The purpose of this paper is to determine how much variance in vacancy duration can be explained by the architectural attributes of apartments and to illuminate strategies to reduce vacancy duration utilized by non-profit housing providers.
This is a sequential mixed methods research study with a qualitative variable-gathering phase followed by a quantitative variable-testing phase. Vacancy duration in days was the dependent variable and the attributes of the apartments were the independent variables. Each building functioned as a separate case, with its own results, and the cases were compared to draw conclusions about the strongest predictors for vacancy duration.
Each case study project has a significant linear regression equation with multiple variables contributing to the variance in tenancy duration. The R2 statistic varied for the case study projects from a low of 10.2 percent to a high of 36.9 percent. Factors that resulted in longer vacancies for two or more of the projects include: unit mix, floor level, road proximity and length of tenancy for the tenant moving out. Factors resulting in shorter vacancies include: corner position in the building and relatively larger size of the apartment.
The geography of the study is limited to Washington State in the USA. However, the case study projects represent three metropolitan statistical areas, with distinct climates and economic conditions. There are limitations to the stepwise analysis method because the degrees of freedom limit the complexity of models that can be estimated.
This paper highlights influences on vacancy duration and proposes conceptual models for measuring the periods of vacancy duration.
Through this study, architectural contributions to vacancy were uncovered and tested so that subsidized housing, a public good, can be distributed more efficiently.
This research is the first known study to compare vacancy durations on a unit-by-unit basis.
