Population Change in Wildfire-Affected Areas in the United States: Evidence from U.S. Postal Service Residential Address Data

Abstract

We examine the utility of data on active and vacant residential addresses to inform local and timely monitoring and assessments of how areas impacted by wildfires and extreme weather events more broadly lose (or not) and subsequently recover (or not) their populations. Provided by the U.S. Postal Service to the U.S. Department of Housing and Urban Development and other users, these data are an underutilized and potentially valuable tool to study population change in disaster-affected areas for at least three reasons. First, as they are aggregated to the ZIP + 4 level, they permit highly local portraits of residential and, indirectly, of population change. Second, they are tabulated on a quarterly basis starting in 2010 through the most recent quarter, thereby allowing for timely assessments than other data sources. Third, one mechanism of population change—namely, underlying changes in residential occupancies and vacancies—is explicit in the data. Our findings show that these data are sufficient for detecting signals of residential and, indirectly, of population change during and after particularly damaging wildfires; however, there is also noticeable variation across cases that requires further investigations into, for example, the guidance the U.S. Postal Services provides its postal offices and carriers to classify addresses as vacant.