가계동향조사를 이용한 주택가격과 가계소비의 관계 분석

Abstract

이 논문은 Attanasio et al.(2009)과 Campbell and Cocco(2007)의 ‘주택가격 내생성 가설’과 ‘세대간 부의 재분배 가설’이 대립되고 있음을 배경으로, 2003-16년간 한국의 분기별 “가계동향조사”를사용하여 주택가격과 가계소비의 관계를 분석하였다. 분석결과 가계소비와 주택가격 사이에서나타나는 양(+)의 상관관계는 주택가격 내생성 가설과 부합한다고 해석할 수 있는 증거들이발견됐다. 연령과 주택보유 기준으로 가계를 구분하였을 때 소비의 주택가격에 대한 반응도는40 미만 청장년층 주택보유가계, 청장년층 무주택가계, 다음으로 55세 이상 노년층가계의 순서였다. 특이하게도 40~55세 중년층 가계의 소비는 주택가격과 작지만 유의한 음(-)의 관계를 보였다. 본 논문은 중년가구 소비의 특이성에 대하여 한국 특유의 ‘자녀부양비용 역할’ 가설을 제안하였다.

The empirical fact that house prices tend to positively commove with household consumption has been attracting increasingly larger attention both from academia and policy commentators. A long-standing question came to existence, as a result, whether housing wealth affects consumption. The current state of literature is yet to reach a consensus. To resolve the question, it is essential to explore household-level microeconomic data, which is rare. Because of the data availability problem, there is only a small literature drawing on the UK Family Expenditure Survey. Even the small literature yields conflicting results. Campbell and Cocco (2007) present that older homeowners raise their consumption in response to a rise in house prices, while younger homeowners and renters decrease, in line with the wealth effect hypothesis. In contrast, Attanasio and Weber (1994), and Attanasio et al. (2009) report that the positive relationship of house price growth and consumption is larger for young households, but insignificant for old, which contradicts the redistributive wealth effect hypothesis. They put forward the endogeneity hypothesis, namely the apparent positive association between house prices and aggregate consumption in the UK data merely captures a common macroeconomic factor such as future income prospects of households that affects both house prices and consumption. At the backdrop of the debate on what underlies the positive relationship between house prices and consumption, this paper examines the Korean Family Expenditure Survey over the period of from 2003Q1 to 2016Q4. Applying the pseudo-panel methodology developed by Deaton (1985), which is now a standard empirical tool in the literature, I investigate plausibility of the two competing views for the Korean data. Following Attanasio et al. (2009) and Campbell and Cocco (2007), I construct two sets of cohort samples: the birth-year cohort sample and the homeownership cohort sample. The regression model is formulated by adopting the standard framework of the life cycle model of consumption. Main findings are as follows. In the first place, I confirm that the positive relationship indeed exists in the Korean data as well. When controlled by homeownership, I do not find any empirical evidences that effects of house prices on consumption is stronger for homeowners. Regression analyses show that responses of household consumption to house prices differ by age rather than homeownership. Specifically, young households tend to increase their consumption responding to a rise in house prices while old households do not. Overall empirical findings of this study suggest that the existing explanations developed in the context of the UK or the US economy lacks relevancy for the Korean data.