Nature's bad editorial is a small part of a bigger problem

www.slowboring.com

Nature's bad editorial is a small part of a bigger problem

Matthew Yglesias

14 - 17 minutes

Hey Chicago! Come hang out at an informal Slow Boring happy hour tonight, starting at 5:30 p.m. at Radio Room, 400 North State Street.

In October 2020, the scientific journal Nature took the unusual step of endorsing Joe Biden’s election — the first time the publication had ever endorsed a candidate for office.

“No US president in recent history has so relentlessly attacked and undermined so many valuable institutions, from science agencies to the media, the courts, the Department of Justice — and even the electoral system,” they wrote. “Trump claims to put ‘America First’. But in his response to the pandemic, Trump has put himself first, not America.”

Their argument was not particularly novel, but I was broadly sympathetic to the points and certainly found it understandable that the journal’s editors wanted to do what they could to contribute to the functioning of society. It also did not, at the time, strike me as an intervention that was likely to be successful. And this fall, Nature’s sister publication, Nature Human Behavior, did what science journals do best and published an experiment. Floyd Jiuyun Zhang conducted a pre-registered exercise in which he randomly exposed participants to information about the Nature endorsement. He found that doing so “caused large reductions in stated trust in Nature among Trump supporters” and a big reduction in Trump supporters’ level of interest in Nature as a source of information on Covid-related topics like vaccine efficacy. He reports that “the endorsement also reduced Trump supporters’ trust in scientists in general,” but there is “little evidence that the endorsement changed views about Biden and Trump.”

I do not find this result shocking.

Changing minds about presidential politics is an undertaking with a very high degree of difficulty. The Nature article was not well-designed to persuade anyone, but my guess is that more conventional newspaper endorsements probably matter a fair amount for things like city council primaries and not at all for presidential general elections. What really did surprise me was Nature’s reaction to the follow-up study, which was to publish a second editorial arguing “political endorsements might not always win hearts and minds, but when candidates threaten a retreat from reason, science must speak out.”

This second editorial is really dumb, and I can’t explain why much better than Josh Barro does here. It also, unfortunately, speaks to a much broader pathology: modern technology has greatly increased the size of the audience that the average person can reach and also greatly increased the number of people who can reach genuinely large audiences. And due to the stochastic nature of virality, it’s at least possible that something a very obscure person says will end up circulating widely. As Uncle Ben says, with great power comes great responsibility — people need to learn to at least try to practice responsible posting. Nature is in the unusual position of having specific evidence that what they did was counterproductive and they persisted anyway, which is really bad.

Holden Thorp, who edits the Science family of journals, liked the Nature editorial and explained his view in a series of badly misguided tweets. Flagging this poll result from Pew, Thorp argued that “following the admonition to stick to science is conceding the idea that scientists can be sidelined in policy decisions. ‘Stick to science’ infantilizes scientists and tells us to sit at the kids table and let the adults decide.”

He followed that with a complaint that the public doesn’t actually want science, it wants “scientific information they can use as they see fit.”

I think any left-of-center person can very clearly see the problem with this by thinking about a community of experts whose political opinions they generally disagree with.

For example, the mayor of a large city cannot completely ignore the opinions of the police department’s leadership or rank-and-file officers or other law enforcement experts. These professionals have information, expertise, and practical know-how that is invaluable. At the same time, “just do whatever the cops want” is not a good approach to policing either. Some of that is because the cops may be sincerely mistaken about things — practitioners have unique insights but also blind spots — but it’s also because like any interest group, they have selfish desires that are not in the public interest.

The military is another obvious example. Congress can’t draw up the budget for the Navy without talking to naval officers and taking their views seriously. But “information they can use as they see fit” is ideally what the officers would be giving Congress. Congress probably can’t actually get that idealized information in the real world; the information provided is invariably colored by self-interest, by bias, by blind spots. That’s life, and we should be realists about it. But if we started seeing admirals tweeting about how “information they can use as they see fit” is a bad ideal, and actually the way it should work is that Annapolis graduates receive absolute deference, that would be alarming.

And note that these dynamics can be poisonous in both directions.

I think progressive city officials really did go overboard in ignoring cops’ professional expertise in the wake of George Floyd’s murder. But that happened in part because police unions spent the prior five years being unduly dismissive of urban progressives’ concerns and failing to distinguish their specific expertise from their all-around conservative political opinions.

By the same token, I don’t see how anyone could read Thorp’s tweets and not come away with a somewhat reduced level of confidence in anything politically relevant that comes out of his journals. Scientists are, as a whole, well to the left of the American public in their political opinions. That doesn’t mean their outputs will be biased by groupthink, peer pressure, and ideology. But it certainly means those are reasonable things to guard against. When a top editor dismisses those concerns, dismisses the idea of worrying about the appearance of bias, and then says that providing people with “information they can use as they see fit” isn’t a good regulative ideal, public confidence naturally falls. That’s bad, and it’s especially bad if you think scientific knowledge has something important to offer the political domain.

I often feel like actually trying to persuade people is underrated.

It’s a tough assignment because it involves doing something a little unnatural. When I wrote “The Case for Bernie Sanders” for Vox, I tried to write an article that didn’t just repeat the normal pro-Bernie talking points because I wanted to address moderate Democrats’ fears.

So I noted, for example, that exactly zero of Bernie’s far-fetched policy ideas would come to fruition in a Sanders administration because the congressional pivot points would be in the hands of someone to the right of Joe Biden. But I also argued that Bernie is a veteran member of Congress who’s actually been a reasonably effective legislator. He knows how to get bills and amendments passed, and he has a good track record of saying yes to other people’s incremental reforms. He’s absolutely cast “no” votes dissenting from bipartisan legislation he regards as too moderate or otherwise inadequate. But whenever his vote is absolutely needed to get something important done, he casts it in support of the Democrats’ agenda, including on bills like Dodd-Frank and the Affordable Care Act. Of course, Bernie didn’t become president. But his role during the Biden era has been exactly what I expected: he’s been a pragmatic legislator who didn’t throw bombs and blow up the IRA or the American Rescue Plan.

