Fiscal Redistribution Dataset, Version 2, February, 2008

The second release of our fiscal redistribution data set reflects changes in the LIS since our data were first made available in August, 2005.  Specifically, we have added figures from nine newly released income surveys: Australia, 1995 (which replaced Australia, 1994); Australia, 2001; Australia, 2003; Denmark, 1995 (which had been withdrawn but has since been re-released); Denmark, 2000; Denmark, 2004; Finland, 2004; Switzerland, 2002; and USA, 2004.  We have also added data from Ireland, 1987, the only Irish survey that includes information on pre-tax income. 

Our data have been adjusted to reflect revisions made to existing Wave V data sets in LIS’s release 2, which became available in 2007.  These include Canada, 1998; Canada, 2000; Finland, 2000; Germany, 2000; Netherlands, 1999; Norway, 2000; Sweden, 2000; the United Kingdom, 1999; and the United States, 2000.  In these revisions, the LIS sought better to reflect the boundary between public and private pensions, and to incorporate recent changes in child and family programs.  More details are available at http://www.lisproject.org/techdoc/bigsplash.htm

Finally, we re-calculated all of the other figures in the data set in late 2007 and early 2008.  The reason for this was to incorporate any revisions made by the LIS since we ran our original jobs in 2005.  Also, we have corrected several errors in our previously reported figures for post-government Gini indexes for 25-59 year old household heads.  

There are two technical issues that should be briefly mentioned.  First, in keeping with current LIS practices, we no longer include the Stata command “replace hweight=1 if hweight==0.”  This makes no difference in the vast majority of results, and only a tiny difference in the rest.

Second, on the advice of Teresa Munzi of the LIS staff, the earlier version of our data set reclassified certain pensions in Finland, the Netherlands and Switzerland from private to public.  These included LIS variable v32 in Waves I-III and variable v32sl in Wave IV.  (Wave V changes are reflected in release 2.)  Details are offered in footnote 10 of the Socio-Economic Review article.  We have continued to make these adjustments, but have also reported non-adjusted figures, should any user wish to compare them to the adjusted results.  The surveys affected are FI87, FI91, FI95, NL83, NL87, NLl91, NL94, CH82 and CH92.

Dataset Description 

This data set offers a number of measures of fiscal redistribution in the developed countries, drawing upon data from 60 Luxembourg Income Study surveys conducted in 13 countries between 1979 and 2002.  Specifically, we have computed:

1)  A measure of overall fiscal redistribution, as reflected in the difference between the Gini indexes of pre-government private sector income and post-government disposable income.  We offer measures of both absolute fiscal redistribution (Gini _pre – Gini _post) and relative fiscal redistribution ((Gini _pre – Gini _post)/ Gini _pre).  [Table A.1 in Excel Spreadsheet]

2)  The shares of absolute and relative fiscal redistribution resulting from direct taxes and social transfers. [Table A.2 in Excel Spreadsheet]

3)  The average size of social transfers as a proportion of households’ pre-tax income. [Table A.3 in Excel Spreadsheet]

4) A summary index of the degree to which transfers are targeted toward low-income groups.  Our measure ranges from -1.0 (the poorest recipient receives all transfer income) to +1.0 (the richest recipient receives all transfer income). [Table A.3 in Excel Spreadsheet]

5)   A measure of the extent of fiscal redistribution that is associated with public sector pensions; unemployment compensation; and other social transfers. [Table A.3 in Excel Spreadsheet]

6)  Measures of pre- and post-government poverty levels, using a composite measure that reflects both the headcount of those in poverty (the percentage of all persons whose equivalized household income falls below 50 percent of their country’s median) and the depth of their poverty (the difference between the median income of the entire population and the mean income of the poor).  [Table A.4 in Excel Spreadsheet]

7)  A measure of overall fiscal redistribution across households headed by persons between the ages of 25 and 59.  [Table A.5 in Excel Spreadsheet]

In measuring income, we have employed an equivalency scale that divides household size by the square root of the number of household members, weighting households by the number of members they include.  As to missing data, we have included households which report zero private sector income (i.e., all of their income is derived from the state) but have excluded households that report zero post-government income.  We have employed standard LIS top- and bottom-coding conventions, top-coding income at 10 times the median of non-equivalized income and bottom-coding income at 1 percent of equivalized mean income.

A more detailed description of these data is available in Vincent A. Mahler and David K. Jesuit “Fiscal Redistribution in the Developed Countries: New Insights from the Luxembourg Income Study,” Socio-Economic Review 4 (forthcoming, 2006). 

Please cite this article when referring to the data set, along with the web address http://www.lisproject.org/publications/fiscalredistdata/fiscred.htm.  You may also refer to LIS Working Paper #392 for additional details.

Any questions may be addressed to:

David K. Jesuit
Department of Political Science
Central Michigan University
Mt. Pleasant, MI 48859
David.Jesuit@cmich.edu

or

Vincent A. Mahler
Department of Political Science
Loyola University Chicago
6525 N. Sheridan Road
Chicago, IL 60626
vmahler@luc.edu

DOWNLOAD Fiscal Redistribution Data (Excel File)