Pyramidal structure, political intervention and firms' tax burden: Evidence from China's local SOEs

Journal of Corporate Finance 36 (2016) 15–25

Contents lists available at ScienceDirect

Journal of Corporate Finance journal homepage: www.elsevier.com/locate/jcorpfin

Pyramidal structure, political intervention and firms' tax burden: Evidence from China's local SOEs Min Zhang a,b, Lijun M a, Bo Zhang a,⁎, Zhihong Yi a a b

School of Business, Renmin University of China, 59 Zhongguancun Street, Haidian District, Beijing 100872, China The Collaborative Innovation Centre for State-owned Assets Administration 33 Fucheng Road, Haidian District, Beijing 100048, China

a r t i c l e

i n f o

Article history: Received 20 June 2015 Received in revised form 4 October 2015 Accepted 9 October 2015 Available online 17 October 2015 JEL classification: G32 G38 H25

a b s t r a c t Using a sample of Chinese firms, we examine the influence of state-pyramids on corporate tax burden. We find results that state-pyramidal layers are significantly and negatively associated with effective tax rates, indicating that pyramids formed by the state protect local state-owned enterprises (SOEs) from political intervention. The results hold after controlling for potential endogeneity. We further find evidences suggesting that taxation is one of the channels through which state-controlled pyramids increase firm value. Our study contributes to both corporate finance and corporate tax literatures by documenting the role of pyramidal organizational structures in reducing local SOEs' tax burden. © 2015 Elsevier B.V. All rights reserved.

Keywords: Pyramidal structure Local SOEs Political intervention Tax burden

1. Introduction A pyramidal ownership structure is generally referred to as a corporate structure in which the ultimate controlling shareholder at the top of the pyramid controls the firm indirectly via several layers of intermediate companies. This ownership structure is pervasive throughout the world and has been existing for several centuries (Chernykh, 2008; Claessens et al., 2000; de Jong et al., 2010; Faccio and Lang, 2002; Khanna and Yafeh, 2007; La Porta et al., 1999). Since pyramidal structure began to be examined in academic research for the first time, its causes, evolution, and economic consequences have been extensively studied. Pyramidal structure is perceived as an agency problem of the second type by most of the literature, which finds that pyramids can result in the expropriation of minority shareholders (Bebchuk, 1999; Bebchuk et al., 2000; La Porta et al., 1999) and a discount on performance and firm value (Claessens et al., 2002; La Porta et al., 2002; Lemmon and Lins, 2003), thus inhibiting the efficiency of resource allocation and economic growth (Morck et al., 2004). Prior research also documents the benefits of pyramidal organizational structures (Almeida and Wolfenzen, 2006; Almeida et al., 2011; Khanna and Palepu, 2000; Khanna and Rivkin, 2001; Khanna and Thomas, 2009). For example, studies find that privately controlled pyramids allow the controlling shareholders to build internal capital markets so to help them overcome capital market frictions (Khanna and Palepu, 2000; Khanna and Rivkin, 2001). On the other hand, the other type of pyramids, pyramids formed by the state, is relatively unexplored. Not until very recently have studies begun to examine state-pyramids (Fan et al., 2013; Liu et al., 2011). Besides agency costs, state pyramids also face ⁎ Corresponding author. Tel.: +86 10 62514992; fax: +86 10 82509169. E-mail addresses: [email protected] (M. Zhang), [email protected] (L. M), [email protected] (B. Zhang), [email protected] (Z. Yi).

http://dx.doi.org/10.1016/j.jcorpfin.2015.10.004 0929-1199/© 2015 Elsevier B.V. All rights reserved.

16

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

political costs associated with government's incentive to interfere in the firm (Fan et al., 2013). Political costs refer to the policy burdens, such as excessive taxation and redundant employment, on the SOEs in order to assist governments in achieving certain social and political objectives (Fan et al., 2013; Lin and Li, 2008; Qian, 1996; Qian and Wu, 2003). Fan et al. (2013) document that local governments build more extensive pyramidal layers when they have more incentives to protect SOEs through reducing political costs, and they further find a positive relationship between state pyramidal structures and firm performances. In this paper, we extend this emerging literature on the impact of state-pyramids on alleviating political intervention. Specifically, we use corporate tax burden as a proxy for government intervention and directly explore the role of pyramids in local SOEs in lessening excessive taxation. Taxation is the most common way through which the government interferes in SOEs, especially in emerging markets (Lin and Li, 2008; Lin et al., 1998). Because of their natural ties with governments, SOEs are often required to pay for the social expenses of governments through excessive taxation. As a result, a higher tax burden implies a higher extent of government intervention. We first test the influence of pyramids on corporate tax burden in a sample of Chinese listed firms controlled by local governments. The Chinese setting has several obvious advantages. First, pyramids formed by the state are particularly common in China and political intervention is an important institutional characteristic in China, thus offering us a great opportunity to explore the role of pyramids in alleviating government intervention. In addition, the central government of China has delegated fiscal rights to local governments since the 1980s. Meanwhile, local officials' promotion depends on regional gross domestic product (GDP) growth (Chen et al., 2005; Jin et al., 2005; Li and Zhou, 2005; Whiting, 2001), resulting in political competition among local governments to boost regional tax revenues (Besley and Case, 1995; Green and Stokey, 1983; Nalebuff and Stiglitz, 1983). This special institutional arrangement provides us an environment of political intervention through taxation. We hand-collect pyramidal structure data over 2004–2011. Using effective tax rates (ETRs) as a proxy for corporate tax burden, we find that pyramidal layers are negatively associated with ETRs for Chinese local SOEs, suggesting that state-controlled pyramids protect SOEs from political intervention through reducing local SOEs' taxation. In addition, our results hold when we use a two-stage least square (2SLS) approach to address the potential endogeneity in pyramidal ownership structure. We further find that corporate taxation partially mediates the positive impact of pyramidal layers on local SOEs' performance, suggesting that taxation is one of the channels through which state-controlled pyramids increase firm value. In addition, consistent with our expectations, we find that pyramidal layers in non-SOEs are not related to ETRs. Our paper makes several contributions to the existing literature. First, building upon prior research (Fan et al., 2013; Liu et al., 2011), we find that pyramidal layers in local SOEs are negatively associated with tax burden, thus providing direct evidences on the role of pyramids in alleviating political intervention. We further document that taxation is one of the important channels through which pyramidal layers affect firm value. Our study extends the emerging literature on state-controlled pyramids and enriches our understandings of pyramidal organizational structures. We also contribute to the corporate tax literature. Prior studies document that ownership structure affects corporate tax strategies (e.g., Bradshaw et al., 2013). Bradshaw et al. (2013) find that Chinese SOEs exhibit less tax avoidance than do non-SOEs. Our results further show that there are also variations among SOEs, and pyramidal structures influence local SOEs' tax rates. In addition, our study relates to the literature on the size–ETR relationship. Wu et al. (2012) find that ownership structure is an important factor determining the relationship between firm size and ETRs. In particular, Wu et al. (2012) document that firm size is negatively associated with ETRs for Chinese SOEs but the relationship is different for non-SOEs. Political power theory explains the negative relation for SOEs: larger SOEs possess more resources and abilities than smaller SOEs in lessening corporate tax burden. Different from Wu et al. (2012), the focus of our paper is pyramidal structure formed by the state but not firm size, and our study shows that pyramidal layers can protect local SOEs from tax burden. Finally, our results also provide some implications of the state-owned economy reform. Government can credibly decentralize real authority using pyramidal structures, thus reducing political costs and increasing the degree of autonomy and marketization of state-owned organizations. Because political intervention is a concern not only for China but also for many other developing and developed countries (Faccio, 2006; Khwaja and Mian, 2005; La Porta et al., 1999; Leuz and Oberholzer-Gee, 2006), these implications are not likely to be unique to China. The remainder of this paper is structured as follows: Section 2 reviews the related literature and develops the hypothesis. Section 3 describes the sample, data, and research design. Section 4 presents the empirical results. Additional tests are reported in Section 5. We conclude in Section 6. 2. Literature review and hypothesis development Pyramidal structure refers to the organization structure that the controlling shareholder at the top of the pyramid controls a firm indirectly through layers of intermediate companies (Fan et al., 2013). The organizational structure is very common around the world (Chernykh, 2008; Claessens et al., 2000; Faccio and Lang, 2002; La Porta et al., 1999). La Porta et al. (1999) suggest that the level of investor protection influences the choice of pyramidal structure. Specifically, pyramidal structure is more prevalent in poor investor protection countries as firms in those countries need concentrated control. However, concentrated ownership can induce controlling shareholders to expropriate the interest of minority shareholders, and the agency conflicts between controlling owners and minority shareholders occur due to a separation in voting rights and cash flow rights (Fan and Wong, 2002). Prior research further finds that the pyramidal ownership structure reduces firm value (Claessens et al., 2002; Lemmon and Lins, 2003). At the economy level, Morck et al. (2004) suggest that pyramidal control structure can distort capital allocation and slower economic growth.

