Devolution (Further Powers) Committee, 11 December 2014 - Evidence from Professor David Bell and Mr David Eiser, Stirling University
The Draft Scottish Budget 2018-19
It is a pleasure to introduce the Draft 2018-19 Budget to the Scottish Parliament. This will be the first year in which the full proposals of the Smith Commission can be implemented. It has taken three years to put these proposals in place.
Thus, following the Scotland Act 2015, the Scottish Parliament is now able to set both the bands and rates of income tax; it can set Air Passenger Duty; it will receive a share of VAT revenues raised in Scotland. Together these measures will take Scotland's revenue raising capability to around £26 billion this financial year.
The potential costs and benefits to the Scottish budget of differences in macroeconomic performance have been amplified by the transfer of full income tax powers under the Scotland Act 2015 and their implementation in the current financial year. Any difference between Scottish and UK rates of tax revenue growth will have a larger consequence for the Scottish budget now than under the Scotland Act 2012 because the block grant adjustment has been increased in line with the increased revenue associated with the extended income tax powers.
The relative changes in the growth of tax revenues in Scotland and rUK since the introduction of the Scottish Rate of Income Tax (SRIT) have been small compared to the size of the Scottish Budget, and therefore easy to accommodate. The decision to hold SRIT at 10p has also maintained broad equality between the budgets that have been set under the Scotland Act 2012 and the outturn that could have been expected under the Barnett Formula.
VAT revenues will now be shared with the rest of the UK. Though common in other countries, this form of revenue allocation is relatively new to the UK. It does not significantly increase the accountability of the Scottish Parliament because the Parliament will have no control over the setting of the tax base (including exemptions such as food and children’s clothing) or the rates of tax. Thus, it will not add significantly to the incentives for the Scottish Parliament to improve Scotland’s economic performance.
Reductions in Air Passenger Duty might have given Scotland a competitive advantage in air travel. However, the change implemented by the UK Government in 2015, whereby children were exempt from the tax, has perhaps reduced the potential for competitive advantage. Whereas business air travellers are probably not very price sensitive, the holiday trade, which inevitably involves children, is more likely to respond to differences in the costs of air fares.
The Scotland Act 2015 has also given Scotland some control over some welfare spending. The new spending responsibilities are grouped in three main categories covering (1) disability benefits, (2) those benefits covered by the so-called Regulated Social Fund and (3) discretionary housing payments. The specific benefits are:
1) Attendance Allowance, Disability Living Allowance/PIP, Carer's Allowance Industrial Injuries Disablement Allowance and Severe Disablement Allowance.
2) Cold Weather Payment, Funeral Payment, Sure Start Maternity Grant and Winter Fuel Payment.
3) Discretionary Housing Payments
Transfer of these benefits will add around £2.5 billion to Scotland's spending power in 2018-19. There is a recent precedent for the transfer of welfare powers from the UK government to local government (in the case of Scotland to the Scottish government). That was due to the "localisation" of council tax benefit which took place in 2013-14. One argument in favour of this change was that councils had previously no incentive to reduce claims for council tax benefit, since these would always be met by the Department for Work and Pensions.
The change to the administration of council tax benefit was effected by a transfer from the DWP Annually Managed Expenditure (AME) to local government funding which is drawn from the Departmental Expenditure Limit (DEL) budget. One important implication of this change is that, as part of DEL, council tax benefit now has to compete within local government budgets against other priorities.
