這篇論文的撰寫目的是要在當前盛行的「極大化序數總效用典範」與林忠正提倡的「序數邊際效用分析法」下，對照式的方式探討課徵均一所得稅導致商品需求量減少的原因。 首先，我們在一個「稅後所得」與「所得稅率」變化會直接影響人們對商品的偏好的極大化序數總效用模型中，發現有3種符合序數總效用概念的效果會使課徵均一所得稅影響商品需求量的增減，並且，據此我們發現這3種效果會組合形成19種課徵均一所得稅導致商品需求量減少的原因。然後，在一個一般化設定的序數邊際效用模型中，我們能確定地指出有23種課徵均一所得稅會影響商品需求量增減的「只有兩股力量運作的單純搭配方式」；至於課徵均一所得稅會導致商品需求量減少較複雜力量的搭配方式，我們發現其數量相當多(應該會超過一萬種)，多到不易輕鬆指出其準確數量。這與在極大化序數總效用分析架構中的19種原因的結論，有非常大的差異。藉此，我們體會到在不同的兩種分析架構下，我們會看到相當不一樣的世界，也就是，會建構出非常不一樣的世界觀。 其次，我們在新的個體選擇理論─「序數邊際效用分析法」─探討為什麼課徵均一所得稅會減少商品需求數量的其中一種可能原因。此原因乃源自「均一所得稅率提高商品邊際效用下降」與「商品邊際效用遞減」兩股力量搭配而成。先假設消費者原先處於商品邊際效用等於支出邊際損失的均衡情況，現在均一所得稅率提高時，因「均一所得稅率提高商品邊際效用下降」的特性，使得商品的邊際效用下降，改變商品邊際效用等於支出邊際損失的均衡情況，形成商品邊際效用小於支出邊際損失的失衡狀況，此時消費者會減少該商品的消費數量。後續因該消費數量減少，由於「商品邊際效用遞減」的特性，而使得該商品的邊際效用上升，此有助於消費者所面對的商品邊際效用小於支出邊際損失的失衡狀況，朝向商品邊際效用等於支出邊際損失的均衡狀況調整，進而使消費者達到新的均衡點。這個推論過程，既合乎直覺又很合理。正常的消費者選擇理論，應是可呈現此種經濟直覺的推測才是。然後我們要告訴你，在林忠正提倡的序數邊際效用分析法中，的確可以輕輕鬆鬆地呈現此經濟直覺。而且，事實上，本文將是在經濟學的歷史上第一次由序數效用理論的觀點成功地表達出這項課徵均一所得稅會導致商品需求數量減少的原因的紀錄。然後，我們對照性地指出，在極大化總效用理論卻不能合宜地呈現此項論述。藉由這分析，我們會讓大家了解為何經濟學家普遍採用的「極大化總效用分析法」做不到的事，在新的序數邊際效用分析法中卻是可以輕輕鬆鬆地做到的。你可以輕鬆地體會到現代理論的不合理性與侷限性與新理論的合理與自然性。新舊理論的高下其實可以再次清楚地呈現出來，這證實林忠正一向的說法：「對於任何一個由效用理論出發的個體選擇模型而言，新理論都優於舊理論」。 最後，我們在討論另一個為什麼課徵均一所得稅會減少商品需求數量的可能原因。此原因乃源自「均一所得稅率提高支出邊際損失上升」與「商品邊際效用遞減」兩股力量搭配而成。透過這個例子，我們再次呈現這原因可以在新理論中和里現身，但卻不能在舊理論中現身。再一次地，我們會讓大家了解為何經濟學家普遍採用的「極大化總效用分析法」做不到的事，在新的序數邊際效用分析法中卻是可以輕輕鬆鬆地做到的。新舊理論的高下其實可以再次清楚地呈現出來。 / The purpose of this paper is to discuss reasons of reduction in demand of goods when a flat income tax is imposed under two different models, the traditional classic of maximum ordinal utility model in contrast to the Lin’s ordinal marginal utility model. First of all, we can find three effects making a deduction in demand as a flat income tax imposed on goods in a maximum ordinal utility model, which assumes changes in after-tax income and in income tax rate will directly have an impact on consumer’s preference. These three effects combined to have 19 reasons causing the deduction of demand in goods. In contrast, we can firmly recognize 23 combinations by merely two forces in a general ordinal marginal utility model. As for all reasons of imposing a flat income tax that incur deduction in demand, there might be hundreds of thousands that are not to be easily counted exactly. A Huge difference between maximum ordinal utility model and marginal ordinal utility model is undeniable. Therefore, we see things totally different through two diverse models and construct different points of view. Secondly, under new micro selection theory, marginal ordinal utility analysis model, this paper discuses about one reason that might reduce consumption amount by flat income tax rate. This may be caused by two forces, tax that reduces marginal utility of goods and the law of diminishing marginal utility combined. Suppose there is a consumption equilibrium in the beginning. Raising the income tax rate will make the marginal utility of goods reduced and change the status quo, making consumer to buy less since marginal utility is no match with marginal loss of wealth or income. Afterwards, because of the law of diminishing marginal utility, the marginal utility of goods will increase gradually as the consumption amount is declining slowly, making a new equilibrium emerge. The above reasoning is logical and instinctual from human being and should be shown by a normal consumer’s theory. However, this paper shows the