A counterpart survey of southeast China was initiated in 2003, under collaboration with the Institute of Population and Labor Economics, Chinese Academy of Social Science (IPLE-CASS). The surveyed regions included Shanghai, Zhejiang, and Fujian. For ease of cross-regional comparison, the targeted population of the China survey covered adults born between 1935 and 1976 (aged 25-68 in 2003), with the range of birth years similar to that of the first three groups of the main sample of the Taiwan survey. In addition, the contents of the first-wave questionnaire of the China survey were similar to those of the Taiwan survey. The targeted sample size was 5,000, with the targeted complete interviews being 1,000, 2,000, and 2,000 for Shanghai, Zhejiang, and Fujian respectively. The stratified multiple-stage PPS sampling method was applied in each of these three regions. The primary selection unit (PSU) was counties and districts. The counties and districts of Shanghai, Zhejiang, and Fujian were stratified by the proportion of nonfarm population, based on statistics from the 2000 Population Census of China. In Fujian and Zhejiang, 22 counties and districts were drawn; while in Shanghai, 11 were drawn using the PPS principle. For each selected county or district, three townships (subdistricts) were drawn. In each selected township (subdistrict), three communities were drawn. For each community, 12 households were randomly selected. When the interviewers contacted the selected households, a pre-designed Kish table was used to choose the respondent if more than one family member met the birth year criterion. The targeted sample size for each community was ten cases. If the number of complete questionnaires was less than the targeted sample size, the failed cases were not replaced by others.
Follow-up interviews were conducted for the respondents who finished the first-wave questionnaire. The questions contained in the questionnaires of the follow-up survey mostly originated from the contemporaneous survey of Taiwan. From the first-wave to the latest China surveys, face-to-face paper-and-pencil interviews (PAPI) were used to collect the data.
The fieldwork of the China PSFD Survey was jointly implemented by the IPLE-CASS and local government units or private survey companies in Shanghai, Zhejiang, and Fujian. The local units recruited interviewers, while interviewer training and fieldwork management were jointly executed by tIPLE-CASS and local units. The IPLE-CASS was in charge of sampling design and sample maintenance, as well as key-in and management of data. Data verification and cleaning were jointly conducted by IPLE-CASS and the PSFD team.