According to The Times, Rishi Sunak is considering a proposal for a UK-wide carbon tax. The Chancellor can see two advantages to such a scheme. It would help towards the goal of zero net emissions by 2050 while also raising substantial revenue (a figure of £27 billion by 2030 is quoted), thus reducing the amount by which other taxes, particularly income tax, would have to rise to offset the massive coronavirus spend.

A carbon tax is an economist’s favourite idea for mitigating climate change, plugged once again this week by Professor Dieter Helm and others in a very interesting Think Tent discussion at the virtual Conservative Party Conference. But I’m not so sure that it would be as attractive a proposition in practice as Mr Sunak’s advisers may be suggesting.

It is a special case of a Pigouvian tax, named after A. C. Pigou, a Cambridge contemporary of Keynes. Pigou was concerned with the way in which ‘negative externalities’, for example those caused by pollution, can mean that the output of a factory exceeds the ‘socially optimal’ level, while the consumer pays less than the ‘socially  optimal’ price. As generations of A level students will recall, a standard textbook diagram shows that the problem is that marginal social costs exceed marginal private costs. Setting a tax equal to the marginal external cost will in principle resolve the problem.

I say ‘in principle’. Because governments have no idea what the marginal private costs and social costs of hundreds of the activities of hundreds of thousands of businesses are. They are in no position to set a tax equal to marginal external costs. To use a favourite term of the late Ronald Coase, this is ‘blackboard economics’ – it is a theoretical rather than a practical solution.

We can certainly pluck a plausible-sounding figure out of the air: the Treasury seems to favour £75 per tonne of CO₂ emissions. But this is an average rather than a marginal figure, and something of a blunderbuss. Moreover, as Jamie Whyte pointed out in the Think Tent discussion, existing taxes on petrol and diesel already exceed this level. Does this mean they should be reduced? Probably not, if the government is interested in tax revenue.

In fact, there are a very large number of taxes and subsidies (for example, zero duty on aircraft fuel, the pensioners’ winter fuel allowance, controls on energy prices) which distort incentives and impact on the level of carbon emissions. While scrapping all of these and just having a carbon tax might well be an improvement, this is not what the Chancellor seems to have in mind.

Carbon taxes have been tried. Australia had one from 2012 to 2014, when Tony Abbott’s incoming government scrapped it. An assessment by IEA author Alex Robson found that there was little benefit in terms of either emission reductions or to the country’s fiscal position. For one thing, strong external competition meant that many heavy energy-using businesses could not pass the tax on to consumers, so consumption was not reduced – although some business may have been lost to overseas producers. For another, the burden of higher electricity prices fell disproportionately on poorer people, so the personal income tax system was adjusted to partially offset this, reducing any net increase in tax revenue.

An alternative to a carbon tax is the ‘cap and trade’ system, which draws inspiration from Coase’s classic paper ‘The Problem of Social Cost’. Coase argues in this paper that many externality problems can be resolved by markets if property rights are clear and unambiguous, and transactions costs are low. The ‘cap and trade’ policy in effect creates a property right in carbon emissions, a permit allocated by the government. Businesses are then free to trade permits; those which can reduce emissions below their allocation can sell their remaining emission rights to new businesses or those existing businesses which continue to produce excessive emissions.

Over time, government reduces the cap on emissions and thus moves closer to the zero net emissions target. At any one time, the discovery process of the market generates an implicit price of carbon emissions.

The European Union Emissions Trading System is the best known of such cap-and-trade systems: the UK has long been a member. The EU scheme has had many problems with setting caps and with the coverage of the scheme. Fraud has been a not insignificant problem. A recent assessment suggests it has had some positive results, although it admits it is difficult to disentangle the effects of the scheme from other factors operating on the European and world economy. The biggest effect seems to have been on the electricity sector.

Brexit offers an opportunity to switch to a carbon tax instead. Would it be any better? If the Chancellor chooses to go down this route, he needs to be aware that a carbon tax is far from a perfect answer to his problems. Its introduction should also be accompanied by efforts to tackle the complex web of taxes and subsidies which currently distort efforts to reduce emissions; this will inevitably generate opposition by vested interests. He must also be aware that a carbon tax which bites will be passed on to consumers, including some disadvantaged groups. On past experience there will be political pressure for fiscal compensation, reducing the gains to the Exchequer.

There is also danger in trying to meet two policy objectives, emission reduction and tax revenue, with one instrument. It is not difficult to imagine a level of tax which severely reduces emissions but as a consequence reduces the tax paid to the Exchequer – a sort of carbon tax Laffer curve effect. In those circumstances would there be a temptation to reduce the tax to boost revenue? I wouldn’t bet against it.

On the other hand, if the government seriously wants to make a significant contribution to the global effort to reduce carbon emissions, even a very high UK-based carbon emission tax would not be enough.