每年各國都有許多報章雜誌以不同的績效指標來進行企業排名，檢視企業前一年度的經營成果，並就排名結果解析未來產業趨勢。一般認為排行榜中的企業可代表標竿企業，同時排行榜符合統計概念中的樣本，加上排行榜為利用前一年已知的財務資料，為落後指標，應可代表整體產業的發展，故有不少研究是直接使用排行榜的企業作為研究對象，但少有直接就排行榜長期資料進行分析的研究。 　　而本研究旨在研究歷年排行榜的資料與產業變遷的關係，以及台灣產業的特性。研究範圍使用台灣經濟新報Top3000資料，研究1991~2008年製造業營收前五百大企業，並針對有企業家數有明顯變化，且為台灣重點發展之半導體、光電、紡織產業深入探討。 結果發現台灣前五百大企業的入榜變動率與美國前五百大企業的入榜變動率接近，每年約28家企業入榜；前一百大的入榜變動率16.59%比前五百大的入榜變動率5.53%還高出許多；半導體、光電、紡織產業的入榜變動率深受到總體環境變化影響。而我們也證實獲利率和入榜家數有正向的線性關係，表示當該產業的平均獲利率上升時，該產業的入榜家數也會增加。最後，在規模報酬的結果中看到，台灣製造業五百大企業和美國Fortune500大一樣呈現規模報酬遞減，也意味著大企業的生產效率不見得是最好，大不代表強，不代表有競爭力。在討論加入研發費用變數探討後，我們也看到研發費用變數對五百大企業的營收貢獻是正向顯著的，從係數的比重變化來看，資本密集的產業也逐漸的走向知識密集產業。 排行榜可說是個量化的產業史，也充份具備了資訊提供的價值。 / For every country, every year, there are lots of newspapers and magazines report the achievements of enterprises according to many different financial performance indicators, making a ranking by examining the operating results of the previous year and analysis future industry trends. The enterprises in the ranking generally considered to represent the benchmark enterprises, while the ranking conforms to the statistical concept sample. Owing to the ranking which uses the financial information known to the previous year, as lagging indicators, should be possible to represent overall industrial development. Therefore, many researches take the ranking directly for the object of study, but only few researches take the long-term ranking data as an analysis. 　　However, this research is for the purpose of studying the relationship between the long-term ranking data and the industrial vicissitude, as well as the characteristics of Taiwan industries. The study using the Taiwan Economy Journal Top3000 financial data, tries to study the top 500 enterprises of manufacturing industry from year 1991 to 2008, and aims to thoroughly discuss the semiconductor industry, the electro-optic industry, the textile industry which are the focus industries and their entrepreneur numbers has obvious change. The results show that the entry/exit change rate of Taiwan top 500 enterprise ranking is close to the change rate of U.S.A top500 enterprise ranking. The change rate shows that approximately 28 enterprises enter the top500 enterprises ranking every year. Besides, the change rate of top100 enterprises rank, 16.59%, is higher than the change rate of top500 enterprises rank 5.53%. The change rates of semiconductor, the electro-optic, the textile industry are deeply influenced by the overall environmental variation. Furthermore, we also confirm that the profit rate and the number of entry top500 ranking have the positive linear relationship. It means that when the industrial average profit rate rises, the entry number of this industry increases as well. Finally, we can find out the return on scale of the top 500 enterprises in Taiwan manufacturing industry, the same as American Fortune500, presents the decreasing returns on scale, which also means that production efficiency of large enterprises is not necessarily the best. Big corporate does not mean that they are stronger or more competitive. After adding the research & development expense variable into discussion, we also see the research & development expense variable for top500 enterprises’ revenue is a significant positive contribution. Looking from the change of coefficient proportion, the capital-intensive industries have gradually transformed into knowledge-intensive industries.. The ranking can be said a quantitative industrial history, but also fully equipped to provide the value of information. Key words: The ranking, the change rate, the profit rate, the return on scale

廠商規模要多大才好呢？而在整体經濟來看，廠商規模報酬的趨又是如何呢？這些均為學者們所關心的問題。在產業經濟上，多方面的研究規模報酬為一大趨勢，了解廠商規模報酬的變動則為一首要的問題。 本論文以台閩地區八十年工商普查調查磁帶中，製造業二欄位的產業母体資料，建立非線性的規模報酬和体現型技術變動之Solow生產函數為實証模型，對製造業作一實証分析，並探討廠商有無出口和研究發展行為下廠商規模報酬之異同。 由結果顯示： １、台灣製造的規模報酬並非固定不變的，為變動的規模報酬，規模報酬會隨著規模變動時而有所變動，但在有出口廠商的情形不明顯，除食品外均為固定規模報酬。 ２、比較有無出口行為之廠商時，出口廠商之規模報酬高，沒出口廠中成衣服飾業為規模報酬遞減。 ３、整体來看時，製造業以較大和較小的廠商規模之報酬較高，但在從事研究發展時，以中型廠商來促進產業轉型與成長較有利。 ４、製造業為体現型的技術進步，技術進步會体現於機器設備的異同，並存在學習效果，但在有研究發展行為廠商並不明顯。

國際貿易理論有二大模型：即Heckscher-Ohlin-Samuelson 模型，以及強調特殊生產 因素之特殊要素模型（Special Factor Model）。 本文擬對特殊生產要素模型加以探討，並改變模型一些假設，分析模型之各種效果。 本文第一部分及乃就此模型，建立統一模型，比較Jones、Ikemoto以及Amano 等人之 主張，並敘述其一般性質。第二、三部分則擬改變假設，分別討論技術進以及規模報 酬變動存在情形下，對模型各種效果之影響。