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

17

On the other hand, prior research documents the positive effects of pyramids. Studies examining privately controlled pyramids find that pyramids allow the ultimate shareholders to construct an internal capital market so to help them conquer the frictions in capital markets (Khanna and Palepu, 2000; Khanna and Rivkin, 2001). Pyramids can act as an alternative financing mechanism (Almeida and Wolfenzen, 2006; Almeida et al., 2011), and there are information sharing advantages among business group members affiliated through interlocking directors (Khanna and Thomas, 2009). Nonetheless, pyramids formed by the state remain relatively unexplored. Pyramids controlled by governments are very common in transitional economies such as China (Fan et al., 2013), Russia (Chernykh, 2008), Hungary (Voska, 1993), and the Czech and Slovak Republics (OECD, 2005) as well as in developed economies such as France (de Jong et al., 2010) and Austria (OECD, 2005). Different from privately controlled pyramids, state pyramids face two types of organizational costs: political costs associated with government's incentive to interfere in the firm, and agency costs associated with manager's incentive to expropriate wealth from the firm (Fan et al., 2013). The political cost factor explains the differences between the two types of pyramids. Political costs refer to the policy burdens, such as excessive taxation and redundant employment, on the SOEs in order to assist governments in achieving certain social and political objectives (Fan et al., 2013; Lin and Li, 2008; Qian, 1996; Qian and Wu, 2003). Prior research documents that political costs lead to the poor performance of SOEs (Shleifer and Vishny, 1994, 1998), and impede corporate innovation and economic development (Chen et al., 2011; Cull and Xu, 2005; Faccio, 2010; Lin et al., 1998). Since the 1990s, the Chinese government has adopted a decentralized organization of SOEs in which state assets are stripped away from government agencies, spun off from parent SOEs, and injected into newly established subsidiaries (Fan et al., 2013).1 As a result, pyramid-like business groups are built where firms' decision-making rights are decentralized and productive assets are better allocated with most of the subsidiaries remaining majority-owned by governments (Aghion and Tirole, 1997). This decentralized pyramidal structure has become extremely prevalent in China's SOEs. Prior studies suggest that the cost of information transmission in a pyramidal structure reduces local governments' incentives to interfere (Baker, 1992; Cremer, 1995; Rajan and Wulf, 2006). Therefore, Fan et al. (2013) argue that a pyramidal structure is a credible mechanism to reduce government intervention. Focusing on the case of China, Fan et al. (2013) is the first one examining the economic determinants of the pyramids built by the state: insulating firms from government intervention. Specifically, Fan et al. (2013) find that local governments establish more extensive pyramids when they have incentives to protect SOEs by reducing the political costs associated with interference. Prior research also finds that state-pyramidal layers have positive performance effects (Fan et al., 2013; Liu et al., 2011). As one way to impose political costs is through corporate tax system (Holland, 1998), we use corporate tax burden as a proxy for government intervention. Built on Fan et al. (2013), we expect that pyramids built by the state can effectively alleviate corporate tax burden. In other words, we predict that the number of pyramidal layers in a local SOE is negatively associated with its tax burden. Similar to Fan et al. (2013), we focus on local SOEs and exclude firms controlled by the central government as central governments have less influence over SOEs than do local governments (Wang et al., 2008). Our main hypothesis is thus as follows: H1. The number of pyramidal layers in a local SOE is negatively associated with its tax burden. 3. Sample, data, and research design 3.1. Sample and data We hand-collect pyramidal layers data from the Chinese Securities Market and Accounting Research (CSMAR) over 2004– 2011. Our sample period starts from 2004 as pyramidal ownership data became largely available in the CSMAR since 2004. We obtain the financial data of listed firms in China from the CSMAR. Our initial sample contains all listed companies in Shanghai and Shenzhen Stock Exchanges for the period 2004–2011. We first delete firms controlled by the central government (Fan et al., 2013). We exclude firms in financial industries as their financial reporting and corporate tax practices are likely to be very different from those of other firms. Following prior studies (Gupta and Newberry, 1997; Kim and Limpaphayom, 1998; Stickney and McGee, 1982; Zimmerman, 1983), we also restrict our sample to firm-years with ETRs between zero and one. Last, we remove firms with missing information for calculation of empirical variables. The sample selection process yields 4016 firm-year local SOEs observations over the sample period. 3.2. Research design To test H1, we estimate the following regression model of the determinants of corporate tax decisions: ETR ¼ β0 þ β1 LAYER þ β2 LNASSET þ β3 LEV þ β4 CAPINT þ β5 INVINT X X þ β6 ROA þ β7 MB þ β8 DEFICIT þ βi Industryi þ β j Year j þ ε:

1

ð1Þ

See Fan et al. (2013) for a detailed introduction to China's reform of state-owned organizations and the formation of state-owned pyramidal structures in China.

