The name is absent

Brian Nolan, Ive Marx and Wiemer Salverda

Cross section analysis

a. Cross-section analysis with significant coverage of OECD countries going back to the 1980s can be un-
dertaken with micro-data from LIS for certain years at 5-year intervals, both for income inequality and
earnings dispersion, though with only a limited number of non-income variables available for analysis.

b. From 1995 to 2001 and for 14 “old” EU member states only, data from ECHP has a much wider set of
non-income variables available in the dataset, with a high degree of harmonization of definitions and
variables.

c. For the mid-2000s, data for a much wider set of EU countries is available from EU-SILC, which also
has a very substantial set of non-income variables; once again the “output-harmonised” nature of EU-
SILC has to be kept in mind.

d. Data from the European Social Survey (ESS) have the advantage of ready availability and broad topic
and country coverage, but the income measurement in this survey is very rudimentary and cannot be re-
lied on; if the other variables being employed in the analysis are available only in this source, then some
validation of the income variables by reference to other sources would be important.

In concluding, it is essential that the different contributions to the GINI project are explicit about the type of data
they are employing, the principal properties of these data, and the practicalities of access etc. The representativeness
of datasets employed—even if only for descriptive purposes — is key and it may be helpful to refer to external aggre-
gates — for example from national accounts — in assessing what is well-covered and what appears to be under-stated.
Any deviation from what is being suggested as ‘default’ options for GINI purposes should also be noted.

In each case, the possible effects on comparisons over time and cross-sectional comparisons should be drawn out, if
only speculatively. Participants are encouraged to extract (GINI-default) data from national and other sources if they
can and make these available, and also to share their experiences with datasets via the GINI website data portal.

Page • 36

<<   33   34   35   36   37   38   39   >>

More intriguing information