An eminent group of American economists, including all living former heads of the Federal Reserve, has called for a carbon tax. Despite a growing global scepticism about any recommendations from the economics profession, the proposal deserves serious consideration.

The central idea is that the production of goods and services that cause carbon emissions should result in a tax payment to the government. Use a megawatt hour of electricity and your company or household will pay a price that includes a fee related to the amount of CO2 released to the atmosphere as a result of the production of this power. A company producing a tonne of steel and will owe a carbon tax related to the emissions from the coal used to make it.

A carbon tax therefore raises the price of goods and services that have burnt fossil fuels in their production. The US economists behind the latest proposal suggest that the revenue raised from the tax is then entirely redistributed back to individuals as a flat rebate. Each person, for example, might get $200 annually as his or her share of the total funds raised by the tax.

Almost all economists like carbon taxes. This mechanism avoids the need for many regulations and disruptive market interventions. Taxes push both producers and consumers towards low carbon ways of providing goods and services without absolutely obliging anybody to change their behaviour.

Electorates are more equivocal. The ‘gilets jaunes’ movement in France sprang into existence partly as a reaction to the rise in the price of vehicle fuels as a consequence of an increased carbon tax. Badly designed levies penalise the less well-off because poorer households may spend a larger portion of their income on energy-intensive purchases. The people of the Canadian province of Alberta were not happy when a carbon tax was introduced there in 2017. Opinions are still very divided even though the average cost to a single person earning under C$95,000 is calculated by the government to be C$286 but the yearly rebate slightly more at C$300.[1]

The UK needs to start discussing whether a carbon tax would work here. The discussion should particularlexamine how a carbon tax could be designed so that it doesn’t affect poorer people adversely.

This article looks at the patterns of household expenditure in the UK and assesses how a government could make sure that the net effect of a carbon tax might be broadly redistributive towards the less well-off. This is an important issue: households in the top 10% of expenditure spend 9% of their income on high carbon goods and services while those in the lowest decile devote over 15% to these products. However, if a 10% carbon tax were to be employed, and the proceeds then redistributed equally to households, the highest spending group would lose about £190 a year and the lowest would gain about £160.

I discuss more complex - and much more obviously redistributive - schemes in the note below but I tentatively conclude that a simple tax, working similarly to VAT, would probably work best. The taxation income then needs to be handed back to households as an equal lump sum for all.

Patterns of household expenditure

I use the Living Costs and Food Survey (LCF), a long established and well-regarded annual report on how a very large sample of UK households spend their money. This survey divides households into ten deciles (10% groups), ranging from the lowest to the highest spending. The decile spending most had a weekly expenditure of £1222 in the 2018 survey, which is almost six times as great as the lowest decile (£213). Richer households tend to include more people but the expenditure differences per person are still very marked.

In summary, analysis of the LCF shows that the poorest decile spends more of its income on items I have defined as typically ‘high carbon’. A flat rate carbon tax would therefore hit poorer households harder and the article assesses how it might be possible to avoid this politically suicidal problem.

I look at four types of expenditure

a) Energy for the home

b) Purchase of meat and meat products for eating

c) Petrol and diesel for running a car or a motor bike

d) Air tickets

Energy for the home

Most households buy gas and electricity. Some only buy electricity and heat their home with it. Others use electricity and another fuel for heat, such as oil.

This is the pattern for how much homes spend on domestic energy per week. The left hand axis is the number in pounds and on the right hand side this figure is expressed as a percentage of the household’s total expenditure.

Chart 1

Living Costs and Food Survey for 2017-18, published January 2019

This chart shows that the amount households spend on domestic energy is only weakly related to income. The homes in deciles 3 to 9 typically pay between £20 and £25 per week for their power and heat. The lowest spending households do spend less than richer homes but the difference is not especially marked.

As a percentage of income, the bottom decile spends over 8% of its income on energy but the richest group spends just 2.6%.

This means that any carbon tax needs to be designed with particular care. Adding 20% to gas and electricity bills would cost the lowest spending decile almost £3.50 a week, or over 1.5% of their expenditure. The richest 10% would lose just half a percent of their expenditure.

The key question is: how could consumption of household electricity be taxed in a way that reduced consumption but did not particularly affect the less well-off?

In parts of the US, for example, electricity gets more expensive the more the household consumes. The first few thousand kilowatt hours per year are priced at one level but as consumption rises, the price per unit increases. (There are no carbon taxes involved of course). One way of discouraging high levels of consumption of gas and electricity would be to have a pricing schedule in the UK and elsewhere that got more expensive as usage increased. This would mean that richer people, who generally live in larger houses, would typically pay much more for their power and gas. And if the carbon tax were imposed as a percentage of the price of the energy, it would disproportionately affect the better off. Theoretically, the tax could be made progressive.