18

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

Table 1 Sample distribution. Panel A: Distribution of pyramidal layers of local SOEs by year

Layer Layer Layer Layer Layer Layer Total

=1 =2 =3 =4 =5 ≥6

2004

2005

2006

2007

2008

2009

2010

2011

Total

Percentage

58 394 91 19 4 1 567

51 367 94 21 4 1 538

43 338 99 24 9 1 514

33 303 106 16 6 1 465

21 278 122 22 3 0 446

22 267 131 23 5 0 448

26 303 136 41 8 0 514

27 276 159 50 11 1 524

281 2526 938 216 50 5 4016

7.00% 62.90% 23.36% 5.38% 1.25% 0.12% 100.00%

Panel B: Distribution of ETRs of local SOEs by pyramidal layer

Layer Layer Layer Layer Layer Layer

=1 =2 =3 =4 =5 ≥6

Obs

ETR1

ETR2

ETR3

ETR4

281 2526 938 216 50 5

0.2503 0.2291 0.2168 0.2033 0.1717 0.1496

0.2696 0.2496 0.2398 0.2253 0.1927 0.1561

0.2354 0.2137 0.1973 0.1881 0.1490 0.1585

0.2475 0.2380 0.2289 0.2158 0.1775 0.2245

Table 1 presents the sample distribution. Panel A presents the distribution of pyramidal layers of local SOEs by year, and Panel B reports the distribution of ETRs by layers. LAYER is the number of pyramidal layers, and it is defined as the number of intermediate layers between the company and its controlling shareholder through the longest pyramidal chain. ETR1–ETR4 are the effective tax rates. ETR1 = Income tax expenses / Pretax profit. ETR2 = (Income tax expenses − Deferred tax expenses)/Pretax profit. ETR3 = Income tax expenses / (Pretax profit − Deferred tax expenses / Statutory tax rate). ETR4 = (Income tax expenses − Deferred tax expenses) / (Pretax profit − Change in deferred tax expenses / Statutory tax rate).

In Model (1), the dependent variable is the effective tax rate (ETR), which is used to measure firms' tax burden. Following prior research (Porcano, 1986; Shevlin, 1987; Stickney and McGee, 1982), we use four measures of ETR: ETR1 measured by Income tax expenses / Pretax profit, ETR2 measured by (Income tax expenses − Deferred tax expenses) / Pretax profit, ETR3 measured by Income tax expenses / (Pretax profit − Deferred tax expenses / Statutory tax rate), and ETR4 estimated by (Income tax expenses − Deferred tax expenses) / (Pretax profit − Change in deferred tax expenses / Statutory tax rate). Our main test variable, LAYER, is defined as the number of intermediate layers between a company and its controlling shareholder through the longest pyramidal chain (Fan et al., 2013). We hypothesize that the number of pyramidal layers in a local SOE is negatively associated with its tax burden. Therefore, we expect a significantly negative coefficient on LAYER. Following prior research, we add several control variables to Model (1). LNASSET is the natural logarithm of a firm's total assets, which is a proxy for firm size. Since larger firms have more resources for tax planning and political lobbying, they can enjoy a lower tax rate than smaller firms (Porcano, 1986; Siegfried, 1972). In addition, prior studies find that large firms in relationship-based economies, like China, are subject to lower ETRs for economic or “industrialization” reasons (Derashid and Zhang, 2003; Kim and Limpaphayom, 1998). Wu et al. (2012) also find that firm size is negatively associated with ETRs for state-controlled firms in China. Hence, we expect the coefficient on LNASSET to be negative. LEV is the overall debt levels of enterprises and is measured by the ratio of total liabilities to total assets. Because of tax-deductible interest payments, firms with higher leverage have relatively lower ETRs (Gupta and Newberry, 1997). However,

Table 2 Summary statistics of key variables.

ETR1 ETR2 ETR3 ETR4 LAYER LNASSET LEV CAPINT INVINT ROA MB DEFICIT

Obs

Mean

Std

Max

Median

Min

4016 4016 2989 2989 4016 4016 4016 4016 4016 4016 4016 4016

0.2255 0.2466 0.2087 0.2343 2.3135 21.8676 0.4975 0.3078 0.1732 0.0484 1.5494 0.0701

0.1280 0.1524 0.1269 0.1365 0.7433 1.0688 0.1786 0.1916 0.1560 0.0407 0.9413 0.0791

0.8581 0.9987 0.9855 0.9782 8.0000 25.1191 0.9487 0.7774 0.7399 0.2299 9.3106 1.1610

0.2054 0.2203 0.1847 0.2099 2.0000 21.7444 0.5106 0.2855 0.1357 0.0378 1.2216 0.0446

0.0000 0.0000 0.0000 0.0000 1.0000 18.5430 0.0487 0.0024 0.0000 0.0005 0.7907 0.0084

Table 2 presents summary statistics of key variables. LNASSET is the natural logarithm of firm's total assets. LEV is firm's leverage, measured by the ratio of total liability to total assets. CPAINT is the intensity of capital measured by the ratio of net fixed assets to total assets. INVINT is inventory intensity measured by the ratio of net inventory to total assets. ROA is firm's profitability, measured by the net profit divided by total assets. MB is market-to-book ratio, measured by firm's market value over total assets. DEFICIT is fiscal deficit of a province, measured by the province's total fiscal expenditure minus its fiscal revenue, divided by its gross GDP. Other variables are defined in Table 1.

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

19

Table 3 The effect of pyramidal layers on ETRs.

Intercept LAYER LNASSET LEV CAPINT INVINT ROA MB DEFICIT Industry dummies Year dummies Clustering by firms Adj-R2 N

