The Budget announced an abolition of the higher rates of Air Passenger Duty (APD). Fair enough, given that average rates of APD – a duty which is generally justified in environmental terms – are already higher than the upper estimates of the social cost of carbon. The change announced in the Budget will also result in a simplification, and an elimination of some of the anomalies produced by the current rate structure.

APD rates are meant to be related to the distance flown, which makes sense given that, other things equal, longer flights produce more carbon emissions than shorter flights. The relationship between flight distance and APD payment is not very strong, but this can also be justified because an exact matching would be administratively complex, as it would require a separate APD rate for every airport in the world. The APD system therefore makes two simplifications. Firstly, ‘distance’ is defined not as the distance actually flown, but the distance from London to the capital of the destination country. That, on its own, would still require a separate APD rate for every country in the world except Liechtenstein and North Korea, so the system is simplified further by grouping countries into four distance brackets.

Excessive complexity is thereby avoided, but this comes at the expense of a greatly weakened relationship between distance and APD payment, and systematic distortions. The west coast of the USA is about a thousand miles further away from London than the Caribbean, yet Caribbean destinations fall into the higher band C, while destinations at the American west coast fall into the lower band B. This is because the relatively short distance between London and Washington sets the APD rate for all US destinations. For large countries, the APD rate is determined by what city happens to be their capital. If Porto Alegre became the capital of Brazil, the price of a ticket from the UK to any Brazilian destination would soar.

The removal of bands C and D will put an end to this, but it will only end the most obvious quirks, and at the expense of weakening the relationship between distance and APD payment even further. The overall structure of APD remains illogical and unfit for purpose. If APD was a proper Pigouvian tax, it would have the following features:

– For identical planes with identical load factors, APD payments would be: APD = C + (d*MSC)

C is the ‘fixed social costs’ of a flight, that is, the social cost of the carbon emissions produced during take-off and landing. The variable component is the product of the marginal social cost (MSC) of carbon and the distance flown (d). Total APD payments would thus start from a fixed level and rise linearly with flight distance, while the average APD payment per mile would fall with flight distance. But as the graph below shows, that is not at all how APD currently works, nor how it will work after the reform.



– APD would differentiate between different types of aircrafts according to their carbon intensity, changing incentives so that over time, aircraft fleets becomes less carbon-intensive than they otherwise would have been. This is currently not the case: APD provides no dynamic incentives, as nothing can be gained by increasing a fleet’s efficiency.

– APD would reflect the fact that one full plane produces much less in emissions than two half-filled planes. This is currently not the case either: the passengers in the two half-filled planes would pay exactly the same amount in APD as the passengers in the full plane.

There is a reason why APD has not evolved that way: an APD system that differentiated by aircraft type, aircraft load factor, and the exact distance travelled, would be a lot more accurate than the current system, but it would also be a bureaucratic nightmare. We would soon have to employ more APD administrators than pilots.

But there is a way of making the system more accurate while simplifying it at the same time: levying APD on aircraft fuel, not passengers. That would solve the problems stated above; it would rid the system of cliff-edges and other anomalies, and it would be less complex to administer. The main reason why this has not happened is that it would go against international aviation treaties. But there is no reason why these treaties could not be renegotiated, given that other signatories face similar problems, and given that changing the treaties does not mean increasing the tax burden on aviation (which, again, is already too high in the UK, relative to the modest social cost of flying). It just means shifting the existing tax burden – or, better still, a lower one – to a different source; a source over which airline companies have some degree of control, while they have no control whatsoever over current APD rates. Therefore, in the long run, APD revenue would probably fall under this system, as planes become more fuel-efficient and as aircraft capacity utilisation increases. That, though, is as it should be. A Pigouvian tax should not be a cash cow.

Kristian Niemietz is the author of Depoliticising Airport Expansion.