The evolution of durable goods demand during China’s transition. An empirical analysis of household survey data from 1989 to 2006

Durable goods ownership is commonly seen as a ‘defining gauge’ for the stage of development of a country. Its unprecedented economic growth and the rise of a strong and steadily growing class of consumers make China a formidable case study for the investigation of durable goods diffusion. Drawing on a household-panel with a survey period from 1989 to 2006, the empirical analysis of the driving forces behind the diffusion of durable goods shows that growth of disposable income was not equally important for all goods in their diffusion process. Rather it was the fall of individual preference thresholds (explained in part by falling durable prices) that proved to have a significant influence on the diffusion process of some goods. As it turned out, this tendency was significantly stronger in rural areas and could have counterbalanced, therefore, welfare patterns in terms of ownership contrary to the stable urban-rural gap in economic performance. Apart from changes in income and durable prices, it was found, that improvement of public services had particularly strong effects for urban poor and in rural areas. A forecast exercise up to 2030 revealed that growth in ownership rates is expected to be particularly strong for durable goods like refrigerators and cars for which households already show (or are about to do so in the case of cars) high sensitivity towards further increases in their disposable income. For other durables, like colour TVs, that are already well spread in the population there are signs of saturation with lower expected growth rates of ownership. Additionally, ownership rates are expected to pick up stronger in rural areas were households are less saturated and show higher income elasticities. As a comparison with figures from the literature demonstrates, actual and projected ownership rates depend, to some degree, also on the choice of the data set. The projections based on CHNS data could, therefore, build a reference to other commonly used data sets from the Chinese National Bureau of Statistics.