Home » Background » The tax year

The tax year

The British genius for insanely obscure tax rules is nowhere more evident than in our choice of tax year, which starts on the obviously convenient date of 6 April. I’m hoping that after reading this you may come round to my view that a much more logical date would be 7 April.

It’s all the fault of a ragbag gang of worthies including Julius Caesar, the First Council of Nicaea, a monk called Little Dennis, a Pope called Gregory and an English Earl; with walk on parts for the Virgin Mary and her donkey, circumcision and a leather sofa.

In 46 BC Julius Caesar reformed the Roman Republican Calendar. This was an utterly insane system with a basic year of only 355 days alternating with years of either 377 or 378 days. This was achieved by the extraordinary mechanism of inserting a leap month of 27 days after the 23rd (or sometimes 24th) of February and then scrapping the last 5 (or sometimes 4) days of February.

It’s hard to imagine a more confusing system, and what was worse it didn’t work. The power to insert a leap month was regularly manipulated by politicians as a way of lengthening, or more usually shortening, the period of office of their allies and rivals. So by the time of Caesar’s reforms the calendar was completely out of line with the seasons and he had to insert 3 whole leap months into 46 BC (one in February and two between November and December), and create a year that dragged on for 445 days.

The following year, 45 BC, marked the first of the Julian calendar. It began on 1 January and all the months were as we now have them giving a standard year of 365 days. Every 4 years a leap day was added, although bizarrely it wasn’t originally tacked onto the end of February but squeezed in between February 23rd and 24th so as to give two consecutive days both confusingly labelled 24th February.

Then in AD 325 the First Council of Nicaea met to thrash out the doctrine of the Trinity. By way of light relief they also standardised the method for calculating Easter (and while they were at it decided to ban a few things: notably the faithful from kneeling while praying on Sundays, and priests from practising self-castration). Because the crucifixion took place at the Jewish Passover festival, which is linked to the moon, the church decided to define Easter as the first Sunday, after the first full moon, after the spring equinox. Depending on where you are on the earth the date of the equinox varies slightly but it’s taken to be 21 March.

All went well until the 16th century when the Roman Catholic church, despite their uneasy relationship with scientists in general, and astronomers in particular, noticed that the actual date of the spring equinox was gradually moving away from the 21 March. This was alarming because it meant the church might be celebrating Easter on the wrong date. In 1582 Pope Gregory XIII introduced the Gregorian calendar which was designed to solve the problem. It made two changes to the Julian calendar. First it moved the calendar forward 10 days by skipping from the 4th to 15th October 1582 and then it changed the pattern of leap years by introducing a new rule that years divisible by 100 would only be leap years if they were also divisible by 400. So 1600 and 2000 would be leap years but not 1700, 1800 or 1900.

Unfortunately Europe was at the time embroiled in several centuries of Catholic v Protestant bickering so the Pope’s new calendar, which he’d modestly named after himself, wasn’t universally welcomed. In 1700, Britain, still using the Julian calendar, had a leap year, while much of Europe using the Gregorian calendar didn’t.

The Union of Scotland and England took effect in 1707 and because both countries were still on the Julian calendar, they did at least agree on the date of Easter. Unfortunately they didn’t agree on the date of New Year’s day.

The problem was that when Dionysius Exiguus (Dennis the short) sat down, in the 6th century, to ‘mis-calculate’ the year of Jesus’ birth, and introduce the AD system, he started the year on 25 March. The argument being that since Jesus was born without sin he must have been carried for exactly 9 months and therefore been conceived on 25 March. No doubt any mother who has ever been compelled to ride hundreds of miles on a donkey while heavily pregnant would confirm it’s a sure way of inducing an early birth. But never mind. England had adopted this ‘Annunciation’ style dating – so-called because 25th March was the feast of the Annunciation, the day the angel announced to Mary that she was with child. Meanwhile, back in 1600, Scotland had the good sense to adopt the rival ‘Circumcision’ style dating – 1 January being the day the church marks the circumcision of Jesus. An event the Scots continue to celebrate enthusiastically, if not entirely appropriately, by drinking, singing and dancing in the streets, and giving their neighbours lumps of coal.

The result was endless confusion because in England the day after 31 December 1749 was 1 January 1749, and the day after 24 March 1749 was 25 March 1750, New Year’s Day. For almost a quarter of each year England and Scotland knew what day and month it was but disagreed about what year it was.

To end this mess, and bring the UK into line with the rest of Europe the Earl of Chesterfield introduced the Calendar (New Style) Act of 1750. He was appropriately rewarded by having a leather sofa named after him.

The Act did two things. It changed England’s New Year from 25 March to 1 January and it introduced the Gregorian calendar. By this stage the UK, using the Julian calendar, was now 11 days behind much of Europe and on top of that the English didn’t know what year it was. So in England we were given a shortened 1751. It started on 25 March and ended on 31 December. Then throughout the UK 11 days were removed from the following October, and Wednesday 2 October 1752 was followed by Thursday 14 October.


