The problem with experts

What we've learned about expertise, the dangers of too much agreement, and why politicians are bad at politics.

Caprice Bourret (Rachell Smith / CC BY-SA

Welcome to the first Ruffian of 2021. For reasons I don’t fully understand, this edition is really long.


One of the most striking video clips of the past year was this encounter on the Jeremy Vine Show between the show’s resident medical expert Dr Sarah Jarvis, and the former supermodel Caprice. It took place on March 16, a week before the UK went into lockdown. Caprice argues that the UK should be following the lead of Asian countries and taking tough measures to contain the virus. Dr Jarvis rips into her for talking “nonsense”. “With the greatest respect”, she says, implying the opposite, “unless you’ve read every scientific paper, you cannot argue with me on that.” When Caprice persists, Vine nervously suggests that since Jarvis is “the expert” we should only listen to her. What makes this clip so excruciating is not so much that Jarvis was wrong - it’s OK to be wrong - but her aggressively condescending tone. I see this tone a lot on Twitter and have been guilty of it myself - the swollen self-regard of one in possession of the facts; the contempt for those who dare to contradict. This pandemic should be a reminder that even the smartest, most well-qualified people can be wrong. A common assumption, made by Vine here, is that “experts” must always be deferred to. In fact, a big lesson of this crisis is that while we owe a huge amount to experts, we do not owe them deference. Many public health experts were wrong on lockdown, on masks, on border controls - and wrong in the face of evidence, not in the absence of it. The truth is that every field of expertise has its own systemic biases and irrational orthodoxies (since academics usually only have their opinions tested by each other, and not by consumers and voters and real world events, they can easily become unmoored from reality). I value expertise immensely, but it’s easy to confuse it with things with which it merely intersects - critical thinking, wisdom, moral worth. We non-experts have a responsibility to listen to and value experts, and experts have a responsibility to deliver opinions in a manner that does not imply their audience is stupid - and the best way to do that is not to assume it. (I returned to the topic of experts and Covid a few times this year - see here, here and here).

One of the themes of my forthcoming book, CONFLICTED (which you can read about here) is that disagreement is crucial to unlocking collective intelligence. Groups of people generally make better decisions than individuals, but only when they air conflicting opinions, so that those opinions can test and modify each other. But since most people like to get along and to be seen as co-operative, groups often behave as if the goal of a discussion is to reach an agreement rather than a good decision. In November the American public health committee responsible for immunisation met to discuss how Covid-19 vaccines should be distributed. In contrast to the UK, they decided to prioritise key workers over the very elderly - even though doing so would mean 6% more deaths. The reason? Racial equity. Since key workers were more racially diverse than the elderly, and Covid seems to impact ethnic minorities disproportionately, it was deemed important to level the playing field. The decision passed on a unanimous vote. A month later, after a public outcry, the committee reversed its decision, also in a (nearly) unanimous vote. In this excellent op-ed on the affair in the Washington Post, Megan McArdle points out that when the 14-person committee deliberated, nobody even made the opposing case. It’s not as if it was a difficult one to make: ‘How about we follow the plan which means the fewest people get killed?’ Even if the initial decision was the correct one - I doubt it, but I may be wrong - the way it was taken was unforgivable. As James Surowieicki (author of The Wisdom of Crowds) puts it, “If a group of 14 people trying to solve a challenging, complex problem all agree on the solution, that’s a sign the group is not working well.”

If you’re enjoying The Ruffian, please be superspreaders on its behalf. Always include the magic link - - or share this post.

I’m really grateful to all of you who have done so in 2020. You’ve helped this newsletter grow like Zoom’s stock price (OK maybe not quite that much). Thanks also for all the lovely messages of appreciation and support, they mean a lot.

Oh and please pre-order my book.


  • The mutant strain is like a really awful plot twist - the discovery of a bomb about to explode. Yet at the same time, it feels like we’re closer to the end than ever. The UK’s single dose strategy is an act of desperation, but it’s the right thing to do and I suspect other governments will follow. This should, supply allowing, speed up our vaccine rollout, which is already pretty fast. If things go well the pressure on the NHS will start being relieved by the end of this month. We were hard on the government’s slowness to act in March; we should recognise they are moving swiftly and boldly now.

