Income inequality is one of the most controversial issues in Turkey. What we know, generally speaking, is that Turkey has the most unequal income distribution among European Union countries. However, public opinion is not in agreement as to whether inequality is on a decline since the Justice and Development Party (AK Party) came into power in 2002. Political parties, columnists and regular citizens belonging to the opposition believe and argue firmly that income inequality worsened in recent years without, indeed, providing any statistical evidence. Their positions are guided much more by ideology than dispassionate statistical analysis. However, this belief is somewhat puzzling since it makes it difficult to explain the increase in AK Party votes in the July 2007 and June 2011 general elections.
The Turkish Statistics Institute (TurkStat) publishes a new survey, “Income and Living Conditions Survey” (ILCS), with the direct aim of estimating income inequality as well as material deprivation. TurkStat began the survey in 2006, but because incomes reported in its surveys refer to the year before, the resulting information on income inequality relates to the years 2005 to 2010. Before this new survey, research on income inequality was conducted based on Household Budget Surveys (HBS), whose main purpose is to estimate household's expenditures rather than income inequality. The ILCS report is based on Eurostat standards and its sample is enough large to allow regional estimation on income inequality. I would like to summarize the main results of ILCS and comment on them following the recent research note of on ILCS findings of Bahçeşehir University's BETAM research center (“Income Inequality in Turkey and regions: 2005-2010”).
First of all, let me remark that from a methodological point of view, the estimation of income inequality is a difficult task. There are different measures of income inequality, each of them pointing out a different dimension or feature of income inequality. Moreover, and here lies the major difficulty, they can evolve in opposite directions within the same period. TurkStat computes two measures: The Gini Index (GI) and the rate of Poverty Risk (PR). The GI measure is particularly sensitive to individual income changes at large while PR is to the changes at the bottom of the distribution. PR counts individuals with incomes below 60 percent of the median income for a given year, showing the share of these individuals, under risk of poverty, in total population. An increase in both GI and PR indicates worsening inequality, while the opposite indicates an improvement.
So what happened in Turkey regarding income inequality between 2005 and 2010? GI decreased from 42.8 percent to 40.4 and PR from 25 percent to 22.6. The improvement is evident albeit limited. Having said that, the picture is quite different regarding the three sub-periods defined within the study. Almost the entire improvement in income inequality occurred during the high economic growth period from 2005 to 2007. Moreover, the improvement in the bottom bracket-- the decrease shown by PR-- has been more pronounced than GI. Obviously, the 10 percent increase in income per capita in Turkey during this period has profited the lower rather than the highest incomes. This is also widely the case at a regional level. Except Northeast and Central Anatolia, the two measures of inequality decreased in ten other regions, confirming the improvement in income equality within the majority of regions. Given these facts, one can admit that the electoral victory of the AK Party in the elections of July 2007 was no surprise.
Then the global economic crisis occurred. The Turkish economy started to contract from the second quarter of 2008 and the contraction reached its deepest point in the first quarter of 2009, pushing unemployment up dramatically. From 2007 to 2008, GI increased by one percentage point but PR stayed almost fixed. Nevertheless, in 8 regions out of twelve, the two measures increased together. These facts might be considered evidence of worsening income inequality and living conditions during the economic crisis. Here also the political outcome can be easily observed: AK Party votes decreased from 46 percent to 38 percent in the local elections held in March 2009.
The period of strong economic recovery from 2009 to 2010 is a bit complicated regarding income inequality. There is a very limited increase in the GI measure from 40.2 percent to 40.4 but a quite clear decrease in PR from 23.5 percent to 22.6. Nevertheless, the tendencies in regions are highly confused. We have an improvement in four regions but a worsening in three. Regarding the five remaining regions, we observe that GI increases but PR decreases in two of them, while GI decreases but PR increases in three of them. Unfortunately, we don't have yet the figures for 2011, during which the growth rate reached 8.5 percent. Considering the relation between growth and income inequality in the previous periods, it would not be too risky to predict that income inequality would have decreased during 2011, which, in turn, can explain why AK Party could get back its lost votes in the June 2011 elections.