Abstract: Along with the changes of urban residents' social space during China's institutional transition period, a wide concern has been aroused on the even more complicated spatial organization of residence and employment. Based on residential survey data of Beijing in 2010, several spatial analysis methods were applied to analyze characteristics of spatial distribution of residence and employment of different social groups in Beijing, such as the buffer analysis, the nearest neighbor clustering and Ripley K functions. The results indicated that spatial differentiation of residence and employment in Beijing was significant. Furthermore, relative differences on proportions of spatial distribution of residence and employment of different social groups also existed in the interior space of residence or employment. Degree of spatial mismatch showed that the circle of about a 15-km distance from Tian'anmen was the boundary of function dominant areas of employment and residence. Meanwhile, degree of jobs-housing balance presented a decreasing trend from inner city to outer layers as a whole, while that of high income groups and young white-collar workers was lower than the other two groups. Besides, spatial agglomeration of residence and employment for different groups also presented different characteristics. For common citizens, residential and employment locations were scattered in distribution. For young migrant workers, cluster locations of their living and working spaces were quite adjacent to each other, which were mainly concentrated in the eastern region of the city. Residential locations of high income groups were mostly separated, but their employment locations were relatively centered in the eastern and northwestern parts of the city. At the same time, distribution of young white-collars' living spaces displayed a band-like clustering in the west but scattered in the east, while their working spaces were mainly agglomerated in two axis regions centered on Zhongguancun and CBD respectively. Spatial agglomeration characteristics of residential and employment locations for different groups were similar in case of scales, revealing a trend of "increase at first but decrease later" with the increase of their distance from clustering center, although their distance thresholds were different. This difference suggested that high income groups had limited optional range on their residential locations and common citizens had minimum optional range on their employment locations, but the young white-collars showed maximum optional range on both residential and employment locations.

Key words: living space, jobs-housing relationship, spatial analysis, employment space, Beijing, spatial cluster