Council of Economic Advisers: Chair's report 2016-2018

An overview of how the Council operates, and the areas they focused on from 2016 to 2018.


5 - Workstream 3: Fiscal Matters

Remit:

  • To provide expert advice to the Scottish Government ahead of each Draft Budget on the revenue risks and possible mitigating actions associated with a rise in the Additional Rate (now termed 'Top Rate') of income tax in Scotland from 45p to 50p, and
  • To advise on wider fiscal matters should that be sought by the First Minister or Cabinet Secretary for Finance and the Constitution.

Key activities: advice on Scottish Government analysis; review and sign-off of the Scottish Government's analytical paper published prior to Draft Budget 2018-19.

The amount of fiscal autonomy devolved to Scotland has increased since the Scotland Act 1998 and the establishment of the Scottish Parliament. In 1999, the only tax powers devolved were local taxation and the ability to vary the basic rate of income tax (known as the Scottish variable rate). The Scotland Act 2012 devolved two further taxes: the Land and Buildings Transaction Tax (LBTT) and the Scottish Landfill Tax (SLfT). Also introduced was the Scottish Rate of Income Tax, whereby the UK Government reduced the three rates of income tax by 10p (e.g. the basic rate fell from 20p to 10p) and the Scottish Parliament could then choose whether to increase these rates. Income tax policy was the same in Scotland and rUK until 2017-18, at which point policy began to diverge.

More recently, due to the Scotland Act 2016, the Scottish Parliament has the power to set all the rates and bands for non-savings non-dividend (NSND) income tax in Scotland. The tax-free personal allowance, and reliefs and other allowances, remain reserved to the UK Government. Prior to the 2016 Scottish Parliament elections, the Scottish Government announced its proposals for Scottish income tax in 2017-18. This announcement included the publication of an analytical note in March 2016 [29], which concluded that there were significant risks to revenues due to potential behaviour effects, were the additional rate of income tax in Scotland to be set at 5 pence above the corresponding UK rate. The Government's conclusion was that the additional rate in Scotland should remain at 45p. However, the Scottish Income Tax system did diverge from that of the rUK as the Scottish and UK parliaments took different decisions on uprating the bands. At that time, the First Minister announced that the Council of Economic Advisers would be invited to update this analysis on an annual basis to inform decisions in future budgets.

In January 2017, the Scottish Government provided Council members with a comprehensive overview of the existing evidence base - including initial analysis of the additional rate taxpayer population based on detailed outturn data - and set out potential avenues for future research. Following advice from the Council, the research agenda was narrowed down to the following three broad aspects:
A. Further analysis of Scottish data
B. On-going engagement with the academic literature
C. Deepening understanding of potential behaviours

Over the course of 2017, the Council reviewed a series of further analytical papers addressing each of these three elements, and provided further advice and comments. These discussions led up to the publication of an analytical note [30] by the Scottish Government in December 2017. The minutes of the January 2017 meeting and the conference calls over the course of 2017 summarise our comments on these papers, as well as areas we felt should be explored further.

Please note: this chapter references the analysis presented to the Council in 2017 and early 2018, since then some figures may have changed due to new data being published.

5.1. Context

Additional Rate Taxpayers in Scotland

In Scotland, taxpayers who pay the additional rate (AR) of income tax account for a disproportionate share of income and tax liabilities despite being a relatively small proportion of the taxpayer population. It was expected that around 20,000, or less than 1 per cent of Scottish adults, would pay the AR of income tax in 2018-19. However, their contribution to total income tax liabilities would account for around 19 per cent of the total tax take in 2018-19. This means that total revenue from income tax is relatively reliant on these top earners and that the behaviour of a small number of individuals may have a disproportionate impact, positive or negative, on overall Scottish income tax liabilities.

Figure 6: Estimated Contributions to Income Tax Liabilities by Taxpayers and Band, 2018-19
Figure 6: Estimated Contributions to Income Tax Liabilities by Taxpayers and Band, 2018-19
Source: Scottish Government

This dependency of income tax revenues on high income earners is not unusual in advanced economies, however, it is worth noting that AR taxpayers make up a smaller proportion of the taxpayer population in Scotland than they do in the UK as a whole.

The data reviewed by the Council for the AR taxpayer population was for 2014-15, the latest available at the time.

