This study examines the effect of the device that allows so-called "Scellier" tax benefits when buying housing for rent. We use counterfactual analysis and the fact that the provision applies only in certain areas to assess the impact of the scheme on land prices. Our estimation method uses difference in difference mobilizing data from the database BNDP between 2004 and 2010. We develop two strategies for the control and treatment groups. The first uses data on both sides of the boundary between affected and unaffected area to the device. The second is based on the decomposition of the control group in percentile in which the temporal evolution before the implementation of the device is most similar to quartiles of treatment. Results show that the device impact the price of building land. This impact may vary from 5 euros to 30 euros per square meter. However, we show that this impact is heterogeneous both structurally and locally.
