China has been searching for suitable corporate forms since 1978 (Nee, 1992). Reformers studied Japanese and Korean business groups and were impressed by their evident capacity to absorb new technology, deliver stable financial performance, and achieve international competitiveness (Ma & Lu, 2005). Reformers believed that business groups might accomplish the same objectives for China. Beginning in 1987, the state signalled that it would favour the reorganization of SOEs into recognized business groups. What followed was a rampant business group fever (Hahn & Lee, 2006), resulting in a dramatic growth in the number of business groups. While the number of registered business groups reached some 7,000, most were small and lacked coherence (Wu, 1990).
To achieve reformers' policy goals, it was evident that significant consolidation was required. In 1991, China's central legislative body, the State Council, identified fifty-seven large groupings described as the National Trial Groups. These state-owned groups were entrusted with a complex socio-economic mission of leading a particular sector into international markets and, at the same time, absorbing a number of underperforming enterprises in return for favourable access to capital and protection from competition (Nolan, 2001). The experiment was judged a success and encouraged the State Council to select a second batch of sixty-three trial groups in 1997. Together these groups are colloquially known as the ‘National Team’ (Nolan, 2001). Yet, despite reformers' efforts at consolidation, business groups continued to proliferate. China is a decentralized federal state with significant responsibility for economic affairs delegated to provinces and large municipal governments. Each provincial government sought to mimic the national policy initiative by organizing local enterprises into a second tier of regional business groups. As reforms proceeded, SOE managers were frequently able to buy-out their enterprises, often at very low prices, and de novo groups founded by private entrepreneurs appeared. In this fashion, numerous private business groups began to emerge on the fringe of the economy. Private business groups (PBGs) are controlled and operated by founder-entrepreneurs, their families, and trusted business partners. As relative newcomers on the economic scene, they have not yet received much attention from researchers, and, due to differences in their ownership, PBGs merit separate consideration (Almeida & Wolfenzon, 2006).
Group Affiliation and Ownership Hypotheses
Mature industrial economies typically benefit from high-quality legal and property rights institutions and a ‘soft infrastructure of the market economy’ (Niskanen, 1991: 233, italics added). The former institutions comprise an institutional matrix of formal laws and regulations and informal normative and cognitive rules and scripts about basic economic relationships in capitalist societies (North, 1990). A soft infrastructure is comprised of a diverse array of organizations and actors, such as technical standards committees, consumer watchdogs, market research firms, executive recruitment agencies, financial institutions, logistics providers, business schools, and training and accreditation agencies that facilitate economic specialization and market efficiency (Khanna & Palepu, 1997). Together, a robust property rights regime and strong soft infrastructure permit independent, freestanding firms to reliably and efficiently acquire key assets and resources through market transactions. In these circumstances, widely diversified and overly integrated vertical firms will underperform more narrowly focused rivals (Williamson, 1985).
Emerging markets are characterized by institutional voids in the form of undefined or unenforced property rights and poorly developed soft infrastructure. In these conditions, transactions costs in external markets will be high for freestanding firms. Diversified business groups have an advantage in the context of institutional voids because they can provide an internal market or quasi-hierarchical governance mechanism that reduces transaction costs for member firms that trade with one another. For example, business groups can provide credible information about their members that reduces the risk of opportunism and lowers contract enforcement as well as search and screening costs. Larger groups can also attain sufficient scope and scale to internalize soft infrastructure and offer services such as management training, finance, technology, marketing, and logistics to their affiliates (Fisman & Khanna, 2004).
As China's enterprise managers gained autonomy, they faced decisions about with whom to trade for the first time (Naughton, 1995). In place of state resource allocation and production targets, managers had to acquire resources in markets characterized by incomplete information and shortages of capital, skilled personnel, and material inputs. Due to the weak soft infrastructure, finding reliable trading partners became a key concern. Financial markets were particularly slow to develop due to restrictions placed on state and foreign banks. Keister (2000) argues managers responded to the uncertainties of imperfect markets by forming stable relations with business partners who could credibly assure the provision of critical resources. To identify credible partners, managers relied upon their contacts and prior social relations with former bureaucrats and party cadres. In this way, hundreds of debt, equity, and trade ties spontaneously developed among newly autonomous enterprises (Keister, 2000). Linkages formed in this manner are at the heart of the spontaneous emergence of China's business groups because these links quickly solidified as firms became de facto group affiliates. Hence, we posit a ‘baseline positive group affiliation’ effect:
Hypothesis 1: Firms affiliated with a business group will be more profitable than independent firms.
