Excerpt from China’s Trapped Transition - The Limits of Developmental Autocracy by Minxin Pei
A Question of the State: Developmental or Predatory
The sustained economic development achieved under authoritarian rule in Singapore, Taiwan, Hong Kong, and South Korea from the 1960s to 1980s has provided the factual basis for the claim that a neoauthoritarian mode of development—state-guided rapid economic growth under authoritarian rule—is a superior and proven strategy.53 To be sure, in the Western academic community, the concept of an East Asian model is a subject of debate, especially because of the controversy over the efficacy and degree of state intervention in East Asian countries. For some scholars, the East Asian experience is proof of the centrality of state intervention in the rapid growth of late-developers.54 For others, right public policies were the key to East Asian success.55
Unfortunately, most leading scholars of East Asian political economy have skirted the issue of regime and development. Only Robert Wade, author of one of the most influential studies on the role of the state in East Asia’s economic development, explicitly identified the development “of effective institutions of political authority before the system is democratized” as a key to East Asia’s success.56 Within the Chinese political and intellectual elite, the East Asian model has been essentially reduced to a simple formula: strong government authority + promarket policies = superior economic performance. It has further been argued that strong government authority would be difficult to obtain under democratic political systems.57 In fact, when asked about his views on neoauthoritarianism by Zhao Ziyang in a private conversation in 1988, Deng admitted that such a strategy, “relying on a political strongman to maintain stability and develop the economy,” was exactly what he was advocating even though “it is not necessary to use the term (neoauthoritarianism).”58
Such a preoccupation with the efficacy of the state in the context of economic development overlooks one crucial issue: the relationship between economic growth and the predatory behavior of the state. In other words, the real East Asian puzzle is not how sustained rapid economic development occurred under strong states, but why and how the predatory, practices of the state were held in check. Based on the assumption of the state as a “helping hand,” much of the literature on the political economy of development in East Asia has all but ignored the possibility that a strong state can also be a “grabbing hand.”59 Peter Evans’s influential Embedded Autonomy: States and Industrial Transformation may be the only exception. By identifying the nature of the state as the critical variable in explaining the variations in the success of industrialization in developing countries, Evans shows that a predatory state is incapable of nurturing new engines of growth (in his case, the information industry).60
But Evans’s explanation of why some states are nurturers while others are predators addresses only part of the puzzle. His formulation of the “embedded autonomy” of the state—the idea that developmental states gain autonomy and efficacy only when they arc “immersed in a dense network of ties that bind them to societal allies with transformational goals”—provides a useful answer, but appears to restate the well-known: states counterbalanced by strong societal coalitions are less likely to be predatory.61 Such explanations, built on the perspective of state-society relations, fail to probe the internal organizational dynamics and norms of the state. While few would deny the desirability and benefits of having strong societal forces committed to economic development, the most important challenge in the real world is that, in the overwhelming majority of developing countries, such forces are extremely weak or absent altogether. The transformativc project has to begin inside the state.
Yet, there is another conundrum: as some scholars have argued, no evidence exists to show that institutions can be devised to make the state an effective protector of property rights, but at the same time prevent it from abusing its power. In other words, there is no guarantee that the same helping hand will not become a grabbing hand.62
Indeed, as the experience of most developing countries shows, states as helping hands are the exception. Sustained developmental successes probably number fewer than ten, with most of them concentrated in the East Asian region.63 At the same time, predatory states have caused disastrous failures in a majority of poor countries, the most egregious examples being the Philippines under Marcos, Zaire under Mobutu, and Haiti under the Duvaliers. As the collapse of the Indonesian economy in 1997-1998 demonstrates, without adequate institutional controls imposed on predatory states, even initial successes could end up as catastrophic failures.
