Creating an HRS-Adapted Simplified SHARE Distribution File

点击翻译按钮获取中文摘要

OTHER PROJECT INFORMATION - Project Summary/Abstract SHARE, the Survey of Health, Ageing and Retirement in Europe, is modeled after the US Health and Retirement Study (HRS). SHARE is also tightly harmonized with the English Longitudinal Study on Aging (ELSA) and the Japanese Study of Aging and Retirement (JSTAR). As opposed to HRS, ELSA and JSTAR, SHARE is a multi-country study which currently involves 20 countries with 27 language versions. SHARE has collected three waves of data in 2004, 2006 and 2008 and is currently in its 4th wave. The third wave in 2008 deviated from the strict panel design and collected retrospective life histories. The multitude of countries with their institutional variety and languages, combined with a large interdisciplinary set of variables stored in about 25 modules per wave, and the differentiation into individual-, couple- and household-level data, make SHARE a very complex data set. This complexity appears to reduce the usage of SHARE in non-European countries, in particular the US, and limits its potential as a widespread tool for cross- national comparisons in order to better understand how health outcomes in older ages are related to the economic, healthcare and social policy environment. The aim of this project is to create a simplified and HRS-adapted dataset for researchers who are less experienced in the quantitative analyses of complex panel data and for student training and teaching. This dataset will include the same number of respondents and proxy reports as the main distribution of SHARE but will be restricted to a subset of variables. This will allow storing the database as a single file without the need for complex merges of database modules across the individual-, couple- and household-levels. In addition, we will add generated variables, such as indexes and recoded health, demographic, social and economic measures, which allow for direct analyses without extended data preparations. All variables will be labeled and documented using common descriptors adapted to the American language and the US institutional environment, with a synopsis targeted to comparisons with HRS. The simplified and HRS-adapted SHARE data set will be made available free of charge in the format of various statistical packages.