However, there are both benefits and costs associated with group affiliation, and it is far from clear whether all affiliated firms participate equally in the distribution of group benefits and costs. Theoretical approaches to business groups typically focus on their complex governance and ownership structures comprised of multiple financial and operational linkages (Khanna & Rivkin, 2006). Indeed, a similar complexity is evident among China's business groups, described by Keister (1998: 408) as ‘coalitions of firms from multiple industries . . . distinguished by elaborate interfirm networks of lending, trade, ownership, and social relations’. Despite the variation in the strength of the linkage with which firms are connected to a group, the vast majority of empirical studies distinguish simply between independent and group-affiliated firms. Researchers typically rely upon directories such Dodwell's Industrial Groupings in Japan, Business Groups in Taiwan, and the Center for Monitoring the Indian Economy that classify firms as either freestanding or group-affiliated firms. However, variation in the degree to which a firm is connected to the group suggests that the group effect will be larger for some affiliate firms than for others (Kim, Hoskisson, & Wan, 2004). China's business groups are characterized by a core or parent firm known as the group company, which is linked to affiliates through equity, debt, personnel, and trading links. For example, the parent group company may hold a majority or minority equity stake in an affiliate, which may in turn hold equity in third companies. While one firm may be tightly coupled in a group's activities via numerous linkages, another firm may be more loosely coupled, playing only a marginal role within the group's affairs.
In this regard, group affiliation is likely to be more beneficial for tightly coupled than for loosely coupled affiliates. However, the categorical or dichotomous measure of group affiliation cannot adequately capture these differences in the extent to which a firm is central or peripheral in the group's affairs. In particular, dichotomous measures are unable to differentiate between firms that participate in the benefits of group affiliation and those that bear the costs. Power dependence perspectives predict that centrally located firms will more likely enjoy access to the benefits of group affiliation while peripherally located groups will be more likely to bear the burdens of group affiliation. Similarly, research that views business groups as a pyramid device (Morck et al., 2005) suggests that intergroup transfer mechanisms, such as related-party transactions, permit value to percolate from the bottom of the pyramid, where a dominant owner's cash flow rights are low, into peak firms, where a dominant owner has greater rights over cash flows. Both power dependence and pyramid perspectives suggest that there is a hierarchy of affiliation in business groups in which core or peak firms are better positioned to accrue benefits while lower order or peripheral affiliates bear the costs of group membership. The percentage of an affiliate's equity owned by the group may indicate this hierarchical aspect of business group structure. We propose a ‘tight coupling’ hypothesis stated in terms of equity ownership:
Hypothesis 2: The greater the group ownership of an affiliate's equity, the greater the performance impact of group affiliation will be.
Researchers are divided about the impact of continuing state ownership on firm performance. On one hand is a ‘grabbing hand’ perspective on the effects of state ownership, which suggests state officials and executives will divert firm resources to their own purposes at the expense of firm performance (Shleifer & Vishny, 1998). Much research on Chinese firms aligns with this view. Clarke (2003) believes China's SOEs are burdened by a syndrome of state ownership problems such as bureaucratic interference, multiple conflicting objectives, and weak incentives, a view supported by other researchers. For example, Nee, Opper, and Wong (2007) find that involvement and direct intervention in the governance of SOEs harms their economic performance. Yiu et al. (2005) argue that, due to factors such as politically motivated appointments and an outdated managerial mindset, continuing state ownership inhibits a firm's ability to develop market-oriented capabilities and harms their performance.
In contrast, developmental state theorists (Amsden, 1989; Wade, 1990) propose that firms in transitional markets are latecomers to industrialization and that, unassisted, they will be unable to catch up with global leaders. Developmental state theorists believe that the state can provide a ‘helping hand’ to their domestic enterprises by curbing competition, guiding firms, allocating resources, and assisting in the acquisition of foreign technology to promote comparability with global leaders. In China, this ‘helping hand’ is likely the motivational force behind the establishment of the National Team (Nolan, 2001). However, the helping hand may reach much further down the industrial hierarchy. Because much responsibility for industrial development in China has been decentralized to more local levels of government, Guthrie (2005) argues that provincial and municipal governments have developed the administrative capacity to effectively monitor and to provide resources and guidance to a relatively small portfolio of SOEs. In this regard, local authorities have been able to promote organizational learning and productivity increases in local SOEs.
