The background variables, GDP per capita, marketization, industrialization and urbanization are used to control different levels of development across provinces. The role of the state is measured in the following ways. First, financial capacity, administrative capacity and coercive capacity are used to measure the role of state capacity in BMI extension. The study examines whether there is a difference in choosing different agencies to collect social insurance premiums: one is local taxation agency and the other is social insurance agency. Third, the performance of BMI is measured through the deposit rates of BMI funding which reflects governments' ability to manage the BMI program. In the current policy, employers are charged largely the social insurance fees. So their willingness and capabilities to pay will affect BMI coverage. The study investigates two kinds of employers: loss making State-Owned-Enterprises (SOE hereafter) and Foreign Invested Enterprises (FIE hereafter). On the employee's part, the percent of informal employment in total urban employment is used to measure the effect of adverse employment conditions on BMI coverage. Trade union density is used to estimate the labor organization strength. / The complicated process of extending coverage is related to three major stakeholders: state, employers and employees. These three stake-holders influence BMI progress. Also, the background factors (such as the economic growth) should be taken into account for the regional variations in development level. Since BMI is a typical social policy field, this study reviews major theories about social policy development: logic of industrialism, power resource theory and state-centered approach and so on. These theories help organize pieces of phenomena into a unified framework and testable hypotheses are also derived. / The contributions of this study can be twofold. First, from the theoretical aspect, this research tests several welfare state development theories using Chinese data. In this way, it does not only expand the scope conditions of theories, but also improves our understanding of the social policy development in China, an outlier of traditional western democracies. Second, this study tests some controversial issues on BMI development and the research findings provide knowledge support for the policy practice in the real world. / The low coverage of social health insurance is one of the causes of the problems in Chinese health care system which is criticized for the rising health cost, large share of out-of-pocket payments and health inequality issue. The Basic Medical Insurance for Urban Employees (BMI hereafter) was chosen as the subject of my investigation. It was established in 1998 for the working population and till now it has not achieved universal coverage yet. The Basic Medical Insurance for Urban Residents (BMI-R hereafter) was started in 2007 and it is still in pilot stage, therefore data are still inadequate. In rural areas, the New Cooperative Medical Scheme (NCMS hereafter) achieved almost full coverage in 2008. Thus extending coverage is not issue at concern for NCMS. Besides, the NCMS data at province level are quite limited. Considering the stages of policy development and data access, BMI-R and NCMS are not included in this study. / The proportion of winning lawsuit in labor disputes is used to measure the function of labor protection system. This study adopts the panel method. Data is ranging from the year 1999 to 2005 and the unit of analysis is province/year. They were collected from various official statistics and constructed into a panel database which can trace the development of BMI from its origin to most recent situation. / The research question is what are the determinants of BMI's coverage? It is originated from some puzzling observations: the NCMS achieved full coverage in four years and it is a voluntary participation insurance program. On the contrary, why the mandatory BMI did not reach universal coverage after almost ten years' development? Besides, the progress of BMI across different provinces varied greatly. Given the policy designing and starting points are rather similar, how can we explain these variations? / The research yields several interesting results. First, the roles of financial capacity and administrative capacity in BMI development are supported by data, especially the social insurance agency. Second, results show that using local taxation to collect social insurance premiums has better effects in extending coverage than the alternative approach. This result will give an end to the decade-long debate on choice of social insurance premiums collection agencies. Third, the deposit rates of BMI funding are negatively related with BMI coverage. It implies that governments should improve the performance of BMI so as to attract more people to enroll in this program. Fourth, the union density in the private sector is positively related with BMI coverage. This result disagrees with the conventional wisdom that the Chinese trade unions are useless. It implies that strengthening the organization of employees (even through the official channel) can protect the rights of employees in some degree. / 劉軍強. / Adviser: Cheek-Kie Wong. / Source: Dissertation Abstracts International, Volume: 73-03, Section: A, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2009. / Includes bibliographical references (p. 198-222) / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts in Chinese and English. / Liu Junqiang.