These short essays should not be read as explicit policy recommendations, but rather as a set of provocations in terms of what is possible.

Almost all economists would subscribe to two fundamental principles of taxation. 

The first is that the tax system should be straightforward, unambiguous and devoid of loopholes. The rate of tax levied on a gingerbread man should not change because chocolate buttons are added, nor should a deer-skin belt’s tax status be dependent on whether the skin has been tanned. A tax system that fails this test is problematic for two reasons. It distorts the real economy, as people put effort into making products that are desirable only because they benefit from a tax exemption, rather than because they satisfy an underlying need. In addition, a complex tax system is expensive to operate, both for the government and for taxpayers. 

The second is that it is right and proper to have a tax system that distorts choice when the government wants to distort choices and where taxes are an efficient way to achieve this distortion. Thus, for example, economists would generally approve of taxes on cigarettes, alcohol and so on. These taxes induce people to purchase less products that have externalities. 

As we will see, a wealth tax may defy both of these principles. It is relatively expensive to collect, and would distort behaviour away from saving, and towards spending. The former is unambiguously bad, the latter depends on whether we think that people are saving too little for the good of the overall economy. 

Royal Commission on Taxation 1951–55 rejected a wealth tax. The Labour Party promised to introduce an annual wealth tax in their 1974 manifesto and issued a green paper once in government. The newly created Institute for Fiscal Studies (IFS) opposed the idea, as did many others, and it did not happen. The 1978 Meade review supported a wealth tax in principle, but did not develop the idea. The 2011 Mirrlees review looked briefly at the idea, and concluded that a wealth tax “is costly to administer, might raise little revenue, and could operate unfairly and inefficiently”. Most OECD countries that have had wealth taxes have since abandoned them. 

The most serious work on wealth taxes has been undertaken by Advani et al. They estimate that a wealth tax of 1% over £10 million would raise £12 billion a year. That is equivalent to around a penny on income tax or on VAT.

The Tax Justice Network and Taxpayers Against Poverty have argued that a 2% tax would raise £24 billion, which is in line with Advani et al. Unite the Union want 1% over £4 million (claiming £25 billion), the Trade Union Congress 1.7% over £3 million (which is the top 1% of the wealth distribution), and 3.5% over £10 million (£10 billion, which seems low) and the Green Party want 1% over £10 million and 2% over £1 billion (£15 billion). 

These numbers are the opposite of robust. We do not have good data on the wealth of individuals, only estimates that are very imprecise. No-one is required to tell the government what they own, and items like unlisted companies are hard to value. Nor do we have any good idea how people subject to this tax will respond. British tax lawyer Dan Neidle estimates, however, that 80% of the wealth tax revenue would come from 5000 people (each with £50 million) and 15% of the total revenue from 10 people. Many of these people already have citizenship of other countries, and all would find it easy to obtain citizenship and residency. Telling someone that they have to pay hundreds of millions of pounds every few years to remain resident here may induce that person to leave. If they do, not only would we fail to get the expected wealth tax revenue, but we would also risk losing their current tax contributions, as well as any benefits of their entrepreneurial activity, charitable contributions, and so on.

It is therefore plausible that a wealth tax would raise a lot less than these estimates, both in its own terms, and in terms of the overall effect on government revenue.

A 1% wealth tax may sound low, but it is a large amount relative to investment returns. Imagine, for example, an investor buying 10-year UK government stock. The yield is currently 4.4%. If the owner is rich enough to pay the wealth tax, they will surely pay income tax at 45%. This means that their after tax return is currently 2.42%. With a wealth tax of 1% as well, their return falls to 1.42%. The effect of a 1% wealth tax is equivalent to raising income tax to 68%. Few people advocate for 68% tax rates. 

