It’s personal (taxation)

In the weeks surrounding the Budget we’re likely to hear a lot of tax discussion. That reflects a deterioration in the public finances – with an IFS estimate of £22 billion – and still more will be needed if the Chancellor wants to build extra headroom or spend on her policy priorities. Our recent paper set out specific suggestions to help fill this gap in the least damaging ways, but this Spotlight steps back and explores the context of current levels and methods of income taxation in particular.

Overall, partly driven by the need to fund much higher interest costs, taxes as a share of GDP are set to rise over this Parliament to all-time highs of over 37 per cent, up from 35 per cent last year and 33 per cent in 2019-20. This puts us mid-table across the G7, with the UK having a lower tax to GDP ratio than recent levels in France, Italy or Germany, but higher than in the US, Canada or Japan.

Zooming in on personal taxes (and this Spotlight includes the effects of payroll taxes), we show that the tax wedge for an average full-time earner in the UK was the lowest in the G7 in 2024, and will very likely remain below the OECD average even once 2025’s employer National Insurance (NI) rise is incorporated into estimates. UK effective tax rates for low- and middle-earning employees – including employer NI – have reached their highest points since the early 2010s, but remain below pre-financial-crisis levels.

While some elements of the tax system, like the personal allowance, have risen and then fallen back – driving those tax trends – the make-up of taxation has changed in important ways. For a typical employee, employer NI is now a record-high share of their tax wedge. This has important implications, particularly for the tax difference between employee and self-employment income. A typical employee will now pay a record 55 per cent more tax than an equivalent self-employed worker, and could be just as well off in a self-employed role even if this were 13 per cent less productive. Overall, the cost of the self-employment tax break has now topped £9 billion a year. Policy change at the Budget would be wise to stop this growing further.

UK personal taxes remain low by international standards

Given the rise in the overall tax ratio, and cost of living pressures, it is natural to worry about what might have happened to personal taxes. Sir Tony Blair has reportedly argued that “if you look at the levels of direct taxation for people, they’re really high, and in historical terms, they’d be considered extremely high”. But internationally the UK stands out for its low personal income taxes. In OECD comparisons, which look at the equivalents of Income Tax and employee NI paid by a full-time employee on mean pay (around £50,000 in the UK), our effective tax rate was the lowest in the G7 last year, with the UK on 21 per cent – compared to 24 per cent in the US and 37 per cent in Germany.

But this is not a complete measure: if we add in employer NI, then the total tax wedge for this employee rises to 29 per cent (as a share of total labour cost). But this was, again, the lowest in the G7, and among the lowest in the OECD, as Figure 1 shows, with an OECD average of 35 per cent.

Figure 1: In 2024, the UK had the lowest taxes on an average employee of any G7 country, though that position may slip in 2025

UK taxes for lower earners – on 67 per cent of average pay – were also the lowest in the G7, at 25 per cent versus an OECD average of 31 per cent. It’s only at higher earnings levels, at 167 per cent of the mean full-time employee pay, that the UK moves to the middle of the G7 – higher than the US, Canada or Japan but below our European peers – although we are still below the OECD average (37 versus 39 per cent).

There are always caveats to international comparisons. One is that the numbers cited above are for 2024, meaning they don’t reflect the UK’s recent rise in employer NI. We will need to wait for the OECD’s next publication for a definitive take on 2025 values (because other countries’ tax systems also change) but the employer NI rise would meaningfully raise the UK’s tax figure for the average earner. Our estimate is that it might rise from 29 per cent to 32 per cent, which would overtake at least the US’s 2024 value – but surely still leave it below the OECD average and far below Italy, France and Germany.

Another caveat is that these figures are for single adults, but many countries have lower taxes for those with children or lower-income partners. Taking into account both taxes and cash benefits (as the line between tax reliefs and cash relief is not ultimately meaningful here), the UK has an above-OECD-average effective tax rate on a lower-paid single parent with two children, and a very typical rate for an average-wage single earner couple with two children. What people get in return for their taxes also varies by country, of course. In some countries, workers would need to pay for private, non-compulsory health insurance on top of these tax rates (which doesn’t need to happen in the UK); on the other hand, many workers in the UK make private, non-compulsory pension contributions on top of these taxes, which doesn’t need to happen in some other countries where ‘taxes’ include substantial pension contributions.