(1) ETR1

(2) ETR2

(3) ETR3

(4) ETR4

0.4117 (8.81)⁎⁎⁎ −0.0132 (−5.10)⁎⁎⁎ −0.0070 (−3.19)⁎⁎⁎

0.3939 (7.11)⁎⁎⁎ −0.0149 (−4.83)⁎⁎⁎ −0.0052 (−2.01)⁎⁎

0.3502 (6.27)⁎⁎⁎ −0.0119 (−3.99)⁎⁎⁎ −0.0067 (−2.67)⁎⁎⁎

0.3585 (5.92)⁎⁎⁎ −0.0100 (−3.10)⁎⁎⁎ −0.0052 (−1.91)⁎

0.0292 (2.16)⁎⁎ 0.0175 (1.35) 0.1004 (5.48)⁎⁎⁎

0.0253 (1.58) 0.0007 (0.05) 0.1073 (4.94)⁎⁎⁎

0.0306 (1.94)⁎ 0.0307 (2.05)⁎⁎ 0.0715 (3.40)⁎⁎⁎

0.0253 (1.48) 0.0267 (1.64) 0.1022 (4.49)⁎⁎⁎

−0.7648 (−12.92)⁎⁎⁎ 0.0098 (3.59)⁎⁎⁎ 0.1092 (4.36)⁎⁎⁎ Yes Yes Yes 0.1401 4016

−1.0071 (−14.35)⁎⁎⁎ 0.0105 (3.26)⁎⁎⁎ 0.0859 (2.89)⁎⁎⁎ Yes Yes Yes 0.1467 4016

−0.3730 (−5.45)⁎⁎⁎ 0.0067 (2.27)⁎⁎ 0.1089 (3.87)⁎⁎⁎ Yes Yes Yes 0.1217 2989

−0.6645 (−8.95)⁎⁎⁎ 0.0079 (2.49)⁎⁎ 0.0848 (2.78)⁎⁎⁎ Yes Yes Yes 0.1073 2989

Table 3 presents the OLS regression results examining the impact of pyramidal layers on effective tax rates (ETRs). The dependent variable is ETRs, measured by ETR1 to ETR4 in column (1)–(4), respectively. All variables are defined in Tables 1 and 2. The OLS model is estimated with clustering error by firm. Absolute t-values are in parentheses. ⁎ 10% level of statistical significance. ⁎⁎ 5% level of statistical significance. ⁎⁎⁎ 1% level of statistical significance.

Table 4 The effect of pyramidal layers on ETRs after controlling for potential endogeneity. (1) ETR1

(2) ETR2

(3) ETR3

(4) ETR4

Intercept

0.4383 (8.74)⁎⁎⁎

0.4146 (6.97)⁎⁎⁎

0.3567 (5.92)⁎⁎⁎

FLAYER

−0.0186 (−2.91)⁎⁎⁎ −0.0072 (−3.29)⁎⁎⁎ 0.0303 (2.24)⁎⁎

−0.0187 (−2.47)⁎⁎ −0.0055 (−2.09)⁎⁎ 0.0264 (1.64) 0.0034 (0.22) 0.1070 (4.92)⁎⁎⁎ −0.9980 (−14.19)⁎⁎⁎

−0.0122 (−1.72)⁎ −0.0068 (−2.70)⁎⁎⁎ 0.0315 (1.99)⁎⁎ 0.0335 (2.22)⁎⁎ 0.0707 (3.36)⁎⁎⁎ −0.3637 (−5.31)⁎⁎⁎

0.3741 (5.73)⁎⁎⁎ −0.0120 (−1.57) −0.0054 (−1.96)⁎⁎ 0.0262 (1.53) 0.0286 (1.75)⁎ 0.1015 (4.45)⁎⁎⁎ −0.6566 (−8.84)⁎⁎⁎

0.0104 (3.19)⁎⁎⁎ 0.0882 (2.94)⁎⁎⁎ Yes Yes Yes 0.1430 4016

0.0065 (2.23)⁎⁎ 0.1092 (3.82)⁎⁎⁎ Yes Yes Yes 0.1179 2989

0.0078 (2.45)⁎⁎ 0.0863 (2.79)⁎⁎⁎ Yes Yes Yes 0.1052 2989

LNASSET LEV CAPINT INVINT ROA MB DEFICIT Industry dummies Year dummies Clustering by firms Adj-R2 N

0.0194 (1.50) 0.1000 (5.45)⁎⁎⁎ −0.7565 (−12.75)⁎⁎⁎ 0.0096 (3.50)⁎⁎⁎ 0.1123 (4.43)⁎⁎⁎ Yes Yes Yes 0.1363 4016

Table 4 presents results examining the impact of pyramidal layers on ETRs after controlling for potential endogeneity. In the first stage, we estimate pyramidal layers (LAYER) as a function of industry average of pyramidal layers (INDLAYER) and other control variables in model (1). We then use the fitted value of pyramidal layers (FLAYER) obtained from the first stage estimation as our main independent variable. The dependent variable is ETR, measured by ETR1 to ETR4 in column (1)–(4), respectively. Other variables are defined in Tables 1 and 2. The 2SLS models are estimated with clustering error by firm. Regression results of the first stage are not reported. Absolute t-values are in parentheses. ⁎ 10% level of statistical significance. ⁎⁎ 5% level of statistical significance. ⁎⁎⁎ 1% level of statistical significance.

20

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

Table 5 Pyramidal structure, tax burden and firm value. Panel A: The effect of pyramidal layers on firm value

Intercept LAYER LNASSET LEV CAPINT INVINT ROA MB Industry dummies Year dummies Clustering by firms Adj-R2 N

Coefficient

t-Value

−3.7977 0.0214 0.8430 −0.9905 −0.0122 0.1070 4.1423 0.3325 Yes Yes Yes 0.9035 4016

(−28.02)⁎⁎⁎ (2.86)⁎⁎⁎ (133.84)⁎⁎⁎ (−25.29)⁎⁎⁎ (−0.33) (2.02)⁎⁎ (24.14)⁎⁎⁎ (42.04)⁎⁎⁎

Panel B: The effect of ETRs on firm value

Intercept ETR LNASSET LEV CAPINT INVINT ROA MB Industry dummies Year dummies Clustering by firms Adj-R2 N

(1) ETR1

(2) ETR2

(3) ETR3

(4) ETR4

−3.6881 (−27.32)⁎⁎⁎ −0.1401 (−3.07)⁎⁎⁎ 0.8418 (133.59)⁎⁎⁎

−3.6996 (−27.48)⁎⁎⁎ −0.1145 (−2.97)⁎⁎⁎ 0.8421 (133.70)⁎⁎⁎

−3.8386 (−23.5)⁎⁎⁎ −0.1190 (−2.21)⁎⁎ 0.8474 (115.15)⁎⁎⁎

−3.8317 (−23.47)⁎⁎⁎ −0.1321 (−2.66)⁎⁎⁎ 0.8475 (115.25)⁎⁎⁎

−0.9857 (−25.16)⁎⁎⁎ −0.0193 (−0.52) 0.1191 (2.24)⁎⁎ 4.0280 (23.03)⁎⁎⁎

−0.9868 (−25.19)⁎⁎⁎ −0.0216 (−0.58) 0.1175 (2.21)⁎⁎ 4.0194 (22.86)⁎⁎⁎

−0.9755 (−21.02)⁎⁎⁎ 0.0149 (0.34) 0.1597 (2.58)⁎⁎⁎ 3.9234 (19.38)⁎⁎⁎

−0.9758 (−21.04)⁎⁎⁎ 0.0148 (0.34) 0.1648 (2.66)⁎⁎⁎ 3.8798 (19.01)⁎⁎⁎

0.3335 (42.11)⁎⁎⁎ Yes Yes Yes 0.9035 4016

0.3333 (42.10)⁎⁎⁎ Yes Yes Yes 0.9034 4016

0.3262 (37.70)⁎⁎⁎ Yes Yes Yes 0.9004 2989

0.3265 (37.74)⁎⁎⁎ Yes Yes Yes 0.9005 2989

Panel C: The effect of pyramidal layers and ETRs on firm value

Intercept LAYER ETR LNASSET LEV CAPINT INVINT ROA MB Industry dummies Year dummies Clustering by firms Adj-R2 N