Similarly, a 2% wealth tax is equivalent to arguing for a 90% income tax rate for someone holding 10-year government debt. It is also equivalent to a 103% tax rate on people holding one year government bonds - who would receive a negative return. A wealth tax along these lines would mean that people were asked to pay the government for the privilege of lending the government money. At one level that is just the policy working, and investors could seek riskier and probably higher return alternative assets, but still, it is plausible to think that a tax above 100% of the interest earned will not pass a common sense test. It would also be likely to require the government to offer a higher rate of interest on government bonds, in order to induce people to hold them. This would reduce the return to the government from imposing the wealth tax. 

A wealth tax also cumulates over time to a significant extent. A person who inherits and owns wealth for one generation, say 32 years, will pay the government 27% of their inheritance if there is a 1% annual wealth tax, and 48% of their wealth if there is a 2% annual wealth tax. 

As the IFS set out, there has been a decline in the number of countries that administer wealth taxes over the last 35 years. In 1990, 12 European OECD countries levied personal wealth taxes. Most were repealed in the 1990s and 2002. France is the most recent country to abolish the wealth tax, doing so in 2018, although they replaced it with a tax on high-value immovable property - which is a form of wealth tax (albeit one that introduces distortions across asset classes). 

At present, Norway, Spain and Switzerland are the only OECD countries that levy individual wealth taxes.

• Switzerland: raises 1.1% of GDP through a wealth tax. But the high figure in Switzerland needs to be seen in the context of the country having no capital gains tax, and no inheritance taxes on gifts to descendents. In the UK those yield £12 billion and £9 billion respectively, so a crude estimate is that a Swiss approach would raise £11 billion extra. Switzerland has a number of other aspects that make it an appealing place for very rich people to live - including very generous rules for incomers that the UK does not have. £11 billion therefore seems optimistic for a UK wealth tax modelled on the Swiss system. 
• Norway: raises 0.4% of GDP, around £12 billion in the equivalent context. This comes from a wealth tax that taxes people with assets of around £140,000 and up. It is therefore a much more extensive wealth tax than any proposed for the UK. No-one is proposing such an approach here, with all proposals starting at a far higher level of wealth. We should not say, therefore, that the UK could raise £12 billion because Norway does, because no-one is willing to advocate for a Norwegian system. 
• Spain: raises 0.2% of GDP in wealth taxes, around £6 million in the UK context. It also taxes people from around the same level of assets, although with exemptions for houses, pensions among other things. It does seem likely that the UK could raise £6 billion. But it is not at all clear that more could be raised, given that wealth taxes are likely to induce changes in behaviour, including people choosing to leave.

It is hard to imagine a UK wealth tax bringing in more than, say, £6 billion a year, an economy-size adjusted equivalent to the Spanish system. That is less than the amount raised in Norway and Switzerland, but as we have seen, Norway taxes far more people, and Switzerland has other desirable features to attract the super-rich that we lack. Spain is therefore a more plausible level than Switzerland or Norway, but even then Spain is taxing people with far lower assets than anyone is suggesting in the UK. Therefore, £6 billion is probably best seen as an upperbound. £6 billion is not nothing, but it is less than 1p on income tax or VAT, and just 3% of annual NHS spending.

Furthermore, these are gross numbers, and do not take account of the falls in tax revenues that come from the economic downsides set out above. For these reasons it is not clear that a wealth tax on the superrich would, overall, boost tax revenues, and may worsen the fiscal position.

In contrast a one-off wealth tax - particularly if there was an accepted rationale - would be much less likely to induce people to leave. This sort of tax at the time of Covid, or soon thereafter, might well have worked. Similarly, a much lower wealth tax on a much higher number of people would be likely to yield revenue. For example, a Norwegian with a £1m house and £500,000 in cash assets is unlikely to leave Norway because they are required to pay around £7,400 as a wealth tax each year, any more than someone living in Manton in Rutlandshire, which has the highest council tax in the UK (£5456), is likely to leave the UK. Broadbased taxes do not induce movement: highly concentrated taxes do. 

