The public’s surprising choice of tax increase, and why we should ignore it

Gabriel Milland of Portland Communications has published polling conducted by the Portland research team in early October. Gabriel takes some interesting political conclusions from the polling (and I’d recommend his article). However, my focus is what the polling says about tax policy, and about polling on tax policy.

Amended to include the answers – apologies for missing this out first time

What do people want?

Here’s the result of asking the public which tax they’d prefer to put up:

It’s a surprising win for taxes on betting, with Conservative voters in particular strongly in favour. Capital gains tax runs a close second, and bank taxes and VAT on private schools are tied for third.

And who do the public think should take the burden of any tax rises?

People who are not them. Specifically, people earning £75k1Who already face a top marginal rate of almost 60%, second home-owners, non-doms, businesses and shareholders.

About a quarter want to increase taxes on business. But, if we focus on business tax specifically (and ask the question in a certain way), there’s a large majority that’s against raising business tax, because they think the cost will be passed on to them:

Should we pay attention to these results?

In my view we should not, for two reasons.

First, there is widespread misunderstanding of tax and spending.

Portland asked respondents to list the top three things Government spends money on:

Total Government spending is around £1,200bn. Spending on asylum seekers and migrants amounts to about £4bn (0.3% of all spending)2It comes out of the foreign aid budget. Spending on MPs’ expenses and staff costs amounts to about £150m each year (0.01% of all spending).

Here’s an approximate chart with the actual figures, thanks to ten minutes of very fast work by Maxwell Marlow. Maxwell stresses that we shouldn’t treat it as more than indicative:

So why do people think that MPs’ expenses are in the top three spending items, when they’re not in the top three hundred?

Because most people don’t understand the size of the Government budget. Only one in six people got the right order of magnitude:3This is a better result than we’d expect if people were guessing (a z-score of approximately 3.355 corresponds to a p-value less than 0.001), but still depressing.

The correct answer is £1 trillion (more precisely, around £1.2 trillion).

And this result in turn may be because most of the population don’t understand billions and trillions:

There are a thousand millions in a billion4The UK used to use a “long billion” of a million millions, but that ceased being used for Government statistics in 1975. We can probably discard the possibility that some of the wrong answers are from people remembering the old billion, because the results vary little across age groups., and a thousand billions in a trillion.

And there is also widespread misunderstanding of one of the simplest feature of the income tax system – the way that bands work:5If you look at the data, linked in Gabriel’s article, people on higher incomes, the self-employed and graduates are more likely to know the correct answer, but not by much. Some of the subgroups have samples too small to be statistically significant.

The second answer is correct. You pay 40% tax on your income above the 40% band (which for most people is £50,000, but that threshold drops for high earners as the personal allowance is reduced).

This reflects our earlier polling – the question is intentionally different from last time, but the result is consistent.

I’d repeat our earlier point that there is a risk this is changing peoples’ real world behaviour – for example causing some people to turn away work for fear of entering the 40% tax band. It would seem prudent for HMRC to investigate if there is a real effect here.

Second, these are three second conversations

Tony Blair talked about the fallacy of polling individual policies and thinking that you’re learning something from peoples’ three second take on complex issues. In the real world, they may have a different view after a thirty second conversation, and a very different view after a three minute conversation. Blair’s point was that election campaigns, where policies are contested, are likely to reflect the three minute view more than the thirty second view (where policies are presented without context).

So, for example, we suspect that the popularity of raising tax on people earning £75k would fall upon realisation that their marginal rate is already often pushing 60%. The consistent popularity of wealth taxes would be unlikely to survive presentation of the history of faliure of previous wealth taxes. People may be attracted to the idea of charging VAT on private school fees, but the amount it raises is economically insignificant. And that interesting finding about the unpopularity of tax on business may well be significantly influenced by the very “three second” framing with which it was presented.

If most people don’t understand the tax system, or government funding, should tax policy simply ignore popular views of tax and the tax system?

That would be a mistake, and perhaps even dangerous. Our political system, and our tax system, needs popular consent. And ignoring popular views on tax would give space for populists, of Left and Right, to pursue tax policies that could be highly destructive.

Tax policy therefore has to be informed by the public mood – but it shouldn’t be led by three second conversations, which confuse public sentiment for public opinion. And tax policy, and tax educators, should seek to challenge popular misapprehensions.

