View profile

TiB 140: How to buy an election; accelerating scientific progress; "Shy Trump voters"; and more...

Revue
 
This week: Why "Shy Trump Voters" don't explain polling failures; how to buy an election (or even an
 
November 10 · Issue #140 · View online
Matt's Thoughts In Between
This week: Why “Shy Trump Voters” don’t explain polling failures; how to buy an election (or even an entire political system); one model for accelerating scientific progress; and more…

Welcome new readers! Forwarded this email? Subscribe here. Enjoy this email? Forward it to a friend.
Don't blame the "Shy Trump Voters"
I started last week’s TiB with an incorrect prediction: that the US presidential election wouldn’t be close. Although Joe Biden’s popular vote margin will turn out to be impressive, it’s clear that the polls (and I) were very wrong again. As this useful piece shows, you’d have been better off just averaging recent election results than looking at poll averages. This is of more than academic interest: effective governance requires understanding public opinion and if polls can’t help us do that, a lot of things break.
So what went wrong? Importantly, the most frequently cited explanation - “Shy Trump Voters” - people who didn’t want to tell a pollster that they supported Trump because of social desirability bias (SDB) - appears to be wrong. First, Republican congressional candidates did better than Trump relative to their polls - the opposite of what we’d expect if there was a strong SDB against admitting to supporting Trump. Second, the polling error was highest in the most Trump-supporting states - again the opposite of what we’d expect under the “Shy Trump Voter” explanation.
A better explanation is something more fundamental: declining social trust. Voters who have the lowest trust in society are less likely to respond to polls and more likely to vote Republican. Pollsters thought they’d fixed this by weighting samples by education - but the trust problem persists even after controlling for this. If the 2020 election has an intellectual hero, it’s David Shor, who’s been pointing this out for a while. His interview with New York back in July is the single best piece on US politics this year (see also his talk on election analytics here). Polling failure is a big problem - and declining social trust is even worse. 
How to buy an election (... it's harder than it looks)
Another piece of conventional wisdom that’s hard to reconcile with the US election results is the idea that money matters a great deal in American politics. Democrats dramatically outraised Republicans, particularly in key Senate races, but failed to make a dent. In fact, the one place Democrats were outspent was Georgia, where they made some of their most impressive gains. This is less surprising viewed through the lens of political science rather than political commentary. Academics have long been unimpressed by the ability of raw cash to swing election results. 
Partly this is because demography is destiny: there are tight limits to the persuadability of the electorate. Even much of the money that Silicon Valley has poured into politics has focused on traditional levers, like ads, data and organising. But could money, with a little imagination, also shape demographics? We talked back in TiB 23 about the idea of “conspiracies” in the Peter Thiel sense of secret, audacious attempts to achieve change through focused, determined action. It’s fun to speculate about what this might look like.
Given that small states are disproportionately powerful and small margins common in American politics, one idea is simply to pay activists to relocate from big, safe states. Some commentators have already pointed this out. It’s been tried before - see the libertarian Free State Project, focused on New Hampshire - and with bigger budgets could have meaningful impact. It might sound crazy - but as more people on all sides perceive the political stakes to be getting higher, we should expect increasingly radical attempts to work around the system. 
"Scientific Freedom" - the key to progress?
A favourite TiB topic is how to increase the rate of scientific and technological progress, which seems to be in decline (see, e.g., TiB 92, TiB 65, TiB 103) One of the most important books exploring this problem is Donald Braben’s Scientific Freedom, originally published in 2008, but reissued this year by Stripe Press. It’s worth reading, but if you’re short on time, super-blogger José Luis Ricon has an excellent review and points to this report and these slides as good summaries. (Our friends at Interintellect are running a salon on the book - book here)
We asked back in TiB 114, “WTF happened in 1971?” - a year where many economic indicators apparently started to stagnate. Braben’s argument is simple: before 1970(ish), academics largely had freedom; afterwards, they had to compete for grants via peer-reviewed processes. For Braben, the key to scientific progress is to identify the people most likely to have an outsized impact and liberate them (financially and organisationally) to work on whatever they want. Braben actually did this, somewhat successfully, at BP, where he ran the Venture Research Unit throughout the 1980s.
One big challenge for anyone who would wish to replicate this model, of course, is how to identify individuals early enough to know which ones to liberate. I’m actually an optimist on this. As we discussed back in TiB 125, there’s evidence that in mathematics extreme talent shows up very. Closer to my day job, I often reflect that if a VC had made an open ended commitment to fund every medallist in the EU Contest for Young Scientists, they’d have done extremely well (scroll down to the Second Prizes section).  
Quick links
  1. The vaccine zooms off. Very good vaccine news (though not for everyone). And a nice chart on which countries will get it first. And a great short thread on the history of vaccines.
  2. The map is not the territory. Wonderful short piece on the cartography of alternative histories (with a rabbit warren of links to explore…)
  3. High time. This was the drugs liberalisation election for America.
  4. Crypto in the wild. Where is Bitcoin most used for actual payments?
  5. Chips with everything. Interesting piece on a breakthrough in post-Von Neumann chip architecture.
  6. BONUS: How to find a co-founder. Podcast with me, my co-founder Alice, and our friend and investor Reid Hoffman on how making it easier for entrepreneurs to find business partners creates economic opportunity.
How you can help
It’s my birthday, so as a present, I’d love you to share the link to Thoughts in Between on Twitter / Facebook / LinkedIn - or just forward this email to a friend.
I’m always happy to hear from readers. Just reply to this email or message me on Twitter.
Until next week,
Matt Clifford
PS: Lots of newsletters get stuck in Gmail’s Promotions tab. If you find it in there, please help train the algorithm by dragging it to Primary. It makes a big difference.


Did you enjoy this issue?
If you don't want these updates anymore, please unsubscribe here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue