Implementation of the Scotland Act 2016: second report

Annex B: Joint UK Government / Scottish Government Paper on Direct Spillover Effects

Background

1. JEC(O) agreed in July that UKG and SG officials should set up a working group to consider how to account for policy spillover effects, in line with paragraphs 44-53 and 98–104 of the fiscal framework. This is where a policy decision from one government has a financial effect on the other – either through a direct, mechanical change (direct effects); or a change in associated behaviour (behavioural effects).

2. JEC(O) agreed that the working group should focus on direct effects first, with the aim of agreeing a process to account for these by October. Behavioural effects will be considered separately; there may be some elements of the process for direct effects that influence discussions on behavioural effects.

3. The working group – comprising DWP, HMRC, HMT officials from UKG; and social security, tax, finance officials from SG – has since worked to reach a shared understanding of:

• A definition of direct effects.
• A process to account for direct effects.

4. This paper sets out where the working group has reached agreement on these, for JEC(O) approval; and it notes the areas that will need further consideration.

5. The working group's aim is to agree a process for spillover effects in relation to tax and welfare. However, there may be issues in other policy areas that the UK and Scottish Governments need to account for under the 'no detriment' principle. Any agreement on a process for tax and welfare should be made without prejudice to how issues may need to be accounted for in other policy areas.

Definition of a direct effect

6. The fiscal framework states that all direct effects should be accounted for. The working group has agreed that:

• A direct effect is defined as an automatic change in one government's expenditure or revenue as a result of a change in another government's policy.
• Direct effects exclude any effects that do not occur as a result of the mechanics of a policy decision and any associated behavioural changes.

7. The group has considered various examples in relation to the above, including policy decisions that result in an automatic change to eligibility criteria for devolved and reserved benefits; and policy decisions in relation to personal tax that are not fully captured by the block grant adjustment mechanism.

8. The working group has agreed that any automatic change is considered a direct effect and will need to be accounted for. For welfare, an automatic change to the number of eligible claimants as a result of a policy decision would be considered a direct effect.

9. In some cases, there may also be behavioural changes that occur alongside automatic changes (for instance, more people might be incentivised to claim a benefit following a policy change). In such cases, UKG and SG analysts should seek to separate the direct effect from any behavioural assumptions, as far as possible, when calculating the effect. Any consideration of behavioural effects will be subject to the criteria set out in the fiscal framework and spillovers process for behavioural effects, which remains to be agreed.

Process to account for direct effects

10. The fiscal framework agreement on direct effects means that both governments should agree a process whereby direct effects can be raised and calculated, with the aim of agreeing a financial transfer if needed.

11. The working group has agreed there should be a single, centralised process every year to examine and agree direct effects, with any transfers ultimately signed off via JEC. The group has agreed that UKG and SG should seek to implement this process in year 1 ( i.e. for policy decisions coming into effect from April 2018), noting that it will need to be reviewed once it has been tested against a year of actual policy decisions and effects.

12. This process is set out in the annexed model, and covers the following steps:

November-December – SG and UKG announce policy decisions

• UKG are likely to announce most tax or welfare policy decisions at the Autumn Budget, and the SG will do the same at their draft Budget (although changes may not necessarily come into effect from the start of the next financial year). Where changes are announced at other times in the year, UKG and SG will follow the same process, with mutually agreed revised timelines if needed.

January-February – SG and UKG officials start considering effects

• SG and UKG officials should start internally reviewing potential spillover effects from policy changes that will come into effect that financial year.

March-April – SG and UKG agree which effects should be modelled

• There should be an initial official-level forum for SG and UKG to jointly review policy changes coming into effect from April, and agree the scope of any spillovers work, so that there is a shared understanding of which effects will be calculated, and by whom. The governance of this forum will need to be considered further and agreed between SG, DWP, HMRC, and HMT officials.
• Either government should be able to raise an effect for consideration (whether a cost or saving). However, the government that wishes to raise the effect should be predominantly responsible for doing the work to calculate it; although a methodology should be agreed between analysts first. Where appropriate, the other government will assist in calculating the effect.
• Once agreed at official level, the effects that each government wishes to raise should be put to ministers for clearance. These should then be agreed via correspondence between UK and Scottish Government finance ministers.

May-August – spillover effects to be calculated

• Once ministers have agreed which effects will be raised, there will be a need for engagement between SG and UKG analysts as part of any modelling process to calculate effects.
• As far as possible, there should also be a shared understanding between UKG and SG analysts of how effects will be calculated (which may vary dependent on the effect). The analysts in the government leading the analysis to calculate the effect should engage with analysts in the other government as early as possible on the proposed approach to calculation, e.g. the models and the data used.
• This should include consideration of whether the effect will need to be recalculated every year, or whether there is a way to index the effect in future years (assuming no further UKG/ SG policy changes). This will be reviewed once the process has been tested.
• UKG and SG analysts will need to share their calculations, data and modelling with each other, in order to reach an agreed costing of the effect.
• Once effects have been calculated and agreed by analysts, SG and UKG officials should agree the spillovers they wish to progress, and modelling treatment of these effects in future years ( i.e. whether they will need to be recalculated each year).

September-October – spillover effects and transfers agreed

• Effects and corresponding transfers will then be agreed via JEC(O), to be put to SG and UKG ministers to sign off via JEC.
• All agreed transfers should then be made via a block grant adjustment at Autumn Budget. This means that agreement will be needed on all transfers by end October.
• There may be some effects which DWP, HMRC, and the SG should prevent from taking place at source, in line with paragraph 89 of the fiscal framework. These are separate from the direct effects that will be dealt with through the spillovers process above and block grant adjustments.

Defining a policy decision

13. Using the process above, either government should be able to raise an effect for consideration. However, the working group also noted that in some instances, it might be difficult to establish which government has made a policy decision, or which decision has caused the effect. The group has agreed that we would not try to define a policy decision now, but would review this once we have tested the process in year 1.

Transfers and review

14. The initial transfers agreed via JEC will be based on forecasts from the SG and UKG modelling process. These effects should be reviewed once outturn data is available, as part of the annual process. Both governments will need to jointly consider outturn data, to agree how far it is possible to agree an improved estimate of the effect; and if so, how this should be calculated. As far as possible, UKG and SG should seek to reconcile any effects and corresponding transfers to outturn. Revised effects should be agreed via the annual process, to agree a further block grant adjustment at Autumn Budget.

15. However, the working group noted that it may not always be possible to calculate an effect exactly – either through the forecast, or the outturn data. In these instances, SG and UKG analysts should therefore work together to reach a best estimate of the effect based on the available data.

16. If UKG and SG are unable to agree a sensible estimate of an effect based on the forecast for a given year, there will be no transfer. However, SG and UKG officials will be able to revisit the effect once outturn data is available, with a view to deciding whether this should be raised again via the process based on outturn.

Process to account for direct effects