The Theory of the Predatory State
In its simplest formulation, the theory of the predatory state is based on a conception of the state as a grabbing hand. It envisions the central role of the state as the expropriation of wealth from society through taxes for the preservation of the state’s own power.64 The recent growing appeal of the theory is due, in large part, to the application of the institutionalist approach to the research on the relationship between political institutions and economic performance.65 The revival of institutionalism has again elevated the role of the state in economic development.66
To be sure, the institutionalist perspective on the state differs from that of the developmental state, which, as popularized by scholars of East Asia, has a direct role in correcting market failures. Institutionalists see the state as the provider and enforcer of rules and norms that underpin market transactions. This distinction is significant because the perspectives on the state behind it are fundamentally different. Unlike the helping hand envisioned in the developmental state perspective, the state is seen by institutionalists as a force both for good and evil. As Douglass North puts it, “The existence of a state is essential for economic growth; the state, however, is the source of man-made economic decline.”67 Although the state may be a helping hand that specifies and protects efficient property rights, it can also be a grabbing hand that expropriates the wealth of its people.68
The grabbing-hand perspective appeals to students of development because the theory of the predatory state provides a persuasive explanation for the weakness of the state and the overall poor performance of government. In applying this perspective, however, we need to make the distinction between centralized predation and decentralized predation because such a distinction is crucial to understanding the different dynamics behind a state’s institutional performance. In the earlier formulations of the theory of the predatory state, the focus is on the aggregate level of state predation and treats predation as the political imperative of the ruler. There is no distinction between the principal and the agent. As a result, predation is conceived as an act of the principal. This formulation assumes, first, that state predation is universal. Because rulers are monopolists of both violence and public goods, state predation, in the form of taxes, is simply the price private producers of wealth pay for such monopolistic services. Second, the most important factor that limits the level of a ruler’s predation is his self-interest. To use Mancur Olson’s colorful analogy, a self-interested ruler behaves like a “stationary bandit” who is unlikely to risk his future revenue streams by looting the current stock of wealth of his subjects. He will raise taxation only up to the level that maximizes his tax revenues.69 Third, rulers are supposed to have an encompassing interest that is akin to the national interest.
In both theory and practice, centralized predation can spiral out of control. The ruler’s encompassing interest may diverge fundamentally from that of the state. For example, the ruler’s personal greed may become insatiable. He and his close cronies may loot wealth, not to provide public goods, but to line their own pockets, thus creating a kleptocracy. 70 The ruler’s encompassing interest may also become too ambitious for his nation’s good. Desire to acquire a larger territory (hence tax base) or international prestige may motivate the ruler to extract excessively from society to build a strong military.71 In addition, the ruler’s monopolistic position is always insecure because a domestic or foreign rival can seize his monopoly by force.72 This structural insecurity affects the ruler’s time horizon and the rate of discount on future revenues, incentivizing behaviors that result in short-term gains but long-term revenue losses.73 Finally, the absence of a third-party enforcer makes the ruler’s commitment to self-restrained predation not credible. Temptations for the ruler to break his promise and increase predation always exist, and most rulers have honored their promises in the breach.74
In the theoretical literature on decentralized predation, the emphasis is on predation by agents of the state. Although agency costs have been identified as a constraint on the ruler’s ability to maintain a desired level of extraction, the effects of such costs on state predation have not been explored until recently.75 Scholars who focus on the role of agents in state predation see decentralized predation as more harmful to the interest of the state. Andrei Shleifer and Robert Vishny demonstrate that centralized corruption, which is a form of monopolist predation, generates higher aggregate revenue for the state—because the state keeps its rate of extraction at the optimal level—than decentralized corruption (a form of predation by state agents acting as independent monopolists), which not only raises the overall level of theft (that is, making corruption more widespread), but also reduces the aggregate amount of income for the state. Since predatory agents simultaneously compete with one another for the same revenue, they have the incentive to steal everything, behaving essentially like Olson’s “roving bandits.”76 The welfare loss from decentralized predation is much greater than that from centralized predation.77 Decentralized predation, moreover, has emerged as a more prevalent problem today, as regime transition in many countries has restructured some of the key institutions governing principal-agent relations (more on the effects of transition on decentralized predation below).
