本研究採用橫跨20年期間的台灣家庭收支調查個體資料,以分量迴歸 (quantile regression) 分析以下三個主題:一、預期所得與非預期所得變動對於家計消費的影響。二、不同所得群組與家計消費水準之間的關係。三、住宅資本利得對於家計消費的影響,強調擁屋家庭的財富效果 (wealth effect) 及租屋家庭的怯志效果 (discouragement effect)。實證結果顯示,非預期所得的邊際消費傾向大於預期所得的邊際消費傾向,此結果與Friedman (1957) 所提出的恆常所得假說 (permanent income hypothesis) 不一致,也與Miles (1997) 的實證發現相反。其次,高所得家庭相對於低所得家庭有較高的消費水準,此所得對消費的邊際效果隨著消費的分配而不同。此外,住宅價格上漲對租屋家庭產生放棄購屋減少儲蓄的怯志效果;房價上升對擁屋家庭帶來財富效果,增加家庭消費而抑制儲蓄。2000年資料下房價的怯志效果與財富效果大於1980年和1990年資料,顯示1980年代中期以來的房價大漲,到了2000年資料才反應出來。
The main purpose of the research is to employ micro data over twenty years in Taiwan to analyze the following three topics with quantile regression: first, the influence of anticipated and unanticipated income on consumption of the household; second, the relationship between consumption and income of across different income groups; third, the impact of housing capital windfall gain on consumption focusing on wealth effect for homeowners and discouragement effect for renters, respectively. Empirical results show that the marginal propensity to consume out of unanticipated income is greater than that of anticipated income. The result is inconsistent with Friedman’s permanent income hypothesis (1957), and is opposite to Miles (1997) as well. In addition, higher income household show higher consumption propensity and the marginal effects are significantly different across the conditional distribution of consumption. The rises of housing price will discourage saving propensity of renters who give up buying houses and consume more. Analogously, owners save less due to standard wealth effect. Furthermore, the wealth effect and discouragement effect in year 2000 are both greater than that in 1980 and 1990. It is possible that the effect of the rise of housing prices in the late-80’s takes time to appear in household consumption behavior in year 2000.