The difference in housing prices between non-arm’s length and arm’s length transactions using propensity score matching

This study uses a social capital approach to examine the differences in housing prices between non-arm’s length and arm’s length transactions. It employed propensity score matching to generate comparable matched pairs and eliminate selection bias, thus facilitating the discernment of housing price differences between non-arm’s length and arm’s length transactions. It used the actual housing prices registered from January 1, 2012, to December 31, 2019, in Taipei City and New Taipei City. The data were divided into two groups: a treatment group of non-arm’s length transactions and a control group of arm’s length transactions. The empirical results indicated that transactions involving first-, second-, and third-degree relatives were priced 53.3%, 51.4%, and 47.2% lower than arm’s length transactions, respectively. Urgent buying/selling transactions were priced 21.7% lower than arm’s length transactions. Transactions influenced by obligations, liabilities, or debt settlements were priced 12.0% lower than arm’s length transactions. Purchases from government agencies were priced 45.1% lower than arm’s length transactions. Transactions between friends were priced 26.3% lower than arm’s length transactions. These empirical results indicate the importance of non-arm’s length transactions on housing prices, with significant rebates due to implicit social capital.