Decentralized Privatization and Change of Control Rights in China

Decentralized Privatization and Change of Control Rights in China The design and implementation of privatization in China is unique in that both are decentralized and administered by the local governments. Based on a proprietary survey data set containing 3,000 firms in over 200 cities, this paper studies privatization choices and outcomes, as well as the mechanism behind the outcomes. We find that less political opposition to labor downsizing and greater fiscal capacity prompt cities to choose direct sales to insiders (MBOs). This method transfers control rights to private owners, retains limited government supports, imposes hardened budget constraints, allows for restructuring, and achieves performance improvement. Received September 8, 2015; editorial decision February 3, 2017 by Editor Andrew Karolyi. Privatization in China between the late 1990s and the mid-2000s was arguably the largest in the world and still influences governance in the Chinese economy in a profound way.1 Our understanding of this vast transformation, however, remains limited, because there is little data, other than that available from the small fraction of firms that underwent share issue privatization (SIP) and became publicly listed. A distinct feature of privatization in China is that both its design and its implementation are highly decentralized and are administered by the local governments. This feature is in contrast to privatization in most other nations that followed a nationwide policy and implemented privatization in a top-down manner.2 No de jure national privatization policy took place in China. Instead, a few city governments first initiated China’s de facto privatization at a time when the central government was cautious about privatization. Later, after the central government endorsed the practice of selling state-owned enterprise (SOE) assets to private owners, for most SOEs, city governments decided whether to privatize, and, if the decision was yes, what privatization approach to adopt. As a result, privatization methods across Chinese cities widely varied. This decentralized feature of privatization in China is not only critically important for understanding the Chinese economy but also provides a rich setting in which to study privatization and institutions in general. We design and conduct a large-scale nationwide survey of 3,000 firms in more than 200 cities. These proprietary survey data allows us to carry out a systematic study of decentralized privatization in China, in an attempt to draw implications for privatization design and, more generally, the design of economic institutions. We seek to understand how local governments choose different privatization methods and how these various methods are responsible for transferring control rights of the firms and lead to diverse mechanisms with respect to restructuring and performance. Specifically, we ask the following questions: How do different privatization methods reallocate control rights among the stakeholders of the firm? Why do city governments choose a particular privatization method? Do firms still obtain favorable treatment and soft budget constraint after privatization? Which methods result in more effective post-privatization restructuring and which better enhances performance? We collect comprehensive information about control rights reallocation, including details about distribution of eight distinctive decision rights among five parties before and after privatization. Our data shows that, while privatization in China has made substantial progress in reallocating control rights from the government to private owners, the degree of the remaining government influence on corporate decisions significantly varies across privatization methods. These methods include direct sales, either to insiders (through management buyouts, or MBOs hereafter) or to outsider private owners, public offerings, joint ventures, leasing, and employee shareholdings. Among these methods, MBOs transfer the most control rights to private owners and account for close to half of all privatization programs. Accordingly, the government provides the least support, in the forms of subsidies, bank financing, and protected entry, to these MBOs, while imposing the most hardened budget constraint. Our analysis further indicates that the decisions city governments make about how to privatize are critically determined by political and fiscal constraints, and a city government’s choice among privatization approaches profoundly affects the governance and performance of privatized firms. Specifically, when cities face less political opposition to labor downsizing and have stronger fiscal capacity, they tend to choose MBOs. Consistent with private owners’ enhanced incentives to make changes, MBOs are most effective in implementing restructuring measures, including a change of core management teams, strengthening of managerial incentives through compensation policies, establishing boards of directors, and introducing international accounting and independent auditing. Not surprisingly, the performance of MBO firms significantly improves after privatization by 4.4% in return on assets (ROA) and close to 6,000 RMB, or 750 USD, per employee per year. For other privatization methods, the government tends to retain its influence in key corporate decisions. These firms are less effective in restructuring and do not achieve statistically detectable improvement in performance. A common challenge in the privatization literature on performance comparison is selection bias, which arises because certain types of firms that are likely to have better future performance (e.g., due to stronger fundamentals or better government support) might be purposely chosen for MBOs. A distinctive advantage of our study is that our detailed data allows us to better handle the selection concern, by explicitly examining why firms are chosen for MBOs, as well as the mechanisms of performance improvements, which is perhaps the strongest guard against endogeneity. To rule out the selection bias even further, we conduct a number of additional analyses, including examining whether there is any preexisting trend in performance, fully accounting for city-level economic prospects by including city-year fixed effects, explicitly controlling for product market competition, and adopting an IV estimation using city characteristics (such as fiscal capacity and private sector development) as the instruments. Our Chinese survey contributes to the literature in a number of ways. First, it supports and significantly extends an important theme in the literature; that is, the types of owners have disparate effects on restructuring and performance; thus looking only at aggregate results without knowing why could be misleading (Frydman et al. 1999; Estrin et al. 2009). A well-known, but puzzling, result from other transition economies is that privatization to managers does not result in efficiency gains in transition economies. This result appears to be in contrast with our findings that MBOs are the most effective means of privatization in China. The difference lies in the extent to which managerial ownership is market based. Frydman et al. (1999) propose an explanation for the ineffectiveness of privatization to managers in CEE and CIS nations, that is, the two “special characteristics” of managerial ownership. Specifically, managers are selected under the old regime and they are offered to buy the shares at preferential prices, but with restrictive terms designed to favor existing employees. Chinese MBOs do not share these characteristics and have much more in common with managerial ownership in market economies. In this sense, Chinese MBOs constitute a nice counterfactual analysis for other transition economies (and vice versa). They confirm the conjectures in the literature regarding why managerial ownership does not work in CEE-CIS nations. Our paper extends beyond the question of the type of owners and illustrates how the market-based managerial ownership in China improves performance by aligning other economic forces, namely, the role of managers, product market competition, and hardened budget constraint, that have been found to be important in shaping privatization outcomes (e.g., Djankov and Murrel 2002). To our knowledge, no prior work has answered, in one study, these questions as comprehensively as we have. Moreover, our Chinese survey offers advantages in dealing with measurement and identification challenges. Our second contribution is that we explore two important aspects of privatization that the previous literature has not examined. Most notably, enabled by our detailed data, we shed new light on the privatization mechanism through the reallocation of control rights. As Jones and Mygind (1999) and Gupta (2005) point out, a common feature of privatization around the world is that transferring of control rights is incomplete, meaning that the state retains a significant ownership during privatization. Thus, our finding regarding the impact of remained state ownership and control is quite general. Another aspect is the role of political factors in shaping the design of privatization programs. Despite that theoretical work and anecdotes that all suggest a significant influence of political factors (e.g., Biais and Perotti 2002), few formal empirical papers study this important topic. Our paper joins a more recent effort (e.g., Dinc and Gupta 2011) in this regard. In the Chinese setting, political economy considerations, specifically a lack of fiscal resources and political opposition to unemployment, prevent the state from withdrawing control and adopting a more effective privatization method. Finally, our analysis extends earlier work on privatization in China and deepens our understanding of the Chinese economy. Previous work has documented the ineffectiveness of share issue privatization (SIP) (Sun and Tong 2003; Deng, Gan, and He 2010), a lack of a significant effect of privatization on performance (Jefferson and Su 2006), and the importance of reducing state ownership in privatized firms for performance improvement (Bai, Lu, and Tao 2009).3 Our data permit us to cover a wide spectrum of privatization methods and to extend beyond performance comparison by identifying the mechanisms of performance improvement (or a lack of it). Equally importantly, the decentralized privatization studied in this paper contributes to a growing literature on China’s regionally decentralized authoritarian regime, particularly on local governments’ decisions and career concerns (Maskin, Qian, and Xu 2000; Li and Zhou 2005; Jin, Qian, and Weingast 2005; Xu 2011; Jia, Kudamatsu, and Seim 2015; Persson and Zhuravskaya 2015). 1. Institutional Background of Decentralized Privatization in China In this section, we first discuss how decentralized privatization in China evolved alongside the country’s political and economic background in the 1990s. Then we introduce the different privatization methods adopted by the local governments. Finally, we discuss government considerations about MBOs, the most effective privatization method. 1.1 Political and economic background In the governance system of the Chinese economy, political and personnel decisions are highly centralized, and the central government appoints and assesses local government officials, whereas administrative and economic matters, including those of the SOEs, are mostly decentralized to local governments. Such a system is termed by some scholars as “regionally decentralized authoritarianism (RDA)”.4 Under the RDA regime, the control rights of SOEs, except for the very large ones, are assigned to municipal governments, giving them the residual claims to enterprise earnings (Granick 1990; Li 1997). This means that local SOEs were very important for city government officials, both as a source of fiscal revenue and as a contributor to growth in local gross domestic product (GDP), a critical criteria used by upper-level governments in personnel promotion decisions (Maskin, Qian, and Xu 2000; Xu 2011). Endowed with the “local” ownership of SOEs, China’s state sector reforms have been mostly driven by local experiments, sometimes even before the central government’s official mandates (Xu 2011). By early 1990s, the deteriorating performance of SOEs placed increasing pressure on the fiscal conditions of local governments. A few cities “quietly” initiated de facto privatization, without explicit approval from upper-level governments. One of the first local privatization attempts was made in Zhucheng, a city in Shandong province. In 1992, more than two-thirds of the SOEs experienced losses amounting to over 18 months of the city government’s fiscal revenue. The city government then sold many SOEs within its jurisdiction to the employees of these SOEs. Another example is Shunde in Guangdong, where the city government encountered a serious debt problem before it privatized most of its state and collective firms in 1992. When these experiments became publicly known, the central government did not prohibit the practice, which was interpreted as an implicit approval (Garnaut, Song, and Yao 2008). The continued deterioration of the state sector’s financial performance severely strained the country’s banking system.5 The central government gradually accepted privatization as a remedy for the country’s ailing SOEs, as indicated in a number of progressively market-based reform policies. In 1993, the 3rd Plenum of the 14th Communist Party Congress endorsed a principle of diversifying the ownership structure of state-owned firms. In 1995, the central government announced the famous policy of “retaining the large, releasing the small” (zhuada fangxiao). That is, the state was to keep a few hundred of the largest SOEs in strategic industries; for the remaining smaller local SOEs, which constituted the vast majority of SOEs, the stated intention was to let competitive forces make them more efficient. Finally, 15th Communist Party Congress (1997) further approved privatization, granting de jure ownership of local SOEs to local governments and authorizing the “owners,” mostly city governments, of SOEs to design and implement privatization on their own.6 Thus, China has no centrally designed nationwide privatization program, a fact that makes its brand of privatization distinctively different from that in the rest of the world. This wave of privatization ended in 2005, both because the vast majority of SOEs had been privatized by then and because of the publicized controversies over some of the privatization programs in 2004 and in 2005. Explicit statistics on the percentage of all SOEs privatized by 2005 is not available but, according to various reports of National Statistics Bureau (NSB), close to three-quarters of large and medium industrial SOEs were privatized.7 Consistent with the policy of “retaining the large, releasing the small,” our reading of available city-level statistics shows that about 85% of SOEs were privatized by 2005. If we use three-quarters as a conservative estimate of proportion of firms privatized, given that total industrial SOE assets at the end of 1999 was 7.6 trillion RMB, we estimate that the total privatized assets amounted to 5.7 trillion RMB, or roughly 700 billion USD, based on the exchange rate at the time. 1.2 Privatization methods Our data shows that the most popular method was direct sales (or open sales), to insiders or to outside private owners, which, respectively, accounted for 47% and 22% of all privatization programs. Other methods included public offering(1%), joint ventures(2%), leasing(8%), and employee shareholding(10%). These patterns are consistent with those in Garnaut, Song, and Yao (2008).8 Under direct sales, the firm was openly sold to insiders (through MBOs) or outside private owners through auctions or negotiations between the local government and the potential buyers. Although we later find that MBOs were the most effective in improving efficiency, it was the most controversial method, mainly because of its lack of transparency and public concern that state assets may have been sold too cheaply. Public offering refers to share issue privatization (SIP). Under the policy of “retaining the large, releasing the small,” large SOEs were privatized through SIP. By design, SIP did not involve transferring control rights, and only noncontrolling shares were sold in the public capital market. SIP accounted for a tiny proportion (1% according to our survey) in terms of the number of firms, and we estimate that SIPs accounted for around 10% of privatized assets.9 Nevertheless, SIPs have been the most-studied type of privatization in China simply because of the availability of data. Joint venture or merger involved privatization in which an SOE formed a joint venture or merged with a private domestic or foreign firm. Under leasing, the company was leased to the management, employees, outside private firms, or other SOEs. In most cases, it involved inside managers as the lessees, and the firms are often privatized later through MBOs. Employee shareholding converted the company into a limited liability company or cooperative. It was one of the most important gaizhi measures employed at the early stage of local experiments, both because the central government required that each privatization plan be approved by employees (other than corporate executives) and because shares were often offered as a compensation for removing employees’ “tenured” state-employment status. As our data verify, at later stages of gaizhi, managers often purchased the majority of employee shares, thereby qualifying the firms as MBOs. 1.3 Government considerations for MBOs To further understand the government’s considerations regarding MBOs, we choose 32 cities with the most MBOs and the least MBOs and reviewed all the publically available documents related to MBOs decisions. Across all the cities, the governments shared similar concerns and, as a result, they typically stipulated against MBOs in three types of firms: (1) firms with government-granted monopolistic permits to operate; (2) firms with government subsidies because of their responsibilities for social welfare; and (3) firms that obtained land or other resources whose value could not be easily assessed. As a result, small firms were often targeted to be “liberalized” and encouraged to be sold to managers. These patterns were perfectly consistent with what we later find in the data about post-privatization government support of MBOs (Section 3.2) and determinants of MBO choices (Section 4). 2. Nationwide Survey and Sample 2.1 Nationwide survey Our large-scale nationwide survey was conducted in 2006. The sampling procedure involved two steps. We started with the 2004 National Bureau of Statistics (NBS) census, which contained all industrial firms with sales above 5 million RMB as the population and drew a random sample of 11,000 firms stratified by region, industry, size, and ownership type. Given that only 20% of firms in the 2004 population were SOEs and our intention was to study privatization, we supplemented the main survey sample with an additional random sample of 5,500 from the 1998 NBS database, again stratified based on region, industry, and size. We chose to use the 1998 NBS data because 1998 was the first year the database was available, and large-scale privatization in China started in the late 1990s. Thus, the 1998 population maximized our chance of including SOEs not yet privatized. In total, we have 16,500 firms in the survey. We designed the questionnaires through an “iterated” process. We started with a pilot survey of 720 firms in nine cities, including Beijing, Laizhou in Shandong province, Taizhou and Changxing in Zhejiang province, Changchun and Jilin in Jilin province, Shijiazhuang, Pingshan, and Tangshan in Hebei province. It was conducted through both on-site interviews and telephone interviews. This pilot survey improves our survey design and later guides our empirical analysis. For example, because of the controversy surrounding MBOs, many of the MBO firms “disguised” themselves by reporting themselves as other less controversial methods, such as employee shareholdings. Thus, in our empirical analysis, we verify each firm’s self-reported privatization methods with its answers to questions on changes in ownership. In soliciting certain sensitive financial variables, instead of asking for the information directly, we experimented with using multiple-choice questions (of percentage intervals), and the response rate substantially increased. The main survey was conducted through telephone interviews. We hired a professional survey company that had a close relationship with the NBS and had previously helped it conduct its own surveys. We spent a week training the survey company’s staff to understand each question. Throughout the survey, we closely worked with the staff and carefully supervised the process. The chief executives of the firms (or their representatives), the chief accountants, or the heads of human resources answered the questions. To facilitate a difference-in-differences analysis, we prepared two sets of questionnaires: one for privatized firms (the “treatment” group) and one for all other firms (including the “control” group). The survey asked every firm whether it was privatized and accordingly used the appropriate questionnaire. The two sets of questionnaires were identical, except that for privatized firms (1) we asked questions related to privatization, for example, the year in which the firm was privatized and the privatization method; and (2) for questions on ownership and control, we asked the firms to provide information on both the pre- and post-privatization periods. Appendix 1 contains the survey questions relevant to this study. We obtained 3,132 responses, yielding a response rate of 19%. Our survey sample contains 899 privatized firms, 475 nonprivatized SOEs and collectively owned enterprises (nonprivatized SOEs hereafter), and 1,758 de novo private firms. In our survey, we do not notice any systematic selection bias of firms that responded to our survey. Indeed, as reported in Table 1, our survey sample matches the distribution of the population reasonably well in terms of both region and industry. The size distribution of our sample is skewed toward larger firms because we purposely oversampled SOE firms, which tend to be larger for this study. Figure 1A further shows the regional distribution of the privatization sample is roughly in line with the presence of SOEs in the country. Figure 1B reports the staggered nature of privatization by region (Appendix 2 shows the breakdown by province). Figure 1 View largeDownload slide Regional distribution of privatized firms in the survey Figure 1 View largeDownload slide Regional distribution of privatized firms in the survey Table 1 Sample distribution of ownership, size, location, and industry Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% This table compares the distribution of our survey sample with that of the population by size, location, and industry. North China includes: Beijing, Tianjin, and Hebei; North-east: Heilongjiang, Jilin, and Liaoning; North-west: Xinjiang, Qinghai, Ningxia, Gansu, Shaanxi, and Innermongolia; North-central: Shanxi, Henan, and Shandong; South-west: Xizang, Yunnan, Guizhou, Sichuan, and Chongqing; East: Shanghai, Jiangsu, and Zhejiang; South: Guangxi, Guangdong, Fujian, and Hainan; and South-central: Hubei, Hunan, Jiangxi, and Anhui. Table 1 Sample distribution of ownership, size, location, and industry Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% This table compares the distribution of our survey sample with that of the population by size, location, and industry. North China includes: Beijing, Tianjin, and Hebei; North-east: Heilongjiang, Jilin, and Liaoning; North-west: Xinjiang, Qinghai, Ningxia, Gansu, Shaanxi, and Innermongolia; North-central: Shanxi, Henan, and Shandong; South-west: Xizang, Yunnan, Guizhou, Sichuan, and Chongqing; East: Shanghai, Jiangsu, and Zhejiang; South: Guangxi, Guangdong, Fujian, and Hainan; and South-central: Hubei, Hunan, Jiangxi, and Anhui. 2.2 Data We obtain the financial information of surveyed firms from the NSB database, which is equivalent to Compustat for U.S.