Is the decentralization of authority necessarily good while centralization necessarily bad in land use governance regime regardless the economic and political environment of a county? Following conventional wisdom, dispersing decision-making governance closer to the people and/or citizens is often seen as a better choice for a representative regime because decision made with greater participation will be better informed and more relevant to diverse interests in society than those made by national political authorities. But how is about in an authoritarian regime? Why we see a centralized trend on the delegation of land use governing authority? Why we see a separation of authority among three ministries who have strong hands in land use governance in China?

My study demystifies the puzzle that China's land use governing regime has changed from a decentralized and united structure toward a central-controlled and authority-separated structure by applying transaction cost theory to illuminate China's unique path of institutional adaptation in the reform era. The main focuses of my study are the central-local relationship and inter-ministries relationship within China's land use governing system. From a comparative institutional perspective, I see land use plans as contracts (incomplete) made between parties and see land use governing organizations as contractual relations. Under China's special political and social situation, these parties consist of the State Council, local Government Offices, three land governing ministries in central government, and the sub-divisional bureaus of the three ministries in local governments, hierarchically organized within China's command and control administrative system. My study shows how and why land governing authority is delegated into different level of governments and separated into three ministries. The theoretical reasoning of my research bases on transaction cost theory and China's political economic environment. The empirical part of my research combines quantitative tests and qualitative tests drawing on a variety of land use policies since 1979, interviews with local cadres across different regions and governmental sectors, as well as strong statistical evidence.

In the first part of my research, I develop two theories to explain the institutional change of China's land governance. In the first theory, I focus on the central-local inter-government relationship. I develop a theoretical framework to explain why we see some land use governing authority-but not others-been delegated to local government level, and argue that decentralization is not always efficient to minimize transaction cost under China's political environment and land property right legislation. I identify two characteristics of planning authority-the profitability and information intensity-as two dimensions for ramifying different type of planning authority into a two-by-three matrix, where a different category indicates a different governance structure (centralized, semi-centralized or decentralized).

Specifically, in the context of incomplete contracts, transaction cost stems from bounded rationality and opportunism. In China's land use governing regime, opportunism is associated with the profitability of a particular governing authority, because the more profits a governing body can obtain from the authority, the more possibility the governing body abuses the authority and violates the land use plan. Bounded rationality is associated with the information intensity of a particular governing authority, because the more details a land use plan is dealing with, the more information is needed to make the plan, and hence the more bounded rational the plan making body become. Under this context, when a particular planning authority generates no profits, delegating authority to the local level is the optimal choice. When a particular planning authority is associated with profits, decentralization is not optimal. Whether or not to employ a centralized structure or some semi-centralized structure depends on minimizing the transaction costs-economizing on the cost of making a detailed plan (solving the problem of bounded rationality) and simultaneously reduce the cost to monitor a plan's enforcement (safeguarding the enforcement in question against the hazard of opportunism). Holding the governance structure constant, these two objectives are in tension, since a reduction in one commonly results in an increase in the other. However, different governance structure has different ability to reduce transaction cost, so that choosing governance structure (semi-centralized or centralized) can be regarded as a means of optimization. In this way, whether a particular planning authority should be centralized or decentralized is clarified.

To empirically test the theoretical reasoning, I review the institutional changes regarding central-local relationship in China's land use governing regime in the past 30 years. I find that the theoretical model can explain the centralized trend of land use governance very well-the centralized land use governing authority are those highly profitable and information simple kinds.

In the second theory, I focus on the inter-ministries relationship. I develop another theoretical framework to explain why we see a separation of land use governing authority among three ministries-the National Development and Reform Commission (NDRC), the Ministry of Land and Resources (MLR), and the Ministry of Housing and Urban-rural Development (MHURD), and argue that the separation of authority is necessary for central government to promote better land governance while it is unwelcome among local governments because it impedes local development progress. In a phase, the relationship (perfectly substitute, substitute at the margin or complementary) between the tasks into which land use governance can be broken is a key to determine the optimal governance structure (separated or integrated) for promoting governing bodies' incentive to specialize on the task to which they are given access.

