A lawyer once told me about a legal technique that he called “the CYA.” A typical manifestation of the CYA is when a district court judge (the first judge to render judgment in a lawsuit) decides the case on multiple, independent reasons. That way, if the judgment gets appealed and one line of reasoning is deemed incorrect, the overall ruling can still hold.
Here’s an example of the CYA that I noticed in a competition law case we were reading called Streetmap v. Google. The case had to do with whether Google favored its own product (Google Maps) to kick a competitor (Streetmap) off the market. The essence of the theory is that if you have a “dominant position” in a certain market (here, Google’s internet search engine), you can’t leverage that dominant position into a different market (here, the market for internet mapping software) to privilege your own product over another. Why? Because doing so enables you to artificially out-compete other market players.
Here, Streetmap claimed that Google added a “OneBox” function so that if you search for any location (or even anything related to a location — say, Korean BBQ) in Google, the first thing that shows up is a map. And that map, of course, is Google Maps — not Streetmap. The typical consumer will use the map right in front of them, going directly to Google Maps rather than Streetmap. So even if Streetmap is a better product than Google Maps, and even if most consumers would prefer Streetmap in a head-to-head comparison against Google Maps, Google redesigning its search engine can enable Google Maps to outcompete Streetmap. Such conduct, Streetmap claimed, amounts to an “abuse . . . of a dominant position” that violates Article 102 of the Treaty on the Functioning of the European Union, which is one of the core competition provisions in EU law.
The court in Streetmap said that to succeed on the lawsuit, Streetmap would have to show a three facts: (1) Google had a “dominant position” on the market for internet search; (2) Google used that dominant position in a way that is “reasonably likely to harm the competitive structure” of the market for internet maps; and (3) the likely anticompetitive effect is “appreciable” (it is more than a “de minimis” effect). Additionally, even if Streetmap can prove all three facts, Google can still prevail in the lawsuit by demonstrating that there were justifications for its conduct — that its conduct has benefits for consumers that counter-balance its anti-competitive effects and that its conduct is proportionate to achieving those pro-consumer benefits.
The way that the court went about assessing Streetmap’s claims are interesting from a legal perspective, but for the sake of this post, I’ll focus on the structure of the court’s argument. In essence, the court assumes without deciding fact #1 (that Google in fact has a “dominant position”) because it is going to conclude that Streetmap failed to show #3 (that the likely anti-competitive effect was appreciable). The court spends some eleven pages making the argument that the effects of the conduct were not reasonably likely to be appreciably anti-competitive (a mouthful, I know!).
The court writes of that conclusion: “That is sufficient to dispose of the allegation of abuse.” But, the court continues, “in case I should be wrong in that conclusion . . . I proceed to consider the issue of objective justification.” The opinion goes on for another nine pages to find that just in case Google’s actions were in fact reasonably likely to have an appreciable anti-competitive effect, there were sufficient objective justifications for Google’s conduct. CYA.
In this example, I think the CYA is a positive tool. There are of course other times in which the CYA isn’t so positive.
Take this silly Twitter thread between two Yale professors (one of whom is a Sterling professor!). As far as I can tell, one professor (Stanley) tweeted about how few Black philosophy colleagues he has had in his time in academia. Another professor (Christakis) tagged him and asked him about whether Yale (or its philosophy department?) is racist, and whether Stanley would leave his job since “I myself would not work for a racist institution, nor tolerate racial prejudice if I saw it.” Stanley’s response was that he “simply stated facts” and neither called Yale racist nor absolved it of racism: “I’m sticking to facts and not making normative claims.”
Before I sound like I’m creating a false equivalence, let me note that I think Christakis is being absurd and obtuse, and I think he’s looking for a fight for no apparent reason. That being said, I think Stanley’s response is a typical CYA answer: I didn’t say anything beyond facts, and I haven’t taken any normative position; you’re just assuming what I’m saying. It’s clear that Stanley’s comment about the few Black colleagues he’s had is meant to be an indictment of the departments he’s been in, as well as the wider field. No reasonable person would think he’s “just making an observation,” even if he leaves the normative point unsaid. But by leaving the actual call-to-action, the real normative argument, implied and unsaid, he accomplishes a CYA — that way, when pushed, he can simply say, “I have stated facts” and “don’t know what to think . . . .”
One last example. I’ve been amassing books on the Voting Rights Act for my dissertation, and I was perusing the CV of a Caltech historian, J. Morgan Kousser, since I knew he was active in the field. It turns out that last year, he reviewed a book I really enjoyed — Acharya, Blackwell, and Sen’s Deep Roots, which argues that the prevalence of slavery in 1860, measured by county, can predict racial attitudes of White southerners living in those counties today. It is a provocative thesis. Indeed, Acharya, Blackwell, and Sen claim that an institution that was outlawed a century-and-a-half ago is a better predictor of contemporary White racial attitudes than the current demographics of those counties. To me, that’s an incredibly powerful statement, and their evidence for it is quite persuasive.
But we should be specific. Note what they aren’t claiming. Acharya, Blackwell, and Sen never said that current demographics don’t predict White racial attitudes in the Deep South. Instead, they use statistical tools to show that the proportion of a county’s population that was enslaved in 1860 is a better predictor than the contemporary proportion of a county’s population that is Black for White racial attitudes. From a theoretical perspective, they’re trying to show that certain institutional decisions (here, the unfolding of Reconstruction and the post-Civil War political economy of the South) matter more for White racial attitude formation than racial threat theory, which posits that White people become more racially conservative when the proportion of racial minorities living near them increases out of a fear of having their power displaced.
