The cash for Cameron affair, says Hopi, reignites the arguments for state funding of political parties. The economist in me finds the case ambiguous.

The case for state funding lies in the fact that private funding of parties is prone to a market failure identified by Robert Frank - that it produces too much competition.

A party's spending on an election campaign is a positional good, in that its payoff depends not on the absolute sum spent, but upon how much it spends relative to its rival; spending ''20m succeeds (ceteris paribus) if your opponent spends ''15m, but fails if it spends ''25m.

This rank-dependence means that party funding becomes like an arms race, in which both parties compete too hard to raise funds, with socially sub-optimal consequences such as the danger of corruption. As Frank writes:

The dependence of reward on rank eliminates any presumption of harmony between individuals and collective interests, and with it, the foundation of the libertarian's case for a completely unfettered market. (The Darwin Economy, p11)

There's an analogy with drugs in sport. Under free competition, every athlete would have an incentive to take drugs. But if everyone does so, each individual's chances of winning doesn't change. All that does change is that athletes jeopardize their lives.

A cap on party donations works like a drugs ban in sport. It removes the unhealthy element of competition. The case for state funding is that the tax-payer must make up the lost money.

There are, however, three other considerations:

1. State funding would change politicians' incentives. The Telegraph says that 'forcing taxpayers to pay for party politics risks making politicians lazy and out of touch.' This, though, is not clear. If parties had to appeal for many small donations rather than a few large ones, they might actually try harder to stay in touch with the public.

Not that this is necessarily a good thing. If you take the Burkean view that politicians owe us not just their industry but their independent judgment, it might be no bad thing if they become out of touch with the populist mob.

The only point here is that incentives would change, and we should try to think how.

2. Political parties are not so much a public good as a public bad. They tend to limit the frame of debate - squeezing out worthwhile positions such as Marxism, libertarianism or anti-managerialism from public discourse. They encourage a narrow-minded tribalism. And they provide a career path for would-be MPs to become researchers and special advisors, thus encouraging the creation of a political class separate from the rest of society. These are negative externalities that should not be promoted with tax-payers' money.

3. A legal cap on donations is not the only way of preventing an arms race. As we saw on Sunday, a free, investigative press can also police party funding.

The economic case for state funding of parties is, therefore, not clear.

But the economic case is not all there is. There's also the matter of freedom. This speaks against capping donations - a man should be free to donate to a party, as long as he's not buying policy - and against using the tax-payers' money. I'm not sure if any argument is powerful enough to trump this.