-listed firms. NSB data are available from 1998 to 2007. While it is the most comprehensive data about Chinese firms, some scholars have questioned its data quality (e.g., Nie, Jiang, and Yang, 2012). Appendix 3 examines the NSB data in detail and demonstrated that their weakness does not significantly affect our results. To ensure all privatized firms have at least one year of performance information prior to privatization, we drop 168 firms that were privatized prior to 1999. We then exclude firms without valid financial information. Given the staggered nature of privatization, our final sample for regression analyses is an unbalanced panel of 717 privatized firms, 460 SOEs that have not been privatized, and 1,685 de novo private firms for the period of 1998–2007. In our analysis of the role of government incentives in privatization decisions, we use the China City Statistical Yearbook to obtain city-level (at and above the prefecture level) fiscal and regional economic variables from 1997 to 2007. We note that, while the data may seem old, they are suitable to study the largest wave of privatization in China (and worldwide), for two reasons. First, we conducted the survey in 2006, while this wave of privatization ended in 2005 (see discussions in Section 1.1). Second, the survey data can be merged with 10 years of NSB data during 1998–2007 and doing so allows us to study performance before and after privatization. It is well known among scholars studying China that the quality of data available to researchers is low in 2008 and in 2009, and that, due to tightened control of data, it is almost impossible to obtain the data after 2009. Thus, it is a nice coincidence that privatization occurred before the end of 2005 and quality NSB financial data are available until 2007, a fact enabling us to cover this historical episode well and to the best extent. 2.3 Preliminary observations from our sample Table 2 reports the summary statistics of the main variables used in our empirical analysis. In panel A of Table 2, we report the basic facts about privatization in China. Between 2000 and 2005, the number of privatizations increases steadily. Direct sales to insiders (MBOs) are, by far, the most widely used method, accounting for 47% of all privatized firms. The next is direct sales to outsiders, accounting for 22% of the firms. Thus, direct sales in total account for close to 70% of privatization programs in China. Other privatization methods include public offerings (1%), joint ventures (2%), leasing (8%), and employee shareholding (10%). Table 2 Basic facts and summary statistics A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** In panel A3, differences between the MBO firms and other methods and between direct sales to outsiders and other methods are tested. Profits are defined as earnings before interest, tax, and depreciation. Significance levels are all based on two-tailed tests of differences. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 2 Basic facts and summary statistics A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** In panel A3, differences between the MBO firms and other methods and between direct sales to outsiders and other methods are tested. Profits are defined as earnings before interest, tax, and depreciation. Significance levels are all based on two-tailed tests of differences. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. The ownership structure of Chinese privatized firms is highly concentrated. The largest shareholders, on average, hold 60% of the shares and the second- and third-largest shareholders hold 26% of shares. MBOs have the lowest ownership concentration, with the largest shareholders holding 37% of the shares, whereas the largest shareholder of the firms sold to outsiders has 64% ownership, on average. For firms privatized by other methods, the largest shareholders, on average, hold 91% of the shares.10 Panel B is a summary of the financial variables. We use two measures of operating performance: operating profits (earnings before interest, tax, depreciation, and amortization, or EBITDA) over assets, and operating profits over the number of employees. Panel B1 compares privatized, nonprivatized, and de novo nonstate (private) firms. Compared with nonprivatized SOEs, privatized firms tend to be larger and generally exhibit greater operating efficiency. Later, we will show that this is due to post-privatization performance. Compared with de novo private firms, privatized SOEs tend to be larger and less profitable. Panel B2 of Table 2 compares the financial variables before and after privatization. Assets and sales generally increased after privatization. Firms tend to become less leveraged after privatization, consistent with a hardened budget constraint. While there is generally an improvement in performance (all at the 1% level, except for the mean of Profits/#employee), performance gain appears to be larger for MBOs, consistent with our later findings that MBOs drive performance gains. 2.4 Financial aspects of privatization We will now discuss the financial aspects of privatization, including the issuance method, payment arrangement and sources of funds for top managers. Other than SIPs, which cover large companies in strategic industries, the transfer of ownership is through secondary offerings of existing shares, consistent with the government’s stated intention of transferring of ownership and of “letting go” of these companies. As reported in Table 3, in 77% of privatization cases, the government receives a lump sum payment, as opposed to multiyear installments. Reflecting a greater transfer of ownership, MBOs are significantly more likely to be paid with lump sum payments (80%), whereas leasing is least likely to use this arrangement. If multiple installments are used, the first payment, on average, accounts for one third of the total proceeds and it takes about 5 years to pay the full amount. Table 3 Financial aspects of privatization A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% In panel A, significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5% and 10% level is indicated by ***, ** and *, respectively. In panel C, for each source of funds, firms are asked to specify the percentage of funding from this source. The possible answers are: 0%, 1%-20%, 21%-40%, 41%-70% and 71%-100%. To estimate the monetary share of each of the financing source, we assume that the median of the range is the actual percentage. Table 3 Financial aspects of privatization A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% In panel A, significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5% and 10% level is indicated by ***, ** and *, respectively. In panel C, for each source of funds, firms are asked to specify the percentage of funding from this source. The possible answers are: 0%, 1%-20%, 21%-40%, 41%-70% and 71%-100%. To estimate the monetary share of each of the financing source, we assume that the median of the range is the actual percentage. Personal savings are predominately the most important source of funds by the top managers, used in 99% of firms. Ninety-five percent of firms report that personal savings account for at least 70% of financing (panels A and B), and we further estimate that they contribute to 96% of all privatization payments (panel C). Other sources of financing include borrowings from friends and relatives, bank loans, and future salaries by 8%, 5%, and 6% of firms, respectively and each account for 1% to 2% of total payments. 3. Mechanisms of Efficiency Gain The essence of ownership structure is its allocation of control rights among the firms’ stakeholders (Grossman and Hart 1986; Hart and Moore 1990). This section investigates reallocation of control rights as the mechanism of performance gain, and the resultant government support and freedom to restructure. 3.1 Reallocation of control rights and performance We find that the government retains, on average, 20% ownership of the privatized firms. Although this figure is much lower than that for share issue privatization, in which the government retains more than half of the ownership, 20% is still a figure significant enough for the state to exert influence. Reflecting the concept of property rights as a bundle of rights, we focus on a set of eight decision rights, including the appointment of senior managers, investment, hiring and laying off of employees, salary and bonus, distribution of profits, production and marketing, financing, and use of funds. We ask how these control rights are allocated, before privatization and after privatization, among five parties, including the government, the party committee at the firm, board of directors, general manager, workers representative committee, board of supervisors, and shareholder committee in making the above-mentioned key corporate decisions. The firms rank, for each of the corporate decisions, the importance of each decision maker on a five-point scale (0 $$=$$negligibly unimportant, 5 $$=$$indispensably important). As shown in Figure 2 and Table 4, the most prominent change in control rights is the reduction of government influence. For nonprivatized SOEs and pre-privatization SOEs, local governments exercise fairly strong control over these firms’ major decisions, with average scores of 2.3 and 1.8, respectively, and the government’s control rights are particularly strong in the appointment of top management, scoring 3 and 2.4 (panel A of Table 4). By contrast, the government has no control power over decisions within de novo private firms. After privatization, both the overall government control and its control in personnel drop substantially, from 1.8 to 0.4 and from 2.4 to 0.6, respectively. Moreover, the government control decreases the most for MBOs, with the average score dropping from 1.8 to 0.1. Direct sales to outsiders come the second, with average government control decreasing from 1.9 to 0.4. Figure 2 View largeDownload slide Reallocation of control rights before and after privatization Figure 2 View largeDownload slide Reallocation of control rights before and after privatization Table 4 Privatization and change of control rights Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 This table reports allocation of control rights in Chinese firms. The importance of each decision maker is given a score from 0 to 5, where 0 means negligibly unimportant and 5 indispensably important. Average and median scores across firms are reported. Significance levels in Columns (4), (6) and (8) are based on two-tailed tests of differences before and after privatization. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively, and na refers to not applicable. Table 4 Privatization and change of control rights Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 This table reports allocation of control rights in Chinese firms. The importance of each decision maker is given a score from 0 to 5, where 0 means negligibly unimportant and 5 indispensably important. Average and median scores across firms are reported. Significance levels in Columns (4), (6) and (8) are based on two-tailed tests of differences before and after privatization. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively, and na refers to not applicable. A unique feature of corporate governance in China is that almost all firms in China have a committee of the Chinese Communist Party. As shown in panels A and B of Table 4, the influence of party committees is similar to that of the government. After privatization, the party committees’ control generally decreases less than the government’s control. Given that the government may influence corporate decisions through both its direct control rights and its intervention via firm-level party committees, we use the max of these two as the score for overall state influence (State influence score). Despite a drop in the score from 2.8 to 1.4 after privatization, state influence is still quite important in a significant proportion of firms, with 39% of firms having a score above 2 (somewhat important) and 15% above 3 (moderately important). In the following analysis, we consider firms with State influence score above 2 as under significant state influence in corporate decisions. Across privatization methods, MBO firms have the lowest level of state control. Only 1% of MBO firms have government ownership above 20%, the mean, significantly lower than the sample average of 50% (Table 5). The state is also much less likely to intervene in MBOs’ major decision-making (16% vs. 59% sample mean). Direct sales to outsiders are under substantially more state influence than MBOs, but, compared to privatization methods, they receive less state intervention, though the difference is only significant for corporate decision-making, not for state ownership. Table 5 State control in privatized firms A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% This table reports the percentage of firms that are under strong state influence post-privatization by privatization method. State-influence score is defined as the max of the importance of local government and that of party committee in corporate decision making based on a five-point scale (0=negligibly unimportant, 5=indispensably important). Panel B uses principal component analysis (PCA) to form additional variables of state control. The source of state influence is from government or party communist. PCA State Control is defined as 1 if either the first component of government influence or the first component of party influence is above the mean. Significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 5 State control in privatized firms A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% This table reports the percentage of firms that are under strong state influence post-privatization by privatization method. State-influence score is defined as the max of the importance of local government and that of party committee in corporate decision making based on a five-point scale (0=negligibly unimportant, 5=indispensably important). Panel B uses principal component analysis (PCA) to form additional variables of state control. The source of state influence is from government or party communist. PCA State Control is defined as 1 if either the first component of government influence or the first component of party influence is above the mean. Significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Given that corporate decisions are multidimensional, we further examine government and party influences using principal component analysis (PCA). PCA is effective in shrinking dimensionality: the first principal component accounts for 90% and 75% of the government and party influences respectively, whereas the second component accounts for only 4% and 6%, respectively. Thus we report, in panel B of Table 5, the first components and PCA state control, defined as either the first component of government influence or the first component of party influence is above the mean. Consistent with panel A of Table 5, PCA state control is significantly lower in MBOs. Other notable changes in control rights include the increased decision power of the board of directors and shareholder meetings, suggesting a general trend of professionalization of management in privatized firms. This change is most prominent among MBOs. 3.1.1 State control and post-privatization performance This subsection further investigates the impact of state control on post-privatization performance, by estimating the following model on the sample of all privatized firms: \begin{equation} \textit{Performance}_{it} = \alpha_{i} + \beta_{t }+ \gamma \textit{Post}_{it} +\lambda \textit{State Control}_{i} \times \textit{Post}_{it} + \delta X_{it} + \varepsilon_{it}, \end{equation} (1) where Performance$$_{it}$$ is measured by both ROA and earnings per employee. Post$$_{it}$$ is a dummy variable indicating years after privatization (it is set to zero for those SOEs that have never been privatized). State control is one of the three binary variables: state ownership above 20%; State influence score above 2; and PCA state control, defined in the same way as in Table 5. X$$_{it}$$ are firm control variables, including size (measured as log of assets) and leverage (debt over assets). $$\alpha_{i} $$ is a firm fixed effect that controls for time-invariant firm characteristics. $$\beta_{t}$$ is a year fixed effect. Coefficient $$\gamma $$ is the difference-in-differences estimate of the effect of state control on post-privatization performance. Linking detailed measures of government control rights to performance improves upon the existing literature which typically assigns a linear relationship between ownership and performance. Our analysis is similar in spirit to (López-de-Silanes, 1997), who finds, in privatization in Mexico, that transferring of controlling share packages is associated with a higher price premium, an ex ante measure of future performance. Table 6 demonstrates that state control significantly hinders performance of privatized firms. In Columns (1) and (2) of Table 6, higher state ownership is associated with significantly worse post-privatization performance, for both operating efficiency measures (at the 1% levels). In Columns (3) - (6), both the measure based on State influence score above 2 and PCA State control are associated with significantly lower operating efficiency. The results are all economically significant. Take the example of the point estimates of State influence score above 2 (Columns (3) and (4)). They imply that, all else equal, state control in decision making reduces ROA by 6% and earnings per employee by 6,252 RMB (close to 800 USD) per employee per year. Both are substantial, especially considering the sample mean is 12.8% for ROA and 15.9K RMB for earnings per employee. Table 6 State influence and performance Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 This table presents the effect of state control on post-privatization performance like in Equation (1). It is based on the sample of all privatized firms during 1998 to 2007. Table 5 defines variables related to state control. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 6 State influence and performance Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 This table presents the effect of state control on post-privatization performance like in Equation (1). It is based on the sample of all privatized firms during 1998 to 2007. Table 5 defines variables related to state control. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. 3.2 Government support While ownership and decision rights are perhaps the most straightforward measures of government influence, there may be a tangled web of relation between the firm and the government. Specifically, it is possible that government exerts influence through other channels, such as connection of the manager to the party, implicit or explicit subsidies, and regulatory barriers to entry. In our survey, we design questions that allow us to further explore these aspects of state influence. Given the dramatic control change via the MBO route, we mainly focus on the comparison between MBOs and other privatization methods. This analysis will also help understand our later results on MBO performance. Panel A of Table 7 displays the firm’s political connections along three dimensions, namely, whether top officials are appointed by the government, whether the firm has are government officials on the board, whether the top manager is a former government official. It turns out that the strongest form of political connection in China’s privatized firms is through personnel appointment: in 23% firms, the chairman or top manager is appointed by the government. Such connection, however, is much weaker in MBOs involving only 0.3% of the firms and the difference is significant at the 1% level. Political connection in the form of government officials on the board or being the top manager is not common and is in only 4% and 2% of privatized firms respectively. The numbers are even lower among MBOs, involving, respectively, 0.3% and 1% of firms, and the difference is significant at the 1% and 5% levels. Table 7 Comparison of political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatized firms All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 This table presents the comparison of post-privatization political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatization methods. Panel E is based on answers to our survey question: “how many competitors does your firm have?” The possible answers are: no, few, some and many competitors. We categorize the firm as a monopoly if it reports no competitor. It is defined to have market power if it has no or few competitors. It is considered to be in a competitive market if it has some or many competitors. Significance levels are based on two-tailed tests of differences between MBO and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 7 Comparison of political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatized firms All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 This table presents the comparison of post-privatization political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatization methods. Panel E is based on answers to our survey question: “how many competitors does your firm have?” The possible answers are: no, few, some and many competitors. We categorize the firm as a monopoly if it reports no competitor. It is defined to have market power if it has no or few competitors. It is considered to be in a competitive market if it has some or many competitors. Significance levels are based on two-tailed tests of differences between MBO and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Panel B of Table 7 shows that MBOs receive less government subsidies. Land is the most important government subsidy. MBOs are significantly less likely to obtain land subsidy, 59% versus 67% (a significant difference at the 1% level). The composition of land subsidy is also telling: MBOs are less likely to obtain direct allocation of land (19% versus 31%, significant at 1% level), which represents a large subsidy, whereas they are slightly more likely to purchase land at substantially subsidized prices (40% versus 36%). Government funded R&D projects are not common, involving 3% of the firms. The number is even lower for MBOs, 1%, and the difference is significant at the 1% level. Panel C of Table 7 demonstrates that MBOs receive less government support in financing. While MBOs have a similar likelihood to have bank loans, their loan applications are significantly more likely to be rejected, 26% versus 22% (a significant difference at the 10% level). When asked about the reasons for loan rejection, MBOs are more likely quote bank credit rationing (4% versus 3%) – state-owned banks typically have quarterly or annual limits imposed by their regulatory agents – and a lack of relationship with the government (4% versus 3%). The differences are significant, respectively, at the 5% and 10% level. Finally, there is no difference in the chance of obtaining government loan guarantees between the two groups of firms. Panel D of Table 7 examines soft budget constraints in privatized firms. It should be noted that soft budget is not easy to measure, because the empirical measure has to meet two criteria. One is that it has to capture the expectation of future bailout; the other is that the expectation is contingent on financial distress. Neither is available in standard company financial statements. As noted by Djankov and Murrell (2002), a survey method provides measures that come closest to theoretically prescribed ones. In our survey, we ask about a number of expected supports in case of financial distress, including tax reduction, subsidies, capital injection, and subsidized loans. The data shows that Chinese privatization is very effective in hardening soft budget constraints: each individual form of soft budget involves less than 1% of the firms, and the proportion of firms with any one form of the soft budget is 0.6%. MBOs are even less likely to have soft budget in terms of all forms of support, except for subsidized loans, arguably the weakest form of support. Panel E of Table 7 reports government support in the form of protected entry, based on the question “How many competitors does your firm have?” The possible answers are none, few, some, and many. We categorize the firm as in a competitive market if there are some or many competitors. The vast majority of firms (75%) are in competitive markets. MBOs are even more likely to be in competitive markets, 84%, and the difference is significant at 1%. 