For low and middle earners, personal taxes remain low by historical standards, but are now higher than in most of the 2010s

Figure 2 shows the historical context of effective tax rates in the UK, in relation to the median salary of all employees – which new data gives as around £33,000 in April 2025 – and again including the effects of employer NI.[1] For low and middle earners, the tax allowance increases of the 2010s, together with last year’s personal NI cuts, significantly lowered effective tax rates. On the other hand, the employer NI rise and the ongoing tax threshold freeze have pushed them up, with an overall effective tax rate for the median employee in 2025 of 27 per cent. This is lower than before the financial crisis (29 per cent in 2007-08), but is the highest it has been since 2011-12. For lower-paid employees, the tax wedge has risen rapidly in recent years, due to the employer NI rise and the fact that such workers gained little from the employee NI cuts; at 16 per cent, the tax wedge is at the highest level since 2010-11, but also below its pre-financial-crisis level (18 per cent in 2007-08). Meanwhile, tax rates for higher earners (at two or ten times the median) are set to be a little higher than they were in the 2000s or 2010s, but still below the levels that were in place before Nigel Lawson’s tax-cutting 1988 Budget.

Figure 2: Fiscal drag and the employer NI rise have raised effective tax rates, but these remain below pre-financial-crisis levels for low and middle earners

But the chart and discussion above look only at those taxes levied directly on earnings, and at an individual level. Figure 3 explores an alternative measure from the ONS’s recent Effects of Taxes and Benefits on Household Income data, which looks at the incidence of a wider range of taxes. For example, the direct tax line here includes student loan repayments – which contribute to higher marginal rates for some but are still small in aggregate at around £5 billion a year in total – as well as Council Tax, which we have shown in earlier work to be rising in importance. There are also differences in the analytical approach, with the ONS’s series based ultimately on survey responses, with imputation of some taxes, and done on a household level. Despite these differences, there is a broadly similar picture to the above, with direct taxes for middle-income (fifth decile) non-retired households in 2023-24 trending upwards, but remaining below pre-financial-crisis levels. One advantage of the ONS series is that it also covers indirect taxes, such as VAT and Fuel Duty. Superficially, this shows that indirect taxes are also not historically high either. But there are two caveats to note with this data. The first is one of timing: it does not reflect the 2024 cuts to personal NI nor the April 2025 rise in employer NI. The second is more substantial, and stems from the fact that these estimates derive from a household survey that has known issues, and this may mean that the amount of indirect taxes paid by households are being underestimated (and this also has implications for all the statistics in the ONS release about what share of households are “net contributors”). But clearly the ONS’s best estimates do not point to households near the median income paying historically high taxes.

Figure 3: Middle-income household data (up to 2023-24) gives a similar picture of rising, but not historically high, direct taxes – and no indication of high indirect taxes

So although our overall tax to GDP ratio may be higher than it has been, that is not true of the tax wedge for a typical single employee, nor (according to ONS data) for typical non-retired households. But the recent rise in employer NI was an important change that does require some reflection about how we tax.

The make-up of employee taxes has shifted towards employer NI…

One thing that has changed over the long-term is the make-up of taxes. Figure 4 breaks down the direct tax liability of the median earner (as shown in Figure 2 above) into Income Tax, employee NI and employer NI. Effective Income Tax rates have fallen over time, from around 14 per cent in 2007-08 to 9 per cent in 2019-20, but have since ticked up to 11 per cent (driven by changes in the personal allowance). Personal NI has shrunk significantly over time, to a record low of 4 per cent, thanks to its higher allowance and (from 2024) a lower rate. But the effective burden of employer NI of the median employee has grown. If we assume that all employer NI is ultimately paid by workers, then this has risen from 9 per cent in 2007-08 to 11 per cent in 2025-26 – only slightly below our estimate for the early 1980s when a ‘National Insurance Surcharge’ remained in place.[2] And as a share of the tax wedge, employer NI has taken on record importance. Back in 1975, NI was an estimated 22 per cent of an employee’s tax wedge; by 1999 (when the modern employer NI system was created) it accounted for 28 per cent; by 2017-18, this had risen to a third; and in 2025-26 this was 42 per cent.

Figure 4: Employer NI is now a record share of the tax wedge on a typical employee salary

…and this is pushing the tax bias towards self-employment to record highs

If we are correct in our assumption that employer NI is ultimately paid for by the employee, and there is certainly plenty of evidence for that, then an employee shouldn’t care in the long-run whether their tax burden comes from Income Tax, employee or employer NI: they are all just parts of the wedge between what an employer has to pay to employ someone and what the employee gets to take home. But the balance between Income Tax, employee NI and employer NI does have implications elsewhere, as it means that the UK has different tax treatments for different sorts of people and different forms of income. This points to some considerations about how policy should ideally change at the Budget.

