Nate Silver
TALKING POLITICS
Catherine Carr
4.7 • 2.5K Ratings
🗓️ 18 April 2020
⏱️ 42 minutes
🧾️ Download transcript
Summary
David and Helen talk to 538's Nate Silver about how to read the pandemic data and what they mean for politics. What do we know now that we didn't know six weeks ago? How should we model the future trajectory of the disease? Where does it leave the election in November? A conversation about everything from death rates to spring breaks, and from Belgium to Biden.
Talking Points:
Are the COVID models we are using now better than they were before?
- People don’t always understand the conditional predictions behind different models.
- There is still a lot of uncertainty: almost every parameter of this disease remains unknown.
- It looks like the downward tail of this disease might be less steep than the upward tail.
What is the true fatality rate?
- We are still behind on testing—we don’t know how many cases are undiagnosed.
- This number is important for making decisions about opening things up again.
- The number of deaths as a share of the population probably tells you more than deaths as a share of cases.
- A lot of deaths are still being missed.
Places that had cases earlier before there was consensus on social distancing will have worse outbreaks.
- Factors such as age distribution and maybe even weather might also affect things.
- It’s still hard to tease out the effect of different variables, but eventually we should be able to make some better inferences.
International coverage of the US doesn’t reflect how empowered state and local governments are. In some sense, they are the most important units.
- Despite the lack of federal response, state and local responses have been fairly good—at least in a lot of places.
- Most countries locked down at roughly the same point in the disease cycle. The country by country differences may be more felt in the recovery phase.
How will the pandemic affect the upcoming U.S. election?
- Trump’s approval rating improved slightly, but only slightly. It’s a smaller ‘rally around the flag’ effect than in other countries.
- Will the fall bring recovery or a second wave?
- Will the contingencies or the fundamentals explain the outcome?
Mentioned in this Episode:
Further Learning:
- 538 on why it’s so hard to make a good COVID-19 model
- And on the American urban/rural divide
- The Atlantic on the upcoming pandemic summer
And as ever, recommended reading curated by our friends at the LRB can be found here: lrb.co.uk/talking
Transcript
Click on a timestamp to play from that location
| 0:00.0 | Hello, my name is David Ronsman and this is Talking Politics. This extra episode is with |
| 0:08.7 | Nate Silver from 538, the guru of database journalism. He's here to explain the data |
| 0:15.4 | behind the pandemic, the data behind the presidential election and the relationship between the two. |
| 0:23.5 | Talking Politics is brought to you in partnership with the London Review of Books, Europe's |
| 0:27.6 | leading magazine of culture and ideas. Improve the quality of your solitude with a subscription |
| 0:34.2 | to the LRB. They'll send you exceptional analysis of the politics, economics, sociology |
| 0:41.2 | and science behind the crisis and reportage from around the world. But also, gloriously |
| 0:48.0 | unrelated, richly immersive distraction from the world's best authors and critics, writing |
| 0:53.5 | about history and philosophy, art and technology, fiction and poetry. Just go to lrb.me-talk |
| 1:02.7 | and get your first 12 issues for just 12 pounds. That's lrb.me-talk. |
| 1:13.0 | Hello and I spoke to Nate on Thursday evening. Nate is out of the city because of the lockdown, |
| 1:18.5 | so I'm afraid this is not quite in studio quality. We're going to get on to the presidential |
| 1:23.8 | election but we started inevitably with the pandemic. Nate, obviously there's still a huge |
| 1:31.0 | amount that we don't know about the shape of this pandemic. But we know a lot more than |
| 1:36.5 | we did say six weeks ago. Just give us a sense from your point of view. How do the models |
| 1:40.8 | that we have now compare to the models that were being used in the very early days of this? |
| 1:45.5 | Are they noticeably better? It's a difficult question for two or three reasons. I think when |
| 1:51.3 | reason is that people tend to mistake models that assume that nothing was done, that no |
| 1:57.1 | interventions were taken, versus models that tried to account for and predict the impact |
| 2:01.5 | of social distancing. So the kind of famous Imperial College predictions that involved more |
| 2:08.3 | than a million people dying in the UK, I believe, are close to that and I think two million |
| 2:12.0 | in the United States, that assume that if the disease is totally unchecked until you |
... |
Please login to see the full transcript.
Disclaimer: The podcast and artwork embedded on this page are from Catherine Carr, and are the property of its owner and not affiliated with or endorsed by Tapesearch.
Generated transcripts are the property of Catherine Carr and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.
Copyright © Tapesearch 2026.