Co-authored with Jens von Bergmann and cross-posted at MountainMath

Who would benefit from dramatically expanding the supply of housing in the market? How would such an expansion address housing need? What non-market interventions would still be required to meet remaining need? How would it influence household formation and migration? This post is picking up where we left off on our housing targets post (von Bergmann and Lauster 2023a), grappling with the distributional effects of adding new housing to the market. To recap: one way of setting housing targets is to figure out how much housing municipalities would have if they weren’t actively restricting its construction. We can do that by estimating how much additional housing would be profitable to build. Setting that amount of housing as a target works to enable market mechanisms to bring housing supply in line with demand.

Our version of housing targets sets a (high) overall target number for net new housing, and that’s it. There are other versions of targets that attempt to more finely tailor numbers, keyed for instance to specific income bands. Here we explain why we think starting with an overall number is the right approach, insofar as adding housing has distributional effects across the entire system. It is generally understood that increasing the housing supply lowers prices of existing housing, all else being equal. What’s less clear are the systemic effects of adding housing and how even expensive new housing helps address underlying housing need. These broad distributional effects of reaching housing targets are what we want to address here.

TL;DR

Housing is good, and adding a lot more of it in the places people want to live would have broadly beneficial effects. Even if new housing tends to house (slightly) higher income households, the vacancy chain moves these households set in motion have systemic effects, freeing up housing all across the income spectrum. To illustrate, our housing targets are set to reduce the price of new housing by roughly 30% in Metro Vancouver. We estimate this would reduce rents by an equivalent amount, and holding households constant we can further estimate this would reduce the population experiencing Core Housing Need by around 35%. The annual monetary subsidy necessary to make housing affordable for all Metro Vancouver households would drop from $1.48bn to $0.68bn. We further refine this estimate by relaxing the assumption that households would remain constant, considering the effects of meeting housing targets on new household formation and net migration. Our estimates of Core Housing Need still decline, but not by as much. We also consider the effects of rent control on who benefits from a boost to housing supply. Finally we consider the implications of these distributional effects. Finely tailored housing targets can be misleading and counterproductive when they fail to account for distributional effects of adding supply overall.

Continue reading →