John O’Connell is Chief Executive of the TaxPayers’ Alliance.

Tax is back on the agenda in this election. Whether or not Labour could pay for their list of free stuff with tax increases only on the top five per cent of earners was a feature of Andrew Neil’s grilling of Jeremy Corbyn; cuts to national insurance was central to Sajid Javid’s recent media round.

We are, reassuringly, debating the size of the state. The tax burden isn’t a full measure of the size of the state. It doesn’t capture interest or profits of government corporations, the deficit, the regulatory state or how the government influences culture, for example. But it’s arguably the pre-eminent factor and certainly the aspect which most concerns us at the TaxPayers’ Alliance (TPA).

The tax burden is the sum total of taxes in a year divided by the total value of the economy in the same year. It is now at a 50-year high. That’s a widely-known statistic, following TPA research in July last year when we published “Highest tax burden this year since 1969-70”. We thought we’d take that research a step further and look at the average tax burden under each prime minister since 1948, which is as far back as OBR tax burden data goes.

Broadly, the data shows what you might expect on party lines. All Conservative prime ministers until Major left with a lower tax burden than when they entered Number 10. But since Thatcher, things haven’t been so clear. Major, Cameron and May have all left Downing Street with the tax burden higher than when they entered as Prime Minister. Intriguingly, while Wilson and Blair left office with substantially higher tax burdens, the burden fell under Callaghan and Brown.

The numbers need to be treated with caution, however, because many factors interact in ways which can delay or even confound the assumption that cutting taxes means a lower tax burden. That relationship does broadly hold, but the economic cycle means the tax burden rises and falls with the economy.

In a boom, incomes rise fast, pushing more low earners into income tax and more high earners into higher tax bands. Corporate profits and asset values rise as a share of economic output, too, leading to disproportionately higher corporate, capital gain and stamp tax receipts, even if tax rates and thresholds remain unchanged.

The ‘pro-cyclical’ nature of the tax burden explains why it fell so sharply under Callaghan (2.1 percentage points lower in 1979-80 than when he became Prime Minister in 1976-77) and Brown (0.3 percentage points down, despite tax rises).

It also goes some way to explain why it was just 0.7 percentage points lower after Thatcher’s 11-year administration. For readers familiar with the ‘Laffer curve’ it makes some intuitive sense: if tax rates are high enough, as they were when Thatcher took office, cutting them can stimulate activity so much that revenues actually increase.

Similarly, the long boom during Blair’s premiership, arguably a result of the spending restraint of both his predecessor and his own early years in office, may explain some of his substantial rise in the tax burden, too. Despite his image as a moderate, he left it 2.5 percentage points higher.

Perhaps the most interesting aspect, just three days from an election, is what the party manifestos say about what they want to do with taxes, and what implications that will have on the tax burden.

Taking their own revenue assumptions at face value, and assuming the plans have no effect on GDP, we calculated that the tax burden would rise another 0.4 percentage points to 34.8 per cent if the Conservatives win, giving Boris Johnson an average tax burden of 34.7 per cent over his term. That’s the highest burden since Attlee, whose figure (between 1948 and 1951 only) was 35.8 per cent – and falling as the economy demilitarised after the second world war.

Unsurprisingly, Corbyn’s Labour manifesto would be far, far worse from a taxpayer’s perspective. We calculate that he will push up the tax burden to an astonishing 37.9 per cent, the highest ever recorded. And while the OBR tax burden figures only go back as far as 1948, their ‘current receipts’ figures (which include interest, dividends and profits from government corporations) go back further still.

Current receipts were 6.6 percentage points higher than the tax burden in 1948, but before then it was only twice higher than Labour’s 2019 manifesto plans. And at its highest, at 39.5 per cent of GDP in 1945-46, total current receipts were only 1.6 percentage points higher than Labour’s tax burden plans for 2023-24. It is therefore likely that their tax burden will be the highest not just since 1948, when OBR records begin, but ever.

If the Conservatives do win a majority on Thursday, they will of course have several years in which to announce Budgets and adjust plans. Some policy may indeed be reactive, owing to external factors in the global economy. So a victorious Johnson would have the chance to join Churchill and Thatcher as a Prime Minister that reduces the tax burden. Or, he may stay the course and increase it, joining Cameron and May. We know that Johnson is a big fan of Chuchill – so membership of that former club surely holds more appeal.