44 2000s (coefficient for EXPORT=3.134, p-value=0.000, coefficient for LOSS=0.895, p-
Ac

45
46 value=0.027). These results suggest that Chinese tax authorities have now focused on the
47
48
49 activities and financial conditions of FIEs, consistent with the audit targets suggested in the
50
51 1998 and 2008 tax laws (Deloitte 2008).
52
53
54 5.3. Sensitivity Tests
55
56
57
4
58 The multiple regression results on the source and the size of investment in the 2000s are different from the
59 univariate results due to the correlations among the regression independent variables.
60
61
62
63
64 18
Page 20 of 32
65
 1
2 We perform sensitivity tests on our three hypotheses. First, with regard to the form of
3
4 investment, a number of FIEs included in our joint venture sample are corporative joint
5
6
7 ventures who often allow the foreign rather than the local partner to be in charge of daily
8
9 operations of the venture (Chan and Chow, 1997). Because of this difference in management
10
11

t
12 arrangements, we deleted these ventures from our sample and reran the regression analysis.

ip
13
14 The results are essentially the same as our main regression results in Table 2, thus we omit
15

cr
16
17
presentation of this detailed analysis.
18

us
19 Second, because Hong Kong and Taiwan have very different tax systems, we deleted all
20
21 Taiwanese FIEs from our Hong Kong-Taiwan sample and reran the regression as shown in
22
23
24
25
26
an
Table 3. Table 3 outcomes are similar to Table 2 except that the WFOE and PERIOD*WFOE

variables are no longer significant. This change indicates that there was no audit focus on
27
M
28
29 WFOEs in both the 1990s and the 2000s. Many Taiwanese WFOEs were audited both in the
30
31 1990s and the 2000s. Deleting them reduces the difference between WFOEs and joint
32
d

33
34 ventures as audit targets. In addition, we find a significant audit focus on non-Hong Kong
te

35
36 FIEs in the 2000s, as deleting the Taiwanese FIEs reveals a clearer shift of audit focus toward
37
p

38
Western FIEs.
39
ce

40
41 The tax rates in Hong Kong and Taiwan were lower than that of China except in 2008
42
43 and 2009 when Taiwan had the same tax rate as China. 5 To eliminate this potentially
44
Ac

45
46 confounding tax rate effect, we excluded Taiwanese FIEs for the 2008 and 2009 sample years
47
48 and reran the regression. As shown in Table 4, the results are basically the same as that of
49
50
51 Table 3 except that the coefficients on WFOE and PERIOD*WFOE become significant as in
52
53 Table 2. Both Tables 3 and 4 show a significant SOURCE variable in the 2000s (i.e.,2 +6)
54
55
56
57
58 5
The statutory tax rate in Taiwan was 25% in 2006-2009 and 17% in 2010, while China was 33% up to 2007
59 and 25% from 2008 onwards.
60
61
62
63
64 19
Page 21 of 32
65
 1
2 whereas Table 2 does not. It is thus interesting to observe that eliminating or reducing the
3
4 impact of Taiwanese FIEs in the contrast between Hong Kong-Taiwan FIEs vs. Western FIEs
5
6
7 reveals a significant shift of audit focus towards Western FIEs in the 2000s.
8
9 On firm size sensitivity, instead of defining the top 33% FIEs as large FIEs, we use the
10
11

t
12 median. We replaced SIZE by SIZE_M where SIZE_M equals 1 if a FIEs sales is above the

ip
13
14 median of the sample firms of a given period (i.e. 1992-93 or 2006-10), and 0 otherwise.
15

cr
16
17
The results are similar to those in Table 2, thus we omit detailed presentation of this analysis.
18

us
19 [Insert Tables 3 and 4 here]
20
21
22
23 6. Conclusion
24
25
26
an
This paper investigates, in two distinct time periods, the relation between
27
characteristics of FIEs in China and the probability of these FIEs being selected for transfer
M
28
29
30 pricing audits. Similar to Chan and Chow (1997a), based on univariate analysis we find that
31
32
d

WFOEs, Hong Kong and Taiwan sourced FIEs, and small FIEs were more likely to be
33
34
te

35 selected for transfer pricing audits in the 1990s. However, given changes in the business
36
37 environment in China, our results show that Chinese tax authorities have shifted their transfer
p

38
39
pricing audit focus away from wholly foreign-owned FIEs and Hong Kong or Taiwan sourced
ce

