Research on ethnic-based preferences in Africa

I’m happy to report that this article, co-authored with Gwyneth McClendon, was recently published in Comparative Political Studies and is now available for your reading pleasure. I paste the abstract below, but the punchline is that even when we control for all sorts of individual-level and contextual factors, African citizen policy preferences vary quite systematically along ethnic lines. We think this has important implications for how we understand the working of ethnic politics, and the challenges of governance under certain types of ethnic diversity.

Of related interest, a few articles on ethnic categorization and violence, co-authored with Prerna Singh, can be found at our project website.

The Ethnicity–Policy Preference Link in Sub-Saharan Africa

Evan S. Lieberman

Gwyneth H. McClendon


Scholars have begun to investigate the mechanisms that link ethnic diversity to low levels of public goods provision but have paid only minimal attention to the role of preferences for public policies. Some argue that ethnic groups hold culturally distinctive preferences for goods and policies, and that such differences impede effective policy making, but these studies provide little evidence to support this claim. Others argue that preferences do not vary systematically across ethnic groups, but again the evidence is limited. In this article, we engage in a systematic exploration of the link between ethnic identity and preferences for public policies through a series of individual and aggregated analyses of Afrobarometer survey data from 18 sub-Saharan African countries. We find that in most countries, preferences do vary based on ethnic group membership. This variation is not merely an expression of individual-level socioeconomic differences or of group-level cultural differences. Instead, we suggest that citizens use ethnicity as a group heuristic for evaluating public policies in a few predictable ways: We find more persistent disagreement about public policies between politically relevant ethnic groups and where group disparities in wealth are high.

Assessing race-based affirmative action in South African universities

A thoughtful essay today from Max Price, the vice-chancellor of the University of Cape Town about the dilemmas of affirmative action. While acknowledging many adverse consequences, he ultimately concludes the policy is still necessary, and I think rightly assesses broad consensus about this. Nonetheless, the debates in South Africa are obviously reminiscent of those in the United States, where we also continue to wrestle with how to address a history of institutionalized discrimination without contributing to the furthering of group division.

…There are two fundamental arguments against the use of race. The first is that racial categorisation undermines our national commitment to nonracialism.

It forces us, and especially youngsters born at the time of the first democratic elections, to view the world, themselves and others in terms of racial categories.

The second argument against race as a basis for affirmative action is that it may include black students who are certainly not disadvantaged, may come from wealthier homes than most whites and may have had the benefits of 12 years of private-school education.

These are the reasons we ought to move away from a race-based policy. We should accept it in the interim only if there is no ­better solution and only if the advantages outweigh the disadvantages. Unfortunately, the University of Cape Town’s experience is that this is still the case.

If the task were only to identify economic disadvantage, this could be done by asking about income or by looking at the school a potential student attended. But the problem is that educational disadvantage has been the consequence of many determinants — including, but not limited to, economic disadvantage..