Who benefits from new housing supply?

• Applicant: Dr. Andreas Mense
• Subject Area: Economic Policy, Applied Economics
• Term: from 2018 to 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 409193813

Private markets provide housing to the poor primarily through the process of filtering. According to this theory, a new home supplied to the market triggers a series of moves. First, a household moves into the new unit, leaving vacant an older unit. This in turn allows another household to move. In this way, a number of households can move up the housing quality ladder. Thus, new housing supply reduces demand for run-down, low-quality dwellings and thereby lifts pressure off the bottom of the rent distribution. This improves housing conditions for the poor. However, depending on market conditions, it may be that property owners find it beneficial to upgrade vacant moderate-quality units. In that case, rents at the lower end remain unaffected. It is thus an empirical question which parts of the rent distribution react to new housing supply. The project’s main goal is to investigate empirically how and when the supply of new housing units affects the lower tail of the rent distribution, using instrumental variable quantile regression and data from Germany. A unit’s net rent is a perfect indicator for its quality if "Quality" is understood as a combination of all dwelling characteristics (including its location), as is common in the literature. A major innovation of the project is the identification of exogenous housing supply shocks on the local level through unforeseen weather events. In preliminary work, I show that a particularly rainy July in a given location causes substantial decreases in local end-of-year housing completions, potentially because they prolong drying time of unfinished buildings. According to the data, most of these unfinished units are completed about ten to twelve months later. This implies that such weather shocks lead to sizable and economically meaningful changes in local new housing supply. As a second goal, the project seeks to analyze heterogeneity in the filtering process. To understand better the sources of potential heterogeneity, I plan to develop a simple filtering model. The model shall deal explicitly with the effects of moving costs and landlords’ propensity to upgrade moderate quality units. Moving costs and rehabilitation are important because the filtering process implies a series of moves, and because it may stop when many units are rehabilitated before they are put on the market. Subsequently, the predictions of the theoretical model shall be tested empirically by exploiting differences over space in household mobility and in the propensity to rehabilitate.The project builds on three main data sources. It uses building completions, conversions, and demolitions data on the level of individual buildings, provided by the Statistical Offices of the German Länder. Data on individual rental dwellings was collected from three large German online market places in the years 2011—2017. The German Weather Service provides weather data in raster format. All data sets have a monthly time resolution.

• DFG Programme: Research Grants