I will admit that not as many moderates were persuaded by my pro-Bernie argument as I hoped. But I really was trying — trying to imagine a persuadable audience, trying to imagine what they were worried about, trying to imagine how I could set some of those fears to rest.

None of this is as fun as clowning around and dunking on someone your intended audience already dislikes.

And it’s totally fair that people want, to some extent, to just have fun on the internet and not worry too much about whether the things they’re saying or doing are constructive politics. But I do think Eitan Hersh’s concept of “political hobbyism” is important here because politics-adjacent speech invites confusion in a way that speech about a hobby like Minecraft or college basketball or a monthly book club does not. Because part of what makes politics compelling for the minority of people who pay close attention to it is that politics really matters, that the stakes are high in presidential elections, and that alongside the nonsense that accretes to any campaign, we are also debating ideas of consequence.

Under the circumstances, it’s incredibly easy to delude yourself into believing that the forms of political expression you personally enjoy are also the ones that are constructive and useful. Sometimes that might even be true. But generally, you need to be on guard against self-deception.

The upshot of this is that I think everyone, from the editors of science journals to professors who like to tweet about their areas of academic expertise to political obsessives posting pseudonymously while slacking off on their real jobs, would benefit from thinking a little bit harder about what they say.

We’re all habituated to living in a world where there’s a big dichotomy between being a guest on the Today Show and blowing off steam at the bar. You have one way of thinking about your words for that big broadcast television spot (cautious, measured, considered) and another when you’re just shooting the shit with friends. But the nature of social media is to collapse that distinction. It wouldn’t do to be stilted and formal all the time on social media — it’s social! — but at the same time, it really is media. It’s very hard to say how many right-wing frog-meme people there are or what their actual influence is, but their weird, creepy, racist, anti-semitic behavior during the 2016 campaign was incredibly off-putting. I’m sure 90% of the people who posted “defund the police” Instagram memes in the summer of 2020 had never looked into this topic and had no genuine strong feelings about police budgets, they were just trying to express the idea that racism is bad and avow solidarity with its victims. But that didn’t stop “defund the police” from becoming a huge political issue that’s backfired and hurt reform causes.

A lot of scientists felt strongly about the importance of urging people to adopt sound approaches to the pandemic in their personal lives and also about the merits of Biden vs. Trump. Mushing these two ideas together was not constructive any more than tweeting “ban cars!” is a good way to convince people that we should alter our transportation policy priorities. There’s no way to force everyone to act like well-behaved messaging robots. But I do think we all have the ability, at the margin, to control our own behavior and the kind of behavior we incentivize. Fundamentally, we all have to try our best to discourage ill-disciplined and counterproductive behavior, even in the absence of a quick fix or obvious solution.

At this point, I think astute readers are probably thinking “does this asshole realize how often he’s failed to practice what he’s preaching here?”

And honestly — I do! This is a topic I’ve thought a lot about because I love writing at high volume. I loved writing multiple blog posts per day, I love tweeting constantly, and when you do what you love, you sometimes take things too far. I’m also someone who thinks that on some level, there is too much cowardice and too much groupthink in the content game, so I’d hate to be read as simply encouraging everyone to be more fearful and timid. What I would like is for everyone to be more thoughtful and considerate. In Dan Zak’s profile of me for the Washington Post, he highlights two specific instances (my Rana Plaza blog post and my Uvalde tweet) that I think are clear and egregious examples of content that failed to meet this standard. In both cases, I had a kernel of an idea that I thought was valid but the time, place, and manner of the posts and the framing weren’t just “provocative” but obnoxious and counterproductive.

But I would say a more workaday problem is that like everyone, I sometimes have a bad day and feel grumpy and easily annoyed. When you write from a place of peevish annoyance, you can sometimes generate entertaining copy or social media posts that feel therapeutic. But that’s rarely constructive.

As a counterpoint, though, I’d offer my tweeting jag about fake license plates in D.C. This was actually a considered strategy that I knew would prompt some backlash under circumstances where I thought a backlash would be constructive by elevating the salience of a problem nobody was talking about but where I think the solutions are relatively uncontroversial. Despite the perception of a Twitter pile-on, I’ve gotten positive feedback from a number of elected officials around the country. We had ANC members in D.C. start to tweet about the problem, the Department of Public Works is becoming more responsive to citizen complaints about fake tags, Tom Fitzgerald (a Philadelphia Inquirer reporter who follows me on Twitter) published a great investigative piece about the Philly version of this problem, local TV news in D.C. did a story about it last week, and Brianne Nadeau who chairs the relevant committee in the D.C. Council is promising to put forward legislation. I don’t want to claim that my tweets somehow single-handedly caused all this, but I was not just posting because I felt irritated — I felt that relatively easy and uncontroversial solutions were available if only more people paid attention to the issue, so it would be constructive to do something attention-grabbing.

But many times in my life I’ve posted thoughtlessly or self-indulgently, just like I’ve eaten bags of potato chips I should have left unopened or had one more drink at the bar when I should have said “enough.” A lot of people don’t enjoy public writing at all, but for those of us who do like it, I think it’s pretty natural to be sloppy about it sometimes. That said, I think this is the kind of situation where you want to respond to potential accusations of hypocrisy by leveling up and saying we should all try to reach a higher standard rather than avoid these accusations by leveling down and giving everyone a pass. The Nature thing was unusually egregious because they had a study at hand, but that really ought to be a signal for us to all be more diligent.