The housing market is difficult to predict due to the complexities of macroeconomic conditions, government policy interventions, and ever-changing supply and demand. Furthermore, the structure of a household's housing consumption decision-making can cause structural changes over time. When a market undergoes structural change, government policies based on model coefficients estimated from historical data can be misleading. In particular, permanent income is a major variable in the analysis of housing demand. However, if using current income causes an estimation bias for other variables, the model's stability cannot be guaranteed. This study explores the stability of consumer preference using the primary variables of the housing tenure choice function, such as income, housing cost, housing price volatility, and unit size. The analysis confirms that the household preference system for income, housing costs, and housing price volatility conditions was generally stable in housing consumption from 2002 to 2019. However, there was a significant change in the 2020 housing tenure choice model. The change was statistically significant. Furthermore, the estimation bias model analysis shows that using the current income variable as a proxy for the permanent income variable did not result in any significant estimation bias for other independent variables.