Subdividing existing housing units into smaller self-contained rental units has become an increasingly important approach to expanding affordable housing supply. Yet while subdivision increases the supply of low-cost housing, it may also generate congestion, safety, and other residential externalities. This paper studies this affordability–externality tradeoff using the expansion of a large PropTech platform in Beijing that converts apartment living rooms into additional bedrooms and rents them separately. Using longitudinal rental contract, census, and administrative complaint data in an instrumental-variable design, we find that subdivision lowers both bedroom rents and standard apartment rents, but increases complaints related to noise, fire hazards, and other housing-related nuisances. To evaluate the welfare and distributional consequences of subdivision, we develop and estimate a structural housing market model with heterogeneous households and endogenous remodeling decisions. We find that migrants without Beijing hukou, non-college households, and smaller households benefit most from the rent reductions generated by subdivision, whereas higher-educated local households and families with children bear larger externality costs. Although a complete ban on subdivision increases aggregate welfare, it disproportionately harms the most disadvantaged renters. Counterfactual analysis shows that targeted restrictions or Pigouvian taxes with resident rebates can substantially reduce externalities while preserving much of the affordability benefit.

We study a large-scale policy relaxing residential land supply constraints near major Dutch cities in the mid-1990s. Land allocation was determined centrally and was unrelated to pre-policy local market dynamics. Using administrative data, we explore effects on housing, mobility, and labor markets. New supply attracted high-income individuals, primarily from nearby high-income neighborhoods, where

house prices decreased. New supply also increased local employment in surrounding areas, especially in high-earning occupations. We develop a rich quantitative spatial equilibrium model and show that the impact of relaxing supply constraints on housing markets and welfare crucially depends on job reallocation and on in-migration flows.

We show that post-GFC bank regulatory changes intended to curb bank risk-taking contributed to the decline in the supply of new housing. With unique data on the financing and sales of nearly 18,000 homebuilders, we document a sharp decline in construction loans from large banks subject to stress testing following the 2013 reforms, a reduction that was not offset by increases in financing from other sources. Employing a difference-in-differences (DiD) design, we find that, relative to homebuilders that previously borrowed from smaller banks, the credit availability and new home sales of homebuilders that previously borrowed from stress-tested banks declined after the reforms. In a county-level instrumental-variable analysis, we find that counties where stress-tested banks held greater pre-GFC market shares originated fewer construction loans, supplied fewer new homes, and experienced greater house price appreciation after 2013.