台灣邊際消費傾向的估計

Abstract

政府的擴張性財政政策的效果往往取決於乘數，而乘數的大小又取決於邊際消費傾向。本文以2014年至2018年的華人家庭動態資料庫作為樣本，利用其追蹤資料的特性，同時使用個人固定效果模型進行台灣民眾的邊際消費傾向的估計，以及將樣本與收入在進行分類的情況下以相同的模型估計，並根據結果進行討論。實證結果顯示台灣民眾每增加100元收入，平均而言會增加15元的消費，而政府每增加100元的支出，平均會增加117.6元的國民所得。而從收入分類的結果來看，相對穩定的薪資收入與租賃收入變化時，對消費的影響是最大且最顯著的，這個結果與恆常所得假說的說法一致。另外，在樣本分類的結果可以看到沒工作的邊際消費傾向大於有工作、單身大於已婚、而單身的類別中的男性大於女性的結果。最後，本文也發現消費的增加會隨著收入的增加而減少，這個結果也得到基本心理法則的支持。

The effect of expansionary fiscal policy is decided by multiplier effect, and multiplier is decided by marginal propensity to consume (MPC). This paper uses the data from Panel Study of Family Dynamics (PSFD) from 2014 to 2018, take advantage of the properties of its panel data and use a personal fixed-effects model to estimate MPC, and do the same estimate with the same model with the classified of income and the sample. According to the empirical results, it is found that if people in Taiwan earns hundred extra dollars of disposable income, they will spend 15 dollars on average, and if government spends hundred extra dollars, it will increase 117.6 dollars National Income(NI) on average. From the results of income classification, the relatively stable income, like salary income and rental income, have the largest and most significant impact on consumption, this result is consistent with Permanent income Hypothesis. In addition, in the MPC estimation of sample classification, it shows that without a job is larger than that of having a job, being single is larger than being married, and males is larger than females in single. Finally, this paper also finds that the increase in consumption will decrease with the increase of income, and this result is also supported by fundamental psychological law.