Capital gains tax (CGT) is currently both too high and too low. It taxes investors at too high a rate when they’ve put capital at risk, only to see it eroded by inflation. But it enables a very low rate of tax for people who haven’t put capital at risk, but are able to pay capital gains tax (rather than income tax) on what’s realistically employment income.
We can fix both these problems by raising the rate but reforming CGT so investors pay a lower effective rate. That could be pure tax reform, or it could potentially raise up to £14bn. Or – my preference – it could be used to cut the rate of income tax and also raise around £6bn. This article explains how.
This Government was elected on a platform of kickstarting economic growth. It has a large majority, and four or five years until the next election. It’s a rare chance for real pro-growth tax reform. That’s all the more necessary if we are going to see tax rises.
We’ll be presenting a series of tax reform proposals over the coming weeks. This is the fourth – you can see the complete set here.
This article is based heavily on a recent IFS paper by Stuart Adam, Arun Advani, Helen Miller and Andy Summers, and a Centax policy paper by Arun Advani, Andy Summers and Andrew Lonsdale.
The proposals here are also consistent with those in the Mirrlees Review, the magisterial 2010 review of the UK tax system chaired by Nobel laureate James Mirrlees.
The rate of CGT is too low
Capital gains tax will raise about £15bn this year.
The highest rate is usually 20%, but it’s 24% for residential property and 28% for the “carried interest” which private equity fund managers receive from their funds.
That’s a lot less than income tax, where the highest rate is 45% (48% in Scotland).1 There’s an even greater gap with tax on employment earnings, where the 45%/48% top rate of income tax is on top of 2% employee national insurance and 13.8% employer national insurance.
This creates two problems.
First, many people think it’s intuitively unfair for a wealthy person making a large gain on their shares to pay less than half the tax rate of someone with normal employment income.
Second, it creates a massive incentive to transform income into capital gains, to reduce your tax. For people who run a business in their own company, this can be as simple as: don’t take dividends out of the company, take loans from the company for now, and in the long term expect to sell the company and make a capital gain. There are many more complicated schemes.
So we should raise CGT.
The rate of CGT is too high
Economic theory says that, if investors put capital at risk, we shouldn’t tax them on the “normal return” (i.e. the risk free return, broadly equivalent to bank rates). If we do, we discourage investment – the investor has done worse than if they’d put cash in the bank, but we’re taxing it anyway.
We should instead only tax the “super normal return” (i.e. if an investor’s investment pays off).
CGT does the opposite of this, because no allowance is given for inflation.
Take an example where I make a less than normal return. Say I bought an asset for £1,000 in 2014, and I sell it for £1,250 today. On the face of it I’ve made a £250 gain. But inflation since 2014 accounts for £230 of that “gain” – I’ve really only made a gain of £20. So if I pay 20% capital gains tax on £250, that equates to an effective tax on my “real” gain of 250%.
The longer an investor holds an asset, the greater these effects become. If the investment doesn’t pay off, I’ve made no money (in real terms), but pay tax anyway. I may decide I’m better off spending the money.
Another bad feature of CGT is that I’m taxed on any gain, but if I make a loss then I can’t use that loss to reduce my general tax burden.2 HMRC takes a slice out of the upside (fair enough) but won’t share in the downside (unfair).
All of this means that CGT in its present form discourages investment, particularly long term investment. It acts as a disincentive to people putting capital at risk, which is something we want to encourage.3
By contrast, take an example where I make a super-normal return. Say I buy an asset for £1,000 last week, and sell it for £2,000 today. The normal return is (almost) nothing, and I pay 20% CGT on my super-normal return.
This is not how a tax system should work.
So we should cut CGT.
How to raise and lower CGT at the same time
How do we cut CGT for people putting capital at risk, but raise it for others?
Surprisingly that’s simple: we simply increase the rate, but create a new allowance for the “normal return” on the original investment.
We used to have an allowance for inflation; this would work precisely the same way, but with a different rate. And in the internet age, applying an uplift to acquisition prices in peoples’ tax returns is trivially easy.
Say we raise the rate to 40% and create a normal return allowance:
- In the first example above, where my gain is swallowed by inflation, the normal return over 2014-2024 would be something like 40%. So we have to compare my acquisition cost uprated by the normal return (£1,000 x 1.40 = £1,400) with my sale price (£1,250). I’ve made a loss – no CGT to pay.
- In the second example, the rate increases from 20% to 40% – a sensible result.
So we have succeeded in cutting and raising CGT, at the same time.
This is too complicated. Why not just raise the rate?
If you take the current top rates to 45% then simple arithmetic suggests that would raise about £14bn per year. Why not?
Because it would be a very bad mistake.
It would greatly exacerbate the current effect of CGT to discourage long term investment.
It also won’t raise anything like the £14bn figure. It could lose money
Simple arithmetic often fails to sensibly estimate the yield of tax changes, because it doesn’t take into account “behavioural effects” – people doing things differently, in response to the tax change.
