Funding people, not projects April 23, 2009Posted by Sarah in politics, science.
Tags: astronomy, funding, netherlands, policy, politics
Today I attended a talk by Prof. Cornelis van Bochove, who was appointed as Professor of Science Policy Studies at Leiden University in February last year. Van Bochove has had an interesting career: after a number of years in econometrics research, he became Director of the Dutch Central Bureau for Statistics for 5 years until 1999, after that Director of Research and Science Policy at the Ministry of Education, Culture and Science until 2007. In those years he apparently always showed a keen interest in astronomy and was a strong supporter of the Dutch astronomy community, which has a long history of international excellence. Rather than hang up his hat, van Bochove joined Leiden to go back to the bench. The focus of his research is “evidence-based science policy”. So he’s looking at the science behind funding science. A bit of a brain twister, I know. But the talk turned out to be very interesting, and a little bit surreal.
He presented a quasi-scientific model of a country’s research output, starting off with a bunch of scientists of a given talent distribution, a total budget, a learning profile per researcher, a rate of attrrition, a tenure age, a percentage of scientists who will get a permanent job, and some random noise, and poof! Here’s how you maximise your science output. Apparently, “science output is very simple: you just add up your innate talent and what you’ve learnt,” he said. Needless to say, we all felt a little skeptical. Until he started discussing the results.
It was quite bewildering seeing our lives in science broken down into some variables and differential equations. The “random noise” in the model is the crap boyfriend you had at 21 who sucked the life out of you, or a bout of glandular fever that knocked you out for a while, not to mention random luck and coincidences. But in the end, what this model with its crude indicators seemed to show was that scientific output of a country can be maximised by offering a relatively small subset of your scientist-population permanent positions at an ealy stage, say, around age 30 or so. This didn’t seem to be connected to any other variables, like salary jumps at tenure, or how the talent is distributed in the population. The result goes against the notion seemingly held in some policy circles that to spot the best scientists, we need to observe their work for a long period of time before we can judge their talent. No, van Bochove says, the most talented researchers can be spotted at an early age, and we should hire them into permanent positions quickly.
Policy makers need to ensure that the most talented scientists don’t spend their most productive years in project-based positions where they’re more preoccupied with securing their next post than doing creative and ground-breaking science. The message seems to be that investing in people gives a better return than investing in projects. Van Bochove is a strong proponent of this approach and was apparently a major force in the introduction of the excellent Veni, Vidi, Vici fellowship programme in the Netherlands.
The model was nice but of course I suspect such an over-simplified appraoch can give any result you want it to. But then it’s not like there is a host of rigorous models out there telling governments that more projects-based funding is the way to increasing the country’s scientific productivity, and yet that’s the direction many seem to be taking. If it has even the smallest chance of making polticians sit up and listen, this model is just fine by me.