Introduction Tobacco use in China stands out as one of the most significant public health issues in the world today. Currently, there are 350 million smokers in China, making up more than one-third of the world total (Koplan et al. 2013). In China, 52.9 per cent of males and 2.4 per cent of females over 15 years old are regular smokers (Giovino et al. 2012). There are also about 740 million people who are regularly exposed to second-hand smoke (Yang and Hu 2011). Notably, China is also the world’s biggest producer of cigarettes. In 2012, it produced over 2.5 trillion cigarettes – more than one-third of the world total (Reference News 2015) – over 95 per cent of which were consumed domestically (Wang 2013). It is estimated that domestic producers in China have about 98 per cent of the market share (China Internet Information Center 2014). The health consequences of smoking are enormous. It is widely recognised that smoking is one of the major causes of lung cancer, stroke and heart disease. The number of deaths caused by tobacco use in China is currently 1.2 million per year (Li 2012). In 2020, it is expected to hit two million per year (Lee and Jiang 2008). The total economic costs of the adverse health effects of tobacco usage in China were estimated at about US$28.9 billion in 2008 and these have been increasing ever since (Yang et al. 2011). These costs include health expenditure for treating smoke-related diseases and the costs associated with premature death. China ratified the World Health Organization’s Framework Convention on Tobacco Control (FCTC) in 2005. Since then, many tobacco control polices have been initiated to meet conditions set by FCTC. Nevertheless, China did not meet the January 2011 deadline to fulfil its FCTC obligations. Also, as reported by the Chinese Center for Disease Control and Prevention (CDC), the effectiveness of tobacco control policies in China is unsatisfactory, scoring only 37 out of 100 points for the implementation of the six major policies in the WHO MPOWER1 package (Yang and Hu 2011). The ineffectiveness of tobacco control policies is associated with bureaucratic noncompliance with policy targets, which is reflected in two main areas. First, both central and local governments are inefficient in implementing tobacco

control policies, particularly in terms of legislation for tobacco control that has not been enacted and enforced to meet the policy targets. Second, institutional arrangements for implementing tobacco control policies are associated with a ‘regulatory capture’ situation, where the regulators of the tobacco industry actually represent interests of the regulated. The fragmented authoritarianism (FA) analytical model (Lieberthal and Oksenberg 1988) – which argues that the decision-making bodies are fragmented and disjointed among different ministries as well as among different levels of government – has been widely used to understand the role of bureaucrats and bureaucratic noncompliance in policy implementation in China for over two decades. FA is particularly useful for understanding the policy-implementation process, in which negotiation and bargaining between different government departments are involved. Both of the aforementioned observations about the policy ineffectiveness of tobacco control are consistent with the FA model. The inefficiency of central and local government is due to the decentralised distribution of authority among different departments, and the bargaining and negotiation that takes place among these departments leads to inefficiency in policy implementation. Moreover, the structure of authority in policy-making and implementation, as well as regulatory capture, is a result of the fragmented structure of the bureaucracy involved in the policy-making aspects of tobacco control. However, two questions can be raised for the proponents of FA. First, whether there is a fragmented structure in the first place? For example, the structure of policy-making and implementation may be endogenously determined by some institutional reasons. Second, why does central government not respond to the fragmented structure by initiating policy and institution changes? In other words, if the role of the fragmented structure is common knowledge for all stakeholders, to improve policy effectiveness central government has an incentive to respond by initiating new policies or establishing a new regulatory agency. This chapter complements the FA literature on the ineffectiveness of tobacco control policies in China. Instead of arguing that central and local government’s inefficiency and regulatory capture in implementing tobacco control policies are symptoms of the fragmented and decentralised structure of decision-making bodies, it proposes that ineffectiveness of policy implementation is a result of certain institutional arrangements, including government monopoly over the tobacco industry and the performance evaluation system for government officials. In the literature of political economics and institutional economics, bureaucrats are presented as self-interested individuals whose behaviours are constrained by institutional arrangements. From this perspective, we show that the bureaucratic noncompliance behaviours result from the following two institutional factors: First, the fragmented structure of authority and regulatory capture are endogenously determined by the government monopoly of the tobacco industry.