依捐贈多寡與變異程度再探捐贈行為的租稅價格效果

Abstract

了解現行租稅制度如何影響私人捐贈的決定，一直都是經濟學家的研究興趣。一般而言，現行制度主要是透過所得稅法中對捐贈給予減免，降低其租稅價格之後，進而鼓勵私人捐贈的增加。因此，對政府而言，捐贈的租稅價格效果的估計就很重要。文獻上不論是用何種的估計模型，均是估計一個“平均”的價格彈性，但這樣的方式可能會受到少數大額捐贈者的影響，而使估計結果有所偏誤。因此，為了減輕這樣的估計偏誤，本研究將依據捐贈金額的多寡加以分組，分別估計不同金額水準的組別下，捐贈者的價格彈性。另外，我們也將利用分量迴歸的估計結果(quantile regression)，檢驗小額捐贈者與大額捐贈者的租稅價格彈性是否會有顯著的不同。再者，學者認為捐贈具有習慣上的僵持性(habit persistence)，因此當期的捐贈會受到前一期捐贈決策的影響。然而，一般文獻卻未更進一步地了解習慣性的強弱對捐贈價格彈性的估計所造成的影響。因此，本研究亦將利用美國IRS 的長期追蹤資料(panel data)，計算捐贈者在某一期間內捐贈金額的變異程度，進而定義捐贈慣性的強弱，並且分析慣性的捐贈者(persistent donor)與隨機性的捐贈者(stochastic donor)在面對捐贈租稅價格的改變時，其反應是否會有明顯的差異。總而言之，本研究希望能以捐贈金額的多寡與變異程度的大小，分別估計不同水準下捐贈的租稅價格效果；並冀望本研究所獲得的實證結果除了學術上的貢獻之外，在政府利用租稅制度來鼓勵私人捐贈時，能提出政策設計之參考意見。

Understanding how the tax system affects the charitable contributions overall and across different groups of individuals is the main question in the study of giving behaviors. Existing literature usually estimates the price effect based on the conditional mean of charitable giving. However, the conventional regression technique is not adequate for charitable contributions because its conditional distribution has a non-standard shape. There are always a few individuals contributing a huge amount of giving which makes the distribution asymmetric. In order to alleviate the estimation bias due to the tailed samples, we estimate the price elasticities of giving for different groups of individuals with different levels of giving. Further, we also apply the quantile regression method to test whether individuals will respond differently to changes of the tax price of giving when their donations increase. In addition, literature has demonstrated that previous charitable contributions will affect current giving decisions according to the habit persistence hypothesis. However, no studies estimate the price elasticities of giving for donors with different intensity of habit persistence. In this study, we try to use the panel data to test whether individuals will respond differently to changes of the tax price of giving when their giving habit becomes more persistent. We expect this study would provide a better insight into the study of charitable contributions. In addition, we also expect our empirical results could draw some valuable policy implications and suggestions to the government and fund-raisers when they make their effort to stimulate charitable contributions.