各國視生技產業為未來發展的關鍵產業，並積極推動各項政策，使生技產業能快速成長，而生技醫藥市場是促成全球生技產業成長的主要動力，為了增加我國的競爭力，希望藉由與鄰近國家醫藥產業的比較，能更了解台灣生技醫藥產業經營績效。 本研究採用Battese and Coelli (1995)隨機成本邊界法，針對2002-2007年間，日本、南韓、中國、印度與台灣等五個國家，共61家生技醫藥廠商進行實證分析，研究結果如下：(1)研發密集度增加使成本效率降低，五個國家裡，日本最具成本效率。(2)產業平均成本效率值為0.855，且有逐年惡化的趨勢。(3)整體產業平均處於遞增規模報酬階段。(4)整體而言，總要素生產力(TFP)的提升主要是因為規模成分的貢獻，其次為技術的進步，而技術效率變動率對TFP成長率為負影響。(5)各國間雖然TFP變動率不存在顯著性差異，但在規模成分、技術變動率與技術效率變動率等方面存在著顯著的差異。 / Many countries regard biotechnology as a key industry for the future development. Governments often implement a variety of policies to help it grow rapidly. The biotech pharmaceutical industry is the main momentum for the growth of the global biotech industry. The objective of this paper is to measure the productivity and efficiency of the industry among Asian countries, and investigates the sources of the performance changes, and then hope to give some insight into the enhancement of the industry’s productivity. To pursue our goal, we adopt Battese and Coelli’s (1995) stochastic frontier approach to assess 61 biotech pharmaceutical firms during 2002-2007. The main empirical results can be summarized as follows: (1) The R&D intensity is negatively related to cost efficiency; in five countries, Japan has the highest cost efficiency. (2) On average, the cost efficiency is about 0.86, and has become worsen year after year. (3)Most of time, the industry is characterized with the increasing returns to scale. (4) The growth of total factor productivity (TFP) is mainly attributed to the scale efficiency change, and technical progress accounts for a minor source. However, technical efficiency deteriorates over time. (5) Among countries, the TFP growth rates have no significant differences, but the components show apparent differences.

本研究包括3個研究議題。第1個議題旨為估算不同部門別（包括農業、工業、服務業與運輸業）的能源燃燒產生CO2排放之組成因素的貢獻量，係藉由拉氏指數法和算術平均迪氏指數法之加法型態，拆解5種不同的因素（包括：碳密集度、部門結構、能源密集度、人口及經濟規模等），觀察其對於CO2排放變動之影響。本文採用台灣1992-2008年的各部門別的資料作為分析的基礎，研究結果顯示，以上部門的經濟規模對於CO2排放的貢獻呈現巨幅的正向效果；人口因素則呈現微幅的正向或負向的影響；而碳密集度對於CO2排放減量有正面的影響，並發現此乃是構成改善能源結構並導致CO2排放減量的最重要因素；能源密集度因素的影響，除服務業以外，其餘部門均呈現負向影響，此一結果顯示，大部分部門要進一步改善其能源效率頗為困難，是以未來致力於減排的努力，應著重於使用乾淨能源，尤其是以再生能源作為替代能源 (Liaskas et al., 2000)；此外，值得注意的是，部門結構因素對於大部分的產業，如農業、工業和運輸業的CO2排放減量有正向的影響，據此可推論，我國的部門結構已漸趨向於低耗能產業（如服務業）發展。另外，本文採用近似不相關迴歸模型，探討各項政策工具（如環境稅、進口關稅）與經濟變數（如貿易條件及時間趨勢等）透過以上5種不同的組成因素，對於CO2排放變動的影響效果作一分析，其實證結果可供決策者制定減排政策的參考。 第2個議題係為建立節能減排的有效政策工具，須先詳實掌握各項政策工具對節能減排與經濟成長的影響，乃深入回顧相關文獻之理論與實證方法，據以建立適合台灣的3E聯立模型，並進行實證分析，藉以推估多項政策工具（如環境稅、關稅等及能源價格等）與經濟變數（如貿易條件、所得等）對於節能減排與GDP的影響。實證分析結果顯示，台灣之能源消費、CO2排放、及GDP對於各項政策工具與經濟變數之彈性不僅各異其趣，而且有些彈性並非固定不變，可隨時間經過動態調整。 第3個議題係利用台灣1992-2008年之農業、工業、服務業與運輸業等部門別的panel data，仿照Battese and Coelli (1995)提出之隨機邊界(Stochastic Frontier Analysis, SFA)模型，建構隨機生產邊界函數 (stochastic production frontier function)與隨機能源需求函數 (stochastic energy demand frontier function)，利用最大概似法估算出各部門的GDP與能源需求之隨機邊界與技術效率 (technical efficiency, TE)，並據此實證結果提出政策建議。 / The thesis includes 3 issues of research. The first research aims at identifying the factors that have influenced change in the level of various sectors (agriculture, industry, service and transport) CO2 emissions from energy use. By means of both Laspeyres index method and the arithmetic mean weight scheme expressed separately in the additive form, the observed changes are analyzed into five different factors: CO2 intensity, structural change, sectoral energy intensity, sectoral employing population and output level. The application study refer to 4 sectors of Taiwan between 1992 and 2008. The obtained decomposition results indicate that the examined sectors the value calculated for the output level effect present the highest value appearing positive contribution of CO2, and the contribution from population is slightly increased or decreased, while CO2 intensity has beneficially influenced the reduction of CO2 emissions, as well as the improvement of fuel mix found to be the most important factor that lead to the reduction of emissions. In most of the examined sectors for the energy intensity factor present positive effect on CO2 emissions, the only exception is service sector showing negative impact on CO2 emissions, which can be stated as Liaskas et al. (2000) that as further improvements in energy efficiency in most sectors become more difficult, efforts to reduce CO2 emissions will be predominantly directed towards the use of clean energy forms and especially towards the deployment of renewable energies. It also should be noted that structural change has positively influenced the abatement of CO2 emissions for the most sectors such as agriculture, industry and transport. We conclude it shifts towards less energy-intensive service sector, due to have negative influenced the observed decrease in CO2 emissions for higher energy use sectors (industry and transport) and agriculture,. In this article, we also use a seemingly unrelated regression to further investigate the policy tools how to change in CO2 emissions level by the five different factors. The results indicate that policymakers may reduce emissions considerably through various policy instruments. The second issue focuses on initiating effective policy to save energy and reduce emission, one needs to reasonably capture the potential impacts of various policy instruments on energy consumption, CO2 emission and economic growth, the second research, after extensively reviewing the literature, builds a locally ideal empirical model that facilitates the estimation of various policy elasticities. The empirical results indicate that policy elasticities may not only differ from one to the others, but also change dynamically, implying the 3E impacts of some policy instruments might be weakening over time. The main goal of the third article is to provide a detailed analysis of productivity and efficiency measurement for panel data on four different sectors from Taiwan over the period 1992-2008. We use a stochastic frontier model set by Battese and Coelli (1995) to build a stochastic production frontier function and a stochastic energy demand frontier function, which are estimated by maximum likelihood to obtain a stochastic frontier of GDP and energy demand, as well as technical efficiency. On this empirical results, we suggest that policymaker may simultaneously make top-down policies (green tax reform, increasing environmental tax etc.) and bottom-up policies (fuel price in line with prices of gas in global markets) to increase energy efficiency in different sectors.