ownership structure tax avoidance

ARTICLE IN PRESS Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ Contents lists available at ScienceDirect Journal of Contemporary Accounting & Economics j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / j c a e Original Search Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China Q1 Grant Richardson a,*, Bei Wang b, Xinmin Zhang c a 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 School of Accounting and Finance, Business School, The University of Adelaide, 10 Pulteney Street, Adelaide, South Australia 5005, Australia School of Public Administration, University of International Business and Economics, Room 610, Ning Yuan Building, 10 Huixin East Road, Chaoyang District, Beijing 100029, China c School of International Business, University of International Business and Economics, 10 Huixin Dongjie Road, Chaoyang District, Beijing 100029, China b A R T I C L E I N F O Article history: Received June 2014 Received in revised form 10 June 2016 Accepted 10 June 2016 Available online Keywords: Ownership structure Private firms Corporate tax avoidance China A B S T R A C T This study examines the association between ownership structure and corporate tax avoidance in publicly-listed private firms in China We find a significant non-linear association between ownership concentration and tax avoidance that exhibits an inverted U-shaped pattern At a lower level, increased ownership concentration is positively associated with tax avoidance due to the entrenchment effect However, beyond the minimum level necessary for effective control, concentrated ownership through voting rights is negatively associated with tax avoidance because of the alignment effect We also find a significantly positive association between pyramidal ownership structure and tax avoidance due to the entrenchment effect When voting rights and cash-flow rights diverge, a lower level of cash-flow rights fails to offer the controlling owner incentive alignment sufficient to reduce the entrenchment effect and tax avoidance © 2016 Elsevier Ltd All rights reserved 39 Introduction 40 41 42 43 44 This study examines the association between ownership structure and corporate tax avoidance1 in publicly-listed private firms in China.2 We analyze the agency problems that may arise when firms feature: (1) concentrated ownership through voting rights, and (2) concentrated ownership where voting rights exceed cash-flow rights (i.e a pyramidal ownership structure3) 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 Q7 * Corresponding author School of Accounting and Finance, Business School, The University of Adelaide, 10 Pulteney Street, Adelaide, South Australia 5005, Australia Tel.: +61 8313 0582; fax: +61 8313 0170 E-mail address: grant.richardson@adelaide.edu.au (G Richardson) We define corporate tax avoidance broadly in this study as the downward management of taxable income via tax-planning activities Our definition thus covers both legal (and quasi-legal) and illegal activities (e.g Frank et al., 2009; Chen et al., 2010) The benefit of increased tax avoidance is greater tax savings for the firm that can reduce the amount of income tax expenses on its financial statements and boost profits (Scholes et al., 2009) Thus, more funds are made available to the controlling shareholder for the purpose of expropriation or self-dealing transactions where profits are transferred to other controlled firms via transfer pricing (tunneling) arrangements, for example Prior research on tunneling in China shows that the controlling owners of private firms regularly use such practices (e.g making intercorporate loans to related parties) to expropriate wealth from minority shareholders (e.g Jiang et al., 2010) We specifically focus our attention on publicly-listed Chinese private firms instead of state-owned enterprises (SOEs) because they all face the same dominant agency problem which is between the controlling owner and minority shareholders, while the dominant agency problem in SOEs is between managers and shareholders (Fan et al., 2013; Liu et al., 2016; Pan and Tian, 2016; Villalonga and Amit, 2006) This issue is discussed in further detail later in this paper in terms of ownership structure and agency theory A pyramidal structure is defined as ownership of the majority of the stock of one firm that, in turn, holds the majority of that of another firm, a process that can be repeated any number of times (Claessens et al., 2000) http://dx.doi.org/10.1016/j.jcae.2016.06.003 1815-5669/© 2016 Elsevier Ltd All rights reserved Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ In fact, it is well documented in prior research that agency problems are highly associated with ownership structure (e.g Shleifer and Vishny, 1997; Guedhami et al., 2014) This study helps to answer recent calls made in the literature by Scholes et al (2009) and Hanlon and Heitzman (2010) for research on the influence of agency costs and ownership structure on tax avoidance We consider private Chinese firms in this study as prior U.S research investigating the influence of different ownership structures (e.g family ownership and dual-class stock) on tax avoidance provides conflicting results (e.g Klassen, 1997; Chen et al., 2010; McGuire et al., 2014) These conflicting results may be attributable to the diffuse stock ownership environment which prevails in the U.S.4 in contrast to most other countries where concentrated stock ownership is more prevalent (Claessens et al., 2000; Faccio and Lang, 2002; Johnson et al., 2000; La Porta et al., 1998; Shleifer and Vishny, 1997) Hence, the concentrated ownership structure that characterizes private firms in China renders them a potentially more compelling set of firms to examine ownership structure’s influence on tax avoidance We develop three arguments in this study about the association between ownership structure and tax avoidance Our first argument deals with the entrenchment effect of concentrated ownership We argue that increasing the voting rights of controlling owners provides entrenchment against corporate governance input by minority shareholders (Morck et al., 1988; Fan and Wong, 2002) When the controlling owner is entrenched through his voting power, there is a greater likelihood that he will expropriate shareholder wealth through tax avoidance Desai and Dharmapala (2006) claim that tax avoidance facilitates insiders’ ability to expropriate such wealth (i.e extract private benefits) due to its complexity and intention to obscure the underlying economics of a business transaction Our second argument concerns the alignment effect of concentrated ownership Specifically, any further increase in the controlling shareholder’s ownership share can alleviate the problem of entrenchment and reduce the related agency costs by aligning the interests of the controlling owners with those of minority shareholders (Fan and Wong, 2002; Shleifer and Vishny, 1986) Gomes (2000) claims that a high ownership concentration is a signal of a controlling owner’s commitment to building a reputation for not expropriating minority shareholders’ wealth The alignment effect thus suggests that increasing the ownership concentration beyond the minimum level necessary for effective control reduces opportunistic behavior by the controlling owners and their incentive to engage in tax avoidance Chen et al (2010) find that family-controlled firms are less tax avoidant than non-family-controlled firms as family owners are willing to forgo tax benefits to avoid the nontax cost of a potential price discount to the firm’s share price that can arise from minority shareholder concern about managerial rent extraction Our third argument deals with the entrenchment/alignment effect when voting rights exceed cash-flow rights The main outcome of a divergence between voting and cash-flow rights is that the controlling owner becomes entrenched with a high level of control, while at the same time the low equity ownership level provides a negligible degree of alignment with minority shareholders (Fan and Wong, 2002) Such divergence strengthens the entrenchment effect (La Porta et al., 1999) and weakens the alignment effect (Fan and Wong, 2002) Francis et al (2005) find that dual-class ownership structures aid rent extraction as they allow insiders to concurrently extract private benefits and avoid the associated cash-flow concerns.5 So, it is possible that the controlling owners of firms typified by a wide degree of such separation may engage in higher levels of tax avoidance to expropriate wealth from shareholders and avoid the related cash-flow consequences When control exceeds ownership, the entrenchment effect argument suggests that a high ownership concentration is positively associated with tax avoidance Hence, when voting rights and cash-flow rights diverge, a lower level of cash-flow rights may fail to provide controlling shareholders with incentive alignment sufficient to reduce the entrenchment effect and thus tax avoidance Employing a sample of 207 publicly-listed private Chinese firms over the 2005–2010 period (1242 firm-year observations), we find a significant non-linear association between ownership concentration and tax avoidance that exhibits an inverted U-shaped pattern At a lower level, increased ownership concentration is positively associated with tax avoidance because of the entrenchment effect However, beyond the minimum level required for effective control, concentrated ownership through voting rights is negatively associated with tax avoidance due to the alignment effect We also observe a significantly positive association between pyramidal ownership structure and tax avoidance owing to the entrenchment effect When voting rights and cash-flow rights diverge, a lower level of cash-flow rights fails to offer the controlling owner incentive alignment sufficient to reduce the entrenchment effect and tax avoidance Our results are robust to additional analysis which considers the endogeneity of ownership structure (i.e instrumental variables two-stage least squares (2SLS) regression Q2 analysis and difference-in-difference (DID) analysis) This study makes the following contributions First, to the best of our knowledge, this is one of the first studies to simultaneously consider the issues of concentrated ownership through voting rights, concentrated ownership in the case of voting rights that exceed cash-flow rights, and tax avoidance in terms of entrenchment and alignment effects It thus extends the literature on the link between ownership structure and tax avoidance Second, this study documents new evidence on the: (1) non-linear U-shaped association between ownership concentration and tax avoidance which is indicative of entrenchment and alignment effects, and (2) the negative association between the divergence of voting and cash-flow rights 119 120 121 122 123 Gompers et al (2010) finds that dual-class firms (i.e firms in which voting rights exceed cash-flow rights) represent only 6% of all publicly-listed firms in the U.S See also Masulis et al (2009) Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 and tax avoidance which is indicative of the entrenchment effect In our scenario, such a divergence exacerbates the entrenchment effect and tax avoidance as a lower degree of cash-flow rights fails to provide the controlling owner with sufficient incentive to reduce his rent extraction behavior Our results are thus consistent with the view that ownership structure plays an important role in determining whether firms actively participate in tax avoidance Third, this study also contributes to research