A one-off wealth tax is not particularly credible now, because there is no emergency that makes it credible that it will not happen again in a couple of years. A broad-based wealth tax would defeat the political object of making someone else pay.

The Taxpayers Against Poverty organisation argues that “If we can build a system that taxes every payslip with precision, we can build one to tax multi-million pound asset portfolios.” This is a false comparison. Taxes on payslips are easy - HMRC and I agree to the penny how much I earn each month. Indeed, my employer informs both of us. 

It is equally easy to value any shares or government bonds that I own. A valuation is available at the close of each day, and this could be averaged over the year and act as a basis for a tax. 

Many classes of assets are, however, much harder to value. The first is houses. Although Zoopla gives an estimate, its executive director for research and insight, Richard Donnell, is clear that it is indicative, rather than offering any guarantee of accuracy. It is particularly inaccurate for the most expensive properties, which are unique, and those that sell rarely. Errors of 20% or more are plausible. 

A second class of hard to value assets are defined benefit pensions. Although the civil service will give an official valuation for an accrued pension, it will not offer that amount of money in lieu of my foregoing my pension rights, nor allow additional pension to be bought at that price. 

An even more difficult class of assets to value are non-listed companies. Valuing an owner-run small shop is not straightforward. Valuing a start-up company with no revenue is even harder. It might be worth millions. It might be worth nothing. It is very hard to tell, and it is correspondingly hard to tax. 

The final class is jewellery, fine art and so on. For sure, these things have to be valued for probate from time to time, but it is not straightforward. Doing it annually would be expensive  and unreliable. 

Finally we have the issue of trusts: these are often created to avoid inheritance tax, but would presumably be included in the scope of a wealth tax because otherwise the number of trusts would increase dramatically.

Clearly a wealth tax could be implemented: as we have seen, other countries do it. The owners of corner shops, start-ups, art and so on would have to self-declare the value of these assets, and prove that they had taken steps to value them accurately. The wealth tax commission suggests that the compliance burden would be between 0.05% to 1.5% of total wealth - a significant proportion of the 1% raised - indeed, in some cases total compliance costs would exceed the total tax due. This violates one of the core principles of taxation - that it should be cheap and easy to collect, both for government and for taxpayers.

If the wealth tax applies to growing companies without a right to defer, then entrepreneurs will have to find the money each year. This would require them either to grow the company more slowly, or to sell out earlier. Both are bad news for the UK and for entrepreneurship. It will also reduce UK tax revenues in other ways.

Given the sums involved for the richest, outmigration - including selling assets on exit - is likely to occur to some extent, and would clearly harm the UK economy.

Making the UK a less attractive place to grow a business is also likely to reduce economic growth.

The UK already has some wealth taxes, most obviously stamp duty on land and on shares. These are transaction taxes, that is, they are levied at the point of purchase, rather than as ongoing annual taxes. This is both an advantage and a disadvantage. 

The advantages of a transaction tax are two-fold. First, there is a clear and unambiguous value. The house is sold for £750,000, therefore the value is £750,000 for the purposes of tax collection. The tax is £27,500 if the buyer is buying their own home, or £65,000 if they are buying a second home, whether for themselves, or to rent out and there will be no disputes as to the value (these are not small sums). Second, a transaction tax has the advantage that the person has the money, since if they do not, the transaction will not go ahead. In contrast, a classic wealth tax is levied irrespective of ability to pay. An asset rich, income poor person may have to move house - often seen as unfair - or the state would have to offer a loan system. 

Against that, transaction taxes on housing have two disadvantages. First, they discourage moving. This is distortionary, and is bad for economic growth. Second, they are unfair. The person who is lucky enough to always work in one area pays only once, the person whose career takes them from place to place pays more than once. 

The UK also stamp duty on shares. When shares are purchased, 0.5% of the value is due in tax. Again, this is a translation tax, and again it is distortionary. It is not a proper wealth tax, but it has some similarities, in that it is primarily paid by the wealthy. 