40
41
42 FIEs toward Western FIEs. In addition, as Chinese tax authorities gain more experience and
43
44
Ac

45
have developed more audit guidelines, audit focus has also shifted from small FIEs toward
46
47 larger FIEs. Export-oriented and loss firms also received special attention by tax authorities
48
49 in the late 2000s. In sum, the empirical evidence confirms that tax audit focus is not static. It
50
51
52 changes as the business environment and audit expertise change.
53
54 Our study provides evidence on how firm characteristics can affect tax audit focus and
55
56
57 how tax audit focus changes with the environment. As authorities devote more resources to
58
59
60
61
62
63
64 20
Page 22 of 32
65
 1
2 revenue collection, auditing more widely and imposing stricter penalties, MNCs come under
3
4 tighter scrutiny. Hence transfer pricing compliance has become a top priority for MNC
5
6
7 management. Our findings provide useful information for MNCs to evaluate international
8
9 financial management practices in light of their audit risk. Management of MNCs can gain a
10
11

t
12 better understanding of how tax audit focus shifts due to the changing business and

ip
13
14 investment environment and how firms of different characteristics should assess the risk of
15

cr
16
17
being selected for transfer pricing audits as well as future trends in developing economies.
18

us
19 MNCs can then better design their transfer pricing strategies, prepare necessary
20
21 documentation, and reduce the risk of onerous penalties.
22
23
24
25
26
an
27
M
28
29
30
31
32
d

33
34
te

35
36
37
p

38
39
ce

40
41
42
43
44
Ac

45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64 21
Page 23 of 32
65
 Figure 1. Economic Statistics of China, 1991-2010

Panel A: Average Annual Percentage Growth of Real GDP

1
2 16
3 14
4 12
Percentage

5 10
6 8
7 6
8 4
9 2
10 0
11

t
12

ip
13
14 Year
15

cr
16 Average annual percentage growth of real GDP
17
18 Note (a): The average percentages for 1991-2000 and 2001-2010 are 10.1 and 9.1, respectively.

us
19
20
21
22 Panel B: Annual Value of Foreign Trade
23
24
25
26
3500

3000
an
27 2500
M
28
US$ billion

29 2000
30
31 1500
32
d

33 1000
34 500
te

35
36 0
37
p

38
39
Year
ce

40
41
42 Value of foreign trade (US$bn)
43
44 Note (b): The average values for 1991-2000 and 2001-2010 are US$278.8 billion and US$1,623.6
Ac

45 billion, respectively.
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64 22
Page 24 of 32
65
 Figure 1. Economic Statistics of China, 1991-2010

1
2 Panel C: Annual Value of Foreign Direct Investment
3
4 120
5
6 100
7
8
80
US$ billion

9
10
11 60

t
12

ip
13 40
14
15 20

cr
16
17 0
18

us
19
20
Year
21
22
23 Value of foreign direct investment (US$bn)
24
25
26
an
Note (c): The average values for 1991-2000 and 2001-2010 are US$32.2 billion and US$75.1 billion,
respectively.
27
M
28
29
30 Panel D: Number of Foreign Investment Enterprises Contracted during the Year
31
32 90,000
d

33 80,000
34
te

70,000
35
36 60,000
Number

37 50,000
p

38
39 40,000
ce

40 30,000
41
20,000
42
43 10,000
44 0
Ac

45
46
47
48 Year
49
50 No. of foreign investment enterprises contracted during the year
51
52 Note (d): The average numbers for 1991-2000 and 2001-2010 are 33,436 and 34,678, respectively.
53
54
55 Sources: World Economic Outlook (IMF 1992-2011) and China Statistical Yearbook (1992-2011).
56
57
58
59
60
61
62
63
64 23
Page 25 of 32
65
 t
1

ip
2
3
4

cr
5 Table 1. Descriptive Statistics and Univariate Tests for Sample Firms
6
7 Data in 2000s Data in 1990s Univariate Tests
8

us
9
(n = 338) (n = 162)
10 A B C D
11 Audited Non-audited Audited Non-audited t-stat. t-stat.
12 Variables Sample Mean Sample Mean Sample Sample (A vs. B) (C vs. D)

an
13 Mean Mean
14
15
16 WFOE 0.5562 0.5089 0.4691 0.2222 0.871 3.400 ***
17

M
18 SOURCE 0.7870 0.6095 0.9259 0.7654 3.613 *** 2.882 ***
19
20 SIZE 0.2722 0.3964 0.1111 0.5556 -2.435 ** -6.761 ***
21

