| Institutionalization is regarded as a key issue for transitional states. It is as crucial to analyze the process of institutional changes, at least as, even more, to explain the causalities of institutional changes. New institutionalism, as a prominent branch in contemporary comparative politics, has not dealt with the issue on institutional changes as well as interactions between institutions and actors for a long time, especially on changes caused by endogenous factors without fundamental shifts of social structure. Fortunately, a theory of gradual institutional change is growing out of the new institutionalism, which elucidates four models of gradual institutional changes, i.e. displacement, layering, drift and conversion. The theory broadens our research agendas; however, there are still some other aspects for further inquries. More importantly, the theory ignores a fundamental issue on the origins of the new institution, since the starting point for explaining institutional changes must be an account of human learning, as North and his colleagues argue. However, both the new institutionalism in organizational analysis and North's framework do not pay attention to the diversity of sources of the learning and its systematization.;China, as a huge transitional state, provides many valuable cases on institution changes. This research is based on empirical study from in contemporary China, so that it aims to introduce a new model of gradual institutional change, named experiment. I define the experiment, experimental point and practices after clarifying the difference between the institutional and policy experiment firstly. Then I divide the whole process of institutional experiment into six stages, i.e. exploring by the local, initiating the program of experiment point (shidian fang'an), assessing pilot program, terminating the program of experimental point, experiences diffusion from point to surface (congdian daomian), and formulating new institution. Furthermore, I raise the framework on institutional experiment and human learning process based on experimentation.;The case of the study is the process of institutional formulation on the system of urban administrative enforcement in contemporary China. The main part of this dissertation is composed of several chapters. In the third chapter, I retrospect the evolution of the system of urban administrative enforcement in China from accent era to after the foundation of PRC, with comparing two distinguished arrangements before and after 1949, for shedding light on the "institutional jungle" before the experiment. The fourth chapter focuses on the first three stages, in which the exploration by the local provides the experience for orienting the institutional experiment. The institutional arrangements generated by the outside organizations are not excluded from sources of actor's learning. The central government indicates its control power while the local is authorized to innovate institutional options, which ensures the learning process avoid from being limitless. I elucidate the next three stages in the next chapter. The institutional experiences are diffused from pilots to the more areas for formulating the new national institution. The new institution is integrated into the local practices, the domestic traditions, and even the outside arrangements. The mechanisms of learning actions indicate its diversity in these three phases. I classify them from three dimensions, the vertical, the horizontal, and the mismatched by comparing promoters with objectives diffused. I move to discuss advantages and problems resulted from the institutional experiment in the sixth chapter, after summarizing the nature of institutional experiment and the process of human learning on the base of experimentation. The reasons why the central and local intend to engage in the experiment lie in two aspects, to deal with uncertainties from institutional changes and make more institutional innovators involved into institutional building in a limited open political system. Simultaneously, the experiment results some problems by causing conflicts between the new and old institutions, and locating the new organizations in a more complicated environment. Conclusions follow in the last chapter from the theoretical and empirical perspectives. I demonstrate my contribution of the research before I discuss the limitation of my study and the further agenda for more comparisons.;Above all, the new model, institutional experiment, differs from existing ones in two features. On the one hand, the institutional replacement as a target is raised definitely, while new institutional arrangements are proposed only as an idea for responsing to the problem(s) confronting the actors, rather than introduced wholly. On the other hand, the central permits the variations of new institutions across different regions in one country during, even after the experiment. In the mean time, the learning process based on experiment is demonstrated its diversity of sources and characteristics of being target-driven and systematized. Actors implement rules learnt from domestic traditions, local practices or outside arrangements, in different areas, for testing their effectiveness. Then they can amend rules on the base of feedbacks from implementation. The process from learning and drafting rules to implemenatation and amending is recurrent until a set of national institution formulated.;Theoretically, the research proposes a new model for the theory of gradual institutional change, and a new learning process based on the experiment. The former develops a new tool to understand institutional changes, and the latter improve the analysis on human learning in institutional changes. Empirically, on the one hand, the effectiveness of the experiment in institution building is worth being emphasized, especially in transitional states with complex conditions, just like China. On the other hand, we should pay attention to the achievement from building the new system of urban administrative enforcement, when we criticize negative issues on it. After all, it needs comprehensive solutions from a broad approach to resolve problems along with the system. |
