Resale Housing Price Determinants: A Case Study of Canada 2000 – 2009

Abstract

This thesis studies the determinants of selling prices of existing residential

properties in Canada from 2000 to 2009 using data from a nation-wide survey

conducted annually at the household level – the Survey of Household Spending

(SHS).I built the regression models based on survey of existing studies, and used the

SHS data for regressions.

I found that after controlling for the individual characteristics of households,

the time dummy and province dummy variables generate significant results in both

the first and the second models. Their coefficients are consistent with the trend

observed in data taken from real life events. The hedonic variables of the dwellings

also generated significant and reasonable results, especially the ones measuring the

population of the neighborhood, age of the dwelling, and also number of bathrooms

in the dwelling.

For the household related characteristics, I found that household income per

person exerts a positive effect on the selling prices of existing houses, keeping all

else fixed. The proportion of the household’s expenditure on food items increases

the selling prices, but the household income needs to be included for its effect to be

significant. The household’s regular property tax payments also exhibit positive

relationship with the selling prices.

For the variable of tenure status of a household, which is owning their

dwelling with or without mortgage in this paper, I obtained some curious results

that required further examination. Owning the dwelling with mortgage decreases its

selling price at an extremely significant rate. I designed a second model with an

extra set of demographic variables built in to test a hypothesis following analysis of

model one results. In the second model I tested out the hypothesis that the tenure

variable has absorbed other factors that decrease the selling prices, such as the age

cohort effect on the condition of the house. The results from the second model

supported the hypothesis, and also shed light on the impact of underlying

demographic structure on housing prices.