14% firms are monopolies with no competitors, whereas significantly less MBOs, a mere 2%, are monopolies (at the 1% level). While most SOE monopolies in China arise from protected entry, it is theoretically possible that the firm has developed or purchased advanced technology. We find that this is not true: only 4% of monopoly firms have patents, much lower than other firms, 30%, and the difference is significant at the 1% level. We further check the market structure of industries that are often perceived as having protected entry, including energy, utilities, car, and pharmaceuticals. It turns out that only utilities seem to possess monopolistic power: an average of 67% firms report themselves as an monopoly and 13% report that the market is competitive. There is only one firm in oil and gas; although it is a monopoly, there is not a big enough sample to make a reliable inference. Taken together, our analysis demonstrates that, after privatization, the government substantially reduce its support and subsidiaries to all firms and particularly so for MBOs. This, however, is not surprising. It is consistent with the guiding rule of “retaining the large, letting go of the small,” where the small ones, which is the vast majority, were generally in competitive sectors. Moreover, given that the government keeps the least ownership and control in MBOs, it is economically rational to provide even less support. 3.3 Post-privatization restructuring and professionalism We ask about four restructuring measures. The first restructuring measure is whether the firm changed its core management team–-the introduction of new human capital into management is shown to be important in improving efficiency in other privatization settings (e.g., Barberis et al. 1996; Lópezde-Silanes; 1997, who emphasizes CEO changes). The second is whether the firm incentivizes its executives through increased performance-based pay. Regarding corporate governance, we ask whether the firm established a board of directors and whether it adopted international accounting standards. Panel A of Table 8 reports, by privatization methods, the proportion of firms implementing each of the restructuring measures. MBO firms are significantly more likely to use performance-based bonuses (54% versus 47%), to establish a board of directors (84% versus 76%), and to adopt international accounting standards and professional independent auditing (11% versus 8%), all significantly at the 1% or 5% level. MBO firms are not likely to have performance-based share compensation for their executives, which is not surprising, since managers are now owners. Compared with the whole sample, direct sales to outsiders are less likely to establish a board (67% versus 76%) but are more likely to adopt performance-based share compensation (15% versus 7%), both differences are significant at the 1% level. Table 8 Post-privatization restructuring and professionalization A. Post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors Direct sales to insiders (MBO) 64% 54%*** 8% 11%** 84%*** Direct sales to outsiders 61% 51% 15%*** 7% 67%*** Other methods 60% 35%*** 2% 5% 71% All privatized SOEs 62% 47% 7% 8% 76% B. Logit and Tobit regression of post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors (1) (2) (3) (4) (5) Lag of performance –0.073** 0.274* –0.264*** 0.192 0.244*** (0.036) (0.140) (0.080) (0.065) (0.046) Log (sales) –0.223 –0.020* 0.45 –3.570*** –0.069 (0.343) (0.011) (0.773) (0.992) (0.408) Leverage –0.631** 0.057 0.422** –0.522 –0.501*** (0.302) (0.099) (0.187) (0.575) (0.182) Direct sales to outsiders –0.166 0.140*** 1.793*** –0.094 –0.055 (0.171) (0.053) (0.423) (0.369) (0.203) MBO 0.388** 0.202*** –1.253*** 0.991*** 0.782*** (0.151) (0.044) (0.272) (0.318) (0.189) Industry fixed effects Yes Yes Yes Yes Yes Observations 606 553 606 606 606 A. Post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors Direct sales to insiders (MBO) 64% 54%*** 8% 11%** 84%*** Direct sales to outsiders 61% 51% 15%*** 7% 67%*** Other methods 60% 35%*** 2% 5% 71% All privatized SOEs 62% 47% 7% 8% 76% B. Logit and Tobit regression of post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors (1) (2) (3) (4) (5) Lag of performance –0.073** 0.274* –0.264*** 0.192 0.244*** (0.036) (0.140) (0.080) (0.065) (0.046) Log (sales) –0.223 –0.020* 0.45 –3.570*** –0.069 (0.343) (0.011) (0.773) (0.992) (0.408) Leverage –0.631** 0.057 0.422** –0.522 –0.501*** (0.302) (0.099) (0.187) (0.575) (0.182) Direct sales to outsiders –0.166 0.140*** 1.793*** –0.094 –0.055 (0.171) (0.053) (0.423) (0.369) (0.203) MBO 0.388** 0.202*** –1.253*** 0.991*** 0.782*** (0.151) (0.044) (0.272) (0.318) (0.189) Industry fixed effects Yes Yes Yes Yes Yes Observations 606 553 606 606 606 Panel A presents, by privatization method, the percentage of firms that have undertaken restructuring. Significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Panel B presents the logit model (Columns (1) and (3)-(5)) or the Tobit model (Column (2)) of restructuring measures after privatization. The financial variables are the three-year average after privatization. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 8 Post-privatization restructuring and professionalization A. Post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors Direct sales to insiders (MBO) 64% 54%*** 8% 11%** 84%*** Direct sales to outsiders 61% 51% 15%*** 7% 67%*** Other methods 60% 35%*** 2% 5% 71% All privatized SOEs 62% 47% 7% 8% 76% B. Logit and Tobit regression of post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors (1) (2) (3) (4) (5) Lag of performance –0.073** 0.274* –0.264*** 0.192 0.244*** (0.036) (0.140) (0.080) (0.065) (0.046) Log (sales) –0.223 –0.020* 0.45 –3.570*** –0.069 (0.343) (0.011) (0.773) (0.992) (0.408) Leverage –0.631** 0.057 0.422** –0.522 –0.501*** (0.302) (0.099) (0.187) (0.575) (0.182) Direct sales to outsiders –0.166 0.140*** 1.793*** –0.094 –0.055 (0.171) (0.053) (0.423) (0.369) (0.203) MBO 0.388** 0.202*** –1.253*** 0.991*** 0.782*** (0.151) (0.044) (0.272) (0.318) (0.189) Industry fixed effects Yes Yes Yes Yes Yes Observations 606 553 606 606 606 A. Post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors Direct sales to insiders (MBO) 64% 54%*** 8% 11%** 84%*** Direct sales to outsiders 61% 51% 15%*** 7% 67%*** Other methods 60% 35%*** 2% 5% 71% All privatized SOEs 62% 47% 7% 8% 76% B. Logit and Tobit regression of post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors (1) (2) (3) (4) (5) Lag of performance –0.073** 0.274* –0.264*** 0.192 0.244*** (0.036) (0.140) (0.080) (0.065) (0.046) Log (sales) –0.223 –0.020* 0.45 –3.570*** –0.069 (0.343) (0.011) (0.773) (0.992) (0.408) Leverage –0.631** 0.057 0.422** –0.522 –0.501*** (0.302) (0.099) (0.187) (0.575) (0.182) Direct sales to outsiders –0.166 0.140*** 1.793*** –0.094 –0.055 (0.171) (0.053) (0.423) (0.369) (0.203) MBO 0.388** 0.202*** –1.253*** 0.991*** 0.782*** (0.151) (0.044) (0.272) (0.318) (0.189) Industry fixed effects Yes Yes Yes Yes Yes Observations 606 553 606 606 606 Panel A presents, by privatization method, the percentage of firms that have undertaken restructuring. Significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Panel B presents the logit model (Columns (1) and (3)-(5)) or the Tobit model (Column (2)) of restructuring measures after privatization. The financial variables are the three-year average after privatization. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. The logit model in panel B of Table 8 further confirms the effect of privatization methods on restructuring. Moreover, the economic magnitudes are substantial: the odds ratios for MBO firms to restructure their management teams, to adopt international accounting, and to establish a board are, respectively, 1.5, 2.7, and 2.2 times that of non-MBOs.11 Regarding the compensation policy, the cash compensation of MBOs has 20% more in bonuses, whereas the odds ratio of MBOs to adopt share compensation is 71% below that of non-MBOs, consistent with owner-manager alignment. Selling to outsiders is more likely to restructure the compensation policy. It is significantly more likely to use performance-based share compensation (at the 1% level) and the odds ratio is 6 times of other methods; they also use significantly more bonuses in cash compensation (at the 1% level) and the incremental effect is 14%, lower than MBOs. But this privatization method is not more likely to undertake other restructuring measures. These findings are consistent with the fact that MBOs entail the greatest transfer of control rights from the state to the firm and that thus have the most freedom in implementing performance-enhancing restructuring. 4. Political Constraints, Governments’ Incentives, and MBO Choices As we have shown in the previous section, MBOs are most effective in transferring the control rights to the private owners and in promoting post-privatization restructuring. This finding inevitably leads us to ask why many city governments choose not to privatize via the MBO approach. This section examines the political and economic constraints that the local governments face at the time of privatization. Two well-documented factors result in poor SOE performance. One is surplus workers: according to Dong and Putterman (2003), surplus workers ranged from 23.5% to 44% of the SOE labor force during 1993$$-$$1996 and a World Bank survey in 1994 indicates that one-third of firms reported a labor-redundancy rate exceeding 20% (Bai, Lu, and Tao 2006). The other relates to various policy burdens, such as pension, social welfare, and perhaps uncompensated uses of corporate resources by the local governments. As we have shown, MBOs represent a commitment from the local government to relinquish its control. Several factors could affect its incentive to make such a commitment. The first is local political opposition to layoffs. Empirically, we measure it as the share of SOE employment in total urban employment. The implicit unemployment problem is most severe in cities dominated by SOEs, resulting in stronger political opposition to layoffs. Moreover, a greater share of SOE employment indicates slower development of the de novo private sector, which makes finding new jobs harder for the laid-off workers and increases political opposition to layoffs. Finally, the share of SOEs may be negatively related to MBOs for a subtle reason. Development of the de novo private sector is affected by the local governments’ attitudes toward private ownership. In the early days of reform, some local governments provided ad hoc local protections and promises to private firms when the constitution did not protect private ownership, whereas others discouraged the growth of private firms. To the extent that MBOs represent a more “thorough” privatization, city governments that are more pro-private ownership are more likely to implement MBOs. The second factor is the ability of local governments to bear the costs of layoffs and social responsibilities. One measure of such ability is the government’s fiscal resources. The more fiscal resources available, the greater the government’s ability is to pay for the layoffs and/or redeployment of laid-off workers. Moreover, the impact of greater government fiscal capacity is likely to be nonlinear: it is more important in regions in which unemployment is a bigger concern, because larger fiscal capacity allows the government to provide better support for redeployment of laid-off workers in MBOs. Fiscal resources also reduce local governments’ reliance on SOEs to achieve their social and political goals, as well as for uncompensated use of resources. The political pressure against layoffs can be exerted through the Employee Representative Congress. At the early stage of privatization, most SOEs had an Employee Representative Congress, which may have influenced the redeployment of employees and the choice of privatization methods.12, 13 As a result, employment was an important negotiation point between the government and the potential buyers. The city government sometimes provided a monetary subsidy for each additional worker the firm would keep. We also include, as additional measures of government incentives, policy subsidies, namely allocation of land (for free or at below-market price) and the city government’s loan guarantees, two of the most probable types of government support. As we discuss earlier, SOEs obtaining significant government resources are explicitly discouraged to use the MBO method to privatize. Moreover, to the extent that these policy subsidies reflect pre-existing “ties” between the firm and the government, the government may have more difficulty committing to a more complete withdrawal of influence. We estimate the following logit model to quantify the influence of government incentives on the choice of MBOs: \begin{gather} \textit{Prob(MBO} = \textit{1)} = \Lambda \textit{(Y), where}\notag\\ Y= a + b \textit{Government Incentives} + \textit{cX} + \textit{Industry Dummies} +\varepsilon , \end{gather} (2) and $$\Lambda $$(.) is the logistic cumulative distribution function. Government Incentives include government fiscal resources as measured by government revenue as a percent of GDP, the share of SOE employment, government allocation of land, and government guarantee of loans. To capture the impact of fiscal resources in cities where unemployment is a greater concern, we also include an interaction term between fiscal resources and a dummy variable indicating a high share of SOE employment (defined as % of SOE employment greater than the sample median). All Government Incentives variables are measured in the year prior to privatization. $$X$$ contains three sets of control variables: (1) city-level variables, including GDP per capita and population growth; (2) firm-level variables, including profitability (EBITDA over sales), size (log of assets), and leverage–-again all measured in the year prior to privatization; and (3) privatization-year dummies. Panel A of Table 9 presents the summary statistics. Indicative of our later findings, MBOs are significantly more popular among cities with better fiscal balance, or with a lower share of SOE output. Moreover, MBO firms are statistically less likely to have obtained land from the government, though the difference is not economically substantial. Table 9 Government incentives and choices of MBO method A. Summary statistics of government incentives and city-level variables All privatizatized SOEs MBO Government incentives Fiscal resources Mean 0.67 0.70*** Median 0.71 0.71 Share of SOE employment Mean 0.25 0.24 Median 0.17 0.16* % with government land subsidy Mean 69 62$$^{***}$$ % with government guarantee of loans Mean 7 7 City-level controls Log (GDP per capita) Mean 9.72 9.77* Median 9.71 9.78* Population growth Mean 0.03 0.04* Median 0.01 0.01 A. Summary statistics of government incentives and city-level variables All privatizatized SOEs MBO Government incentives Fiscal resources Mean 0.67 0.70*** Median 0.71 0.71 Share of SOE employment Mean 0.25 0.24 Median 0.17 0.16* % with government land subsidy Mean 69 62$$^{***}$$ % with government guarantee of loans Mean 7 7 City-level controls Log (GDP per capita) Mean 9.72 9.77* Median 9.71 9.78* Population growth Mean 0.03 0.04* Median 0.01 0.01 B. Logit regression of MBO choices Dependent variable: MBO (1) (2) Government incentives Fiscal resources –0.979 –1.173 (0.230) (0.159) Share of SOE employment –0.748** –0.754** (0.024) (0.026) Fiscal resources * High share of SOE employment 2.660*** 2.372*** (0.002) (0.008) Government land subsidy –0.142*** –0.142*** (0.000) (0.001) Government guarantee of loans 0.053 0.078 (0.464) (0.314) City-level controls Log (GDP per capita) –0.021 –0.022 (0.568) (0.554) Population growth 0.216 0.233 (0.242) (0.241) Firm-level controls Log (sales) –0.021* (0.054) Performance –0.023 (0.874) Leverage –0.103 (0.330) Observations 708 678 R-squared 0.199 0.207 B. Logit regression of MBO choices Dependent variable: MBO (1) (2) Government incentives Fiscal resources –0.979 –1.173 (0.230) (0.159) Share of SOE employment –0.748** –0.754** (0.024) (0.026) Fiscal resources * High share of SOE employment 2.660*** 2.372*** (0.002) (0.008) Government land subsidy –0.142*** –0.142*** (0.000) (0.001) Government guarantee of loans 0.053 0.078 (0.464) (0.314) City-level controls Log (GDP per capita) –0.021 –0.022 (0.568) (0.554) Population growth 0.216 0.233 (0.242) (0.241) Firm-level controls Log (sales) –0.021* (0.054) Performance –0.023 (0.874) Leverage –0.103 (0.330) Observations 708 678 R-squared 0.199 0.207 This table presents the effect of government incentives on MBO choices. Panel A reports the summary statistics of variables. Significance levels are based on two-tailed tests of differences between the MBO firms and other methods. Panel B presents the logit regression of MBO choices as in Equation (2). Fiscal resources is defined as fiscal revenue over GDP; High share of SOE employment is a dummy variable indicating Share of SOE Employment above the median. Robust standard errors are in parentheses. In both panels, significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 9 Government incentives and choices of MBO method A. Summary statistics of government incentives and city-level variables All privatizatized SOEs MBO Government incentives Fiscal resources Mean 0.67 0.70*** Median 0.71 0.71 Share of SOE employment Mean 0.25 0.24 Median 0.17 0.16* % with government land subsidy Mean 69 62$$^{***}$$ % with government guarantee of loans Mean 7 7 City-level controls Log (GDP per capita) Mean 9.72 9.77* Median 9.71 9.78* Population growth Mean 0.03 0.04* Median 0.01 0.01 A. Summary statistics of government incentives and city-level variables All privatizatized SOEs MBO Government incentives Fiscal resources Mean 0.67 0.70*** Median 0.71 0.71 Share of SOE employment Mean 0.25 0.24 Median 0.17 0.16* % with government land subsidy Mean 69 62$$^{***}$$ % with government guarantee of loans Mean 7 7 City-level controls Log (GDP per capita) Mean 9.72 9.77* Median 9.71 9.78* Population growth Mean 0.03 0.04* Median 0.01 0.01 B. Logit regression of MBO choices Dependent variable: MBO (1) (2) Government incentives Fiscal resources –0.979 –1.173 (0.230) (0.159) Share of SOE employment –0.748** –0.754** (0.024) (0.026) Fiscal resources * High share of SOE employment 2.660*** 2.372*** (0.002) (0.008) Government land subsidy –0.142*** –0.142*** (0.000) (0.001) Government guarantee of loans 0.053 0.078 (0.464) (0.314) City-level controls Log (GDP per capita) –0.021 –0.022 (0.568) (0.554) Population growth 0.216 0.233 (0.242) (0.241) Firm-level controls Log (sales) –0.021* (0.054) Performance –0.023 (0.874) Leverage –0.103 (0.330) Observations 708 678 R-squared 0.199 0.207 B. Logit regression of MBO choices Dependent variable: MBO (1) (2) Government incentives Fiscal resources –0.979 –1.173 (0.230) (0.159) Share of SOE employment –0.748** –0.754** (0.024) (0.026) Fiscal resources * High share of SOE employment 2.660*** 2.372*** (0.002) (0.008) Government land subsidy –0.142*** –0.142*** (0.000) (0.001) Government guarantee of loans 0.053 0.078 (0.464) (0.314) City-level controls Log (GDP per capita) –0.021 –0.022 (0.568) (0.554) Population growth 0.216 0.233 (0.242) (0.241) Firm-level controls Log (sales) –0.021* (0.054) Performance –0.023 (0.874) Leverage –0.103 (0.330) Observations 708 678 R-squared 0.199 0.207 This table presents the effect of government incentives on MBO choices. Panel A reports the summary statistics of variables. Significance levels are based on two-tailed tests of differences between the MBO firms and other methods. Panel B presents the logit regression of MBO choices as in Equation (2). Fiscal resources is defined as fiscal revenue over GDP; High share of SOE employment is a dummy variable indicating Share of SOE Employment above the median. Robust standard errors are in parentheses. In both panels, significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Panel B of Table 9 presents our regression results. In Column (1), the impact of a higher share of SOE employment is negative as expected (at the 5% level). The interaction term between Fiscal revenue/GDP and High share of SOE employment enters with a positive sign (at the 1% level), suggesting that, in cities where political opposition to layoffs is stronger, greater fiscal resources allow the government to provide better support for redeployment of laid-off workers, resulting in more MBOs. Government allocation of land is significantly negative (at the 1% level), suggesting preexisting government-firm ties make committing to MBOs harder for the government. In Column (2) of Table 9B, we further add firm-level variables in the year prior to privatization, including size, profitability, and leverage. Firm size is significantly related to MBO choices with a negative sign (at the 10% level), consistent with our earlier discussion that small firms are targeted for MBOs (Section 1.3). Notably, profitability is not statistically significant in determining the restructuring choices. The results are economically significant for measures of city governments’ incentives. Using the point estimates in Column (2) of Table 9B, a 10% increase in SOE shares reduces the odds ratio of MBOs by 7.3% ($$=$$1-exp(-0.754*0.1)). In cities with high share of SOEs and thus greater political pressure against layoffs, a 10% increase in fiscal resources increases the odds ratio of MBOs by 1.3 times. The odds ratio of MBOs among firms with government land subsidies is 13% lower than those without. 5. Choice of Privatization Methods and Firm Performance This section empirically evaluates the effect of privatization methods on performance. In our sample, firms are privatized in different years, whereas the NSB’s financial information is only available from 1998 to 2007. To fully utilize the data, we use the following panel regression: \begin{equation} \textit{Performance}_{it} = \alpha_{i} + \beta_{t} + \gamma \textit{Post}_{it} +\lambda \textit{MBO}_{i}\times\textit{Post}_{it} + \delta X_{it} + \varepsilon_{it}, \end{equation} (3) where Performance$$_{it}$$ is measured as earnings over assets (or ROA) and earnings per employee. Post$$_{it}$$ is a dummy variable indicating years after privatization. X$$_{it}$$ contains firm control variables, including size (measured as log of assets), leverage (debt over assets). $$\alpha_{i} $$is the firm fixed effect and $$\beta_{t}$$ is the year fixed effect. Coefficient $$\gamma $$ captures performance improvement after privatization, whereas the coefficient $$\lambda $$ is difference-in-differences estimate of the performance gain of MBOs as compared with other methods. 5.1 A first look at the performance of Chinese firms We first present an overall picture of the operating performance of Chinese firms, by estimating Equation (3) without the coefficient $$\lambda $$, on the sample of all firms including privatized firms, nonprivatized SOEs, and de novo private firms. Columns (1) and (2) of Table 10 show that, consistent with popular reports that SOEs are in a much weaker competitive position as compared to de novo private firms, the SOE dummy is significantly negative for both performance measures (at the 1% levels). In Columns (3) and (4), we add a dummy indicating privatized firms, it is not significantly different from zero. Meanwhile, the Post dummy is insignificant, suggesting that, when we pool all privatized firms together, regardless of how they are privatized, we find no evidence of performance improvement. Table 10 A first look at performance of Chinese firms Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) Log (sales) 0.026*** 16.629*** 0.026*** 16.621*** (0.001) (1.415) (0.001) (1.400) Leverage –0.071*** –3.511 –0.070*** –3.473 (0.011) (4.901) (0.011) (4.851) Privatized firms –0.088*** –31.925*** –0.094*** –32.494*** (0.007) (2.672) (0.006) (3.815) SOE –0.097*** –23.972*** –0.098*** –24.032*** (0.006) (2.482) (0.006) (2.561) Post dummy 0.012 1.026 (0.009) (3.220) Year dummies Yes Yes Yes Yes Industry fixed effects Yes Yes Yes Yes Observations 17,339 17,153 17,339 17,153 R-squared 0.072 0.057 0.072 0.057 Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) Log (sales) 0.026*** 16.629*** 0.026*** 16.621*** (0.001) (1.415) (0.001) (1.400) Leverage –0.071*** –3.511 –0.070*** –3.473 (0.011) (4.901) (0.011) (4.851) Privatized firms –0.088*** –31.925*** –0.094*** –32.494*** (0.007) (2.672) (0.006) (3.815) SOE –0.097*** –23.972*** –0.098*** –24.032*** (0.006) (2.482) (0.006) (2.561) Post dummy 0.012 1.026 (0.009) (3.220) Year dummies Yes Yes Yes Yes Industry fixed effects Yes Yes Yes Yes Observations 17,339 17,153 17,339 17,153 R-squared 0.072 0.057 0.072 0.057 This table presents the OLS estimates of the effect of privatization on firm performance as in Equation (3). It is based on the sample containing privatized SOEs, nonprivatized SOEs, and de novo private firms during 1998 to 2007. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 10 A first look at performance of Chinese firms Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) Log (sales) 0.026*** 16.629*** 0.026*** 16.621*** (0.001) (1.415) (0.001) (1.400) Leverage –0.071*** –3.511 –0.070*** –3.473 (0.011) (4.901) (0.011) (4.851) Privatized firms –0.088*** –31.925*** –0.094*** –32.494*** (0.007) (2.672) (0.006) (3.815) SOE –0.097*** –23.972*** –0.098*** –24.032*** (0.006) (2.482) (0.006) (2.561) Post dummy 0.012 1.026 (0.009) (3.220) Year dummies Yes Yes Yes Yes Industry fixed effects Yes Yes Yes Yes Observations 17,339 17,153 17,339 17,153 R-squared 0.072 0.057 0.072 0.057 Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) Log (sales) 0.026*** 16.629*** 0.026*** 16.621*** (0.001) (1.415) (0.001) (1.400) Leverage –0.071*** –3.511 –0.070*** –3.473 (0.011) (4.901) (0.011) (4.851) Privatized firms –0.088*** –31.925*** –0.094*** –32.494*** (0.007) (2.672) (0.006) (3.815) SOE –0.097*** –23.972*** –0.098*** –24.032*** (0.006) (2.482) (0.006) (2.561) Post dummy 0.012 1.026 (0.009) (3.220) Year dummies Yes Yes Yes Yes Industry fixed effects Yes Yes Yes Yes Observations 17,339 17,153 17,339 17,153 R-squared 0.072 0.057 0.072 0.057 This table presents the OLS estimates of the effect of privatization on firm performance as in Equation (3). It is based on the sample containing privatized SOEs, nonprivatized SOEs, and de novo private firms during 1998 to 2007. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. 5.2 Privatization methods and firm performance We now examine the differing effect of privatization methods on firm performance, by estimating Equation (3) on the sample of all privatized firms. In the first two columns of Table 11 the interaction between MBO and the Post dummy is significantly positive for both measures of performance (at the 1% level). The coefficient on the Post dummy itself is not significant, suggesting privatization methods other than MBOs do not improve performance. In Columns (3) and (4), we add firm fixed effects. The coefficient on MBO*Post remains positive and significant. The results are economically significant. The point estimate is -0.044 in Column (3), implying that, all else equal, MBOs outperform non-MBOs by 4.4% in ROA and by close to 6000 RMB, or roughly 750 USD, in earnings per employee per year. Table 11 Influence of privatization methods on post-privatization performance Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Log (sales) 0.021*** 9.256*** 0.084*** 18.263*** 0.084*** 18.262*** (0.002) (0.401) (0.011) (1.521) (0.011) (1.521) Leverage –0.079*** 0.207 0.006 6.756* 0.007 6.739* (0.020) (3.545) (0.018) (3.881) (0.018) (3.857) Post dummy 0.005 –1.178 –0.013 –4.169* –0.009 –4.275 (0.012) (2.074) (0.013) (2.218) (0.015) (2.903) MBO * Post 0.057*** 7.467*** 0.044*** 5.950*** 0.040*** 6.047* (0.011) (1.938) (0.015) (2.615) (0.017) (3.194) Direct sales to –0.010 0.268 $$\quad$$ outsiders * Post (0.018) (3.792) Year dummies Yes Yes Yes Yes Yes Yes Industry fixed effects Yes Yes No No No No Firm fixed effects No No Yes Yes Yes Yes Observations 5,222 5,144 5,245 5,167 5,245 5,167 R-squared 0.083 0.206 0.518 0.549 0.518 0.549 Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Log (sales) 0.021*** 9.256*** 0.084*** 18.263*** 0.084*** 18.262*** (0.002) (0.401) (0.011) (1.521) (0.011) (1.521) Leverage –0.079*** 0.207 0.006 6.756* 0.007 6.739* (0.020) (3.545) (0.018) (3.881) (0.018) (3.857) Post dummy 0.005 –1.178 –0.013 –4.169* –0.009 –4.275 (0.012) (2.074) (0.013) (2.218) (0.015) (2.903) MBO * Post 0.057*** 7.467*** 0.044*** 5.950*** 0.040*** 6.047* (0.011) (1.938) (0.015) (2.615) (0.017) (3.194) Direct sales to –0.010 0.268 $$\quad$$ outsiders * Post (0.018) (3.792) Year dummies Yes Yes Yes Yes Yes Yes Industry fixed effects Yes Yes No No No No Firm fixed effects No No Yes Yes Yes Yes Observations 5,222 5,144 5,245 5,167 5,245 5,167 R-squared 0.083 0.206 0.518 0.549 0.518 0.549 This table presents the influence of privatization methods on firm performance as in Equation (3). It is based on the sample of all privatized firms during 1998 to 2007. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 11 Influence of privatization methods on post-privatization performance Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Log (sales) 0.021*** 9.256*** 0.084*** 18.263*** 0.084*** 18.262*** (0.002) (0.401) (0.011) (1.521) (0.011) (1.521) Leverage –0.079*** 0.207 0.006 6.756* 0.007 6.739* (0.020) (3.545) (0.018) (3.881) (0.018) (3.857) Post dummy 0.005 –1.178 –0.013 –4.169* –0.009 –4.275 (0.012) (2.074) (0.013) (2.218) (0.015) (2.903) MBO * Post 0.057*** 7.467*** 0.044*** 5.950*** 0.040*** 6.047* (0.011) (1.938) (0.015) (2.615) (0.017) (3.194) Direct sales to –0.010 0.268 $$\quad$$ outsiders * Post (0.018) (3.792) Year dummies Yes Yes Yes Yes Yes Yes Industry fixed effects Yes Yes No No No No Firm fixed effects No No Yes Yes Yes Yes Observations 5,222 5,144 5,245 5,167 5,245 5,167 R-squared 0.083 0.206 0.518 0.549 0.518 0.549 Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Log (sales) 0.021*** 9.256*** 0.084*** 18.263*** 0.084*** 18.262*** (0.002) (0.401) (0.011) (1.521) (0.011) (1.521) Leverage –0.079*** 0.207 0.006 6.756* 0.007 6.739* (0.020) (3.545) (0.018) (3.881) (0.018) (3.857) Post dummy 0.005 –1.178 –0.013 –4.169* –0.009 –4.275 (0.012) (2.074) (0.013) (2.218) (0.015) (2.903) MBO * Post 0.057*** 7.467*** 0.044*** 5.950*** 0.040*** 6.047* (0.011) (1.938) (0.015) (2.615) (0.017) (3.194) Direct sales to –0.010 0.268 $$\quad$$ outsiders * Post (0.018) (3.792) Year dummies Yes Yes Yes Yes Yes Yes Industry fixed effects Yes Yes No No No No Firm fixed effects No No Yes Yes Yes Yes Observations 5,222 5,144 5,245 5,167 5,245 5,167 R-squared 0.083 0.206 0.518 0.549 0.518 0.549 This table presents the influence of privatization methods on firm performance as in Equation (3). It is based on the sample of all privatized firms during 1998 to 2007. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. In Columns (5) and (6) of Table 11, when we add an interaction between Direct sales to outsiders and Post, it is not significantly different from zero, suggesting direct sales to outsiders do not improve performance. This result is consistent with state control and a general lack of restructuring measures in this kind of privatization. Finally, the Post dummy itself is insignificant, suggesting that there is generally not any change in performance for non-MBOs. 5.3 Further analysis: The selection concern A common concern about performance evaluation of privatized firms is the selection bias. For example, one may worry MBO firms have significantly better post-privatization performance because better firms are systematically chosen for MBOs; or managers may have incentives to buy out the firms if they have information about government implicit promises (e.g., in the form of protected entry) or future prospects of the firms; or managers may have manipulated the earnings downward prior to MBOs so that they could buy out the firms more cheaply, causing a mechanical increase in earnings post privatization. We should stress that compared with the previous literature, our data allow us to deal with the selection bias more seriously. In fact, the analyses so far have already addressed the selection issue in several ways. First, we do not simply make performance comparisons, but rather, we have identified the mechanism of performance improvement in MBOs, through transfer of control rights to private owners (Tables 4 and 5) and thus avoiding the negative impact of state control on performance (Table 6), through less government support and hardened budget constraint (Table 7), and through more enterprise restructuring (Table 8). Second, we explicitly examine the factors that affect the chances of firms being selected for MBOs. The fact that we find political and fiscal incentives, rather than the above-mentioned economic considerations, determine the choice of privatization method (Table 9) is reassuring. In the following analysis, we perform several additional tests to rule out the selection bias even further. 5.3.1 Preexisting trends in performance If MBO firms were better firms or firms with greater growth potential, one should observe better performance prior to privatization. As Figure 3 shows, there is not any preexisting trend in performance. Figure 3 View largeDownload slide No preexisting trend in performance differences between MBOs and other privatization methods Solid lines are the mean; dashed lines are 90% confidence intervals. Figure 3 View largeDownload slide No preexisting trend in performance differences between MBOs and other privatization methods Solid lines are the mean; dashed lines are 90% confidence intervals. 5.3.2 Controlling for the impact of city-level economic aspects One might worry that MBOs perform better because of city-level economic prospects. Specifically, if a greater share of SOEs and low fiscal resources symbolize a lack of future prospects for the privatized firms, there would be less managerial incentives to buy out the firms, resulting in a positive relationship between MBO and performance. To address this concern, we use Columns (3) and (4) in Table 11 as the base estimation and add the interactions between city and year dummies, thus fully purging of all city-level time varying trends. We note that this is a strong test, as there are 205 cities and ten years of data, resulting in 1,481 dummies being included in the estimation.14 It is also worth pointing out that in our earlier estimation, we already have firm fixed effects and thus have controlled for time-invariant city-level economic prospects. Panel A of Table 12 shows that, while adding 1,383 city-year dummies slightly weakens the significance level in one estimation (not surprisingly), it does not qualitatively change any of the earlier results. Further, the point estimates of MBO*Post, our main variables of interest, remain of similar magnitudes, implying that the time-variant component of city level influence does not drive the results. Table 12 Further analysis of MBO performance improvement A. Impact of city-level economic prospects Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) Log (sales) 0.084*** 16.863*** 0.084*** 16.848*** (0.015) (1.570) (0.015) (1.573) Leverage –0.002 0.092 –0.001 –0.109 (0.024) (5.208) (0.025) (5.219) Post dummy –0.026 –6.543* –0.024 –8.423* (0.017) (3.390) (0.020) (4.297) MBO * Post 0.067*** 8.187** 0.065** 9.971** (0.025) (3.783) (0.026) (4.420) Direct sales to outsiders * Post –0.004 4.732 (0.025) (5.501) City*Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 R-squared 0.647 0.724 0.647 0.724 A. Impact of city-level economic prospects Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) Log (sales) 0.084*** 16.863*** 0.084*** 16.848*** (0.015) (1.570) (0.015) (1.573) Leverage –0.002 0.092 –0.001 –0.109 (0.024) (5.208) (0.025) (5.219) Post dummy –0.026 –6.543* –0.024 –8.423* (0.017) (3.390) (0.020) (4.297) MBO * Post 0.067*** 8.187** 0.065** 9.971** (0.025) (3.783) (0.026) (4.420) Direct sales to outsiders * Post –0.004 4.732 (0.025) (5.501) City*Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 R-squared 0.647 0.724 0.647 0.724 B. Protected entry Performance measures Performance measures Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) (5) (6) (7) (8) Log (sales) 0.084*** 18.192*** 0.084*** 18.192*** 0.084*** 18.197*** 0.083*** 18.203*** (0.011) (1.524) (0.011) (1.525) (0.011) (1.527) (0.011) (1.533) Leverage 0.007 6.803* 0.007 6.803* 0.007 6.816* 0.008 6.805* (0.018) (3.872) (0.018) (3.880) (0.018) (3.877) (0.018) (3.872) Post dummy –0.003 –3.995* –0.002 –3.996* –0.007 –4.019* –0.002 –4.160* (0.012) (2.210) (0.012) (2.253) (0.015) (2.102) (0.013) (2.215) MBO * Post 0.036** 5.766** 0.035** 5.769** 0.041** 5.866** 0.031* 6.121** (0.015) (2.589) (0.016) (2.703) (0.016) (2.492) (0.016) (2.825) Monopoly * Post –0.058** –1.652 –0.061** –1.642 (0.025) (4.907) (0.031) (5.878) Monopoly * MBO *Post 0.016 –0.053 (0.037) (8.656) Market power * Post –0.020 –0.884 –0.038* –0.403 (0.022) (3.266) (0.023) (4.808) Market power * MBO * Post 0.049 –1.246 (0.052) (5.948) Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 5,241 5,163 5,241 5,163 R-squared 0.519 0.549 0.519 0.549 0.518 0.549 0.519 0.549 B. Protected entry Performance measures Performance measures Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) (5) (6) (7) (8) Log (sales) 0.084*** 18.192*** 0.084*** 18.192*** 0.084*** 18.197*** 0.083*** 18.203*** (0.011) (1.524) (0.011) (1.525) (0.011) (1.527) (0.011) (1.533) Leverage 0.007 6.803* 0.007 6.803* 0.007 6.816* 0.008 6.805* (0.018) (3.872) (0.018) (3.880) (0.018) (3.877) (0.018) (3.872) Post dummy –0.003 –3.995* –0.002 –3.996* –0.007 –4.019* –0.002 –4.160* (0.012) (2.210) (0.012) (2.253) (0.015) (2.102) (0.013) (2.215) MBO * Post 0.036** 5.766** 0.035** 5.769** 0.041** 5.866** 0.031* 6.121** (0.015) (2.589) (0.016) (2.703) (0.016) (2.492) (0.016) (2.825) Monopoly * Post –0.058** –1.652 –0.061** –1.642 (0.025) (4.907) (0.031) (5.878) Monopoly * MBO *Post 0.016 –0.053 (0.037) (8.656) Market power * Post –0.020 –0.884 –0.038* –0.403 (0.022) (3.266) (0.023) (4.808) Market power * MBO * Post 0.049 –1.246 (0.052) (5.948) Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 5,241 5,163 5,241 5,163 R-squared 0.519 0.549 0.519 0.549 0.518 0.549 0.519 0.549 C. Comparison between MBO and de novo private firms C1. The regression method Performance measures Profits / assets Profits / #employee (1) (2) Log (sales) 0.110*** 23.402*** (0.013) (1.962) Leverage –0.005 8.037 (0.023) (11.316) MBO * Post 0.023 4.425 (0.014) (2.703) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 7,468 7,385 R-squared 0.632 0.515 C. Comparison between MBO and de novo private firms C1. The regression method Performance measures Profits / assets Profits / #employee (1) (2) Log (sales) 0.110*** 23.402*** (0.013) (1.962) Leverage –0.005 8.037 (0.023) (11.316) MBO * Post 0.023 4.425 (0.014) (2.703) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 7,468 7,385 R-squared 0.632 0.515 C2. The matched sample method Matched de novo firms based on industry and MBO privatization Assets within 20% range Closest in assets Sales within 20% range Closest in sales % Matched na 67 98 71 98 Performance measurement: Profits/assets Performance improvement 0.09$$^{***}$$ 0.06*** 0.06*** 0.09*** 0.07*** Difference-in-differences na 0.07 0.04 –0.04 –0.02 p-values na (.236) (.301) (.241) (.518) Performance measurement: Profits/ #employee Performance improvement 16.70$$^{***}$$ 35.18** 30.08*** 35.12*** 37.38*** Difference-in-differences na –11.16 –8.51 –17.86 –15.27* p-values na (.405) (.385) (.104) (.099) C2. The matched sample method Matched de novo firms based on industry and MBO privatization Assets within 20% range Closest in assets Sales within 20% range Closest in sales % Matched na 67 98 71 98 Performance measurement: Profits/assets Performance improvement 0.09$$^{***}$$ 0.06*** 0.06*** 0.09*** 0.07*** Difference-in-differences na 0.07 0.04 –0.04 –0.02 p-values na (.236) (.301) (.241) (.518) Performance measurement: Profits/ #employee Performance improvement 16.70$$^{***}$$ 35.18** 30.08*** 35.12*** 37.38*** Difference-in-differences na –11.16 –8.51 –17.86 –15.27* p-values na (.405) (.385) (.104) (.099) This table provides further analysis of the influence of privatization methods on firm performance. Panel A and B estimate Equation (3) based on the sample of all privatized firms during 1998 to 2007. Panel C compares MBOs and de novo private firms. Panel C1 estimates Equation (3) based on the sample of MBO and de novo private firms during 1998 to 2007. Panel C2 reports the results by the method of matched sample. Matching is by industry and size, where the size is based on assets or sales within 20% range, or firms that are closest in size to the MBO firm. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Performance improvement is defined as the difference of average performance measures before and after privatization. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, ** and *, respectively, and na refers to not applicable. Table 12 Further analysis of MBO performance improvement A. Impact of city-level economic prospects Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) Log (sales) 0.084*** 16.863*** 0.084*** 16.848*** (0.015) (1.570) (0.015) (1.573) Leverage –0.002 0.092 –0.001 –0.109 (0.024) (5.208) (0.025) (5.219) Post dummy –0.026 –6.543* –0.024 –8.423* (0.017) (3.390) (0.020) (4.297) MBO * Post 0.067*** 8.187** 0.065** 9.971** (0.025) (3.783) (0.026) (4.420) Direct sales to outsiders * Post –0.004 4.732 (0.025) (5.501) City*Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 R-squared 0.647 0.724 0.647 0.724 A. Impact of city-level economic prospects Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) Log (sales) 0.084*** 16.863*** 0.084*** 16.848*** (0.015) (1.570) (0.015) (1.573) Leverage –0.002 0.092 –0.001 –0.109 (0.024) (5.208) (0.025) (5.219) Post dummy –0.026 –6.543* –0.024 –8.423* (0.017) (3.390) (0.020) (4.297) MBO * Post 0.067*** 8.187** 0.065** 9.971** (0.025) (3.783) (0.026) (4.420) Direct sales to outsiders * Post –0.004 4.732 (0.025) (5.501) City*Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 R-squared 0.647 0.724 0.647 0.724 B. Protected entry Performance measures Performance measures Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) (5) (6) (7) (8) Log (sales) 0.084*** 18.192*** 0.084*** 18.192*** 0.084*** 18.197*** 0.083*** 18.203*** (0.011) (1.524) (0.011) (1.525) (0.011) (1.527) (0.011) (1.533) Leverage 0.007 6.803* 0.007 6.803* 0.007 6.816* 0.008 6.805* (0.018) (3.872) (0.018) (3.880) (0.018) (3.877) (0.018) (3.872) Post dummy –0.003 –3.995* –0.002 –3.996* –0.007 –4.019* –0.002 –4.160* (0.012) (2.210) (0.012) (2.253) (0.015) (2.102) (0.013) (2.215) MBO * Post 0.036** 5.766** 0.035** 5.769** 0.041** 5.866** 0.031* 6.121** (0.015) (2.589) (0.016) (2.703) (0.016) (2.492) (0.016) (2.825) Monopoly * Post –0.058** –1.652 –0.061** –1.642 (0.025) (4.907) (0.031) (5.878) Monopoly * MBO *Post 0.016 –0.053 (0.037) (8.656) Market power * Post –0.020 –0.884 –0.038* –0.403 (0.022) (3.266) (0.023) (4.808) Market power * MBO * Post 0.049 –1.246 (0.052) (5.948) Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 5,241 5,163 5,241 5,163 R-squared 0.519 0.549 0.519 0.549 0.518 0.549 0.519 0.549 B. Protected entry Performance measures Performance measures Performance measures Performance measures Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee Profits/assets Profits/#employee (1) (2) (3) (4) (5) (6) (7) (8) Log (sales) 0.084*** 18.192*** 0.084*** 18.192*** 0.084*** 18.197*** 0.083*** 18.203*** (0.011) (1.524) (0.011) (1.525) (0.011) (1.527) (0.011) (1.533) Leverage 0.007 6.803* 0.007 6.803* 0.007 6.816* 0.008 6.805* (0.018) (3.872) (0.018) (3.880) (0.018) (3.877) (0.018) (3.872) Post dummy –0.003 –3.995* –0.002 –3.996* –0.007 –4.019* –0.002 –4.160* (0.012) (2.210) (0.012) (2.253) (0.015) (2.102) (0.013) (2.215) MBO * Post 0.036** 5.766** 0.035** 5.769** 0.041** 5.866** 0.031* 6.121** (0.015) (2.589) (0.016) (2.703) (0.016) (2.492) (0.016) (2.825) Monopoly * Post –0.058** –1.652 –0.061** –1.642 (0.025) (4.907) (0.031) (5.878) Monopoly * MBO *Post 0.016 –0.053 (0.037) (8.656) Market power * Post –0.020 –0.884 –0.038* –0.403 (0.022) (3.266) (0.023) (4.808) Market power * MBO * Post 0.049 –1.246 (0.052) (5.948) Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Observations 5,241 5,163 5,241 5,163 5,241 5,163 5,241 5,163 R-squared 0.519 0.549 0.519 0.549 0.518 0.549 0.519 0.549 C. Comparison between MBO and de novo private firms C1. The regression method Performance measures Profits / assets Profits / #employee (1) (2) Log (sales) 0.110*** 23.402*** (0.013) (1.962) Leverage –0.005 8.037 (0.023) (11.316) MBO * Post 0.023 4.425 (0.014) (2.703) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 7,468 7,385 R-squared 0.632 0.515 C. Comparison between MBO and de novo private firms C1. The regression method Performance measures Profits / assets Profits / #employee (1) (2) Log (sales) 0.110*** 23.402*** (0.013) (1.962) Leverage –0.005 8.037 (0.023) (11.316) MBO * Post 0.023 4.425 (0.014) (2.703) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 7,468 7,385 R-squared 0.632 0.515 C2. The matched sample method Matched de novo firms based on industry and MBO privatization Assets within 20% range Closest in assets Sales within 20% range Closest in sales % Matched na 67 98 71 98 Performance measurement: Profits/assets Performance improvement 0.09$$^{***}$$ 0.06*** 0.06*** 0.09*** 0.07*** Difference-in-differences na 0.07 0.04 –0.04 –0.02 p-values na (.236) (.301) (.241) (.518) Performance measurement: Profits/ #employee Performance improvement 16.70$$^{***}$$ 35.18** 30.08*** 35.12*** 37.38*** Difference-in-differences na –11.16 –8.51 –17.86 –15.27* p-values na (.405) (.385) (.104) (.099) C2. The matched sample method Matched de novo firms based on industry and MBO privatization Assets within 20% range Closest in assets Sales within 20% range Closest in sales % Matched na 67 98 71 98 Performance measurement: Profits/assets Performance improvement 0.09$$^{***}$$ 0.06*** 0.06*** 0.09*** 0.07*** Difference-in-differences na 0.07 0.04 –0.04 –0.02 p-values na (.236) (.301) (.241) (.518) Performance measurement: Profits/ #employee Performance improvement 16.70$$^{***}$$ 35.18** 30.08*** 35.12*** 37.38*** Difference-in-differences na –11.16 –8.51 –17.86 –15.27* p-values na (.405) (.385) (.104) (.099) This table provides further analysis of the influence of privatization methods on firm performance. Panel A and B estimate Equation (3) based on the sample of all privatized firms during 1998 to 2007. Panel C compares MBOs and de novo private firms. Panel C1 estimates Equation (3) based on the sample of MBO and de novo private firms during 1998 to 2007. Panel C2 reports the results by the method of matched sample. Matching is by industry and size, where the size is based on assets or sales within 20% range, or firms that are closest in size to the MBO firm. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Performance improvement is defined as the difference of average performance measures before and after privatization. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, ** and *, respectively, and na refers to not applicable. 5.3.3 Accounting for the role of product market competition Section 3.2 shows that MBOs are rarely in monopolistic industries, which suggests that the better performance of MBOs is not likely driven by regulatory barrier to entry. We now extend performance analysis by explicitly including firms’ self-reported competition. This analysis is interesting and important in its own right. It helps understand the impact of product market competition on privatization outcomes. The answer is not ex ante clear: on the one hand, firms with monopoly power generally do better; on the other hand, firms facing competition has a stronger drive to improve efficiency because if they do not they may not even survive. Evidence from market economy has been mixed and identification is difficult due to endogenous market structure. Since trade liberalization often accompanies privatization, data from transition economies offer an opportunity to better deal with the identification challenge (Djankov and Murrell 2002). This advantage is even stronger in the Chinese setting, because, at the time of privatization, there is a significant variation in market structure across industries. Again using Columns (3) and (4) in Table 11 as the base estimation, we add variables on product market competition (panel B of Table 12). We find that firms with monopoly power perform significantly worse in terms of return on assets (at the 5% level), suggesting that, when firms in monopolistic industries are privatized, they do not have a strong incentive to improve efficiency. This is consistent with results from transition economies that, while there are regional variations, the impact of product market competition on privatization performance is significantly positive overall (Djankov and Murrell 2002). Moreover, the triple interaction Monopoly*MBO*Post is insignificant, while our main variable of interest, MBO*Post, remains significantly positive. This means that better MBO performance is not driven by regulated entry. Finally, when we use a variable indicating market power, i.e., both monopoly firms and those reported to have few competitors, we obtain very similar results. 5.3.4 Controlling for the impact of government supports We now compare MBOs with their private sector benchmark. If they have advantages arising from their government ties they would outperform de novo private firms. In panel C1 of Table 12, we rerun Equation (3) by including only MBOs and de novo private firms. The insignificant coefficient of MBO*post suggests that MBOs have similar performance to the private sector benchmark. In panel C2 of Table, we perform a difference-in-differences analysis based on industry and size matching. The results are highly robust to alternative matching criteria and we do not find any evidence that MBOs outperform their private sector peers. 