Tom Pepinsky

Author: tompepinsky

  • U.S. Politics in the Age of the Babbling Equilibrium

    Ever since Sean Spicer’s press conference in which he insisted, against all evidence, that President Trump’s inauguration crowd was the biggest in history, the Trump administration has faced a problem of credibility. Every time the administration issues a message that is demonstrably false, it undermines trust that any future message can be trusted.

    One illustration of the consequences of diminished administration credibility is the current debate about the American Health Care Act (no link provided because as I write this, no one knows what it actually says). One argument for repealing and replacing the Affordable Care Act is that individual state exchanges are so fragile that the health system will collapse anyway. Here’s a quote from NPR’s Kelly McEvers interviewing Alabama Representative Bradley Byrne yesterday afternoon:

    What we are hearing from people in the health insurance industry is that these plans are deteriorating so rapidly that we cannot wait.

    There was a time when I would have interpreted this piece of information about the fragility of Alabama’s health system as evidence that there is something fundamentally wrong and that needs to be addressed, regardless of my views about Byrne’s own political views and how he might solve such a problem. Yet as I listened to the interview yesterday, I realized that there has been a fundamental change in the way that I process information delivered by President Trump and his surrogates. My first instinct is now that I simply do not believe what they say. I no longer believe that “these plans are deteriorating so rapidly that we cannot wait.” Or more precisely, the fact that President Trump’s surrogate has uttered that phrase no longer has any effect on whether or not I believe that it is true. It could be, it could not be.

    Welcome to the age of the babbling equilibrium in U.S. politics.

    The term “babbling equilibrium” comes from game theoretic models of communication, in which a “sender” takes an action that is meant to convey information to a “receiver,” but both the sender and the receiver realize that there are strategic incentives to act in certain ways. This way of thinking is useful for making sense of all sorts of things: poker, nuclear strategy, buying a used car, dating, and so forth. In many situations, such as those just listed, the actions of the sender can be characterized as “cheap talk” (informal presentation here, formal presentation here [PDF]). As an example, when buying a used car, it is almost certainly irrelevant to base your decision on how clean the car’s exterior is, because it is nearly costless to wash a car and it conveys no information about how good the engine is, and you ought to know this, and so should the used car dealer.* The word “babbling” in babbling equilibrium conjures the image of a 6 month old child babbling—the content of the babbles doesn’t tell you anything, and so you don’t change how you respond based on hearing “goo goo” versus “ga ga.” This is an equilibrium in the sense that neither the sender nor the receiver has any information to behave any differently based on how each expects the other to respond.

    Political speech has always, of course, been strategic. “Talk is cheap,” and babbling equilibria exist in any cheap talk game. But it is important to contrast the difference between a babbling equilibrium world and a world in which the sender and the receiver differ on how to interpret and act based on the same facts. It was supremely important for the George W. Bush administration to use evidence to support its intention to invade Iraq in 2003. The Obama administration similarly relied on evidence and reasoning to develop its case for health insurance based on the costs and benefits of the ACA relative to the status quo. In neither case did their opponents agree, but both administrations benefited from a general consensus that the arguments would need to be evaluated on their merits. They did so, I presume, because they recognized the benefits of establishing their own credibility for future negotiations.***

    In a babbling equilibrium, the Trump administration’s public statements mean nothing. They mean nothing in the precise sense that their interlocutors should learn nothing about what the administration’s actual position is, or what it is willing to do, from the administration’s public statements (see, for example, the negotiations over changes to the draft of the AHCA that are unfolding as I write this). This undermines the administration’s ability to be persuasive, which is bad for the administration’s ability to direct legislation.

    But it is also bad for U.S. politics more generally. By now, it is common to observe that the administration will at some point in time need to tell the truth about something important. That doesn’t just hurt the administration, it hurts everyone who is affected by administration policy. The worrying scenario is something like a national security emergency or a homeland security threat. In that scenario, citizens should want to be able to trust the administration to say true things and not false things. Without that, I cannot see how this administration could ever make good policy about the things that I care about, even if we share the same interests.

    NOTE

    * The precise definition of a babbling equilibrium from Sobel (PDF) is “the sender’s strategy is independent of type and the receiver’s strategy is independent of signal.”
    ** Recall how politically meaningful the “Bush Lied, Thousands Died” slogan was?
    *** For the argument that the credibility motive can sustain truthful communication over time, see (here [PDF]).