Except for one small problem. The tax year was left completely unchanged: it still began on 25 March and stayed on the Julian calendar; only by now everybody else had gone Gregorian and what the taxman thought was 25 March was called 5 April.

So why does the tax year start on 6th April and not the 5th? Simple. In 1800 the taxman’s Julian calendar had a leap year and gained an extra day but the Gregorian calendar used by everybody else didn’t, so 5th April moved on to 6th. In 1900 it should have advanced another day to 7th April, but by then even the tax authorities had forgotten what rules they were supposed to be following.

The curious result is that we are stuck with a fiscal year that starts on a completely ludicrous date. In fact it’s dafter even than that. Corporation tax and government finances are based on the much more sensible financial years which start on 1 April. The fiscal year is only relevant for two main taxes: income tax and capital gains tax. Income tax didn’t exist in 1752; it wasn’t introduced until 1799. And capital gains tax wasn’t invented until 1965.


One comment

  1. Alan O'Brien says:

    Your explanation of the reason for the income tax year starting on 6 April is not, in my view, wholly correct. A day was not added in 1800 as can be seen from an examination of legislation around this date.

    In Pitt’s first income tax act (Duties on Income Act 1799) we have:

    ‘LXXII. And be it further enacted, That the first Assessment to be made by virtue of this Act shall be made for one Year, at the respective Rates before mentioned, from the fifth Day of April one thousand seven hundred and ninety-nine until the fifth Day of April one thousand eight hundred; and every subsequent Assessment shall be made for one Year, from the fifth Day of April yearly ‘.

    If your view was right the draftsman would have recognised the leap day in February 1800 by imposing the charge from the sixth of April 1799. At first sight the 1799 act exhibits an impossibility: you have two fifths of April. The answer in my view is that ‘from the fifth day of April’ means the year that begins on and includes the sixth April and does not include the opening fifth.

    Osborn’s Concise Law Dictionary has the following entry for ‘from’:

    ‘Subject to the context, it ordinarily excludes the day from which time is to be reckoned.’

    The Income Tax Act 1803 has this formulation:

    ‘CCX. And be it further enacted, That every Assessment to be made under this Act, within the Year appointed for making the same, shall be deemed to be for the current Year, and shall be in Force for such Year; and every Assessment made after the Expiration of any Year in which the same ought to have been made, through the Default of any Person, shall be deemed to be for the Whole of the Year current when the Assessment ought to have been made, and such Year shall commence from the fifth Day of April One thousand eight hundred and three, for the first Assessment, and for every subsequent Assessment during the Continuance of this Act, from the fifth Day of April in such Year;’

    The 1803 act does not refer to the closing fifth of April but it does similarly refer to a year ‘from the fifth’ and this confirms in my view that there was no change because of the addition of a leap day in 1800.

    You can see a similar approach before the calendar reform in, for example, the Land Tax Act of 1701 which provided that a charge of:

    ‘£1,484, 948 19s. 3¾d. to be raised in, One Year, from 25th March 1701; in Manner following . . . ’

    And this:

    ‘ . . . shall be raised levied and paid unto His Majesty within the Space of One Year from the Five and twentieth Day of March One thousand seven hundred and one ‘.

    [William III, 1700 & 1701: An Act for granting an Aid to His Majesty for defraying the Expence of His Navy Guards and Garrisons for One Year and for other necessary Occasions.]

    After the calendar reform the draftsman replaced from ’25 March’ with ‘from 5 April’.

    Dr Robert Poole examined the issues surrounding the calendar change in a paper published in 1995 called ‘Give us back our eleven days!’ and subsequently in a book: ‘Time’s Alteration’ (1998). Dr Poole is a Reader in History at the University of Cumbria. He isn’t, naturally, much concerned with tax and the sixth of April issue but he tells me he has examined old national accounts in the National Archives and found that they were made up to end on and include the old quarter days, including 25 March. He believes that this explains the origin of the sixth of April date. That is, if the accounts before the calendar reform ended on and included 25 March, it would follow that after the change they would end on and include 5 April.

    A brief article in Taxation in 2001 by John Jeffrey-Cook about the origin of the 6 April date takes the same line:

    ‘Following the church’s lead with tithes, the land tax and other taxes had been imposed for legal years ending on 25 March. So that the land tax year of assessment 1752-53 covered a full 365 real days, it began on 26 March 1752 but finished on 5 April 1753.’

    Your statement that the taxman continued to use the Julian calendar while the rest of the country used the Gregorian, so the former had a leap year in 1800 and everyone else did not, seems unjustified. What evidence do you have? I know some older books give this explanation but I believe they are wrong.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.