結婚決策對首次購屋決策影響之內生性分析 —台灣地區男性受訪者之實證現象探討The Endogenous Effect of Marital Decision on First-time House Buying -An Empirical Study on Taiwan Male Households

首次購置住宅為家戶重大的消費與投資性決策。此一決策除了受所得財富因素與住宅市場經濟狀況影響之外，亦與生命歷程重大事件之決策具有高度相關性，其中以結婚決策對首次購置住宅之影響效果最為明顯。然而，多數研究在實證分析所採用之計量模型，僅將結婚視為外生變數，並未考量其內生性之影響效果，參數較估結果可能產生偏誤。本研究建立雙變量混合比例危險模型，藉以分析結婚決策對首次購置住宅之內生性影響，並以『華人家庭動態資料庫(PSFD)』之男性受訪樣本為對象，進行實證檢定。實證結果發現雙變量混合比例危險模型中不可觀察項之相關係數顯著為正，顯示結婚變數對於購置住宅之決策的確存在內生性影響效果。即結婚機率越高之男性受訪者，其潛在購置住宅之機率越高。若將結婚變數以外生變數之設定方式進行分析，則參數較估結果將產生偏誤。
關鍵詞：購屋、結婚決策、家庭動態資料庫、雙變量混合比例危險模型、內生性

The decision regarding first-time house buying is one of the most important investment and also consumption decisions for most households in Taiwan. Not only is this decision-making process affected by the individual’s socio-economic factors and housing market conditions, but it is also very much associated with life-course events, such as marital decisions. While previous studies show that the marital decision has a significant affect on the housing decision, they only assume it to be an exogenous factor in their empirical studies, with the result that the estimated results might be biased. In this study, we consider this effect to be endogenous and construct the so-called “bivariate mixed proportional hazard” (BMPH) model to examine the endogeneity of the marital decision on the first-time house buying. The empirical data was retrieved from the Panel Study of Family Dynamics (PSFD) in Taiwan. Our empirical results show that the coefficient of the correlation between unobserved factors in the BMPH model is both significant and positive. It provides empirical evidence in support of our hypothesis that the marital decision for male households might have an endogenous effect on the housing decision. If we were to treat the marital decision as an exogenous factor, the estimated results would be biased. 

Key words: house buying, marital decision, Panel Study of Family Dynamics (PSFD) data,
bivariate mixed proportional hazard (BMPH) model, endogeneity