Commentary: devolution within the UK.

Armstrong, Angus ; Ebell, Monique

The outcome of the Scottish referendum shows that the clear will of the Scottish people is to stay in the UK. For the first time it can be said that the union of the two largest nations is entirely consensual. (1)

However, the referendum process also revealed considerable dissatisfaction. The leaders of the UK's main political parties recognised this two days before the referendum and made a public vow that "extensive new powers for the [Scottish] Parliament will be delivered by the process and to the timetable agreed and announced by our three parties." (2) Not surprisingly, the three main parties, the Scottish National Party and the Scottish Green Party all have different ideas of what these 'extensive new powers' should be. Lord Smith has agreed to oversee the process for reaching agreement between the parties on which new powers should be devolved by November, allowing for draft legislation by January.

We believe that the Scottish referendum revealed two clear preferences. First, there are certain capabilities which the UK provides that are invaluable to all constituent nations: a successful currency union and a seat at the top table of the world's leading international forums, such as the European Union. There may also be other centrally provided capabilities which are highly valued, such as security and diplomacy. Second, there is a clear wish for genuine power to be transferred away from central government to local decision making. This is not simply a matter of more discretion over public spending but the responsibility for real economic choices. Only when politicians have responsibility to make decisions can they be held accountable for the outcomes.

Whether the vow was wise or not is a moot point. The only way for Scotland's politicians to be accountable is for them to be able to take decisions on which they can succeed or fail. This requires changing the Scottish Parliament's near total dependence on a block grant decided by Westminster. In addition to more tax powers (other than fully devolving income tax), a necessary requirement is to allow the Scottish government to borrow in its own name from capital markets. This requires removing the extremely cautious borrowing limits in the Scotland Act (2012). Only by introducing market discipline can the Scottish government be held to account by its electorate.

UK government finances

The structure of government finances differs in every country depending on history and who holds economic power. Some states are described as unitary systems, where most fiscal resources are collected and disbursed by a central government. Other states are federal systems where more public services are delivered and paid for through sub-central tiers of government, such as the US, Germany and Switzerland. The UK is usually described as a unitary system. However, because of the amount of devolved spending in recent years it increasingly resembles a hybrid structure.

The UK is an uneven union of nations. The English make up 84 per cent, Scots make up 8 per cent and the Welsh and Northern Irish are the other 8 per cent of the population. Because of the sheer size of England relative to the other nations, a true federal system of government is only possible with equally devolved regional powers in England. While devolving power to regional governments in England has many desirable features, this is unlikely to be in place in the short term. Yet in recognition of the different preferences of the other nations in the Union, a high degree of regional spending power has been devolved. On the spending side of the ledger, this has similarities to a federal system.

While spending has been devolved, most revenues are still collected centrally by Westminster. Spending by regional assemblies and the Scottish Parliament is funded by central government transfers through a block grant and formula system administered by the Treasury. A grant is transferred to the Scottish government based on the UK Spending Review for those items identified as specifically for Scotland and which are not reserved for central government. It is then for the Scottish authorities to spend the funds according to local demands. The size of the grant changes according to the Barnett formula which links changes in the UK Departmental Expenditure Limits in the Spending Review to relative population size.

The important point to recognise is that through the dominance of the block grant the Scottish government operates a 'balanced budget' by construction. There is much to commend this system. It avoids wasteful tax competition, ensures a single macro economic framework and has maximal risk-sharing and common social protection across the Union. However, it also means that the Scottish government is not responsible for raising or lowering taxes or borrowing to finance its spending decisions. And without these responsibilities it cannot be accountable to its electorate. It is mostly tasked with spending an allocated amount of funding. In Armstrong and Ebell (2014a) we describe this as 'pocket money' devolution.

The Scotland Act (2012) gives the Scottish parliament significant new tax powers. Yet the powers are structured so that the Scottish government still effectively operates in a balanced budget context. If the revenues raised by the Scottish government fall short of target, then it has three options: first, borrow up to 200 million [pounds sterling] in any year (500 million [pounds sterling] in total) from the Treasury for repayment within four years; second, accumulate cash reserves from the new taxes which can be drawn down in future years; and third, cut in-year expenditure. The Act also allows the Scottish government to borrow 240 million [pounds sterling] in any one year and 2.2 billion [pounds sterling] in total for capital spending. Total borrowing allowed by the Scottish government is one hundredth of 1 per cent of the UK projected gross debt in 2015.