(1) ETR1

(2) ETR2

(3) ETR3

(4) ETR4

−3.7456 (−27.41)⁎⁎⁎ 0.0198 (2.63)⁎⁎⁎

−3.7567 (−27.58)⁎⁎⁎ 0.0199 (2.65)⁎⁎⁎

−3.9069 (−23.61)⁎⁎⁎ 0.0218 (2.49)⁎⁎

−3.8996 (−23.58)⁎⁎⁎ 0.0218 (2.49)⁎⁎

−0.1309 (−2.86)⁎⁎⁎ 0.8422 (133.71)⁎⁎⁎ −0.9870 (−25.21)⁎⁎⁎

−0.1069 (−2.77)⁎⁎⁎ 0.8425 (133.83)⁎⁎⁎ −0.9880 (−25.24)⁎⁎⁎

−0.1099 (−2.04)⁎⁎ 0.8479 (115.28)⁎⁎⁎ −0.9770 (−21.07)⁎⁎⁎

−0.1255 (−2.52)⁎⁎ 0.8480 (115.38)⁎⁎⁎ −0.9771 (−21.08)⁎⁎⁎

−0.0103 (−0.28) 0.1197 (2.25)⁎⁎ 4.0433 (23.12)⁎⁎⁎

−0.0124 (−0.33) 0.1182 (2.22)⁎⁎ 4.0353 (22.96)⁎⁎⁎

0.0247 (0.56) 0.1593 (2.58)⁎⁎ 3.9420 (19.48)⁎⁎⁎

0.0247 (0.56) 0.1643 (2.65)⁎⁎⁎ 3.8995 (19.11)⁎⁎⁎

0.3337 (42.17)⁎⁎⁎ Yes Yes Yes 0.9036 4016

0.3336 (42.16)⁎⁎⁎ Yes Yes Yes 0.9036 4016

0.3263 (37.74)⁎⁎⁎ Yes Yes Yes 0.9006 2989

0.3266 (37.78)⁎⁎⁎ Yes Yes Yes 0.9006 2989

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

21

firms whose tax rates are higher may borrow more to ease their tax burden (Derashid and Zhang, 2003; Kim and Limpaphayom, 1998). Thus, we make no prediction on the sign of LEV. CPAINT is capital intensity and is measured by the ratio of net fixed assets to total assets; INVINT is the intensity of inventory and is measured by the ratio of net inventory to total assets (Stickney and McGee, 1982; Gupta and Newberry, 1997). A higher capital intensity (CPAINT) is associated with a lower tax burden because the accelerated depreciation of long-term assets has the effect of tax saving. A higher level of capital intensity (CPAINT) is usually associated with a lower level of inventory intensity (INVINT). Therefore, we predict a negative (positive) coefficient on CPAINT (INVINT). ROA is a proxy for profitability, which equals net profit divided by total assets. Firms have more taxable income when they are more profitable, leading to a positive relation between ETR and ROA (Kim and Limpaphayom, 1998; Spooner, 1986; Zimmerman, 1983). On the other hand, prior research argues that more efficient firms, measured by profitability, pay less effective tax (Derashid and Zhang, 2003). Maker-to-book ratio, MB, is a proxy for firms' investment opportunities, and it is measured by firm's market value over total assets. Spooner (1986) finds that firms with greater investment opportunities have higher ETRs. However, there are studies documenting inconsistent results (e.g., Chen et al., 2010). Therefore, we make no prediction on the sign of ROA and MB. DEFICIT is a proxy for the fiscal deficit of a province where a sample firm is located, and it is measured by subtracting a province's total fiscal revenue from its fiscal expenditures and then dividing the difference by the province's gross GDP. Local governments with larger deficits are more likely to levy taxes to ease their financial pressure. Thus, we expect a positive coefficient on DEFICIT. In addition, year and industry dummies are included in model (1). 4. Empirical results 4.1. Descriptive statistics Table 1 presents sample distribution. Panel A of Table 1 shows the distribution of pyramidal layers of local SOEs by year. Firms with one layer are directly controlled by their ultimate owners, that is, the local State-Owned Assets Supervision and Administration institutions or a local government. There are 7.00% of firm-year observations with only one control layer. Similar to Fan et al. (2013), we document that the majority of our sample are linked to their controlling owners through two-layer pyramids. In addition, 938 observations (23.36%) are linked to their ultimate owners through three pyramidal layers. Panel B shows the average ETRs on each pyramidal layer. It shows that the mean values of ETRs continuously decrease as the number of layers increases, providing preliminary evidences that the ETRs of local SOEs decrease as their pyramidal layers increase. Table 2 presents summary statistics of key variables. The mean value of ETR1 is 0.2255, suggesting that the average level of Chinese local SOEs' ETRs is approximately 23%. In addition, ETRs vary significantly among firms. The other three measures of ETRs show a statistically similar pattern. Similar to Fan et al. (2013), we document that the mean (median) value of pyramidal layers (LAYER) is 2.3135 (2.0000). The mean (median) value of leverage (LEV) is 0.4975 (0.5106). CAPINT and INVINT have mean values of 0.3078 and 0.1732, respectively. ROA has a mean value of 0.0484 and a minimum above zero as firms with negative ETRs are excluded from our sample (Gupta and Newberry, 1997; Kim and Limpaphayom, 1998; Stickney and McGee, 1982; Zimmerman, 1983). 4.2. Regression results Table 3 shows results from estimating Eq. (1). As mentioned above, the dependent variable, ETRs, is measured using four methods ETR1–ETR4. Due to missing values in calculating ETR3 and ETR4, we have a total number of observations of 2989 in column (3) and column (4). Consistent with our expectations, the coefficients on LAYER are significantly negative in all four columns, indicating that pyramidal layers are negatively associated with corporate tax burden. Therefore, our results suggest that pyramids built by local governments can protect local SOEs from corporate tax burden. Economically, we find that a one-standard deviation increase in pyramidal layers corresponds to a 4.35% (4.49%, 4.24%, and 3.17%) decrease in ETR1 (ETR2, ETR3, and ETR4). The results for the control variables are generally consistent with our expectations and prior studies. Consistent with prior studies (e.g., Wu et al., 2012), we find a negative and significant coefficient on firm size (LNASSET), suggesting that firm size is negatively associated with ETRs for local SOEs in China. The positive and significant coefficient on leverage (LEV) indicates that local SOEs with higher tax rates tend to borrow more, resulting in a positive relation between financial leverage and effective tax rates. Consistent with our expectations, we find that both inventory intensity (INVINT) and financial deficit (DEFICIT) are positively associated with ETRs. Furthermore, the negative coefficient on profitability (ROA) suggests that more efficient local SOEs Notes to Table 5 Table 5 reports test results on the role of ETR in mediating the positive impact of pyramidal layers on firm value, LNMV is the natural logarithm of firm market value. Panel A shows the results of the impact of pyramidal layers on firm value. Panel B presents the results of the effect of ETRs on firm value. Panel C reports the results of including both pyramidal layers and ETRs in model (4). Other variables are defined in Tables 1 and 2. The OLS models are estimated with clustering error by firm. Absolute t-values are in parentheses. ⁎ 10% level of statistical significance. ⁎⁎ 5% level of statistical significance. ⁎⁎⁎ 1% level of statistical significance.

