With the locational information provided in the ads, we also gener-ated spatial variables that are well established in the literature to be sig-nificant and to have a large impact on housing values. The distance of housing to the city center is typically one of the largest contributors to a house's value and would be especially important for a physically mono-centric city like Beijing. However, we were not sure if we should measure to the physical center because the Forbidden City and the mon-umental public spaces that surround it are not large sources of employ-ment. So, we also tested whether the financial downtown, the center of Chaoyang district, might actually be the more economically potent cen-ter. However, in our preliminary regression of the first several months of ads we found that indeed the Forbidden City center was significant and Chaoyang was not. While we also entertained other polycentric nodes (Qin & Han, 2013), we found them insignificant in this under-ground market. This is probably due to the major ring roads and subway lines that encircle the Forbidden City and its corresponding land values (Zheng & Kahn, 2008).