How far should we rely on science to make political decisions? What makes a good science advisor — or a good science advice system? What do we do when the evidence is incomplete or controversial? What happens when science advice goes wrong and how can we fix it?
We explore these questions, and many more, in conversation with the researchers, policymakers and communicators who make science advice happen around the world.
The Science for Policy podcast is produced by the Scientific Advice Mechanism, but the varied and interesting opinions expressed by our guests are those of the guests themselves, and don’t necessarily represent our views or those of the European Commission.
Podcast episodes take the form of a friendly one-on-one conversation. Don’t think of it like a conference presentation. Think of it more like an informal interview with a journalist — but one where you can decide what to talk about. You can choose the topic and we agree the questions in advance.
We’ll have the conversation over the internet using Zoom (or another platform if you prefer). It will be recorded, then edited and later published as an episode of the podcast.
The majority of our audience divides into three groups:
At the time of writing, we average about 1200 listeners per episode, but this really varies depending on topic and time of year. So it’s a fairly niche audience, but that’s OK — it’s also a fairly niche topic!
We cover a wide range of topics on the podcast. The best way to get a feel for it is to look through our list of previous episodes.
The two main rules are that the topic needs to be related to science advice somehow (obviously), and something you can speak about with authority. But both of these can be flexibly interpreted. People who are interested in science advice might also be interested in science communication, science diplomacy, policymaking, philosophy of science, science and politics, and many other linked topics. Don’t feel too constrained. If you find it interesting, and it’s linked to science for policy, it’s probably OK.
Following these guidelines, most episodes fall into one of three categories:
There is also a fourth category of episode which we publish occasionally, but it’s much harder to pull off:
There are three things we definitely don’t do:
The final podcast episode will be 30–45 minutes long. But the recording process will take 60–90 minutes. This allows time for us to get set up, warm up, troubleshoot any last-minute technical issues, and record the actual interview.
The style of the podcast is a natural, free-flowing conversation, so we don’t really edit content unless you ask us to, or unless we accidentally go way over time. Occasionally, if our conversation is very twisty and turny, we might rearrange some of the questions and answers to make a more coherent structure for the listener. Other than that, we will mostly edit just for audio quality (‘mastering’), and to cut out long pauses, speech mistakes, technical errors and unexpected noises.
Yes, absolutely. It works best if we can have a quick chat a few days before the recording to agree a topic and a rough outline of what you’d like to cover, so we have a shared understanding. We will then draft some bullet points which we can both refer to as a guide during the conversation. Don’t write yourself a full script, though — it’s supposed to be an informal conversation, not a formal presentation where you read from your notes.
Also, remember that the conversation is recorded, not live. It doesn’t matter if you need to stop and think, or want to have a second try at answering a question. We can edit all that out afterwards. If you’re uncomfortable with anything, we can easily re-record that part, or just cut it if you prefer!
Yes and no.
This isn’t a journalistic interview, and we want you to be happy with how you come across. Not only do we agree the questions with you in advance, but during or immediately after the recording, if there’s anything you said that you regret or would rather not include, no problem — we will simply re-record that part, or cut it.
Once the recording process is over and we have edited the interview into its final episode format (usually a few days before it gets published), you can’t make further changes. You won’t get to re-check the final recording before it goes out. This isn’t so much an issue of principle as an issue of workflow: the editing process is quite involved, and we don’t have capacity to re-edit based on guest feedback.
We have three secret weapons to keep things running smoothly:
Normally at least a few weeks, often a month or two if there are already several recorded episodes in the queue.
We are also quite careful with our programming: we like to prioritise very current topics, we don’t like to publish multiple episodes in a row on similar subjects, and we try to maintain a reasonable gender and geographical balance across multiple episodes. For all these reasons, the delay between recording and publishing might be longer than you’d expect.