Introduction

In his Spring Statement, the chancellor said that he saw light at the end of the fiscal tunnel. We are afraid that this will turn out to have been an illusion. While public borrowing has been reduced to sustainable levels, our analysis of the prospects for the public finances points to severe challenges ahead. Sustainable borrowing has been achieved by cutting back on public services and restraining public sector pay. But 'austerity fatigue' is now setting in, and emerging recruitment difficulties and concerns about the quality of public services mean that there are already strong pressures for public spending to rise. On top of that, the demand for public services is increasing sharply to meet the needs of an ageing population. That provides a difficult background for the 2019 Spending Review. With debt already at an uncomfortable level, increasing public spending by substantially more than currently planned would need to be financed with higher taxation or more direct charging for services. The alternative option of continued spending restraint would not be palatable unless government can find significant ways to improve efficiency and get better value for money.

Seven lean years

In the aftermath of the financial crisis and associated severe recession, public sector net borrowing reached a postwar record high of 153 billion [pounds sterling] in 2009-10 alone (9.9 per cent of national income in that year). At the time, there was widespread agreement across the political spectrum that this borrowing needed to be brought down.

The main approach to fiscal consolidation since 2010 has been to restrict public spending by means of pay restraint and by cutting public services in non-prioritised areas. Adjusted for general inflation, public spending per head of population was 6 per cent lower in 2017 than it had been seven years earlier. As a share of spending in the whole economy, public spending fell from 45.1 per cent of GDP in 2009-10 to an estimated 38.9 per cent of GDP in 2016-17, and is now lower than the postwar average of 39.3 per cent of GDP (figure 1).

The fiscal consolidation was not spread evenly across the public sector. From the outset, certain types of spending were protected and real spending per head on health, overseas aid and pensions (via the so-called 'triple lock') continued to increase.

This meant that non-priority spending areas experienced severe cutbacks. Real spending on defence was 14 per cent lower in 2016-17 than it had been in 2010-11, spending on public order and safety was 17 per cent lower, and spending on housing and community activities was 37 per cent lower. While spending on schools was protected, spending on education as a whole was 13 per cent lower, though this partly reflected a switch in university funding to student fees.

But even some of the priority spending areas have not been able to maintain the quality of service expected. For example, at a time when the elderly population is increasing, it is now generally accepted that not enough resources have been allocated to health, even though its budget has not been cut. And this has been compounded by real cuts in funding for adult social care that have reportedly led to delayed discharges of elderly hospital patients and high levels of bed occupancy.

How much spending is warranted?

This raises the important question of how to judge what level of spending on public services is warranted?

One approach is to compare government spending in the United Kingdom with that of other countries. Government spending as a share of GDP in the United Kingdom was broadly in the middle of the pack of OECD countries in 2016. (1) But one size does not necessarily fit all. And using that yardstick does not take into account the different preferences and characteristics of countries, as diverse as Finland, at the top end of the range, and the United States, towards the bottom, that lead them to choose different levels of spending.

An alternative approach is to suppose that the level of public spending as a share of GDP has evolved over time in accordance with the preferences of the British people, and is determined by a range of factors. In our empirical work we consider the following factors: existing spending commitments, the proportion of people who are over 85(as a proxy for the demand for public services associated with an ageing population), the economic growth rate (to account for counter-cyclical fiscal policy), and the debt to GDP ratio (as a proxy for what is thought to be affordable. We use statistical methods to estimate the importance of these different factors over the postwar period. That allows us to construct an estimate of the amount of public spending that would have been chosen by previous governments if confronted by the circumstances of today. (2) We call this 'warranted' spending as it reflects the historic choices made by democratically elected governments in the UK. As Pain, Weale and Young (1997) point out, this exercise is similar to estimating a Taylor rule for monetary policy, and does not imply that the behaviour that is summarised in this way is in any sense optimal.

Figure 2 shows our estimates of warranted spending as a share of GDP over the seven-year period from 2010-11 to 2016-17 in comparison to actual spending.

On this basis, our analysis suggests that the pace of spending reduction between 2010-11 and 2014-15 was actually a little slower than our estimate of what was warranted (line in figure 2). However, our analysis shows that after 2014-15, actual spending became more stringent than our estimate of what was warranted by previous governments' spending priorities.

We can apply the same approach to individual areas of public spending. This shows that even in areas such as health care and social protection, where spending was in line with our estimates in the early years of the fiscal squeeze, it now appears that consolidation has gone on too long. For instance, our analysis implies that in 2016-17 the health care sector lacked around 440 [pounds sterling] per head in funding. And any room to reduce spending on social protection appeared to be fully exhausted by 2017.

