Great comment in the Guardian from as ever the irrepressible Ross Anderson on an article by Athene Donald. Reality is in the first sentence. [link for this post]7
There’s only one way to do great science, and that’s to hire the brightest people you can find and let them do whatever turns them on. That’s how Bell Labs operated, and it’s how universities like Cambridge used to operate (and still try to). Smart firms like Microsoft and Google try to emulate it in their own research operations.
The UK is screwing up because of increasing micromanagement; because ministers like to stand up and make “announcements”. So Osborne wants a big research centre near Manchester? This is surely unconnected with the fact that he’s an MP for Cheshire; and of course no-one would be so crass as to suggest that Willetts gave Oxford £20m for research in quantum computing because he’s an alumnus. However, as Athene says, the mechanisms used to distribute the research budget should have the confidence of academics. Walport doesn’t.
Were I the legislator, I’d shift some of the funding from places to people. Every year, the thousand research students who produced the best theses would get five years’ postdoctoral funding to go and do their research wherever they wanted. They’d vote with their feet, and ministers would see pretty quickly whether Oxford or Manchester was hot or not.
At present, the Royal Society awards less than 100 junior research fellowships a year, which let lucky young scientists do just that. The EU adds a few more Marie Curie fellowships. These young postdocs are among the most productive we have, because they’re doing what interests them, and they’re much better placed to judge what’s a good investment of effort than decrepit old senior professors (let alone ministers). What’s needed is to scale this up. At steady state, 5000 postdocs at £35k each plus the same again for overheads would be £350m a year or 7.5% of the science budget. What’s more I’d expand it into the arts, humanities and social sciences too.
This would force vice-chancellors to focus on providing an environment in which people can do great research, rather than sucking up to ministers and lobbying for more pork. This should be the natural drift of policy for conservative or liberal ministers, as it would align incentives somewhat better than at present.
I find Anderson disturbing — not because he is wrong— but because he is right about so many things.
Computer security, public health and medical education
There is an interesting article by Cory Doctorow in the Guardian, in which he draws parallels between public health and computer security. It is worth reading in conjunction with some of Bruce Schneier’s stuff (the security guru, as the Economist calls him). If only medical education was a little more agile, this topic would form a great module for some students. I suspect however that students just get a ‘professionalism fix’ on using encrypted USB drives on NHS machines (yes, those ones running IE6). We are missing the chance to talk about big issues: the apparent data breaches by the English NHS (see letter from Ross Anderson and others here); and the inability of the Wellcome Trust—amongst others— to understand the limits to anonymisation, nor the fact that research does not trump all other values. So students, if you wanted to hack medical information, whether paper or digital, how would you go about doing it? I suspect students would find such an approach interesting, and those running the NHS might learn something too (I am not of course suggesting they try to breach security, merely that they are forced to think about some of the tradeoffs involved —scale, security, ease of use etc). For all of this to fly, we need genuine ‘core’ and ‘options’, something that seems as likely as a system immune to hacking. And we need to educate them, so that they do not think data security is something they get told about in a FY1 induction pack. (And of course, we might ask them to appraise some of Doctorow’s metaphors re typhus, cholera and the importance of water in disease spread)