These strict borrowing limits will become highly problematic if significant tax raising powers are devolved. Scotland could easily find itself forced to cut spending and/ or raise taxes in a recession to make up for a drop in income tax or other revenues and balance its budget. It is difficult to see how this paltry amount of borrowing could suffice to smooth out income tax revenues over the cycle, so the Scottish government might find itself forced into running a potentially damaging procyclical fiscal policy.

Reserved central government spending

Certain types of public goods and services are better supplied by central governments. In the UK this is called 'reserved' spending. This includes spending on defence, foreign affairs and membership fees and interest payments on the national debt. These items affect all citizens of the UK, and leaving spending to sub-central governments may result in a sub-optimal level of supply. Reserved spending also includes elements of social protection. Central governments can mitigate regional economic disturbances by pooling resources and sharing risks across a currency union. Shocks which do not affect regions equally can then be offset by making fiscal transfers from one region to another. (3) For this reason transfers related to regional economic conditions, or the regional business cycle, such as social protection and welfare payments, are better managed by central government.

Central government is also the 'risk manager of last resort'. When crises strike, for example an environmental disaster or financial crisis, central government is expected to have the resources to respond. A practical example of the resources the UK central government must command was its ability to insure 1 trillion [pounds sterling] of household deposits in 2007. Acting as insurer of last resort will often require borrowing. Armstrong and Ebell (2013) explain how central borrowing is, ceteris paribus, cheaper than subcentral borrowing. The central government always has a further advantage because it does not have a hard budget constraint. In extreme events the central government controls the supply of the currency.

These items of central government spending require funding. While Scotland is less than one-tenth of the UK tax base, it is important that all constituent nations contribute equally. What does this imply for the revenue raising powers of central government? (4) Average central government revenues of OECD countries are 33.5 per cent of GDP (21.2 per cent for the six high income federal nations). (5) The UK has central government revenues of 38 per cent of GDP. A reasonable case can be made for the UK requiring more central resources than other countries (or at least access to more resources) due to its large financial industry and high debt position. It is worth noting that the financial sector is relatively larger in Scotland than in the rest of the UK.

Devolved spending and revenue powers

To see whether more spending should be devolved requires first looking at the current powers of the Scottish Government and Local Authorities (SGLAs). Table 1 below summarises all public spending in Scotland in 2012-13. The first column shows itemised total spending (current and capital) of 65.2bn [pounds sterling], or 51.6 per cent of onshore GDP. The second column shows the amount of spending identified to be for the exclusive benefit of Scottish residents and businesses. This is spending which has a relatively small spillover into the rest of the UK. The final column shows the amount of the identified public spending controlled by SGLAs.

Almost 60 per cent of all public sector spending in Scotland is controlled by the SGLAs. Of the total spending identified as specifically for Scotland, 69 per cent is controlled by the SGLAs. This is equivalent to 45 per cent of onshore GDP. The areas of direct spending not controlled in Scotland are reserved for the 'benefit' of the whole of the UK such as defence, interest payments and international payments. The only area of identified spending which is not substantively controlled by the SGLAs is for social protection, which includes social security payments such as unemployment and incapacity benefits. This is entirely consistent with the principles of efficient risk sharing and fairness. It is difficult to see what items of spending could be further devolved.

While most UK taxes are collected by Westminster, the Scottish government's estimates of how much revenue was raised in 2012-13 are presented in table 2. The first column shows the estimated revenue and the second column is the estimated amount of revenue raising powers devolved to Scotland. At present only Council Tax and non-domestic rates are devolved, which generate a paltry 8.4 per cent of onshore revenue. Significantly more tax raising powers have been granted by the Scotland Act (2012). Stamp duty and landfill tax are fully devolved and a new Scottish Rate of Income Tax (SRIT) will be introduced in April 2015. All UK income tax rates in Scotland will be reduced by 10p and replaced by a SRIT to be set by the Scottish government. (6) The Scottish government keeps the revenue from the SRIT but the block grant from the rest of the UK is reduced accordingly.

The second column in table 2 shows the Scottish government's estimate of the amount of money which would have been raised in 2012-13 if the devolved taxes had already been in place. (7) This is equivalent to 15 per cent of identified spending and 22 per cent of spending controlled by the SGLAs.

Vertical fiscal imbalance

The difference between devolved revenue and spending is sometimes called the 'vertical fiscal imbalance' (VFI). This is one measure of the degree of responsibility and accountability at the sub-central level of government. The weaker the link between changes in locally controlled taxes and changes in locally controlled public spending, the less accountable are local politicians. Put another way, the greater the VFI the greater the block grant to fund the public spending controlled by the Scottish Parliament. The VFI in Scotland is the difference between the 69 per cent of Scottish-controlled spending and 15 per cent of Scottish-raised revenue as a share of identified spending. In plain words, less than one quarter of the public spending by the SGLAs is funded by revenue controlled by the Scottish government.