When we read Deep Roots, the theoretical contribution we take away is, in essence, that “slavery, not racial threat theory, explains Southern White racial attitudes today.” But the book can’t actually say that. In other words, the book can’t, and doesn’t, “rebut” racial threat theory. It merely shows that a different factor has more predictive ability.
Kousser notes this point and shows how Acharya, Blackwell, and Sen essentially employ the CYA to preempt critiques of the book that defend the validity of racial threat theory. He writes, “Provocative in their subtitle and in some sweeping statements, [Acharya, Blackwell, and Sen] often qualify their thesis that contemporary differences in counties’ black proportions cannot account for observed contrasts in whites’ racial attitudes by inserting such words as ‘in part’ (pp. 14–15), ‘alone’ (p. 43), ‘exclusively’ (p. 78), ‘only’ (p. 101), ‘fully’ (p 107), or ‘solely’ (p. 162).” By including these nebulous qualifiers, the authors can effectively dodge criticisms, even as their theoretical contribution — the thesis that most readers will take away from the book — is meaningfully implicated by those critiques!
To take a general lesson, I think there’s two kinds of CYA — call them “constructive” and “destructive” CYA. With constructive CYA, the arguer bolsters their conclusion by adding superfluous information. Streetmap could have been decided solely on the appreciable anti-competitive effects analysis, but the judge added additional analysis just to cement the conclusion. With destructive CYA, the arguer subtly adjusts the conclusion (I’m not saying racial threat theory is wrong, just that it can’t explain everything!) to mask the limits of the argument. If a solid argument relies on a proportional relationship between the reasoning and the conclusion, constructive CYA bolsters the reasoning to strengthen the “ratio” of reasoning to conclusion; destructive CYA, in contrast, limits the conclusion so as to make it seem like the ratio is more proportionate.
The examples I gave of destructive CYA are, I think, helpful for spotting other CYAs. In the first example, the actual conclusion is left unsaid. By leaving the conclusion amorphous, the arguer can easily but subtly shift the goalpost upon encountering criticism. If an argument lacks a well-identified conclusion, that’s a sign of destructive CYA happening. In the second example, the addition of unquantifiable adjectives to modify the conclusion similarly allows the arguer to shift the goalpost. Weasel words and passive construction are good examples of this second form of destructive CYA.
One of my friends made a sly remark to me about nebulous new year’s resolutions a month ago. At the time, I suspected I was getting some side-eye for including a nebulous resolution about “improving my writing.” But thinking about the CYA has given me a few ideas for improving.
First, I will try my best to be more active in my sentence construction. Passive construction allows for the first kind of destructive CYA since you don’t have to identify a particular actor.
Second, I will try to ensure that my writing always has a central, identifiable takeaway — again to avoid the first kind of destructive CYA. This one is a bit harder since I’m also trying to cut the amount of time I spend on these posts; I find that when I write in a more stream-of-consciousness way, I more often than not find my conclusion by the end of the writing rather than having the conclusion at the beginning. Nevertheless, I will still strive to have articulated somewhere in each of these posts a central argument that is clear.
Third, I will try to ensure that those central arguments are phrased proportionately and precisely. That is, I will try my best to avoid nebulous adjectives (many, often, somewhat) in favor of specific claims to ensure falsifiability. One of my professors, I’m noticing, is really good at identifying the central idea behind a long article, and he’s very willing to distill it into a single statement. The result is that he phrases arguments very bluntly. To take an example, I once asked him about whether he thought the field of regulation formed an “epistemic community” that leads to a narrow way of thinking about regulation and that avoids asking questions about, say, the distributional consequences of regulatory decisions. It was a windy and overly complicated question. His response: “So you’re asking if we’re a cult?” I liked that response a lot because, in hindsight, I realized that I didn’t know that that’s what I was asking — but it definitely was.
* * * * *
Rose: This has been a fairly quiet week. I’ve spent a good amount of time thinking about my dissertation, though, and I made a trek to the LSE campus to get books! It was exciting. And my arms were pretty tired after carrying the stack home. Notice, at the top of the stack, a book edited by one Lorn S. Foster — a Pomona professor who just retired recently (and, in fact, whose successor I played a minor role in selecting). It was cool to find the volume as a kind of collision of a few of my different worlds. I also got to call a good number of friends this week and had a nice chat with my capital markets professor (who is terrifyingly brilliant and is the head of LSE’s law department and yet still made time to explain to me the magic of stocks).
Bud: This Wednesday, everyone in the Regulation program is presenting on their dissertation topic. I’m excited to try to hammer out mine and to see what everyone else is planning on doing.
Thorn: I’ve noticed that there are birds that love to chirp outside my window from around 2 to 3 am. It’s always a sad moment to hear them chirping — and, alas, even worse to hear them stop chirping.
Gratitude: One of the other Marshalls in London reached out recently to go for a walk. It turns out that he’s going back to the US soon until the lockdowns ease up. I felt like I hadn’t kept up with him super well even though both of us were in London, but I do really like his company. So I’m grateful that he reached out, and that we got to go for a walk, before he leaves.
Future topics:
* The question of tying-and-bundling versus refusal-to-deal in competition law — basically, there are two different theories of harm in competition law that are somewhat interchangeable as applied to contemporary questions of competition in technology markets. And the debate on how to distinguish between the two is actually really interesting.