Possible Responses to a Change in the AR of Income Tax

The existing income tax rates, thresholds, exemptions, reliefs, and allowances determine an individual's tax liability. There is uncertainty associated with estimating the potential revenue raised from changing income tax for very high earners. This reflects the uncertain behavioural response relating to the tax policy changes that could either erode the revenues raised from a tax rise or lessen the cost of a tax cut. Potential behavioural responses may include:

  • Avoidance - artificially (but legally) reducing one's tax liability; e.g. incorporation, where individuals form companies to minimise tax;
  • Evasion - which illegally reduces tax liabilities; e.g. failing to declare income;
  • Economic responses - such as individuals choosing to seek work or increase hours worked; and
  • Migration - taxes could also affect migration, both into and out of Scotland.

The Council considered the potential behavioural responses specifically relevant to Scotland. A divergent Scottish income tax from that in the rest of the UK would create new incentives for behavioural responses or may affect incentives to locate in Scotland.

Taxpayers' responsiveness to changes in income tax policy are usually estimated through taxable income elasticities (TIEs) which measure the percentage change in taxable income in response to a one per cent change in the net of tax rate. The more responsive taxpayers are, i.e. the greater the TIE, the larger will be the change in their taxable income and hence tax receipts.

While the magnitude of behavioural responses is a matter of debate, there is a general consensus that those with the highest incomes have the greatest incentives to change their behaviour as they pay a higher level of tax. Top earners will also have greater means to change their behaviour, for example the money and connections to access sophisticated and expensive avoidance schemes. Whilst significant changes in behaviour may be limited to a small number of high income individuals, these individuals pay large amounts of tax revenue, and so present disproportionate risks or opportunities for tax revenues.

5.2. Discussion

Overview of the Analysis Undertaken

The Scottish Government's analysis of behavioural impacts follows the standard approach adopted by HMRC and other fiscal institutions. It uses taxable income elasticities (TIEs), informed by discussions with a number of institutions and a review of the academic literature.

While Council Members approved of this standard approach to modelling behaviour change, their key recommendation was to broaden the existing evidence base in order to refine estimates of TIEs and capture Scotland-specific circumstances by considering the three research avenues set out in the introduction. Each of these is discussed in further detail below, alongside the Council's recommendations.

A: Further analysis of Scottish data

Firstly, Council members were interested in exploring the structure, composition, sector of activity, and residency of the existing additional rate tax base in greater depth, where possible. Scottish Government analysts therefore undertook further analysis of detailed micro-data from the 2014-15 Survey of Personal Incomes. This showed that there are relatively more AR taxpayers who are self-employed, in both Scotland and across the UK, than in the taxpayer population as a whole.

AR taxpayers in Scotland are predominately male, although there has been a marginal improvement in the share of female AR taxpayers over the year to 2014-15. In addition, Scottish AR taxpayers are also more prevalent in certain sectors. As expected, the highest number of AR taxpayers can be found in Professional Services - which includes activities, such as legal, consulting and accounting services as well as engineering and architectural services - and Financial Services and Insurance This could mean that these taxpayers are highly mobile as these sectors tend to compete for global talent.

B: Engagement with the latest academic literature on TIEs

Whilst the academic literature on the behavioural response to income tax changes is vast, there is no overall consensus on the magnitude, as estimates of TIEs vary widely across countries and time and are often focused on national tax reforms.

Council members suggested in January 2017 that reviews of the existing evidence base could be updated, focusing on regional tax competition in federations, such as the United States, Canada, or Switzerland.

In response to Council Members' suggestions, Scottish Government analysts collaborated with the University of Edinburgh [31] to undertake a refresh of the existing evidence base on taxpayers' responsiveness. This included new material published by the Institute for Fiscal Studies in August 2017 [32], alongside a review of empirical studies on tax induced cross border migration in fiscal federations. The academic literature review was complemented by a quantitative study of the behavioural response of Scottish and UK taxpayers to the introduction of the 50p rate in April 2010.

Based on advice by the Council, the Scottish Government also reviewed the limited, but growing, number of empirical studies that attempt to quantify the impact of changes in taxation on people's relocation choices, both in relation to mobility across different countries and mobility across different regions or states within countries. The majority of these studies are focused on (semi) migration elasticities, which measure the percentage change in the number of people migrating in response to
a 1 percentage point change in the average tax rate – a slightly different definition
to the standard TIE. This is because migration decisions are thought to be based on total take home pay rather than marginal tax rates.