It should be noted that all sub-national governments have a VFI. It simply means that they contribute to central government beyond funding national public goods. The issue is whether Scotland's imbalance is at odds with other countries. Scotland's VFI of 54 per cent is considerably higher than an average estimate of 40 per cent by economists at the IMF. (8) Reducing this gap requires allowing the Scottish Parliament to control more of its own revenues. Note that assigning a share of fiscal revenues collected on a UK basis by an assignment formula, for example for VAT revenues, creates an incentive to increase the tax base but there is still no direct link with spending powers.

Will borrowing improve accountability?

Reducing the VFI by devolving more taxation power must go alongside greater borrowing rights. Clearly there will be occasions when spending and revenue are out of kilter. The Scottish government ought to be allowed to borrow in its own name and without bound with the explicit legal statement that the UK government bears no responsibility for the debt. Extending greater tax powers to Scotland but without greater capacity to borrow implies continued balanced budget dependency on Westminster and therefore no responsibility or accountability.

Only with borrowing added to new tax powers will there be a truly accountable government in Scotland. In Alexander Hamilton's famous dictum "the creation of debt should always be accompanied with the means of extinguishment."

If the Scottish government were allowed to borrow, it could take responsibility for its decisions in the sense that they may succeed or fail. This includes being able to finance large infrastructure projects in the hope that the additional tax revenues generated in the future would be sufficient to repay the debt incurred to finance them. This would also reveal the cost of borrowing for Scotland and introduce market discipline for the Scottish government where it is absent today.

The challenge is to make the ability to borrow compatible with the currency union. There are two possible incentives for Scotland to over-borrow. The first is that the real cost of borrowing could be reduced and imposed on the rest of the UK by creating an 'inflation bias' at the Bank of England. But since Scotland is less than one tenth of the UK economy, it is hard to see how the Bank would accommodate any Scottish fiscal largesse. The second incentive is more problematic. There is likely to be a 'bail-out bias' simply because the rest of the UK is large enough to always bail out Scotland. However, this could be mitigated by the UK government making it explicit (perhaps in legislation) that it has no liability for Scottish debt. No fiscal limits or stability pact should be agreed because they imply culpability.

There is no evidence to suggest that the Scottish government would not learn to accept the market discipline imposed by borrowing. Indeed, discovering Scotland's borrowing costs might be an excellent disciplining device for its government. Even if we are wrong and underestimate the risk of Scottish profligacy, the amounts involved do not seem likely to threaten the UK government. After the largest financial crisis in three generations the average debt burden of the six OECD federal states is 23 per cent. A similar debt burden for Scotland is 2 per cent of UK GDR This risk may simply be the price to pay to ensure that devolution is structured to encourage accountability and responsibility and reinforces, rather than undermines, the Union.

Which taxes to devolve?

It is not our aim to describe here the pros and cons of devolving each tax. Yet an emerging consensus in favour of fully devolving income taxes requires some comment in the context of borrowing. This would violate at least three economic principles and possibly undermine the integrity of the union. First, high income earners are particularly mobile and there is a risk of creating inefficient tax competition. (9) Second, income tax has a large yield and is highly dependent on the local economy. Without the capacity to borrow this may create macroeconomic stability problems. An adverse shock leading to a sudden fall in taxes would require fiscal tightening perhaps leading to a deeper downturn. If higher earners migrate then this could worsen the outcome further.

Finally, there is the issue of fairness to Scotland. The morning after the referendum the Prime Minister said that it is now time to face the West Lothian question. (10) If income tax is fully devolved this greatly strengthens the case for 'English votes for English laws'. It would be difficult to justify Scottish MPs voting on income taxes which are in effect exclusively in another jurisdiction. Yet because England is 84 per cent of the Union, its decisions on income tax would have a strong influence on macroeconomic policy for the whole of the UK. Since an English government may not be the same as a UK government, Scotland would have no say on what would effectively be UK-wide macroeconomic policy. Imposing a symmetric solution on an asymmetric union would be undemocratic. (11)

Finally, taxes from North Sea oil and gas are large and clearly of totemic importance. (12) The ability to borrow would also make devolving tax revenues from North Sea oil and gas to Scotland feasible. Tax revenues from North Sea oil and gas are notoriously difficult to predict accurately and highly volatile: year-to-year drops of 4-5bn [pounds sterling] (equivalent to more than 10 per cent of total devolved spending) have occurred twice in the past decade. With the ability to borrow, Scotland could smooth out these fluctuations.

