Does Income Inequality Derive the Separatist Terrorism in Turkey?

Abstract

Separatist terrorism has been a severe problem for Turkey since the mid-1980s. The conventional wisdom contends that economic deprivation in southeastern Turkey is the fundamental reason for the long-running battle against the Kurdish rebels. Considering that there is limited empirical literature on the roots of terrorism in Turkey, yielding conflicting results about the claim that the main cause of terrorism is deprived economic conditions, this study aims to answer whether there is a causal relationship between income inequality and separatist terrorism in Turkey. To this end, the Global Terrorism Data Base for the period of 1973–2006, two Theil indices of pay inequality as proxy for income inequality, and the vector autoregression and Autoregressive Moving Average (ARMA) methods are utilized. The results support the early findings that income inequality, a particular focus in this paper and an essential indicator of economic deprivation, is not a main cause of escalation of separatist terrorism in Turkey.

Keywords:

• Economics of terrorism
• Separatist terrorism
• Income inequality

JEL Classification:

• D71
• D74
• H56

Acknowledgments

We would like to thank the participants of the 16th International Conference on Economics and Security, 21–22 June 2012, The American University in Cairo, Egypt, where the earlier version of this paper was presented and Gülbanu Altunok and an anonymous referee for their invaluable comments.

Notes

1 In this context, the best case would be utilizing an inequality data that are based on household surveys, where the within inequality in the region can be measured. It is a fact that not just inequality between east and west increases but also the one between different social strata within the regions might increase. However, due to the limited number of household surveys it is not possible to reach a different inequality measure for a long enough time period. Nonetheless, it would not be wrong to argue that it is more likely that Kurds pay more attention to discrepancies in income compared to the west, rather than discrepancies in income within the same region between different social classes.

2 In this context, Newman (2006) shows that there is no clear correlation between economic inequality and terrorism between 1996 and 2003 and adds that this linkage is becoming less strong when it comes to ideological terrorist organizations. On the other hand, using panel data from 65 countries for a relatively longer time period of 1975–1999, Krieger and Meierrieks (2010) show that higher inequality is associated with more terrorist occurrences.

3 Despite the recent ‘Kurdish Initiative’, – which has not been implemented due to some internal political dynamics – it can be argued that taking military measure alone is still the main policy response of the state (Alptekin, forthcoming).

4 UTIP-UNIDO and Elveren and Galbraith (2009) Theil Inequality indices are available at http://www.utip.gov.utexas.edu/data.html under UTIP-UNIDO and Turkey sections, respectively.

5 We acknowledge that terrorism is conducted by individuals and/or organizations whereas we take the income inequality between regions in our analysis since there is no data at individual level for such long period of time. However, this does not lead an ‘ecological fallacy’ since inequality trend between the regions a good proxy for the income inequality between individuals from two regions in average term.

6 These provinces are Adiyaman, Agri, Bingol, Bitlis, Diyarbakir, Elazig, Erzincan, Hakkari, Kars/Ardahan/Igdir, Mardin/Batman, Mus, Siirt/Sirnak, Tunceli, and Van. Igdir and Ardahan, Batman and Sirnak who used to be districts of Kars, Mardin and Siirt, respectively, became provinces. However, in order to prevent an arbitrary increase in the Theil’s T statistic due to an increase in the number of provinces (from 67 to 81 throughout the study period) Elveren and Galbraith (2009) recalculated data based on 67 provinces.

7 Controlling for survey type and the share of manufacturing employment in a population, Galbraith and Kum (2005) shows that pay inequality in the manufacturing sector has greater consistency across time and countries than is shown using the Deininger and Squire inequality measure, which is the most common inequality measure.