Explanations of China's post-Mao social policy have concentrated on the political, social, fiscal, and economic goals of the state and its governing elite. In a study of urban health insurance policy, this article argues that bureaucratic interests and institutions within the Chinese state are also influenced. This article first shows how bureaucratic interests within the central government have influenced the adoption of a new national social health insurance framework. It then shows how that framework has been modified following local implementation experiences that have allowed other bureaucratic and non-bureaucratic interests to be expressed. This examination of both central and local interests helps explain the adoption of a basic social health insurance system that provides for only the urban working population, subsidizes civil servants, and is administered locally. The article also shows the policy process in this sector to have been particularly protracted and incremental and argues that further incremental policy changes are likely.
The enormous societal impact of China's post-Mac market-oriented economic policies is now well-recognized. Social and geographical mobility and wealth have all increased along with unemployment and inequality. At the same time, social policies have begun to restructure the old welfare system; for example, introducing in urban areas contributory old-age unemployment and health insurance and broad but minimal poverty relief. To date, studies have explained such post-Mao social policies as the result of the state or its leaders' efforts to prevent social unrest, respond to social needs, and reduce spending and responsibilities. Thus, leaders are said to be attempting to limit the erosion of safety nets because they fear that unrest will threaten the political system (Croll, 1999; Leung and Wong, 1999).
However, in other work, policies emerged from a "black box" unitary and apparently benevolent state to deal with problems that have emerged in the course of market transition (see, for example, Chow, 1995; Lee, 1995). In contrast, others explain policy as the result of a parsimonious state determined to cut back spending and responsibilities with little regard for social need (Wong, 1999; Linda, 1994).
Such discussions offer rich insights into the variety of factors influencing social policy in the post-Mao period but can also seem contradictory. Some of the apparent contradictions can be explained, however, by looking "inside" the state. Because policy-making in China today takes place in a centralized one-party system, there is little public debate or space for open lobbying (Liebenthal, 1995). Although non-state actors can have influence (Bernstein, 2000; White, 1993; Zweig, 1997), policy-making is relatively closed. As studies of other sectors have shown, the actors within the central state involved in making policy decisions include not only the core elite of top leaders and their staff but also "leadership small groups," research centers and commissions, and the line ministries whose activities they coordinate. These individuals and bureaucratic agencies have their own interests and organizational missions and strive to defend them in the policy-making process (Liebenthal and Oksenberg, 1988; Shirk, 1993; White, 1993).
China's top leaders, particularly those on the Politburo Standing Committee, have broad political responsibilities and tend to have an interest in their own political position. They take a broad view and are likely to be concerned with threats to the regime. Studying these leaders may therefore show why new policy initiatives appear and explain the general direction of policy. A second tier of leaders has narrower, more specialized roles and are more likely o represent particular bureaucratic interests (Lieberthal and Oksenberg, 1988:28-29). Because of this and since much of the work of formulating and implementing policy is handled by the bureaucracy, attention to bureaucratic institutions and interest can help explain why particular policies are adopted and how they are developed, revised, and implemented or rejected (Liebenthal and Oksenberg, 1988). …