analyzing the effect of dual-class stocks on firm value and corporate behavior Prior research shows that firm value decreases as the difference between voting rights and cash-flow rights increases (e.g Gompers et al., 2010) We extend prior research exploring the way in which a dual-class ownership structure leads to a reduction in firm value (e.g Francis et al., 2005; Masulis et al., 2009) by analyzing the association between the divergence of voting and cash-flow rights and tax avoidance Our findings suggest that increased tax avoidance by firms whose owners have excessive control rights is one possible explanation for such a reduction Finally, compared to the results of U.S research in this area (e.g Chen et al., 2010; McGuire et al., 2014), our results are likely to be more relevant to those parts of the world where a concentrated ownership structure is more common than a diffuse stock ownership structure which is the situation in the U.S The remainder of this paper proceeds as follows Section describes the institutional background of publicly-listed private firms operating in China in addition to the institutional background of the corporate tax system in China Section provides theory and develops our hypotheses Section explains our research design Section reports our empirical results, while Section carries out additional analysis which considers the endogeneity of ownership structure Finally, Section concludes the paper 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 Institutional background 2.1 Institutional background of private firms in China Over the past three decades, China has enjoyed significant economic growth in the process of moving from a centrally planned to a market-oriented economy Two major reforms have been critical to China’s economic transformation: (1) the redistribution of decision-making powers between the central and local governments, and (2) the emergence of the private sector, which mainly comprises firms controlled by entrepreneurs One of the key events to occur during this transformation has been the relative decline of SOEs and the rise of private firms (Ding et al., 2007) For ideological reasons, private firms were restricted from operating in China until the late 1980s However, in 1987, the 13th National Congress of the Communist Party of China (CPC) recognized the private sector as a necessary extension of the state sector In the years since, China has formulated several major policies to support the development of private firms,6 which have experienced rapid development (Chen et al., 2011) As per the China Statistical Yearbook 2011 (National Bureau of Statistics of China, 2012), there were 273,259 domestic private firms with annual sales greater than RMB5 million in 2011, representing 72% of the total number of domestic firms in China In that year, the profits of private firms totaled RMB1510 billion, accounting for 40% of the profits of all domestic firms in China There has also been a steady increase in the number of private firms listed on the Chinese stock market (Ding et al., 2007) China’s stock market was originally designed as a platform to raise investment capital for SOEs, and stock market financing initially showed a large bias toward these firms in comparison with private firms While the Chinese stock market saw the first listing of a private firm in 1992, the number listed between 1992 and 1997 was small, given the rapid increase in market capitalization and total number of listed firms (Ding et al., 2007) In 1997, fewer than 6% of listed firms were privately owned despite the increasing importance of private firms to the Chinese economy Yet, 1998 saw a surge in the number of such listings and by the end of 2002, there were 216 listed private firms accounting for 18% of all listed firms (Chinese Association of Public Companies) (CAPCO, 2012) Fig shows that between 2002 and 2010, when the number of listed firms in China grew from 1224 to 2063, the number of listed private firms grew from 216 to 970 As of 2010, private firms accounted for 47% of China’s publicly-listed firms (CAPCO, 2012) 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 2.2 Institutional background of the corporate tax system in China Prior to China passing the uniform version of the Corporate Income Tax Law in 2008, domestic firms were subject to a corporate tax rate of 33% (Li, 2007) The central government also provided many preferential tax policies to specific regions and industries with the aim of balancing the development of different regions and enhancing preferred industries (Li, 2007) Q3 Before 2002, local governments also provided tax rebates to local enterprises (Wu et al., 2012a, 2012b) In 1994, China enacted the Corporate Income Tax Code to reform its tax policies Taxes were classified into central and local, and the National Taxation Bureau and provincial bureaus were separately responsible for collecting taxes Since 1994, local revenue has been re-characterized as the revenue from local taxes and the local portion of shared taxes (Wu et al., 2012a, 2012b).7 Since 2002, In 1997, the 15th CPC National Congress officially recognized the private sector as an important part of the economy, thereby accelerating the development of private firms In 2002, the 16th CPC National Congress emphasized its willingness to guide and support the development of the country’s private sector It was during this Congress that the CPC first altered its official constitution to allow private firm entrepreneurs to join The major local taxes are income taxes from all enterprises (except central government enterprises), business taxes from the sales of services and personal income taxes (Wu et al., 2012a, 2012b) Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 2500 2063 Number of Listed Firms 2000 1500 1224 1000 284 370 35% 25% 970 388 462 40% 30% 18% 216 185 40% 22% 500 1434 1377 1381 28% 27% 1287 37% 38% 1718 34% 1625 1550 50% 47% 45% 520 618 687 2002 2003 2004 2005 2006 2007 2008 2009 2010 Year 20% Total Private Firms Ratio 15% 10% 5% 0% Fig Number of listed firms in China by year (2002 to 2010) Source: CAPCO (2012) Available at: http://www.capco.org.cn/zhuanti/cjz/index.html 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 the tax revenue collected from SOEs and private firms has been shared between the central and the local governments (Wu and Yue, 2009) The lack of efficiency in the tax administration in China provides a strong incentive for Chinese firms to engage in tax avoidance (Cai and Liu, 2009; Chan and Mo, 2000) Theory and hypotheses development 3.1 Alternative explanations for concentrated ownership The ownership structure of publicly-listed private firms in China is characterized by the dominance of one primary owner who normally also participates in firm management (Chen et al., 2011) The extant literature offers two alternative explanations for the concentrated ownership of China’s private firms: (1) the enforcement of property rights, and (2) the absence of specialized intermediaries in the capital market (i.e institutional voids) The enforcement of property rights by individual owners plays an important role in economies in which the state does not effectively enforce those rights (Fan and Wong, 2002) Hence, the property rights framework predicts that concentrated ownership will be observed in economies in which the state fails to enforce property rights La Porta et al (1997, 1998) further claim that the benefits of concentrated ownership are greater in less-developed countries In fact, involuntary loss of control, which would result in minority shareholder status, may constitute such a costly proposition in terms of surrendering the private benefits of control that controlling shareholders will prefer to maintain control La Porta et al (1999) consider this supposition by examining the ownership concentration of the three major shareholders of many of the world’s largest firms They find weak legal and institutional environments (“law and enforcement”) to be associated with highly concentrated share ownership It thus seems likely that existing ownership structures are primarily an equilibrium response to the domestic legal environments in which firms operate (La Porta et al., 1999) Consequently, the private enforcement of property rights is a likely reason for the concentrated ownership seen in China’s private firms, which are often confronted by a weak legal system, poor law enforcement and corruption Hence, control rights may be the only way for private shareholders to protect their cash-flow rights (Bertrand et al., 2008) Concentrated ownership can also exist because of institutional voids, namely, the absence of specialized intermediaries in the capital market (Khanna and Palepu, 1997, 2000) For example, most firms in East Asia are not only characterized by concentrated ownership, but are also affiliated with a business group (Claessens et al., 2000) The largest ultimate shareholder with overall control of the group can deploy capital among the affiliated firms; a situation that often benefits the development of those operating in emerging capital markets (Fisman and Khanna, 2004; Khanna and Yafeh, 2005) While the Chinese economy is transforming itself from a government-controlled to market-oriented model, government policies at both the central and local levels still dominate the allocation of resources via licensing, the granting of land-use rights and the control of access to financial capital Banking remains one of the few industries in China still under strict government Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 control and relative to SOEs, private firms remain at a disadvantage in obtaining financial resources (Cai and Liu, 2009) Thus, the largest ultimate shareholder will choose to retain concentrated ownership control and organize the business group into a pyramidal structure due to the financing advantages and resulting payoff when newly created firms are expected to yield low-security benefits relative to investment requirements (Almeida and Wolfenzon, 2006) 3.2 Ownership structure and agency theory The degree of ownership concentration affects the nature of contracting, thereby producing agency problems between managers and outside shareholders When ownership is diffuse, as is typically the case in the U.S., agency problems stem from conflicts of interest between outside shareholders and managers who own an insignificant amount of equity in the Q4 firm (Jensen and Meckling 1976; Shleifer and Vishny, 1997) (“Type I agency problem”) However, when ownership is concentrated at a level at which an owner obtains effective control of the firm, as is commonly the case in Asia and most other locations outside the U.S (Claessens et al., 2000, 2002; Faccio and Lang, 2002; Johnson et al., 2000; La Porta et al., 1998), the nature of the agency problem shifts away from manager–shareholder conflicts to conflicts between the controlling owner (who, effectively, is also the manager) and minority shareholders (“Type II agency problem”) We therefore focus our attention on the Type II agency problem in terms of the entrenchment and alignment effects for a sample of publicly-listed Chinese private firms.8 3.2.1 Entrenchment effect Acquiring effective control of a firm allows a controlling owner to strongly influence firm operations, especially profit distribution among shareholders Although minority shareholders are entitled to cash-flow rights proportionate to their share investments, they face the possibility that an entrenched controlling owner may opportunistically deprive them of their rights (Fan and Wong, 2002) An outcome of