A similar arrangement has been agreed with the UK Government for the transfer of the new welfare responsibilities to Scotland. This has created a substantial difficulty for the Barnett Formula (which operates only on DEL budgets) since these benefits remain part of AME expenditure at UK level. However, a solution has been found that is in a sense a variation on the Barnett Formula[i]. The Scottish Government will receive a transfer into its DEL budget of an amount equivalent to 2018-19 DWP spending on the relevant benefits. In subsequent years, this component of its budget will be adjusted in line with changes in spending on the equivalent benefits in the rest of the UK, adjusted for "population at risk". (Some of these benefits are specific to particular groups within the population. Rather than adjust for overall changes in the population, as is the case for the Barnett formula, it has been agreed to reflect relative changes in the size of the populations that may be eligible for the benefits - the "population at risk")
This mechanism for transferring welfare benefits means that the existing AME spending in Scotland, which is largely used to fund public sector pensions, has not been affected by the transfer of fiscal powers to Scotland. The new welfare benefits will be dealt with through Scottish DEL.
This change will give the Scottish Government the incentive to use the additional resource to deliver better outcomes for welfare recipients, or the same outcomes at lower cost. Having these resources within the DEL budget does enable it to transfer resources to other priorities. For example, the delivery of free personal care might be made more effective by the effective integration of AA and DLA budgets with existing social care budgets at local government level.
It has taken some time, and considerable resource, to put in place the necessary infrastructure to deal with these changes in Scotland's fiscal powers. The changes involved were much more extensive than those needed under the Scotland Act 2012, which involved the creation of Revenue Scotland and payments of around £40 million to HMRC to implement the Scottish rate of income tax (SRIT). We do not yet have the final cost associated with the implementation of these new powers.
Macroeconomic Context for the 2018-19 Budget
Since 2014-15, the performance of the Scottish economy has largely mirrored that of the UK as a whole. As a result, the growth in income tax revenues in Scotland has closely tracked that in the UK. The consequence is that the reduction in the Barnett block grant that accompanied the granting of limited income tax powers under the Scotland Act 2012 has almost exactly offset the revenues arising from SRIT (based on an SRIT of 10p), Land and Buildings Transactions Tax and Landfill Tax.
Thus far, the tax powers associated with Scotland Act 2012 have had no net effect on the Scottish Government's spending power.
Nevertheless, due to the continuing pressure to reduce public sector borrowing, the block grant itself from Westminster has been under considerable strain. However, Scotland has been somewhat sheltered from this pressure by the UK Government’s decision to protect real spending on health and education in England. The Scottish Government allocates more of its overall budget to health and education than does the UK Government. In the jargon of the Barnett Formula, spending on health and education is almost entirely “comparable” and thus a larger share of the Scottish Budget has been protected from spending cuts than that of the UK as a whole.
Our proposed budget for 2018-19 is set out in Table 1 below:
[i] See: Powering Public Services, Local Government Association
In this section we outline the detailed assumptions underpinning the draft budget.
Council Tax grows in line with OBR (Dec 2014) projections for Council Tax in Scotland. This assumes that the Council Tax freeze remains until 2015/16, after which the rates are increased annually in such a way as to raise revenues by 2% per annum. The tax base is assumed to grow by 0.7% per annum, from present.
LBTT and Landfill Tax receipts grow at the rate projected by the OBR (Dec 2014) for those taxes in Scotland.
All other taxes are assumed to grow at the same rate in Scotland as the OBR forecasts for the UK as a whole in its Dec 2014 projections. Income tax revenues are projected to grow 28% from 2013/14 to 2018/19 (having been revised down from 35% growth in March).
Block grant assumptions
Scotland’s block grant in 2013/14 is assumed equal to Scotland’s DEL budget minus revenues from council tax and business rates. Changes to this block grant in future years are calculated as follows:
OBR projections (Dec 2014) provide forecasts of DEL spending for current (RDEL) and capital (CDEL) spending at UK level over the period to 2019/20.
We assume that RDEL will be protected in real terms over the period to 2018-19 for health, schools and international development. We take spending on these items in 2013/14 and uprate nominal spending to maintain the real value of spend. The spending profiles of all other departments is assumed to be cut in such a way as to balance the OBR’s forecast RDEL budget.