A pricing structure for energy that obliges suppliers to charge customers more for larger amounts supplied would be extremely difficult within the UK’s energy market. Suppliers would target customers using more energy. (In the US, rising rates happen in markets with just one supplier, usually either publicly owned or heavily regulated).

The obvious alternative is to impose a rising percentage carbon tax on domestic bills. Users of more energy would pay an increasing proportion of their bill as tax. So for example, the first 2000 kWh of electricity might cost 14 pence each, with no carbon tax, but then the costs rise by 2p per kilowatt hour for each extra 1000kWh. (The average domestic user in the UK not using electricity for heating uses about 3,000kWh). The same ideas would apply to gas as well.

A customer buying 100% renewable electricity, or low carbon gas would not see this cost escalation. This would encourage the purchase, and supply, of renewables and nuclear power. It would boost the production of biomethane from anaerobic digestion and other sources.

Such a tax would raise the incremental cost of energy - possibly sharply - for non-renewable customers and really encourage efficiency.

What are the problems with such a pricing structure? First, it is a blunt tool from an equity point of view. Some less well-off people have large houses and would be penalised. Or they live off the gas grid and so have to use substantial amounts of electricity for heating. Such households would disproportionately suffer from the high price of electricity for larger consumers. Of course, the obvious choice would then be to buy renewable electricity to avoid the higher costs for heavy users. This might very substantially incentivise the development of new renewable sources.

A tax that increased as domestic usage rose could be broadly progressive. The revenues generated by the carbon tax could then be distributed on a flat per capita basis, meaning that lower income households might be net beneficiaries.

b) Purchase of meat and meat products for eating

I don’t know of anywhere that imposes a carbon tax on foods. But, logically, the world probably should try to use taxation to reduce the consumption of high carbon foods, particularly those from ruminant animals such as sheep and cows. (The recent Lancet report on improving the global diet put the share of all food production in total carbon emissions at 30%. Others might be happier with a figure nearer 20% but there’s no doubt that agriculture really matters in the contest to limit climate change).

I looked at the pattern of expenditure on meat and meat products. The Living Costs and Food data shows that there is a steeper gradient to this item; richer households do spend more but the percentage of their income expended on meat is lower than for less well-off groups. Although the bottom decile spends about £6 and the top almost £20, the percentage of expenditure falls from about 3% to 1.6%.

Chart 2

Perhaps it is not appropriate to tax all meat products; the climate change impact of beef and lamb is disproportionate and it may be logical to focus on these foods. The excellent Living Costs and Food survey doesn’t give us perfect data. We can track purchases of the meats themselves but we cannot know how much of a line called ‘meat products’ contains beef or lamb. Nevertheless, I thought it would be helpful to show the chart of beef and lamb purchases

Chart 3

The amounts spent rise slightly more sharply with income and, as a consequence, the share of income spent on beef and lamb is flatter across the deciles.

Personally I think we should investigate the possibility of adding VAT (20% value added tax) on all products containing beef, lamb or the meats of other ruminants, such as deer and goats. Logically, it also makes sense from the point of view of carbon avoidance to tax milk and milk products from ruminants. VAT is already generally charged on meats sold in restaurants.

This policy would cause outrage that any such policy would cause among farming groups and many individual households, even though producers of other meats, such as pork and chicken, would probably benefit. Nevertheless, I think VAT on foods with a very high carbon impact makes very good sense.

Assuming that about 50% of all items classed as meat or meat products in the consumer survey actually contain cow or sheep products, a rich household will pay a tax of about £100 a year and one in the poorest decile will see a £30 surcharge. If all the carbon taxes from meat were then given back to households, the average rebate would be about £60. Poorer households would be net beneficiaries.

c) Petrol and diesel for running a car or motor bike

Motor fuels are already heavily taxed. In the UK, duty and VAT take more than half of the price of petrol and diesel at the pumps.

The percentage of total expenditure spent on motor fuels rises across the deciles, with an unusual exception of the highest income households. Although this group spends more in cash terms than the next decile down, the percentage of income spent nevertheless falls because the total expenditure of the richest families is almost 50% greater than the second decile.

Chart 4

This pattern of expenditure makes it possible, in theory, to increase the duty (the old fashioned word used in the UK for the tax on petrol and diesel) on fuels without increasing income inequality across society.

A carbon tax of, say, 15p per litre on motor fuels would raise the average household’s expenditure on motor fuels by just over £100. If the tax system rebated this equally across all households, the average home in at least the lower four deciles would see a net gain.