opposite. By adopting Lin’s marginal ordinal utility model, this fact can be easily depicted. In other words, this is the first paper that records the reason that imposing a flat income tax will cause a reduction on the amount of consumption. Subsequently, this paper does a contrast to traditional maximum utility model and finds traditional model cannot present this reasoning properly. This paper make it clear what a maximum utility cannot do, though it is prevailing, and what can be done by the new model in the following analysis. One can easily recognize why traditional model is unreasonable and how limited it is, and how reasonable and natural the new model is. The pros of new theory and cons from traditional theory can be shown easily, which proves Lin’s saying “For any model that rooted from micro utility theory, new model is better than traditional one.” Finally, this paper focus on the other reason that flat income tax cause a decrease in consumption amount. This result is caused by two forces, increasing marginal loss from larger expenditure and diminishing marginal utility from goods. This example shows us again that traditional model are not able to explain this logical reasoning, but new theory can easily do. Over and over again, this paper proves that new theory is better than old one.

有鑑於歐洲各國的福利政策與福利支出相較於其他國家高，歐洲國家多會提供窮困與殘障的人民最基本的社會保障，一般人民亦可享受到基礎醫療保障，在失業時也可以領取失業補助；歐洲各政府甚至會收購即將倒閉的企業或者提供補助使企業能夠繼續經營。然而在持續延燒的歐債危機下，為維持歐元區普遍的薪資和福利水平，歐元區內國家開始採取了國家借貸的做法，但由於國際經濟形勢不佳，歐元區經濟增量未達預期，許多國家原有債券陸續到期，若無法借到新貸款，國家將面臨倒閉危機。因此本論文主要以動態隨機一般均衡模型(DSGE)，探討一個封閉經濟體系下，政府支出進入家計單位效用函數後，換言之政府支出變動將會影響家計單位消費的邊際效用，試著尋找能夠極大化社會福利的政府公債占國內生產毛額的最適比例。我們發現在提高政府公債占國內生產毛額比例時：家計單位將預期未來稅賦繳納之金額更高，因而減少消費、增加儲蓄，所以會排擠掉部分私人消費；同時政府必須提撥部分費用以支付債券利息，所以將排擠掉部分政府支出。另外隨政府公債占國內生產毛額比率上升，整體社會福利水準會逐漸下降；因此，我們認為最適政府公債占國內生產毛額比率應為零。 關鍵字：動態隨機一般均衡模型、政府支出、邊際效用、公債、國內生產毛額、社會福利水準 / In Euro zone, large spending obligations needed to support the welfare state and redistribute wealth in an effort to gain greater equality. Most European countries provide liberal social security benefits to the poor, disabled, basic medical needs and very liberal unemployment benefits. They also own and run large public companies. Under the ongoing Euro debt crisis, European governments figure out the way to maintain the high welfare level by increasing the public debt they hold. In our paper, we would like to investigate the optimal ratio of public debt to GDP by constructing a micro-based dynamic stochastic general equilibrium model (DSGE) under a closed economy. We focus on discussing the optimal ratio of public debt to GDP which maximizes social welfare in economy while putting the government expenditure into households’ utility. That is to say, the government expenditure will influence the marginal utilities of households. We find out that as a government decides to raise the debt to GDP ratio; it will crowd out both private consumption and government expenditure. Because households will expect to pay more tax in the future, they will decrease their consumption and increase their saving; high debt ratio means government should have paid more interest payment in the future. Also as the ratio of public debt to GDP rises, the social welfare becomes lower. Thus, in our findings, the optimal debt ratio to GDP should be 0. Keywords: DSGE, Government expenditure, Marginal utility, Public debt, GDP, Social welfare