Public sector output, costs and efficiency

What people care about is what public sector services deliver in terms of improved health, crimes prevented and examinations passed, rather than just the amount of money spent on providing the services. And it is possible for the output of public services to increase even when spending is limited. This can come about in two ways. First, when public sector costs are kept low so that more can be purchased with less. Second, when the efficiency of public service delivery is improved, so that more can be achieved with less.

On the first point, public sector costs have been kept down by public sector pay restraint. During the first two years of the fiscal consolidation, a freeze was imposed on public sector pay. This freeze was then followed by a 1 per cent annual growth cap. An extended period of pay restraint has meant that a gap has built up between public sector pay and what is available in the private sector. Recent NIESR research has estimated that public sector wages had fallen more than 3 per cent below their equilibrium level in 2016-17 (Dolton, Hantzsche and Kara, 2018). A consequence of the growing divergence between pay in the public and private sectors is that the public sector has become a less attractive place to work. This divergence has resulted in increasing recruitment difficulties in a number of areas. For example, in February 2017 it was announced that prison officers would receive immediate pay rises to reverse a'dangerous staffing crisis'. In January 2018, the Public Accounts Committee criticised the government for giving 'insufficient priority to teacher retention and development'. (3) This issue was also examined by Sam Sims and Rebecca Allen in the February issue of this Review (Sims and Allen, 2018). And shortages are apparent in the NHS. By 2017, one in nine nursing posts in England remained unfilled. (4) Partly as a consequence, a new pay deal for NHS staff was announced in March 2018 to be funded by additional NHS spending.

This pattern suggests that continued pay restraint in the public sector is unlikely to be sustainable. Indeed, the government has now lifted the across-the-board 1 per cent pay cap. But that does not mean that the government will provide extra money to fund additional pay. While the Treasury has committed to do this in the case of the NHS pay deal, it has reportedly warned other government departments not to regard the NHS deal as a signal that it will provide extra resources to them to fund pay increases to their staff.

On the second point, there is little evidence of an improvement in measured productivity in the public sector. But measuring public sector efficiency is complex. (5) For some public service activities, the Office for National Statistics (ONS) measures output by the activities and services delivered. But in other areas, such as defence and police, there is no obvious way to measure the output. It is assumed for such services that the volume of output is equal to the volume of inputs used to create them.

The ONS attempt to measure quality improvements in healthcare, education, and public order and safety. The purpose is to reflect the extent to which the services succeed in delivering their intended outcomes and the extent to which services are responsive to users' needs.

Figure 3 shows ONS estimates of the quality of healthcare, education and public order and safety services up to 2015. In education and in public order and safety there has been a marked reduction in measured quality since the onset of the fiscal consolidation. In education, this decline in measured quality may be misleading as it partly reflects changes in examination passes that may not be due to changes in the quality of education provided. In public order and safety the measured decline was due largely to increases in the number of reported self-harm and assault incidents in prisons. Improvements in quality appear to have stalled in health care.

Figure 4 shows the latest official estimates of public service inputs, outputs and productivity, after quality adjustment, covering the period up to 2015. This shows a sharp slowdown in the growth rate of public sector output over the years 2011-15. It is worth noting that output growth would probably have been even weaker had it not been for public sector pay restraint. These estimates show little evidence of an increase in public sector efficiency in recent years, though they pre-date the Efficiency Review that was announced in the 2016 budget.

There are a range of other indicators of the quality of public services, many of which are not yet taken into account in the official ONS statistics. The Institute for Government (IfG) maintains an excellent Performance Tracker that highlights where pressures are building in the provision of public services. (6) For example, in the health sector, there has been a marked deterioration since 2010 in the percentage of A&E visits being dealt with in four hours. And there are pressures building in areas where the ONS makes no quality adjustment. For example, continued cuts in police numbers may be having an effect on the quality of service, with victim satisfaction with the police having fallen since 2014.

Taken together the evidence suggests that, following a sustained period of fiscal restraint, low levels of public spending have had a detrimental effect on the provision of public services. To some extent public sector pay restraint has helped to keep costs down and so allowed the public sector to keep staff numbers and service output higher than would otherwise have been possible. But it is very doubtful that pay restraint can be maintained given the better rewards available in the private sector. This suggests that the emerging pressures on public sector service quality are likely to be stronger in the future.

Seven more lean years

The government's approach to fiscal policy and public spending is set out in the Charter for Budget Responsibility. In his post-2017 election autumn budget, the chancellor confirmed that the government was committed to following fiscal rules that would 'guide the UK towards a balanced budget by the middle of the next decade'.

