Can conflict be predicted?

ETH News: Professor Cederman, your research deals with violent conflict, as well as the ability to predict it. How well can armed conflict be predicted in general?
Lars-Erik Cederman: The risk of future armed conflict can, in fact, be identified at an early stage. The risk of conflict occurring is high in regions with oppressed ethnic groups, for example. In Syria, the situation was already known to be volatile long before the civil war broke out. But conflict is enormously complex. In terms of predictions, conflict research is much like earthquake research, in the sense that scientifically substantiated risk maps can be created. But it is hardly impossible to predict exactly where and when armed conflict will actually occur in a region.

What are the difficulties of making such predictions?
World history does not add up to a linear sequence of logically ordered events. Instead, it is often erratic and unpredictable. This is especially true today, as demonstrated by Brexit or by the election of US president Donald Trump. Predicting the outcome of elections and referendums is already difficult enough, even though they follow established laws and principles. But armed conflict not only occurs much less frequently, it is also much more complex. Although its likelihood is to some extent based on regularities that can be studied, it hardly obeys established legal rules or timelines. In particular in times of major historical change, such as the current situation, projecting the events of the previous decades into the future will hardly work.

Some of your science colleagues are pinning their hopes on data science, believing that intelligent computer algorithms such as those used to analyse social media posts will make it possible to predict conflict in the future.
Nils Weidmann and I commented on this subject in an essay published in Science. There’s no doubt that data science provides new tools that we can use in conflict research. And I am convinced that big data can make our predictions even more accurate. But the optimistic view that predictions can be made much more precise and their temporal and spatial reach increased by amassing large quantities of potentially non-representative, unverified data, is exaggerated in our opinion. That is the main point that we make in our essay.

How, specifically, could conflict research benefit from data science in the future?
Media reports are an important source of data in conflict research. Nationalist tensions and potential conflict situations, for example, can be identified by analysing keywords. However, most of the data collection is still done manually. In his research, my colleague Nils Weidmann, who co-authored the essay with me, shows that new developments in data science have made it possible to analyse such data automatically to a certain extent. Software that interprets the significance of a text can be used to pre-select press article in order to speed up the coding process. This makes it possible to make faster assertions about political developments. However, it is premature to hope that highly complex conflict patterns can also be analysed in a fully automated way without any significant loss of precision. As conflict researchers, we don’t expect to be out of work any time soon.

Why might fully automated analysis be infeasible?
In our experience, computerised analysis is only possible to a limited extent. Software that can prioritise data doesn’t even exist for many of the main languages in our field. What’s more, people are needed to select the media sources. It is important to note that that the media in many regions is not independent, and a naive analysis would paint a distorted picture. When analysing social media, it is also important to bear in mind that some of the data is of dubious quality. In many regions of the world, especially where conflict is highly probable, the internet is censored and only made accessible to a minority.

Are there any other limits?
Data can only be analysed if they are available in the first place. My work covers the situation in Burma, for example, where only few people who live in the jungle there are connected to the internet. If researchers are interested in the views of the people living there, they have to conduct local surveys. However, indirect information can also be obtained in such regions by using computerised methods. In our own research, we use satellite images of light emissions to draw conclusions about economic welfare and inequality. Compared to using official statistics, this method offers the advantage of identifying short-term developments very quickly.