To be sure, the relationship between centralization and corruption is a subject of scholarly dispute. Some scholars believe that decentralization may actually reduce the level of corruption. For example, decentralization can make local officials more accountable to the public because they can no longer hide behind the actions taken by higher authorities. Greater political accountability would help control corruption. 78 Decentralization may also contribute to lower levels of corruption because local officials are deterred from corrupt activities by a higher likelihood of being caught. Moreover, decentralization deprives the central government of the financial resources that otherwise are routinely squandered on grand corruption schemes (such as white elephant projects), thus reducing the aggregate costs of corruption. Even though decentralization can lead to a short-term increase of petty corruption by local officials, the total costs of petty corruption are likely to be much lower.79
Many scholars believe, however, that decentralization can increase corruption for several reasons.
Given the low wages paid to local officials, increased political discretion as a result of decentralization is likely to lead to more corruption.80 If decentralization should lead to the breaking of arm’s-length relationships between clients and government agents, it can cause corruption to rise, especially in cultures where interpersonal connections play an important role. Newly empowered local bureaucrats are thus more likely to reward family friends with various forms of rents.81 Decentralization may exacerbate corruption if it occurs in the context of weak government. When the political authority of the government is weak across the board, decentralization can create independent monopolists who have every incentive to maximize the collection of bribery at the local level.82 On balance, the argument that decentralization can lead to more corruption is more persuasive because its proponents adequately account for the agency problem, while the same problem is simply assumed away by those who believe that decentralization can reduce corruption.83
The distinction between centralized and decentralized predation notwithstanding, the central point of the predatory state perspective is self-evident : without effective political institutions or structural constraints that curb the predatory appetite of the state, a state that is sufficiently strong to promote economic development unhindered by parochial interests is also strong enough to prey upon society without much restraint. The consequences of unrestrained state predation are dire. In such a state, the ruling elites distort markets, create rent-seeking opportunities for self-enrichment, and loot the wealth of society. Sustained economic development under such a state is impossible. The hope that economic development would eventually lead to democratic transition is only wishful thinking because the predatory state and economic development are, logically, mutually exclusive.
Why Decentralized Predation May Emerge during Transition
Research on transitions in the former socialist states indicates a significant increase in decentralized predation immediately following regime changes. Joel Hellman’s study of reform in the former Soviet bloc countries suggested that the ruling elites were able to capture the state and reap all the benefits of partial economic reforms.84 Michael McFaul and Federico Varese found that the communist ruling elites in the former Soviet Union were able to use their institutional privileges and exploit the loopholes in property rights laws to steal public assets in the privatization process.85 In an insightful analysis of the collapse of the former Soviet Union, Steven Solnick showed that transitions that decentralize authority tend to lead to an increase in the number of thefts of state assets.86
Theoretically, the type of post-transition state predation observed by country specialists and journalists is qualitatively different from that which occurred during the pre-transition era. In pre-transition communist countries, state predation was centralized. Two characteristics defined centralized predation under communist rule. First, the aggregate amount of revenues generated was large, reflected in the government revenue as a share of GDP. Second, a significant amount of the revenues was used to provide public goods, mainly national defense, health, and education spending. Consequently, countries ruled by communist regimes enjoyed a higher level of human development relative to their economic development, especially in terms of their literacy rate, infant mortality rate, and life expectancy.87
By contrast, posttransition state predation is decentralized and manifests itself in various forms of official corruption. Decentralized state predation reduces the aggregate amount of state revenue, as agents divert public money into private pockets. It also causes a fall in the provision of public goods, as state agents convert public resources into private consumption or offshore investments. Although the phenomenon of decentralized state predation in post-transition countries has received enormous attention, the causes of decentralized predation are not well understood.