REFERENCES

Armstrong, A. and Ebell, M., (2013), 'Scotland's currency options', National Institute of Economic and Social Research Discussion Paper 415.

--(2014), 'Monetary unions and fiscal constraints', National Institute Economic Review, 228, pp. F4-11.

--(2014a), 'Real devolution the power to borrow', NIESR Discussion Paper 437.

--(2014b), 'Real devolution: a reply', NIESR blog.

Armstrong, A. and McCarthy, D. (2014), 'Scotland's lender of last resort options', National Institute of Economic and Social Research Discussion Paper 426.

Cameron, D., Miliband, E. and Clegg, N. (2014), The Vow, Daily Record, p. 1.

Carney, M. (2014), The economics of currency unions', speech given at the Scottish Council for Development and Industry, 29 January.

Eyrand, L. and Lusingan, L (2011), 'Decentralising spending more than revenue: does it hurt performance?', IMF Working Paper WP/11/226.

Kenen.R (1969),'The theory of optimum currency areas:an eclectic view', in Mundell, R.A. and Swoboda, A.K. (eds), Monetary Problems of the International Economy, Chicago, University of Chicago Press, pp. 41-60.

Scottish Government (2013), Government Expenditure and Revenue Scotland 2011-12, The Scottish Government.

NOTES

(1) There is no similar campaign for other constituent nations to leave the UK.

(2) Daily Record (2014), pi.

(3) This is the important insight from Kenen's (1969) work on the optimal currency area.

(4) Governor Carney noted that effective unions typically have centralised spending of over 25 per cent of GDP.The difference between the two totals is generally the amount of funding allocated to local government.

(5) An average of OECD countries which report central government funding. Data are from 2011 and 2012 except Australia which is from 2007.The six federal states are Australia, Canada, Germany, Spain, Switzerland and US.

(6) The SRIT provides the Scottish government with some discretion to set the overall rate of income tax and to keep part of the revenues from widening the tax base. However, the percentage point difference between the tax rates (e.g. currently the 25p difference between the top and bottom rates) will continue to be set by the UK government. Higher tax rates for the highest earners would also imply higher tax rates for the lowest earners. Of course, the lower earners could then be compensated through new social spending programmes.

(7) This assumes the Scottish government chose to set the SRIT at 10p.

(8) Eyrand andf Lusingan (2011).

(9) One could argue that the risk of human capital flight would be a disciplining device for the Scottish government and thus discourage high taxes.

(10) This question is whether MPs from outside of England should be able to vote on exclusively English matters.

(11) There are many other ways this would violate the fairness principle. Suppose that there is another banking crisis. The UK government is the ultimate backstop but Scotland would enjoy tax revenues from its financial sector just as much as England. If the UK government raised income taxes but the Scottish government chose to cut taxes this would rightly be seen as unfair. Perhaps the Scottish government may simply cut high income taxes to attract the financial sector, knowing that the backstop would not be Scottish taxpayers.

(12) One possibility is that these are awarded back to Scotland on a smoothed basis to make fiscal planning easier.

Angus Armstrong and Monique Ebell, National Institute of Economic and Social Research and Centre for Macroeconomics. E-mail: a.armstrong@niesr.ac.uk.

Table 1. Estimated public spending

                                Total    Identified    SG & LAs

Social protection              22,458        21,969       5,539
Health                         11,284        11,178      11,169
Defence                         3,027             4           4
Education and training          7,651         7,650       7,625
Public order and safety         2,529         2,402       2,274
General public services         6,381         4,998       4,828
Other                           7,856         7,602       7,105
Public sector debt interest     4,020             0           0

                               65,206        55,803      38,544

Source: Scottish Government (2013) and NIESR calculations.

Table 2. Estimated fiscal revenue

                                                    Devolved
                                           Total      2015

Income tax                                10,865       4,231
VAT                                        9,347           0
National insurance contributions           8,521           0
Gross operating surplus                    3,247           0
Corporation tax (excl North Sea)           2,872           0
Fuel duties                                2,258           0
Alcohol and tobacco duties                 2,108           0
Council tax                                2,006       2,006
Non-domestic rates                         1,981       1,981
Vehicle excise duty                          481           0
Stamp duties                                 472         283
Capital gains tax                            292           0
Other taxes on income and wealth             271           0
Inheritance tax                              243           0
Air passenger duty                           234           0
Landfill tax                                 100         100
Other taxes, royalties and adjustments     2,267           0
Total (excl. North Sea revenue)           47,566       8,601
North sea revenue (geographical share)     5,581

Source: Scottish Government (2013) and NIESR calculations.