家庭消費、所得與房價：台灣家庭收支調查的實證研究

Household Consumption, Income and House Price: Evidence from The Survey of Family Income and Expenditure in Taiwan

楊茜文(Chien-Wen Yang)

指導教授 ： 黃朝熙

國立清華大學/科技管理學院/經濟學系所/博士(2018年)

摘要

本論文之二個章節主要涵蓋台灣家庭消費行為的實證議題。本論文利用台灣家庭收支調查資料與準追蹤資料 (pseudo panel approach) 方法，以家戶戶長年齡 25 -75 歲為基準建構本章的家庭消費與收入等資料來實證探討下列主題。第一章為實證台灣 1980 年代末期的金融自由化過程至今，家計部門借貸限制與消費行為的變化，第二章為實證探討房價與消費的關係。

第一章的研究著重於實證探討台灣自 1990 年代初期實施金融自由化至今，信用市場條件改變對家計部門借貸限制與消費對可預期所得變動反應的影響。本章實證以 1982 年至 2015 年台灣家庭收支調查與準追蹤資料方法建構家庭群組的消費與所得等資料，除了以特定時間點的虛擬變數視為判斷金融自由化前與後的影響外，由於金融自由化為一隨政策鬆綁而逐步改變的過程，本章運用不同方法所建構的金融條件指數，捕捉信用市場條件變化對個人借貸限制與消費行為的影響，並藉由財產所得淨額、可支配所得與年齡 (以家戶戶長年齡40歲為比較基準) 進行區分面對借貸限制低與高的群組，以釐清信用條件變化對不同借貸限制群組其所面對借貸限制與消費行為影響的差異。本章實證結果發現：金融市場條件寬鬆有助於降低台灣家戶的借貸限制，此影響為一逐步的過程，且以1980 年代至 1990 年代中期間特別明顯，與台灣信用擴張與經濟快速發展的背景一致。在個別群組部分，本章實證發現面對較高借貸限制的群組，其消費對可預期所得變動反應的敏感係數較低借貸限制群組為高，此反映各群組間所面對借貸限制程度的差異。此外，財產所得淨額或可支配所得較低的高借貸限制群組，其借貸限制受到信用市場條件改變影響的程度特別顯著；但年輕群組所受到的影響程度則與年老群組相當，其原因可能為年輕群組未來所得的成長受惠於金融開放的影響程度較年老群組為多，因此以年齡區分很難斷定低與高借貸限制群組間受到金融自由化影響的差異。

第二章之研究動機起始於資產價格與家計部門消費行為的實證研究，例如：Campbell and Cocco (2007) 與 Attanasio et al. (2009, 2011) 等文獻探討房價與消費的關係為來自財富效果所強調的因果關係，抑或是共同因素假說的爭論。本章實證利用不動產市場具有區域性的特徵，以及 1993 年至 2006 年家庭收支調查家戶戶長的出生年與區位條件進行群組的分類，並將其作為檢驗地區性消費成長的差異是否能由區域性房價漲幅的不同所解釋，以及房價對不同年齡與居住層別群組之間消費影響的差異；依據財富效果所述，房價對消費的影響在上述區域與世代之間應具有顯著的差異。本章實證延伸 Muellbauer and Lattimore (1995) 理性預期生命週期/恆常所得模型 (Rational Expectations Life-Cycle/Permanent Income Model)，一方面納入代表不確定性因素的死亡機率於模型中，另一方面利用區位性差異的特徵以及考量區域與總體經濟的共同因素進行估計。本章實證結果發現：區域性消費的差異亦無法由各區房價漲跌的不同所解釋；房價對消費的影響為正向，但在群組之間均沒有顯著的差異。此外，相較於代表新成屋價格的國泰房價指數，中古屋價格的信義房價指數較能反映房價對消費的影響。依據上述，本章實證結果傾向支持共同因素假說所描述的：觀察到房價與消費的同向變動為受到共同因素所致，此反映本章實證發現與 Attanasio et al. (2009, 2011) 以共同因素解釋英國房價與消費的結果一致。

關鍵字

消費 ； 所得 ； 房價 ； 金融自由化

並列摘要

This dissertation studies the topics on households’ consumption, financial market condition, and house price. To explore the issue, we employ the data from the Survey of Family Income and Expenditure in Taiwan (SFIE). As SFIE data are repetitive cross-section surveys rather than longitudinal panel data, we use the pseudo-panel approach a la Deaton (1985) to construct real consumption and income series based on the birth year of household heads. The first chapter aims to empirical study the role of credit constraint in Taiwan’s household consumption using the event of Taiwan’s early 1990s banking deregulation. The second chapter attempts to explore the relation between house price and household consumption.

