Revisiting international house price convergence using house price level data

The literature on house price convergence largely relies on house price indices, based in an arbitrary year, rather than on actual price levels. This is essentially due to the scarcity of comparable house price level data. However, this severely constrains the analysis. In particular, it often forces to discard a large part of the data sample, which leads to a great loss of information and hampers analysis on sub-periods. In this paper, we combine the price level estimates produced by Bricongne et al. (2019) with long OECD house price series to analyse long-run house price convergence across 18 OECD countries. Applying the Phillips and Sul (2007) procedure, we find convergence in OECD nominal house prices, but two separate convergence clubs in real house prices. We find some convergence in euro area real house prices, but no evidence of strengthening after the introduction of the euro. β-convergence regressions (linking changes in house prices to their initial levels) point to conditional convergence, related to the evolution of gaps in real GDP per capita, long-term interest rates and population growth across countries.