Finally, we have a £2125 addition tax levied on new cars priced at more than £40,000, or £50,000 if the car is electric. This is levied in five installments, as a supplement to vehicle excise duty. Once more, this is a quasi-wealth tax. It is not a particularly well-designed wealth tax in that the tax due is the same whether the car is a top of the range Toyota Yaris, or a £23 million Rolls Royce Droptail, nor does it continue to be levied if the car keeps its value over time. That said, it raises over £1 billion a year (compared to the predicted yield VED for 2019-20 from the 2015 budget and 2017 budget, at £5.1 billion and £6.3 billion).

If the government wishes to ask those with the broadest shoulders to contribute more, there are more effective ways to do so. 

Property taxes

Economists dislike transaction taxes very much, and prefer annual taxes. For that reason stamp duty land tax is strongly disliked - indeed, the former head of the IFS has described it as the worst tax in the country. 

There are currently two good proposals to replace stamp duty land tax (and council tax). The first is a paper from myself for Onward. This sets out a proposal that would be revenue neutral on a net present value basis, although with a short term fiscal hit for the government. The Centre for British Progress then built on this idea, and produced a version that was revenue neutral in the short run, and cash generative in the medium term. Both proposals argue that the new system would only come into operation when a house is sold, to avoid unfairly hitting people who have just paid stamp duty. At that point, the new owner would choose between paying stamp duty up front, or paying the new proportional property tax (PPT). Once the home was subject to the PPT, however, it could not revert in future. The rate of the PPT would vary from 0.2% for homes £125,000-£250,000 up to 1.1% over £5 million, roughly mirroring the current progress nature of stamp duty land tax. That proposal should be attractive to anyone interested in a wealth tax and should be studied seriously by any genuine tax reformer. 

These taxes could be calibrated to raise any particular amount of revenue within reason. Like the Norwegian and Spanish approaches to wealth taxes, however, the most successful approaches are likely to start from relatively low valuations compared with many currently suggested wealth taxes. 

Capital gains tax reform 

At present, like most countries, the UK taxes capital gains, including gains that are simply inflation. Most economists see that as unreasonable. To compensate investors, the tax rate on capital gains tax is somewhat lower than for other forms of income - 24% vs 40% or 45%. In addition there is a small annual tax-free allowance of £3,000. 

There is no a priori logic to whether 24% is above or below the tax neutral rate of compensation for taxing investors on inflationary rather than real gains. There is, however, empirical evidence that it over-compensates. 

Advani et al. estimate that returning to the old UK system - whereby people did not have to pay capital gains tax if the value of their asset rose only in line with inflation, but paid tax at their marginal rate beyond that - would raise £11.5 billion per year. This proposal is broadly, although not exactly, in line with the Mirrlees report, and would be seen as an improvement on the current system by most economists. It could also be introduced quickly, yielding revenue within a year (the best approach would be an announcement, effective as quickly as possible, to prevent people choosing the tax system that minimises their tax obligations). 

Although this is not a wealth tax, it would be an £11.5 billion rise in tax on people who own substantial assets. As such, it probably does achieve the underlying aims of most people who are pro a wealth tax, namely that the rich should pay more. 

Changing the tax-break on ISAs from an annual allowance, to a lifetime allowance, would achieve a similar outcome. It is now possible to have more than £1 million in an ISA, and the tax break on that sized portfolio should be anathema to those who support a wealth tax. Limiting the total amount that can be put into an ISA to, say, £250,000 (either retrospectively, or for future investments) would achieve similar objectives to a wealth tax, but would be easier to operate, since there would be no need to value any assets or create a new tax.

A wealth tax seems an easy win - a tax that most of us can imagine “someone else will pay”. As is so often in economics and politics, if it was that easy, someone would have done it already. There are better approaches for any government that wants to raise more tax from the richest in society.