Housing Supply and Regulation in 35 Chinese Cities

Over the past two decades China’s housing market has seen a remarkable transition, in part due to major institutional changes and in part due to urbanization, economic growth, and demographic changes (Long and Wang 2009; Renaud 2009; Zang, Man and Ren 2009). The annual total investment in the real estate industry is about 20 percent of the stock of fixed property investment, and 10 percent of national GDP. According to Shanghai Bureau of Statistics, the real estate industry contributed 39 percent of Shanghai’s GDP growth in 2009. Perhaps most remarkably, within the space of only two decades China’s floor space per capita more than doubled (Chow and Niu, 2011, p. 48). Analysis of Chinese housing prices can shed light on some of these fundamental processes, as well as short-run concerns such as the existence, or not, of a housing “bubble” in major cities, and whether such a bubble is bursting as we write (Gough (2015) and Jim (2015), to give just two examples from the English press).

To be clear, we are not, in this paper, pronouncing on the existence of a Chinese housing bubble, or forecasting the future time path of China’s housing prices. These exercises are difficult enough with large amounts of high quality data, and credible tests of bubbles and price forecasts are simply not possible, in our view, without a longer time span of high quality housing prices. Instead we focus on the related issue of housing supply, in particular the price elasticity of supply of housing, and its determinants. Papers by Wheaton (1999) and Malpezzi and Wachter (2005) have presented the argument that inelastic markets are necessary, though not sufficient, conditions for “bubbles.” It’s the necessary condition that we investigate in this paper. Further, international research has shown that demand patterns, e.g. how housing demand responds to changes in incomes and prices, are surprisingly regular and predictable; but supply responsiveness varies considerably from place to place. Other previous research suggests this supply response would depend partly on natural constraint (physical geography), and partly on the regulatory regime for land use and real estate development. Investigating these relationships is the main contribution of our paper.

Our focus is China, of course. Using a panel data set of 35 major cities in China from the period 2000 to 2011, we examine several simple alternative housing price models that can be linked to a simple supply-demand framework similar to that used in Follain (1979), Malpezzi and Mayo (1997) and Malpezzi (1999). Until very recently most studies of Chinese housing prices to date use aggregate data, e.g. Chow and Niu (2011). However there is now an emerging literature using panel data, developed independently but in the same spirit as our work; see Wu, Gyourko and Deng (2011), and especially excellent papers by Wang, Chan and Xu (2012) and Wang, Yang and Lin (2011), which take an approach related to that we take in this paper.

We also draw on analyses of individual household decisions in China’s housing market, e.g. Fu, Tse and Zhou (2000), and Zax (1997). Such analyses of microdata complement the approach we take in this paper; in particular such studies can be the source of identifying restrictions, as shown in Malpezzi and Mayo (1997).