controlling owner entrenchment is direct expropriation (e.g the controlling shareholder benefits from self-dealing transactions in which profits are transferred to other controlled firms via transfer pricing/tunneling arrangements) The controlling owner can also engage in indirect expropriation by pursuing non-profitmaximizing activities in return for personal benefits (e.g related-party transactions pertaining to investment projects with a negative net present value) The entrenchment problem created by a controlling owner is similar to the managerial entrenchment problem discussed in prior research (e.g Morck et al., 1988; Shleifer and Vishny, 1997; McConnell and Servaes, 1990) Increased managerial ownership can entrench managers, who become increasingly less subject to governance by the firm’s board of directors and to discipline by the market for corporate control (Fan and Wong, 2002) 3.2.2 Alignment effect However, any further increase in the controlling shareholder’s ownership share could possibly alleviate the problem of entrenchment and reduce the related agency costs by aligning his interests with those of minority shareholders (Fan and Wong, 2002; Shleifer and Vishny, 1986) A higher ownership stake in a firm gives the controlling owner stronger voting and cash-flow rights Stronger cash-flow rights make it costlier for the controlling shareholder to divert the firm’s cash flow for private gain, whereas voting rights beyond those necessary for effective control fail to produce further entrenchment In fact, an ownership concentration in excess of effective control sends a signal to investors that the controlling shareholder is committed to alignment with minority shareholders rather than to direct or indirect expropriation from them (Gomes, 2000) Ownership concentration thus has an incentive alignment effect: increasing that concentration beyond the minimum level needed for effective control improves the interest alignment between the controlling owner and the minority shareholders, thereby reducing the damaging effects of entrenchment on the latter (Fan and Wong, 2002) 3.2.3 Entrenchment effect when control exceeds ownership Ownership arrangements are further complicated by pyramidal structures that allow controlling owners to commit themselves to a small amount of equity investment while maintaining tight control of the firm, thereby creating a separation between control (voting rights) and ownership (cash-flow rights) (Claessens et al., 2000, 2002; Fan and Wong, 2002; La Porta et al., 1999) A major outcome of the divergence between voting and cash-flow rights is that the controlling owner becomes entrenched with a high level of control while his low equity ownership level provides only a minimal degree of alignment with minority shareholders In this scenario, a controlling owner can extract wealth from the firm, receiving the entire benefit while assuming only a fraction of the cost (Fan and Wong, 2002).9 In short, once the controlling owner has achieved effective control of the firm, ownership concentration has two incentive effects: (1) the entrenchment effect, and (2) the alignment effect If there is no separation between voting and cash-flow See also footnote above To further explain, consider the following example of a simple pyramidal structure A business person owns 30% of the stock in publicly-listed Firm X, which in turn owns 20% of the stock in Firm Y Thus, this business person controls 20% of Firm Y, which is the weakest link in the chain of voting rights At the same time, he owns around 6% of the cash-flow rights in Firm Y (i.e the product of the two ownership stakes down the chain) Under this ownership structure, the business person “pays” only $6 for every $100 expropriated from Firm Y If stock pyramids or cross-shareholdings are utilized to consolidate control, they will also give rise to a separation between ownership and control, thus aggravating the entrenchment problem of the controlling owner (Fan and Wong, 2002) Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS 282 283 284 G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ rights, concentrating ownership beyond the minimum level necessary for effective control improves interest alignment and thus alleviates the entrenchment effect When these rights diverge, however, a lower level of cash-flow rights may fail to offer the controlling owner incentive alignment sufficient to reduce the entrenchment effect 285 286 287 288 289 290 291 3.3 Ownership structure and corporate tax avoidance We now consider the hypothesized associations between ownership structure and tax avoidance, positing three arguments that possibly explain them Our first and second arguments consider the entrenchment and alignment effects, respectively, as discussed in the foregoing paragraphs in terms of concentrated ownership Our third argument considers the entrenchment effect/alignment effect when control exceeds ownership (i.e pyramidal structure) 292 293 294 295 296 297 298 299 300 301 302 303 3.3.1 Entrenchment effect argument: concentrated ownership Although a firm’s share ownership structure defines its agency problem, it can also affect its tax agenda When an owner has effective control of a firm, he also controls the formulation of its tax policies When the controlling owner is entrenched through his voting power, he is more likely to expropriate shareholder wealth by means of tax avoidance Klassen (1997) suggests that firms with concentrated ownership are more concerned with tax avoidance, and thus are more likely to initiate tax-reduction strategies Desai and Dharmapala (2006) argue that tax avoidance can facilitate insiders’ ability to expropriate shareholder wealth (i.e extract private benefits) because it is complex and designed to obscure the underlying economics of a business transaction Thus, in firms where the controlling owner is entrenched through his voting power, it is possible that he may engage in higher levels of tax avoidance as insiders view such avoidance as a means of expropriating wealth from minority shareholders The entrenchment effect argument thus suggests that a high ownership concentration is positively associated with tax avoidance 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 3.3.2 Alignment effect argument: concentrated ownership One possible way to alleviate the problem of controlling owner entrenchment is to further increase his ownership stake (Fan and Wong, 2002) A higher ownership stake gives the controlling owner stronger voting and cash-flow rights in the firm Once he has achieved effective control of the firm, however, a further increase in voting rights does not lead to additional entrenchment, but his stronger cash-flow rights mean that it becomes costlier to divert the firm’s cash-flow for private gain (Fan and Wong, 2002) A high ownership concentration can also serve as a credible signal of the controlling owner’s commitment to building a reputation for not expropriating minority shareholders via tax avoidance This commitment is credible as minority shareholders understand that if the controlling owner suddenly extracts high levels of private benefits through tax avoidance while holding a significant number of shares, the stock price will be discounted, thus reducing the majority owner’s share value In equilibrium, the majority shareholder will hold a large ownership stake and the firm’s stock price will be higher (Chen et al., 2010) Chen et al (2010) investigate whether family-controlled firms are more tax avoidant than non-family-controlled firms They find the former to be less tax avoidant, as family owners are willing to forgo tax benefits to avoid the non-tax cost of a potential price discount to the firm’s share price that can arise from minority shareholder concern about managerial rent extraction masked by tax avoidance Thus, ownership concentration has an incentive alignment effect: increasing the ownership share beyond the minimum level needed for effective control enhances the alignment of interests between the controlling owner and minority shareholders and reduces the effects of entrenchment and tax avoidance The alignment effect argument thus suggests that an ownership concentration beyond the minimum level is negatively associated with tax avoidance 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 3.3.3 Entrenchment effect/alignment effect argument: when control exceeds ownership As previously noted, a major outcome of a divergence between voting and cash-flow rights is that the controlling owner becomes entrenched with a high level of control, but his low equity ownership level provides little alignment with the interests of minority shareholders (Fan and Wong, 2002) Such divergence thus aggravates the entrenchment effect (Claessens et al., 2002; La Porta et al., 1999) and weakens the alignment effect (Fan and Wong, 2002) Francis et al (2005) find that dual-class ownership structures encourage rent extraction because they allow insiders to simultaneously extract private benefits and avoid the related cash-flow effects Masulis et al (2009) also observes that insiders are more likely to pursue private benefits as the separation between voting rights and cash-flow rights widens They further find that as this separation increases, CEOs receive more compensation, cash holdings have less value to outside shareholders, managers become more likely to engage in acquisitions that destroy shareholder value, and capital expenditures contribute less to shareholder value Hence, in firms in which there is a large degree of separation between voting and cash-flow rights, it is possible that controlling owners may engage in higher levels of tax avoidance since insiders view such avoidance as a method of expropriating shareholder wealth and avoiding the related cash-flow consequences The entrenchment effect when control exceeds ownership argument suggests that a high ownership concentration is positively associated with tax avoidance When voting rights and cash-flow rights diverge, a lower level of cash-flow rights may fall short of providing controlling shareholders with sufficient incentive alignment to reduce the entrenchment effect and thus tax avoidance Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 341 342 343 344 345 346 347 348 349 350 351 352 353 3.4 Predicted associations Following from the foregoing discussion, we now formalize our hypotheses for empirical testing H1 and H2 examine the entrenchment and alignment effects when there is concentrated ownership, while H3 and H4 investigate these effects when control exceeds ownership: H1: All else being equal, a concentrated ownership structure obtained through voting rights is positively associated with corporate tax avoidance H2: All else being equal, beyond the minimum level needed for effective control, a concentrated ownership structure obtained through voting rights is negatively associated with corporate tax avoidance H3: All else being equal, a divergence between voting rights and cash-flow rights is positively associated with corporate tax avoidance H4: All else being equal, when control exceeds ownership, cash-flow rights are not associated with corporate tax avoidance 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 Research design 4.1 Sample selection and data source Our sample consists of 207 private firms listed on the Chinese (Shanghai and Shenzhen) A-share stock market over the 2005–2010 period (1242 firm-years) On August 2, 2001, the China Securities Regulatory Commission (CSRC) issued a new standard for information disclosure The Information Disclosure Content and Format Standard Number – Annual Report Content and Format requires publicly-listed firms to disclose detailed information on their ultimate owners (including information on pyramid ownership structures) in their annual reports Notwithstanding the implementation of this standard, many publiclylisted firms failed to disclose the required ownership information To remedy this deficiency, the CSRC issued a revised version of the standard on December 22, 2003, which requires all publicly-listed firms to disclose their ownership structure trees and other relevant ownership information in their annual reports, starting from 2004 It is this revised standard that provides the ownership structure data for our study Our sample selection period begins in 2005, the first year in which nearly all publicly-listed firms disclosed the requisite ownership information, and ends in 2010, the latest year for which ownership information is currently available Our final sample comprises 207 private firms (1242 firm-years) listed on the Chinese (Shanghai and Shenzhen) A-share stock market after excluding firms that fall into the following categories: (a) firms ultimately controlled by the government (i.e SOEs), (b) firms controlled by universities, investment funds or collective enterprises, (c) firms with no ultimate controller, (d) firms with a change in the ultimate controller, (e) firms in severe financial distress/undergoing bankruptcy, (f) firms in the financial industry sector, and (g) firms with missing financial and/or ownership data Table provides a summary of the sample reconciliation procedure All of our financial and tax data were collected from the China Stock Market and Accounting (CSMAR) and Sinofin databases, respectively 377 378 379 380 381 382 4.2 Dependent variable Our dependent variable is denoted by corporate tax avoidance (TA) To improve the robustness of our empirical results, we employ two proxy measures of tax avoidance that have been widely used in the extant literature: effective tax rates (ETRs) and the book–tax gap (BTG) (e.g Gupta and Newberry, 1997; Manzon and Plesko, 2002; Rego, 2003; Desai and 383 384 385 Table Sample reconciliation 386 387 All firms listed on the Chinese (Shanghai and Shenzhen) A-share stock market at the beginning of the 2005 financial year 388 389 390 391 392 393 394 395 396 397 Less: Firms ultimately controlled by the government: SOEs Firms controlled by universities, investment funds or collective enterprises Firms with no ultimate controller Firms with a change in ultimate controller Firms under severe financial distress/bankruptcy Firms in the financial industry sector Firms with missing financial and/or ownership data Final sample (number of firms) Final sample (firm-years) 1353 (622) (75) (186) (136) (56) (10) (61) 207 1242 Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ Dharmapala, 2006; Dyreng et al., 2008; Chen et al., 2010; McGuire et al., 2012) Both are reflections of tax planning that reduces a firm’s taxable income without necessarily reducing its accounting income.10 ETRs are normally calculated as current income tax expenses scaled by pre-tax accounting income (Gupta and Newberry, 1997) They thus measure a firm’s ability to reduce its income tax expenses compared with its pre-tax accounting income and show the relative tax burden across firms (Rego, 2003) Tax avoidance influences ETRs in at least two ways First, it often generates book–tax differences that cause ETR variation because the numerator is based on taxable income, whereas the denominator is based on financial accounting income Tax-motivated transactions (e.g foreign sales, tax-exempt income, tax credits and the deferral of income recognition) generally lower ETRs Second, firms often use their foreign operations to avoid paying income tax, and ETRs capture this form of tax avoidance For instance, shifting income from a high-tax jurisdiction to a low-tax jurisdiction reduces a firm’s ETR.11 Firms that avoid income tax by reducing their taxable income while maintaining their financial accounting income generally have lower ETRs, which suggests that ETRs are a suitable measure of tax avoidance (Rego, 2003) Our first measure of tax avoidance (ETR1) is calculated as current income tax expenses scaled by pre-tax accounting income (Gupta and Newberry, 1997) Our second measure (ETR2) is computed as cash taxes paid scaled by pre-tax accounting income (Dyreng et al., 2008).12 The BTG (or book–tax difference) has been proposed as a measure of tax avoidance (e.g Manzon and Plesko, 2002) since large differences between accounting (or book) income and taxable income are typical among firms that exhibit greater tax avoidance behavior (Lisowsky, 2010; Wilson, 2009).13 Firms can structure transactions to generate large temporary or permanent differences between accounting and taxable income.14 Our third measure of tax avoidance (BTG1) is based on the raw book–tax gap, computed as pre-tax accounting income less taxable income15 (Manzon and Plesko, 2002) scaled by total assets Our fourth measure (BTG2) is calculated as the BTG residual, in keeping with Desai and Dharmapala (2006).16 We adjust the BTG in the same way that Desai and Dharmapala (2006) to control for earnings management activities that may be responsible for the gap In particular, the BTG component that is attributable to earnings management is removed to leave a residual value which is inferred to measure tax avoidance 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 4.3 Independent variables Our independent variables are represented by cash-flow rights (CR), voting rights (VR), voting rights squared (VR2) and voting rights scaled by cash-flow rights (VR_CR) Following La Porta et al (1999), we analyze ultimate ownership: we trace back the owners of immediate shareholders, the owners of those owners and so on If more than one individual in a family has an ownership stake in the firm, we use total ownership by each family group – defined as a group of people related by blood or marriage – as the ultimate ownership We use both ownership and control variables to capture the severity of the potential agency issues for each firm in our sample Our definition of ownership relies on cash-flow rights (CR), that is, on the CR share held by the largest ultimate owner of the firm, CR is calculated as the sum of the products of direct cash-flow rights along the different ownership chains in line with La Porta et al (1999) and Claessens et al (2000, 2002) Thus, the CR measure increases with the largest owner’s cash-flow rights and positively captures the degree of ownership in the firm The larger the CR measure is, the fewer incentives the largest owner has to extract private benefits from the firm Our definition of control relies on voting rights (VR), that is, the share of voting rights held by the largest ultimate owner Q5 of the firm VR is calculated as the sum of the weakest links in the chain of voting rights is consistent with La Porta et al (1999) and Claessens et al (2000, 2002) Hence, this measure increases with the largest owner’s voting rights and positively captures the degree of control in the firm The larger the VR measure is, the greater the largest owner’s incentives to extract private benefits from the firm Our definition of control also relies on voting rights squared (VR2), specifically, the largest ultimate owner’s share of voting rights in the firm squared VR2 is calculated as the sum of the weakest links in the chain of voting rights squared We follow prior research and add a square term of concentrated ownership in our model (e.g McConnell and Servaes, 1990; Ding et al., 2007)17 and estimate the non-linear regression accordingly Thus, this measure decreases with the largest owner’s voting Q8 Q9 10 We note that proxy measures of tax avoidance based on ETRs and BTGs have been effectively applied in prior Chinese studies (e.g Wu and Yue, 2009; Wu et al., 2012a, 2012b; Cai and Liu, 2009; Tang and Firth, 2011a, 2011b; Tang and Firth, 2012) 11 The ETR is reduced because the denominator remains constant (pre-tax accounting income does not change), whereas the numerator becomes smaller (currently payable income taxes decrease) 12 ETRs are truncated to the 0–1 range in this study, in line with prior research (e.g Chen et al., 2010) In addition, as a higher ETR is a proxy measure for a lower level of tax avoidance, we transform ETR1 and ETR2 by multiplying them by −1 to obtain increasing measures of tax avoidance for our empirical analysis (e.g Lanis and Richardson, 2012) 13 For instance, based on confidential tax shelter and tax return data obtained from the IRS, Lisowsky (2010) shows that BTG proxy measures are particularly sound measures of tax avoidance 14 For example, the use of depreciation expenses can generate temporary book–tax differences, whereas the use of R&D tax credits can produce permanent book–tax differences 15 Taxable income is computed as income tax expenses scaled by the statutory tax rate of 33% before 2008 and 25% after 2008 16 The Desai and Dharmapala (2006) method to calculate the BTG residual is explained further in the Appendix 17 McConnell and Servaes (1990) find an inverted U-shaped pattern of association between concentrated ownership and firm value for U.S firms and Ding et al (2007) find the same pattern between concentrated ownership and earnings management for Chinese firms Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 rights and negatively captures the degree of control in the firm The larger the VR2 measure is, the fewer incentives the largest owner has to extract private benefits from the firm When control (VR) exceeds ownership (CR), as in the case of a pyramidal structure, we define the level of separation (VR_CR) as the ultimate owner’s voting rights scaled by his cash-flow rights, in line with prior research (e.g Lemmon and Lins, 2003; Lins, 2003; Harvey et al., 2004; Masulis et al., 2009) Thus, the VR_CR measure increases with the largest owner’s voting rights and decreases with his cash-flow rights, and positively captures the degree of separation between ownership and control The larger the VR_CR measure is, the stronger the ultimate owner’s incentives to extract private benefits from the firm.18 4.4 Control variables To control for other effects, we include the following control variables in our regression model: firm size (SIZE), leverage (LEV), capital intensity (CINT), R&D intensity (RDINT), inventory intensity (INVINT), absolute discretionary accruals (ABS_DA), the market-to-book ratio (MKTBK), return on assets (ROA), regional effects (REGION), industry-sector effects (INDSEC) and year effects (YEAR) We employ SIZE (measured as the natural log of total assets) to control for firm size effects Prior research suggests that larger firms are more complex in nature and are able to achieve economies of scale in tax avoidance to significantly reduce their tax burden (Rego, 2003) LEV is long-term debt scaled by total assets; CINT is net property, plant and equipment scaled by total assets; and RDINT is R&D expenditure scaled by total assets Prior research shows LEV, CINT and RDINT to be positively associated with tax avoidance (Stickney and McGee, 1982) We also include INVINT (inventory scaled by total assets) in our regression model To the extent that INVINT is a substitute for CAPINT, inventory-intensive firms should be less tax avoidant than capitalintensive firms (Stickney and McGee, 1982) ABS_DA (measured as the absolute value of discretionary accruals where discretionary accruals are computed using the modified Jones model as per Dechow et al., 1996) is included as a control variable to ensure