A few years ago, I saw this Wait But Why Cartoon that perfectly encapsulated a phenomenon I’ve been thinking about more and more. Two stick figures, blissfully united in their agreement on points A and B, suddenly find themselves at odds when a third belief enters the picture. For many people, it’s as if the entire foundation of their relationship with someone crumbles over a single disagreement.
To explore this, let’s play a little game I like to call “Who Would You Admire?” The rules are simple: make a list of people you admire most. Now, imagine revisiting that list in a decade. How many names would survive the culling?
In 2015, Marginal Revolution readers played this game. They voted Elon Musk as the person they most admired. Tyler Cowen fared even worse here. The only name he could list was Aung San Suu Kyi as his sole admirable figure. Since then, both Musk and Suu Kyi have been disgraced, having their reputations tarnished.
Noah Smith’s 2015 list provides an even starker example. Of his top ten, at least half (including Musk, Elizabeth Warren, Shinzo Abe, Marc Andreessen, and Jon Stewart) are likely people that Noah would view with contempt today.
It’s as if the past decade has been a cosmic game of reputation whack-a-mole.
So what’s going on here? Have people suddenly become less admirable? Are we witnessing a mass moral deterioration? I don’t think so.
Instead, I think the following have happened:
- The Opinion Surface Area Expansion: We now expect people to have stances on an ever-widening array of issues. Your local dog catcher? Better have the right take on cryptocurrency and Gaza!
- The Great Opinion Unveiling: Social media and the documentation and indexing of everything have made it so every opinion and action becomes knowable to all, and forever relevant. In the context of public figures, this has meant that every action, statement, and misstep is immortalized, ensuring that each person’s most objectionable beliefs or actions are accessible to everyone as one of the first things you will learn about that person.
We’re simultaneously increasing the number of potential disagreements and our awareness of them.
One could argue that this increased visibility is a net positive. After all, shouldn’t we want to know the true nature of those we associate with? Isn’t it better to curate our relationships and communities based on a more complete picture rather than a less complete one?
On a gut level, this feels wrong to me. There’s something to be said for the blissful ignorance of not knowing every opinion held by every person in our lives and for spending less time dwelling on divisive topics. If you got along well with your work colleague until you discovered their view on topic X, then maybe discussing topic X isn’t wise.
The easy solution might be to simply avoid discussing topic X with your colleague, but perhaps the more meaningful approach is to focus less on topic X altogether. By doing so, we avoid contributing to its salience and polarization and should take a more disinterested stance when others engage in these debates, whether online or in person. In this context, it would also be helpful to recalibrate our expectations, allowing for greater tolerance when people make mistakes or hold “bad” views.
Yet, I remain very uncertain about all of this. It’s undeniably problematic when people do harmful things or cling to misguided beliefs, and it seems reasonable to consider these factors when judging others, but the consequences of this on a macro level seem awful. I’m curious to hear what others think about this, because I feel deeply confused…
I think you are right that the frequency of the behaviour described in the cartoon, which I’m going to dub Reversal Over Single Issue (ROSI, yes I’m going to be annoying and use acronyms), has increased. I also think the two potential causes you identified, Opinion Surface Area Expansion (OSAE) and Great Opinion Unveiling (GOU), are plausible explanations of at least part of the increase in ROSI frequency. I also personally find OSAE annoying because I think it leads to significantly more repeated and superficial conversations, but that’s not the topic.
I take you to be asking two questions:
1) Is ROSI a reasonable reaction?
2) Given the increased observed frequency in ROSI, how should I modify my interactions with others?
On 1), while you can imagine extreme cases where it would be a reasonable response, in almost all cases it isn’t reasonable. That’s the joke of the comic. Your evaluation of someone should be a considered judgement based on the quality of the belief, the strength of the conviction, the impact the belief has on their actions, moral uncertainty, the enduring nature of moral and aesthetic pluralism, etc. Your opinion of a dependable and kind musician friend with moderately bad political takes that they don’t act on in any way probably shouldn’t change. Your opinion of a person with detestable ideals who treats people badly should rightly be low even if they share some moral or aesthetic ideas with you.
But I think the emergence of OSAE in particular makes ROSI even less reasonable. There seems to be an expectation for people to hold more opinions on a wider range of topics, and to express less doubt about those opinions. I think this leads people to hold a bunch of superficial opinions they can’t really defend, and don’t act on in any way. So I think you now need to spend more effort determining if the person understands their position or holds it with any real conviction before in order to make a properly considered judgement.
