There are two insights about human behaviour that are pretty fundamental to economics:
1. People respond to incentives.
2. Knowledge is limited.
These sound elementary, and they are, but it's often astonishing how easily they're forgotten. Here's an example, carried in yesterday's Financial Times :
Total, the French oil and gas group, and one of the biggest producers in the UK, has become the latest company to warn about the effect of last week’s Budget tax rise on future exploration investment in the North Sea. . . .
Meanwhile, Valiant Petroleum, the explorer, said it had put on hold a £93m project. Centrica, the utility, said it was seeking meetings with the government in order to press its case for gas to be excluded from the tax increase. On Tuesday, Statoil, the Norwegian oil and gas group, said it was postponing $10bn (£6.2bn) worth of investment in two projects after the chancellor raised the supplementary tax on production from 20 per cent to 32 per cent.
People respond to incentives. (Stephen Landsburg thinks that this statement alone  summarizes most of economics.) If government makes the incentives to do something worse then, other things being equal, people will do less of it. So it is with the North Sea oil companies – tax them more, and some of them will leave if there are better opportunities in operating elsewhere. That means less tax revenue overall.
It's not as simple as that, of course, because the precise response is hard to pinpoint. They might cut their profits. Or they might pass on the cost of business to their customers. The BBC's Stephanie Flanders wondered about this too :
I asked Chief Secretary to the Treasury, Danny Alexander, what would stop the oil companies passing on the cost of the extra tax onto consumers, at the pump. His answer, give or take, was that he didn't think they would, but they'd be looking out for it. Hmmm.
Knowledge is limited. It's hard to predict the consequences of any action that we take in our own lives. Now think about a society with millions of people, each with their own goals and preferences (which are constantly shifting). Governments hire armies of economists and other consultants to try to figure this sort of thing out, but they're still naked in the dark. Even small government actions can have big unforeseen consequences, because only individuals know their precise response to a change in circumstances – and even then, they only know in their actions.
People respond to incentives. Knowledge is limited. They're almost self-evidently true, but politicians (and quite a few economists) don't seem to recognise this. We can't make water run uphill and we can't design a chaotic, cloud-like system like society. I wish our rulers would stop trying to do so.