David Yanagizawa-Drott on Propaganda and Conflict

Interviewed by Molly Lanzarotta on February 24, 2012

There has long been anecdotal evidence that propaganda and mass media play a role in violence and conflicts. But to what extent? And how significant is the role of mass media during chaotic periods? David Yanagizawa-Drott, an assistant professor of public policy at the Harvard Kennedy School, examines these questions in a recent study. His research interests include economic development and political economy, with special focus on civil conflict, health, information and mass media.

Q: How did you become interested in the impact of mass media on violence and conflict?

Yanagizawa-Drott: My interest began when I went to Rwanda on a scholarship a few years ago. When I first read up on the 1994 Rwandan genocide, I was quite stunned by the amount of mass participation by ordinary citizens in the genocide. And so I became very interested in what drove people to participate in these killings. As I spoke to scholars and others in Rwanda and read up on this topic, it became quite apparent that many believed that mass media played a key role in fueling the riots. And so to me that raised the question of whether mass media in general can have an impact on conflict, on whether it can cause conflict, and whether it can induce people to participate in conflict.

But as I looked at the existing evidence and the literature on the impact of mass media on conflict, there really was not a lot of empirical evidence out there, so I thought it was worthwhile to investigate it.

Q: Could you please tell us more about that research?

Yanagizawa-Drott: The Rwandan genocide took place over the period of about three months in 1994. At least half a million of the Tutsi minority population was killed during this period, and a lot of the violence was performed by ordinary citizens of the Hutu majority population. During this period there was a radio station called RTLM that broadcast inflammatory propaganda calling for the extermination of the Tutsi minority population. So what I did to investigate the impact of these broadcasts was collect data on violence across villages in Rwanda and also whether the villages had access to these broadcasts. And what I found in my study was that these broadcasts increased violence, increased participation in these killings. And these effects were quite substantial when we looked at the magnitudes of the impact. This radio station was an important factor in fueling the violence in Rwanda.

Q: One of the challenges of understanding the impact of mass media in general and propaganda in specific is that much of the evidence is anecdotal. Explain your methodology for this research and how you have tried to make the evidence of cause and effect more concrete.

Yanagizawa-Drott: The main empirical challenge for a paper like this is to properly address the question: what would have happened in the absence of these broadcasts?

Rwanda is called “the land of a thousand hills.” There are literally hills and valleys all over the country. And how radio reception is determined in that kind of environment is essentially random. So some villages will have good radio reception and other villages will not have good radio reception because there are hills between the antennae and the village. That gives me a so-called “natural experiment” that allows me to test the impact of having access to these broadcasts. Using this methodology, what we find is that the villages that had good access to these broadcasts, that could listen to this radio station – they had higher levels of violence. And if we aggregate these effects across the country, we can attribute about 10 percent of the violence to these broadcasts. So this radio station had a significant impact on the genocide.

Q: How might this research inform policymaking?

Yanagizawa-Drott: My study shows that mass media can have adverse effects. In terms of policy making, I think it is relevant when you are thinking about restricting mass media in similar situations of ethnic conflict in the future.

For example, at the time of the Rwandan genocide, there were discussions about jamming the radio station. This never happened. One of the main arguments against jamming the radio station was that it would be a violation of national sovereignty. But in addition, calculations by the U.S. department of defense indicated that jamming this radio station would have a high monetary cost. What my paper shows is that not jamming the station – permitting these types of broadcasts – has a potentially high human cost as well. So in future, similar situations of ethnic conflict where we see similar types of broadcasts, the case for restricting mass media in different ways is potentially stronger.

Q: What are the larger questions that you seek to examine in your research? What future studies are you contemplating?

Yanagizawa-Drott: This research forms part of a broader research agenda looking into the impact of information on both political outcomes and economic outcomes. In particular, I am currently working on understanding anti-malarial drug markets in East Africa and agriculture markets in India. There is a separate research agenda that I am also currently working on, which is understanding the impact of political protest on public opinion and policymaking. There are actually surprisingly few studies that use data to investigate this relationship of how political protests affect political preferences, public opinion and ultimately policy. I am currently in the process of collecting data on historical political protests in different countries over the last decade or so to investigate how this affected public policy.

Print print | Email email