Can political theories be tested like scientific ones? And if they can, does it matter? Alexis Papazoglou argues in the New Republic that, even if they can be tested, it is less important than other factors in the success of the political theory. In his signal case, the conflict between anti-globalist populists and the conventional international order is questioned as resulting in clear outcomes that somehow will determine the viability of one theory versus the other. It’s an ongoing experiment. Papazoglou breaks down the conflict as parallel to the notion that scientific processes ultimately win on falsifiability and rationality. In science, as per Kuhn’s landmark The Structure of Scientific Revolutions, the process is more paradigmatic agendas, powerful leaders, and less calculated rationality.
The scientific process may have been all of those things, of course, and may continue to be so in the future, but there are ongoing developments that make it less likely that sociological factors will dominate. And this is why the comparison with political theories is perhaps wrongheaded. There may be a community of political theorists but they are hardly the primary architects and spectators of politics, unlike science and scientists. We are all political actors, yet very few have the time or inclination to look carefully at the literature on the threat of successful authoritarian Chinese civilization versus Western liberal democracy, for instance. But we are not all scientific actors, despite being governed by the reality of the world around us. Politics yells and seethes while science quietly attends a conference. Even the consequences of science are often so gradualistic in their unfolding that we barely notice them; see the astonishing progress on cancer survival in the past decades and note the need for economic discounting for global climate change, where the slow creep of existential threats are somehow given dollar values.
But the most important changes in scientific progress are in the gradual professionalization and globalization of science itself. The communities that serve as peer reviewers are quite unlike anything in politics. Imagine if a politician’s statements and claims were reviewed by peers for coherence and rationality prior to being fielded to the public. Imagine if that was the primary method for communications between politicians and their subjects.
Second, there are newer tools that change the falsifiability narrative. Take Judea Pearl’s Structural Causal Models (SCM) that formalize methods in associational statistics with experimental and counterfactual arguments. The goal here is to help explain why pure data-driven model formulation (for instance, by an artificial intelligence) needs a larger ladder of logical methods in order to explain the model and to ensure that it is less wrong than it could otherwise be. This has much the same flavor as hidden or latent variable modeling and the analysis of confounds in experimental design. While there may be some modest parallels in political theorization, there is no overarching desire for accuracy and accountability in politics per se.
Finally, politics is self-fulfilling. Hate globalization and want to put up high walls of tariffs around your country? It may have sufficient buoyancy for a subset of local manufacturers to offset the broader complaints about import pricing. The voices and human inputs wash out any stronger signal that overall GDP is on a slide. Science can’t and doesn’t operate that way. The data ultimately intervenes with incommensurabilities, to use Kuhn’s term, that must be reckoned with, and a trail of papers and textbooks that can’t be ignored. We may ignore history in politics at our peril, but we never ignore it in science.
Still, Papazoglou may be right that an alternative political narrative is more powerful than experimental evidence or the test of time. Politics is the art of impossible persuasion, it seems, in modern America, if not the world.