長江三角洲的迅速發展已經成為改革開放之後中國大陸經濟發展的主要表徵，然而在面對全球化的挑戰下，長江三角洲整合的問題也成為各方關注長江三角洲能否持續發展的關鍵。 本文以區域創新系統理論解析改革開放以來的長江三角洲區域經濟發展變遷，並提出對長江三角洲區域整合議題的看法。 第一章為緒論─本文簡介，介紹本文之研究動機、目的，研究方法，研究範圍及研究流程安排。 第二章為新區域主義理論及文獻評述─本文的主要理論基礎為區域創新系統，而區域創新系統屬於新區域主義的重要流派。因此針對新區域主義的源流、相關理論基礎以及發展做一評述，並歸結出新區域主義重要的三個特色：在地化的地域空間與產業組織─地理空間；產業群聚與生產網絡的形成─網絡；區域發展的絕對優勢─創新。 第三章為區域創新系統理論─區域創新系統是新區域主義理論當中，結合空間、網絡、創新的重要流派，本章從創新理論到創新如何影響地理空間與產業的轉型，歸納出區域創新系統的立論基礎、結構、以及在方法論上的重要意義。 第四章為長江三角洲區域發展─應用區域創新系統理論作為分析框架，對於長江三角洲自改革開放以後的發展進行分析。 長江三角洲區域創新系統主要的特色在於地方政府治理性格上的差異性，以及地方廠商與產業的發展模式與專業市場，這些現象的產生都與地方背景與環境有關。研究建議，長江三角洲整合的目標應該是市場化、促進區域內網絡的交流以及持續產業群聚的發展。 / The rapidly development of Yanzi River Delta has become one of the main characteristics of China economic reform. Under the trend of globalization, the regional integration issues about this area has also become a main perspective whether Yanzi River Delta could ever continually develop. In this article, we use “Regional innovation system” to analyse the economic development and evolution of Yanzi River Delta, thus providing some perspectives of regional integration of Yanzi River Delta under the basis of our analysis. The first chapter is introduction, we introduce our motivation, the main purposes, the research methods and tools, the definitions, the arrangement and the process of the reasearch. The second chapter is concept backgound; we introduce and generalize the theory of New Regionalism. New Regionalism has been recognaized as the oringin of regional innovation system. New Regionalism is also the main stream in nowtime economic geography. We introduce its development, theory basis, and related theories development. We conclude that New Regionalism has three main pillar: localized geographical space, industry networks, and innovation. The third chapter is the introdcution of theory; we sum up the development, the structrue, and the methodology of regional innovation system. The fouth chapter is the analysis; we use regional innovation as framework to analyze the development and evolution of Yanzi River Delta afer the China economic reform began. The fifth chapter is the conclusion; the main characteristics of Yanzi River Delta is the defferent governence of local governments; the clusters, develop patterns and local market of local industries. These developments were highly influenced by local context. And we sujest the main objects of regional integration in this area is common market, the developments of regional networks, and industrial clusters.