that the association between ownership structure and tax avoidance is not driven by earnings quality We expect ABS_DA to have a positive sign (Frank et al., 2009) We also incorporate MKTBK (measured as the market value of equity scaled by the book value of equity) as a growth control variable (Gupta and Newberry, 1997) However, due to the conflicting results obtained in prior research, we make no sign prediction for MKTBK ROA (measured as pre-tax income scaled by total assets) is also included as a profitability control variable (Gupta and Newberry, 1997), but again, owing to the inconsistent results obtained in prior research, we not predict its sign REGION dummy variables are incorporated as control variables to capture the uneven spread of regional development across China The Chinese government has used preferential tax policies as a tool to enhance the economic development of Q6 specific regions to reduce the regional development gap (Shevlin et al., 2012; Tang and Firth, 2011a, 2011b) We generate 30 provincial dummy variables that are coded as if the firm is located in a specific region, otherwise No sign predictions are made for the REGION dummies INDSEC dummy variables are also included as control variables because it is possible that tax avoidance fluctuates across different industry sectors (Dyreng et al., 2008) We include 21 INDSEC dummy variables based on the CSRC industry classifications The INDSEC dummies are coded as if the firm is represented in a particular CSRC category, otherwise No sign predictions are made for the INDSEC dummies Finally, YEAR dummy variables are included to control for possible differences in corporate tax avoidance activities in China over the 2005–2010 financial years The YEAR dummies are coded as if the year falls within a specific year category, otherwise No sign predictions are made for the YEAR dummies 4.5 Regression models The regression model used to examine the association between corporate ownership structure (i.e voting rights) and tax avoidance is represented as follows: TA it = α + β1VR it + β2 VR 2it + β3SIZEit + β 4LEVit + β5CINTit + β6RDINTit + β7INVINTit + β8 ABS_DAit + β9MKTBK it + β10ROA it + β11−40REGIONit + β 41−61INDSECit + β62−67 YEAR i + εit , (1) where: i = firms 1–207, t = financial years 2005–2010, TA = a proxy measure of corporate tax avoidance (ETR1, ETR2, BTG1 and BTG2), VR = the ultimate owner’s voting rights, VR2 = the ultimate owner’s voting rights squared, SIZE = the natural logarithm of total assets, LEV = long-term debt scaled by total assets, CINT = net property, plant and equipment scaled by total assets, 18 We also use a different measure of VR_CR in this research, namely, the largest owner’s voting rights minus his cash-flow rights (the “wedge”), as per Claessens et al (2002), Villalonga and Amit (2006) and Gompers et al (2010) Given that the wedge measure of the separation between control and ownership provides similar results throughout our analysis, herein we report only empirical results based on the ratio measure Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 10 520 521 Table Descriptive statistics 522 Variable N Mean Std Dev Median Minimum Maximum 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 ETR1 ETR2 BTG1 BTG2 VR CR VR_CR SIZE LEV CINT RDINT INVINT ABS_DA MKTBK ROA 1716 1716 1716 1716 1716 1716 1716 1716 1716 1716 1716 1716 1716 1716 1716 0.190 0.163 0.005 0.009 0.331 0.218 2.027 21.286 0.075 0.116 0.001 0.191 −0.015 4.370 0.048 0.166 0.241 0.101 0.401 0.152 0.131 1.528 1.053 0.104 0.153 0.004 0.161 0.119 8.814 0.395 0.158 0.082 0.011 −0.005 0.297 0.195 01.421 21.271 0.033 0.040 0.000 0.148 −0.012 3.334 0.044 0.000 0.000 −0.327 −0.417 0.092 0.014 0.441 17.344 0.000 0.005 0.000 0.001 −0.336 −11.94 −0.336 1.000 1.000 0.153 0.426 0.916 0.644 10.026 24.599 0.802 0.600 0.020 0.758 0.244 23.579 0.292 538 539 540 541 542 543 544 545 Q12 Variable definitions: ETR1 = current income tax expenses scaled by pre-tax accounting income, ETR2 = cash taxes paid scaled by pre-tax accounting income, BTG1 = pre-tax accounting income less taxable income (where taxable income is computed as income tax expenses scaled by the statutory corporate tax rate (i.e 33% before 2008 and 25% after 2008) using the method developed by Manzon and Plesko (2002) scaled by total assets, BTG2 = book–tax gap residual, calculated using the method developed by Desai and Dharmapala (2006), VR = the ultimate owner’s voting rights, CR = the ultimate owner’s cashflow rights, VR_CR = the ultimate owner’s voting rights scaled by his cash-flow rights, SIZE = the natural logarithm of total assets, LEV = long-term debt scaled by total assets, CINT = net property, plant and equipment scaled by total assets, RDINT = R&D expenditure scaled by total assets, INVINT = inventory scaled by total assets, ABS_DA = absolute value of discretionary accruals where discretionary accruals are computed using the modified Jones model as per Dechow et al (1996), MKTBK = the market value of equity scaled by the book value of equity, and ROA = pre-tax income scaled by total assets 546 547 548 549 550 551 552 553 554 555 556 557 558 RDINT = R&D expenditure scaled by total assets, INVINT = inventory scaled by total assets, ABS_DA = the absolute value of discretionary accruals where discretionary accruals are computed using the modified Jones model as per Dechow et al (1996), MKTBK = the market value of equity scaled by the book value of equity, ROA = pre-tax income scaled by total assets, REGION = a regional dummy variable, coded as if the firm is located in a specific region, otherwise 0, INDSEC = an industry sector dummy variable, coded as if the firm is represented in a particular CSRC category, otherwise 0, and YEAR = a year dummy variable, coded as if the year falls within a specific year category, otherwise The regression model used to investigate the association between corporate ownership concentration (i.e a pyramidal structure) and tax avoidance is denoted as follows: TA it = α + β1CR it + β2 VR_CR it + β3SIZEit + β 4LEVit + β5CINTit + β6RDINTit + β7INVINTit + +β8 ABS_DA it + β9MKTBK it + β10ROA it + β11−40REGIONit + β 41−61INDSECit + β62−67 YEAR i + εit , (2) where CR = the ultimate owner’s cash-flow rights, and VR_CR = the ultimate owner’s voting rights scaled by his cash-flow rights 559 560 561 562 563 564 565 566 567 568 Empirical results 5.1 Descriptive statistics Table presents the descriptive statistics for the dependent variables (ETR1, ETR2, BTG1 and BTG2), independent variables (VR, CR and VR_CR) and control variables (SIZE, LEV, CINT, RDINT, INVINT, ABS_DA, MKTBK and ROA) For VR, we find that private Chinese firms have a high level of control, with a mean (median) of 0.331 (0.297) In terms of CR, they have a low degree of ownership, with a mean (median) of 0.218 (0.195), and for VR_CR, they have a high level of separation between voting and cash-flow rights when control exceeds ownership, with a mean (median) of 2.027 (1.421) 569 570 571 572 573 574 575 576 577 578 5.2 Correlation results The Pearson correlation results are reported in Table We find significantly positive correlations between ETR1, ETR2, BTG1 and BTG2 and VR (p < 0.05 or better), showing that increased levels of voting rights increase the likelihood of tax avoidance Table also reports significant negative correlations between ETR1, ETR2, BTG1 and BTG2 and CR (p < 0.05 or better), indicating that increased levels of cash-flow rights reduce the likelihood of tax avoidance We also find significantly positive correlations between ETR1, ETR2, BTG1 and BTG2 and VR_CR (p < 0.05 or better), showing that a greater separation between voting and cash-flow rights increases the likelihood of tax avoidance We also find significant correlations between ETR1, ETR2, BTG1 and BTG2 and some of the control variables such as SIZE, LEV, CINT, RDINT, INVINT, ABS_DA and ROA (p < 0.10 Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 579 580 581 582 583 Q14 ETR2 BTG1 BTG2 VR CR VR_CR SIZE LEV CINT RDINT INVINT ABS_DA MKTBK ROA 0.031** −0.046*** 0.046*** 0.040*** 0.011 0.021* 0.061*** −0.026** 0.135*** 0.010 0.005 0.029** −0.030** 0.034** 0.052*** 0.033*** 0.012 0.024** −0.142*** 0.105*** 0.014 0.005 0.069*** −0.064*** 0.027** 0.046*** 0.009 0.068*** 0.034*** −0.028** 0.498*** 0.004 0.186*** 0.071*** −0.066*** 0.038*** 0.061*** 0.008 0.028** 0.006 −0.052*** 0.561*** 0.007 0.012 – 0.672*** 0.064*** 0.277*** 0.024* 0.016 −0.043*** −0.018 0.070*** 0.025* −0.014 – −0.555*** 0.105*** −0.013 −0.053* −0.070** 213*** 0.089*** 0.022 −0.006 – 0.107*** 0.021 0.015 0.018 0.008 −0.027* 0.005 0.016 – 0.036** 0.073*** −0.169*** −0.015 0.214*** 0.006 0.035** – 0.001 −0.039*** −0.022 0.000 −0.016 0.135*** – −0.035*** −0.015 −0.045*** −0.072*** 0.064*** – 0.065*** −0.010 0.063*** 0.110*** – 0.201*** 0.113*** 0.169*** – −0.024 0.448*** – 0.001 – See Table for variable definitions N = 1716 for all variables ARTICLE IN PRESS VR CR VR_CR SIZE LEV CINT RDINT INVINT ABS_DA MKTBK ROA ETR1 G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ Q13 11 Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 Table Pearson correlation results ARTICLE IN PRESS 12 595 596 Table Regression results: concentrated ownership structure 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ Predicted sign Intercept +/− VR + VR2 − SIZE + LEV + CINT + RDINT + INVINT − ABS_DA + MKTBK +/− ROA +/− REGION INDSEC YEAR Adj R2 (%) N +/− +/− +/− ETR1 ETR2 BTG1 BTG2 −0.175 (−1.65)* 0.140 (1.37)* −0.241 (−1.95)** 0.008 (1.78)** 0.019 (0.34) 0.053 (0.88) 0.194 (0.39) −0.334 (−1.21) 0.355 (1.43)* −0.534 (−1.77)** 0.031 (3.00)*** 0.143 (1.41)* 0.010 (0.09) 0.279 (0.84) −0.261 (−2.64)*** 0.083 (1.52)* −0.066 (−1.10) 0.013 (3.02)*** 0.037 (0.91) 0.020 (0.82) 0.240 (0.85) −0.293 (−2.74)*** 0.097 (1.70)** −0.096 (−1.56)* 0.014 (2.79)*** 0.020 (0.49) 0.010 (0.39) 0.136 (0.45) −0.099 (−2.63)*** 0.150 (3.22)*** 0.001 (1.88)** 0.012 (1.06) Yes Yes Yes 8.51% 1242 −0.102 (−1.24) 0.258 (2.57)*** 0.001 (0.05) 0.016 (0.44) Yes Yes Yes 5.36% 1242 −0.099 (−3.87)*** 0.398 (3.97)*** 0.001 (1.39) 0.015 (0.37) Yes Yes Yes 27.52% 1242 −0.093 (−3.44)*** 0.513 (4.74)*** 0.001 (1.40) 0.006 (0.16) Yes Yes Yes 35.63% 1242 Variable definitions: REGION = a region dummy variable, coded as if the firm is located in the specific region, otherwise 0, INDSEC = an industry sector dummy variable, coded as if the firm is represented in the particular CSRC category, otherwise 0, and YEAR = a year dummy variable, coded as if the year falls within the specific year category, otherwise See Table for the other variable definitions * , ** and *** indicate significance at the 10, 05 and 01 levels, respectively The p-values are one-tailed for the directional hypotheses and two-tailed otherwise Note 1: Coefficient estimates with t-statistics are in parentheses 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 or better with predicted signs, where appropriate) Finally, Table reports that only moderate levels of collinearity exist between the explanatory variables 5.3 Regression results 5.3.1 Concentrated ownership structure for publicly-listed private firms – H1 and H2 Table presents the results of ordinary least squares (OLS) regression analysis in terms of concentrated ownership structures for publicly-listed private Chinese firms.19 The regression coefficient for VR is significantly positively associated with ETR1, ETR2, BTG1 and BTG2 (p < 0.10 or better), thus providing support for H1 (the entrenchment effect) This finding is consistent with Desai and Dharmapala’s (2006) claim that tax avoidance can facilitate insiders’ ability to expropriate shareholder wealth because it is complex