  • An Interpretive Ethnography of Interpretive Ethnography

    While reading Lisa Wedeen‘s “Reflections on Ethnographic Work in Political Science” I was struck by this description of ethnography due to my grad school friend Tim Pachirat*:

    Ethnography as a method is particularly unruly, particularly undisciplined, particularly celebratory of improvisation, bricolage, and serendipity, and particularly attuned to the possibilities of surprise, inversion, and subversion in ways that other methods simply are not. If we think of the range of research methods in political science as a big family, ethnography is clearly the youngest, somewhat spoiled, attention-seeking child, always poking fun at and annoying her more disciplined, goal-oriented, and outwardly-successful older siblings. Ethnography is the method who [sic] comes home to family reunions with the new mermaid tattoo, with the purple hair, with yet another belly button ring, and with a moody, melancholic artist for a girlfriend. At the dinner table, she is the method who interrupts her older brother’s endless description of his stock portfolio with tales of the last full moon party on Phi Phi Island in Thailand. Given that kind of unruliness, it’s no wonder that the older siblings and father figures of our discipline often revert to the language of “disciplining” and “harnessing” ethnography, of bringing her wild and unruly impulses under control by making her abide by the rules of the dinner table. In short, ethnography may be fun and exciting, but she might also get you excommunicated from the family.

    It is an exciting characterization! Perhaps coming from my background in Asian studies, however, is strikes me as strange. In the community of anthropologists, qualitative sociologists, “Indologists,” critical political economists, and others who together constitute the undisciplined world of “Southeast Asian political studies,” nothing could be more established or conventional than ethnography or interpretivism.

    This raises interesting questions when read next to the rest of Wedeen’s essay about ethnography, specifically her invocation of “work” in the Foucauldian sense:

    Ethnographers beholden to Foucault do this by analyzing the “work” discourses do—their underlying assumptions, omissions, implications, and effects, as well as their historical conditions of possibility.

    What I find interesting is the “work” that the self-understanding of ethnography as an unruly outsider method does, both in the context of the broader discipline of political science and to those scholars who produce texts and train students. I am quite certain that an interpretive, ethnographic approach would be the right one here. What is being conveyed—to whom, for whom—with the bit about “purple hair”? What models of the social world are implicit when we suppose that an academic discipline is like a family? What are the “conditions of possibility”** for ethnographic methods to be understood as “young”? In the true Foucauldian sense, how does this discourse of unruly outsiderness challenge, reconfigure, or even construct relations of power; and between whom?

    I mean this completely seriously. The importance of actually understanding how different scholars create meaning is a theme that Andrew Little and I took up in our discussion of critiques of formal models of comparative politics. As an illustrative exercise about the work that discourse can do, I tried to recreate Pachirat’s unruly outsider paragraph from the lens of a different group who consider themselves something of an unruly, insurgent outsider group*** who read voraciously from across multiple disciplines and borrow promiscuously from them:

    Causal inference is particularly unruly, particularly undisciplined, particularly celebratory of critique of established scholars and their findings, and particularly attuned to the possibilities of surprise and subversion in ways that other methods simply are not. If we think of the range of research methods in political science as a big family, causal inference is clearly the youngest, somewhat spoiled, attention-seeking child, always poking fun at and annoying her more disciplined, established, and outwardly-successful older siblings. Causal inference comes home to family reunions with a copy of Cryptonomicon under one arm and a TV on the Radio LP under the other, and snickers at her father’s admonition to “use probit or else!” At the dinner table, she interrupts her older brother’s endless description of his last full moon party on Phi Phi Island in Thailand by interjecting “how utterly conventional” and leaving it at that. Given that kind of unruliness, it’s no wonder that the older siblings and father figures of our discipline often revert to the language of “disciplining” and “harnessing” causal inference, of bringing her wild and unruly impulses under control by making her abide by the rules of the dinner table. In short, causal inference may be fun and exciting, but she might also get you excommunicated from the family.

    I will note that this wasn’t particularly hard to do. I wonder if it rings true.

    I’ll conclude by clarifying what I fear might not be clear from above: If you are reading this as anti-interpretivist snark, you have missed my point. I take it as incontrovertible that interpretivist methods and ethnography have value and place in political science. I am suggesting that we use these methods to understand ourselves better.

    NOTE

    * Tim’s Every Twelve Seconds is the closest thing to a new classic of anything written by anyone near to my grad school cohort. At least, that’s what I think.
    ** I have never been able to pin down this term satisfactorily. It is not good that I can still use it in a sentence.
    *** For example, “the credibility revolution” and “randomista.”