This is expected to mean that spending restraint will extend further into the future. Departmental spending levels for the period up to 2019-20 were fixed in the 2015 Spending Review, (7) and will be agreed for the following five years in the 2019 Spending Review. The future path of government spending will be set to achieve the government's fiscal objective. The latest Spring Statement forecast by the Office for Budget Responsibility (OBR), representing its interpretation of the government's declared policy, shows spending continuing to fall from 38.8 per cent of GDP in 2017-18 to 37.6 per cent in 2022-23. But even with this level of fiscal restraint, the OBR's central forecast of public sector net borrowing goes no lower than 0.9 per cent of GDP. (8) That means that the government would have to tighten its purse strings even further if it wants to balance the budget by the middle of the next decade.

In this section we set out two possible 'what-if' scenarios to enable us to estimate the scale of the future fiscal challenges that the government faces. In doing so we are very aware that the future is uncertain and that circumstances may change dramatically for good or bad. For example, there may be a surge in economic growth and tax revenue that makes it easy to finance more public services. But it would not be prudent to assume such an outcome. Instead, the purpose of different scenarios is to enable us to think through the risks of different policies in the interests of making better choices today.

Continued austerity

The first scenario we consider is the business-as-usual case given by NIESR's February 2018 central forecast. This is our benchmark and assumes that the government sticks to its current spending plans in cash terms. Because the NIESR forecast for GDP is a little more optimistic than the OBR's, public spending as a share of GDP falls further to 36.6 per cent in 2022-23, a percentage point lower than the OBR assumption for that year. This is shown by the black line in figure 5.

Easing austerity

The second scenario is where we look at the consequences for the public finances of an easing of fiscal austerity, driven by the need to deal with existing pressures on public services and, on top of that, the consequences of further ageing of the population.

Our analysis has established that by 2017-18, spending in a number of areas had fallen below warranted levels. These areas included education, public order and safety, and more recently health and social care. The government will find it hard to resist pressure to restore the quality of public services in these areas. We have also shown that some of the fiscal squeeze was achieved by capping public sector pay and this has resulted in building pressures on recruitment. In order to alleviate increasing recruitment difficulties in sectors like the NHS, pay will need to increase by more than in comparable private sectors to catch up with lost ground.

In addition, the pressure from further ageing of the population is gradually building. And this will mean that extra resources will be needed to cover rising health and care costs and serve a larger number of pensioners.

We have created a fiscal easing scenario where to ease these pressures spending grows in line with our estimate of its warranted level (dotted red line in figure 5). In that case, spending would rise to 40.2 per cent of GDP in 2022-23, thereby opening up a substantial gap over what is currently planned.

Consequences for the public finances

Figure 6 shows public sector borrowing in the two scenarios we have described. This is calculated on the assumption that there are no changes to taxes, other than those which have already been announced.

In the first scenario, where the government continues with austerity and sticks to its current spending plans, it would achieve its objective of a balanced budget by the middle of the next decade. But, given existing pressures on public spending, we doubt that such a path is tolerable unless it is backed by substantial efficiency improvements in the delivery of public services.

In the second scenario, where spending restraint is relaxed to accommodate more spending on public services, higher public sector pay and increased demands from an ageing population, the deficit would rise to over 4 per cent of GDP by 2022-23, breaching the government's fiscal rules. That clearly would not be acceptable to the government and would mean that public sector debt would remain at uncomfortable levels.

The unpalatable consequences for government borrowing could be avoided by raising taxes. But the increase would have to be substantial to raise 4 per cent of GDP, currently worth around 80 billion [pounds sterling], bearing in mind that an increase of lp in the standard rate of income tax would currently raise only around 4 [pounds sterling] billion.

How to deal with the challenges ahead

There are no easy answers to the fiscal challenges the government currently faces. There certainly does not appear to be much light at the end of the tunnel.

Of course it is possible that economic circumstances surprisingly improve, lifting tax revenue and so the amount of public spending that could be afforded within the existing fiscal rules. But disappointing economic performance over the past ten years suggests that might be just a pipe dream, and that it would be better to consider other options. It does not appear credible to continue on the existing path of deteriorating public services when needs are increasing.

The options available to ease fiscal austerity fall into three main categories:

Amending the fiscal rules

There is widespread acceptance across the political spectrum that debt levels are uncomfortably high and so there is a reluctance to increase borrowing significantly. Nevertheless the government's fiscal objective of balancing the budget in the medium term is seen by many experts as overly stringent. There are other rules that would allow more borrowing and still maintain fiscal sustainability in the medium term.

For example, the government could adopt the so-called 'golden rule' that it borrows no more than it invests. On current trends that would allow public spending to be sustained at 39 per cent of GDP, rather than falling to 37.6 per cent of GDP by 2022-23 as in the government's current plans.