In each year, a Barnett consequential for Scotland is calculated by applying the relevant departmental comparability factor to the change in nominal spend. Scotland is allocated a population share of the consequential changes summed over all departments; population proportions are taken from ONS population projections.
Similarly, OBR projections for CDEL spending are combined with relevant departmental comparability factors to estimate the Barnett consequential associated with capital spending.
Taxes devolved in 2015/16 result in a one-off deduction to the block grant. The value of the estimated SRIT is assumed to be deducted from the block grant in 2016/17. We assume that the ‘Indexed Deduction’ method is used to determine the ‘Block Grant Adjustment’ in 2017/18 and 2018/19.
Revenues from other taxes (the remainder proportion of income tax, the share of VAT, APD and the Aggregates Levy) are assumed to be deducted from the block grant in 2018/19.
Devolved benefit spending
Expenditure associated with benefits devolved under the Smith proposals are estimated by taking outturn expenditure on these benefits in Scotland in 2013/14, and assuming that future spending grows on each of these benefits grows at the same rate as the DWP forecasts for the UK as a whole.
RDEL/ CDEL split
We assume that the Scottish Government increases its DEL budget in proportion to the CDEL element of its Barnett consequential. Thus our estimate of CDEL can be thought of as a lower bound estimate of the likely CDEL budget (the Scottish Government can vire RDEL into CDEL as it wishes, but has to seek the approval of HM Treasury to vire CDEL to RDEL). We also assume that the Scottish Government uses its borrowing powers under the Scotland Act 2012 to fund additional capital spend in 2015/16 – 2017/18, which effects interest payments in 2018/19.
Borrowing and other costs
The estimated cost associated with PFI/ NDR is taken from a SPICe research paper (Campbell and Aiton, April 2014). We also assume that the Scottish Government has borrowed £250m for capital investment in each of 2015/16, 2016/17 and 2017/18 in line with its Scotland Act powers, and that it faces a 5% rate of interest on this debt (covering principal and interest), in line with the estimate in the Scottish Government Draft Budget 2015/16, (p.157).
It is assumed the Scottish Government faces a one-off payment of £45m to the HMRC in respect of administration arrangements for devolved taxes21[i].
Forecast AME expenditure on elements associated with Scotland’s existing devolution settlement (largely pensions) are taken from the Draft Budget 2015/16, and assumed to grow at the same rate as UK AME spending excluding the debt interest component of AME. This may underestimate the growth rate of the Scottish Government’s AME budget, given that it relates largely to pensions rather than welfare benefits.
In Appendix 1, we explore the implications of varying some of these assumptions described above, while in Appendix 2 we illustrate how forecast errors may lead to a significant borrowing requirement for the Scottish Parliament.
Appendix 1: Alternative Scenarios
In this section we explore the effects of changes to some of the key assumptions made above for the size of the Scottish budget. We present three scenarios that have varying degrees of plausibility and indicate the effects they would have on our central budget estimates in Table 2.
- Scenario 1: No protection of health and education at UK level
We maintain the OBR assumptions for the path of RDEL, but assume that after 2015/16 these cuts are distributed evenly across UK spending departments. This results in a reduction in Scotland’s block grant of some £2.5bn. The decision to ring-fence real spending in health and education at a UK level is highly favourable to the Scottish Government, generating increased revenue equivalent to the cost of the new welfare benefits that have been assigned to Scotland.
- Scenario 2: A slower path to closing the UK deficit
The current UK Government aims for a budgetary tightening of 5% of GDP between 2013/14 and 2018/19. It expects TME to fall from 41% of GDP to 36% of GDP, with this adjustment virtually all coming from DEL.
In this scenario we assume a slower path to deficit reduction. Specifically, we assume that there is a fiscal tightening of 2% of GDP; spending on RDEL is assumed to fall from 18 to 16% of GDP, rather than under the status quo where RDEL is expected to fall from 18% to 13%. Not surprisingly, this results in an increase in Scotland’s block grant of £1.5bn. The increase is perhaps not as great as might be expected. This is because of the core assumption that health and schools spending is protected; the additional RDEL spending is assumed to go to departments that are not currently protected, and these are associated with a smaller Barnett consequential than are health and schools.
