自民國76年實施天空開放政策以來，國內航空市場與營運環境受到相當程度的衝擊與影響。面臨與日俱增的競爭壓力，各航空官司若無法積極改善其經營效率和健全其成本結構時，勢必島市場所淘汰。故本研究透過成本效率的觀點，使用符合全面凹性更具彈性之對稱性一般化影子成本函數來檢視國內各航空官司民國78-86年期間之分配扭曲對密度報酬、投入要素需求量(勞動、燃油、固定要素及其他)及成本的影響，以了解各航空公司無效率之所在。其主要結論如下：(1)藉由相對燃油分配效率探討各投入要素之分配扭曲情形發現:勞動和固定投入要素之分配無效率因官司別而有所不同，且隨時間而嚴重，唯民國86年稍有改善；其他要素之分配無效率則較不因官司別而有所不同，但亦隨時間發生變化，初期為無效率，在民國的、86年則較具效率；(2)若考慮分配扭曲(無效率)情形下，國內航空業在民國85年之前具有較高之密度經濟，民國86年則趨向固定密度報酬；就整體投入要素之技術效率而言，在天空間放政策後其技術效率隨時間變化而下降，但就燃油投入要素平均技術進步率而言，則其技術效率逐年增加且幅度明顯；(3)航空公司因分配無效率所造成成本增加的情形相當明顯；(4)準價格工具變數對成本之影響並不明顯，此可能反映在此期間國內航空官司似乎並未趨向於高品質高價格之現象。

Since the announcement of the so-called open-air policy in 1987, the domestic airline market environment in Taiwan has been significantly changed. All airline carriers must improve their managerial performance and maintain a healthy cost structure in order to meet the rising competition. This study examines the effects of allocative distortions on returns to density and input demand in domestic airlines during 1989-1997 by using a non-minimum cost function from the cost efficiency point of view. We generalize a globally concave and flexible cost function by using the symmetric generalized McFadden (SGM) cost function to accommodate allocative distortions for domestic airlines, and estimating the density economies with and without allocative distortion as well as its effect on cost efficiency. The major findings include: (1)By using fuel as the base indicator of relative efficiency, we found the following input factors’ distribution inefficiency for domestic airlines: The distribution inefficiency of labor and capital input factors differs from airline to airline, and is becoming more severe with time; the distribution in efficiency of other factors does not differ very much among airlines but varies with time. It is inefficient from 1989 to 1995 but becomes more efficient during 1996 to 1997; (2) Airline companies have higher economies of density before 1996, but tend to show a constant return to density in 1997 when considering allocative distortions. The technical efficiency of input factors as a whole has declined after the open-air policy; (3)The costs of airlines are increasing significant due to their distribution inefficiencies; (4) The instrument variable generated through a quasi price unction does not have significant influence on cost; this may imply that airline companies in this time period did not exhibit the phenomena of high­quality high-price.