The £14bn figure comes from an Office of Tax Simplification paper, which makes clear that the behavioural effects would be extreme:
What are these behavioural effects? Some mixture of:
- If you announce a CGT change in advance of it taking effect (which is what’s always happened in the past), then people sitting on large unrealised gains can sell their property before the rate changes. They “accelerate” their gain.
- Once the change comes into effect, people who want to sell could take the view that the rate is bound to come down again soon, and “defer” their gain. This is particularly likely if there have been many recent changes in tax rates, or the Government is not expected to last many more years.
- What if someone has a very large gain (e.g. an entrepreneur about to sell their company for billions of pounds), but can’t sell before the rate changes? They have another option. They can leave the UK for somewhere that doesn’t tax foreign gains (and many countries fit the bill here). Then sell and pay no tax.
- There’s a very similar result if, instead of leaving the UK, a taxpayer dies. When the children (say) inherit the asset (which will often be exempt from inheritance tax), all the historic gain is wiped out. Any future sale by the children is taxed on the capital gain from the point they inherited. So an elderly investor facing a CGT bill has a powerful incentive to simply not sell their assets.
This isn’t theory – we can see these effects with each of the many, many, changes to capital gains tax over the last 25 years. Massive taxpayer responses:
CGT is particularly susceptible to these issues, because people control when they sell assets.
There’s more evidence of that in a recent paper by Arun Advani, Andy Summers and others. People who’d received income into companies were liquidating them to make a capital gain. Then the Government announced that, from 6 April 2016, this structure would no longer work. That prompted a huge rush of people liquidating companies to beat the deadline:
And this is why HMRC’s “ready reckoner“, showing the effect of changing tax rates, shows that an increase to the top rate of CGT will lose significant amounts of revenue. That was the problem with the Green Party’s general election proposal to raise the rate.
The lesson is: any increase in capital gains needs to be made very carefully indeed.
How to avoid leaking tax
There is a significant risk of tax leakage, as taxpayers think they’ll lose out and take steps to avoid the new higher rate. So any rise in CGT needs to be accompanied by a policy package:
- There should be an “exit tax”, like many other countries already have, so that gains made in the UK are taxed in the UK. And, to be fair and coherent, we should stop taxing people who’ve moved to the UK on gains they made before they came to the UK (in the jargon, we should “rebase” their assets when they arrive here). I talked about exit taxes and entry rebasing in more detail here. But, in short, people would usually opt for the exit tax to be deferred to the point they actually sell the asset. Given the number of other countries that have exit taxes – the US, Australia, France, Germany – for us to create an exit tax shouldn’t put the UK at a competitive disadvantage.4
- Instead of death erasing historic capital gains, anyone inheriting an asset should also inherit its embedded capital gain. So, when they come to sell, they pay on the asset’s original capital gain as well as the capital gain during their own period of ownership. It is only fair that any inheritance tax is reduced to reflect this practical reduction in the value of the asset (so there is no double tax here).
- Capital losses should be fully utilisable against other income/profits.
- Abolish the existing Business Asset Disposal relief, whose initials tell you precisely what its reluctant creators thought of it..
And there has to be an allowance for the normal return, otherwise the rise in tax will harm investment.
The rate change, and the reforms, need to apply immediately after the Budget speech, so people can’t accelerate sales to keep the old rate.
It would be a very serious mistake to increase CGT without these changes. Investment would suffer, and CGT revenues would likely fall. Winners from the new system would benefit; potential losers would avoid the tax.
The authors of the recent Centax paper agree. Andy Summers said:
“Our proposed package of reforms is about much more than just raising rates. In fact, there’s a big risk that if this is all the government does in the upcoming Budget, it will seriously backfire. There’s big money available, but only if the government is bold and takes on major reform.”
Arun Advani is more blunt:
“if they hike the rate without doing anything else that is a terrible idea. It would be easy to avoid and be bad for growth.“
Winners and losers
Tax reforms almost always have winners and loses.
The clear winners are people who hold an asset which has risen in cash terms, but fallen in real terms (after inflation). Right now, they pay CGT when they sell. They’ll pay less tax. Better than that; they’ll get a loss they can use to shelter other income/profits.
Also winners: people whose assets have risen in real terms, but beat the normal return by a sufficiently small amount that they benefit more from the normal return allowance than they lose from the rise in rates. They’ll pay less tax.
So for many investors, large and small, and in shares and property, these reforms represent a tax cut.5
But people who significantly beat the normal return will pay more tax.
We can quantify this. Say the new rate is 40% and the normal return is 5%. Anyone making an annualised return of less than 10% wins from the proposal; anyone making a return of more than 10% loses.6
You’d be forgiven for thinking that this is such a high break-even point that hardly anyone would be paying capital gains tax under our proposal. The surprising answer (from this Advani/Summers paper) is that nearly half of all capital gains are from shares in private companies where the annualised return was over 100%:
The reason is simple: these are cases where someone starts a company with little or no capital of their own, works for it for many years, and then sells it at a large gain. This wasn’t a return on their financial capital; it was a return on their human capital – remuneration for their labour.7 But it’s currently taxed as a capital gain, at a lower rate than tax on normal income. It shouldn’t be.
