氣候變遷問題對全球經濟發展與人類生存均造成極大威脅，促使人們高度關注環境保護議題，故消耗地球資源最多的企業是否充分揭露環境資訊因此顯得相當重要。 以往金融機構自詡為無煙囪產業，但金融業身為許多企業資金的貸放者，卻同時也是造成全球暖化日益嚴重的罪人之一，此時，因應而出現的是許多二氧化碳碳排量計算的標準或是溫室氣體盤查的議定書， 溫室氣體盤查便是能夠使這些企業藉由量化的數據來參與環境保護的議題及省思。它們的出現不僅幫助企業自行檢視二氧化碳碳排量以達環境保護的要求，亦透過揭露來符合國際公約或該國法規規定以減少企業所可能面臨的罰鍰成本。 文獻部分利用前人文獻來討論金融機構進行溫室氣體盤查的動機進而做為本研究主題的成因，實證則利用廣義卜瓦松之單變量與多變量迴歸來探討模型的配適程度以及假說檢定，結果發現該國的聲音與責任程度、該金融機構有無上市以及該金融機構之員工數對於金融機構重視溫室氣體盤查影響最為顯著的三個因子，除此之外利用分年迴歸也發現，聲音與責任以及該金融機構有無上市在2009年呈現顯著。 最後針對有符合假說檢定的因素進行討論並給予後人如何延伸本研究的建議，能夠使本研究更臻於完美。 ;Climate change is always a big threat for human existence and global economic development, so it makes human start to focus on environment protection issue considerably. Thus, it is very important for those enterprises which consumed resources at most on earth to disclose their activities or information of environment fully. In the past, financial institutions were viewed as harmless industry to environment, but actually they are not only suppliers of funds for enterprises which were resource of carbon emission, but also a sinner of global warming. Owing to this situation, many accounting methods of carbon emission and greenhouse gas inventory protocols are published and launched. Greenhouse gas inventory can make firms inspect quantitative data to understand what they did for environment. At the same time, these tools not only help these industries to calculate and inspect their carbon emission, but also associate them to meet limitation of international agreements and domestic regulations. I cited some references to talk about why financial institution will executate Greenhouse gas inventory. In the analysis of regression, we found the “Voice and Accountability”, “No. Employees” and “Listed” are significant via univariate and multivariate Generalized Poisson regression. Yearly regression models aslo find the “Voice and Accountability” and “Listed” are significant in 2009. Finally, I discussed these variables in accordance with hypothesis in detail and gave some practical suggestion to extend this research and make it more perfect.