On balance, the empirical evidence is inconclusive but points towards potentially small, tax induced migration responses across state and regional borders for top earners in the United States and Spain, with estimates of the percentage of taxpayers migrating in response to tax changes ranging from close to zero to 0.23. These studies suggest that standard TIEs - which often only consider nationwide changes in income tax - may have to be uplifted, albeit by a small amount, to account for differential tax rates across regions or states and the larger behavioural response associated with this.

One analytical contribution of particular relevance was the August 2017 Institute for Fiscal Studies (IFS) publication mentioned above. The IFS investigated how high income earners in the UK responded to the introduction of the 50p rate in April 2010. The researchers concluded that, despite careful analysis, it has proven impossible to obtain precise or robust estimates of the responsiveness of top earners in the UK. Instead, different methodologies, data sources, and assumptions result in estimates of the TIE that range from 0.31 to around 1. The IFS considered HMRC's TIE of 0.48 as a reasonable central estimate for policy-costing purposes.

The 2017 IFS study also showed that there is variation within the AR taxpayer group: taxpayers just above the £150,000 AR threshold appear to be less responsive to tax changes than the AR taxpayer group as a whole.

The Council also indicated that we were interested in further analysis on the distinction between short run and long run behavioural responses. Although some evidence suggests that TIEs decrease over time, more recent research has indicated TIEs are likely to be higher over time as the costs associated with reducing taxable income are lower; meaning a more pronounced behavioural response over the longer term.

The Council's input also guided the analysis considering the impact of the rise in the AR on businesses and the wider economy.

C: Deepening understanding of potential behaviours

Council Members suggested that Officials could investigate one particular channel of avoidance highlighted by the Office for Budget Responsibility (OBR) and the Institute for Fiscal Studies (IFS): the risk that tax motivated incorporations may have on non-savings non-dividend (NSND) liabilities. Research on this issue is relatively scarce. However, analysis undertaken by the Institute for Fiscal Studies [33] suggests that the tax related benefits of individuals incorporating businesses are greatest for high-income earners. This is of particular relevance to Scotland's public finances, as taxpayers who incorporate would pay tax on earned income to the Scottish Government but pay corporation tax, capital gains tax and income tax on dividends to the UK Government.

The risk that incorporations pose to NSND liabilities, and hence the Scottish Budget, depend on two factors:

  • 1. The number of people incorporating and whether trends differ in Scotland and the rUK; and
  • 2. Their income before incorporating and profits after incorporating, which determine the amount of income tax liabilities the Scottish Government would forego, on average, if these individuals no longer paid NSND income tax.

In response to the Council's recommendation, Scottish Government analysts undertook further analysis. This showed that there appears to have been relatively stronger growth in the number of people incorporating in Scotland, relative to the UK, since 2014 [34]. However, it is not clear what is driving this trend and whether it is likely to persist. Structural factors may be at work, such as underlying sectoral shifts and challenges in the oil and gas industry.

In addition, individuals most likely to pursue incorporation have higher income levels in Scotland than in the UK. This could mean that there is a relatively stronger incentive for Scottish taxpayers to incorporate to save tax even before any future changes in income tax policy, such as a rise in the Additional rate, is considered. Whilst the data sources point towards strong growth in incorporations in both Scotland and the UK, it is impossible to say how much of this is entirely due to tax reasons. A 2014 HMRC survey, as well as research by the Institute for Fiscal Studies [35], indicate that at least some of this behaviour is a direct result of differential tax treatment. As such, increasing the AR, now termed the 'Top Rate', to 50p would further exacerbate the revenue risk.

Council members welcomed this new evidence, which illustrated that in reality, taxpayers' responses are likely to be nuanced. The analysis also highlighted the interdependencies between income tax and the wider UK tax system, an element not fully captured by some of the national studies.