近幾十年來，受到全球化與資訊化的影響，知識的創新能力表現成為競爭力優劣的重要指標，國家或地區經濟發展競爭力高低，與其全球化優勢息息相關，而全球化競爭優勢能力又取決於其創新成效表現高低。台灣在全球化的浪潮及大陸開放競爭下，地方傳統產業在此衝擊下面臨倒閉或出走。在這充滿變數與挑戰的階段，本文從區域創新系統角度探討地方創新環境與其創新能力表現對競爭力的影響。 本研究藉文獻探討及台灣地區鄉鎮市之基礎數據資料分析研究，以創新系統觀點探討影響地方創新成效的關鍵因素，並據此選定衡量創新成效指標，再針對台灣區域地方環境、產業發展及國內創新環境現況，探討目前地方創新活動的狀況與設施，建構台灣地區地方創新系統指標。再利用多變量分析方法進行實證分析，利用因子分析取得代表地方創新環境與區域發展特徵之因素，並將其萃取出創新能力因子，作為劃分創新能力集群之根據，分為高、低創新成效地區；最後印證創新成效表現優異地區，其所具有創新能力之特質與創新成效表現優劣間之相關性。 本研究實證結果如下：一、影響創新成效主要因素為製造業產業群聚與支援性服務設施；二、高度創新成效地區皆坐落於新竹以北地區；三、創新能力因子指數與專利數產出量不一定成正相關，創新能力因子指數高僅代表該地區環境具有利於產業間分享知識技術交流的能力與潛力，但此種能力不一定表現在專利數產出量多寡，而須倚賴觀察該地區的各項創新活動狀況；四、創新能力因子指數高地區，近3年（93-95）來的專利數產出遠超過過去10年(83-92)，推論地區內產業及創新設施已成熟，再加上智慧財產權保護觀念普及，促使專利數量大增；五、鄰近創新成效較高之地區，使其創新成效表現低卻屬於高度創新成效地區，推論其因位置直接毗鄰高度創新成效地區，受地理鄰近性影響甚大。 最後建議政府應協助各地區地方政府鼓勵產業投資發展，提高地方創新研發動能，並積極輔導大學與育成中心提高其育成與輔導的效率，未來更應建立媒合大學及育成中心與產業界之合作機制，以推動產、官、學、研各種人才流通，提高研發創新能力。 / In the past few decades, the influence of globalization and informationization has made the ability to generate innovations in knowledge a key benchmark in competitiveness. The competitiveness of a country or region's economy is now closely linked to its global advantage. This global competitiveness, in turn, hinges on its ability to innovate. In Taiwan, the impact of globalization and China's market liberalization has seen local traditional industries either shut down or move elsewhere. In a time filled with uncertainties and challenges, this study seeks to examine the effect of a regional innovation environment and its innovation capability on competitiveness from a regional innovation system perspective. In this study, we review and analyze the basic data from Taiwan's regional areas and examine the key factors influencing the effectiveness of regional innovation. This was then used to select the indicators for evaluating innovation performance. Taiwan's regional environment, industry development and the current state of the domestic innovation environment were then examined to determine the current state and infrastructure of regional innovation activities to construct a set of regional innovation system indicators for Taiwan. Empirical analysis was performed utilizing multivariate analysis. First, factory analysis was used to extract the factors that represent the regional innovation environment and regional development characteristics. From these, we determine the innovation capability factors, and use these to define the innovation capability clusters. These clusters were then separated into high and low innovation effectiveness regions. Finally, we examine the areas with high innovation effectiveness and determine the correspondence between their innovation characteristics and their innovation performance. The empirical results from this study were as follow: 1. The main factors influencing innovation effectiveness were industry clustering and support services & infrastructure; 2. The regions with high innovation effectiveness were all located north of Hsinchu; 3. The innovation capability factor index may not correspond directly to patent output. A high innovation capability factor index only indicates that a region possesses the capability and potential to facilitate the sharing of knowledge and technology between industries. This capability may not necessarily be reflected in the patent output, but is instead, depends on the innovation activities being conducted in the region; 4. In regions with high innovation capability factor indices, their patent output in the past 3 years (2004~2006) far exceed the preceding decade (1994~2003). We hypothesize that the maturity of their local industries & innovation infrastructure, as well as widespread understanding of intellectual property rights, encouraged the rapid increase in the number of patents; 5. Some regions near highly innovative regions perform less effectively even though they are part of the same area. We hypothesize that this is directly caused by their geographical proximity to highly innovative regions. Finally, we recommend that the government provide assistance for regional governments to encourage investment in industrial developments and boost local innovative R&D capabilities. Assistance should also be provided to universities and incubation centers to improve their incubation & assistance effectiveness. In the future, a cooperative mechanism for linking universities, incubation centers and the industry should be set up as well to boost the capability for innovative R&D by encouraging the free flow of expertise between the industry, government, academic and research sectors.