Research on internal management processes in SOEs also lends support to the positive view of state ownership. One group of scholars concludes that contemporary Chinese SOEs have re-engineered their organizational cultures to become more market oriented ‘dynamos’ (Ralston, Terpstra-Tong, Terpstra, Wang, & Egri, 2006). Others see rapid learning and confident entrepreneurship among listed SOEs (Tan, 2005).
Between 1987 and 1998, the state actively promoted the formation of business groups, and the National Team enjoyed protection from domestic and foreign competition. Groups with proximity to powerful state actors enjoyed access to soft bank credit, some were allowed to create internal finance companies, and yet others were granted permission to make initial public offerings on the Hong Kong and New York stock exchanges. Moreover, in contrast, private business groups were dependent on self-generated resources or capital provided by families and friends. Private business groups also operated in unrestricted and more competitive markets. Given the division of opinion, the impact of state ownership on firm performance is ultimately an empirical question. We suggest that, on balance, state ownership will moderate the business group effect in a positive way, at least in the initial stages of reform. Hence our ‘helping hand’ hypothesis states:
Hypothesis 3: The performance impact of group affiliation will increase if the firm is affiliated with an SOE-owned business group.
Hahn and Lee (2006) argue that SOEs responded to the encumbrance of forced mergers by diverting assets and resources out of the parent firm to form spin-off enterprises in their group affiliates. We expect this asset diversion to favour affiliates in which the parent has a greater ownership linkage. Ma et al. (2006) argue that state ownership through business groups represents a superior monitoring and control device, relative to alternatives such as state asset ownership agencies, because groups fill ownership voids. Other things being equal, the greater the ownership, the greater the incentive to monitor and support the performance of the affiliate is, and we propose an ‘amplified helping hand’ effect:
Hypothesis 4. The performance benefits of SOE affiliation will increase with the ownership stake of the state.
Group Ownership and Temporal Hypotheses
The temporal hypotheses are also based upon the idea of institutional voids (Khanna & Palepu, 1997). A corollary of the theory that business groups emerge (or are created) to solve market failures is that the logic for their existence will disappear when market institutions and soft infrastructure are established. Two mechanisms are activated by institutional development. First, the benefits associated with business group affiliation will gradually erode as market institutions emerge to fill institutional voids. For example, as alternative sources of finance materialize, the advantage of a group finance company lessens. Second, the development of market institutions facilitates the appearance of more focused freestanding firms that will compete away the excess returns of group-affiliated firms (Peng, 2003).
For example, Haier, a domestic manufacturer of refrigerators and air conditioners, grew rapidly through the 1990s due to its establishment of a diverse group of firms dedicated to warehousing, freight and logistics, retailing, and a service network to serve markets in China's interior. In the absence of a well-developed national distribution system, Haier's proprietary distribution network offered a competitive advantage over more focused freestanding firms such as Whirlpool and Electrolux because Haier's distribution network filled an important market infrastructure void. However, Haier executives recognize that the value of their proprietary distribution network is likely to erode as the quality of China's market distribution infrastructure improves and provides better access to the interior for freestanding firms (Palepu, Khanna, & Vargas, 2006).
We do not expect business groups to adapt smoothly and immediately to changes in their institutional environment. Rather we anticipate that business groups will display considerable inertia against a trend of institutional development. Keister (2001) believes the exchange ties that developed in the initial period may become enduring features of China's corporate landscape akin to those found in Japan and Korea. Importantly, Keister finds that, even when less expensive alternative sources of goods and capital became available, these early trading relationships persisted. If members continue to trade with one another within the group as less expensive and better quality sources are available from outside the group, then performance will worsen. Hence, she conjectures that ‘while business groups may be advantageous early in reform, increasing internalization of ties may create inefficiencies that have negative long-term consequences’ (Keister, 2001: 356).