Mitigating Strategies for Legal Avoidance Behaviour

In addition to advising on revenue impacts, the Council also discussed ways in which the Scottish Government might mitigate against revenue loss arising from both legal avoidance behaviour and illegal evasion behaviour.

i. Incorporation

As mentioned above, the difference in tax liabilities between different taxes levied (e.g. income tax vs. corporation tax) is often cited as a rationale for incorporation – where an individual may be able to pay lower taxes by switching remuneration from a salary from an employer to either dividends or profits from a business. As this behaviour is legal, and the Scottish Government has limited tax powers, it would suggest that the Scottish Government could be limited in the actions it could take to prevent or mitigate against it. However, it should be noted that tax is only one component of a complex decision to incorporate.

ii. Forestalling

If the intention to increase the AR was announced in advance of implementation (e.g. in a Draft Budget) there may be a one-off effect, as individuals look to bring forward income into the financial year before the increase takes effect. This will positively affect tax revenues in the preceding year, and negatively affect them in the implementation year. However, Scottish income tax powers relate to NSND income only and it is much more difficult to bring earned income forward or postpone salary payments or pensions.

Since forestalling largely moves income tax receipts across years, the net loss to the Scottish Budget is likely to be fairly limited. However, due to the way the Fiscal Framework currently operates, the Scottish Government would see the full loss in income tax receipts from forestalling in 2018-19 but would not receive any potential uplift in 2017-18 receipts until re-conciliation occurred in Autumn 2019 (for inclusion in the Draft Budget 2020-21). Forestalling behaviour therefore represents a short-term risk to the Scottish Budget unless this impact can be mitigated through other means.

Few direct actions can be taken by the Scottish Government to mitigate the potential revenue loss. The late announcement of tax changes may reduce potential forestalling behaviour by reducing the amount of time that affected individuals would have to respond. However, this could contradict key principles of the Scottish approach to taxation relating to providing certainty and engaging with stakeholders.

iii. Cross Border Migration

A divergence in the AR could incentivise AR paying individuals to move to a nearby jurisdiction to minimise their tax liability. As income tax is collected on a residence basis, anyone who already owns, or has the wealth to acquire, a property in a different tax jurisdiction is more likely to be able to undertake this behaviour. When considering a 50 per cent tax rate, it is possible that some AR taxpayers could shift their circumstances so that they become rUK taxpayers (which currently have an AR of 45 per cent). Under this circumstance, the Scottish Government would lose all the income tax revenue from these individuals.

Again, the Scottish Government will not have any formal mechanisms to mitigate against this behaviour, under the current or any future constitutional settlement. However, as with incorporation, there are likely to be many elements involved in a decision to relocate, of which tax will be just one. As well as the tax considerations, there are many other elements of Scottish life that will likely factor into an individual's decision on location. Therefore, work by the Scottish Government to make Scotland an attractive place to live and work can be an important, if indirect, mitigation strategy against any cross border migration motivated by differing AR tax policies across the UK.

iv. Labour Supply and Retirement Decisions

As mentioned, increases in the AR affect both marginal and average tax rates for AR taxpayers. These could lead to individuals reducing the number of hours they work, as the reward for working has reduced. This could lead to people leaving the labour market (e.g. retirement) or an individual choosing to work fewer hours (or to not increase their hours) to balance their tax liability against working and earning income. It is also possible that some people will choose to work longer hours, to make up for the additional income that they will pay in tax.

Again, although the Scottish Government will not have any direct mechanism to mitigate this, it is likely that making formal arrangements with employers to reduce hours will be costly for the individual both to implement and reverse (in terms of time etc.).

Mitigating Strategies for Illegal Evasion Behaviour

In addition, there are a number of illegal tax evasion schemes. Under the terms of the Scotland Act (2016), and associated Fiscal Framework, the collection and management of income tax remains the responsibility of HMRC. This includes the management of risks to tax revenues from illegal tax evasion. HMRC are committed to providing the same tax enforcement and compliance standards to Scottish income tax as to that in the rest of the UK and have an extensive programme of enforcement and compliance work across the UK tax base and report on this regularly.

Finally, there are indirect actions that the Scottish Government can take to discourage illegal tax evasion, and promote a strong and consistent approach to tackling this issue. For example, the recent Scottish General Anti- Avoidance Rule (GAAR), introduced as part of the Revenue Scotland and Tax Powers Act (2014), allows Revenue Scotland to take counteraction against tax avoidance arrangements considered to be artificial, even if they otherwise operate within the letter of the law. Whilst this only applies to the fully devolved taxes and not Scottish income tax, it still contributes to a wider culture of tax compliance.

Revenue Risks of Changes to the Additional Rate of Income Tax from 45p to 50p

Following a re-consideration of the evidence base and Council advice over 2017, Scottish Government analysts refreshed their TIE estimates. The behavioural analysis used in the November 2017 publication adopted a range of 0.35 to 0.75 for the TIEs of those earning more than £250,000 and slightly smaller ranges for taxpayers earning less than that.

