Saturday, January 23, 2010

On the democracy of science

Should science obey to democratic principles? This is a question that I recently saw discussed on Scienceblogs, and that is also mentioned in the latest strip of PHD Comics.

Democracy is about letting people decide on their fate by choosing between objectives. Concretely, this has translated to choosing between policies, but that is another question. There is, however, a disturbing trend in the media to let people vote ("reader polls") on things they have no business deciding on and then present this as evidence. "Is this person guilty?" Leave that to the courts who have the full information and know the laws. "Is the theory of evolution true"? Leave that to scientist to figure out. "Is climate change real?" Idem. "Should the Feds bail out banks?" Leave that to economists. Etc.

While science is built on consensus, it does not mean consensus among people who do not know what they are talking about. The wisdom of crowds is often wrong. Even among economists, conventional intuition is sometimes wrong, this is what makes great papers in top journals. Instead of turning to the opinion of the people in the street to fill airtime, why not actually ask experts. And by experts, I do not mean political pundits.


Agent Continuum said...

True, but you also need to ensure that the scientists have the right incentives, if they're to be independent and the final authority on things.

K----- said...

I couldn't agree more about the need to remove some of the perverse incentives present in academic science. Moreover, the scientist I work with are usually atrocious at knowing where their expertise ends and policy begins. Couple that with big egos ("I'm brilliant, so my opinions on XYZ must be correct"), and things get ugly fast.