5.3.5 Instrumental variable estimation We use city government’s political incentives as instruments to estimate the effect of MBO choices on performance. The instruments include %SOE employment, Fiscal revenue/GDP, government allocation of land, and loan guarantees. The first-stage regression is the same as that in Column (1) of Table 9B in Section 4. We employ the limited information maximum likelihood (LIML) estimation of the two-stage least-squares (TSLS) regressions, which is more robust to weak IV problems.15Table 13 reports the results. The IV difference-in-differences estimates are quantitatively similar to our OLS estimates, further confirming that selection does not drive MBO performance. Table 13 Instrumental variable estimates of the effect of MBO on performance Performance measures Profits/assets Profits/#employee (1) (2) Log (sales) 0.081*** 17.442*** (0.011) (1.307) Leverage –0.001 5.594 (0.019) (4.091) Post dummy –0.085** –17.426** (0.039) (6.941) MBO * Post 0.186** 30.588** (0.080) (13.070) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 4,869 4,805 R-squared 0.520 0.566 Performance measures Profits/assets Profits/#employee (1) (2) Log (sales) 0.081*** 17.442*** (0.011) (1.307) Leverage –0.001 5.594 (0.019) (4.091) Post dummy –0.085** –17.426** (0.039) (6.941) MBO * Post 0.186** 30.588** (0.080) (13.070) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 4,869 4,805 R-squared 0.520 0.566 This table presents the instrumental variable (IV) estimates of the effect of MBO on performance as in Equation (3). It is based on the sample of all privatized firms during 1998 to 2007. The model is estimated using limited information maximum likelihood (LIML) estimation. Government incentives, defined in Table 5, are used as instruments. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets, sales, and number of employees, respectively. The number of observations is less than that in Table 11 because of missing numbers in some of the instruments including SOE shares and fiscal capacity. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 13 Instrumental variable estimates of the effect of MBO on performance Performance measures Profits/assets Profits/#employee (1) (2) Log (sales) 0.081*** 17.442*** (0.011) (1.307) Leverage –0.001 5.594 (0.019) (4.091) Post dummy –0.085** –17.426** (0.039) (6.941) MBO * Post 0.186** 30.588** (0.080) (13.070) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 4,869 4,805 R-squared 0.520 0.566 Performance measures Profits/assets Profits/#employee (1) (2) Log (sales) 0.081*** 17.442*** (0.011) (1.307) Leverage –0.001 5.594 (0.019) (4.091) Post dummy –0.085** –17.426** (0.039) (6.941) MBO * Post 0.186** 30.588** (0.080) (13.070) Year dummies Yes Yes Firm fixed effects Yes Yes Observations 4,869 4,805 R-squared 0.520 0.566 This table presents the instrumental variable (IV) estimates of the effect of MBO on performance as in Equation (3). It is based on the sample of all privatized firms during 1998 to 2007. The model is estimated using limited information maximum likelihood (LIML) estimation. Government incentives, defined in Table 5, are used as instruments. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets, sales, and number of employees, respectively. The number of observations is less than that in Table 11 because of missing numbers in some of the instruments including SOE shares and fiscal capacity. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. 6. Generality of the Chinese Experience We now discuss the generality of the Chinese experience. At a high level, as pointed out by Frydman et al. (1999) and Estrin et al. (2009), reallocation of ownership and control to different types of owners have disparate effect on restructuring and performance; thus looking at aggregate results without knowing why could be misleading. The Chinese experience is perfectly consistent with this insight: when we pull all privatization methods together, we do not see any performance gain. In fact, our survey is designed to advance the literature by asking detailed questions on transfer of ownership and control, as well as restructuring. At a more detailed level, the previous literature has pointed to a common set of economic factors governing the success of privatization, including type of owners, management turnover, product market competition, and hardening of budget constraints (Djankov and Murrell 2002). Our findings support the importance of all four factors. The first factor, the types of owners, is worth a special note. There is a well-known result in the literature that privatization to insiders does not lead to efficiency gains in transition economies, which appears to be inconsistent with our findings that Chinese MBOs are the most effective privatization method. This, however, is a surface contrast. Our results do support findings in other countries that privatization to employees does not work, a quite intuitive result given that drawbacks of employee ownership are well recognized (e.g., Hansmann 1996). The ineffectiveness of managerial ownership, however, is somewhat puzzling, as it has usually been found to be effective in market economies since Morck, Shleifer, and Vishny (1988). The difference lies in the extent to which it is market based. Frydman et al. (1999) point out two “special characteristics” of managerial ownership in transition economies. First, in CEE-CIS nations, managers are selected under the old regime and thus may not have the skill set needed in the market economy. The second characteristic is that managers are offered to buy the shares at preferential prices but with restrictive terms, designed to favor existing employees. Chinese MBOs do not share these characteristics. While employment is also a concern at the time of privatization, once the firms are privatized via MBOs, they operate according to market practices. In this sense, Chinese MBOs constitute a nice counterfactual analysis for the studies of other transition economies (and vice versa). They confirm the conjectures in the literature regarding why managerial ownership does not work in CEE-CIS nations. In addition to supporting and extending the existing literature, our paper explores two important aspects of privatization not previously examined, both of which have general implications. Most notably, enabled by our detailed data on re-allocation of control rights, we show that withdrawal of state control in MBOs is the driving force behind restructuring and performance improvement. This has important implications for other privatization programs. So far, despite that most privatization around the world is partial, that is, the state incompletely transfers control, the role of retained state influence has not been thoroughly studied.16 The second area is the role of political factors in shaping the design of privatization programs. Despite that theoretical work and anecdotes all suggest a significant influence of political factors (e.g., Biais and Perotti 2002), few formal empirical papers study this topic and our paper joins a more recent effort (e.g., Dinc and Gupta 2011) on this important topic. In the Chinese setting, a lack of fiscal resources and political opposition to unemployment, prevent the commitment to withdrawal of state control and adoption of an effective privatization method. 7. Conclusion Privatization in China is unique in that, instead of being designed by the national government, it is initiated, designed, and implemented by city governments. Consequently, privatization methods and outcomes substantially vary across cities within one country. This distinctive experience provides a rich laboratory for studying choice of privatization methods and outcomes, and the mechanisms behind the differences in outcomes. We conduct a large-scale nationwide survey of over 3,000 firms from nearly one-third of China’s cities, based on a random sampling stratified by size and industry. The survey collects detailed information on ownership structure, reallocation of five corporate decision rights among five parties, remaining government tie and support, and four measures of post-privatization restructuring. Thus, the data we collect are, to our knowledge, the most comprehensive data available to researchers in studying a single country’s privatization. We find that, while privatization in China has made substantial progress in reallocating control rights to private owners, the degree of remaining government influence in corporate decisions differs significantly across privatization methods. The MBO method, which accounts for about half of all privatization programs, represents the strongest commitment to withdrawing state both control and support. Our evidence further suggests the city governments’ incentives and political constraints are the key determinants of their choices of privatization methods. In cities where political opposition to layoffs is weaker and where the city government has more fiscal resources to bear the cost of layoffs and to fill the gap in social welfare, the government is more likely to choose the MBO method. Finally, MBOs restructured more effectively and improved their performance significantly. In contrast, in direct sales to outsiders and other methods, the state retains substantial control, resulting in less restructuring and a lack of post-privatization performance improvement. The dynamics of privatization provide an important perspective for understanding the Chinese economy. Political constraints and state intervention are the main reasons why some privatization programs fail to enhance performance. The same dynamics govern the Chinese economy till today. During the period of rapid economic growth, the state has no urgency to push for further economic reforms, and political compromises result in greater state influence and thus economic inefficiencies in many sectors of the economy. Given the current economic slowdown, however, resolving these inefficiencies is important for future economic growth. Indeed, reforming the remaining often ultra-sized SOEs is back on the agenda and insights garnered from our study have important implications for SOE reforms in the future. Acknowledgement Comments from Tom Easton, Sudipto Dasgupta, Peter Mackay, Gary Jefferson, Tom Rawski, David Scharfstein, and Andrei Shleifer are greatly appreciated. We are also grateful for the helpful comments and suggestions from the Editor and two anonymous referees. We have benefited from comments on early versions of the paper by conference and seminar participants at NBER, Yale, Michigan, Hong Kong, and Seoul. We thank Ziyi Chen, Harry Leung, Tianshuo Shi, and Jin Wang for excellent research assistance. Special thanks go to the ESRC under the World Economy and Finance Research Program (award # RES-156-25-0009), which made the survey possible. Gan acknowledges the financial support from Hong Kong Research Grants Council (project # HKUST6490/06H). Xu acknowledges the financial support from the WCU program through the Korea Science and Engineering Foundation funded by the Ministry of Education, Science, and Technology (grant # R32-2008-000-20055-0) and the Hong Kong RGC TRS grant (T31-717 112-R). We are grateful for help from the governments of the nine cities in which we conducted our pilot survey. References Bai, C. E., Lu, J. and Tao. Z. 2006 . Property rights protection and access to bank loans. Economics of Transition 14 : 611 – 28 . Google Scholar CrossRef Search ADS Bai, C. E., Lu, J. and Tao. Z. 2009 . How does privatization work in China? Journal of Comparative Economics 37 : 453 – 70 . Google Scholar CrossRef Search ADS Barberis, N., Boycko, M. Shleifer A. and Tsukanova. N. 1996 . How does privatization work? Evidence from the Russian shops. Journal of Political Economy 104 : 764 – 90 . Google Scholar CrossRef Search ADS Biais, B., and Perotti. E. 2002 . Machiavellian privatization. American Economic Review 92 : 240 – 58 . Google Scholar CrossRef Search ADS Boubakri, N., Cosset, J. C. and Saffar. W. 2008 . Political connections of newly privatized firms. Journal of Corporate Finance 14 : 654 – 73 . Google Scholar CrossRef Search ADS Boycko, M., Shleifer, A. and Vishny. R. 1996 . A theory of privatization. Economic Journal 106 : 309 – 19 . Google Scholar CrossRef Search ADS Cuervo, A., and Villalonga. B. 2000 . Explaining the variance in the performance effects of privatization. Academy of Management Review 25 : 581 – 90 . Deng, J., Gan, J. and He. J. 2008 . The dark side of concentrated ownership in privatization: Evidence from China. Working Paper , Hong Kong University of Science and Technology . Deng, J., Gan, J. and He. J. 2010 . Political constraints, organizational forms, and privatization performance: Evidence from China. Working Paper , Hong Kong University of Science and Technology . Google Scholar CrossRef Search ADS Dinc, I., and Gupta. N. 2011 . The decision to privatize: Finance and politics. Journal of Finance 66 : 241 – 69 . Google Scholar CrossRef Search ADS Djankov, S., and Murrell. P. 2002 . Enterprise restructuring in transition: A quantitative survey. Journal of Economic Literature 40 : 739 – 92 . Google Scholar CrossRef Search ADS Dong, X., and Putterman. L. 2003 . Soft budget constraints, social burdens, and labor redundancy in China’s state industry. Journal of Comparative Economics 31 : 110 – 33 . Google Scholar CrossRef Search ADS Dong, X., Putterman, L. and Unel. B. 2006 . Privatization and firm performance: A comparison between rural and urban enterprises in China. Journal of Comparative Economics 34 : 608 – 33 . Google Scholar CrossRef Search ADS Estrin, S., Hanousek, J. Kočenda, E. and Svejnar. J. 2009 . Effects of privatization and ownership in transition economies. Journal of Economic Literature 47 : 699 – 728 . Google Scholar CrossRef Search ADS Frydman, R., Gray, C. Hessel, M. and Rapaczynski. A. 1999 . When does privatization work? The impact of private ownership on corporate performance in the transition economies. Quarterly Journal of Economics 114 : 1153 – 91 . Google Scholar CrossRef Search ADS Garnaut, R., Song, L. and Yao. Y. 2008 . Impact and significance of state–owned enterprise restructuring in China. In Policy reform and Chinese markets: Progress and challenges, eds. Fleisher, B. M. Hope, N. C. Pena, A. A. and Yang, D. T. 38 – 70 . Cheltenham, UK : Edward Elgar Publishing . Google Scholar CrossRef Search ADS Granick, D. 1990 . Chinese state enterprises: A regional property rights analysis . Chicago : University of Chicago Press . Grossman, S., and Hart, O. 1986 . The costs and benefits of ownership: A theory of vertical and lateral integration. Journal of Political Economy 94 : 691 – 719 . Google Scholar CrossRef Search ADS Guedhami, O., Pittman, J. A. and Saffar. W. 2009 . Auditor choice in privatized firms: Empirical evidence on the role of state and foreign owners. Journal of Accounting and Economics 48 : 151 – 71 . Google Scholar CrossRef Search ADS Guo, K., and Yao. Y. 2005 . Causes of privatization in China– Testing several hypotheses. Economics of Transition 13 : 211 – 38 . Google Scholar CrossRef Search ADS Gupta, N. 2005 . Partial privatization and firm performance. Journal of Finance 60 : 987 – 1015 . Google Scholar CrossRef Search ADS Hart, O., and Moore. J. 1990 . Property rights and the nature of the firm. Journal of Political Economy 98 : 1119 – 58 . Google Scholar CrossRef Search ADS Hansmann, H. 1996 . The Ownership of Enterprise . Cambridge, MA : Harvard University Press . Huyghebaert, N., and Quan. Q. 2009 . Share issuing privatizations in China: Sequencing and its effects on public share allocation and underpricing. Journal of Comparative Economics 37 : 306 – 20 . Google Scholar CrossRef Search ADS Jefferson, G. H., and Su. J. 2006 . Privatization and restructuring in China: evidence from shareholding ownership, 1995–2001. Journal of Comparative Economics 34 : 146 – 66 . Google Scholar CrossRef Search ADS Jia, R., Kudamatsu, M. and Seim. D. 2015 . Political selection in China: The complementary roles of connections and performance. Journal of the European Economic Association 13 : 631 – 68 . Google Scholar CrossRef Search ADS Jin, H., Qian, Y. and Weingast. B. R. 2005 . Regional decentralization and fiscal incentives: Federalism, Chinese style. Journal of Public Economics 89 : 1719 – 42 . Google Scholar CrossRef Search ADS Jones, D., and Mygind. N. 1999 . The nature and determinants of ownership changes after privatization: Evidence from Estonia. Journal of Comparative Economics 27 : 422 – 41 . Google Scholar CrossRef Search ADS Li, H., and Rozelle. S. 2000 . Savings or stripping rural industry: An analysis of privatization and efficiency in China. Agricultural Economics 23 : 241 – 52 . Google Scholar CrossRef Search ADS Li, H., and Zhou. L. 2005 . Political turnover and economic performance: The incentive role of personnel control in China. Journal of Public Economics 89 : 1743 – 62 . Google Scholar CrossRef Search ADS Li, T., Sun, L. and Zou. L. 2009 . State ownership and corporate performance: A quantile regression analysis of Chinese listed companies. China Economic Review 20 : 703 – 16 . Google Scholar CrossRef Search ADS Li, W. 1997 . The impact of economic reform on the performance of Chinese state enterprises, 1980–1989. Journal of Political Economy 105 : 1080 – 1106 . Google Scholar CrossRef Search ADS López-de-Silanes, F. 1997 . Determinants of privatization prices. Quarterly Journal of Economics 112 : 965 – 1025 . Google Scholar CrossRef Search ADS Maskin, E., Qian, Y. and Xu, C. 2000 . Incentives, information, and organizational form. Review of Economic Studies 67 : 359 – 78 . Google Scholar CrossRef Search ADS Megginson, W., and Netter. J. 2001 . From state to market: A survey of empirical studies on privatization. Journal of Economic Literature 39 : 321 – 89 . Google Scholar CrossRef Search ADS Morck, R., Shleifer, A. and Vishny. R. 1988 . Management ownership and market valuation: An empirical analysis. Journal of Financial Economics 20 : 293 – 315 . Google Scholar CrossRef Search ADS Nie, H., Jiang, T., and Yang, R., 2012 . A review and reflection on the use and abuse of Chinese Industrial Enterprises Database (Chinese). World Economy 5 , 142 – 158 . Persson, P., and Zhuravskaya. E. 2015 . The limits of career concerns in federalism: Evidence from China. Journal of the European Economic Association 14 : 338 – 74 . Google Scholar CrossRef Search ADS Porta, R., and López-de-Silanes. F. 1999 . The benefits of privatization: Evidence from Mexico. Quarterly Journal of Economics 4 : 1193 – 1242 . Google Scholar CrossRef Search ADS Sun, Q., and Tong. W. 2003 . China share issue privatization: The extent of its success. Journal of Financial Economics 70 : 183 – 222 . Google Scholar CrossRef Search ADS Tian, L., and Estrin. S. 2010 . Retained state shareholding in Chinese PLCs: Does government ownership always reduce corporate value? Journal of Comparative Economics 36 : 74 – 89 . Google Scholar CrossRef Search ADS Wang, X., Xu, L. and Zhu. T. 2004 . State-owned enterprises going public: The case of China. Economics of Transition 12 : 467 – 87 . Google Scholar CrossRef Search ADS Xu, C. 2011 . The fundamental institutions of China’s reforms and development. Journal of Economic Literature 49 : 1076 – 1151 . Google Scholar CrossRef Search ADS Yusuf, S., Nabeshima, K. and Perkins. D. 2005 . Under new ownership: Privatizing China’s state-owned enterprises . Washington, DC : World Bank . Google Scholar CrossRef Search ADS Footnotes 1 As will be discussed in Section 1.1, 5.7 trillion RMB, or roughly 700 billion USD based on the exchange rate at the time, is a conservative estimate of total amount of privatized industrial assets. 2 For example, see the surveys by Megginson and Netter (2001) and Estrin et al. (2009) for privatization in transition economies, such as Central and Eastern Europe and Commonwealth of Independent States (CEE-CIS), Mexico, India, and Brazil. 3 Our findings on the role of transferring control rights in performance improvement are consistent with these results, because the SIP does not transfer control rights (see the next section). The literature, however, disagrees on the impact of the remaining state shares on performance (Sun and Tong 2003; Li, Sun, and Zou 2009; Tian and Estrin 2010). Estrin et al. (2009) summarize that “in China the results to date are less clear cut.” The mixed results highlight two identification challenges. First, other than Deng, Gan, and He (2010), who emphasize expropriation as the driver of impaired performance, the studies do not identify the mechanism and are subject to endogeneity problems. Second, the studies often cannot sharply identify privatization. Some infer privatization from census data by looking at changes in the registration of the firms, which, as our survey reveals, may suffer from type II errors (see the Appendix), whereas others have to rely on small and/or nonrepresentative samples (e.g., Li and Rozelle 2000; Wang, Xu, and Zhu 2004; Guo and Yao 2005; Yusuf, Nabeshima, and Perkins 2005; Dong, Putterman, and Unel 2006). 4 This term is first used by Xu (2011) in summarizing the literature on the political economy of China. It has then been used in the subsequent literature (e.g., Jia, Kudamatsu, and Seim 2015). 5 Nationwide, in 1998, the state sector incurred a total loss of 307 bn RMB, and the overwhelming bad-loan problem associated with these losses was regarded as the biggest threat to the economy (Xu 2011). 6 Because of an ideological aversion to capitalism, the term “privatization” was never used in the official documents; instead, government documents used the term “gaizhi,” meaning “transforming the system.” 7 Relevant reports include 1996 and 2001 Bulletin of Nation Census of Basic Economic Units, and 2004, 2008, and 2013 Bulletin of Nation Economic Census. They can be found at the NSB website http://www.stats.gov.cn/tjsj/pcsj/. 8 Another often-mentioned gaizhi measure was internal restructuring, including incorporation, spinning off, introducing new investors, and debt–equity swaps, as well as bankruptcy/reorganization. Internal restructuring often involved partial privatization but also may involve no privatization when the restructuring occurs among state-owned firms. The latter case was concentrated in large-scale SOEs owned by the central government, and they enjoy monopolistic powers in such markets as oil, electricity, and telecommunication. 9 Based on numbers reported by Huyghebaert and Quan (2009), SIP (excluding financial firms) between 1995 and 2005 involved 539 billion RMB of assets. As discussed in Section 1.1, 5.7 trillion RMB of industrial assets is privatized, implying that SIP account for less than 10% of privatized assets. 10 For the other method, the total ownership shares of the largest shareholders and the second- and third-largest shareholders are above 100%, because the ownership of the latter is based on the subsample that reports this information. 11 The odds ratio is exp($$\beta )$$ times that of non-MBO firms, where $$\beta $$ is the coefficient on MBO. 12 See http://china.findlaw.cn/lawyers/article/d28876.html and http://finance.sina.com.cn/chanjing/b/20120117/185911224916.shtml for rules (in Chinese) governing the power of Employee Representative Congress in Shanghai and Shijiazhuang. In both cities, the Employee Representative Congress must approve layoffs. 13 In our interviews, we found that employment concern, at the initial period of privatization, is also part of the reason why a significant portion of SOEs were privatized through employee shareholding, which avoids dispute between the firm and the employees. Later, because employee shareholding could not achieve efficiency, many of these firms introduced a second round of privatization through MBOs. 14 We find that in the NSB data, the names of the cities in which firms are located are sometimes missing, involving 149 firms and 1040 firm-years. We manually identify the cities based on company names or addresses. Sometimes only county information is available; we then manually match the county with the city to which it belongs to. In a few cases in which the addresses do not contain city or county information, we search the Internet for information, assuming there is no change in company location. 15 We report LIML results because they are more robust to weak IV. A simple IV estimation yields very similar results and is available upon request. 16 The exception is perhaps more recent studies finding that political connections in privatized firms hinder performance enhancement (Boubakri, Cosset, and Saffar 2008) and that state ownership is associated with less accounting transparency (Guedhami, Pittman, and Saffar 2009), both of which consistent with the Chinese experiences. © The Author 2017. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Review of Financial Studies Oxford University Press

Decentralized Privatization and Change of Control Rights in China

Loading next page...