22

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

pay less tax (Derashid and Zhang, 2003). We find that market-to-book ratio (MB) is also positively associated with ETRs, indicating that firms with greater investment opportunities have higher ETRs (Spooner, 1986). However, we find no association between capital intensity (CPAINT) and effective tax rates for most of the ETR measures. In addition, the results are generally consistent across all four columns. Taken together, consistent with our hypothesis, results suggest that pyramids built by local governments protect local SOEs from political intervention by reducing their tax burden. 4.3. Endogeneity As both pyramidal layers and ETRs may be correlated with some omitted variables, we employ a two stage-least square (2SLS) approach to control for potential endogeneity issue. We choose the industry average of pyramidal layers, INDLAYER, as our instrumental variable. INDLAYER is positively correlated with LAYER but it is less likely to directly affect the firm-level ETRs. In the first stage, we estimate pyramidal layers (LAYER) as a function of INDLAYER and other control variables in Eq. (1). We then use the fitted value of pyramidal layers (FLAYER) obtained from the first stage estimation as our main independent variable in the second stage. Table 4 shows that the coefficient on FLAYER is negative and significant for most of the ETRs measures. Therefore, we find consistent evidences after controlling the potential endogenity in pyramidal organizational structure. 5. Additional tests 5.1. Pyramidal structure, tax burden and firm value Prior studies document that pyramids built by the state increase firm value (Fan et al., 2013; Liu et al, 2011). In this section, we further test whether the positive impact of pyramidal layers on firm value is mediated by reduced tax burden. Following Baron and Kenny (1986) and Banker et al. (2008), we estimate the following models: LNMV ¼ β 0 þ β1 LAYER þ β2 LNASSET þ β3 LEV þ β4 CAPINT þ β5 INVINT X X þ β6 ROA þ β7 MB þ βi Industryi þ β j Year j þ ε;

ð2Þ

LNMV ¼ β0 þ β1 ETR þ β2 LNASSET þ β3 LEV þ β4 CAPINT þ β5 INVINT X X þ β6 ROA þ β7 MB þ βi Industryi þ β j Year j þ ε;

ð3Þ

LNMV ¼ β0 þ β1 LAYER þ β2 ETR þ β3 LNASSET þ β4 LEV þ β5 CAPINT X X þ β6 INVINT þ β7 ROA þ β8 MB þ βi Industryi þ β j Year j þ ε:

ð4Þ

In model (2), we first estimate the impact of pyramidal layers (LAYER) on firm value. Firm value is measured by the natural logarithm of firms' market value (LNMV). We next use mode (3) to estimate the impact of the mediating variable, effective tax rates, on firm value. The two models estimate the independent impacts of pyramidal layers and effective tax rates on local SOEs' firm value. Finally, we include both LAYER and ETR in model (4), which represents a partial mediating effect where the impact of LAYER is partially mediated through ETR. In addition, we control for other variables in the three models, including firm size (LNASSET), financial leverage (LEV), capital intensity (CAPINT), inventory intensity (INVINT), profitability (ROA), and market-to-book ratio (MB). Panels A, B, and C in Table 5 show the regression results for Models (2), (3), and (4), respectively. Consistent with prior studies (Fan et al., 2013; Liu et al., 2011), we find results that pyramids built by the state increase firm value. Specifically, in panel A of Table 5, the estimated coefficient on LAYER is significantly positive. Panel B of Table 5 shows the regression results of model (3). Using four different measures of ETRs, the estimated coefficients on ETR are significantly negative in all columns, indicating that the tax burden on corporate reduces firm value. In addition, Panel C shows that the coefficient on LAYER (ETRs) continues to be significantly positive (negative) after including both pyramidal layers and ETRs. Therefore, the results indicate that lessening the tax burden on corporate is one of the channels through which pyramidal layers increase firm value. We acknowledge that there are other potential channels, like employment efficiency and productivity (Fan et al., 2013). 5.2. Pyramidal structure in non-SOEs In this session, we examine the impact of pyramidal layers on corporate tax burden for non-SOEs. Different from SOEs, nonSOEs are subject to less political intervention. Therefore, we expect that there is no such effect of pyramidal structure in reducing corporate tax burden for non-SOEs. In Panel A of Table 6, we first report the distribution of non-SOEs by year and by pyramidal layer. Among the 2881 observations, 1444 (50.12%) non-SOEs are linked with their controlling owners through two layers and 639 (22.18%) are linked through three layers. These two groups together cover 72.30% of the non-SOE sample. In addition, the number of observations continues to increase during our sample period. Panel B of Table 6 shows the average effective tax

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

23

Table 6 Distribution of non-SOE sample. Panel A: Distribution of pyramidal layers of non-SOEs by year

Layer Layer Layer Layer Layer Layer Total

=1 =2 =3 =4 =5 ≥6

2004

2005

2006

2007

2008

2009

2010

2011

Total

Percentage

20 115 68 23 6 1 233

28 145 60 22 7 3 265

25 154 66 24 8 3 280

35 153 76 34 7 4 309

63 177 77 30 9 3 359

85 200 80 26 11 4 406

110 248 101 30 12 5 506

108 252 111 29 16 7 523

474 1444 639 218 76 30 2881

16.45% 50.12% 22.18% 7.57% 2.64% 1.04% 100.00%

Panel B: Distribution of ETRs of local SOEs by pyramidal layer

Layer Layer Layer Layer Layer Layer

=1 =2 =3 =4 =5 ≥6

Nobs

ETR1

ETR2

ETR3

ETR4

474 1444 639 218 76 30

0.1659 0.2088 0.2181 0.2278 0.2159 0.2260

0.1851 0.2313 0.2387 0.2545 0.2345 0.2457

0.1549 0.1971 0.1992 0.2094 0.2353 0.2172

0.1833 0.2224 0.2205 0.2453 0.2178 0.2557

Table 6 presents the sample distribution of non-SOEs. Panel A presents the distribution of pyramidal layers of non-SOEs by year, and Panel B reports the distribution of ETRs by layers. All variables are defined in Tables 1 and 2.

rates for each pyramidal layer. Unlike local SOEs, there is no evidence suggesting that effective tax rates decrease as the number of layers increases. Table 7 shows the regression results from the estimation of model (1) for non-SOEs. Again, we use four measures of effective tax rates. Consistent with our expectations, the coefficients on LAYER are insignificant in all four columns. The results suggest that pyramidal layers in non-SOEs have no effect on reducing corporate tax burden. 6. Conclusions In this study, we examine the effect of pyramids formed by the state on corporate tax burden. We hand-collect pyramidal layers data from the Chinese Securities Market and Accounting Research (CSMAR) over 2004–2011. Using effective tax rates as Table 7 The effect of pyramidal layers on ETRs for non-SOEs.

