city, then, on the basis of the NBS statistical definition of urban, these manufacturing employees should be classified in the data as urban manufacturing workers.
Indeed, China’s census of November 1, 2000, reported that the nation had 41.96 million manufacturing workers in cities and 18.41 million in urban towns (zhen), for a total of 60.36 million urban manufacturing workers, constituting 68 percent of all the manufacturing workers in China. The census also counted 28.07 million rural (xiangcun) manufacturing workers, 32 percent of the enumerated manufacturing workers.35 As it is, the 2000 census may have underestimated the urban proportion of manufacturing employment because rural workers who had moved to towns or cities within the previous 6 months and who were working in manufacturing there would be counted for the census back in their villages and therefore might be called rural manufacturing workers or even rural agricultural workers.
Table 1 shows, however, that the annual statistics for yearend 2000 recorded urban manufacturing employment at only 39 million (fewer than the 2000 census counted in the cities alone), just 49 percent of the reported national total, while rural manufacturing employment, at 41 million, constituted more than half of China’s manufacturing employment. The category of manufacturing employment in urban units was reported at only 33 million that year and that of urban manufacturing staff and workers were reported at 32 million. China’s 2000 census used a comparatively careful definition of urban population and employment that has been refined by the NBS during the last two decades. The NBS official statistical definition of urban and rural populations and employment is arguably the best standard for other Chinese statistics. Interestingly, the breakdown of China’s annual manufacturing employment statistics into the rural classification and the various urban categories is inconsistent with China’s own statistical definitions of urban and rural.
The inconsistencies between census data and annually reported data on urban and rural manufacturing employment arise in part because annual data are “administrative.” The regular statistics are based on administrative geographical boundaries, rather than on statistical distinctions between rural and urban employment. In the administrative data, urban encompasses only cities and perhaps also the political county seat (called the “county town”) of each county in China, while rural denotes everywhere else, including all other towns that are officially established as urban.36 In the annual manufacturing employment statistics, however, the word urban appears to be a misnomer. The data appear to refer to manufacturing employment only in China’s cities and perhaps some of their immediate suburbs and the county towns. Apparently, almost all of China’s manufacturing employment in urban towns (zhen) and rural areas is lumped together as rural manufacturing employment.
In truth, the discrepancy between the annual administrative data and the census data is even greater than the preceding paragraph asserts. According to the NBS, some “urban” data on manufacturing employment include data from units that are not in fact located in any urban area.37 Specifically, if a state-owned factory is located in a remote rural area, its data might still be included in urban data on employment and wages in manufacturing. This categorization is a legacy of the planned-economy practice of reporting data by administrative subordination rather than geographic locality. In addition, according to the NBS, there is at least one case of a rural county in Guangdong province that was reclassified and established as a city, after which all the factories in the