and designed to obscure the underlying economics of a business transaction The regression coefficient for VR2 is significantly negatively associated with ETR1, ETR2 and BTG2 (p < 0.10 or better), supporting H2 (the alignment effect) This result is consistent with research by Chen et al (2010), showing that family-controlled firms are less tax avoidant as family owners are willing to relinquish tax benefits to avoid the non-tax cost of a potential price discount to the firm’s share price that can arise from minority shareholder concern about managerial rent extraction We also find that some of our control variable regression coefficients are significantly associated with ETR1, ETR2, BTG1 and/or BTG2 (p < 0.10 or better) with predicted signs (where applicable), namely, SIZE, LEV, INVINT, ABS_DA and MKTBK Overall, our regression results show that there is a significant non-linear association between ownership concentration and tax avoidance Specifically, this association displays an inverted U-shaped pattern, which means that, at a lower level, an increased ownership concentration is positively associated with tax avoidance (the entrenchment effect) However, beyond the minimum level necessary for effective control, a concentrated ownership structure obtained through voting rights is negatively associated with tax avoidance (the alignment effect) 19 Coefficient estimates with t-statistics are reported in parentheses Standard errors are corrected based on the Huber/White/Sandwich estimator of standard errors for all our regression models (see e.g Wooldridge, 2010) We also compute variance inflation factors (VIFs) when estimating our regression models to test for signs of multicollinearity among the explanatory variables As none of the VIFs exceeds ten, we conclude that multicollinearity is not a problem in our study (Hair et al., 2006) Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 660 661 Table Regression results: pyramidal ownership structure 662 Predicted sign 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 13 Q15 Intercept +/− CR − VR_CR + SIZE + LEV + CINT + RDINT + INVINT − ABS_DA + MKTBK +/− ROA +/− REGION INDSEC YEAR Adj R2 (%) N +/− +/− +/− ETR1 ETR2 BTG1 BTG2 −0.105 (−1.01) −0.054 (−1.08) 0.008 (2.41)*** 0.007 (1.48)* 0.010 (0.18) 0.060 (0.99) 0.059 (0.11) −0.093 (−2.44)*** 0.150 (3.24)*** 0.001 (1.91)* −0.215 (−0.80) −0.166 (−1.12) 0.009 (1.79)** 0.030 (2.87)*** 0.137 (1.24) 0.018 (0.17) 0.281 (0.81) −0.093 (−1.15) 0.266 (2.62)*** 0.001 (0.01) −0.223 (−2.39)** −0.012 (−0.76) 0.001 (0.41) 0.013 (2.96)*** 0.027 (0.69) 0.016 (0.43) 0.316 (1.03) −0.075 (−2.77)*** 0.389 (3.97)*** 0.001 (1.44) −0.205 (−2.92)*** −0.072 (−0.25) 0.055 (2.04)** 0.076 (2.28)** 0.192 (0.63) 0.255 (0.64) 0.545 (1.11) −0.300 (−0.95) 0.236 (0.91) 0.004 (1.31) 0.012 (1.05) Yes Yes Yes 8.49% 1242 0.018 (0.48) Yes Yes Yes 5.12% 1242 0.096 (2.15)** Yes Yes Yes 27.35% 1242 0.111 (1.04) Yes Yes Yes 33.67% 1242 See Tables and for variable definitions * , ** and *** indicate significance at the 10, 05 and 01 levels, respectively The p-values are one-tailed for directional hypotheses and two-tailed otherwise Note 1: Coefficient estimates with t-statistics are in parentheses 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 5.3.2 Pyramidal ownership structure for publicly-listed private firms – H3 and H4 Table reports the results of OLS regression analysis of the pyramidal ownership structure of publicly-listed private Chinese firms The regression coefficient for VR_CR is significantly positively associated with ETR1, ETR2 and BTG2 (p < 0.05 or better), furnishing support for H3 (the entrenchment effect) This result is consistent with research by Francis et al (2005) and Masulis et al (2009) who find that insiders are more likely to pursue private benefits as the separation between voting rights and cash-flow rights widens because such separation allows them to extract such benefits while avoiding the related cash-flow consequences The regression coefficient for CR is not significant, supporting H4 (no alignment effect) and is consistent with the claim made in prior research that the divergence between voting and cash-flow rights exacerbates the entrenchment effect (e.g La Porta et al., 1999; Claessens et al., 2002) while producing minimal or no alignment between the controlling owner and minority shareholders (e.g Fan and Wong, 2002) We also find that several of the regression coefficients for our control variables (e.g SIZE, INVINT, ABS_DA, MKTBK and ROA) are significantly associated with ETR1, ETR2, BTG1 and/or BTG2 (p < 0.10 or better) with predicted signs (where applicable) On the whole, our regression results show that there is a significantly positive association between ownership concentration in the form of pyramidal ownership structure and tax avoidance (the entrenchment effect) When voting rights and cash-flow rights diverge, a lower level of cash-flow rights fails to offer an incentive alignment to the controlling owner (the alignment effect) which is sufficient to reduce the entrenchment effect and thus tax avoidance Additional analysis – endogeneity of ownership structure 6.1 Instrumental variables 2SLS regression analysis Because it is possible that ownership structure is an endogenous variable (e.g Demsetz and Lehn, 1985; Hermalin and Weisbach, 1988; Loderer and Martin, 1997; Cho, 1998; Demsetz and Villalonga, 2001), we carry out additional analysis that takes the endogeneity of that structure into account Specifically, we initially perform instrumental variables 2SLS regression analysis, whereby the predicted ownership structure variable produced by the first-stage regression is used in place of the original variable in the second-stage regression Consistent with prior research (e.g Demsetz and Lehn, 1985; Demsetz and Villalonga, 2001; Gompers et al., 2010), we select four instrumental variables for ownership structure: (1) the market value of equity (MVEQUITY), (2) Tobin’s Q (TOBINSQ), (3) the listing date of the firm (LISTDATE), and (4) regional market share (REGMKTSHR) The aforementioned research has Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 14 724 725 726 727 728 729 found all of these variables to be indicative of ownership structure, but there is no evidence of reverse causation, which makes them suitable instruments for use in our empirical analysis (Larcker and Rusticus, 2010) MVEQUITY is measured as the market value of equity, TOBINSQ as the book value of debt plus the market value of equity scaled by total assets, LISTDATE as the number of years the firm has been listed on the Chinese stock exchange, and REGMKTSHR as the ratio of the focal firm’s sales to the total sales of all listed firms operating in the same region The 2SLS regression analysis results are presented in Table 730 731 732 733 Table Instrumental variables (2SLS) regression results: endogeneity of ownership structure First-stage Regressions Concentrated Ownership Structure 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 Predicted sign Intercept +/− MVEQUITY + TOBINSQ + LISTDATE + REGMKTSHR + Control Variables REGION INDSEC YEAR Adj R2 (%) N +/− +/− +/− +/− 751 752 753 Second-stage Regressions 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 Intercept +/− VR + 783 784 785 786 787 788 789 790 VR VR2 CR VR_CR −0.247 (−1.90)* 0.026 (3.39)*** 0.003 (2.19)** 0.006 (4.32)*** 0.001 (0.94) Yes Yes Yes Yes 14.11% 1230 −0.432 (−3.79)*** 0.022 (3.57)*** 0.002 (1.49)* 0.004 (3.05)*** 0.001 (0.24) Yes Yes Yes Yes 14.98% 1230 0.125 (1.27) 0.024 (3.74)*** 0.005 (3.35) 0.007 (7.16)*** 0.001 (0.74) Yes Yes Yes Yes 15.37% 1230 1.074 (1.16) 0.009 (0.11) 0.025 (0.97) 0.049 (4.23)*** 0.001 (0.63) Yes Yes Yes Yes 13.00% 1230 Concentrated Ownership Structure Predicted sign − CR − VR_CR + SIZE + LEV + CINT + RDINT + INVINT − ABS_DA + MKTBK +/− ROA +/− REGION INDSEC YEAR +/− +/− +/− Pyramidal Ownership Structure ETR1 ETR2 BTG1 BTG2 −1.257 (−2.58)*** 3.359 (1.87)** −5.585 (−2.27)** −3.250 (−2.68)*** 7.338 (1.62)* −13.177 (−2.14)** 0.433 (0.84) 1.489 (1.00) −2.859 (−1.31)* 0.409 (1.00) 1.618 (1.43)* −2.885 (−1.73)** 0.045 (3.34)*** 0.070 (0.82) 0.047 (0.48) 2.476 (1.31)* −0.324 (−2.91)*** 0.127 (1.61)* 0.001 (1.92)** 0.001 (0.01) Yes Yes Yes Pyramidal Ownership Structure VR 0.131 (3.80)*** 0.297 (1.46)* 0.271 (1.23) 10.361 (1.32)* −0.710 (−2.46)*** 0.152 (0.74) 0.002 (1.54)* 0.009 (0.23) Yes Yes Yes 0.011 (0.69) 0.004 (0.07) 0.056 (0.94) 1.724 (1.09) −0.238 (−2.95)*** 0.380 (3.68)*** 0.001 (0.64) 0.010 (0.24) Yes Yes Yes 0.008 (0.78) 0.014 (0.39) 0.064 (1.16) 1.604 (1.36)* −0.221 (−3.52)*** 0.500 (4.71)*** 0.001 (1.56)* 0.002 (0.06) Yes Yes Yes ETR1 BTG1 BTG2 0.720 (4.15)*** ETR2 1.578 (2.43)** −0.731 (−3.28)*** −0.716 (−2.98)*** −1.793 (−0.81) 0.282 (2.78)*** 0.013 (1.70)** 0.213 (0.93) 0.128 (0.94) 12.601 (1.51)* −0.162 (−2.21)** 0.178 (2.09)** 0.003 (2.57)*** 0.015 (0.70) Yes Yes Yes −4.097 (−0.37) 0.560 (2.64)*** 0.046 (4.75)*** 0.287 (0.90) 0.138 (0.53) 27.122 (1.58)* −0.273 (−1.34)* 0.284 (1.81)** 0.004 (2.10)** 0.027 (0.78) Yes Yes Yes −1.227 (−0.16) 0.166 (1.67)** 0.008 (1.93)** 0.161 (1.50)* 0.050 (0.50) 7.088 (1.19) −0.146 (−2.93)*** 0.344 (3.32)*** 0.001 (0.60) 0.001 (0.03) Yes Yes Yes −1.105 (−0.84) 0.160 (1.53)* 0.009 (2.11)** 0.146 (1.35)* 0.036 (0.38) 7.102 (1.15) −0.133 (−2.73)*** 0.468 (4.14)*** 0.001 (0.46) 0.006 (0.14) Yes Yes Yes Variable definitions: MVEQUITY = the market value of equity, TOBINSQ = the book value of debt plus the market value of equity scaled by total assets, LISTDATE = the number of years the firm has been listed on the Chinese stock exchange, REGMKTSHR = the ratio of the focal firm’s sales to the total sales of all listed firms operating in the same region, and see Tables and for other variable definitions Instrumented: VR, CR and VR_CR Instruments: MVEQUITY, TOBINSQ, LISTDATE and REGMKTSHR * , ** and *** indicate significance at the 10, 05 and 01 levels, respectively The p-values are one-tailed for directional hypotheses and two-tailed otherwise Note 1: Coefficient estimates with t-statistics are in parentheses Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 15 In terms of the first-stage regressions for the concentrated ownership structure of publicly-listed private Chinese firms, we find that the regression coefficients for MVEQUITY, TOBINSQ and LISTDATE are significantly positively associated with VR and VR2 (p < 0.10 or better), which is consistent with prior research (e.g Demsetz and Lehn, 1985; Demsetz and Villalonga, 2001; Gompers et al., 2010) In addition, for the second-stage regressions we find the regression coefficient for VR to be significantly positively associated with ETR1, ETR2 and BTG2 (p < 0.10 or better), thus providing additional support for H1 (the entrenchment effect) The regression coefficient for VR2 is significantly negatively associated with ETR1, ETR2, BTG1 and BTG2 (p < 0.10 or better), thus again supporting H2 (the alignment effect) We also observe that some of the regression coefficients for control variables, such as SIZE, LEV, RDINT, INVINT, ABS_DA and MKTBK are also significantly associated with ETR1, ETR2, BTG1 and/or BTG2 (p < 0.10 or better with predicted signs, where applicable) Concerning the pyramidal ownership structure of these firms, for the first-stage regressions, we find that the regression coefficients for MVEQUITY and LISTDATE are significantly positively associated with VR and/or VR2 (p < 0.01), which is in line with prior research (e.g Demsetz and Lehn, 1985; Demsetz and Villalonga, 2001; Gompers et al., 2010) Moreover, for the second-stage regressions we find the regression coefficient for VR_CR to be significantly positively