In particular, because of bounded rationality and opportunism, land use plans can be regarded as incomplete contracts, especially when the term of plans is 20 or 15 years. Hierarchically controlled governance is needed to reduce transaction costs and to promote the enforcement of incomplete contracts, because the power which forms the hierarchical relationship (principal and agents) can protect and foster relationship-specific investments (investments which are useful only when the relationship continues and are helpful for fulfilling the incomplete contract). The purpose of optimizing governance structure is to maximize the aggregate specific investments (in other words, to economizing on transaction costs), and hence maximizing the principal's gains. Since power stems not only from the ownership of a crucial resource but also (and more importantly) from the access to the resource (the ability to use or work with the resource), optimizing governance structure necessarily leads to a choice on the structure of allocating access (to one agent or several agents). If the tasks of using a crucial resource are perfect substitute, a principal giving access to one agent is the best choice, because giving access to multiple agents will decrease each agent's incentive to invest. If the tasks are substitute at the margin, give multiple accesses will decrease agents' incentive, but the principal can achieve the greatest aggregate investments because he creates a "rat-race" among agents. If the tasks are complementary, then giving access to one agent becomes better again. Multiple accesses will grant too much hold-up power to each agent. In China's land use governing regime, the crucial resource is authority. From the State Council's (principal) point of view, land use governance consists several tasks which are substitute at the margin-urban development, farmland protection and environment conservation (a bit better farmland protection can be pursued at the cost of a bit worse urban development, but both farmland protection and urban development are necessary and un-interchangeable). Therefore, in order to promote aggregate specific investments, it is better to separate authority into three ministries. On the other hand, from local governments' point of view, urban development is the main and possibly the only concern. Thus, the tasks of local land governance are complementary kinds: (1) decide the boundary of urban built-up area (2) decide land use functions (3) regulate FAR and Building Height, (4) approve development applications. Separating the tasks into different agents will grant too much hold-up power to each agent, thereby lagging the development process. As a result, the local Government Offices (principal) are unwilling to separate authority into multiple local land governing bureaus (agents).

The theoretical reasoning follows by another empirical investigation on the institutional change about the inter-ministries relationship in the reform era. The evidence demonstrates the predictions very well. In China's land governing system, there is an obvious trend of separating land governing authorities between three ministries. On the other hand, since the separation of authorities does not match local governments' interests as predicted in the theoretical model, they tend to integrate three local bureaus. Evidence from several local land governing practices also demonstrates such a trend.

In above two theoretical models, the main point is that the institutional change of land governance is an adaptation process where governance structure for a particular authority moves toward the optimal one (centralized or decentralized, separated or integrated). In the first theory, optimizing governance structure is a choice between more hierarchical levels and less hierarchical levels, while in the second theory the optimization is a choice between more agents and less agents at one hierarchical level.

In the second part of my research, I analyze the advantage and disadvantage of China's land governing system. The advantage of the centralized and authority-separated structure, as argued, is that it helps to prevent local land from excessive use. Drawing on above theoretical reasoning, I generalize a structure of China's land governance, where the horizontal jurisdictional governing bodies and the vertical divisional governing bodies weave an intricate bureaucratic net of China's authorities from the top to the bottom. By drawing two models (decentralized model and centralized model), I argue that the land use behaviors are different among local governments with different administrative structures. Specifically, under decentralized model, in which horizontal relationship between parallel ministries or bureaus dominates vertical relationship between higher level divisional bureaus and lower level divisional bureaus, local government officials tend to change land use plan to use more land in order to promote economic development; on the contrary, under centralized model, in which vertical relationship overwhelms horizontal relationship, government officials usually have difficulty on changing land use plan and then use less land.

To evaluate the theoretical argument, this research tests the discrepancy of locals' land use from two ways. First, I investigate the role of the Ministry of Land and Resource (MLR) on restricting local land use. The argument is that the decentralization index represented by the coverage of local Bureaus of Land and Resource (BLR) in a province will have positive effects on controlling local land used of that province. The greater the index is, the less land the province uses. Employing an Ordinary Least Square with fixed effect model on a panel database on 29 provinces during the period 1999-2008, a total 290 observations (China Statistics Year Book, National Land and Resource Statistics Year Book), I found the results are in favor of the hypothesis. In the second test, I use an Ordinary Least Square with fixed effect model on municipalities (prefectures) level land use data in 287 cities in China from 2006 to 2007 (China Cities Statistics Year Book, National Land and Resource Statistics Year Book). The analysis finds confirm my argument that the land use behaviors are really different between prefectural level cities and province level cities, where the administrative structures are different.

As for the disadvantage of the central controlled and authority separated structure, I argue that the asymmetric information in planning making process will lead to large local discretion. I conduct interviews with local cadres across different regions and governmental sectors, and investigate the pattern of local discretion as a response to the central controlled and authority separated structure in detail. The empirical evidence demonstrates my arguments very well.