Intercept LAYER LNASSET LEV CAPINT INVINT ROA MB DEFICIT Industry dummies Year dummies Clustering by firms Adj-R2 N

(1) ETR1

(2) ETR2

(3) ETR3

(4) ETR4

0.3181 (5.60)⁎⁎⁎ 0.0022 (1.00) −0.0041 (−1.47) 0.0333 (2.12)⁎⁎ −0.0005 (−0.03) 0.0712 (3.87)⁎⁎⁎

0.2887 (4.34)⁎⁎⁎ 0.0023 (0.90) −0.0027 (−0.85) 0.0548 (2.95)⁎⁎⁎ −0.0173 (−0.90) 0.0757 (3.50)⁎⁎⁎

0.3412 (4.94)⁎⁎⁎ 0.0039 (1.46) −0.0049 (−1.46) 0.0221 (1.13) −0.0021 (−0.10) 0.0567 (2.50)⁎⁎

0.3038 (4.24)⁎⁎⁎ 0.0012 (0.46) −0.0044 (−1.28) 0.0594 (2.93)⁎⁎⁎ 0.0064 (0.30) 0.0729 (3.10)⁎⁎⁎

−0.7446 (−13.86)⁎⁎⁎ 0.0090 (3.88)⁎⁎⁎ −0.0716 (−3.16)⁎⁎⁎ Yes Yes Yes 0.2143 2881

−1.0132 (−14.53)⁎⁎⁎ 0.0068 (2.49)⁎⁎ −0.0958 (−3.59)⁎⁎⁎ Yes Yes Yes 0.2120 2881

−0.4449 (−6.09)⁎⁎⁎

−0.7688 (−10.16)⁎⁎⁎ 0.0079 (2.77)⁎⁎⁎ −0.0690 (−2.43)⁎⁎ Yes Yes Yes 0.1700 2188

0.0049 (1.78)⁎ −0.0502 (−1.83)⁎ Yes Yes Yes 0.1578 2188

This table reports test results of the relation between pyramidal layers and ETRs for non-SOE sample. The dependent variable is ETRs, measured by ETR1 to ETR4 in column (1)–(4), respectively. All the variables are defined in Tables 1 and 2. The OLS model is estimated with clustering error by firm. Absolute t-values are in parentheses. ⁎ 10% level of statistical significance. ⁎⁎ 5% level of statistical significance. ⁎⁎⁎ 1% level of statistical significance.