 
/lp/ou_press/decentralized-privatization-and-change-of-control-rights-in-china-7ILEqvnkNj
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.
ISSN
0893-9454
eISSN
1465-7368
D.O.I.
10.1093/rfs/hhx100
Publisher site
See Article on Publisher Site

Abstract

The design and implementation of privatization in China is unique in that both are decentralized and administered by the local governments. Based on a proprietary survey data set containing 3,000 firms in over 200 cities, this paper studies privatization choices and outcomes, as well as the mechanism behind the outcomes. We find that less political opposition to labor downsizing and greater fiscal capacity prompt cities to choose direct sales to insiders (MBOs). This method transfers control rights to private owners, retains limited government supports, imposes hardened budget constraints, allows for restructuring, and achieves performance improvement. Received September 8, 2015; editorial decision February 3, 2017 by Editor Andrew Karolyi. Privatization in China between the late 1990s and the mid-2000s was arguably the largest in the world and still influences governance in the Chinese economy in a profound way.1 Our understanding of this vast transformation, however, remains limited, because there is little data, other than that available from the small fraction of firms that underwent share issue privatization (SIP) and became publicly listed. A distinct feature of privatization in China is that both its design and its implementation are highly decentralized and are administered by the local governments. This feature is in contrast to privatization in most other nations that followed a nationwide policy and implemented privatization in a top-down manner.2 No de jure national privatization policy took place in China. Instead, a few city governments first initiated China’s de facto privatization at a time when the central government was cautious about privatization. Later, after the central government endorsed the practice of selling state-owned enterprise (SOE) assets to private owners, for most SOEs, city governments decided whether to privatize, and, if the decision was yes, what privatization approach to adopt. As a result, privatization methods across Chinese cities widely varied. This decentralized feature of privatization in China is not only critically important for understanding the Chinese economy but also provides a rich setting in which to study privatization and institutions in general. We design and conduct a large-scale nationwide survey of 3,000 firms in more than 200 cities. These proprietary survey data allows us to carry out a systematic study of decentralized privatization in China, in an attempt to draw implications for privatization design and, more generally, the design of economic institutions. We seek to understand how local governments choose different privatization methods and how these various methods are responsible for transferring control rights of the firms and lead to diverse mechanisms with respect to restructuring and performance. Specifically, we ask the following questions: How do different privatization methods reallocate control rights among the stakeholders of the firm? Why do city governments choose a particular privatization method? Do firms still obtain favorable treatment and soft budget constraint after privatization? Which methods result in more effective post-privatization restructuring and which better enhances performance? We collect comprehensive information about control rights reallocation, including details about distribution of eight distinctive decision rights among five parties before and after privatization. Our data shows that, while privatization in China has made substantial progress in reallocating control rights from the government to private owners, the degree of the remaining government influence on corporate decisions significantly varies across privatization methods. These methods include direct sales, either to insiders (through management buyouts, or MBOs hereafter) or to outsider private owners, public offerings, joint ventures, leasing, and employee shareholdings. Among these methods, MBOs transfer the most control rights to private owners and account for close to half of all privatization programs. Accordingly, the government provides the least support, in the forms of subsidies, bank financing, and protected entry, to these MBOs, while imposing the most hardened budget constraint. Our analysis further indicates that the decisions city governments make about how to privatize are critically determined by political and fiscal constraints, and a city government’s choice among privatization approaches profoundly affects the governance and performance of privatized firms. Specifically, when cities face less political opposition to labor downsizing and have stronger fiscal capacity, they tend to choose MBOs. Consistent with private owners’ enhanced incentives to make changes, MBOs are most effective in implementing restructuring measures, including a change of core management teams, strengthening of managerial incentives through compensation policies, establishing boards of directors, and introducing international accounting and independent auditing. Not surprisingly, the performance of MBO firms significantly improves after privatization by 4.4% in return on assets (ROA) and close to 6,000 RMB, or 750 USD, per employee per year. For other privatization methods, the government tends to retain its influence in key corporate decisions. These firms are less effective in restructuring and do not achieve statistically detectable improvement in performance. A common challenge in the privatization literature on performance comparison is selection bias, which arises because certain types of firms that are likely to have better future performance (e.g., due to stronger fundamentals or better government support) might be purposely chosen for MBOs. A distinctive advantage of our study is that our detailed data allows us to better handle the selection concern, by explicitly examining why firms are chosen for MBOs, as well as the mechanisms of performance improvements, which is perhaps the strongest guard against endogeneity. To rule out the selection bias even further, we conduct a number of additional analyses, including examining whether there is any preexisting trend in performance, fully accounting for city-level economic prospects by including city-year fixed effects, explicitly controlling for product market competition, and adopting an IV estimation using city characteristics (such as fiscal capacity and private sector development) as the instruments. Our Chinese survey contributes to the literature in a number of ways. First, it supports and significantly extends an important theme in the literature; that is, the types of owners have disparate effects on restructuring and performance; thus looking only at aggregate results without knowing why could be misleading (Frydman et al. 1999; Estrin et al. 2009). A well-known, but puzzling, result from other transition economies is that privatization to managers does not result in efficiency gains in transition economies. This result appears to be in contrast with our findings that MBOs are the most effective means of privatization in China. The difference lies in the extent to which managerial ownership is market based. Frydman et al. (1999) propose an explanation for the ineffectiveness of privatization to managers in CEE and CIS nations, that is, the two “special characteristics” of managerial ownership. Specifically, managers are selected under the old regime and they are offered to buy the shares at preferential prices, but with restrictive terms designed to favor existing employees. Chinese MBOs do not share these characteristics and have much more in common with managerial ownership in market economies. In this sense, Chinese MBOs constitute a nice counterfactual analysis for other transition economies (and vice versa). They confirm the conjectures in the literature regarding why managerial ownership does not work in CEE-CIS nations. Our paper extends beyond the question of the type of owners and illustrates how the market-based managerial ownership in China improves performance by aligning other economic forces, namely, the role of managers, product market competition, and hardened budget constraint, that have been found to be important in shaping privatization outcomes (e.g., Djankov and Murrel 2002). To our knowledge, no prior work has answered, in one study, these questions as comprehensively as we have. Moreover, our Chinese survey offers advantages in dealing with measurement and identification challenges. Our second contribution is that we explore two important aspects of privatization that the previous literature has not examined. Most notably, enabled by our detailed data, we shed new light on the privatization mechanism through the reallocation of control rights. As Jones and Mygind (1999) and Gupta (2005) point out, a common feature of privatization around the world is that transferring of control rights is incomplete, meaning that the state retains a significant ownership during privatization. Thus, our finding regarding the impact of remained state ownership and control is quite general. Another aspect is the role of political factors in shaping the design of privatization programs. Despite that theoretical work and anecdotes that all suggest a significant influence of political factors (e.g., Biais and Perotti 2002), few formal empirical papers study this important topic. Our paper joins a more recent effort (e.g., Dinc and Gupta 2011) in this regard. In the Chinese setting, political economy considerations, specifically a lack of fiscal resources and political opposition to unemployment, prevent the state from withdrawing control and adopting a more effective privatization method. Finally, our analysis extends earlier work on privatization in China and deepens our understanding of the Chinese economy. Previous work has documented the ineffectiveness of share issue privatization (SIP) (Sun and Tong 2003; Deng, Gan, and He 2010), a lack of a significant effect of privatization on performance (Jefferson and Su 2006), and the importance of reducing state ownership in privatized firms for performance improvement (Bai, Lu, and Tao 2009).3 Our data permit us to cover a wide spectrum of privatization methods and to extend beyond performance comparison by identifying the mechanisms of performance improvement (or a lack of it). Equally importantly, the decentralized privatization studied in this paper contributes to a growing literature on China’s regionally decentralized authoritarian regime, particularly on local governments’ decisions and career concerns (Maskin, Qian, and Xu 2000; Li and Zhou 2005; Jin, Qian, and Weingast 2005; Xu 2011; Jia, Kudamatsu, and Seim 2015; Persson and Zhuravskaya 2015). 1. Institutional Background of Decentralized Privatization in China In this section, we first discuss how decentralized privatization in China evolved alongside the country’s political and economic background in the 1990s. Then we introduce the different privatization methods adopted by the local governments. Finally, we discuss government considerations about MBOs, the most effective privatization method. 1.1 Political and economic background In the governance system of the Chinese economy, political and personnel decisions are highly centralized, and the central government appoints and assesses local government officials, whereas administrative and economic matters, including those of the SOEs, are mostly decentralized to local governments. Such a system is termed by some scholars as “regionally decentralized authoritarianism (RDA)”.4 Under the RDA regime, the control rights of SOEs, except for the very large ones, are assigned to municipal governments, giving them the residual claims to enterprise earnings (Granick 1990; Li 1997). This means that local SOEs were very important for city government officials, both as a source of fiscal revenue and as a contributor to growth in local gross domestic product (GDP), a critical criteria used by upper-level governments in personnel promotion decisions (Maskin, Qian, and Xu 2000; Xu 2011). Endowed with the “local” ownership of SOEs, China’s state sector reforms have been mostly driven by local experiments, sometimes even before the central government’s official mandates (Xu 2011). By early 1990s, the deteriorating performance of SOEs placed increasing pressure on the fiscal conditions of local governments. A few cities “quietly” initiated de facto privatization, without explicit approval from upper-level governments. One of the first local privatization attempts was made in Zhucheng, a city in Shandong province. In 1992, more than two-thirds of the SOEs experienced losses amounting to over 18 months of the city government’s fiscal revenue. The city government then sold many SOEs within its jurisdiction to the employees of these SOEs. Another example is Shunde in Guangdong, where the city government encountered a serious debt problem before it privatized most of its state and collective firms in 1992. When these experiments became publicly known, the central government did not prohibit the practice, which was interpreted as an implicit approval (Garnaut, Song, and Yao 2008). The continued deterioration of the state sector’s financial performance severely strained the country’s banking system.5 The central government gradually accepted privatization as a remedy for the country’s ailing SOEs, as indicated in a number of progressively market-based reform policies. In 1993, the 3rd Plenum of the 14th Communist Party Congress endorsed a principle of diversifying the ownership structure of state-owned firms. In 1995, the central government announced the famous policy of “retaining the large, releasing the small” (zhuada fangxiao). That is, the state was to keep a few hundred of the largest SOEs in strategic industries; for the remaining smaller local SOEs, which constituted the vast majority of SOEs, the stated intention was to let competitive forces make them more efficient. Finally, 15th Communist Party Congress (1997) further approved privatization, granting de jure ownership of local SOEs to local governments and authorizing the “owners,” mostly city governments, of SOEs to design and implement privatization on their own.6 Thus, China has no centrally designed nationwide privatization program, a fact that makes its brand of privatization distinctively different from that in the rest of the world. This wave of privatization ended in 2005, both because the vast majority of SOEs had been privatized by then and because of the publicized controversies over some of the privatization programs in 2004 and in 2005. Explicit statistics on the percentage of all SOEs privatized by 2005 is not available but, according to various reports of National Statistics Bureau (NSB), close to three-quarters of large and medium industrial SOEs were privatized.7 Consistent with the policy of “retaining the large, releasing the small,” our reading of available city-level statistics shows that about 85% of SOEs were privatized by 2005. If we use three-quarters as a conservative estimate of proportion of firms privatized, given that total industrial SOE assets at the end of 1999 was 7.6 trillion RMB, we estimate that the total privatized assets amounted to 5.7 trillion RMB, or roughly 700 billion USD, based on the exchange rate at the time. 1.2 Privatization methods Our data shows that the most popular method was direct sales (or open sales), to insiders or to outside private owners, which, respectively, accounted for 47% and 22% of all privatization programs. Other methods included public offering(1%), joint ventures(2%), leasing(8%), and employee shareholding(10%). These patterns are consistent with those in Garnaut, Song, and Yao (2008).8 Under direct sales, the firm was openly sold to insiders (through MBOs) or outside private owners through auctions or negotiations between the local government and the potential buyers. Although we later find that MBOs were the most effective in improving efficiency, it was the most controversial method, mainly because of its lack of transparency and public concern that state assets may have been sold too cheaply. Public offering refers to share issue privatization (SIP). Under the policy of “retaining the large, releasing the small,” large SOEs were privatized through SIP. By design, SIP did not involve transferring control rights, and only noncontrolling shares were sold in the public capital market. SIP accounted for a tiny proportion (1% according to our survey) in terms of the number of firms, and we estimate that SIPs accounted for around 10% of privatized assets.9 Nevertheless, SIPs have been the most-studied type of privatization in China simply because of the availability of data. Joint venture or merger involved privatization in which an SOE formed a joint venture or merged with a private domestic or foreign firm. Under leasing, the company was leased to the management, employees, outside private firms, or other SOEs. In most cases, it involved inside managers as the lessees, and the firms are often privatized later through MBOs. Employee shareholding converted the company into a limited liability company or cooperative. It was one of the most important gaizhi measures employed at the early stage of local experiments, both because the central government required that each privatization plan be approved by employees (other than corporate executives) and because shares were often offered as a compensation for removing employees’ “tenured” state-employment status. As our data verify, at later stages of gaizhi, managers often purchased the majority of employee shares, thereby qualifying the firms as MBOs. 1.3 Government considerations for MBOs To further understand the government’s considerations regarding MBOs, we choose 32 cities with the most MBOs and the least MBOs and reviewed all the publically available documents related to MBOs decisions. Across all the cities, the governments shared similar concerns and, as a result, they typically stipulated against MBOs in three types of firms: (1) firms with government-granted monopolistic permits to operate; (2) firms with government subsidies because of their responsibilities for social welfare; and (3) firms that obtained land or other resources whose value could not be easily assessed. As a result, small firms were often targeted to be “liberalized” and encouraged to be sold to managers. These patterns were perfectly consistent with what we later find in the data about post-privatization government support of MBOs (Section 3.2) and determinants of MBO choices (Section 4). 2. Nationwide Survey and Sample 2.1 Nationwide survey Our large-scale nationwide survey was conducted in 2006. The sampling procedure involved two steps. We started with the 2004 National Bureau of Statistics (NBS) census, which contained all industrial firms with sales above 5 million RMB as the population and drew a random sample of 11,000 firms stratified by region, industry, size, and ownership type. Given that only 20% of firms in the 2004 population were SOEs and our intention was to study privatization, we supplemented the main survey sample with an additional random sample of 5,500 from the 1998 NBS database, again stratified based on region, industry, and size. We chose to use the 1998 NBS data because 1998 was the first year the database was available, and large-scale privatization in China started in the late 1990s. Thus, the 1998 population maximized our chance of including SOEs not yet privatized. In total, we have 16,500 firms in the survey. We designed the questionnaires through an “iterated” process. We started with a pilot survey of 720 firms in nine cities, including Beijing, Laizhou in Shandong province, Taizhou and Changxing in Zhejiang province, Changchun and Jilin in Jilin province, Shijiazhuang, Pingshan, and Tangshan in Hebei province. It was conducted through both on-site interviews and telephone interviews. This pilot survey improves our survey design and later guides our empirical analysis. For example, because of the controversy surrounding MBOs, many of the MBO firms “disguised” themselves by reporting themselves as other less controversial methods, such as employee shareholdings. Thus, in our empirical analysis, we verify each firm’s self-reported privatization methods with its answers to questions on changes in ownership. In soliciting certain sensitive financial variables, instead of asking for the information directly, we experimented with using multiple-choice questions (of percentage intervals), and the response rate substantially increased. The main survey was conducted through telephone interviews. We hired a professional survey company that had a close relationship with the NBS and had previously helped it conduct its own surveys. We spent a week training the survey company’s staff to understand each question. Throughout the survey, we closely worked with the staff and carefully supervised the process. The chief executives of the firms (or their representatives), the chief accountants, or the heads of human resources answered the questions. To facilitate a difference-in-differences analysis, we prepared two sets of questionnaires: one for privatized firms (the “treatment” group) and one for all other firms (including the “control” group). The survey asked every firm whether it was privatized and accordingly used the appropriate questionnaire. The two sets of questionnaires were identical, except that for privatized firms (1) we asked questions related to privatization, for example, the year in which the firm was privatized and the privatization method; and (2) for questions on ownership and control, we asked the firms to provide information on both the pre- and post-privatization periods. Appendix 1 contains the survey questions relevant to this study. We obtained 3,132 responses, yielding a response rate of 19%. Our survey sample contains 899 privatized firms, 475 nonprivatized SOEs and collectively owned enterprises (nonprivatized SOEs hereafter), and 1,758 de novo private firms. In our survey, we do not notice any systematic selection bias of firms that responded to our survey. Indeed, as reported in Table 1, our survey sample matches the distribution of the population reasonably well in terms of both region and industry. The size distribution of our sample is skewed toward larger firms because we purposely oversampled SOE firms, which tend to be larger for this study. Figure 1A further shows the regional distribution of the privatization sample is roughly in line with the presence of SOEs in the country. Figure 1B reports the staggered nature of privatization by region (Appendix 2 shows the breakdown by province). Figure 1 View largeDownload slide Regional distribution of privatized firms in the survey Figure 1 View largeDownload slide Regional distribution of privatized firms in the survey Table 1 Sample distribution of ownership, size, location, and industry Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% This table compares the distribution of our survey sample with that of the population by size, location, and industry. North China includes: Beijing, Tianjin, and Hebei; North-east: Heilongjiang, Jilin, and Liaoning; North-west: Xinjiang, Qinghai, Ningxia, Gansu, Shaanxi, and Innermongolia; North-central: Shanxi, Henan, and Shandong; South-west: Xizang, Yunnan, Guizhou, Sichuan, and Chongqing; East: Shanghai, Jiangsu, and Zhejiang; South: Guangxi, Guangdong, Fujian, and Hainan; and South-central: Hubei, Hunan, Jiangxi, and Anhui. Table 1 Sample distribution of ownership, size, location, and industry Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% Survey sample Population (1) (2) A. Size distribution Large 3% 1% Medium 17% 11% Small 80% 88% B. Regional distribution North 10% 8% North-east 7% 7% North-west 5% 4% North-central 16% 15% South-west 6% 5% East 34% 35% South 14% 18% South-central 8% 8% C. Industry distribution Mining 9% 12% Food, beverage, and tobacco 9% 9% Textiles 12% 15% Timber and paper products Petroleum and chemical 17% 15% Metals 21% 21% Machine and electonics 17% 16% Electricity, gas, and water 6% 3% This table compares the distribution of our survey sample with that of the population by size, location, and industry. North China includes: Beijing, Tianjin, and Hebei; North-east: Heilongjiang, Jilin, and Liaoning; North-west: Xinjiang, Qinghai, Ningxia, Gansu, Shaanxi, and Innermongolia; North-central: Shanxi, Henan, and Shandong; South-west: Xizang, Yunnan, Guizhou, Sichuan, and Chongqing; East: Shanghai, Jiangsu, and Zhejiang; South: Guangxi, Guangdong, Fujian, and Hainan; and South-central: Hubei, Hunan, Jiangxi, and Anhui. 