Land Use and Housing Production in California

Abstract

My dissertation studies the economics of housing production using data from California to understand how cities grow, change, and increase in density over time. To do this, I examine cities by thinking about buildings as durable goods and look to data to examine the implications of durability and how it interacts with housing policies. The durable goods approach models buildings as having a large cost of construction and then generate a steady profit until they deteriorate to the point that they are replaced. Replacement occurs due to low quality of current building or improvements to the value of alternative buildings. From this approach, a number of predictions follow: Policies that increase the benefits of existing buildings will decrease the chances they are replaced with new building; policies that reduce future profits from buildings can reduce the value of new buildings; and the destruction of existing buildings will lead to more different buildings being built than there would be otherwise. I test each of these implications in the context of California’s current housing crisis and show if and how these implications are reflected in the real world. In chapter 1, I leverage differential exposure to the Proposition 13 tax laws to understand the impact of this policy on the production of housing in Southern California. Proposition 13 restricts property tax growth as long as the owner doesn’t sell or redevelop the property, raising the benefits of not redeveloping, which allows me to exploit differences in market conditions at the time of prior purchase to identify the effect of these property tax limits on property redevelopment. I find that Proposition 13 discourages redevelopment and sales. In a dynamic discrete choice model of land use, I find that adopting a land value tax that replaces Proposition 13 would increase housing production by 15-32% generating a similar or greater amount of new housing as other policies under consideration in California. In chapter 2, I estimate the cost of a recent rent control policy on the value of new housing. Rent control is a classic example of a price control. I evaluate the effect of AB 1482, a recently introduced rent control law in California. Using sales prices of multiple family housing, I find that the value of new controlled rental housing did not decline after the passage of rent control using a differences in differences approach. In this case, the rent control did not substantially effect the landlord’s value of rental housing because the law is generally non-binding. AB 1482 is an example of a non or rarely binding rent control policy that is unlikely to substantially impact landlords incentives to provide rental housing to the market but could still provide insurance value to renters. In chapter 3, I use variation in wildfire location and timing to estimate the impact of durability on land use. I build a dynamic model of land use that shows how durability, land use laws, and natural disasters interact together and that wildfires will increase the rate of land use changes. I test this prediction on the universe of properties in Los Angeles county over the past twelve years. My results show that land owners do not change between broad categories (housing to commercial, for example) but that the characteristics of single-family homes change faster after a wildfire. The three chapters show that at all three implications due appear to matter but are often moderated by context and policy details. Using a durable goods model of housing allows for an improved understanding of housing markets.