國家整體經濟發展向來與產業本身有著相互依存的關係，產業園區的設立更是被賦予經濟、政策上的特定目的。然而過去以計畫、政策為導向進行園區的設立已難以因應國際景氣波動及創新、知識型產業競爭衝擊，情勢是轉為考量產業根本及新興產業軸線的發展，由產業需求決定園區轉型、再生方向。 本研究一方面探究我國具備優勢性的產業聚落以產業群聚的角度著手，呈現出產業發展上最有效率的模式，另一方面是以典型探討的區域創新理論，反映區域網絡間創新效果以提高區域競爭力。針對兩層面的的正向能力評估目前國內各個產業園區其中的新興產業發展潛力，進而將產業適地適當的置入、整體的布局，作為促進地區新興產業發展啟動點，擴散至周遭的區域產業發展。 研究結果是得以將區域創新及產業群聚兩層面，轉化為實質架構體系做為衡量績效的評估模式，並可由編定工業區環域做為未來研擬產業發展分區的基礎。另外，在經評定下新興產業的高發展潛力地區仍是以六都所在之環域範圍，並劃分出策略核心發展區及主力發展區做為在有限的資源分配上可行的新興產業實質發展地區。 / A country’s overall economic development and its industries are interdependent; the establishment of industrial parks bears the specific goals of economy and policy. However, the establishment of an industrial park bases on plans and policies is getting more difficult to tackle the competition impacts of global economic fluctuation and innovation, knowledge industry. It’s about time to consider the basis of industry and emerging industrial developing axis, determining industrial park transformation and regeneration direction by industrial demand. There are two discussing aspects in this research, one of which is to investigate Taiwan’s advantaged industrial cluster, In terms of industrial cluster, the research tries to present the most efficient model of the industrial development; another discussion is about the classic discussion of regional innovation theory, through an overall discussion the research tries to reflect the innovative effect among regional networks so as to increase regional competitiveness. According to the positive ability of these two aspects, the research assess the developing potential of emerging industry existed in the current industrial parks in Taiwan so as to find the start of promoting local emerging industrial development and bring a comprehensive, proper industry in the local, extending the industrial development to its periphery. The result of this research is to turn the aspects of regional innovation and industrial cluster to a physical framework so that the evaluation model of performance measurement can be established; furthermore, an industrial surrounding area can be set up being a basis for planning industrial development segment in future. In addition, after the assessment, the emerging industries with high developing potential are mainly in the scope of surrounding areas of six municipalities, and due to an issue of limited resources distribution, the strategic essential development areas and the major development areas are divided becoming specific development areas of feasible emerging industries.

在創新系統中，產、學、研為新技術與知識積蓄的點，創新網絡中節點與節點之間的互動扮演重要的角色。區域內領導性廠商為了取得技術優勢，提供經費購買技術、引進技術等外部資源建立核心能力，在區域內形成產、學、研的技術聯盟組織，形塑區域特殊產業。各單位所交織形成的網絡動態為有助於區域創新活動的決定性因素。 透過這樣的角度來檢視台灣，本研究以區域創新系統的角度出發，觀察「數位遊戲」這項新興產業在技術與知識網絡中，行動者之間彼此的互動連結。將廠商、相關研究機構、產業公協會、人才培育機構、大專院校與育成中心視為區域創新網絡的驅動角色，透過問卷調查與社會學中的網絡分析工具，觀察網絡中技術與知識的互動關係。本研究的主要結論如下： ㄧ、由本研究實證結果發現，台灣北部區域所形塑遊戲產業產官學研合作所環繞的創新氛圍十分強烈。北部區域中，又以台北縣中和「遠東世紀廣場科技園區」與台北市「南港軟體工業園區」分別為遊戲產業高度密集聚點。 二、由於台灣數位遊戲產業各廠商之間的封閉性，各單位間技術與知識的互動關係偏低。廠商之間的網絡活動主要是以垂直性的關係為主，而非水平性的關係。廠商傾向倚賴國際之間網絡技術資源。 三、知識網絡中越是有能力的單位，越可能處於技術網絡的核心位置。也就是原本在遊戲產業中處於優勢地位的單位，基於資訊流通與知識傳遞方便的優勢，技術能力的掌握能力亦強，少數特定單位重複扮演技術與知識互動的重要角色。本研究從網絡結構中證實了技術與知識的高度關係。 四、技術網絡的連結模式呈現環環相扣的環狀結構，彼此交流與技術互動的相當頻繁與快速。知識網絡的星狀連結模式，具有明顯的核心單位，此類結構有利於知識的傳遞，處於散佈節點的單位居處特別重要的地位。 / In the innovation system, firms, schools and research institutes are new technology and knowledge storage places. The interactions between nodes play important roles in the innovation networks. In order to gain technology advantages, leading firms provide expenses to purchase and introduce technology to establish core ability. They form consortia with other firms, schools, and research institutes to build up regional specific industries in the regions. The dynamic network formed from the interaction of units is the main factor to facilitate regional innovative activities. This thesis is on the basis of Regional Innovation System (RIS) in Taiwan and takes “Digital Games” for example. Each actor interacts in the technology and knowledge networks in the RIS. Regarding firms, associate research institutes and industry associations as driving roles in the innovative networks, this thesis applies innovation networks surveys and the methodology of Social network analysis in Sociology to observe technology and knowledge interactions in the innovative networks. The major conclusions are as follows: First, “Far East Century Scientific Square” in Jhonghe City and “ Nan-Gang Software Industrial Park” in Taipei City are surrounded extremely strongly innovation milieu in the north of Taiwan. Secondly, owing to self-contained, digital games firms are deficient in interaction with each other. Activities in the networks are vertical relations almost, not horizontal ones. Firms tend to depending on international technology resources in the networks. The third, many competent actors basing on the advantage of information and knowledge transfer very quickly and conveniently in knowledge networks are located in the core of technology networks. A small number of firms repeatedly play the important role in technology and knowledge networks. From the network structure, this thesis proved technology and knowledge have a close relationship. Finally, technology networks display circular structures which exchange and transfer quite frequently and quickly; knowledge networks display star structure which have apparent core units and which are advantageous to transfer knowledge.

本研究以區域創新系統觀點，分析在兩岸區位優勢差異下，影響台灣製造業廠商兩岸研發區位選擇及廠商研發網絡治理情形，強化既往研究未同時關注台商、兩岸地區及偏重個案研究的不足之處；經實證結果發現，區位優勢、研發部門類型、市場規模、政策條件、產業群聚、知識流通及研發經費投入等變數為台商兩岸研發區位佈點的影響變數，除了市場規模及市場型研發部門與研發區位選擇呈現負向關係外，其他皆與區位選擇呈現正向關係，顯示除因台灣市場規模太小造成台商選擇至對岸設立研發部門外，台灣在其他表現上皆較對岸具有吸引力，台灣仍具有不可忽視的區域創新能量，建議政府部門應在產業政策上有新作為，提供廠商各項研發所需要的協助，以吸引更多本土企業於台灣設立研發部門，以不斷累積及提升台灣的研發競爭力。 區域創新系統強調區域內行動者的網絡連結及互動，形塑該地區無可取代的區域特色，為彌補研發區位選擇的量化研究，無法表達出區域創新系統內廠商在地化網絡連結及互動的情況，本研究在研發網絡治理部分以華碩企業集團為研究對象，探討其如何運用本身及當地資源，建立在地化網絡及維持網絡關係運作等網絡治理議題，經研究結果發現，由於華碩企業集團具有豐富的研發經驗、高水準的研發人員及雄厚的資金，扮演在地化網絡主要行動者的角色，掌握與其他行動者（例如政府部門、研究機構、大專院校及其他廠商等）間關係建立的主導權，在研發技術的傳遞上係以公司內部的垂直組織網絡為主要傳遞路徑，與區域創新系統內各行動者的網絡關係及在地鑲嵌程度不強。 / With the viewpoint of regional innovation system, this paper analyzes the impacts of R&D locational choices and enterporises’ R&D network governance that manufacturing firms in Taiwan have under the differentiation of locational advantages between China and Taiwan as well as consolidates some points which are incompletely focused in the past on Taiwanese firms, two sides of Taiwan Strait, and overweighing case-studies. The empirical evidences bring that the impacts of variable to locational distribution of Taiwanese manufacturing firms across the strait are many: locational advantages, the types of R&D unit, the scales of market, policy conditions, industrial clusters, knowledge flows and R&D investment. Except for the negative correlations between the scale of market/type of R&D unit and R&D locational choices, the rest variables have the