  • A good post on the problem of “pandemic perfectionism”. Governments and scientists have spent too long looking for the perfect solutions instead reaching for “good enough” ones.

  • An excellent books of the year list with many books that aren’t on any of the other lists.

  • You know when you’re falling asleep and your body jerks? What’s that about?

  • McCartney III is a delight. Apart from anything, it’s an incredibly modern-sounding record for a self-produced album by a 78-year-old. While most of the credit should to go to Macca, he works with a couple sound engineers who know him well and they’re interviewed here. While most of the gear talk flies over my head, I found some fascinating insights into his working processes here too. Bonus: I just found out the artwork for the album is by Ed Ruscha, probably the greatest visual artist alive. Not a bad guy to have in your phone when you need a job doing.

  • You have a responsibility to be happy.

  • If the dinosaurs or some other ancient species had an advanced industrial civilisation millions of years ago, would we know about it today? Not necessarily, say these scientists, in what they call the Silurian Hypothesis. So when humans disappear, everything we’ve built may disappear with us. There is reassurance in insignificance.

    As I’ve mentioned before, I think David Shor, the young US political analyst and former Dem operative, is the most interesting commentator on US politics - and by extension, politics generally - around. So I’m super excited that he got together with another of my favourite minds, Julia Galef, for a conversation. I had to keep pausing this podcast to digest the insights, which just keep coming; what follows is a mere taster. Julia asks a fascinating question - why aren’t political campaigns more rational? There’s a lot of evidence for what works and what doesn’t, whether on strategy (eg which issues to focus on) or tactics (eg ads vs canvassing) but politicians and parties often make terrible, self-harming decisions. This happens despite those involved having a powerful incentive to remain grounded in reality - the desire to win elections. Shor says that one reason that campaigns go wrong is that campaigns are run by people who are essentially quite odd yet believe they’re normal. Activists and campaign strategists are on average much more educated, much more liberal and much younger than the median voter (the effect is asymmetric btw: it pulls the GOP towards the centre, and the Dems away from it). Politics nerds have different priorities to most voters: the political scientists Eitan Hersh and Ryan Enos carried out a survey of Obama activists in 2012, and found that about a third of them named income inequality as the most important issue facing the country. Less than 1% of voters agreed. Activists and staffers tend to hear from people, on Twitter and elsewhere, who only reinforce their basic intuitions (journalism is also increasingly dominated by young, highly educated hyper-liberals). Shor mentions a memorable ad produced by Hillary Clinton’s campaign in 2016, called “Mirrors”, in which shots of adolescent girls are intercut with Trump saying horrible things about women. I myself remember everyone saying how powerful it was. A study of responses to the ad found that it actually made voters slightly more likely to vote for Trump. They felt patronised by it - that while Trump was talking about real issues like immigration and trade, Clinton had only emotional manipulation to offer. The ad only works if you already hate Trump, which means it doesn’t really work at all. One of Shor’s themes is that the political classes have become much worse at speaking to voters. In the old days of politics - the days that Biden grew up in - you often encountered voters who weren’t on your side, since so many voters were non-aligned. That made it easier to learn how to speak to voters (as opposed to activists or donors). You had to learn, through trial and error, about what matters to people and what doesn’t, how voters think and what they feel. Hiring experts in data analytics does not substitute for this. Shor himself was a 20 year old data analyst on the 2012 Obama campaign, and he was contemptuous of the old-school consultants with their intuition-based voter talk. He later came to realise that most of the time, they had been on the right side of internal arguments. As I said in my recent piece, Biden won not despite his longevity but because of it - which, if true, is quite worrying for the future of the Democrats. Not everyone is as capable of understanding the limits of their own worldview as Shor.

    Among the music clips I posted in The Ruffian last year, my favourite is the one of Aretha and Smokey at the piano. I found it in an excellent article from The Believer which also links to a clip of Stevie Nicks singing backstage, while having her make-up done for a photo-shoot. I find it quite moving and I’m not sure why - something to with the spontaneity of it, a great performer caught in the act of pure singing, carried away by the song. Happy New Year.

Share The Ruffian