ed
22 EXPORT 0.9172 0.3373 0.9012 0.6296 13.734 *** 4.280 ***
23
24
INDUSTRY 0.2604 0.2308 0.3457 0.3210 0.630 0.331
25
26
27 ROS -0.0005
pt 0.0469 -0.0601 -0.4311 -2.924 *** 1.040
28
29 LOSS 0.3195 0.1420 0.5556 0.2840 3.950 *** 3.620 ***
ce
30 *** and ** indicate significance at the 1 percent and 5 percent levels respectively.
31
32 Definitions:
33 WFOE = 1, if a foreign investment enterprise (FIE) is a wholly foreign-owned enterprise, 0 otherwise
Ac

34
SOURCE = 1, if a FIE is sourced from Hong Kong or Taiwan, 0 if the FIE is sourced from the U.S., Japan or other countries
35
36 SIZE = 1, if a FIE's sales is above one third of the sample firms of a given period (1990s or 2000s), 0 otherwise
37 EXPORT = 1, if a FIE is export-oriented, 0 otherwise
38 INDUSTRY = 1, if a FIE is in an industry perceived as having a higher risk of transfer pricing manipulation, 0 otherwise
39 ROS = return on sales (i.e., profit after tax divided by total sales)
40
41
LOSS = 1, if profit before tax is less than 0, 0 otherwise
42
43
44
45
46
47
48 24
Page 26 of 32
49
 1
2 Table 2. Logistic Regression on the Change of Audit Focus on Transfer Pricing Audit: 1990s vs. 2000s
3
4
5
Regression equation:
6 Audit = 0 +1 WFOE +2 SOURCE +3 SIZE +4 PERIOD +5 PERIOD*WFOE
7 +6 PERIOD* SOURCE +7 PERIOD*SIZE +8 EXPORT + 9 INDUSTRY
8 +10 ROS +11 LOSS +
9
10
11 Variable Predicted Regression Z-Stat. P-value
12 Sign Coefficient

t
13 Intercept -3.963 -6.562 0.000

ip
14 WFOE + 1.106 2.627 0.009 ***
15
16 SOURCE + 2.657 5.247 0.000 ***

cr
17 SIZE - -2.473 -4.919 0.000 ***
18 PERIOD ? 2.046 3.956 0.000 ***
19 PERIOD*WFOE - -1.090 -2.163 0.031 **

us
20
21
PERIOD*SOURCE - -3.118 -5.841 0.000 ***
22 PERIOD*SIZE + 2.694 4.579 0.000 ***
23 EXPORT + 2.542 8.337 0.000 ***
24 INDUSTRY + -0.556 -1.953 0.051

an
25 ROS - 0.366 1.012 0.312
26
27 LOSS + 0.220 0.666 0.505
28
M
29 N 500
30 McFadden R-squared 0.375
31
32
33 Value F-statistic P-value
d

34 1 + 5 0.016 0.002 0.956

35 2 + 6 -0.462 2.114 0.147
te

36 3 + 7 0.220 0.532 0.465

37
38
p

39
40 ***, ** indicate significance at the 1 percent, 5 percent levels, respectively (two-tails).
41
ce

42
43
Definitions of variables:
44 AUDIT = 1, if a FIE was audited by Chinese tax authority due to transfer pricing
45 manipulation, 0 otherwise;
Ac

46
47 WFOE = 1, if a FIE is a wholly foreign-owned enterprise, 0 otherwise;
48
49
SOURCE = 1, if a FIE is sourced from Hong Kong or Taiwan, 0 if the FIE is sourced
50 from the U.S., Japan or other countries
51 SIZE = 1, if a FIEs sales is above one third of the sample firms in a period, 0
52 otherwise
53 PERIOD = 1, if the observation belongs to the 2006-2010 sample, 0 otherwise
54
55 EXPORT = 1, if a FIE is export-oriented, 0 otherwise
56 INDUSTRY = 1, if an FIE is in an industry perceived as having a higher risk of transfer
57 pricing manipulation, 0 otherwise
58 ROS = return on sales (i.e., profit after tax divided by total sales)
59 LOSS = 1, if profit before tax is less than 0, 0 otherwise
60
61
62
63
64 25
Page 27 of 32
65
 1
2
3 Table 3. Sensitivity Test on Hypothesis 2 (Source of Investment: Excluding Taiwanese FIEs)
4 Logistic Regression on the Change of Audit Focus on Transfer Pricing Audit: 1990s vs. 2000s
5
6
7 Regression equation:
8 AUDIT = 0 +1 WFOE +2 SOURCE +3 SIZE +4 PERIOD +5 PERIOD*WFOE
9 +6 PERIOD* SOURCE +7 PERIOD*SIZE +8 EXPORT + 9 INDUSTRY
10 +10 ROS +11 LOSS +
11
12

