Survey Research in the Study of Contemporary China: Learning from Local Samples

A widely publicized 1983 Chinese survey found 43 per cent of all "specialized households" in a Shanxi county were households of cadres or former cadres. In what sense, if any, is this finding significant? More generally, what can be concluded about Chinese society, politics and the economy based on findings from survey research conducted there? This article sets out what can (and what cannot) be inferred from the unrepresentative samples of the Chinese population that are the basis for most survey research conducted in mainland China. Sampling is not the only serious obstacle to survey research in the People's Republic of China (PRC). Many other problems greatly challenge the ingenuity of social scientists in adapting standard methods to distinctly non-standard conditions.' The nature of the problem considered here is different, however. The most fundamental reality in sampling in the PRC is not the obstacles survey researchers strive to overcome, but a situation most accept as quite tolerable: for most researchers, obtaining a nation-wide probability sample of the Chinese population is both impossible and impractical.2 Interviews, however skilfully and cleverly designed, must practically always be conducted with Chinese in localities selected not according to any principle of random selection but chosen partly for convenience and always subject to the approval of the Commu-