But, as the French found three months ago, some people with relatively low incomes, particularly those living in areas with poor public transport, spend a much larger proportion of their total expenditure on driving a car. They would therefore disproportionately suffer from a tax increase and a £100 rebate would not be sufficient to recoup their costs. Very, very roughly any household having to travel more than 11,000 miles/18,000 km a year in a small modern car would see a net cost from a 15p/litre tax.

I cannot think of an easy way to make a carbon tax on motor fuels easy to implement politically. Some people will suffer heavily from a duty imposed on fuels. It can be argued that governments should invest in better public transport in rural areas or increase the subsidies for electric cars but neither of these measures will work rapidly enough to avoid the cash costs to poorer households necessarily using a car for high mileage.

d) Air tickets

The situation is much easier when we come to look at air travel. The Living Costs and Food survey does not separately list the expenditure on ticket purchases but the ONS very kindly provided me with the data for 2017. This shows a very steep increase in money spent on air travel as household expenditures rise. Households in the top decile spend almost £1,000 a year on tickets, and the bottom group spends less than a tenth of this figure. So a tax on air travel will disproportionately affect the rich.

Chart 5

Increasing air ticket prices by an average of 10% would produce a sum equal to about £35 a household compared to the typical cost across the bottom three deciles of about £12. A tax would therefore have a positive effect on income inequality.

International rules forbid the levying of taxes on fuel for international air travel. So any tax cannot directly be on the carbon emissions from burning the aviation fuel. However a carbon tax can certainly be levied on the price of the ticket, as indeed it already is in the form of air passenger duty (APD).

The simplest way of attempting to decrease air travel is to increase the cost of flying through a higher rate of tax. But a case can be made that a better route forward might be to allow each person one or two flights a year and then to increase substantially the cost of extra travel beyond these flights. The idea is that people going on holiday to Spain once a year shouldn’t be burdened with extra tax.

Others have already suggested a voucher scheme to achieve this. Each individual would get the rights (presumably expressed in digital form) to perhaps a couple of flights a year and would be able to sell these rights to others for cash. Just under 50% of UK adults do not take a single flight in any twelve month period and these people would be able to cash in their vouchers and make some money from not contributing to the destruction of the planet.

The value of these vouchers on the open market would be heavily influenced by the volumes issued. If more vouchers were issued than flights taken, then the value would zero, or close to it. But if the number of vouchers were adjusted to achieve a price of, say, £20 a voucher then people taking no air flights might gain a reasonable benefit.

Summary

A carbon tax that increased the price of each of these four categories of goods and services by 10% would increase inequality if the revenue was simply added to the existing government budget. 15% of the average expenditure of the least well-off decile % goes on high carbon goods, of which much is spent on home energy. This compares to a total of 9% of the richest households.

Chart 6

However if all the revenue was simply handed back equally to households, perhaps by something as simple as a reduction in council taxes, carbon tax could become redistributive. Speaking personally, I prefer this outcome to the more complex ideas I briefly discussed above.

As the chart below shows, the richest decile households spend over £5,000 a year on the four categories I discuss. The least well-off decile spends about £1,600. A 10% tax on the four categories would add about £165 to yearly bills in the bottom expenditure decile but £517 to the richest group. If the money raised were then redistributed equally, each household would get about £329 a year back. This means that the bottom four deciles would gain income, the next two see a roughly neutral result and the richest four deciles suffer an income loss.

Chart 7

However if all taxes were rebated to households as a lump sum, perhaps as a reduction in council tax, then the net impact would be redistributive. Across all expenditure deciles, the average impact of a 10% rise in the price of the high carbon goods specified here would add £320 to typical bills. But lower expenditure groups would face a smaller absolute increase in their bills. The bottom group will be asked to pay an average of £160 in carbon tax. A rebate of £320 would therefore save them money as shown in the chart below.

Chart 8

Living Costs and Food Survey for 2017-18, published January 2019. Calculations by the author.

It seems to me that a relatively simple scheme like this could work. But it doesn’t solve the problem of poorer households with high energy bills or which need to travel long distances by car. People in these groups cannot easily be protected. All the more reason to plan now for major improvements in public transport and better insulation in homes. To be slightly more specific, a fair carbon tax must be accompanied by transport improvements that benefit the less well-off, and not simply make it easier to get to the capital from other cities (I am referring to HS2, the vanity project that provides a new London to Birmingham link). It must also target substantial and inexpensive improvements to the quality of the UK building stock rather than the ill-designed schemes that governments have been playing with over the last decades.

[1] https://www.alberta.ca/climate-carbon-pricing.aspx

[2] Council tax in the UK is a highly regressive property tax