design overlay
War and diplomacy viewed through granular data

War and diplomacy viewed through granular data

Sharing Tools

Link copied!

By Peggy McInerny, Director of Communications

A quantitative scholar of international relations, Eric Min joined UCLA in fall 2018. He begins teaching “Globalization: Governance and Conflict,” a core course in the Global Studies program, in winter quarter 2020.

UCLA International Institute, October 7, 2019 — Eric Min, a new member of the teaching faculty of the Global Studies program, has lived all over the country: Kansas, New York City, Palo Alto and now, Los Angeles. He came to love New York despite his initial culture shock and has since developed a deep affection for California.

“My graduate program was at Stanford, so I had a little time to get used to California and the West Coast,” he shares. “Now I'm starting to realize the aspect that I didn't appreciate: Northern versus Southern California and how it’s sacrilege to consider them the same thing.”

Min completed his Ph.D. in 2017 and joined UCLA in fall 2018 after a year as a Zukerman Postdoctoral Fellow at Stanford’s Center for International Security and Cooperation.

The long, comprehensive route

Min’s open, friendly demeanor is a sunny complement to his prodigious appetite for research, which focuses on the intersection of interstate war and diplomacy, international security and conflict management, and elite decision making. A quantitative specialist, the young scholar has found his stride in creating robust, granular datasets to test core tenets of established international relations (IR) theory.

In graduate school, Min began to find the formal models of IR conflict theory overly reductive. “You can’t learn anything about how a war actually happens if you reduce it to one observation,” he comments. Early on, he made a deliberate choice to take the time needed to create robust datasets in order to examine issues that interested him. “It’s a potentially higher-risk, higher-reward approach,” he says. That is, he has been slower to publish, but his work relies on far more comprehensive data than is usually the case, adding weight to his arguments.

Min has spent the last five years creating three comprehensive databases, and is now at work creating a fourth. One database, mostly derived from existing compendia, captures the outcomes and dates of 1,708 battles across 94 interstate wars between 1823 and 2003. Another captures the dates of negotiations that took place near in time to those battles and consists of data points that Min found by manually skimming thousands of articles and books on all 94 conflicts. These two databases were the foundation of his dissertation.

Royal Irish Rifles ration party during the Battle of the Somme, World War I, July 1916.
Photo: Royal Engineers No. 1 Printing Company, UK. Courtesy of Wikimedia Commons.

Yet another database, co-created with Azusa Katagiri for a recent article (see next section), consists of coded data from 18,000 declassified documents about the Berlin Crisis (U.S. State Department incoming cables, White House memoranda and media translations by the Foreign Broadcast Information System).* 

Today, Min is working with a group of political scientists and some 20 research assistants on perhaps his most ambitious project yet: collecting as many declassified National Security Council (NSC) files as possible from every presidential administration from Truman through Ford. The enormous project seeks to investigate three main issues: how information circulates within the government, what factors influence decisions taken by the NSC “in the room” and whether the decision-making structure and process varies from president to president. To date, the group has collected close to 600,000 pages of documents.

The upshot of his deep-dive approach to data has been an “eye-opening understanding of the brute force work that goes into making scholarly work happen,” says Min. “The first time I created one of these databases, I was naive. Now, I’m a little more realistic about it. But based on my past experience, I’m convinced that it’s worth it, as hard as it is.”

Publications and works in in progress

The UCLA professor published his first peer-reviewed article, co-authored by Azusa Katagiri of Nanyang Technological University in Singapore, in the American Political Science Review earlier this year.** The article grew out of their shared view that IR theory gave short shrift to the value of diplomacy.

After taking a course in machine learning, the two researchers decided to apply the technique to a deeply documented crisis in international relations in order to test the relative importance of public communications, private communications and material actions in how one state perceives the threat posed by another state. Their choice for the case study was the Berlin Crisis of 1961, viewed within the larger time frame of 1958–63. The paper analyzes how U.S. policymakers used these three types of information to judge the threat of military conflict with the Soviet Union over Berlin.

Berlin Crisis of 1961. Three Naval reservists at the end of the American Sector, where East Berlin began.
Photo courtesy of Wikimedia Commons.

In addition to the large dataset based on original diplomatic and administration documents described above, the co-authors created an additional, smaller dataset by manually coding 1,601 articles published in The New York Times about actions taken by the Soviet Union that reflected a willingness to use force in Berlin.