t
13 Variable Predicted Regression Z-Stat. P-value

ip
14 Sign Coefficient
15 Intercept -4.060 -6.273 0.000
16 WFOE + 0.461 1.045 0.296

cr
17 SOURCE + 2.366 4.555 0.000 ***
18 SIZE - -2.255 -4.287 0.000 ***
19
PERIOD ? 2.033 3.738 0.000 ***

us
20
21 PERIOD*WFOE - -0.618 -1.139 0.255
22 PERIOD*SOURCE - -3.088 -5.437 0.000 ***
23 PERIOD*SIZE + 2.332 3.647 0.000 ***
24 EXPORT + 3.026 8.113 0.000 ***

an
25 INDUSTRY + -0.593 -1.853 0.064
26 ROS - 0.688 1.289 0.198
27 LOSS + 0.242 0.679 0.497
28
N 410
M
29
30 McFadden R-squared 0.381
31
32 Value F-statistic P-value
33 1 + 5
d
-0.157 0.201 0.653
34
35 2 + 6 -0.721 3.941 0.048 **
te

36 3 + 7 0.077 0.046 0.830

37
38
***, ** indicate significance at the 1 percent, 5 percent levels, respectively (two-tails).
p

39
40
41 Definitions of variables:
ce

42 AUDIT = 1, if a FIE was audited by Chinese tax authority due to transfer pricing
43 manipulation, 0 otherwise;
44
45 WFOE = 1, if a FIE is a wholly foreign-owned enterprise, 0 otherwise;
Ac

46 SOURCE = 1, if a FIE is sourced from Hong Kong, 0 if the FIE is sourced from the U.S.,
47
Japan or other countries
48
49 SIZE = 1, if a FIEs sales is above one third of the sample firms in a period, 0
50 otherwise
51 PERIOD = 1, if the observation belongs to the 2006-2010 sample, 0 otherwise
52 EXPORT = 1, if a FIE is export-oriented, 0 otherwise
53 INDUSTRY = 1, if an FIE is in an industry perceived as having a higher risk of transfer
54 pricing manipulation, 0 otherwise
55 ROS = return on sales (i.e., profit after tax divided by total sales)
56 LOSS = 1, if profit before tax is less than 0, 0 otherwise
57
58
59
60
61
62
63
64 26
Page 28 of 32
65
 1
2
3 Table 4. Sensitivity Test on Hypothesis 2 (Source of Investment: Excluding Taiwanese FIEs for 2008
4 and 2009)
5
Logistic Regression on the Change of Audit Focus on Transfer Pricing Audit: 1990s vs. 2000s
6
7
8 Regression equation:
9 AUDIT = 0 +1 WFOE +2 SOURCE +3 SIZE +4 PERIOD +5 PERIOD*WFOE
10
+6 PERIOD* SOURCE +7 PERIOD*SIZE +8 EXPORT + 9 INDUSTRY
11
12 +10 ROS +11 LOSS +

t
13

ip
14 Variable Predicted Regression Z-Stat. P-value
15 Sign Coefficient
16

cr
17 Intercept -4.096 -6.480 0.000
18 WFOE + 1.003 2.352 0.019 **
19 SOURCE + 2.513 4.879 0.000 ***

us
20 SIZE - -2.453 -4.879 0.000 ***
21
22 PERIOD ? 2.233 4.170 0.000 ***
23 PERIOD*WFOE - -1.236 -2.334 0.020 **
24 PERIOD*SOURCE - -3.283 -5.857 0.000 ***

an
25 PERIOD*SIZE + 2.411 3.956 0.000 ***
26
EXPORT + 2.806 8.206 0.000 ***
27
28 INDUSTRY + -0.469 -1.531 0.126
M
29 ROS - 0.422 1.132 0.258
30 LOSS + 0.293 0.856 0.392
31
32
33
N 470
d

34 McFadden R-squared 0.413

35
te

36 Value F-statistic P-value

37 1 + 5 -0.233 0.474 0.491
38
2 + 6 -0.770 4.897 0.027 **
p

39
40 3 + 7 -0.042 0.015 0.903
41
ce

42 ***, ** indicate significance at the 1 percent, 5 percent levels, respectively (two-tails).

