Does Nicola Sturgeon’s stance on the 50p tax rate undermine the case for tax devolution?
To increase the additional rate or not to increase the additional rate? That is the question that has been placed centre stage of the Scottish election, ahead of the full devolution of income tax bands and rates to the Scottish Parliament in 2017.
The Additional Rate is paid by those with incomes over £150,000. The additional rate was introduced at 50% in April 2010, having been announced by Alasdair Darling in the 2009 Budget, before being reduced by George Osborne in the 2013 Budget to 45%.
The SNP administration has on several occasions affirmed a commitment to raise the additional rate back to 50%. In the lead up to the 2015 General Election they pledged to support a 50p additional rate. More recently, they have been critical of their existing income tax powers which allow them only to vary all three income tax bands in lockstep – so the Additional Rate can only be raised to the same extent that the basic rate is. They have indicated that they would raise taxes on the higher paid only when they can do so without increasing basic rate tax. A commitment to raising the tax burden of the highest paid has always been part of their social democratic mantra – and the basis for criticism of UK Government tax policy.
Scottish Greens publish their plans for income and property tax
The tax debate needs to move beyond the narrow confines of incremental rises and falls
Public sector workers face pay cut under new state pension rules
But now, having gained the power to vary the Additional Rate from next year, the Scottish Government has balked at the idea. The reason? A fear that ‘behavioural effects’ – the responses of top tax rate payers to the increase – might mean that Scottish tax revenues could actually fall. The key behavioural effect that the Scottish Government are concerned with is the idea that a higher tax rate might result in outmigration of top earners to England, if rates remain lower there.
In FMQs, Nicola Sturgeon said that raising the Additional Rate to 50p would be ‘reckless’ and ‘daft’. The SNP still support a rise to 50p in principle, but only if this is matched south of the border.
Opposition parties had a field day, pointing out the inconsistency in arguing for devolution of tax powers if those powers will only be used to mirror UK Government policy. Some of Sturgeon’s supporters were also critical of the policy stance. The Scottish Government has previously argued for a higher top rate in order that ‘those with the broadest shoulders should bear a slightly larger share of the burden’; what does it say about a party’s principles if it is prepared to renege on a position of principle because that might have a cost in terms of lost revenue?
What is clear is that raising the Additional Rate will not raise a great deal of revenue. There are around 17,000 Additional Rate taxpayers in Scotland. Even with an average salary of £270,000, raising the tax rate on income above £150,000 by 5% would raise £102m, less than 1% of what is spent on the NHS in Scotland. But the big question up for debate is how these taxpayers would respond to a higher rate of tax in Scotland.
Evidence from the UK
Incomes respond to tax changes in a variety of ways. People might vary the amount they work, or avoid tax in any number of ways, including taking advantage of the large number of tax reliefs available in the UK. At UK level, the effect of tax changes on migration (into and out of the UK as a whole) is likely to be dwarfed by other types of behavioural response.
But evidence on how people respond to Additional Rate tax changes is limited. This is partly because there is never any counterfactual against which we can evaluate the impacts of policy change. But it is also because previous changes to the Additional Rate have been pre-announced.
The 50p was announced by Alasdair Darling in 2009, but did not come in to force until the following year. Similarly, the reduction from 50p back to 45p was announced by George Osborne in March 2012 but did not come into force until 2013. As a result, top earners brought forward income from 2010 to 2009, and delayed receiving income from 2012 until 2013. In retrospect therefore, incomes (and tax receipts) appear disproportionately low when the Additional Rate was at 50p, and disproportionately high when the Rate has been at 45p. This has enabled George Osborne to make outlandish claims about how much revenue has been ‘raised’ by cutting the Additional Rate.
With a lack of concrete evidence, the Treasury’s assumptions about the revenue effects of changing the Additional Rate in the UK come with large confidence intervals. If there had been no behavioural response, the Treasury estimates that cutting the Additional Rate to 45p in 2013 would have reduced revenues by £3.5bn. After factoring in its best estimate of the likely behavioural response, the Treasury estimate that the tax cut cost just £100 million. But the possibility that the tax cut either increased revenues slightly or reduced them are also within reasonable confidence limits.
