What Parcel Tax Records Tell Us about Homeownership Measurement in Surveys

This research aims to understand the measurement error in self-reported homeownership data collected by surveys. The analysis focuses on Detroit as a case study. We use legal ownership status in administrative records (the city of Detroit parcel tax records) as the benchmark to validate self-reported ownership status collected from a survey (the Detroit Metro Area Communities Study). We compare data from two question formats, which measure ownership at the household level and at the individual level, respectively. We also study the associations between sociodemographic characteristics and measurement errors in the self-reported ownership. The results suggest that 1) respondents do not always interpret the ownership questions as was intended, 2) the reported ownership status is sensitive to question formats, 3) the risk of measurement error appears to be heterogeneous in the population. The results challenge the assumption that homeownership is a standard fact, the reporting of which is not impacted by how it is measured. The findings are useful for understanding discrepancies across survey results and for advising how to craft homeownership questions in surveys.