Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Since the Great Recession, central banks and macroprudential authorities have been devoting much more attention to the housing market. To properly assess trends and risks, policymakers need detailed, timely, and granular information on demand, supply, and transactions. This information is hardly provided by traditional survey or administrative data. In this paper, we argue that data coming from housing sales advertisements (ads) websites can be used to overcome some existing deficiencies. Using a large data set of ads in Italy, we provide the first comprehensive analysis of the problems and potential of these data. We show how machine learning tools can correct a specific bias of online listings, namely the proliferation of duplicate ads that refer to the same housing unit, increasing the representativeness and reliability of these data. We then show how the timeliness, granularity, and online nature of these data make it possible to monitor in real-time housing demand, supply, and prices.