Centralized predation becomes decentralized when the state, as the principal, loses effective control over its agents. Of course, different types of regime transitions generate different dynamics that affect principal-agent relations. In communist states that saw a quick collapse of the ancien regime, state agents were afforded great advantages by even the temporary decline of the principal’s authority. In those societies, the agents’ theft of state assets was completed within a relatively short period of time. However, the patterns of post-transition agent predation diverged dramatically in those post-communist states that experienced dual transition. As Hellman’s work shows, new regimes with a higher degree of democracy and more complete market reforms tend to restrain such predation, while new regimes with less democracy and partial economic reforms are beset by increased levels of agent-predation.88
By comparison, agent-predation followed a different dynamic in post-communist systems that have seen market liberalization but no political transition, such as in China and Vietnam. In these societies, the political authority of the state remains unchallenged. However, the decentralization of decision making, needed to reincentivize state agents, led to a restructuring of the contracts between the state and its agents, which proved to be extremely advantageous to the latter. Therefore, the key to understanding the rise of decentralized predation is to examine both the preexisting and the transition-related institutional changes that have structured and restructured principal-agent relations. Specifically, changes in the control of property rights, mechanisms of monitoring, exit options, and institutional norms arc the critical variables responsible for the decentralization of state predation in postcommunist societies.
Decentralization of property rights
In theory, the degree of centralization of property rights is negatively correlated with decentralized predation. In countries with a high level of centralization of property rights, the loss of state money through agent theft or misappropriation tends to be small. Under the prereform communist system, despite the lack of clarity of property rights, the high degree of centralization of such rights was the decisive institutional factor that limited agent-predation. In practice, the centralization of property rights prevented large-scale theft of state property. Of course, the mono-property rights regime and the high degree of centralization of property rights caused low efficiency because this system provided few incentives for agents to improve the financial performance of state assets.89 The decentralization of property rights during the transition phase in most state-socialist systems was originally designed to increase agent incentives so that state assets could become more productive. In some countries, such as China, the decentralization of property rights also involved the transfer of formal ownership of state assets from the central government to local governments. Such decentralization granted the state’s managerial agents more discretion in operating SOEs, especially regarding investment and compensation. Although no evidence shows that decentralization of property rights alone contributes to efficiency gains, the combination of lack of clarity of property rightsand decentralization of such rights has led to widespread asset-stripping and other forms of theft by state agents.90
Given the institutional changes entailed in regime transition, the old system of monitoring state agents is likely to break down. Rule changes are frequent and confusing during transition, resulting in poor coordination among various state agencies monitoring agent behavior, such as the secret police, tax authorities, auditors, and financial controllers. The breakdown of monitoring becomes even more likely if those state agents in charge of monitoring other state agents detect the latter’s theft but decide to divide the spoils with the thieving agents, instead of reporting their malfeasance to the principal. (Unsurprisingly, one of the most corrupt government bureaucracies in China and Russia is the anticorruption agency.) Transition frequently entails changes in political values and erodes the authority of the principal. Agents face negligible risks in defying the authority of the principal. Erosion of the principal’s authority makes effective monitoring of agents ineffective. The monitoring of agents has also become more difficult under reform because of an increase in transaction channels. As X. L. Ding observed, interfirm transactions almost did not exist under the old system, in which ministries directly controlled SOEs’ sales and purchase processes. As a result, the monitoring of agents’ business deals was easier under the old system. In the transition era, the advent of marketization replaced the firm-ministry-firm transactions chain with the more efficient firm-to-firm transactions chain. Consequently, the number of transactions exploded, making effective government monitoring nearly impossible.91
New exit options
Large-scale theft of state assets was made less likely under the old communist system by the absence of exit options from the state sector for nearly all agents. Regime transition has opened numerous exits for these agents, including ownership stakes and managerial positions in the new private or semiprivate firms, and overseas investment opportunities. 92 These exits effectively allowed “stationary bandits” under the old system to become “roving bandits” because they can steal and then store their loot safely elsewhere. The time horizon of state agents with lucrative exit options is likely to be short, resulting in more intense theft of state assets.
Erosion of institutional norms
Institutionalists have long recognized the role of institutional norms in constraining agent opportunism and the free-rider problem.93 As a concept, institutional norms are vague and difficult to define. In practice, institutional norms may derive much of their legitimacy, appeal, and binding power from the prevailing ideology of the political system. In the case of communist systems, it may be controversial to claim that the communist ideology had any appeal. One can point to the widespread cynicism among the ruling elites of the ancien régime. It is nevertheless conceivable that even residual ideological appeals of communism, socialism, or nationalism might have played a role in constraining the predatory instincts of the agents under the old regime. During transition, the total bankruptcy of the communist ideology meant that state agents were under no constraints imposed by institutional norms.