One implication of the rise of employer NI is that we have ramped up the incentive for employees to be remunerated through employer pension contributions rather than salary or personal pension contributions. Opinions may vary on whether that incentive is a good use of over £20 billion a year. But the different treatment of personal and employer pension contributions adds unnecessary unfairness: firms and workers who are able to restructure their remuneration to maximise employer contributions will pay less tax than those that are not big enough to administer salary sacrifice arrangements or that simply follow the Government’s default auto-enrolment rates, for example. This is why we have proposed that the Chancellor could dial down the generosity of pension salary sacrifice tax relief at the Budget, so as to reduce these arbitrary tax differences.

More importantly, though, the rise of employer NI has changed the tax gap between the self-employed and employees. We have seen that effective tax rates for employees have ticked up, but we estimate that the effective tax rate for a self-employed worker equivalent to the median employee (in terms of overall output or labour cost) in 2025-26 was the second lowest in the past 50 years, with only 2024-25 being lower (see Figure 5), thanks to the previous Government’s decision to cut their NI rates and abolish compulsory, flat Class 2 payments. The fact that employer NI doesn’t apply to the self-employed, together with their slightly lower basic personal NI rate added up to a tax break of £8 billion in 2024-25. We estimate this will have grown to £9 billion in 2025-26, given higher employer NI, and will top £10 billion by the end of the Parliament unless policy changes. With a self-employed tax rate of 17 per cent and employee tax rate of 27 per cent in 2025-26, the latter is now paying a record 55 per cent more direct tax than the former. For this typical worker that is a tax gap of over £3,500 a year.

Figure 5: Taxes for the median employee are now a record 55 per cent higher than those for an equivalent self-employed worker

Such tax differences are not just for the tax purists, but influence how our economy is structured. For example, should you eat in a restaurant, where the staff serving you will attract employer NI, or have the food delivered to your home, with delivery drivers not paying that tax? Should you go to a hair salon that employs staff or one that merely hosts self-employed hairdressers? Giving an unwarranted tax break is not only unfair to employees, but also means that self-employment can be more lucrative to a worker than an employee job even when that job is more productive. For example, if a worker had to choose between employment with productivity of £40,000 per year and self-employment of £35,000 per year, the NI system would push them to accept the latter, all else equal, despite productivity being 13 per cent lower. Similarly, a very high earner would be better off generating £900,000 of output through self-employment than £1 million through employment. For a country striving to eke out more growth, this distortion of incentives is a significant design problem, and one that policy makers keep making worse. Conversely, beginning to tackle the issue could bring productivity benefits that might offset any of the usual labour-supply hit from higher tax rates. As a very rough illustration, if you thought 1 per cent of the workforce (340,000 workers, out of 4 million self-employed) were currently mis-allocated due to this tax bias and could be approximately 10 per cent more productive elsewhere, that would be worth a 0.1 per cent boost to GDP – similar to the long-run benefit of the recent, well-trumpeted UK-India trade deal.

In many other countries, the self-employed must pay the equivalent of both employee and employer social security contributions parts. A good starting point for the UK, then, would be to at the very least close the gap in personal NI rates (6 per cent for the self-employed; 8 per cent for employees), which would still leave self-employed tax rates lower than in 2023 – while also raising the basic rate of dividend taxation; and to extend a form of employer NI to at least some partnership income (especially for big Limited Liability Partnerships). Changes such as these could prevent the self-employment tax break rising beyond £10 billion a year, and would also be consistent with the Government’s intention to extend a new Unemployment Insurance benefit to the self-employed.

Voters remain concerned about the cost of living (59 per cent of voters think the Government should prioritise ‘the cost of living and the economy’ in the Budget), and projections for living standards growth remain bleak. But, within this, we should not overdo concern that taxes for typical employees or households are unusually high in the UK either by international or historic standards (only 13 per cent of voters in the same survey said that ‘cutting taxes’ should be a priority). But we should be worried that Budget 2024, together with previous trends, has delivered tax rises disproportionately through employer NI – increasing damaging biases that Budget 2025 would do well to correct, at the same time as balancing the books.

[1] For series excluding employer NI, see A Corlett, Call of duties: Revenue and reform for Autumn Budget 2025, September 2025. https://doi.org/10.63492/urc611

[2] The incidence of employer NI is not uncontested. For example, the OBR has said that “we assume, based on demand and supply elasticities for labour, that 76 per cent of the total cost [of the 2025 tax rise] is passed through lower real wages, leaving 24 per cent of the cost to affect profits. Of the long-run pass-through of this cost to employees’ real wages, we assume four-fifths comes through lower nominal wages and one-fifth via higher prices.” As indicated, an assumption of lower incidence on employees would further reduce current personal tax levels relative to international and historical comparisons. The incidence is less important when considering the economic effects of the bias between self-employment and employment taxes.