Development of a panel dataset from the IAB Establishment Panel cross-sectional data

"The IAB Establishment Panel now consists of 12 waves from 1993-2004, which are available for research use in the Research Data Centre of the Federal Employment Agency at the Institute for Employment Research. Due to repeated surveys of the same establishments, individual processes within the establishments can be studied in both cross-sectional and longitudinal terms. Cross-sectional evaluations of various waves and the production of a longitudinal or panel dataset require different tasks to be solved. Researchers must check across the individual survey years whether the variable values have remained identical or whether values are absent or new ones have been added. In addition, the measurement levels of the variables must be checked and any differences must be harmonised. Further, the time period in question can alter (e.g. financial year or first half-year). The purpose of such checks is to guarantee the comparability of the variables beyond the analysis period. The above examples illustrate that producing a panel dataset involves time-consuming processing of the individual datasets. This report has been written as a tool to support users in producing a panel dataset on the basis of data from the IAB Establishment Panel. The Datenreport consists of the development of a panel dataset containing variables that tend to be frequently used, in our experience. The panel dataset is available in the Research Data Centre and can be used both via remote data access and during research visits. The STATA and SPSS program code used to produce the dataset is available as a download, enabling users to integrate additional variables into the program code used with relative ease. The program codes can also be modified with the aid of the test data. This Datenreport is structured as follows: following a brief description of the IAB Establishment Panel, the structure of the panel dataset is presented. Next, Chapter 4 presents the variables for the dataset in the form of tables and details the use of the program codes. Appendix 1 describes each of the variables in the panel dataset, also detailing the values and any restrictions or problems occurring. Appendix 2 presents frequency counts1 of the variables used. The program codes in SPSS and STATA are contained in a separate document." (Author's abstract, IAB-Doku) ((en))