The above theoretical analysis suggests that a temporary partial reform equilibrium, or a transition trapped in semireformed economic and political institutions, is a product of a confluence of factors. The most important among them includes the initial conditions of the transition process in post-totalitarian regimes such as the CCP, which retains unchallenged political supremacy over society and maintains its rule through a mixture of coercion, co-optation, and adaptation. Additional factors contributing to the emergence of such a trap are embedded in the political and economic logic of market transitions and authoritarian politics. Indeed, the Chinese experience provides an intriguing example that demonstrates why gradualist economic reform pursued under a neoauthoritarian regime, even after achieving impressive initial results, may lose momentum. Instead of moving toward an even more open economy and society, such a system may be heading toward long-term stagnation amid widespread symptoms of state incapacitation and deterioration of governance.
At the theoretical level, one can construct an argument that incorporates the insights from the theories of democratization, economic reform, and predatory state to explain the phenomenon of trapped transitions.
Gradual democratic transitions in post-totalitarian regimes face higher hurdles than those in authoritarian regimes. The connection between economic development and political liberalization is likely to be weak in these regimes because the initial conditions are far more adverse. The institutionalized curbs on the power of the ruling elites in a post-totalitarian regime are negligible. The ruling elites thus have far greater ability to defeat societal challenges. The presence of the post-totalitarian ruling party in state bureaucracies, economic entities, the military, and the judiciary provides it with the instant ability to convert political monopoly into economic rents during economic transition. Consequently, economic growth, rather than creating exits for peaceful withdrawal from power and lowering the costs of political transition, may perversely increase the stakes of exiting power because the ruling elites risk losing not only political power, but also economic rents. In addition, such rents become more valuable in an open and fast-growing economy, and, more important, the material wealth accurnulated by the ruling elites can be consumed openly, extravagantly, and without fear when the prereform codes of austerity are no longer operative. Therefore, even though economic growth may have a long-run positive impact on democratization, its short-run impact can be decidedly negative.
Gradualist transitions can further help entrench the post-totalitarian party-state and thwart efforts to both deepen market reforms and initiate democratic transition. Gradualism allows the ruling elites to make selective withdrawals and maintain their control in the most lucrative high-rent sectors; this development tends to make the ruling elites even less inclined to give up political power during transition. The control over sectors with rich rents also facilitates the emergence of -political alliances with stakes in a semireformed system but with no interest in political reform, as the ruling elites use such control to co-opt emerging social elites individually, include them in a collusive network of rent-sharing, and preempt potential political challengers. To the extent that initial reform efforts may be successful, gradualist reforms buy the regime a new, albeit temporary, lease on life, removing the pressures for political reform. Gradualism becomes eventually unsustainable because of the problem of dissipation of rents. The regime’s strategy of protecting rents in key sectors ultimately fails when such rents are distributed and consumed by the agents of the regime, critically weakening the health of the economy.
However appealing the concept of a developmental state, successful economic development under neoauthoritarianism may be the exception. A self-restrained developmental state can materialize only under rare circumstances that force the ruling elites to choose between curbing their predatory appetite or risking their own survival, a choice that is not always correctly made. In post-totalitarian political systems where the ruling elites possess overwhelming advantages vis-à-vis societal oppositions, operate under ineffective institutional restraints on their power, and face no credible external threat, the state is most likely a grabbing hand, not a helping hand. Thus, despite its promarket rhetoric and policies, a post-totalitarian regime may likely degenerate into a predatory autocracy, rather than evolve into a developmental neoautocracy. Economic development under a predatory autocracy is ultimately unsustainable. Ironically, a democratic opening may emerge in the end, not as a regime-initiated strategy undertaken at its own choosing, but more likely as the result of a sudden crisis brought on by years of corruption, mismanagement, and institutional decay.