Research Highlights Podcast
May 14, 2025
The cultural roots of rebellion
Eleonora Guarnieri discusses the role of cultural distance in driving civil conflict in Africa.
Yida refugee camp in South Sudanese territory hosts roughly 68,000 refugees from the Nuba Mountains.
Source: Marco Gualazzini, CC BY-SA 4.0
蝌蚪网
Civil conflict has plagued much of Africa, with ethnically diverse countries experiencing particularly high rates of violence. Yet within these nations, patterns vary, leading to questions of why some groups rebel while others do not and why a given group rebels at certain times but not at other times.
In a paper in the 蝌蚪网 Review, author untangles the factors that drive groups to rebel against their central government. She shows that when ethnicities become more culturally distant from those in power, their likelihood of engaging in civil conflict increases significantly.
Her research suggests that conflicts arise as a result of ethnic favoritism in resource distribution and from fundamental disagreements over the types of public goods that central governments should provide.
Guarnieri recently spoke with Tyler Smith about how she estimated the impact of cultural distance on civil conflict, and what her findings may mean for reducing violence across Africa's diverse societies.
The edited highlights of that conversation are below, and the full interview can be heard using the podcast player.
Tyler Smith: What does cultural distance mean in the context of your research?
Eleonora Guarnieri: For economists, culture is this big bundle of preferences, beliefs, and tastes that shape behavior and that members of a given group tend to share and agree on. Based on this definition, you can think about cultural distance between a pair of groups as the difference in preferences, beliefs, and tastes and issues that they tend to not agree on. That's what I mean by cultural distance in the context of this paper. The challenge is how do we measure this bundle of differences. One way, which is quite common, is to use linguistic distance as a measure of cultural distance.
Smith: If you're thinking about preferences and beliefs, it's not obvious that this similarity between languages would be a good proxy. Can you elaborate on why linguistic distance is a good proxy for cultural distance?
Guarnieri: The idea is that languages as they are today are basically the result of horizontal separations between populations that have happened throughout history. So, languages we know belonging to the same linguistic family can be reconciled to a common ancestor basically. And if you think about these horizontal separations between populations, you can also think that this has gone hand in hand with some form of cultural divergence and divergence of cultural values. Also, of course, at the core of believing that this measure is a good measure of cultural distance, there is the idea that language is an important carrier of culture. And it's a carrier of culture that is also transmitted intergenerationally within populations, within cultural groups.
Smith: I think it's worth elaborating a little bit on the types of conflicts you have in mind. Can you describe what these civil conflicts look like?
Guarnieri: In my paper, I focus specifically on ethnic civil conflicts, meaning those that are fought along ethnic lines. You can imagine a civil conflict breaking out whenever one or more armed groups in a country, whose fighters and soldiers are recruited along ethnic lines and whose identity is really tied to ethnicity, start to rebel against the central government. The focus of this piece of work is conflicts that happen over government power, so overthrowing the government or changing its composition. In a nutshell, these are conflicts that are fought over public goods. The idea is that these diverging preferences, these cultural differences between groups, may matter the most for conflicts over public policies and public goods that everyone in a country must share, whether they like them or not.
Smith: Your hypothesis here is that if we increase the linguistic distance between the groups out of power and the groups in the central government, we're likely to see more fights to overthrow that government. How do you go about testing this idea empirically?
Guarnieri: The first step is really to assemble data that will allow me to test whether it's the case that whenever linguistic distance to the government increases for a given ethnicity, that ethnicity will be more likely to fight. I start by tracking all ethnic groups in the continent of Africa. I track all politically relevant ethnic groups over time from 1961 to 2017. Then I measure two key things. First, I construct a variable that captures whether, in a given year, a given ethnic group is engaging in conflict against the central government or not. This is my outcome variable of interest. And the second thing I measure is the linguistic distance, a proxy for cultural distance, between each ethnic group in the country and the ethnicities that are forming the central government in that specific year. With this panel data set in hand, I exploit the fact that ethnic groups enter in and out of power. There are government transitions over time, meaning the ethnic identity of the government changes. If the government changes, an ethnic group may experience either an increase or a decrease in cultural distance to the central government. This is precisely the type of variation I exploit to understand whether changes in cultural distance affect an ethnicity's propensity to fight. This design is very powerful because I can include a set of what we call fixed effects to basically really be able to isolate the role of cultural distance in shaping conflict from other characteristics at the ethnicity level that may also affect the group’s propensity to fight.
Smith: You find a standard deviation increase in linguistic distance increases conflict with the government by 0.4 standard deviations. What does this mean in practice? How big are these effects?
Guarnieri: These effects are relatively large. Think about a typical government transition in my sample and the consequent change in cultural distance that this government transition would generate for a given ethnic group in the country. Such a median government transition would basically increase an ethnicity’s propensity to fight whenever linguistic distance increases by 39 percent compared to the sample mean.
What is really driving this increase in the propensity to fight is triggered by cultural distance, but it's triggered by cultural distance due to the fact that culturally distant groups have different preferences and disagree over public goods.
Eleonora Guarnieri
Smith: You make this distinction between not only the distribution of public goods that the government is providing, but also the type of public goods that are being provided. Why does this distinction matter?
Guarnieri: I try to show that what is really driving this increase in the propensity to fight is triggered by cultural distance, but it's triggered by cultural distance due to the fact that culturally distant groups have different preferences and disagree over public goods. I use survey data to show that, indeed, respondents that are culturally distant from the government tend to disagree or dislike what the government is doing. Groups that are very distant from one another also have different ideas about what the government should do, irrespective of who is in power at the time of the survey. What we want to understand, especially in the context of Africa, is what they disagree on. Are they disagreeing on the allocation of public goods? This is a big issue in the continent, as ethnic favoritism has been widely documented in the literature. Favoritism from the part of the government in distributing public resources not only occurs for in-group members, but it also extends to culturally close groups. So we want to know whether this whole story is about ethnic favoritism—that this is what triggers conflict. On the other hand, it could also be that this disagreement not only refers to the allocation of public goods, but also about the type of public good that should be provided. I try to understand whether in contexts where favoritism is less prevalent—meaning where the government spends more public resources on goods that are more public in nature versus more particularistic, and whether these public goods are equally distributed—cultural distance still plays a role in explaining conflict. And I find that to be the case. I do see that if you distribute resources more equally and if you invest in public goods more, the effect is smaller. This does alleviate the effect of cultural distance on conflict, but not fully, so there is still an effect which suggests that disagreement over the type of public goods provided also plays a role in explaining conflict.
Smith: This seems to raise a new challenge for policymakers. Do you have any ideas on how policymakers might address conflict over the types of public goods provided by the central government?
Guarnieri: I would like to emphasize that there is still a lot of potential to distribute public goods and public resources more equally, especially in the context I'm studying. This does alleviate conflict, and it's an important part of the effects I uncover. But part of the effect seems to be driven by the type of public goods provided. Here I don't have hard empirical evidence at hand on what can be done. Certainly, striving for political inclusion and new institutions that may accommodate diverging preferences within a country when deciding on how to spend public resources could be key in alleviating and preventing conflict. There is some emerging, recent work, especially in Africa, on how to alleviate ethnic tensions. I think contact, both cultural and economic, may help along this dimension, but the evidence is still very mixed and it's not clear what could work, especially in the longer run.
♦
“Cultural Distance and Ethnic Civil Conflict” appears in the April 2025 issue of the 蝌蚪网 Review. Music in the audio is by .