Vacancy durations in the office market

Abstract

The durations of other indicators have been researched extensively in real estate studies, primarily the time on market and the duration of residence in housing units. Despite their importance, empirical research on the duration of vacancies is relatively limited and focused mainly on the housing sector. This paper aims to fill this gap and analyze the determinants of vacancy durations in the office sector. The analysis is based on a dataset of individual office suites located in New York City, NY that became vacant between 2012 and 2015. Vacancy durations are a form of time-to-event data and as such can be examined using survival analysis. We present several parametric and non-parametric survival models. Four key characteristics -- unit size, asking rent, building height, and floor number -- are found significant across all model specifications. Specifically, vacancy durations are affected the most by unit size and asking rent. Survival probabilities are found to considerably vary over time, which appears to be driven by variations in employment growth.