However there is a wider problem.

Robert Colville recently wrote in The Times that much bad tax policy is driven by politicians’ fear that they can’t tell voters the truth about the real constraints on tax and spending. He may be right. But how can politicians tell voters the truth about what may have to change, if voters don’t understand the basics of how things work at the moment?

It seems trite and inadequate to say that we need more education around finance and tax. Right now I don’t have any other suggestions.


Many thanks to Gabriel Milland and Portland Communications – they ran the tax band question, and shared with polling data with us, entirely pro bono.

Photo by Red Dot on Unsplash.

  • 1
  • 2
  • 3
    This is a better result than we’d expect if people were guessing (a z-score of approximately 3.355 corresponds to a p-value less than 0.001), but still depressing.
  • 4
    The UK used to use a “long billion” of a million millions, but that ceased being used for Government statistics in 1975. We can probably discard the possibility that some of the wrong answers are from people remembering the old billion, because the results vary little across age groups.
  • 5
    If you look at the data, linked in Gabriel’s article, people on higher incomes, the self-employed and graduates are more likely to know the correct answer, but not by much. Some of the subgroups have samples too small to be statistically significant.

We welcome comments from readers, particularly where there are technical errors or omissions in our reports. Please try to keep the comments away from political and personal issues, and focussed on the topic of the article or report. Unfortunately we have to have some moderation to prevent spam; the first time you comment there will be a delay until your post is manually moderated (sometimes minutes; sometimes hours or even days). Once you’ve had a post accepted then all future posts should appear immediately.

23 responses to “The public’s surprising choice of tax increase, and why we should ignore it”

      • I did see that, but doesn’t answer the question about the methodology (sampling/recruitment/survey design).

        For your own articles using their data, I think adding that description will add credibility.

  1. Imho, a key issue is “horizontal” tax equity, as well as the need to extract money from the economy (as tax – which ‘destroys’ money) in order to avoid inflation from too much government spending and too much credit money creation. (Noting that government spending is pretty important and really shouldn’t be cut, given the dire state of so many services and infrastructure).
    Horizontal equity means taxing UNearned income at the same rates as earnings, not burdening employment with more NI.
    How is it that ex PM Sunak paid 23% of his income of £2 million as tax? That’s the same overall rate as someone paying PAYE on earnings of £55,000.

  2. If we had run our finances like the government did such as the incredibly high waste on ‘bad’ spending such as furlough fraud,bounce back loan fraud, government ‘mates’ being paid handsomely for PPE etc, we would feel the full impact of insolvency law. As a Chartered Accountant, it was mind bogling that there were no checks on BB loans and this was a disaster in the waiting. You wonder what ministers do.
    There is ’cause and effect’. If you spend less, you tax less. So you need to cut spending AND lower the tax increase.

  3. Interesting article but would have benefitted from some context on what the accurate answers were. On the top 3 things the government spend money on I could spot some of the obvious ones scoring too highly, though wasn’t sure what the biggest 3 would be. Having tried to find out, it looks like the biggest chunk at 24.3% is ‘other’, so I’m not really any better informed! https://ifs.org.uk/taxlab/taxlab-key-questions/what-does-government-spend-money

  4. No measure affecting only a small cohort will raise much revenue. There is no choice but to hit the largest population with higher Employers NI rates. It’s the only way Labour can spin the manifesto pledge not to increase taxes on working people – it will be paid mainly by Employers & entirely by Off-Payroll workers & subsequently recouped by higher prices/lower wages. Meanwhile base rates have risen due to Reeves meddling with fiscal rules to justify substantially higher borrowing. And she criticised Liz Truss !!!

  5. Hi Dan, I popped this question on the hellsite formerly known as twitter too, but appreciate you get a lot of notifications so may be hard to swim through the sludge.

    On just a couple of these charts, the total %’s are higher than the greater of conservative or labour, and some where the total is lower than the lesser of either. I initially thought that was a possible error in the chart, but I wonder if voters for other parties were included but not split out? If so, we can infer that the “other” categories views are stronger than either of the two largest parties one way or the other. Thanks.

    • Hi – yes, that’s correct – other parties aren’t broken out because the numbers get too small for statistical significance. It’s in the linked spreadsheet.

  6. “It seems trite and inadequate to say that we need more education around finance and tax” – So possibly provide the actual figures for eg government spending alongside the public perception?!