Proponents of the institutional voids explanation of business group change do not specify the time-frame in which costs and benefits of group affiliation might be expected to change, perhaps because the tempo of institutional development is likely to vary across countries. Campbell (2004) suggests that a scale of decades is necessary for the analysis of the formation of capitalist institutions because an interrelated set of legal, normative, and cognitive rules and scripts must co-evolve to produce a coherent and functioning system. Formal laws and rules about property rights can change swiftly, but normative and cognitive elements necessary for their efficacy may take considerably longer. In contrast, a scale of years may be adequate for the analysis of changes in soft infrastructure in the sense defined by Khanna and Palepu (1997).
A co-evolutionary pattern of institutional change has been observed in the context of China (Krug & Hendrishke, 2008). In the wake of the 1997 Asian financial crisis, which implicated poor governance in business groups as a causal factor, China's reformers became concerned that their business groups might share similar problems and set a course for correction in the direction of reform. Reformers were determined to accelerate the development of China's market institutions and rushed through a slate of legislation designed to establish international best-practices in corporate governance. Initiatives included bank reform, the establishment of a state asset supervisory administration commission, privatization of small- to medium-sized SOEs, establishment of internal controls through mandatory boards of directors and supervisory councils, a legal code for companies, a bankruptcy procedure, and principles of protection for minority stakeholders (Clarke, 2003). Most importantly, China's accession to the World Trade Organization strongly commits China to a prescribed timetable of market-based institutional development.
While China has made progress in institutionalizing market mechanisms and implementing its World Trade Organization commitments, we do not suggest that reforms have had an immediate and full effect in establishing a robust property rights regime, although the cumulative effect of change may eventually do so. Rather, we propose that the increasing depth and improved quality of China's soft infrastructure is driving changes in the business environment. The period between 1992 and 1998, when GDP growth in China was typically over 20 percent per annum and reached 35 percent one year, was a particularly turbulent era (Tan, 2005) that would promote group affiliation. However, the heavy investment in market infrastructure in this period would thereafter enable entry by freestanding firms that, by 2004, could exert increasing competitive pressure upon business group affiliates. Hence, our baseline temporal hypothesis states:
Hypothesis 5: The positive impact of business group affiliation effect will decline over time.
Here, we propose that, with the progress of institutional reform, the value of the state's helping hand will diminish and the ‘grabbing hand’ deficiencies of weak SOE governance will become increasingly salient. Several analysts suggest that inherent governance deficiencies have begun to surface in the ranks of SOE business groups. Initial reforms successfully cultivated a dynamic market orientation in the senior management of many enterprises. Charismatic chief executive officers who are closely identified with the rise of the particular enterprise have become a common phenomenon in China. Yet within a relatively short period of time, these powerful CEOs have become entrenched in their positions and are difficult to dislodge, even as the performance of their enterprise deteriorates (Clarke, 2003). Lin (2001) argues that officials who hold monitoring positions have few incentives to pursue their duties with any real vigour; heads of state ministries and senior bureaucrats are compensated according to standardized public sector payment systems that bear no relationship to the performance of the SOEs under their control. Lin (2001) concludes that the supervision of state firms is beset with serious moral hazard problems.
Hahn and Lee (2006) propose that, due to inadequate supervision, business groups are characterized by large-scale asset diversion as managers seek to shield more valuable assets. Within these non-transparent insider structures, it is likely that senior managers may engage in self-serving behaviour, such as taking perks or extracting rents for personal use. Hence, while business group governance of SOEs may have filled ‘ownership voids’ (Ma et al., 2006) during the early stages of reform, we suggest that inadequately monitored SOE managers subsequently exploited these voids in a manner that negatively impacts firm performance.
Further, the protected SOE business environment has liberalized. Specifically, whereas state-owned business groups had previously enjoyed favourable access to financial resources and protection from competition prior to the 1997–1998 Asian financial crisis, thereafter, the government began to tighten their soft budget constraints. Product market competition sharpened due to the gradual dismantling of competitive restrictions in sectors previously reserved for national champions. The confluence of these contextual effects suggests a ‘negative amplification’ effect:
Hypothesis 6a: In later stages of reform, the performance impact of group affiliation will decrease if the firm is affiliated with an SOE-owned business group.
Hypothesis 6b: In later stages of reform, the performance benefits of SOE affiliation will decrease with the ownership stake of the state.