"Structural Density and Homeownership"

Abstract :

One of the most salient stylized facts about homeownership is that multifamily units—typically located in more central parts of cities—are much less likely to be owner-occupied than single-family units. This fact is consistent with landlord production efficiency advantages associated with structural density, arising from economies of scale and coordination frictions. To derive testable predictions and guide our empirical analysis, we develop a simple modified monocentric city model that incorporates these efficiency advantages and spatial variation in structural density due to centrality and soil conditions. To identify the causal effect of structural density, and to disentangle it from other determinants of homeownership, we employ a novel instrumental variable (IV) strategy derived from the engineering literature, exploiting differences in soil quality. Using data at the ward level in England and grid level in Germany, our IV-estimates show that a ten–percentage-point increase in the share of multifamily housing reduces the local homeownership rate by about eight to nine percentage points. The magnitude of this effect aligns with observed spatial differences in homeownership within England and Germany (in-sample) and across Europe (out-of-sample). In England, about half of the effect reflects sorting, whereas in Germany sorting appears negligible. We provide direct evidence for our proposed mechanism and confirm our findings using an alternative empirical strategy based on a large European household panel.« »