So at UK level, HMRC’s working assumption is that raising the Additional Rate will raise little additional revenue, or possibly cost revenue – once the behavioural effect is factored in.
Would a Scottish response be stronger?
The fear of Scottish politicians is that some of these behavioural responses could be more marked if tax rises are implemented in Scotland, but not in the rest of the UK (rUK).
How well placed is this fear?
Firstly, it is important to realise that much of the income response of Additional Rate taxpayers to tax changes in the UK comes not from changes in employment income, but from changes in dividend income. This is important in the Scottish context because the Scottish Government’s income tax powers apply only to non-savings, non-dividend income. In other words, even if the Scottish Government increases the Additional rate to 50p, Scottish taxpayers will still pay the UK Additional Rate on savings and dividend income.
Of course, there is the possibility that some top earners might be able to convert employment income to dividend income to take advantage of a lower UK rate. There is probably some truth to this, but for many this option is limited.
Might Scottish taxpayers physically relocate to England on the basis of a slightly higher additional rate in Scotland? In a majority of cases, this seems unlikely. For someone earning £300,000, a 5p rise in the Additional Rate would cost around £7,500. A lot of money perhaps, but enough to go through the upheaval of moving? Clearly this will depend on individual circumstances. It will also depend on whether the difference between Scottish and rUK tax rates are anticipated to remain in the longer term.
But there is some evidence that differences in tax rates within a country can be the cause of income-shifting (Canada) and relocation (Switzerland). Higher tax rates may discourage in-migration of high-earning individuals from other countries more than they incentivise existing residents to leave. Indeed this latter effect (disincentivising potential future in-migration to Scotland) is probably the bigger risk than the possibility of outmigration of existing residents.
Uncertainty as a rationale for inertia
In summary, the best evidence at UK level is that a rise in the Additional Rate will raise little additional revenue once one takes account of behavioural effects. Behavioural responses to unilateral tax changes in Scotland could be slightly stronger – on the basis of migration possibilities – but equally slightly weaker – on the basis that non-savings, non-dividend income will remain subject to the UK rate. The overall impact is highly uncertain.
One of the rationales for devolution is the ability to undertake policy experimentation – what economists sometimes refer to as ‘laboratory federalism’. It is understandable that the Scottish Government does not want to make a guinea pig of itself, but the only way realistically we will improve our understanding of the behavioural response to unilateral Additional Rate tax changes in Scotland is to implement the policy and see what happens.
There were always going to be constraints on the extent to which devolved fiscal policy in Scotland could diverge from that in rUK. But Nicola Sturgeon’s suggestion that she would only countenance any change to the Additional Rate should it be matched in rUK does seem to undermine the case for fiscal devolution. It also weakens the case for independence, given the level of fiscal interdependence between countries in an increasingly globalised world. (Although if Scotland was independent it could tighten up on income tax reliefs, the proliferation of which contributes to tax revenues being more responsive to tax rate changes in the UK than in many other countries).
As the Smith Commission recognised with its infamous ‘no detriment’ principles, virtually any policy decision by one government can have knock on effects for the other. But as Scotland’s Fiscal Framework makes clear, these behavioural effects have to be accepted as part and parcel of the devolution of policy responsibility.
Is it reasonable for Nicola Sturgeon to say that she supports a 50p Additional Rate but will only implement this policy if it is implemented in rUK? This depends in part on what the main motivation for supporting a 50p rate is. If the main motivation is to raise revenue, then increasing the Additional Rate to 50p won’t be a very effective policy – and it is unclear how much more or less effective this policy would be if it was implemented in Scotland but not in rUK. But if the main motivation is to achieve a (modest) reduction in income inequality, then perhaps this policy objective should be thought of in the same way that most other policy objectives are – it implies a fiscal cost (although in this case, a highly uncertain one).
Whether a future Scottish Government thinks that the risk of a fall in revenue is a risk worth taking will depend on how strongly it believes in the principle of a 50p Additional Rate as part of its wider social democratic objectives.
David Eiser is a Research Fellow at the University of Stirling. This blog first appeared on Democratic Audit Scotland on 29th March 2016