And private equity executives’ “carried interest” is also in this category. They pay next-to-nothing for interests in their funds which become incredibly valuable if the funds succeed. Currently taxed as a capital gain at a lower rate than income but, again, it shouldn’t be.
People in these scenarios would pay significantly more tax than at present. If rates were equalised, it would potentially raise £14bn – and most of this new tax comes from these people.
What does this do to incentives for entrepreneurs?
I would say: very little. Very few people starting a company today think about what the CGT consequences would be when they sell it in say fifteen years time. If they did think about it, they’d sensibly conclude from history that CGT rates today are no guide to where they will be in fifteen years time.
There is very little evidence that lower rates of CGT influence entrepreneurship/company start-up rates – our literature search found none. There is, however, evidence that few entrepreneurs consider CGT when they start up a new business – see this IPPR report.
There is also detailed analysis in the Centax paper (starting on page 32), looking at studies of historic CGT changes in Canada and the US.
But there would be very real and positive effects for investors who put capital at risk. There is good evidence that has a significant and positive effect on startups, because it makes it easier for them to access capital.
What should the rate be?
There are three approaches:
One answer is to simply equalise rates to the appropriate income tax rate (45% for most assets; 39.35% for shares) whilst creating an investment allowance. The Advani/Summers paper plausibly estimates that would raise £14bn (and that’s a real £14bn, which fully accounts for behavioural effects).
Another answer would be to raise the rate of CGT less than this. Enough to make the proposals break even, or further – but not to 45%. We’d improve incentives to invest, and reduce the incentive to avoid tax. Raise some lesser sum than £14bn.
But that loses the wider benefits of equalising rates – an end to income-to-capital avoidance, and considerable simplification (given all the anti-avoidance rules that would become irrelevant).
So, if the aim is (at least in part) tax reform rather than revenue-raising, a better idea is to equalise rates but reduce the rate of income tax. For example, cutting all income tax rates by 2% would cost about £14bn – so doing this, and raising CGT to that point would, therefore, be broadly neutral.
Or we could cut all income tax rates by 1% and raise CGT to that point. This would raise around £6bn of additional tax: that would probably be my preference, given the fiscal constraints.
So why not?
There are political challenges here.
The first is that a headline CGT rate of 45% would be the highest in the developed world:
The effective rate would be much lower, because of the normal return allowance, but that may be too subtle a point to affect perceptions.
The second is that the losers, entrepreneurs who start a company with nothing and make a very large gain, have a powerful political voice. I don’t see how these proposals would change their incentives – but I expect many entrepreneurs will disagree.
The third is the risk that politicians see the large figures in the Advani/Summers paper, and think that a simple rise in CGT is the answer, without reform. That would be a disaster.
The fourth is that the numbers are dependent on the Advani/Summers research. I’ve been through it in detail and am convinced; in fact it appears overly conservative in places.8 But I am not an economist. HM Treasury would need to undertake a very serious analysis before committing to this kind of reform.
With these caveats, I’m strongly in favour of proceeding with reform. In the present environment, I’d cut income tax by 1%, raise CGT to the new income tax levels, and then book the c£6bn of proceeds as additional tax revenue.
Many thanks to Arun Advani and Andy Summers.
Photo by Austin Distel on Unsplash
Footnotes
Leaving aside for now the very high top marginal rates that can apply. ↩︎
You can use a capital loss against a capital gain, but only a capital gain – not normal income. ↩︎
NB this is only for UK residents; non-residents aren’t subject to UK capital gains tax, except on land and property. ↩︎
Great care would need to be taken structuring the exit tax, so that people couldn’t escape it using double tax treaties. Treaties are supposed to stop people being taxed twice on the same income/gain, but are frequently used to stop people being taxed even once on an item of income/gain. ↩︎
An interesting consequence is that anyone in this category who tried to “beat the Budget” by selling property ahead of it, to book a gain before the rate went up, will actually have lost money. ↩︎
The formula to give the breakeven return for a new tax rate t and normal return n is: (n x t) / ( t – 20%). ↩︎
It’s sometimes suggested that this is a special kind of labour, because an entrepreneur is taking a small salary in the hope of a large future payoff, but with a significant risk that payoff never comes. However, this situation is by means unique to entrepreneurs. There are many careers where people are gambling that their small initial income will be compensated for by a large payoff in later years, and therefore they accept the opportunity cost of failing to receive an income from a “normal” job. Think: actors, sportspeople – even some legal careers. ↩︎
In particular, I think they underestimate the value of the US “buy, borrow, die” strategy because they assume that assets are subject to US estate tax. In practice, nobody of means pays US estate tax – it is even more full of loopholes than our inheritance tax. ↩︎
Leave a Reply to Andrew Mortimer Cancel reply