Bravo to IainB for the above discussion. I'd like to extend it a bit farther, if I might.
Some factions in our culture take the position that scientific results should directly drive public policy. So, for example, a finding that income inequality has increased, or that obesity is a health threat, should
automatically lead us to institute policy to combat that problem.
But this belief is mistaken. Identifying a problem, or finding a correlation, is science. Determining whether something ought to be done, and indeed, whether that action should be executed by the government, is another question entirely, and one that cannot be answered by that same science. Before we can make that determination, we need to consider (I don't mean to argue on either side of these issues, I just mean to show that there are questions that must be answered before action is appropriate):
- Is it actually a problem that society as a whole should have any say in? This generally is determined by personal values, and so reasonable people are likely to have different opinions. For example, given that obesity threatens my health, is it not my personal decision whether I prefer to have a shorter life of gustatory pleasure, or a longer ascetic life? Thus, the scientific outcome doesn't automatically mean that something *must* be done.
- Would the costs of doing something exceed the costs of the problem itself? In the climate change debate, for example, we hear a lot of bickering about the evidence. But I don't see so much cost/benefit discussion about the likelihood of various outcomes, and the actual human cost of each, especially discussions that directly compare the cost of implementing greenhouse gas reductions.
- Is the government the right agent to affect the change? Forcing everyone to act a certain way is a very blunt tool to use, and thus possibly ineffective. Even when the government can do something, it may be that the best approach is just to set up an appropriate system of incentives so that the private sector can work out the details.
Thus, a statement of the form "scientific studies show X, therefore the government must implement regulation Y" are flawed.
And I'd like to take that a step farther, too. In political discussions we frequently hear things like "candidate X isn't intelligent enough to be President", or even outright name-calling intended to disparage a candidate's intelligence (e.g., "dumbya"). I submit that science is the job of the scientists, and not the job of the President. There's no need for the President to understand biochemistry or chaos theory or orbital mechanics; as the country's Chief Executive, the holder of that office needs to be able to
execute, and that involves being a good manager: knowing how to find the best people to handle an issue, and delegating to them. Beyond a basic threshold, raw IQ points aren't what we need in our Chief Exec, we need a specific management skillset, one that has nothing to do with the sciences.