Should people like Richard Murphy be taxed more heavily than the likes of Tim Worstall?

There's a simple reason to think so. The revenue-maximizing tax rate will depend on, among other things, people's attitudes to tax.

If you think taxes are state-sanctioned theft used to line the pockets of idiots, tyrants and con-men you'll be likely to try to avoid tax by retiring, emigrating, exploiting legal loopholes, cooking the books or whatever. This means a high tax rate will in fact be expensive; it will lead to a loss of revenue and GDP and to high compliance costs as governments employ lots of tax men to chase down evaders.

If, however, you think, with Oliver Wendell Holmes, that tax is the price we pay for a civilized society, then you'll be more willing to pay and so a high tax rate will be feasible.

This means the optimal tax rate depends upon tax morale. The happier people are to pay tax, the higher is the optimal tax rate. A society of Murphys, then, would have a higher optimal tax than a society of Worstalls.

The empirical evidence bears this out. A new paper shows that, controlling for income, 'groups with high tax morale systematically face higher average and marginal tax rates.'

Now, I'm not advocating that individuals should be taxed according to their political views; this has obvious drawbacks.

Instead, the point is that the response of tax revenues to tax rates - the shape of the Laffer curve - is not a primitive, fundamental datum. Instead, it depends upon tax morale. A rise in tax morale shifts the Laffer curve right, a fall shifts it left.

For this reason, Tim is not quite right to say that we'll find out whether the 50p tax rate raises revenues when this year's self-assessment returns are analyzed. Instead, the revenue response to higher taxes is subject to the Duhem-Quine problem. It tells us about a joint hypothesis - the interaction of tax rates and tax morale - not a single one.

Let's imagine the numbers were to show that the 50p rate has led to a reduction in tax revenue. The left could reply that this shows not that taxes are too high, but rather that tax morale is too low. Years of neoliberal ideology - and declining trust in government - have increased people's willingness to dodge taxes, they might say.

Conversely, if it appears that the 50p rate has led to increased revenues, libertarians might say that this just shows that people are still duped by statist ideology into a supine acceptance of high taxes.

In other words, we have a problem of multiple belief equilibria. If people believe that high tax rates are acceptable then they will be consistent with high revenues. And if they don't, they won't.

In other words, ideologies can be self-fulfilling. Anti-statist ideology leads to lower tax morale and so the claim that high taxes are damaging becomes self-fulfilling. And conversely, statist ideology raises tax morale and thus optimal tax rates.