The independent resource on global security

The misfortune of visibility in violent conflicts

Dr Damir Esenaliev and Susan Steiner

Economic disparities between ethnic groups are often seen as a driver of ethnic conflict. Understanding the extent of these disparities, and their sources, can help inform policies that aim to avoid further conflict. However, our research on economic disparities between Kyrgyz and Uzbek households in southern Kyrgyzstan suggests that the reality is far more complex.

Ethnic conflicts in Kyrgyzstan are uncommon but violent

In 1989 Kyrgyzstan was among the most ethnically heterogeneous of the Soviet republics. The three largest ethnic groups at that time were the Kyrgyz (totalling 60 per cent of the population), Russians (16 per cent), and Uzbeks (14 per cent). Today, Kyrgyz people make up 71 per cent of the population, Uzbeks 14 per cent, Russians 8 per cent, and all other ethnicities are each around 1 per cent or less.

In June 1990 mass riots broke out in the southern Kyrgyz city of Osh and neighbouring areas. About 170 people were killed and many more injured. One of the triggers for the violence was the claim by Uzbeks that the local government had disproportionately awarded plots of land and housing to Kyrgyz, and that the Kyrgyz people had prematurely occupied the land.

In June 2010 a second major episode of violence occurred in Osh, in which around 470 people, mostly Uzbeks, were killed. About 400 000 fled their homes (with many temporarily moving to neighbouring Uzbekistan), and a large number of properties were destroyed. The cause of the violence is not entirely clear, but it may have been triggered by a series of coordinated attacks carried out by separate groups of armed men.

In the aftermath, academics and journalists referred to the economic disparity between the two ethnic groups, with some even inferring that this was the root cause underlying the outbreak of violence. For example, the New York Times reported as follows: 

The most notable distinction, the one that is most responsible for the animosities that led to the recent violence, Central Asian experts say, is economic: Kyrgyz are traditional nomads, while Uzbeks are farmers. That divide has translated today into a wide class distinction, as Uzbeks have prospered and now own many of the businesses in southern Kyrgyzstan, which has engendered resentment.

No one has ever analysed welfare differentials in Kyrgyzstan  

Despite these claims, and the growing literature on inter-ethnic relations in Kyrgyzstan—most of it written by non-economists—no effort has been made to quantitatively investigate the extent and the sources of economic disparity between Kyrgyz and Uzbeks. Our paper, published in the Journal of Comparative Economics, contributes to closing this gap by measuring welfare differentials between ethnic Kyrgyz and Uzbeks involved in the violent clashes in Osh in June 2010.

Specifically, we apply the Oaxaca–Blinder decomposition to determine how much of the gap in mean welfare between the two groups can be attributed to differences in group characteristics and how much to differences in the returns to these characteristics.

Our research uses data from the 2005 Kyrgyz Integrated Household Survey to show that, in contrast to the public perception, Kyrgyz households have slightly higher expenditure than Uzbek households in urban areas, and about the same expenditure level in rural areas. Furthermore, our data shows that Uzbek households are better off than Kyrgyz households in terms of property values.  

Looking at the values of assets and housing

To the best of our knowledge, previous studies on economic disparity between ethnic groups in developing and transition countries have only analysed mean differentials and have mostly focused on differences in household expenditure. We go beyond the earlier literature in two respects. First, we complement the Oaxaca–Blinder decomposition with a quantile regression decomposition. This allows us to explain welfare differentials across the entire distribution.

Second, unlike many other studies, we use two other welfare measures besides expenditure, namely ownership of assets and housing property value. Our reason for doing so is based on the fact that assets and housing information capture different aspects of the standard of living. By considering different measures of welfare, we allow for several perspectives on economic disparity. Furthermore, visible aspects of welfare—such as the ownership of assets and housing property—might be more likely to provoke envy and grievances among a relatively poorer group.

Household survey data indicates that economic disparity is mainly observed in urban areas, where Kyrgyz have higher per capita expenditure and less valuable houses than Uzbeks. Differences in asset ownership are not so clearly established. Much of the expenditure gap can be explained by the fact that Kyrgyz tend to have higher education and fewer adults in the household.

The house value gap, which is the largest welfare gap in economic terms, is mostly due to large baseline differences that remain unexplained. We argue that Uzbeks have more valuable houses because their houses are generally of a larger size—being home to a higher number of people—and serve as bases for running small businesses. In sum, economic disparities between Kyrgyz and Uzbek households are not as predictable as would be expected.

Policy implications

Understanding the extent and the sources of economic disparity is important because of concerns that the disparity may have led to violent clashes between the groups in the past and may possibly lead to further tensions in the future.

Apocryphal evidence in Kyrgyzstan has long suggested that Uzbeks are more prosperous than Kyrgyz people. In contrast to the common perception, Kyrgyz have slightly higher per capita expenditure than Uzbeks in urban areas. The gap is mostly due to smaller households and higher education among the Kyrgyz, which is an indication that policymakers may have to find ways to increase education levels among the Uzbeks.

Furthermore, we present clear evidence that the Uzbeks are better off than the Kyrgyz in terms of the value of housing property. The house value gap remains by and large unexplained by the decomposition. This indicates that an economic perspective may be helpful but not sufficient for studying complex phenomena such as ethnic conflict. An interdisciplinary approach that takes into account political, social as well as economic aspects may be required.


This blog post is published as part of a collaborative partnership between SIPRI and Economists for Peace and Security (EPS).


Dr Damir Esenaliev was a Senior Researcher in the Peace and Development Programme.