Internal Migration and Health in China

Despite our best intentions, and a previous announcement on this page, we have not yet been able to make the data set described here publicly available. However, work in cleaning and documenting the data is proceeding rapidly. Our new target date is the end of February 2013. When the data are publicly available, we will announce it on this page, together with instructions regarding how to download the data. [Last revised January 14, 2013]


In 2008, with funding from the National Science Foundation (SES 0551279), a group of researchers from UCLA and the Capital Medical University, Beijing, carried out a national probability sample survey of 3,000 Chinese adults age 18-64, which included an oversample of residents in high in-migration areas.  The fieldwork was done by Millward Brown ACSR, a leading Chinese survey research firm.  The focus of the survey was on the determinants and dynamics of internal migration and its consequences for well-being. To migrantworkersstudy these topics, we collected extensive biographical information, including educational, residential, work, and marital histories and child rosters, and, for migrants, the migration experience and the nature of contact with other family members. We also collected demographic and socioeconomic information on the respondent and all other family members, and for the respondent psychological measures, health behaviors and health histories, biometric measures, and information about the household living environment and socioeconomic circumstances.

Project team

  • Yao LU, Assistant Professor of Sociology, Columbia University
  • Yaqiang QI, Assistant Professor of Sociology, Renmin University, Beijing
  • Shige SONG, Assistant Professor of Sociology, Queens College, CUNY
  • Wei WANG, Dean and Professor, School of Public Health and Family Medicine, Capital Medical University, Beijing
  • Jing (Amy) ZENG, Project Director, Millward Brown ACSR, Beijing




The questionnaire consisted of several sections.  The CP (Cover Page) section includes information on the characteristics of the place where the interview took place and on the disposition of attempts to contact the household.  The S (Respondent Selection) section includes a household roster, information on the relationship of household members to each other, and information on which household member(s) were selected to be interviewed.  The main questionnaire includes sections BI (Basic Information) through GH (General Health) plus a Health Data Collection Form.  Finally, in the Chinese language version only there is a Feedback Form, used to provide information to respondents regarding their health status.  The English-language version of the questionnaire is shown as a single document.  The Chinese-language version is shown as four separate documents: the CPS, and main portions of the questionnaire and the Feedback Form.


In order to ensure that we would have a sufficient number of internal migrants to sustain analysis, but lacking any way of listing the population so as to distinguish migrants from non-migrants in advance, we resorted to over-sampling areas with a high proportion of in-migrants (defined for these purposes as those born elsewhere) and out-migrants (defined as the fraction of households with members who had been away for at least six months).  We used as a sampling frame the “township” file constructed by the Chinese National Bureau of Statistics in conjunction with the Chinese census of 2000.  Townships are the lowest administrative level that covers the entire land area of China; in 2000 there were about 50,000 townships in China. Using a design that stratified townships on the basis of the proportion of in-migrants (with some distinctions made by province for very high in-migration areas) and also, for townships with few in-migrants, the proportion of out-migrants; whether the township was an urban district (jiedao), rural area with a town (zhen), or rural area lacking a town (xiang); and the average level of education of the population, we created 75 strata, from each of which we selected two townships at random but with probability proportional to size.  Because townships are too large to permit enumeration of households (they average 25,000 people and often are much larger, especially those with large fractions of in-migrants), we then created secondary sampling units—enumeration districts (EDs)—within each township. In urban areas, EDs were defined by overlaying township maps with a 250-by-250 meter grid and then drawing ED boundaries to conform as closely as possible to the grid while following roadways and other natural divisions between buildings.  In rural areas, we defined administrative villages as EDs. We then randomly selected four EDs within each township; completely enumerated all dwelling units (defined as units with an entrance to a public space—that is, to a corridor or the exterior of a building); and randomly sampled dwelling units from the resulting lists.  A list of eligible dwelling units, shown in random order, with addresses and location information, was provided to each interviewer, who moved sequentially through the list, attempting to contact each household, listing all adults in the household, and randomly selecting one adult (or more, depending on the size of the household) to be interviewed. The interviewer continued, in the order specified, until five interviews were completed in each ED. This procedure yielded 3,000 interviews (75 strata × 2 townships × 4 EDs × 5 interviews per ED).  (Unfortunately, due to a communication error between the fieldwork agency and its subcontractors, the identification of EDs was lost, which means that the only clustering information we have is at the township and household levels.)  See Sample Design Documents for details on the sampling procedure.

Fieldwork for the survey was conducted from Oct. 2007 through April 2008, by Millward Brown ACSR, a leading international commercial survey research firm with the Chinese branch (ASCR) headquartered in Beijing.  The overall response rate (AAPOR’s RR1 [AAPOR 2008]) for the 2008 survey was 47% which, while low by U.S. standards, is typical of response rates in nongovernmental Chinese surveys.  Only surveys in which interviewers are accompanied by local officials or police elicit substantially higher response rates. Our response rate also is not out of line with those experienced in some European nations by high quality surveys such as the European Social Survey, a repeated cross-section similar to the US GSS.  In that survey, the response rate to the 2006 round was 51% for Denmark, 46% for France, 54% for Germany, 60% for the Netherlands, and 55% for the United Kingdom (ESS 2007). In the 2008 round, the corresponding response rates were 54%, 49%, 48%, 50%, and 50% (ESS 2009).