associated with ETR1, ETR2, BTG1 and BTG2 (p < 0.10 or better), which provides further support for H3 (the entrenchment effect) The regression coefficient for CR is not significant, and thus H4 is also supported (no alignment effect) Finally, the regression coefficients for the SIZE, LEV, RDINT, ABS_DA, INVINT and MKTBK control variables are also significantly associated with ETR1, ETR2, BTG1 and/or BTG2 (p < 0.10 or better with predicted signs, where applicable) 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 6.2 DID analysis As a final test of the potential issue of the endogeneity of ownership structure unduly influencing our empirical results, we perform a DID analysis (see e.g Gippel et al., 2015) Specifically, an exogenous event which affected all firms in China was the introduction of the general anti-avoidance rule (GAAR) by the National People’s Congress (NPC) of China as part of the Enterprise Income Tax Law (EITL) on January 1, 2008 to curb aggressive forms of corporate tax avoidance.20 The GAAR fundamentally altered the corporate tax landscape in China through the introduction of the new rules including “catch all” general anti-avoidance provisions that have the effect of examining all business transactions and arrangements of a firm based on a “reasonable business purpose” test (Li and Huang, 2008) For instance, Article 47 of the EITL allows the State Administration of Taxation (SAT) to make tax adjustments relating to transactions or arrangements in which the main business purpose of a firm is to avoid, reduce, defer or make exempt the payment of corporate income taxes Overall, the GAAR was designed to target abusive, aggressive or illegal tax structures of firms that were not able to be appropriately dealt with prior to its introduction, and thus is expected to increase a firm’s marginal costs of carrying out tax avoidance.21 To capture the introduction of the GAAR in China on January 1, 2008 in our study, we construct a dummy variable, coded as if the observations correspond to the post-GAAR period (i.e 2008–2010), and otherwise (i.e 2005–2007) consistent with prior research (e.g Chan et al., 2013) We also construct several interaction terms in accordance with a DID research design which is adapted for our exogenous event in China (see e.g Liu et al., 2016) In particular, VR*GAAR and VR2*GAAR are the interaction terms for concentrated ownership structure, whereas CR*GAAR and VR_CR*GAAR are the interaction terms for pyramidal ownership structure The DID analysis results are reported in Table For concentrated ownership structure, we find the regression coefficient for VR to be significantly positively associated with ETR1, BTG1 and BTG2 (p < 0.10), thus providing further support for H1 (the entrenchment effect) The regression coefficient for VR2 is significantly negatively associated with ETR1, ETR2 and BTG2 (p < 0.10 or better), thus yet again supporting H2 (the alignment effect) For the interaction regression coefficients, we find VR*GAAR is significantly negatively associated with ETR1 and ETR2 (p < 0.10), while VR2*GAAR is not significant in any of the regression model specifications These results suggest that the GAAR tax reform in China did not have a major influence on the general association between concentrated ownership structure and tax avoidance Finally, we also find that several of the regression coefficients for our control variables (e.g SIZE, LEV, CINT, INVINT, ABS_DA and MKTBK) are significantly associated with ETR1, ETR2, BTG1 and/or BTG2 (p < 0.10 or better) with predicted signs (where applicable) In terms of pyramidal ownership structure, Table shows that the regression coefficient for VR_CR is significantly positively associated with ETR1, ETR2 and BTG2 (p < 0.10 or better), providing additional support for H3 (the entrenchment effect) The regression coefficient for CR is once again not significant, supporting H4 (no alignment effect) For the interaction regression coefficients, we find that CR*GAAR is not significant, while VR_CR*GAAR is only significantly positively associated with ETR1 (p < 0.10) These findings suggest that the GAAR tax reform in China did not have a major effect on the general association between pyramidal ownership structure and tax avoidance Finally, we observe that some of the coefficients for our control variables (e.g SIZE, LEV, CINT, INVINT, ABS_DA and MKTBK) are significantly associated with ETR1, ETR2, BTG1 and/or BTG2 (p < 0.10 or better) with predicted signs (where applicable) 845 846 847 848 849 850 20 The NPC of China promulgated the new EITL on March 16, 2007 to take effect on January 1, 2008 The marginal costs associated with tax avoidance include the potential tax penalty imposed by the tax authorities, implementation costs (e.g time/ effort and transaction costs of executing tax transactions), and the agency costs (e.g managerial rent extraction) of tax avoidance activities (Chen et al., 2010; Scholes et al., 2009) 21 Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 16 851 852 853 Table DID regression results: endogeneity of ownership structure Second-stage Regressions 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 Predicted sign Intercept +/− VR + VR2 − VR*GAAR +/− VR2*GAAR +/− CR − VR_CR + CR*GAAR +/− VR_CR*GAAR +/− SIZE + LEV + CINT + RDINT + INVINT − ABS_DA + MKTBK +/− ROA +/− REGION INDSEC Adj R2 (%) N +/− +/− Concentrated Ownership Structure Pyramidal Ownership Structure ETR1 ETR2 BTG1 BTG2 ETR1 ETR2 BTG1 BTG2 −0.146 (−1.42) 0.163 (1.57)* −0.266 (−2.08)** −0.150 (−1.93)* 0.186 (1.38) −0.242 (−0.90) 0.286 (1.14) −0.422 (−1.37)* −0.378 (−1.89)* 0.552 (1.49) −0.253 (−3.86)*** 0.082 (1.43)* −0.064 (−0.99) −0.003 (−0.06) 0.001 (0.01) −0.286 (−2.75)*** 0.097 (1.64)* −0.097 (−1.45)* −0.005 (−0.10) 0.014 (0.17) −0.081 (−0.79) −0.129 (−0.49) −0.246 (−2.54)** −0.271 (−2.62)*** −0.040 (−0.76) 0.008 (2.16)*** −0.002 (−0.54) 0.078 (1.71)* 0.006 (1.29)* 0.023 (0.43) 0.108 (3.06)*** 0.083 (0.15) −0.105 (−2.78)*** 0.147 (3.15)*** 0.001 (1.75)* 0.012 (1.03) Yes Yes 8.28% 1242 −0.177 (−0.90) 0.011 (1.29)* −0.098 (−0.86) 0.012 (1.20) 0.026 (2.55)*** 0.170 (1.58)* 0.161 (2.25)** 2.429 (0.71) −0.128 (−1.61)* 0.255 (2.48)*** 0.001 (0.21) 0.017 (0.47) Yes Yes 5.11% 1242 −0.018 (−1.02) 0.001 (0.11) −0.001 (−0.01) 0.011 (0.18) 0.013 (3.17)*** 0.040 (1.00) 0.013 (0.63) 0.324 (1.00) −0.096 (−3.94)*** 0.396 (3.97)*** 0.001 (1.43) 0.015 (0.37) Yes Yes 27.48% 1242 −0.002 (−0.10) 0.048 (1.54)* −0.001 (−0.05) 0.001 (0.19) 0.014 (2.93)*** 0.022 (0.53) 0.006 (0.28) 0.210 (0.62) −0.091 (−3.55)*** 0.511 (4.75)*** 0.001 (1.45) 0.006 (0.15) Yes Yes 33.82% 1242 0.007 (1.55)* 0.032 (0.60) 0.102 (2.90)*** 0.347 (0.69) −0.113 (−2.97)*** 0.146 (3.11)*** 0.001 (1.67)* 0.012 (1.06) Yes Yes 8.32% 1242 0.027 (2.67)*** 0.180 (1.69)** 0.163 (2.28)** 2.420 (0.74) −0.137 (−1.71)** 0.247 (2.44)*** 0.001 (0.24) 0.016 (0.44) Yes Yes 5.24% 1242 0.013 (3.06)*** 0.039 (0.98) 0.011 (0.55) 0.204 (0.74) −0.097 (−4.01)*** 0.396 (3.97)*** 0.001 (1.43) 0.015 (0.38) Yes Yes 27.66% 1242 0.014 (2.86)*** 0.022 (0.53) 0.005 (0.23) 0.104 (0.35) −0.091 (−3.58)*** 0.511 (4.75)*** 0.001 (1.44) 0.007 (0.16) Yes Yes 33.95% 1242 Variable definitions: VR*GAAR = an interaction term comprising VR multiplied by the GAAR dummy variable, VR2*GAAR = an interaction term comprising VR2 multiplied by the GAAR dummy variable, CR*GAAR = an interaction term comprising CR multiplied by the GAAR dummy variable, VR_CR*GAAR = an interaction term comprising VR_CR multiplied by the GAAR dummy variable, and see Tables and for other variable definitions * , ** and *** indicate significance at the 10, 05 and 01 levels, respectively The p-values are one-tailed for directional hypotheses and two-tailed otherwise Note 1: Coefficient estimates with t-statistics are in parentheses 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 Overall, our empirical results for H1, H2, H3 and H4 based on the 2SLS regression and DID analysis (see Tables and above) are consistent with those reported in Tables and 5, which confirms that the endogeneity of ownership structure does not appear to drive our empirical findings Conclusion This study examines the association between ownership structure and corporate tax avoidance among publicly-listed private firms in China We find a significant non-linear association between ownership concentration and tax avoidance that displays an inverted U-shaped pattern At a lower level, an increased ownership concentration is positively associated with tax avoidance due to the entrenchment effect However, beyond the minimum level required for effective control, voting rights-induced concentrated ownership is negatively associated with tax avoidance because of the alignment effect We also find a significantly positive association between pyramidal ownership structure and tax avoidance due to the entrenchment effect When voting rights and cash-flow rights diverge, a lower level of cash-flow rights fails to offer the controlling owner incentive alignment which is sufficient to reduce the entrenchment effect and thus tax avoidance Our study provides unique insights into the association between ownership structure and tax avoidance in the context of both concentrated and pyramidal ownership In so doing, it extends the extant literature on ownership structure and tax avoidance Our results are generalizable to parts of the world other than China, given that concentrated ownership structure is a more representative corporate stock structure worldwide than is diffuse stock ownership Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 17 This study is subject to several limitations First, our sample is drawn from publicly-listed Chinese firms, as data unavailability rendered it impossible to include unlisted firms in our sample Second, we constructed our tax avoidance measures based on financial statement data as tax return data of Chinese firms are not in the public domain and are thus unavailable The extant literature questions the accuracy of financial statement-based tax avoidance measures (e.g Plesko, 2003; Hanlon and Heitzman, 2010), so our results should be interpreted with some caution Future research could be carried out to address the way in which ownership structure affects tax policies in other emerging markets and transitional economies Acknowledgments This paper was presented at the Journal of Contemporary Accounting and Economics (JCAE) Symposium in Kuala Lumpur, Malaysia on 9–10 January 2015 It won the Best Paper Award at the Symposium We are indebted to the discussant, Professor Simon Fung, and the anonymous reviewer for the helpful comments and suggestions that have significantly improved the paper We would also like to thank the co-editors, Professors Ferdinand Gul and Zoltan Matolcsy, for their insightful comments and suggestions All errors and omissions are our own Q16 Uncited references 945 946 947 948 949 Richardson, Lanis, 2007 Appendix: Description of Desai and Dharmapala’s (2006) method for calculating the BTG residual Applying Desai and Dharmapala’s (2006) methodology, taxable income is calculated as TIjt = accounting income tax expense scaled by the Chinese statutory tax rate of 33% before 2008 and 25% after 2008 The BTG is calculated by subtracting TI from pre-tax accounting income (AI): BTGjt = AIjt − TIjt The BTG is scaled by beginning-of-period total assets Total accruals (TA) are calculated for each firm in each year using the measure of total accruals developed by Healy (1985) Total accruals are considered to measure the earnings management component of the BTG TA it = EBEIit − (CFOit − EIDOit ) (A1) where: i = firms 1–207, t = financial