On question 2), I do think a partial personal rollback of GOU and OSAE is advisable to try and reduce the frequency of being on the receiving end of ROSI.
I think as you alluded to there is value in identifying people who share beliefs which are important to you so that you can engage in shared projects. So I don’t think you should refrain from expressing core ideas which you feel most confident in. And I don’t think you need to withhold opinions from people who know you well and you can trust to hear you out and put in the effort to understand the context of what you’re saying.
But I think withholding unimportant or irrelevant opinions on controversial subjects from acquaintances or people who you don’t trust to make a reasonable effort to understand you is probably a good idea. As is avoiding public or social media statements on unimportant beliefs with high uncertainty or little impact on your actions, as there’s usually little surrounding context for you to explain yourself, little upside for you, and high chance of being on the receiving end of ROSI.
As always, I really appreciate your thoughtful comment.
A few thoughts and questions for you:
Re: “Your opinion of a person with detestable ideals who treats people badly should rightly be low even if they share some moral or aesthetic ideas with you.”
I think that’s the crux for me — this is less about a person having one specific bad take, but more so about, on a more meta-level, a given person having bad values (ie being a bully) or having a very poorly functioning thought process.
I guess there’s this broader question to grapple with: should we have thoughts and views on people we don’t even know? Like maybe I’m better off having the general stance that I should have zero thoughts, opinions, or feelings towards someone like Joseph Stiglitz (https://x.com/NathanPGoodman/status/1805964845851791448) or Marc Andreessen (honestly, way too many examples to link of him being a terrible human being).
As it relates to morality – in the pre-documentation era, we could hold people to a near-perfect moral standard. But now that we have more visibility into every bad thing a person has done, how do we lower our standards? Do we still expect perfect behaviour, do we downgrade to a handful of small offences being okay or one major bad act? Or do we get rid of character judgments altogether, with the view that everyone has done bad things? I think this is a pretty important problem and really thorny to grapple with.
Also, as it relates to expertise — we have so much more knowledge about how various institutions, politicians, or systems are making errors. And while things in fact are quite suboptimal, in the “before” days, we would be oblivious to most of these things and be more satisfied with lots of things in society.
I just saw this tweet:
https://x.com/RealTimeWWII/status/1826922480541331525
Can you imagine if people were on Twitter at the time, complaining about this?
I guess to summarize all this — now that we have so much more information, how should we adjust our standards for various things? While the question is simple, I don’t think I’ve reached a satisfying answer. Do you feel you have?
With respect to people you don’t know, I do think you should just have some prior (mine would be assume modestly good), and be slow to adjust it.
With respect to prominent people you don’t know, I think it’s better to be compartmentalize a bit. For the Stiglitz example (I listen to CWT but pick and choose episodes), I know he’s an economist and pitching a book so I’m not looking to see if he’s a good person. I’m looking to see if I learn anything about his economic reasoning or if his book sounds interesting.
Now that’s tricky with Tyler because he always tries to push people outside their core expertise. So for that particular YIMBY example, I see that he’s trying to bring it back to a point in his book and it’s in a challenging conversation environment. So I’m more inclined to be generous than say if he had written an extended NIMBY defence. And I would want to listen to the whole conversation and everything they talked about and see how that made me feel about his skills as an economist.
I’m not sure I agree that pre-documentation era that prominent were held to a higher perceived standard because people didn’t know anything about them. I think it was fairly standard for prominent people to sound off about many of their opinions. Consider Henry Ford and the parallel to Musk. Or Charles Lindbergh. And it’s at least my perception that the lack of information in the past was supplemented with large amounts of rumours and speculation.
To the extent that you do think more information has shown prominent people to be less good than you’d expect, I think that should just cause you to estimate that prominent people in the past were worse than their reputation.
I would be inclined to keep character judgements as is, with just a more realistic understanding that success, financial or otherwise, is not that correlated with good character.
With respect to expertise, I think it’s true that people have less trust in experts due to more information. I’m not sure whether that means people were more satisfied in the past. If I read biographical or semi-biographical stories from the past, say Orwell’s Homage to Catalonia, or A Farewell to Arms, they don’t seem satisfied with the people in charge.