The uncertainty in estimating the revenues from a 5 pence rise in the AR, based on the new TIEs, is illustrated in the below chart. If top earners did not change their behaviour in response to the rise in the AR, the policy is expected to raise around £145 million in Scotland in 2018-19. With a low level of behavioural responsiveness, the policy is forecast to raise £53 million, whilst with a high level of behavioural response, the policy could potentially result in a £24 million loss in revenues for the Scottish Government.

Figure 7: Potential Revenue from 5 pence rise in Additional Rate 2018-19
Figure 7: Potential Revenue from 5 pence rise in Additional Rate 2018-19
Source: Scottish Government

The evidence presented in the December 2017 paper relates to a 5 pence change in the AR. However, the analysis also suggested that changes below this level will result in a proportionately lower behavioural response, as at the margin the differential relative to the rUK is lower. It is reasonable to conclude that a smaller divergence in tax rates will create a lower incentive for taxpayers to change their behaviour. The Scottish Government policy is outlined in the text box below.

Scottish Government Policy Action

On 14 December 2017, the Cabinet Secretary for Finance and the Constitution set out the Draft Budget 2018-19 and his proposals for income tax policy. These were later amended as part of the Stage 1 process. Table 2 illustrates the new tax bands and rates which came into effect in April 2018. The Additional Rate of Income Tax, now termed the Top Rate, sits at 46%.

Table 2: Rates and bands for 2018-19

Name Rate Band
Starter Rate 19% Over £11,850* - £13,850
Basic Rate 20% Over £13,850 - £24,000
Intermediate Rate 21% Over £24,000 - £43,430
Higher Rate 41% Over £43,430 - £150,000**
Top Rate 46% Above £150,000**+

*Assumes individuals are in receipt of the Standard UK Personal Allowance
**Those earning more than £100,000 will see their Personal Allowance reduced by £1 for every £2 earned over £100,000.
Source: Scottish Government

It is important to note that forecasting responsibilities now sit with the Scottish Fiscal Commission and their forecasts, including their assessment of the extent of taxpayers' behavioural responses, determine the amount of resources the Scottish Government can draw down from HMRC.

Table 3 below compares the assumptions about taxpayers' responsiveness in the Scottish Government's analytical paper and the assumptions used by the Scottish Fiscal Commission in their Economic and Fiscal Forecasts [36] for the Draft Budget 2018-19.

Table 3: Overview of TIEs: Comparison of assumptions underpinning the SG Paper (informed by CEA advice) and the SFC forecasts

Income ranges SG Paper: Low Responsiveness SG Paper: High Responsiveness SFC Assumptions (SFC report, Table 3.10 )
From To
150,000 200,000 0.25 0.35 0.35
200,000 250,000 0.35 0.45 0.35
250,000 300,000 0.35 0.75 0.35
300,000 500,000 0.35 0.75 0.55
500,000 1,000,000 0.35 0.75 0.75
1,000,000   0.35 0.75 0.75

Source: Scottish Government

Reflections and Recommendations for Fiscal Matters

Officials have carried out detailed and rigorous assessment, using available data, of the revenue impact of a five pence increase in the additional rate of income tax. The Council commend the quality and thoroughness of the analysis, as well as the responsiveness to constructive comment provided by members. Going forward, we recommend analysing the impact of changes to income tax announced in the Draft Budget 2018-19 (and the broader fiscal landscape) on taxpayers' behaviour, when data becomes available though this is unlikely to be the case until May 2021. Maintaining an awareness of how these changes potentially affect the whole economy is important, as well as ensuring transparent clarity about these potential impacts on income taxpayers.

When carrying out analysis on income tax, it should be highlighted that this tax forms only one of a wider basket of taxes, most of which the Scottish Government do not have devolved power over.

Officials are encouraged to continue working in partnership with the Scottish Fiscal Commission, given their role in forecasting the impact of tax policy changes. In particular, we encourage further investigation of the behavioural impacts of a tax change in the devolved context and opportunities for Scotland, and the potential to facilitate positive behavioural change.

In addition, we recommend officials consider the broader economic context and economic cycle when scrutinising the impacts of tax structures and changes.

Contact

Back to top