In a phase, China's current central controlled and authority separated land use governing structure is an institutional adaptation in response to China's political economic environment-economic reform, the authoritarian regime and the land property right legislation. It is not without merits, but the disadvantages are also obvious. Moreover, from my theoretical and empirical analysis, I find that even though the political economic environment can hardly been changed, there is still a large room for us to improve the current governing system.

This research is an initial effort to model the political structure of land use governance in an authoritarian regime by adapting the transaction-cost theory from economics. It also provides theoretical explanations and empirical evidence on the institutional arrangement of land use governance in China. Practically speaking, it provides valuable suggestions on the improvement of China's land governance. In a broad sense, it is also a good reference for people who would like to study China, China's land use, authoritarian regime, organizational modeling of urban governance, as well as quantitative analysis of land use governing organizations.

中国では、市場経済化政策による経済発展政策のもとで、開発重視の都市政策が進められている。この結果、立ち退き問題、農地の過剰な都市用地への転換、投機目的の開発、過剰融資問題など、土地開発にともなうさまざまな課題が山積し、土地開発管理の強化の必要性が指摘されているが、都市計画制度ならびに土地管理行政システムの観点からの研究は十分とは言えない。このような背景のもとで、本研究の目的は以下の3点である。

(1)土地管理に関わる権限の中央政府と地方政府間における垂直的な権限移譲と土地管理に関わる省庁間の水平的な権限の分離について理論的に検討する。

(2)土地管理行政システムの機能と課題について実証的に分析する。

(3)上記の分析を踏まえて、中国における土地管理システムの改善方向について提言する。

研究の方法は、理論的なモデル構築をベースとして、土地利用転換に関わる統計データの量的分析ならびに各レベルの地方政府担当官に関するインタビューをもとにした質的分析に基づいている。

本論文の概要は以下のとおりである。研究の目的、既往研究のレビューならびに分析枠組みを示した第1章に続いて、第2章では、研究全体の前提条件となる中国の土地管理システムの現状と課題について分析している。第3章では、中国における中央政府と地方政府間の権限移譲のプロセスに関して、取引コストの概念に基づき、理論的なモデルを構築し、近年における中央・地方間関係の変化の要因を解明している。第4章では、同様に取引コストの概念をベースとしつつ、中国に土地管理行政システムに特徴的な省庁間の権限分離システムの形成プロセスを理論的に解明した。第5章では、中央政府・地方政府間の垂直的関係と地方政府内における水平的関係の両者の土地管理に対する作用メカニズムについてゲーム理論をベースとして理論的な検討を加えている。第6章および第7章において、理論的検討をベースとして、土地管理システムがいかに機能しているかに関して、統計的な分析とインタビューをもとにして質的な分析をもとに実証的な分析を行った。第8章では、以上の検討をもとに結論を提示するとともに、今後の中国における土地管理行政システム改善の方向性に関する政策的な示唆を示している。

本研究により得られた知見を以下のようにまとめることができる。

(1)現在の中国における政治経済的なシステムを前提とした場合、現在観察されるような土地管理システムの中央集権化へ向けての動向は理論的にも実証的にも意義があることが確認された。

(2)また、同様に土地管理行政システムにおける省庁間の機能分離政策についてもポジティブな意義を有していることが確認された。

(3)一方、現状では、地方政府における恣意的な判断が大きく容認されるシステムとなっているが、この恣意性が発揮されうる程度が地方政府間で違いがあり、この違いが土地開発の程度に関係していることが実証的に明らかとなった。

以上の知見に基づき、本研究では、土地管理行政システム改善方向への政策的な示唆として、中央政府と地方政府間の情報取得に関わる条件の違いが恣意的な土地開発を引き起こす要因として機能することから、中央政府による土地管理に関わる情報コストの低減、モニタリング機能の強化、透明性の拡大が必要であること、また、土地管理に関わる省庁間の機能分担の輻輳を整徐化し、役割分担を明確化することで、レント・シーキング的な行動を抑制することが重要であることを提言している

本研究は、中国における土地行政システムの機能について、理論的、実証的なな分析を加えたうえで、今後の改善のための政策課題を明確に提示しているという点で、他に類例のない先駆的研究であり、学術的に優れた価値を有していると同時に、土地管理システムを改善するにあたってきわめて有益な提言となっている。

よって本論文は博士(工学)の学位請求論文として合格と認められる。