- Scenario 3: A more rapid growth of income tax revenues in Scotland
This scenario assumes that income tax revenues in Scotland grow at the rate forecast for the UK in the OBR’s March 2014 forecast (as opposed to the December forecast). This results in Scottish income tax revenues being around £0.6bn higher by 2018/19, as noted previously. Note however that the Scottish Budget only benefits if this rate of growth is greater than the growth of UK income tax revenues overall. If higher Scottish revenue growth is matched by higher revenue growth in the UK, then the value of Scotland’s Block Grant Adjustment simply grows more rapidly, and a larger deduction is made from Scotland’s block grant when the rest of Scottish income tax is devolved in 2018/19.
In this scenario therefore, we assume that the SRIT grows between 2016/17 and 2018/19 at the rate forecast for UK income tax growth by the OBR in March 2014, whilst UK revenues grow at the rate forecast by the OBR in December 2014. This results in SRIT revenues growing from £4.8 - £5.35bn, while the value of the BGA grows from £4.8 - £5.3bn. Thus the Scottish Budget grows by some £50m in this scenario as opposed to the core scenario.
Of course, the remaining Scottish income tax revenues (those outside the SRIT) have also grown at a more rapid rate than those in rUK, but the higher SRIT in 2014/15 was deducted from the block grant in that year, and the remaining tax revenues were deducted from the block grant in 2018/19. The only ‘benefit’ to the Scottish Budget of the higher income tax growth is the difference between the growth of the SRIT between 2016/17 and 2018/19 relative to the growth of the equivalent tax in rUK.
- Scenario 4: Continuation of Council Tax freeze
Another scenario (not shown in the table) is to assume the Scottish Government maintains the Council Tax freeze into the next parliament. This would result in a reduction of Council Tax revenues of around £250m by 2018/19.
Appendix 2: Forecast errors
Our main budget contains estimates of tax revenues in Scotland for 2018-19. These are, of course, subject to forecast error. Income tax revenues can only be known with certainty well after the end of the financial year. Yet, forecasts for income tax are of particular interest to the Scottish Government since this is the largest source of revenue for the Scottish Parliament.
Income tax forecasts have been particularly prone to error in recent years due to the unusual behaviour of the Scottish and UK local labour markets following the recession of 2008-09. In particular, the substantial increase in employment that has occurred since 2008-09 has mostly been in low-paid jobs (both employees and self-employed) which generate relatively small amounts income tax revenue given the generosity of the personal allowance.
This change was recognised by the Office of Budget Responsibility in December 2014. This led to a substantial downward revision in its average earnings forecast and consequent reduction in expected income tax revenues.
To illustrate the magnitude of the forecast errors we have conducted an experiment with our microsimulation model. The experiment shows the difference between forecast income tax revenues for Scotland in 2018 based on the March 2014 OBR forecast compared with the December 2014 forecast. Table 3 below shows the OBR forecasts of average earnings made on these dates.
Aggregated over the period 2014-15 to 2018-19, these seemingly small reductions in the average earnings forecast between March and December 2014 would have led to a reduction of more than £37 billion in income tax receipts at a UK level. For Scotland in 2018-19, our model suggests that the reduction in revenues would have been around £600 million: this is clearly a considerable sum. It is worth around one third of net council tax revenues in Scotland. It is also well in excess of the £300 million borrowing facility agreed under the Scotland Act 2012 to cover forecast errors associated with SRIT.
[i] Based on the fact that The Second Annual Report on the Implementation of the Financial Provisions in the
Scotland Act 2012 (Scottish Government, April 2014) identifies that HMRC costs for the implementation of the
SRIT are likely in the order £35-£40m.