References cited in the discussion of the sample design

  • American Association for Public Opinion Research [AAPOR]. 2008. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 5th edition. Lenexa, KN: AAPOR.
  • European Social Survey [ESS]. 2007. ESS-4 2006 Documentation ReportEdition 3.2. London: Centre for Comparative Social Surveys, City University.
  • European Social Survey [ESS].  2009.  ESS-4 2008 Documentation Report.  Edition 2.0. Bergen: European Social Survey Data Archive, Norwegian Social Science Data Services.


Presentations and Working Papers


  • Zhang, Zhuoni, and Donald J. Treiman.  2013.  “Social Origins, Hukou Conversion, and the Wellbeing of Urban Residents in Contemporary China.”  Social Science Research 42(1):71-89.  [An earlier version is available as “Registration (Hukou) Status, Status Mobility, and Well-being in Contemporary China.”  Los Angeles: UCLA, California Center for Population Research, Population Working Paper PWP-CCPR-2011-009.]
  • Lu, Yao, Peifeng Hu, and Donald J. Treiman.  2012.  “Migration and Depressive Symptoms in Migrant-sending Areas: Findings from the Survey of Internal Migration and Health in China.”  International Journal of Public Health 57(4):691-698.
  • Treiman, Donald J.  2012.  “The ‘Difference Between Heaven and Earth’: Urban-rural Disparities in Well-being in China.”  Research in Social Stratification and Mobility: Special Issue on Inequality Across the Globe 30(1):33-47.  [An earlier, more comprehensive, version is available at Los Angeles: UCLA, California Center for Population Research, Population Working Paper  PWP-CCPR-2011-006.]
  • Treiman, Donald J., Lu Yao, and Yaqiang Qi.  2011.  “New Approaches to Demographic Data Collection.”  In Frontiers of Demography, edited by Zai Liang.  Beijing: People’s University Press, Beijing (in Chinese).  [Also, in English at Los Angeles: UCLA, California Center for Population Research, Population Working Paper  PWP-CCPR-2009-022.  A slightly shortened English-language version appeared in the Spring 2012 issue of the Chinese Sociological Review 44(3):56-.]
  • Treiman, Donald J.  2011.  “Migration and the Quality of Life: Lessons from China.”  NUSS Professorship Lecture, National University of Singapore, 20 January.  [PowerPoint presentation.]
  • Treiman, Donald J., Perry Peifeng Hu, Yao Lu, William M. Mason, Yaqiang Qi, and Shige Song.  2009.  “Determinants and Consequences of Peasant Labor Migration in Contemporary China.”  Paper presented at the Spring Meeting of  the International Sociological Association Research Committee on Social Stratification and Mobility (RC28), Beijing, 14-16 May.  [Also presented as Treiman, Donald J., and Lu Yao.  2009.  “The Effect of Internal Migration in China on Socioeconomic Outcomes and the Level of Living.”  Paper presented at the Annual Meeting of the Population Association of America, Detroit, 29 April – 2 May.]  [Powerpoint]
  • Treiman, Donald J.  2009.  “Types of Internal Migration in Contemporary China.”  Paper presented at the Annual Meeting of the Population Association of America, Detroit, 29 April – May 2. [Powerpoint]
  • Hu, Peifeng, and William M. Mason.  2009.  “The Relationships between Migration Status and Cardiovascular Disease Risk Factors in China.”  Poster presented at the Annual Meeting of the Population Association of America.  Detroit 29 April-2 May.
  • Lu, Yao.  2009. “Labor Migration and People Left Behind in China: Evidence from the 2008 China Migration and Health Survey.”  Poster presented at the Annual Meeting of the Population Association of America.  Detroit 29 April-2 May.
  • Qi, Yaqiang, Peifeng Hu, and William M. Mason.  2009.  “A Comparative Analysis of Self-Rated General Health in Three Developing Countries.”  Paper presented at the Annual Meeting of the Population Association of America.  Detroit 29 April-2 May.
  • Song, Shige, and Jingjun Qiu.  2009.  “Initiation of Cigarette Smoking among Chinese Adults: Evidence from the 2008 China Internal Migration and Health Survey.”  Poster presented at the Annual Meeting of the Population Association of America.  Detroit 29 April-2 May.
  • Treiman, Donald J.  2009.  “Doing Survey Research in Contemporary China.”  Seminar presented at the Division of Social Sciences and Humanities, Hong Kong University of Science and Technology, 18 March 2009.
  • Treiman, Donald J.  2009.  “Internal Migration in Contemporary China.”  Seminar presented at the Division of Social Sciences, Singapore Management University, 13 March 2009.  Also presented at the Division of Social Sciences and Humanities, Hong Kong University of Science and Technology, 19 March 2009; and the Center for Demography and Ecology and Center for Demography of Health and Aging, University of Wisconsin, Madison, 14 April 2009.  [PowerPoint presentation.]
  • Treiman, Donald J.  2008.  “Migration, Health, and Well-being in China: Preliminary Results from a New Survey.”  Paper presented at the Spring Meeting of the Research Committee on Social Stratification and Social Mobility (RC 28) of the International Sociological Association, Florence, Italy, May.