The excel spreadsheet pengrid_var_correspondence.xls' (which is also available as part of the ELSA Wave 2 documentation) shows how the variables in the pension-level dataset relate to those contained in the main individual-level dataset. In most cases several variables with the same name stem but different numerical suffixes will have been combined into one single variable. For example, in the individual-level dataset there are three variables with the stem wpsps: wpsps, wpsps2 and wpsps3. These relate to pension types 1, 2 and 3 respectively (see table 2 below for a description of each pension type). Each of these variables was not applicable (i.e. took the value -1) for all other pension types (e.g. wpsps2 was "not applicable" unless pentype_wave2=2). Therefore, these three variables have been combined into the single variable called wpsps in the pension-level dataset. Table 1a shows an example of how some of the data is structured in the individual-level dataset (elsa_wave2_archive_v2.dta), while table 1b shows how the same data is structured in the pension-level dataset.

Full version (external link)