43
44
45 Definitions of variables:
Ac

46 AUDIT = 1, if a FIE was audited by Chinese tax authority due to transfer pricing
47 manipulation, 0 otherwise;
48
49 WFOE = 1, if a FIE is a wholly foreign-owned enterprise, 0 otherwise;
50 SOURCE = 1, if a FIE is sourced from Hong Kong, 0 if the FIE is sourced from the U.S.,
51 Japan or other countries
52
53
SIZE = 1, if a FIEs sales is above one third of the sample firms in a period, 0
54 otherwise
55 PERIOD = 1, if the observation belongs to the 2006-2010 sample, 0 otherwise
56 EXPORT = 1, if a FIE is export-oriented, 0 otherwise
57 INDUSTRY = 1, if an FIE is in an industry perceived as having a higher risk of transfer
58 pricing manipulation, 0 otherwise
59 ROS = return on sales (i.e., profit after tax divided by total sales)
60 LOSS = 1, if profit before tax is less than 0, 0 otherwise
61
62
63
64 27
Page 29 of 32
65
 1
2
3
4 References
5
6
7 Ambler, T., and M. Witzel. (2000). Doing Business in China. New York, NY: Routledge.
8
9 Beck, P., and W. Jung. (1989). Taxpayers reporting decisions and auditing under
10
11
information asymmetry. The Accounting Review 64 (July), 468-87.
12

t
13 Bell, K. A. (2012). One-third of multinationals audited over three years, Deloitte webcast

ip
14 finds. Tax Management Transfer Pricing Report 21(2), 74.
15
16
Bloomberg News. (2010). People's Bank of China statement on Yuan exchange-rate

cr
17
18 flexibility. (June 19). Available at: www.bloomberg.com
19

us
20 Chan, K. H., and L. Chow. (1997a). An empirical study of tax audits in China on
21 international transfer pricing. Journal of Accounting and Economics (May), 83-112.
22
23
24 Chan, K. H., and L. Chow. (1997b). International transfer pricing for business operations

an
25 in China: Inducements, regulation and practice. Journal of Business Finance and
26 Accounting, 24 (October & December), 1269-1289.
27
28
Chan, K. H., K. Lin, and P. Mo. (2006a). A political-economic analysis of auditor
M
29
30 reporting and auditor switches. Review of accounting Studies 11(1), 21-48.
31
32 Chan, K. H., and A. Lo. (2004). The influence of management perception of
33
d

34 environmental variables on the choice of international transfer pricing methods.

35 International Journal of Accounting 39, 93-110.
te

36
37 Chan, K. H., and A. Lo. (2005). International transfer pricing in China: Post WTO.
38
Hong Kong: Sweet & Maxwell Asia.
p

39
40
41 Chan, K. H., and A. Lo., and P. Mo. (2006b). Managerial autonomy and tax compliance:
ce

42 An empirical study on international transfer pricing. The Journal of the American

43 Taxation Association 28 (2), 1-22.
44
45
Ac

46 China Commerce Yearbook. (2011). Beijing, China. China Commerce and Trade Press.
47
48 China Statistical Yearbook. (1992-2011). Beijing, China. China Statistics Press.
49
50
51 Deloitte. (2008). China Master Tax Guide. CCH Asia Pte Limited: Hong Kong.
52
53 Deloitte. (2009). New transfer pricing requirements in China. Hong Kong: Deloitte.
54
55
56 Deloitte. (2011). Taxation and Investment in China 2011. PRC: Deloitte Touche
57 Tohmatsu Limited.
58
59
60 Ernst & Young. (2010). 2010 Global Transfer Pricing Survey. New York, NY: Ernst &
61
62
63
64 28
Page 30 of 32
65
 1
2
3
4 Young International Limited.
5
6
7 Ernst & Young. (2011). China Transfer Pricing Developments. Available at:
8 http://tmagazine.ey.com.
9
10
11
Ernst & Young. (2012). 2012 Global Transfer Pricing Tax Authority Survey. New York,
12 NY: Ernst & Young International Limited.