Their paper quantitatively refutes of the concept of “audience costs” — a core tenet of Cold War–era conflict IR theory first intimated by Thomas Schelling and elaborated by James Fearon. In a nuclear-armed world, audience cost theory assumes that during a crisis, public communications by a state are more costly to disavow than private communications (i.e., private diplomatic exchanges) and therefore more important in judging its intentions. By contrast, the authors found that the U.S. relied first on material events, then on private communications and then on public communications to judge the threat level posed by the Soviet Union during the Berlin Crisis.

Moreover, they make an important theoretical argument about information overload. “There is more to these two streams of communication — public and private — than just costs,” says Min. “Public diplomacy is a lot more chaotic; it’s a lot easier to miss and a lot easier to misinterpret public information because it's not necessarily directed toward you.

“In contrast,” he continues, “private diplomacy is much less frequent, so it's much easier to keep up with and much more direct… There's no misinterpretation.”

At present, Min is working on both a slew of working papers as well as restructuring his dissertation into a book. Using the first two databases described above, his dissertation argues that negotiations during war are important and highly strategic — often used to support a state’s war effort by stalling for time on the battlefield. The quantitative analysis discounts theoretical IR assumptions about wars resulting from the misrepresentation of parties’ strength or resolve, as well as the idea that diplomacy is of little value other than for ending wars.

“Negotiation is not just an afterthought, it’s not just a reaction to what’s happening on the battlefield,” explains the scholar. “It’s an integral part of how states try to navigate conflict and even try to fight back in conflict.”

Two of Min’s working papers are close to the publication stage: “Negotiation as an Instrument of War,” which distills the core argument of his dissertation, and “Talking while Fighting,” which looks at trends in the battle and negotiation datasets created for his dissertation. The latter finds that negotiations during conflicts have been more frequent after 1945, but are far less predictive of a war’s termination.

Other working papers, still being refined, address such topics as how Chinese citizens seek information about the South China Sea dispute during moments of high tension; an agenda for using machine learning research methodology to test long-held IR theories; when and how terrorists claim credit for attacks; and why military regimes that transition to civilian rule often “relapse” into military rule.


An indefatigable researcher, Min is also a dedicated teacher. In addition to having been inspired by wonderful teachers, he notes, “I also feel that some of the best researchers I know are also some of the best teachers.

“I enjoy teaching because I feel that I am imparting something that students absorb,” remarks Min. “For example, when you read something a student wrote and say, 'Oh, it sank in!’ That energizes me to keep at it and put as much effort as I can into it.”

In his introductory political science course, “World Politics,” he explains, “I talk about current events and try to map out how these theories and concepts that we’ve developed in political science can help explain or add structure to, the seeming chaos that is out there.” His graduate political science seminar, “Strategic Interaction,” is devoted to facilitated discussions of the most recent contemporary literature on the topic.

In winter 2020, Min will begin regularly teaching one of the three core courses required of Global Studies majors: “Globalization: Governance and Conflict.” The course addresses globalization, international law and current international institutions. Given that the current world order and its multilateral institutions are under severe strain, it’s a unique historical moment to be teaching the course. “Until recently,” says Min, “‘What is the future of the world order?’ was more a rhetorical question, but now we are starting to ask the question in earnest.”

“The Strategic Logic of War and Diplomacy,” a Global Studies senior seminar, examines how conflicts work, why they happen and how diplomacy interacts with conflicts — a topic at the core of Min’s research. “I spend real time deeply diving into individual conflicts in that course to help students understand how context can help explain why certain conflicts proceed as they do, and why others have been more successfully stopped,” he comments.

Min reflects that the unipolar moment of U.S. hegemony after the end of the Cold War may be ending, opening the way for greater engagement with great power politics once again. “I think we are seeing that American hegemony is not a fixed thing — we shouldn’t take it for granted,” he comments.

* The documents were transformed into digital form, broken down into 300-word segments (to account for the range of topics covered), and then pre-processed so as to eventually reduce commonly used words to “tokens” (e.g., for example, “talks,” “talking” and “talked” all became “talk”). The tokens were subsequently used as a key variable in statistical analysis, using machine learning and text-to-text techniques.

** Eric Min and Azusa Katagiri. February 2019. “The Credibility of Public and Private Signals: A Document-Based Approach,” American Political Science Review 113 (1): 156–72.