But that would provide relatively limited scope for additional spending and still leave public sector debt at an uncomfortably high level. High debt levels would be less of a concern if real interest rates were certain to remain at currently low levels. One possible approach would be to carry out debt management operations to lock in more borrowing at current low real interest rates.

Raising tax revenue in new ways

There appears to be little appetite for a significant increase in general taxation to fund higher spending. But there are other ways to increase revenue such as charging users for more parts of services provided, or changing the nature of services provided by shifting some functions currently carried out by government further down the delivery chain. Any moves in that direction would need to be clearly communicated in advance to give people time to prepare for their introduction.

In addition, there have been renewed calls for a hypothecated tax to fund the NHS. There are good reasons to be wary of such taxes. (9) Nevertheless, if they are the only politically feasible way of raising extra revenue then they should be considered as an alternative to continued austerity. That said, the amount of money that could be raised in this way without such taxes coming to be seen as identical to general taxation is probably limited.

Improving public sector efficiency

The third option is to improve the efficiency with which public services are delivered. As we have seen, measured public sector productivity gains have been relatively modest in the past, despite attempts over many years to get more from less. But that does not mean that efficiency gains are not available, nor that they would be associated with worse services. Productivity improvements in the private sector are generally achieved by a process of 'creative destruction' where competition from new businesses with new ideas and using new technologies forces incumbent businesses to keep up or close down. That process may not be appropriate throughout the public sector, but its absence may explain why ideas to improve public sector efficiency are not adopted more readily.

That would suggest a range of options be investigated that might help improve public sector efficiency by transforming public services. These would include providing incentives to make changes, innovate and improve efficiency; encouraging private sector discipline and incentives to appropriate parts of public services; changes to organisation and workforces, including adopting best management practices; redesigning services and providing alternative delivery mechanisms that empower users to choose most efficient providers; using technology and data to target or deliver services; more widespread sharing of best practices.

These three options for ending fiscal austerity are not mutually exclusive. But in the context of ongoing Brexit negotiations and the associated elevated level of uncertainty, the government is unlikely to want to consider introducing new taxes or amending the fiscal rules. That suggests that seeking radical ways to improve public sector efficiency may be the only palatable option for ending fiscal austerity in current circumstances and should be urgently considered.

NOTES

(1) OECD (2018), general government spending indicator.

(2) The estimated relationship is:

[mathematical expression not reproducible]

where TME/Y and Debt/Y are total managed expenditure and UK national debt, both as a share of GDP, Age 85+/Pop is the proportion of the population who are over 85, and growth is the annual GDP growth rate.

N=65, [R.sup.2]=0.86, Newey-West standard errors in brackets.

(3) 'Retaining and developing the teaching workforce', Public Accounts Committee.

(4) Estimates by Royal College of Nursing, May 2017.

(5) Methods of improving estimates of public sector productivity are being considered as part of the research programme of the Economic Statistics Centre of Excellence at NIESR.

(6) https://www.instituteforgovernment.org.uk/publications/performance-tracker-autumn-2017.

(7) An exception was made for those departments whose overall budgets are protected. Their Resource DEL was set to 2020-21. Capital budgets for every department were set to 2020-21.

(8) Economic and Fiscal Outlook, March 2018.

(9) See Peter Dolton's blogpost https://www.niesr.ac.uk/blog/realremedies-or-hypothetical-hypothecation-fund-nhs.

REFERENCES

Dolton, P., Hantzsche, A. and Kara, A. (2018), 'Follow the leader? The interaction between public and private sector wage growth in the UK', presented at Royal Economic Society annual conference, March 2018.

Pain, N., Weale, M. and Young, G. (1997), 'Britain's fiscal problems', Economic Journal, 107 (403), pp. 1142-56.

Sims, S. and Allen, R. (2018), 'Identifying schools with high usage and high loss of newly qualified tecahers', National Institute Economic Review, 243, February, R27-36.

Arno Hantzsche and Garry Young, NIESR. E-mail: a.hantzsche@niesr.ac.uk, g.young@niesr.ac.uk. We are grateful to Jagjit Chadha and Philip Craig for comments on an earlier version, and to Sopra Steria for funding the underlying research.

Caption: Figure 1. Total managed expenditure as per cent of GDP

Caption: Figure 2. Warranted public spending as a share of GDP

Caption: Figure 3. Quality of public services

Caption: Figure 4. Growth in public service outputs, inputs and productivity

Caption: Figure 5. The outlook for public spending under different assumptions about austerity

Caption: Figure 6. The outlook for public borrowing under different assumptions about austerity