Two readers have written with pertinent observations about the surprising data adduced by Tom Smith purporting to show that one out of three people in Sweden and the UK were impoverished, while a mere 17% in the U.S. were.
Chris Bertram (Bristol Philosophy) writes:
"I think the key line to notice in the Blackburn paper referenced by Tom Smith is 'To the extent that differences across countries in the living conditions of those at the low end of the income distribution are not fully reflected in the LIS measure of disposable income, those poverty comparisons may be misleading .....' (p. 374). See also (and Table 1 gives a very different picture ....), 'Do Social-Welfare Policies Reduce Poverty? A Cross-National Assessment', Lane Kenworthy, Social Forces, Vol. 77, No. 3. (Mar., 1999), pp. 1119-1139."
Meanwhile, Chris Edmond (UCLA Economics) observes:
"1. The paper referred to by Tom Smith comes from the once-a-year *non-refereed* 'papers and proceedings' issue of the AER (some papers from that year's AEA meetings).
"2. The very first footnote of the paper says that because the unit of analysis in the Swedish data is the individual tax unit which treats individuals aged 18+ as separate income-receiving units (when households are typically considered more relevant -- households pool the income of several individuals...) the paper's measures of absolute and relative poverty are likely to be overstated for Sweden relative to other countries."
The second point is surely an understatement--by itself, it may well explain the screwball results for Sweden.
My thanks to the two Chris's for sharing their expertise.
UPDATE: More from Smith here. I'm heading out the door, so my further thoughts--if I have any useful ones on the subject--will have to wait.
Recent Comments