t
13

ip
14 Green, W. (2008). Econometric Analysis. Upper Saddle River, NJ: Pearson/Prentice Hall.
15
16
Gujarati, D. (1999). Essentials of Econometrics. 2nd Edition. New York, NY:

cr
17
18 McGraw-Hill.
19

us
20 Guoshuifa No. 237. (1992). Tax Circular No. 237: Definition of Related Party and
21 Disclosure Requirement. Beijing, PRC: China Taxation Publishing House.
22
23
24 Guoshuifa No. 242. (1992). Administration of Tax on Business Transactions Between

an
25 Affiliated Enterprises Rules. Beijing, PRC: China Taxation Publishing House.
26
27
28
Guoshuifa No. 59. (1998). Tax Circular No. 59: Tax Regulations for Related Party
Transactions. Beijing, PRC: China Taxation Publishing House.
M
29
30
31 Guoshuifa No. (118). 2004. Tax Circular No. 118: Advance Pricing Agreement (APA)
32 Implementation Rules. Beijing, PRC: China Taxation Publishing House.
33
d

34
35 Guoshuifa No. 2. (2009). Tax Circular No. 2: Disclosure and documentation obligations.
te

36 Beijing, PRC: China Taxation Publishing House.

37
38
IMF (International Monetary Fund). (2004). World Economic Outlook. Washington D.C.:
p

39
40 IMF.
41
ce

42 IMF (International Monetary Fund). (2010). World Economic Outlook. Washington D.C.:
43 IMF.
44
45
Ac

46 IMF (International Monetary Fund). (2011). World Economic Outlook. Washington D.C.:
47 IMF.
48
49
Klassen, K. (1997). The impact of inside ownership concentration on the trade-off
50
51 between financial and tax reporting. The Accounting Review 72(3), 455-474.
52
53 Liu, T. S. (1999). Hong Kong Chinese Merchants: Corporate Management. Hong Kong,
54 PRC: Joint Publishing Co.
55
56
57 MOC (Ministry of Commerce), PRC. (2012). Import and export statistics for FIEs in
58 China. Available at: www.mofcom.gov.cn.
59
60
61
62
63
64 29
Page 31 of 32
65
 1
2
3
4 Pappas, L. A. (2012). Practitioners note are of audit scrutiny as China steps up
5
6 international enforcement. Tax Management Transfer Pricing Report 2012 21(2),
7 73-74.
8
9 Rhoades, S.C. (1999). The impact of multiple component reporting on tax compliance
10
11
and audit strategies. The Accounting Review 74 (January), 63-85.
12

t
13 Sansing, R. (1993). Information acquisition in a tax compliance game. The Accounting

ip
14 Review 68(October), 874-84.
15
16
SAFE (State Administration of Foreign Exchange). (2012). Circular No. 59 on further

cr
17
18 improving and adjusting foreign exchange administrative policies.
19

us
20 SAT (State Administration of Taxation). (2004a). GuoShuiHan No. 70.
21
22
23 SAT (State Administration of Taxation). (2009). 2009 OECD transfer pricing training.
24 Available at: http://www.chinatax.gov.cn/n8136506/n8136563/n8136889/n8747853/.

an
25
26 Scholes, M.S., G. P. Wilson, and M. A. Wolfson. (1992). Firms responses to anticipated
27
28
reductions in tax rates: The tax reform act of 1986. Journal of Accounting Research
(30): 161-185.
M
29
30
31 UNCTAD. 2011. World Investment Report 2011. Available at: www.unctad.org/wir.
32
33
d

34 Wang, B. and H. Wang. (2011). Chinese manufacturing firms overseas direct investment:
35 Patterns, motivations and challenges. Rising China: Global Challenges and
te

36 Opportunities, Jane Golley and Ligang Song, eds. (ANU Press).

37
38
Watts, R. L., and J. L. Zimmerman. (1990). Positive accounting theory: A ten year
p

39
40 perspective. The Accounting Review 65, 131-156.
41
ce

42 Yin, Q., and C.S. Cheng. (2004). Earnings management of profit firms and loss firms in
43 response to tax rate reductions. Review of Accounting & Finance 3(1), 67-92.
44
45
Ac

46 Zhang, K. H. (2006). Foreign direct investment in China. Canadian Foreign Policy, 13(2),
47 35-50.
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64 30
Page 32 of 32
65