Taxation of cross-border commuters
University of Stirling
One of the reasons that Professor Andrew Hughes-Hallett uses to claim that the Scottish budget will be stronger with independence is that there are “repatriated taxes from cross-border commuters (£1 billion)”. In a later post he then goes on to explain his reasoning “My figures on taxes repatriated from cross-border commuters come from a 2008 report by Oxford Economics”. This seems to be a reference to a series of reports that were undertaken by the City of London to show London’s importance in the UK economy. These reports break down tax receipts on a residence and on a workplace basis for different parts of the UK. Commuting causes differences in tax receipts at place of work relative to place of residence.
Oxford Economics use income tax data on a residence basis derived from the HM Revenue and Customs (HMRC) Survey of Personal Incomes (SPI). This is combined with the Annual Population Survey (APS), from which employment can be derived on both a workforce and residential basis. These data can therefore be used to compare workers who claim to be working and living in Scotland, with those living in Scotland and working elsewhere. If more workers say that they are resident of Scotland than work in Scotland, this suggests that there is a commuter effect – a discrepancy between the location of tax revenues and place of residence. The LSE methodology is less clear. In 2006-07 they get similar results for the commuting effects in respect of National Insurance and income tax.
Table 1: Results from City of London reports, difference between Scotland residence based and work based. 2005-06 to 2010-11
|Total of all taxes
|NI and income tax
The commuting effect increased between 2005 and 2011. The reason seems to be that work based tax revenues have decreased more than residence based revenues since 2007-2008. The City of London reports have stopped and the publicly-available SPI has not been updated since 2009, meaning that it is not possibly to update this research. However, estimates of the commuting effect can be inferred from ASHE (Annual Survey of Households Earning) data. One caveat on the ASHE data is that it does not include the self-employed. The Labour Force Survey (LFS), was also used to construct commuting effect estimates and the results are compared in table 2. It seems quite clear that it would take much more commuting to get a £1bn increase in income tax.
Table 2: Comparing ASHE results to LFS to necessary number to receive £1bn of income tax
||LFS Quarter 4 2013
||To get £1bn of income tax
|Net income (£Million) of those people who commute
|Net number of people who commute and work outside Scotland
The ASHE data suggest that the net effect of the income of commuters and residents is larger than for just residents in most years; however that has varied across the past 10 years. 2002 and 2007 stand out as years where rUK-resident workers in Scotland paid more in tax than Scottish resident workers in rUK. This implies that the income of rUK commuters into Scotland was greater than the net income of Scottish residents working in rUK.
Would this level of commuting be able to justify an extra £1 billion of revenues? According to Government Expenditure and Revenue in Scotland (GERS) total income tax raised in Scotland is £10.8bn while national insurance receipts are £8.5bn. An extra 9% of income tax would have to be raised by the 1% of workers who work outside Scotland but live in Scotland. The wages of those workers, according to ASHE is only £530 million, while according to my own calculations £5,200 million would be needed or approximately 10 times the current level. Additionally, it is unclear how such taxes might be repatriated, it is unlikely that the tax regimes of the countries that the workers commute to, would view this as a favourable relationship. In fact, regular workers in London, on just one of the tests, would automatically taxed by the rUK, if the same rules applied. It also seems that the Scottish white paper on independence, would agree with this definition and that only the proportion of workers from Scotland, who spend less than a qualifying period of time in UK, would be liable for Scottish taxes, on income and insurance contributions.
Who would be liable to pay Scottish taxes?
According to the Scottish Government white paper on independence, an independent Scotland would build on the existing definition set out in the Scotland Act 2012 and general international protocols to establish a definition of a Scottish taxpayer based on residence. In general, this means that people who live in Scotland for most of the year will pay their taxes here.
The Scotland Act 2012 – Section 80F says that to be considered Scottish for tax purposes:
“The number of days in the year on which an individual is in Scotland ……equals or exceeds the number of days in the year on which an individual is in any other part of the UK .. “
Additionally if Scotland was to become a member state of the EU there are some general conventions that might apply:
Each country has its own definition of tax residence; yet: you will usually be considered tax-resident in the country where you spend more than 6 months a year and if you spend less than 6 months a year in another EU country, you will normally remain tax-resident in your home country.
Overall, it seems that there is no clear trend in the number of Scottish resident workers who commute. It is also unclear which of these workers would be taxable in Scotland and how large would be the resultant revenues. The number of workers seems to balance over recent years, to a fairly low number. Will this relationship continue to hold? If there are higher levels of unemployment or low wages in Scotland this might lead workers to commute more from Scotland. Alternatively if Scotland becomes more desirable due to higher incomes, then more workers will commute in. Under these circumstances, it is difficult to see that the net revenue to the Scottish Treasury from cross-border commuters would amount to £1bn.
 Not a City of London report
 Wages in the last quarter of the LFS might be skewed
 Author’s calculations