2.2 Data We obtain the financial information of surveyed firms from the NSB database, which is equivalent to Compustat for U.S.-listed firms. NSB data are available from 1998 to 2007. While it is the most comprehensive data about Chinese firms, some scholars have questioned its data quality (e.g., Nie, Jiang, and Yang, 2012). Appendix 3 examines the NSB data in detail and demonstrated that their weakness does not significantly affect our results. To ensure all privatized firms have at least one year of performance information prior to privatization, we drop 168 firms that were privatized prior to 1999. We then exclude firms without valid financial information. Given the staggered nature of privatization, our final sample for regression analyses is an unbalanced panel of 717 privatized firms, 460 SOEs that have not been privatized, and 1,685 de novo private firms for the period of 1998–2007. In our analysis of the role of government incentives in privatization decisions, we use the China City Statistical Yearbook to obtain city-level (at and above the prefecture level) fiscal and regional economic variables from 1997 to 2007. We note that, while the data may seem old, they are suitable to study the largest wave of privatization in China (and worldwide), for two reasons. First, we conducted the survey in 2006, while this wave of privatization ended in 2005 (see discussions in Section 1.1). Second, the survey data can be merged with 10 years of NSB data during 1998–2007 and doing so allows us to study performance before and after privatization. It is well known among scholars studying China that the quality of data available to researchers is low in 2008 and in 2009, and that, due to tightened control of data, it is almost impossible to obtain the data after 2009. Thus, it is a nice coincidence that privatization occurred before the end of 2005 and quality NSB financial data are available until 2007, a fact enabling us to cover this historical episode well and to the best extent. 2.3 Preliminary observations from our sample Table 2 reports the summary statistics of the main variables used in our empirical analysis. In panel A of Table 2, we report the basic facts about privatization in China. Between 2000 and 2005, the number of privatizations increases steadily. Direct sales to insiders (MBOs) are, by far, the most widely used method, accounting for 47% of all privatized firms. The next is direct sales to outsiders, accounting for 22% of the firms. Thus, direct sales in total account for close to 70% of privatization programs in China. Other privatization methods include public offerings (1%), joint ventures (2%), leasing (8%), and employee shareholding (10%). Table 2 Basic facts and summary statistics A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** In panel A3, differences between the MBO firms and other methods and between direct sales to outsiders and other methods are tested. Profits are defined as earnings before interest, tax, and depreciation. Significance levels are all based on two-tailed tests of differences. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 2 Basic facts and summary statistics A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A. Basic facts about privatization in China, 1999–2005 A1. Year of privatization Year # firms Percentage 1999 60 8 2000 103 14 2001 102 14 2002 109 15 2003 129 18 2004 95 13 2005 119 17 A2. Methods # firms Percentage Direct sales $$\quad$$ To insiders (MBO) 338 47 $$\quad$$ To outsiders 157 22 Other methods $$\quad$$ Public offering 8 1 $$\quad$$ Joint venture 11 2 $$\quad$$ Leasing 56 8 $$\quad$$ Employee holding 70 10 $$\quad$$ Others 77 11 Total 717 100 A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% A3. Ownership of privatized firms MBO Direct sales to outsiders Others All Ownership by the largest shareholder Mean 37%*** 64% 91%*** 60% Median 30%*** 70% 100%*** 51% Ownership by the second and third Mean 27%** 20%*** 30%* 26% $$\quad$$ largest shareholder Median 22%** 15%*** 30%** 20% B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** B. Financial information from Chinese firms, 1998–2007 B1. Overview of financial information of Chinese firms State-owned enterprises (SOEs) Whole sample Privatized Nonprivatized Difference Non-SOEs Difference (1) (2) (3) (2)$$-$$(3) (4) (2)$$-$$(4) Assets (RMB ’000) Mean 181,801 354,285 252,388 101,897$$^{***}$$ 51,996 $$-$$302,289$$^{***}$$ Median 26,250 58,023 45,903 12,120$$^{***}$$ 15,926 42,097$$^{***}$$ Sales (RMB ’000) Mean 135,102 239,621 155,860 83,761$$^{***}$$ 64,706 174,914$$^{***}$$ Median 23,911 31,060 23,311 7,749$$^{***}$$ 21,395 9,665$$^{***}$$ Leverage Mean 0.085 0.129 0.136 –0.006 0.040 0.090*** Median 0.001 0.051 0.041 0.010 0.000 0.051*** Profits/Assets Mean 0.132 0.091 0.069 0.022*** 0.180 –0.088*** Median 0.077 0.051 0.041 0.010*** 0.109 –0.058*** Profits/# employee Mean 30.658 13.446 23.769 –10.323 43.393 –29.948*** $$\quad$$(RMB ’000) Median 11.876 8.387 5.837 2.550*** 17.000 –8.613*** Number of firm-years 17,609 5,340 3,351 8,918 B2. Financial variables before and after privatization All privatized SOEs MBO Before After Difference Before After Difference (1) (2) (2)$$-$$(1) (3) (4) (4)$$-$$(3) Assets (RMB ’000) Mean 260,276 449,856 189,580$$^{***}$$ 117,114 195,703 78,589$$^{***}$$ Median 54,706 61,084 6,378$$^{***}$$ 44,237 43,215 $$-$$1,022 Sales (RMB ’000) Mean 155,549 325,057 169,509$$^{***}$$ 77,595 178,131 100,536$$^{***}$$ Median 24,685 40,235 15,551$$^{***}$$ 22,121 30,390 8,269$$^{***}$$ Leverage Mean 0.143 0.115 –0.028*** 0.132 0.102 –0.030*** Median 0.072 0.031 –0.041*** 0.069 0.021 –0.047*** Profits/Assets Mean 0.054 0.128 0.074*** 0.047 0.153 0.106*** Median 0.039 0.068 0.030*** 0.036 0.078 0.043*** Profits/# employee Mean 10.963 15.898 4.934 7.901 2.659 –5.241 $$\quad$$ (RMB ’000) Median 5.242 14.616 9.374*** 4.449 14.743 10.293*** In panel A3, differences between the MBO firms and other methods and between direct sales to outsiders and other methods are tested. Profits are defined as earnings before interest, tax, and depreciation. Significance levels are all based on two-tailed tests of differences. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. The ownership structure of Chinese privatized firms is highly concentrated. The largest shareholders, on average, hold 60% of the shares and the second- and third-largest shareholders hold 26% of shares. MBOs have the lowest ownership concentration, with the largest shareholders holding 37% of the shares, whereas the largest shareholder of the firms sold to outsiders has 64% ownership, on average. For firms privatized by other methods, the largest shareholders, on average, hold 91% of the shares.10 Panel B is a summary of the financial variables. We use two measures of operating performance: operating profits (earnings before interest, tax, depreciation, and amortization, or EBITDA) over assets, and operating profits over the number of employees. Panel B1 compares privatized, nonprivatized, and de novo nonstate (private) firms. Compared with nonprivatized SOEs, privatized firms tend to be larger and generally exhibit greater operating efficiency. Later, we will show that this is due to post-privatization performance. Compared with de novo private firms, privatized SOEs tend to be larger and less profitable. Panel B2 of Table 2 compares the financial variables before and after privatization. Assets and sales generally increased after privatization. Firms tend to become less leveraged after privatization, consistent with a hardened budget constraint. While there is generally an improvement in performance (all at the 1% level, except for the mean of Profits/#employee), performance gain appears to be larger for MBOs, consistent with our later findings that MBOs drive performance gains. 2.4 Financial aspects of privatization We will now discuss the financial aspects of privatization, including the issuance method, payment arrangement and sources of funds for top managers. Other than SIPs, which cover large companies in strategic industries, the transfer of ownership is through secondary offerings of existing shares, consistent with the government’s stated intention of transferring of ownership and of “letting go” of these companies. As reported in Table 3, in 77% of privatization cases, the government receives a lump sum payment, as opposed to multiyear installments. Reflecting a greater transfer of ownership, MBOs are significantly more likely to be paid with lump sum payments (80%), whereas leasing is least likely to use this arrangement. If multiple installments are used, the first payment, on average, accounts for one third of the total proceeds and it takes about 5 years to pay the full amount. Table 3 Financial aspects of privatization A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% In panel A, significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5% and 10% level is indicated by ***, ** and *, respectively. In panel C, for each source of funds, firms are asked to specify the percentage of funding from this source. The possible answers are: 0%, 1%-20%, 21%-40%, 41%-70% and 71%-100%. To estimate the monetary share of each of the financing source, we assume that the median of the range is the actual percentage. Table 3 Financial aspects of privatization A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% A. Details of privatization payment schemes All privatized SOEs MBO Direct sales to outsiders Leasing % lump-sum cash payment 77 80** 74 55*** % first payment if by installment 33 34 28** 33 # years to pay if by installment 4.9 4.7 5.5 4.7 B. Sources of managers’ funds Personal savings Borrowings from friends and relatives Future Bank loans Future salaries % firms using this method 99 8 5 6 % as an above 20% source 99 2 2 3 % as an above 70% source 95 0 0 0.2 C. Estimated monetary share of each source of funds Personal savings 96% Borrowings from relatives 1% Bank loans 1% Future salaries 2% In panel A, significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5% and 10% level is indicated by ***, ** and *, respectively. In panel C, for each source of funds, firms are asked to specify the percentage of funding from this source. The possible answers are: 0%, 1%-20%, 21%-40%, 41%-70% and 71%-100%. To estimate the monetary share of each of the financing source, we assume that the median of the range is the actual percentage. Personal savings are predominately the most important source of funds by the top managers, used in 99% of firms. Ninety-five percent of firms report that personal savings account for at least 70% of financing (panels A and B), and we further estimate that they contribute to 96% of all privatization payments (panel C). Other sources of financing include borrowings from friends and relatives, bank loans, and future salaries by 8%, 5%, and 6% of firms, respectively and each account for 1% to 2% of total payments. 3. Mechanisms of Efficiency Gain The essence of ownership structure is its allocation of control rights among the firms’ stakeholders (Grossman and Hart 1986; Hart and Moore 1990). This section investigates reallocation of control rights as the mechanism of performance gain, and the resultant government support and freedom to restructure. 3.1 Reallocation of control rights and performance We find that the government retains, on average, 20% ownership of the privatized firms. Although this figure is much lower than that for share issue privatization, in which the government retains more than half of the ownership, 20% is still a figure significant enough for the state to exert influence. Reflecting the concept of property rights as a bundle of rights, we focus on a set of eight decision rights, including the appointment of senior managers, investment, hiring and laying off of employees, salary and bonus, distribution of profits, production and marketing, financing, and use of funds. We ask how these control rights are allocated, before privatization and after privatization, among five parties, including the government, the party committee at the firm, board of directors, general manager, workers representative committee, board of supervisors, and shareholder committee in making the above-mentioned key corporate decisions. The firms rank, for each of the corporate decisions, the importance of each decision maker on a five-point scale (0 $$=$$negligibly unimportant, 5 $$=$$indispensably important). As shown in Figure 2 and Table 4, the most prominent change in control rights is the reduction of government influence. For nonprivatized SOEs and pre-privatization SOEs, local governments exercise fairly strong control over these firms’ major decisions, with average scores of 2.3 and 1.8, respectively, and the government’s control rights are particularly strong in the appointment of top management, scoring 3 and 2.4 (panel A of Table 4). By contrast, the government has no control power over decisions within de novo private firms. After privatization, both the overall government control and its control in personnel drop substantially, from 1.8 to 0.4 and from 2.4 to 0.6, respectively. Moreover, the government control decreases the most for MBOs, with the average score dropping from 1.8 to 0.1. Direct sales to outsiders come the second, with average government control decreasing from 1.9 to 0.4. Figure 2 View largeDownload slide Reallocation of control rights before and after privatization Figure 2 View largeDownload slide Reallocation of control rights before and after privatization Table 4 Privatization and change of control rights Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 This table reports allocation of control rights in Chinese firms. The importance of each decision maker is given a score from 0 to 5, where 0 means negligibly unimportant and 5 indispensably important. Average and median scores across firms are reported. Significance levels in Columns (4), (6) and (8) are based on two-tailed tests of differences before and after privatization. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively, and na refers to not applicable. Table 4 Privatization and change of control rights Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 Privatization methods All privatizatized SOEs MBO Direct sales to outsiders Nonprivatized SOEs De novo private firms Before After Before After Before After (1) (2) (3) (4) (5) (6) (7) (8) Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median Mean Median A. Control rights of government Appointment of top management 3.0 4.0 0.0 0.0 2.4 2.0 0.6*** 0.0*** 2.4 3.0 0.1*** 0.0*** 2.6 2.0 0.4*** 0.0*** Employment/Layoff 2.2 2.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 2.2 2.0 0.5*** 0.0*** Wages/Compensations 1.9 2.0 0.0 0.0 1.6 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Investment 2.6 3.0 0.0 0.0 2.0 2.0 0.4*** 0.0*** 2.0 2.0 0.1*** 0.0*** 1.9 2.0 0.4*** 0.0*** Fund raising 2.4 2.0 0.0 0.0 1.9 0.0 0.4*** 0.0*** 1.9 0.0 0.1*** 0.0*** 1.8 1.0 0.4*** 0.0*** Fund using 2.1 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.6 0.0 0.1*** 0.0*** 1.8 1.0 0.3*** 0.0*** Distribution of profits 2.0 2.0 0.0 0.0 1.7 0.0 0.4*** 0.0*** 1.7 0.0 0.1*** 0.0*** 1.8 0.0 0.4*** 0.0*** Production and marketing 1.8 1.0 0.0 0.0 1.6 0.0 0.3*** 0.0*** 1.5 0.0 0.0*** 0.0*** 1.7 0.0 0.3*** 0.0*** Average 2.3 2.3 0.0 0.0 1.8 0.8 0.4*** 0.0*** 1.8 0.9 0.1*** 0.0*** 1.9 1.1 0.4*** 0.0*** Number of firms 454 1550 717 714 338 337 89 88 B. Control rights of party committee Appointment of top management 2.7 3.0 2.0 2.0 2.8 3.0 1.7*** 2.0*** 2.9 3.0 1.5*** 2.0*** 2.5 3.0 1.3*** 1.0*** Employment/Layoff 2.7 3.0 2.2 2.0 2.8 3.0 1.7*** 2.0*** 3.0 3.0 1.6*** 2.0*** 2.4 3.0 1.3*** 1.0*** Wages/Compensations 2.4 3.0 2.2 2.0 2.7 3.0 1.7*** 2.0*** 2.8 3.0 1.6*** 2.0*** 2.3 2.0 1.3*** 1.0** Investment 2.5 3.0 2.0 2.0 2.2 2.0 1.3*** 1.0*** 2.2 2.0 1.2*** 0.0*** 2.1 2.0 1.1*** 1.0** Fund raising 2.4 3.0 1.7 2.0 2.1 2.0 1.3*** 1.0*** 2.1 2.0 1.2*** 1.0*** 2.2 2.0 1.1*** 1.0*** Fund using 2.3 2.0 1.6 2.0 1.9 2.0 1.2*** 1.0*** 1.9 2.0 1.1*** 0.0*** 2.1 2.0 1.0*** 1.0*** Distribution of profits 2.4 3.0 1.8 2.0 2.5 2.0 1.6*** 2.0*** 2.6 3.0 1.4*** 1.0*** 2.3 2.0 1.2*** 1.0** Production and marketing 2.2 2.0 1.8 2.0 2.4 2.0 1.5*** 1.0*** 2.5 2.0 1.3*** 1.0*** 2.2 2.0 1.1*** 1.0*** Average 2.5 2.8 1.9 2 2.4 2.4 1.5*** 1.5*** 2.5 2.5 1.3*** 1.1*** 2.2 2.3 1 1 Number of firms 320 181 611 611 285 285 67 67 C. Control rights of CEOs Appointment of top management 3.9 4.0 4.3 5.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.1 4.0 4.3** 5.0* Employment/Layoff 4.1 4.0 4.3 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.5 4.0 4.1 4.0 4.3 5.0 Wages/Compensations 4.0 4.0 4.2 5.0 3.7 4.0 3.6** 4.0 3.6 4.0 3.6 4.0 4.2 4.0 4.3 5.0* Investment 3.8 4.0 4.3 5.0 3.2 4.0 3.3* 4.0 2.9 3.0 3.3*** 4.0** 4.1 4.0 4.3 5.0* Fund raising 3.8 4.0 4.0 5.0 3.1 4.0 3.3** 4.0 2.9 3.0 3.2*** 4.0** 3.9 4.0 4.2 5.0* Fund using 3.8 4.0 4.2 5.0 3.1 4.0 3.2 4.0 2.9 3.0 3.2** 4.0** 4.0 4.0 4.2 5.0 Distribution of profits 3.9 4.0 4.2 4.0 3.6 4.0 3.6 4.0 3.4 3.0 3.5 4.0 4.2 4.0 4.4 5.0 Production and marketing 4.0 4.0 4.1 5.0 3.8 4.0 3.7 4.0 3.6 4.0 3.6 4.0 4.3 4.0 4.5 5.0 Average 3.9 4.0 4.2 4.9 3.5 4.0 3.5 4.0 3.3 3.4 3.4 4.0 4.1 5.0 4.3 5.0 Number of firms 466 1667 717 716 338 338 89 88 D. Control rights of boards of directors Appointment of top management 4.5 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.6 5.0 Employment/Layoff 3.9 5.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.3 5.0 Wages/Compensations 3.9 5.0 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 3.9 4.0 n.a. n.a. 4.0 4.0 Investment 4.3 5.0 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 n.a. n.a. 4.7 5.0 Fund raising 4.3 5.0 4.4 5.0 n.a. n.a. 4.5 5.0 n.a. n.a. 4.6 5.0 n.a. n.a. 4.7 5.0 Fund using 4.3 5.0 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.6 5.0 Distribution of profits 4.4 5.0 4.5 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 4.0 n.a. n.a. 4.7 5.0 Production and marketing 3.9 4.5 3.6 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.0 4.0 n.a. n.a. 4.2 4.0 Average 4.2 4.9 4.2 4.6 n.a. n.a. 4.3 4.6 n.a. n.a. 4.3 4.5 n.a. n.a. 4.5 4.8 Number of firms 103 756 n.a. 545 n.a. 285 n.a. 42 E. Control rights of shareholder meetings Appointment of top management 3.4 4.0 3.7 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.5 4.0 n.a. n.a. 3.7 4.0 Employment/Layoff 2.5 3.5 3.1 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.4 4.0 n.a. n.a. 3.6 4.0 Wages/Compensations 2.8 3.0 2.9 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.3 4.0 n.a. n.a. 3.3 4.0 Investment 3.7 4.0 4.0 4.0 n.a. n.a. 4.1 5.0 n.a. n.a. 4.2 5.0 n.a. n.a. 4.1 5.0 Fund raising 3.4 4.0 3.9 4.0 n.a. n.a. 4.3 5.0 n.a. n.a. 4.4 5.0 n.a. n.a. 4.3 5.0 Fund using 3.5 4.0 3.9 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.7 4.0 n.a. n.a. 3.8 4.0 Distribution of profits 3.4 4.0 3.8 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 n.a. n.a. 3.6 4.0 Production and marketing 2.7 3.0 2.8 3.0 n.a. n.a. 3.2 4.0 n.a. n.a. 3.1 3.0 n.a. n.a. 3.4 4.0 Average 3.2 3.8 3.5 3.9 n.a. n.a. 3.7 3.9 n.a. n.a. 3.6 3.9 n.a. n.a. 3.8 4.0 Number of firms 48 376 n.a. 428 n.a. 286 n.a. 91 This table reports allocation of control rights in Chinese firms. The importance of each decision maker is given a score from 0 to 5, where 0 means negligibly unimportant and 5 indispensably important. Average and median scores across firms are reported. Significance levels in Columns (4), (6) and (8) are based on two-tailed tests of differences before and after privatization. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively, and na refers to not applicable. A unique feature of corporate governance in China is that almost all firms in China have a committee of the Chinese Communist Party. As shown in panels A and B of Table 4, the influence of party committees is similar to that of the government. After privatization, the party committees’ control generally decreases less than the government’s control. Given that the government may influence corporate decisions through both its direct control rights and its intervention via firm-level party committees, we use the max of these two as the score for overall state influence (State influence score). Despite a drop in the score from 2.8 to 1.4 after privatization, state influence is still quite important in a significant proportion of firms, with 39% of firms having a score above 2 (somewhat important) and 15% above 3 (moderately important). In the following analysis, we consider firms with State influence score above 2 as under significant state influence in corporate decisions. Across privatization methods, MBO firms have the lowest level of state control. Only 1% of MBO firms have government ownership above 20%, the mean, significantly lower than the sample average of 50% (Table 5). The state is also much less likely to intervene in MBOs’ major decision-making (16% vs. 59% sample mean). Direct sales to outsiders are under substantially more state influence than MBOs, but, compared to privatization methods, they receive less state intervention, though the difference is only significant for corporate decision-making, not for state ownership. Table 5 State control in privatized firms A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% This table reports the percentage of firms that are under strong state influence post-privatization by privatization method. State-influence score is defined as the max of the importance of local government and that of party committee in corporate decision making based on a five-point scale (0=negligibly unimportant, 5=indispensably important). Panel B uses principal component analysis (PCA) to form additional variables of state control. The source of state influence is from government or party communist. PCA State Control is defined as 1 if either the first component of government influence or the first component of party influence is above the mean. Significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 5 State control in privatized firms A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% A. Ownership and state-influence score State ownership above mean State-influence score above 2 Direct sales to insiders (MBO) 1%*** 16%*** Direct sales to outsiders 15% 25%* Other methods 50% 59% All privatizatized SOEs 19% 31% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% Panel B. Principal component analysis of state control First component of government influence % first component of government influence above mean First component of party influence % first component of party influence above mean PCA state control Direct sales to insiders (MBO) 2.71*** 21$$^{***}$$ 5.82*** 44$$^{***}$$ 42%$$^{***}$$ Direct sales to outsiders 3.54* 28$$^{*}$$ 6.18* 47$$^{*}$$ 49% Other methods 3.61 32 6.03 50 59% All privatizatized SOEs 3.40 26 5.96 47 49% This table reports the percentage of firms that are under strong state influence post-privatization by privatization method. State-influence score is defined as the max of the importance of local government and that of party committee in corporate decision making based on a five-point scale (0=negligibly unimportant, 5=indispensably important). Panel B uses principal component analysis (PCA) to form additional variables of state control. The source of state influence is from government or party communist. PCA State Control is defined as 1 if either the first component of government influence or the first component of party influence is above the mean. Significance levels are based on two-tailed tests of differences between a particular privatization method and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Given that corporate decisions are multidimensional, we further examine government and party influences using principal component analysis (PCA). PCA is effective in shrinking dimensionality: the first principal component accounts for 90% and 75% of the government and party influences respectively, whereas the second component accounts for only 4% and 6%, respectively. Thus we report, in panel B of Table 5, the first components and PCA state control, defined as either the first component of government influence or the first component of party influence is above the mean. Consistent with panel A of Table 5, PCA state control is significantly lower in MBOs. Other notable changes in control rights include the increased decision power of the board of directors and shareholder meetings, suggesting a general trend of professionalization of management in privatized firms. This change is most prominent among MBOs. 3.1.1 State control and post-privatization performance This subsection further investigates the impact of state control on post-privatization performance, by estimating the following model on the sample of all privatized firms: \begin{equation} \textit{Performance}_{it} = \alpha_{i} + \beta_{t }+ \gamma \textit{Post}_{it} +\lambda \textit{State Control}_{i} \times \textit{Post}_{it} + \delta X_{it} + \varepsilon_{it}, \end{equation} (1) where Performance$$_{it}$$ is measured by both ROA and earnings per employee. Post$$_{it}$$ is a dummy variable indicating years after privatization (it is set to zero for those SOEs that have never been privatized). State control is one of the three binary variables: state ownership above 20%; State influence score above 2; and PCA state control, defined in the same way as in Table 5. X$$_{it}$$ are firm control variables, including size (measured as log of assets) and leverage (debt over assets). $$\alpha_{i} $$ is a firm fixed effect that controls for time-invariant firm characteristics. $$\beta_{t}$$ is a year fixed effect. Coefficient $$\gamma $$ is the difference-in-differences estimate of the effect of state control on post-privatization performance. Linking detailed measures of government control rights to performance improves upon the existing literature which typically assigns a linear relationship between ownership and performance. Our analysis is similar in spirit to (López-de-Silanes, 1997), who finds, in privatization in Mexico, that transferring of controlling share packages is associated with a higher price premium, an ex ante measure of future performance. Table 6 demonstrates that state control significantly hinders performance of privatized firms. In Columns (1) and (2) of Table 6, higher state ownership is associated with significantly worse post-privatization performance, for both operating efficiency measures (at the 1% levels). In Columns (3) - (6), both the measure based on State influence score above 2 and PCA State control are associated with significantly lower operating efficiency. The results are all economically significant. Take the example of the point estimates of State influence score above 2 (Columns (3) and (4)). They imply that, all else equal, state control in decision making reduces ROA by 6% and earnings per employee by 6,252 RMB (close to 800 USD) per employee per year. Both are substantial, especially considering the sample mean is 12.8% for ROA and 15.9K RMB for earnings per employee. Table 6 State influence and performance Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 This table presents the effect of state control on post-privatization performance like in Equation (1). It is based on the sample of all privatized firms during 1998 to 2007. Table 5 defines variables related to state control. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 6 State influence and performance Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 Performance measures Performance measures Performance measures Profits/assets Profits/# employee Profits/assets Profits/# employee Profits/assets Profits/# employee (1) (2) (3) (4) (5) (6) Lag of performance Log (sales) 0.084*** 18.290*** 0.083*** 18.256*** 0.084*** 18.100*** (0.011) (1.523) (0.011) (1.522) –0.012 –1.525 Leverage 0.003 6.432* 0.007 6.877* 0.004 6.672* (0.018) (3.881) (0.018) (3.891) –0.018 –3.901 Post dummy 0.032*** 1.781 0.018* –0.258 0.031** 0.447 (0.012) (1.833) (0.011) (1.660) –0.013 –1.935 State share above –0.074*** –9.655*** $$\quad$$ mean * Post (0.014) (2.669) State influence score –0.060*** –6.252* $$\quad$$ above 2 * Post (0.020) (3.539) PCA state –0.063*** –5.397** $$\quad$$ control * Post –0.014 –2.589 Year dummies Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,245 5,167 5,245 5,167 5,214 5,136 R-squared 0.518 0.549 0.520 0.550 0.519 0.547 This table presents the effect of state control on post-privatization performance like in Equation (1). It is based on the sample of all privatized firms during 1998 to 2007. Table 5 defines variables related to state control. Performance measures are calculated as operating profits (earnings before interest, tax, and depreciation) over assets and number of employees, respectively. Robust standard errors are in parentheses. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. 3.2 Government support While ownership and decision rights are perhaps the most straightforward measures of government influence, there may be a tangled web of relation between the firm and the government. Specifically, it is possible that government exerts influence through other channels, such as connection of the manager to the party, implicit or explicit subsidies, and regulatory barriers to entry. In our survey, we design questions that allow us to further explore these aspects of state influence. Given the dramatic control change via the MBO route, we mainly focus on the comparison between MBOs and other privatization methods. This analysis will also help understand our later results on MBO performance. Panel A of Table 7 displays the firm’s political connections along three dimensions, namely, whether top officials are appointed by the government, whether the firm has are government officials on the board, whether the top manager is a former government official. It turns out that the strongest form of political connection in China’s privatized firms is through personnel appointment: in 23% firms, the chairman or top manager is appointed by the government. Such connection, however, is much weaker in MBOs involving only 0.3% of the firms and the difference is significant at the 1% level. Political connection in the form of government officials on the board or being the top manager is not common and is in only 4% and 2% of privatized firms respectively. The numbers are even lower among MBOs, involving, respectively, 0.3% and 1% of firms, and the difference is significant at the 1% and 5% levels. Table 7 Comparison of political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatized firms All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 This table presents the comparison of post-privatization political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatization methods. Panel E is based on answers to our survey question: “how many competitors does your firm have?” The possible answers are: no, few, some and many competitors. We categorize the firm as a monopoly if it reports no competitor. It is defined to have market power if it has no or few competitors. It is considered to be in a competitive market if it has some or many competitors. Significance levels are based on two-tailed tests of differences between MBO and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Table 7 Comparison of political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatized firms All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** All privatizatized SOEs MBO (1) (2) A. Political connections % with chairman or top manager appointed by government 23 3*** % with government officials on the board 4 0.3*** % with top manager being a former government official 2 1** B. Government subsidies % with government land subsidy 67 59*** $$\quad$$ % with direct allocation of land by government 31 19*** $$\quad$$ % with purchases at below-market prices 36 40** % with government-funded R&D projects 3 1*** C. Bank loans % with bank loans 82 84 % with loan rejection 22 26** $$\quad$$ % rejected due to constraints on bank credit supply 3 4** $$\quad$$ % rejected due to a lack of relations with the government 3 4* % with government guarantee of loans 7 7 D. Soft budget constraints % expected tax reduction in case of financial distress 0.2 0 % expected government subsidies in case of financial distress 0.3 0 % expected capital injection in case of financial distress 0.4 0 % expected subsidized loans in case of financial distress 0.1 0.3 % with any of the above expectations 0.6 0.3 E. Protected entry E1. Reported competition by MBOs and other privatized firms All privatized SOEs MBO % monopoly 9 2*** % has market power 22 16*** % competitive market 78 84*** E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 E2. Reported competition in industries perceived as protected industries All privatized SOEs MBOs # obs % firms % monopoly % competitive market # obs % firms % monopoly % competitive market Energy 35 4 23 69 10 2 10 80 $$\quad$$ Coal 34 4 21 71 10 2 10 80 $$\quad$$ Oil and natural gas 1 0.1 100 0 0 0 n.a. n.a. Utilities 83 9 67 13 10 2 60 20 $$\quad$$ Power supply 47 5 68 13 7 2 43 29 $$\quad$$ Fuel gas 9 1 44 33 0 0 n.a. n.a. $$\quad$$ Water 27 3 74 7 3 1 100 0 Car 54 6 11 74 18 4 6 78 Pharmacy 37 4 0 86 21 5 0 86 This table presents the comparison of post-privatization political connections, government subsidies, soft budget constraint, and protected entry between MBO and other privatization methods. Panel E is based on answers to our survey question: “how many competitors does your firm have?” The possible answers are: no, few, some and many competitors. We categorize the firm as a monopoly if it reports no competitor. It is defined to have market power if it has no or few competitors. It is considered to be in a competitive market if it has some or many competitors. Significance levels are based on two-tailed tests of differences between MBO and other methods. Significance at the 1%, 5%, and 10% level is indicated by ***, **, and *, respectively. Panel B of Table 7 shows that MBOs receive less government subsidies. Land is the most important government subsidy. MBOs are significantly less likely to obtain land subsidy, 59% versus 67% (a significant difference at the 1% level). The composition of land subsidy is also telling: MBOs are less likely to obtain direct allocation of land (19% versus 31%, significant at 1% level), which represents a large subsidy, whereas they are slightly more likely to purchase land at substantially subsidized prices (40% versus 36%). Government funded R&D projects are not common, involving 3% of the firms. The number is even lower for MBOs, 1%, and the difference is significant at the 1% level. Panel C of Table 7 demonstrates that MBOs receive less government support in financing. While MBOs have a similar likelihood to have bank loans, their loan applications are significantly more likely to be rejected, 26% versus 22% (a significant difference at the 10% level). When asked about the reasons for loan rejection, MBOs are more likely quote bank credit rationing (4% versus 3%) – state-owned banks typically have quarterly or annual limits imposed by their regulatory agents – and a lack of relationship with the government (4% versus 3%). The differences are significant, respectively, at the 5% and 10% level. Finally, there is no difference in the chance of obtaining government loan guarantees between the two groups of firms. Panel D of Table 7 examines soft budget constraints in privatized firms. It should be noted that soft budget is not easy to measure, because the empirical measure has to meet two criteria. One is that it has to capture the expectation of future bailout; the other is that the expectation is contingent on financial distress. Neither is available in standard company financial statements. As noted by Djankov and Murrell (2002), a survey method provides measures that come closest to theoretically prescribed ones. In our survey, we ask about a number of expected supports in case of financial distress, including tax reduction, subsidies, capital injection, and subsidized loans. The data shows that Chinese privatization is very effective in hardening soft budget constraints: each individual form of soft budget involves less than 1% of the firms, and the proportion of firms with any one form of the soft budget is 0.6%. MBOs are even less likely to have soft budget in terms of all forms of support, except for subsidized loans, arguably the weakest form of support. Panel E of Table 7 reports government support in the form of protected entry, based on the question “How many competitors does your firm have?” The possible answers are none, few, some, and many. We categorize the firm as in a competitive market if there are some or many competitors. The vast majority of firms (75%) are in competitive markets. MBOs are even more likely to be in competitive markets, 84%, and the difference is significant at 1%. 14% firms are monopolies with no competitors, whereas significantly less MBOs, a mere 2%, are monopolies (at the 1% level). While most SOE monopolies in China arise from protected entry, it is theoretically possible that the firm has developed or purchased advanced technology. We find that this is not true: only 4% of monopoly firms have patents, much lower than other firms, 30%, and the difference is significant at the 1% level. We further check the market structure of industries that are often perceived as having protected entry, including energy, utilities, car, and pharmaceuticals. It turns out that only utilities seem to possess monopolistic power: an average of 67% firms report themselves as an monopoly and 13% report that the market is competitive. There is only one firm in oil and gas; although it is a monopoly, there is not a big enough sample to make a reliable inference. Taken together, our analysis demonstrates that, after privatization, the government substantially reduce its support and subsidiaries to all firms and particularly so for MBOs. This, however, is not surprising. It is consistent with the guiding rule of “retaining the large, letting go of the small,” where the small ones, which is the vast majority, were generally in competitive sectors. Moreover, given that the government keeps the least ownership and control in MBOs, it is economically rational to provide even less support. 3.3 Post-privatization restructuring and professionalism We ask about four restructuring measures. The first restructuring measure is whether the firm changed its core management team–-the introduction of new human capital into management is shown to be important in improving efficiency in other privatization settings (e.g., Barberis et al. 1996; Lópezde-Silanes; 1997, who emphasizes CEO changes). The second is whether the firm incentivizes its executives through increased performance-based pay. Regarding corporate governance, we ask whether the firm established a board of directors and whether it adopted international accounting standards. Panel A of Table 8 reports, by privatization methods, the proportion of firms implementing each of the restructuring measures. MBO firms are significantly more likely to use performance-based bonuses (54% versus 47%), to establish a board of directors (84% versus 76%), and to adopt international accounting standards and professional independent auditing (11% versus 8%), all significantly at the 1% or 5% level. MBO firms are not likely to have performance-based share compensation for their executives, which is not surprising, since managers are now owners. Compared with the whole sample, direct sales to outsiders are less likely to establish a board (67% versus 76%) but are more likely to adopt performance-based share compensation (15% versus 7%), both differences are significant at the 1% level. Table 8 Post-privatization restructuring and professionalization A. Post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors Direct sales to insiders (MBO) 64% 54%*** 8% 11%** 84%*** Direct sales to outsiders 61% 51% 15%*** 7% 67%*** Other methods 60% 35%*** 2% 5% 71% All privatized SOEs 62% 47% 7% 8% 76% B. Logit and Tobit regression of post-privatization restructuring measures Performance-based compensation Change of core management team Ratio of bonus in cash compensation Shares International acccounting and independent auditing Establishing board of directors (1) (2) (3) (4) (5) Lag of performance –0.073** 0.274* –0.264*** 0.192 0.244*** (0.036) (0.140) (0.080) (0.065) (0.046) Log (sales) –0.223 –0.020* 0.45 –3.570*** –0.069 (0.343) (0.011) (0.773) (0.992) (0.408) Leverage –0.631** 0.057 0.422** –0.522 –0.501*** (0.302) (0.099) (0.187) (0.575) (0.182) Direct sales to outsiders –0.166 0.140*** 1.793*** –0.094 –0.055 (0.171) (0.053) (0.423) (0.369) (0.203) MBO 0.388** 0.202*** –1.253*** 0.991*** 0.782*** (0.151) (0.044) (0.272) (0.318) (0.189) Industry fixed effects Yes Yes Yes Yes Yes Observations 606 553 606 606 606 A. Post-privatization restructuring measures Performance-based compensation Change of core management team Ratio