Secondary housing supply

I estimate the impact of new housing supply on the local rent distribution, exploiting delays in housing completions caused by weather shocks during the construction phase. Increasing the flow of new supply by 1% lowers average rents by 0.2%, and increases disproportionately the number of second-hand units offered for rent. The supply shock affects the entire rent distribution. Employing a quantitative model, I explain this pattern by secondary supply: New supply triggers a cascade of moves that frees up units in all segments of the local market. The impact on rents is similarly strong in locations experiencing growing housing demand.