    It is claimed that there is a direct correlation between income tax and democracy (ie if you’re parting with your hard earned cash, you want to know how it is spent.) That doesn’t appear to be bourne out by this polling.

    • The median worker in the UK actually pays a fairly low rate of tax compared to other countries.

      There’s a good article by the IFS that demonstrates this: https://ifs.org.uk/taxlab/taxlab-key-questions/how-do-uk-tax-revenues-compare-internationally

      “The average tax rate (incorporating income tax and SSCs) on median full-time earnings in the UK was 28% in 2016–17 (the year for which this analysis was undertaken). This is much lower than it would have been under the tax systems of the other countries shown in the chart, for which the average was 44%.”

      So I think there’s a good argument to make that increasing the tax base (which is going to happen anyway with frozen thresholds until 2028+) is going to help in terms of the public educating themselves on how their tax is spent as more people are pulled into paying tax/at a higher rate.

      It’s also a good argument for scrapping the nil-rate tax band and introducing a nominal 5-10% rate of tax on all earnings – if everyone in the country is paying in, everyone would be more engaged with what they are getting out.

  7. Very interesting charts. So that I can check my answers and educate myself, could you possibly publish the correct answers to the interesting questions you pose about tax and Government spending? (the top three things on which the government spends the most money from tax, the size of government spending, how much tax someone earning £100000 pays).

  8. I guess the confusion over billions and trillions is uniquely intrinsic to being British. After all a British billion/trillion isn’t the same as the billions/trillions that are used in finance (occasionally referred to as US billions/trillions)

    • the “British billion” ended official use in 1975. I think that would be a good explanation if we saw older people disproportionately making the mistake… in fact it’s broadly equal across age groups. See the spreadsheet.

  9. Thanks Dan. Unfortunately you don’t show us details of what is actually spent. I suspect the waste is staggering. Handing your money over as a blank cheque to a here today gone tomorrow ideologue is not an ideal scenario. The truth is we should be given a choice where our money is spent not be vulnerable to to the arbitrary diktats of Rachel from Accounts. I also sense a condescending undertone to your article. The politicos adopt a similar snobbery. Perhaps why they are so despised.

  10. Hi Dan, it’s estimated that 1 in 50 cars and rising doesn’t pay ved which costing £120m pa, if the Gov’t scrapped ved and put say £0.05 on a litre of fuel then high mileage drivers like me would pay more, i currently do 23k @55mpg paying £20pa ved but my son does 1k @40mpg paying £140pa, putting it on fuel ensures everyone pays proportionately to the environmental damage of emissions and may help road safety.

  11. If you have a funded pension, and there is legislation that your employer provides you with one, its assets will be held offshore, not to avoid tax but to delay it

  12. Interesting charts & insights thanks Dan.
    Personally I’d like to see higher tax on super high earners to allow those on lower incomes to keep more of their earnings but need your help with the maths. So:
    If base rate was increased from £12,570 to £16k and 40% & 45% tax bands stayed as they were but a new tax band of 50% was created for those earning the highest salaries, where would that 50% band need to kick in to balance the tax take? Any feedback greatly appreciated. Thanks

    • unfortunately that doesn’t remotely work. Increasing the personal allowance to £16k costs more than £25bn. Even if you turn the entire 45p band into 50p it’s unlikely to raise more than £1bn. The 50p rate was a political sideshow, not a tax-raising measure.

    • You can get a rough idea of the effects of various changes to the tax system here:

      https://www.gov.uk/government/statistics/direct-effects-of-illustrative-tax-changes/direct-effects-of-illustrative-tax-changes-bulletin-june-2024#corporation-tax

      Based on those figures, increasing the additional tax rate from 45p to 50p would roughly pay for a £100 increase in personal allowance.

      There is no way to raise the sorts of sums of tax the government is talking about raising without doing one of two things. Either you ask a very, very large number of people to pay a little bit more or you ask a small number of people to pay a very, very large amount more.

      The first is politically hard because taxes are generally unpopular. So you end up having to hide who is really bearing the cost of taxes (such as increasing employer’s NI, which is largely borne by employees through fewer jobs or lower salaries).

      The second is economically hard because you are likely to have very large behavioural effects (early retirement, emigration, reduced investment, etc).

Leave a Reply

Your email address will not be published. Required fields are marked *