Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
When connected through economic, financial, and policy shocks, cities can be considered to be neighbors in the social-economic sense. Therefore, a spatial correlation may exist in cities that are not geographically contiguous. The paper simultaneously studies the spatial correlation within each city and the spatial interactive effects among different cities on housing prices. We build a spatial autoregressive hedonic pricing model in a system of interrelated networks, which allows for multiple spatial interactive effects among housing units inside the same city and from other cities. For estimation, we propose a two-stage least squares (2SLS) method and the maximum likelihood estimation (MLE). Finite sample properties of these two methods are investigated in a Monte Carlo simulation. After applying these steps to March 2016 housing prices in Shanghai and Beijing, the empirical findings show that the spatial correlations within each city are significantly large, i.e., approximately 0.6–0.8, and spatial interactive effects between the two cities exist, although the magnitude is smaller, i.e., approximately 0.1–0.2.