Study Introduction: China Health and Retirement Longitudinal Study (CHARLS)
Written by: Hunter Green and Drystan Phillips
Published on: Jul 29, 2021
The China Health and Retirement Longitudinal Study (CHARLS) is a nationally and urban-rural representative longitudinal survey of people aged 45 years and over and their partners, regardless of age, living in China. The baseline survey of CHARLS was fielded in 2011 and 2012 and follow-up surveys have been conducted in 2013, 2015, and 2018. To facilitate cross-country comparisons, CHARLS was designed to be comparable to the U.S. Health and Retirement Study (HRS) and its expanding global Family of Surveys.
The initial CHARLS sample was recruited in 2011 and 2012 using a stratified, multistage sampling of households at all county-level units of China except for Tibet. At baseline, if a sampled household had multiple people older than 40, one of them was randomly selected. If the selected person was aged 45 or older, they became a respondent. If the selected person was between age 40 and 44, they were reserved for a refreshment sample and added to the CHARLS once they reached age 45. To date, CHARLS has introduced refreshment samples at Waves 2, 3, and 4 (Figure 1).
CHARLS elicits a wide variety of information from respondents, including demographics, family structure, health, cognition, health care utilization, job history, retirement, pension, and household and individual income and assets. In addition to the core interviews, CHARLS collected non-blood biomarkers, such as anthropometric measurements, blood pressure, grip strength, and lung function measurements, at Waves 1 through 3, as well as venous blood samples at Waves 1 and 3. CHARLS has also conducted exit surveys on deceased study participants starting at Wave 2, a Life History interview between Waves 2 and 3, and an in-depth Harmonized Cognitive Assessment at Wave 4.
Due to the wealth of information collected by the study, researchers who are new to CHARLS may find it difficult to navigate. To assist researchers with this process and encourage cross-national analyses, the Gateway has produced a harmonized dataset for CHARLS. Harmonized datasets provide Gateway users with research-ready variables that are intuitively named and comparably defined across waves and between studies. Figure 2 displays one of the many benefits of using harmonized datasets like the Harmonized CHARLS. Instead of dealing with 72 different files to look at respondents who have participated in all four waves and the Life History Survey, you can use one file which has many of the most important pieces of information already combined.
It is important to note that the Harmonized CHARLS dataset does not include all of the survey questions that are asked by CHARLS. Instead, harmonized datasets focus on survey measures which are most often used or requested by researchers and which are most comparable between waves of CHARLS and between CHARLS and its sister studies.
The Harmonized CHARLS data and documentation can be downloaded from the CHARLS website by any user who has registered with CHARLS. CHARLS registration is a straightforward process, users can find step-by-step instructions for registration using our CHARLS Data Access Instructions. In the coming weeks, we will be releasing an updated version of the Harmonized CHARLS dataset that adds variables for CHARLS Wave 4 and includes new sections covering physical measures, assistance and caregiving, stress, and psychosocial topics. We also recently hosted a webinar where we discussed the study design of CHARLS and the data structure of the newly released harmonized data. We also provided an example of a cross-section analysis using the Harmonized CHARLS and a cross-national analysis comparing CHARLS data to data from its sister study from South Korea, the Korean Longitudinal Study of Aging (KLoSA). A recording of the session can be found here and the presentation slides can be found here.
- Hunter Green is a Programmer at the University of Southern California.
- Drystan Phillips is a Project Manager at the University of Southern California.