I have been meaning to comment on the results of the Survey of Income and Living Conditions (SILC) 2010, but hadn’t got around to it with all the excitement on fiscal matters. As has been widely reported, the survey showed a dramatic increase in income inequality, with the quintile ratio (the ratio of the incomes of the highest income quintile to the lowest income quintile) rising to 5.5 from 4.3 in 2009. The Gini coefficient also jumped dramatically, from 29.3 to 33.9. (A Gini of 0 represent perfect equality; a Gini of 100 represents perfect inequality – i.e. one household having all the income.) These increases more than reversed the measured declines since the middle of the decade.
Although these increases took place at a time of wrenching adjustment in the Irish economy, the magnitude of the increases in measured inequality is still surprising. All the more so since the ESRI’s analysis of the 2010 budget indicated that the changes were progressive (see here for an analysis of the period 2008 to 2012). (See also page 17 of this European Commission report on Ireland.)
The SILC is an extremely carefully constructed survey. However, as noted in the survery report, sampling variation is unavoidable, and results must be always be interpreted carefully. One indication that sampling variation may be an issue is the fact that the number of “household heads” with a third-level qualification or higher increased to 1,830 from 1,505 (p. 83). This increase occurred despite the total number being surveyed falling to 11,587 from 12,641. Thus the proportion of survey respondents with a household head with a third-level education or higher – a fraction that changes relatively slowly over time – increased to 15.8 percent from 11.9 percent. This suggests that sampling variation may explain part of the measured increase.
My purpose here is most certainly not to diminish the importance of inequality as an issue – quite the contrary. And it may well be true that inequality will turn out to have risen over this recession after balancing market and post-market factors. But those concerned with inequality might be unwise to put too much emphasis on one year’s survey results. To the extent that the increase in inequality partly reflects sampling variation, there is a greater likelihood that it will fall again in the next survey release. This could too easily be interpreted as indicating that recession-related inequality effects are not an issue. Perhaps of greater concern than the results in SILC 2010 is the breaking of the recent trend of progressive budgets in Budget 2012 (see Figure 1 here). However, we will need to see more than one budget to accurately gauge the distributional implications of Government policies.