24

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

a proxy for corporate tax burden, we find that pyramidal layers in local SOEs are significantly and negatively associated with ETRs. The results suggest that pyramids built by local governments can protect SOEs from corporate tax burden. In addition, our results hold when we use a two-stage least square (2SLS) approach to address the potential endogeneity in pyramidal ownership structure. We further find evidences suggesting that taxation is one of the channels through which state-controlled pyramids increase firm value. Our paper makes two main contributions to the existing literature. We first contribute to corporate finance literature on pyramidal organizational structure. Our results suggest that state-pyramidal layers protect local SOEs from excessive tax burden, thus enlarging our understandings of pyramids formed by the state. We also contribute to corporate tax literature on the effect of ownership structure on corporate tax decisions (Bradshaw et al., 2013). There are limitations to our analysis. We document significantly negative associations between pyramidal layers formed by the state and ETRs. However, because of the nature of our empirical tests, we cannot draw conclusions regarding causality. Although we use a two stage-least square (2SLS) approach to reduce omitted variable bias, we acknowledge that it is possible that our results might be affected by a correlated omitted variable that explains both ownership structure and corporate tax policy. Acknowledgments Min Zhang acknowledges the financial support from the National Natural Science Foundation of China (no. 71432008), Ministry of Education in China Project of Humanities and Social Sciences (no. 15YJA630101), and the Collaborative Innovation Centre for Stateowned Assets Administration of Beijing Technology and Business University (GZ20130801). Bo Zhang acknowledges the financial support from the National Natural Science Foundation of China (no. 71402185). Zhihong Yi acknowledges the financial support from the National Natural Science Foundation of China (no. 71572192). References Aghion, P., Tirole, J., 1997. Formal and real authority in organizations. J. Polit. Econ. 105, 1–29. Almeida, H., Wolfenzen, D., 2006. A theory of pyramidal ownership and family business groups. J. Financ. 61, 2637–2681. Almeida, H., Park, S.Y., Subrahmanyam, M.G., Wolfenzon, D., 2011. The structure and formation of business groups: evidence from Korean chaebols. J. Financ. Econ. 99, 447–475. Baker, G.P., 1992. Beatrice: a study in the creation and destruction of value. J. Financ. 47, 1081–1119. Banker, R.D., Bardhan, I.R., Chen, T., 2008. The role of manufacturing practices in mediating the impact of activity-based costing on plant performance. Acc. Organ. Soc. 33, 1–19. Baron, R.M., Kenny, D.A., 1986. The moderator–mediator variable distinction in social psychological research: conceptual, strategic and statistical considerations. J. Pers. Soc. Psychol. 51, 1173–1182. Bebchuk, L., 1999. A Rent-Seeking Theory of Corporate Ownership and Control. NBER Working paper. Number 7203. Bebchuk, L., Kraakman, R., Triantis, G., 2000. Stock pyramids, cross-ownership, and dual class equity: the creation and agency costs of separating control from cash flow rights. In: Morck, R. (Ed.), Concentrated Corporate Ownership. National Bureau of Economic Research Conference Volume. University of Chicago Press, Chicago. Besley, T., Case, A., 1995. Incumbent behavior: vote-seeking, tax-setting, and yardstick competition. Am. Econ. Rev. 85, 25–45. Bradshaw, M., Liao, G., Ma, M., 2013. Ownership Structure and tax Avoidance: Evidence from Agency Costs of State Ownership in China. Working Paper (http://ssrn. com/abstract=2239837). Chen, Y., Li, H., Zhou, L., 2005. Relative performance evaluation and the turnover of provincial leaders in China. Econ. Lett. 88, 421–425. Chen, S., Chen, X., Cheng, Q., Shevlin, T.J., 2010. Are family firms more tax aggressive than non-family firms? J. Financ. Econ. 95, 41–61. Chen, S., Sun, Z., Tang, S., Wu, D., 2011. Government intervention and investment efficiency: evidence from China. J. Corp. Finance 17 (2), 259–271. Chernykh, L., 2008. Ultimate ownership and control in Russia. J. Financ. Econ. 88, 169–192. 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, P.H., Lang, L.H.P., 2002. Disentangling the incentive and entrenchment effects of large shareholdings. J. Financ. 57, 2741–2771. Cremer, J., 1995. Arm's length relationships. Q. J. Econ. 110, 275–295. Cull, R., Xu, L.C., 2005. Institutions, ownership and finance: the determinants of profit reinvestment among Chinese firms. J. Financ. Econ. 77, 117–146. de Jong, Douglas V., Jong, A.D., Mertens, G., Hege, U., 2010. Leverage in Pyramids: When Debt Leads to Higher Dividends. Working Paper. HEC, Paris. Derashid, C., Zhang, H., 2003. Effective tax rates and the “industrial policy” hypothesis: evidence from Malaysia. J. Int. Account. Audit. Tax. 12 (1), 45–62. Faccio, M., 2006. Politically connected firms. Am. Econ. Rev. 96, 369–386. Faccio, M., 2010. Differences between politically connected and non-connected firms: a cross country analysis. Financ. Manag. 39 (3), 905–927. Faccio, M., Lang, L.H.P., 2002. The ultimate ownership of Western European corporations. J. Financ. Econ. 65, 365–395. Fan, J., Wong, T.J., 2002. Corporate ownership structure and the informativeness of accounting earnings in East Asia. J. Account. Econ. 33, 401–425. Fan, J., 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. Green, J.R., Stokey, N., 1983. A comparison of tournaments and contracts. J. Polit. Econ. 3, 349–364. Gupta, S., Newberry, K., 1997. Determinants of the variability in corporate effective tax rate: evidence from longitudinal data. J. Account. Public Policy 16 (1), 1–39. Holland, K., 1998. Accounting policy choice: the relationship between corporate tax burden and company size. J. Bus. Finance Account. 25 (3 and 4), 265–288. Jin, H., Qian, Y., Weingast, B.R., 2005. Regional decentralization and fiscal incentives: federalism, Chinese style. J. Public Econ. 89, 1719–1742. Khanna, T., Palepu, K.G., 2000. Is group affiliation profitable in emerging markets? an analysis of diversified Indian business groups. J. Financ. 55, 867–891. Khanna, T., Rivkin, J.W., 2001. Estimating the performance effects of business groups in emerging markets. Strateg. Manag. J. 22, 45–74. Khanna, T., Thomas, C., 2009. Synchronicity and firm interlocks in an emerging market. J. Financ. Econ. 99, 182–204. Khanna, T., Yafeh, Y., 2007. Business groups in emerging markets, paragons or parasites? J. Econ. Lit. 45, 331–372. Khwaja, A.I., Mian, A., 2005. Do lenders favor politically connected firms? rent-seeking in an emerging financial market. Q. J. Econ. 120, 1371–1411. Kim, K.A., Limpaphayom, P., 1998. Taxes and firm size in pacific-basin emerging economies. J. Int. Account. Audit. Tax. 7 (1), 47–63. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. J. Financ. 54, 471–517. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 2002. Investor protection and corporate value. J. Financ. 57, 1147–1170. Lemmon, M.L., Lins, K.V., 2003. Ownership structure, corporate governance, and firm value: evidence from the East Asian financial crisis. J. Financ. 58, 1445–1468. Leuz, C., Oberholzer-Gee, F., 2006. Political relationships, global financing, and corporate transparency. J. Financ. Econ. 81, 411–439. Li, H., Zhou, L.A., 2005. Political turnover and economic performance: the incentive role of personnel control in China. J. Public Econ. 89 (9), 1743–1762. Lin, J.Y., Li, Z., 2008. Policy burden, privatization and soft budget constraint. J. Comp. Econ. 36, 90–102. Lin, J.Y., Cai, F., Li, Z., 1998. Competition, policy burdens, and state-owned enterprise reform. Am. Econ. Rev. 88 (2), 422–427. Liu, Q., Zheng, Y., Zhu, Y., 2011. On the Chinese Pyramids: A Dichotomy of Causes, Evolution and Consequences. Working Paper. University of Hong Kong.

M. Zhang et al. / Journal of Corporate Finance 36 (2016) 15–25

25

Morck, R., Wolfenzon, D., Yeung, B., 2004. Corporate governance, economic entrenchment, and growth. J. Econ. Lit. 43 (3), 655–720. Nalebuff, B.J., Stiglitz, J.E., 1983. Prizes and incentives: towards a general theory of compensation and competition. Bell J. Econ. 14, 21–43. Organization for Economic Co-operation and Development, 2005. Corporate Governance of State-Owned Enterprises: A Survey of OECD Countries. The OECD Publishing. Porcano, T.M., 1986. Corporate tax rates: progressive, proportional, or regressive. J. Am. Tax. Assoc. 7 (2), 17–31. Qian, Y.Y., 1996. Enterprise reform in China: agency problems and political control. Econ. Transit. 4, 427–447. Qian, Y.Y., Wu, J.L., 2003. China's transition to a market economy: how far across the river. In: Hope, N.C., Yang, D.T., Li, M.Y. (Eds.), How far Across the River: Chinese Policy Reform at the Millennium. Stanford University Press, Stanford. Rajan, R.G., Wulf, J., 2006. The flattening firm: evidence on the changing nature of firm hierarchies from panel data. Rev. Econ. Stat. 88, 759–773. Shevlin, T., 1987. Taxes and off-balance-sheet financing: research and development limited partnerships. Account. Rev. 62, 480–509. Shleifer, A., Vishny, R.W., 1994. Politicians and firms. Q. J. Econ. 109, 995–1025. Shleifer, A., Vishny, R.W., 1998. The Grabbing Hand: Government Pathologies and Their Cures. Harvard University Press. Siegfried, J., 1972. The relationship between economic structure and the effect of political influence: empirical evidence from the federal corporation income tax program (Ph.D. dissertation) University of Wilconsin. Spooner, G.M., 1986. Effective tax rates from financial statements. Natl. Tax J. 39 (3), 293–306. Stickney, C.P., McGee, V.E., 1982. Effective corporate tax rates: the effect of size, capital intensity, leverage and other factors. J. Account. Public Policy 1 (2), 125–152. Voska, E., 1993. Restructuring of Large State Owned Enterprises in Hungary 1988–1993. United Nations Economic Commission for Europe. Wang, Q., Wong, T., Xia, L., 2008. State ownership, the institutional environment, and auditor choice: evidence from China. J. Account. Econ. 46 (1), 112–134. Whiting, S.H., 2001. Power and Wealth in Rural China: The Political Economy of Institutional Change. Cambridge University Press, Cambridge. Wu, L., Wang, Y., Luo, W., Gillis, P., 2012. State ownership, tax status and size effect of effective tax rate in China. Account. Bus. Res. 42 (2), 97–114. Zimmerman, J.L., 1983. Taxes and firm size. J. Account. Econ. 5, 119–149.