years 2006–2010, TA = total accruals, EBEI = income before extraordinary items, CFO = cashflow from operations, and EIDO = extraordinary items and discontinued operations from the cash-flow statement The following OLS regression is performed to account for the BTG component attributable to earnings management BTGit = β1TA it + μit + ε jt 950 951 952 953 954 955 956 where: BTG = the book–tax gap scaled by beginning-of-year assets, TA = total accruals scaled by beginning-of-year assets, μ = the residual, and ε = the error term The residual value of the BTG is considered to reflect tax avoidance activity (TA): TAit = μit + εit 957 958 959 960 961 962 963 964 965 966 967 968 969 Q10 Q11 970 971 972 973 974 975 976 977 978 979 980 981 982 983 Almeida, H., Wolfenzon, D., 2006 Should business groups be dismantled? The equilibrium costs of efficient internal capital markets J Financ Econ 79, 99–144 Bertrand, M., Johnson, S., Samphantharak, K., Schoar, A., 2008 Mixing family with business: a study of Thai business groups and the families behind them J Financ Econ 88, 466–498 Cai, H., Liu, Q., 2009 Competition and corporate tax avoidance: evidence from Chinese industrial firms Econ J (London) 119, 764–795 Chan, K.H., Mo, P.L.L., 2000 Tax holidays and tax noncompliance: an empirical study of corporate tax audits in China’s developing economy Account Rev 75, 469–484 Chan, K.H., Mo, P.L.L., Zhou, A.Y., 2013 Government ownership, corporate governance and tax aggressiveness: evidence from China Account Finance 53 (4), 1029–1051 Chen, C.J.P., Li, Z., Su, X., Sun, Z., 2011 Rent-seeking incentives, corporate political connections, and the control structure of private firms: Chinese evidence J Corp Finance 17, 229–243 Chen, S., Chen, X., Cheng, Q., Shevlin, T., 2010 Are family firms more tax aggressive than non-family firms? J Financ Econ 95, 41–61 Chinese Association of Public Companies (CAPCO), 2012 The development of Chinese public companies Cho, M.H., 1998 Ownership structure, investment and the corporate value: an empirical analysis J Financ Econ 47, 103–121 Claessens, S., Djankov, S., Lang, L.H.P., 2000 The separation of ownership and control in East Asian corporations J Financ Econ 58, 81–112 Claessens, S., Djankov, S., Fan, J.P.H., Lang, L.H.P., 2002 Disentangling the incentive and entrenchment effects of large shareholdings J Finance 62, 2741–2771 Dechow, P.M., Sloan, R.G., Sweeney, A., 1996 Causes and consequences of earnings manipulation: an analysis of firms subject to enforcement actions by the SEC Contemp Account Res 13 (1), 1–36 Demsetz, H., Lehn, K., 1985 The structure of corporate ownership: causes and consequences J Polit Econ 93, 1155–1177 Demsetz, H., Villalonga, B., 2001 Ownership structure and corporate performance J Corp Finance 7, 209–233 Desai, M.A., Dharmapala, D., 2006 Corporate tax avoidance and high-powered incentives J Financ Econ 79, 145–179 Ding, Y., Zhang, H., Zhang, J., 2007 Private vs state ownership and earnings management: evidence from Chinese listed companies Corp Gov Int Rev 15, 223–238 Dyreng, S.D., Hanlon, M., Maydew, E.L., 2008 Long-run corporate tax avoidance Account Rev 83, 61–82 Faccio, M., Lang, L.H.P., 2002 The ultimate ownership of western European corporations J Financ Econ 65, 365–395 Fan, J.P.H., Wong, T.J., 2002 Corporate ownership structure and the informativeness of accounting earnings in East Asia J Account Econ 33, 401–425 Fan, J.P.H., Wong, T.J., Zhang, T., 2013 Institutions and organizational structure: the case of state-owned corporate pyramids J Law Econ Org 29 (6), 1217–1252 (A2) References Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ARTICLE IN PRESS 18 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 G Richardson et al / Journal of Contemporary Accounting & Economics ■■ (2016) ■■–■■ Fisman, R., Khanna, T., 2004 Facilitating development: the role of business groups World Dev 32, 609–628 Francis, J., Schipper, K., Vincent, L., 2005 Earnings and dividend informativeness when cash flow rights are separated from voting rights J Account Econ 39, 329–360 Frank, M., Lynch, L., Rego, S., 2009 Tax reporting aggressiveness and its relation to aggressive financial reporting Account Rev 84, 467–496 Gippel, J., Smith, T., Zhu, Y., 2015 Endogeneity in accounting and finance research: natural experiments as a state-of-the-art solution Abacus 51 (2), 143–167 Gomes, A., 2000 Going public without governance: managerial reputation effects J Finance 55, 615–646 Gompers, P.A., Ishii, J., Metrick, A., 2010 Extreme governance: an analysis of dual-class firms in the United States Rev Financ Stud 23, 1051–1088 Guedhami, O., Pittman, P.A., Saffar, W., 2014 Auditor choice in politically connected firms J Account Res 52, 107–162 Gupta, S., Newberry, K., 1997 Determinants of the variability in corporate effective tax rates: evidence from longitudinal data J Account Public Policy 16, 1–34 Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E., Tatham, R.L., 2006 Multivariate Data Analysis, sixth ed Pearson Prentice-Hall, Upper Saddle River, NJ Hanlon, M., Heitzman, S., 2010 A review of tax research J Account Econ 50, 127–178 Harvey, C.R., Lins, K.V., Roper, A.H., 2004 The effect of capital structure when expected agency costs are extreme J Financ Econ 74, 3–30 Healy, P., 1985 The effect of bonus schemes on accounting decisions J Account Econ 7, 85–107 Hermalin, B., Weisbach, M., 1988 The determinants of board composition RAND J Econ 19, 589–606 Jiang, G., Lee, C.M.C., Yue, H., 2010 Tunneling through intercorporate loans: the China experience J Financ Econ 98, 1–20 Johnson, S., La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2000 Tunneling Am Econ Rev 90, 22–27 Khanna, T., Palepu, K., 2000 Is group affiliation profitable in emerging markets? An analysis of diversified Indian business groups J Finance 55, 867–891 Khanna, T., Palepu, K.G., 1997 Why focused strategies may be wrong for emerging markets Harv Bus Rev 75, 41–51 Khanna, T., Yafeh, Y., 2005 Business groups and risk sharing around the world J Bus 78, 301–340 Klassen, K., 1997 The impact of inside ownership concentration on the tradeoff between financial and tax reporting Account Rev 72 (3), 455–474 La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1997 Legal determinants of external finance J Finance 52, 1131–1150 La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998 Law and finance J Polit Econ 106, 1113–1155 La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999 Corporate ownership around the world J Finance 54, 471–518 Lanis, R., Richardson, G., 2012 Corporate social responsibility and tax aggressiveness: an empirical analysis J Account Public Policy 31 (1), 86–108 Larcker, D.F., Rusticus, T.O., 2010 On the use of instrumental variables in accounting research J Account Econ 49, 186–205 Lemmon, M.L., Lins, K., 2003 Ownership structure, corporate governance, and firm value: evidence from the East Asian Financial Crisis J Finance 58, 1445–1468 Li, J., 2007 Fundamental enterprise income tax reform in China: motivations and major changes Bull Int Tax 61 (12), 519–528 Li, J., Huang, H 2008 The transformation of Chinese enterprise income tax: internationalization and Chinese innovations CLPE Research Paper No 41 Lins, K.V., 2003 Equity ownership and firm value in emerging markets J Financial Quant Anal 28, 159–184 Lisowsky, P., 2010 Seeking shelter: empirically modeling tax shelters using financial statement information Account Rev 85, 1693–1720 Liu, Y., Li, X., Zeng, H., An, Y., 2016 Political connections, auditor choice and corporate accounting transparency: evidence from private sector firms in China Account Finance in press Loderer, C., Martin, K., 1997 Executive stock ownership and performance: tracking faint traces J Financ Econ 45, 223–255 Manzon, G., Plesko, G., 2002 The relation between financial and tax reporting measures of income Tax Law Rev 55, 175–214 Masulis, R.W., Wang, C., Xie, F., 2009 Agency problems at dual-class companies J Finance 64, 1697–1727 McConnell, J.J., Servaes, H., 1990 Additional evidence on equity ownership and corporate value J Financ Econ 27, 595–612 McGuire, S., Wang, D., Wilson, R., 2014 Dual class ownership and tax avoidance Account Rev 89 (4), 1487–1516 McGuire, S.T., Omer, T.C., Wang, D., 2012 Tax avoidance: does tax-specific industry expertise make a difference? Account Rev 87 (3), 975–1003 Morck, R., Shleifer, A., Vishny, R.W., 1988 Management ownership and market valuation: an empirical analysis J Financ Econ 20, 293–315 National Bureau of Statistics of China, 2012 China statistical yearbook 2011 Pan, X., Tian, G., 2016 Family control and loan collateral: evidence from China J Bank Finance 67, 53–68 Plesko, G.A., 2003 An evaluation of alternative measures of corporate tax rates J Account Econ 35, 201–226 Rego, S.O., 2003 Tax-avoidance activities of US multinational corporations Contemp Account Res 20, 805–833 Richardson, G., Lanis, R., 2007 Determinants of the variability in corporate effective tax rates and tax reform: evidence from Australia J Account Public Policy 26, 689–704 Scholes, M., Wolfson, M., Erickson, M., Maydew, E., Shevlin, T., 2009 Taxes and Business Strategy: A Planning Approach, fourth ed Pearson Prentice-Hall, Upper Saddle River, New Jersey Shevlin, T., Tang, T.Y.H., Wilson, R.J., 2012 Domestic income shifting by Chinese listed firms J Am Tax Assoc 34, 1–29 Shleifer, A., Vishny, R., 1986 Large shareholders and corporate control J Polit Econ 94, 461–488 Shleifer, A., Vishny, R., 1997 A survey of corporate governance J Finance 52, 737–783 Stickney, C., McGee, V., 1982 Effective corporate tax rates: the effect of size, capital intensity, leverage, and other factors J Account Public Policy 1, 125–152 Tang, T., Firth, M., 2011a Can book-tax differences capture earnings management and tax management? Empirical evidence from China Int J Account 46, 175–204 Tang, T., Firth, M., 2011b Earnings persistence and stock market reactions to the different information in book-tax differences: evidence from China Int J Account 47, 369–397 Villalonga, B., Amit, R., 2006 How family ownership, control and management affect firm value? J Financ Econ 80, 385–417 Wilson, R., 2009 An examination of corporate tax shelter participants Account Rev 84, 969–999 Wooldridge, J.M., 2010 Econometric Analysis of Cross Section and Panel Data, second ed MIT Press, Cambridge, MA Wu, L., Yue, H., 2009 Corporate tax, capital structure, and the accessibility of bank loans: evidence from China J Bank Finance 33, 30–38 Wu, L., Wang, Y., Gillis, P., 2012a State ownership, tax status and size effect of effective tax rate in China Account Bus Res 42, 97–114 Wu, W., Wu, C., Zhou, C., Wu, J., 2012b Political connections, tax benefits and firm performance: evidence from China J Account Public Policy 31, 277–300 Please cite this article in press as: Grant Richardson, Bei Wang, Xinmin Zhang, Ownership structure and corporate tax avoidance: Evidence from publicly listed private firms in China, Journal of Contemporary Accounting & Economics (2016), doi: 10.1016/j.jcae.2016.06.003 ... set of firms to examine ownership structure? ??s influence on tax avoidance We develop three arguments in this study about the association between ownership structure and tax avoidance Our first argument... 285 286 287 288 289 290 291 3.3 Ownership structure and corporate tax avoidance We now consider the hypothesized associations between ownership structure and tax avoidance, positing three arguments... link between ownership structure and tax avoidance Second, this study documents new evidence on the: (1) non-linear U-shaped association between ownership concentration and tax avoidance which