“We don’t know what we’re doing.” That is what Bear Braumoeller wrote regarding the current state of international relations scholars using statistical methods (Lauretig and Braumoeller 2018, 133-151). Braumoeller’s career was oriented towards addressing the flaws in how international relations scholars used statistical methods, showing both how to use such methods and the big payoffs from doing so. Braumoeller’s efforts were cut short, but his legacy can continue if scholars of international relations take seriously the lessons he offered.
To “know what we’re doing” meant doing something more fundamental than addressing the (albeit important) technical complications of failing to adhere to regression assumptions or ensuring measurement validity. It meant doing empirical work that was theoretically informed. As he and Anne Sartori wrote in a 2003 piece, “we are particularly concerned about…the widespread use of statistics with inadequate attention to the goal of testing theories of international behavior" (Braumoeller and Sartori 2003, 130). If we’re going to empirically take theory seriously, then we need to get serious about theory. A good theory of human behavior is not X then Y. The world is too full of contingency. Such contingencies needed to be incorporated into our empirical work. As Braumoeller wrote in one of his papers, though many theories in the political and social sciences posit complex causal processes, testing them, at least with respect to large-n approaches, has been given “short shrift” (Braumoeller 2003, 209-233). Empirically evaluating the theory was, for Braumoeller, where the proverbial rubber met the road. All too often, political scientists were creating blowouts.
Providing tools for empirically addressing the complexity of our theories was a core theme of his methodological contributions. They ranged from how to include interaction terms, the best ways of utilizing Boolean statistics, and accounting for conditions. He even addressed how such complexity could be accounted for in qualitative comparative analysis.
Though Braumoeller’s methodological contributions are notable, he was never strictly a methodologist. Consistent with his desire for methods to be theoretically informed, he studied methodology because he saw it as crucial for addressing important but empirically vexing issues related to international politics. Th-ese included the agent-structure problem and the persistence of warfare.
With respect to his work on the agent-structure problem, the initial fruits of such efforts were published in 2008 in the discipline’s flagship journal, American Political Science Review, and then his book, Great Powers and the International System (Cambridge University Press, 2011). The goal of this work was to determine if the relations between nation-states were more a function of the choices of those individual states—i.e. the agents—or instead compelled by the environment in which the states found themselves existing—i.e. the structure. Of course, one naturally influences the other, in a perpetual feedback loop. The relationship is highly complex and even circular. That is why empirically evaluating this issue is so hard.
The heart of Bear’s approach is to say that capturing system influence is not a “partial equilibrium” question. Specifically, you can’t rely on simply identifying the direct relationship between, say, the number of major powers in the international system and the incidence of war. This would be captured by understanding how the coefficient in a standard linear regression approach. Instead, one must consider the “general equilibrium.” This means we should try to estimate both the direct and indirect, or second order, relationship of a change in the number of major powers. This would be captured by understanding how changing the value of a single variable influences a system of equations, not just a single equation. Hence, rather than identifying an “effect” of the system, Braumoeller’s approach was to show the “pressure” the system exerted on the behavior of a state and how that pressure steered a state to behave in a particular way.
Braumoeller’s work on the persistence of warfare was a natural extension of this research on how complexity and non-linearity shape international politics. Drawing on the well-known phrase from the early twentieth century poet, essayist, and philosopher George Santayana, the title of his 2019 book, Only the Dead, perfectly captured his thesis: war will always be with humanity. This runs counter to claims, made popular by social psychologist Steven Pinker’s 2011 book The Better Angles of our Nature, that war and violence between nations has been on a steady decline since the middle of the twentieth century.
Braumoeller acknowledged that wars in recent decades have remained relatively small in size compared to the devastation wrought by the First and Second World Wars of the early-to-mid twentieth century. However, Braumoeller argued that there was little reason, statistically or substantively, to believe that this trend would hold. The size of wars follows a complex and non-uniform pattern known as a power-law distribution. This means that a series of smaller conflicts can suddenly produce a massive conflict in terms of scale and scope. Long stretches of time without a major war can pass before a massive conflagration erupts. Hence, even if wars have been smaller in recent decades, this should not deceive humans into thinking that war as a human practice has gone away or that massive wars are no longer possible.
Braumoeller’s work is unfinished, which makes his untimely passing all the more tragic. We still don’t know what we’re doing, and it will be harder for us to figure that out without Braumoeller’s guidance. But international relations scholars should still try to do better. Striving to do better large-n work on the big questions is the best way we can honor his legacy. ■
