Clemson researchers code two centuries of conflict to understand, predict war
CLEMSON, South Carolina — Mining nearly two centuries of data, a Clemson University political scientist hopes to help policymakers and scholars better understand and predict causes of international conflict.
Assistant professor Steven Miller will use a $92,000 grant from the National Science Foundation to work with the University of Alabama on a collaborative research project.
Miller will team with undergraduate students at Clemson over three years to expand and add further detail to a collection of data Miller has already revised with Doug Gibler, a professor of political science in the University of Alabama’s Institute for Social Science Research. The research team will study militarized conflicts from 1816 to 2010. Miller said the results should help political scientists predict how international conflicts unfold.
“An international dispute is a strategic situation like bargaining over the price of a good,” Miller said. “This research will more fully test our theory that all arguments between nations are basically bargaining situations, and the resulting data will allow us to predict how a dispute may unfold.”
Miller added that an illustrative case of bargaining in World War I is an incomplete approach to understanding the bargaining framework for all international conflicts because most don’t resemble that classic case. With this in mind, the Clemson research team will get granular data on all levels of conflict for thousands of disputes, such as the Indo-Pakistani war in the 1960s and the Prussian war of 1865.
While it may be difficult to see the connection between a conflict originating more than 150 years ago and modern disagreements between states, Miller is looking for similarities and trends across data. If the research reveals that several conflicts that began without clear initial positions or demands often result in similar outcomes, that trend may help to predict the outcome of a similar, future conflict.
“We want to look at not just what is demanded but how those demands are communicated and how demands evolve over time during a conflict,” Miller said. “The data we acquire will help us determine if the results of certain responses to a provocation or action are generalizable.”
The more information documented and coded, the better the prediction. The collaboration between Clemson and the University of Alabama has already yielded many gigabytes of data made up of thousands of PDF files and text files just for conflicts occurring between 1950 and 1992.
The undergraduate researchers’ first order of business will be to dig into repositories of historical newspapers, such as the Economist and the Times of London, which date to the mid-1860s so that the researchers can beef up data on 19th century conflicts where existing information is the scarcest. Miller said there will be many centuries-old newspaper clippings scanned and emailed back and forth between Clemson and Alabama over the next few years.
“This will be an exciting opportunity for student researchers and it’s going to entail a great deal of work,” Miller said, “but Doug and I believe this research could significantly influence our discipline and shape research in international conflict in the future.”
The award is currently the only active National Science Foundation political science award in South Carolina, but is rare for other reasons. Jeffrey Peake, professor and chair of Clemson’s political science department, said awards from the National Science Foundation are quite uncommon in political science in general, particularly in departments such as Clemson’s that lack a graduate program. This is the largest grant in the department’s history.
“Our department has grown and consequently it has received multiple awards in the last few years, but this is by far the largest and most prestigious,” Peake said. “We look forward to seeing Steve’s progress along the way and what this will mean for our discipline for years to come.”
This material is based upon work supported by the National Science Foundation under Grant No.1729138. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the view of the National Science Foundation.