Abstract
If remote work has reduced the demand for living in big cities, then why have their rents gone up so much? In this paper, we argue that one key to this question lies in understanding the heterogeneous effects of remote work on housing demand. Consistent with the emerging literature, we show that remote work in expensive and dense places causes less housing demand through out-migration, which helps to explain why big cities have seen declining population and relatively weaker rents and house prices. However, we show that is counterbalanced by remote work also causing a surge in household formation. Methodologically, this paper adds to the growing body of remote work literature by being the first to use actual, post-pandemic remote work rates at a granular level as an instrumental variable. At the individual level, we utilize occupation and industry fixed-effect IVs, and at local housing market level we utilize a shift-share approach. The causal effect of remote work on housing markets is consistent with OLS, suggesting modestly larger effects for both individual and housing market models.
Conclusion
In this paper we explore how remote work affected housing demand at both a micro level and an aggregate level. Methodologically, this paper adds to the growing body of remote work literature by being the first to use actual, post-pandemic remote work rates at a granular level as an instrumental variable. At the individual level, we utilize occupation and industry fixed effect IVs, and at the local housing market level we utilize a shift-share approach. The causal effect of remote work on housing markets is consistent with OLS.
We find that remote work has increased the demand for housing at both the intensive and extensive margins for households and individuals. Remote households spent more on rent and mortgages than otherwise comparable non-remote households. In addition, remote households were more likely to have moved into their home within the last 12 months, and individuals who worked remotely were more likely to head their own households. All individual level effects are larger when using IV approaches.
In general, exposure to remote work led to increases in housing demand as shown, for example, through gross monthly rental payments and home values. However, this effect was smaller for PUMAs with high population densities and expensive housing markets. Indeed, we find a negative effect of remote work exposure on population growth the most dense and expensive PUMAs, suggesting that working from home led to some level of out-migration from these areas. However, in these dense and expensive PUMAs, the positive effect on the household formation helped offset the population loss. In short, one reason places that lost population nevertheless saw robust housing markets was that they had stronger household formation.
This paper makes three main contributions to the literature. First, we build on the existing literature by being the first study to use 2021 ACS data to examine the effects of remote work on individual level outcomes. Second, we use actual post-pandemic remote work rates to construct instrumental variables for working from home. Lastly, we extend the prior research by focusing on household formation as an important extensive margin of household demand.
This paper also opens up potential avenues for subsequent research into the economic geography of remote work and the future of cities. For example, our research has shown that remote work has changed the way that people live within dense and expensive urban areas. Working from home has led to both an out-migration from city centers as well as increased headship rates in those same areas. It remains an open question to estimate the magnitude of the effects of these changes on local economies and governments.