positive correlation with R&D locational choices, which means that, barring that the scale of market in Taiwan is so small that Taiwanese firms choose to institute R&D spots in China, other variables in Taiwan are more attractive than those in China; Taiwan possesses innegligible energy of regional creativity. The suggestion is that the government has to take new actions on industrial policy, providing the assistance to R&D that enterprises need, in order to attract more and more local enterprises to set up their R&D functions in Taiwan for accumulating and advancing Taiwan’s competitive power continuously. Regional innovation system emphasizes on actors’ network linkages and interactions within one region, shaping irreplaceable regional characteristics for that region in order to atone for the quantitative research of R&D locational choices which cannot express the situation of enterprises’ localized network linkages and interactions within regional innovation system. Taking ASUS as example, this research, in terms of network governance, confers how ASUS draws on own and local resources for network governance issues like establishing localized networks and maintaining the operation of network relations. After the analysis, it is found that, on account of ASUS with abundant R&D experiences, high-level R&D staffs and tremendous funds, ASUS plays the role as key actor in the localized network, controlling the predominant power established by relations between other actors (i.e. public sectors, research institutes, colleges/universities, other enterprises and so on). In terms of R&D technology transfer, vertical networks in the inner company as main transfer passage is less attached with every actor’s network relation and its local embeddedness within regional innovation system.

區域技術創新能力是展現區域將知識轉化為經濟的能力，是區域競爭力的重要內容。由於區域創新能力是造成中國各區域經濟發展差距的主要原因之一。本研究採用中國之上市公司為研究對象，探討公司創新活動的揭露與中國各省（自治區、直轄市）創新能力對分析師預測行為及分析師預測跟隨人數之影響。由三個面向檢視區域創新能力：研究與試驗發展全時人員當量增長率(REGTH)、政府研發投入占GDP比例(GIGDP)、每百萬人平均發明專利授權數(AVPAT)。研究結果發現創新活動揭露越多之公司，分析師之預測誤差程度越低、分析師之離散度越高且能吸引越多分析師對公司做預測。區域創新能力指標REGTH越高地區，分析師預測跟隨人數越少；區域創新能力指標GIGDP越高地區，分析師之預測誤差程度越低、分析師之離散度越低、分析師預測跟隨人數越少；區域創新能力指標AVPAT越高地區，分析師之預測誤差程度越低、分析師之離散度越低、分析師預測跟隨人數越多。 / Regional innovation capability is an important competitive ability to improve the economic development in China. Using a sample of Chinese listed companies, this study investigates the extent to which innovation disclosure at the company level and regional innovation capabilities affect analysts’ forecast characteristics in terms of forecast error, and forecast dispersion, and the number of analysts’ following. The indicators of regional innovation capabilities are the growth rate of R&D full-time equivalent personnel (REGTH), regional government's investment in R&D per GDP (GIGDP), and average of innovation patent authorized per millions of people in one region (AVPAT). The results indicate that firms with more innovation disclosures improve analysts’ forecast error, and regions with higher GIGDP and AVPAT also improve analysts’ forecast error. For firms with more innovation disclosures increase analysts’ forecast dispersion, and for regions with lower GIGDP and AVPAT increase analysts’ forecast dispersion. I also find firms with more innovation disclosures attract more analysts’ following, and regions with higher REGTH and GIGDP have less analysts’ following, while regions with higher AVPAT attract more analysts’ following.