In the first chapter, we use the event of Taiwan’s early 1990s banking deregulation to explore the role of credit constraint in Taiwan’s household consumption. We examine how the response of consumption to predictable changes in income varies over the financial market condition. We employ SFIE over the period 1982 to 2015 to construct consumption and income series for cohorts. In our empirical study, on the one hand, as the general financial environment may be influenced by not only banking deregulation but also other factors, we construct the index of financial condition to capture the changing financial environment and examine how it affects credit constraint and household consumption. On the other hand, since the credit constraints across cohorts tend to be heterogeneous, we consider the following measures to differentiate high and low liquidity-constrained cohorts: cohorts’ net property income, cohorts’ disposable income and the age of cohorts’ head (age below 40 vs. age above or equal 40). Our empirical study reveals the evidence for a gradual way of financial liberalization as well as the effect of relaxed financial market condition on the easing of these constraints particularly during the period of the mid-1980s and -1990s. More specifically, we obtain higher estimates for the sensitivity of consumption to predictable changes in income for cohorts facing a higher degree of liquidity constraint. We also find that the relaxed financial market condition has a significant impact on easing degree of credit constraint, especially for the cohorts facing a higher degree of liquidity constraint when we use cohorts’ net property income and cohorts’ disposable income to differentiate. However, there is no significant differences in age cohorts. These finding suggests that the older cohorts have higher income from property and disposable but young cohorts are better capacity for repayment and get more benefits from financial liberalization than older cohorts.

In the second chapter, our study is motivated by the debate in recent literature that the relation between house price and household consumption is explained by wealth effect or the common factor hypothesis, e.g., Campbell and Cocco (2007) and Attanasio et al. (2009, 2011). To explore the relation between house price and consumption in Taiwan, we focus on the characteristics that real estate market is regional and urban leads rural, and employ the sample of age and residence stage households over the period 1993 to 2006 to examine the following questions: (1) can regional house prices explain the difference of consumption growth in regions; (2) whether the difference between the impacts of house price on consumption of cohorts is a direct result of their difference in age (young and older) and residence stage (rural vs. urban). According to wealth effect, the impacts of house price on cohorts’ consumption should be different in region and generation. In our regression model, we extend Rational Expectations Life-Cycle/Permanent Income Model by Muellbauer and Lattimore (1995) and consider mortality rate as an uncertainty factor and regional and macro factors as control variables. In our empirical results, we find that the difference of consumption growth in regions can’t be explained by regional house prices. We also find that house price has positive effects on consumption, but there is no difference between age and residence stage cohorts. Moreover, the effects of house price on consumption using Shinyi house price index that represents pre-owned house price are more significant than Cathay house price index that represents new house property price. Our findings are consistent with Attanasio et al. (2009, 2011) that the relation between house price and consumption is the common hypothesis.

並列關鍵字

Consumption ； Income ； House price ； Financial liberalization

參考文獻

消費、借貸限制與金融條件：台灣的實證研究

Alessie, R., Devereux, M. P., and Weber, G. (1997) “Intertemporal Consumption, Durables and Liquidity Constraints: A Cohort Analysis,” European Economic Review, 41, 37-59.

Aron, J., Duca, J., Muellbauer, J., Murata, K. and Murphy, A. (2012) “Credit, Housing Collateral, and Consumption: Evidence from Japan, the U.K. and the U.S.,” Review of Income and Wealth, 58(3), 397-423.

Aron, J. and Muellbauer, J. (2013) “Wealth, Credit Conditions, and Consumption: Evidence from South Africa,” Review of Income and Wealth, 59, Special Issue, 161-196.

Bacchetta, P., and Gerlach, S. (1997) “Consumption and Credit Constraints: International Evidence,” Journal of Monetary Economics, 40, 207-238.