The state of the public finances.

Pain, Nigel ; Young, Garry ; Westaway, Peter 等

Introduction

The public sector borrowing requirement (PSBR) was |pounds~36.7 billion (6 per cent of GDP) in the last fiscal year and was forecast by the government in the March Budget to rise to |pounds~50 billion (8 per cent of GDP) in the current fiscal year. Yet as recently as 1988/9, the public sector was able to make a debt repayment of |pounds~14.7 billion (3 per cent of GDP). Such a large change has prompted concerns about the present state of the public finances and indeed caused the government to set in place future increases in taxation designed to ameliorate the situation.

The purpose of this note is to assess the sustainability of the current fiscal position. In doing this we pay particular attention to the development of the stock of public sector debt in relation to national income and the public sector's stock of capital. Whereas the PSBR measures the amount of borrowing the public sector needs to do within a particular period, the debt stock measures the total amount of obligations to pay interest that the public sector has outstanding. It is this concept that is most useful in discussing the sustainability of fiscal policy. The ultimate constraint on the budgetary decisions of the public sector is that it is able to pay the interest on its debt. Interest payments will tend to rise in line with the debt stock and, for given interest rates, will tend to rise as a proportion of GDP when the debt stock is rising as a proportion of GDP. Such a situation is not sustainable: ultimately the public sector would be forced to change its budgetary policy or repudiate its debt.

It is also important to examine the relationship between the evolution of the public sector's debt stock and the amount of capital it owns. Public sector capital assets typically provide services over a number of years. Some investments, such as council houses, generate future revenue for the public sector. Other items of capital expenditure, such as expenditure on roads, do not (as yet) provide a direct source of future revenue. In either case it can be argued that it is appropriate for the public sector to finance such capital expenditure by borrowing and that the revenue required to finance the interest payments on the resulting debt will become due in line with the services provided by the capital. For capital that generates revenue, the revenue itself can be used to cover the interest. For capital that does not generate revenue, the tax levied to finance interest payments can be seen as a payment for services provided.

We begin by considering the proximate causes of the sharp change in the state of the public finances since 1988/89 and attempt to quantify the extent to which those changes are related to the general economic cycle. Cyclical changes in the fiscal position are less worrying than other more structural changes since there is a tendency for them to be reversed as the economy recovers, ensuring that they do not have a long term adverse effect on the debt stock.

Our estimates suggest that around 2 1/2 percentage points of the 9 percentage point swing in the PSBR/GDP ratio since 1988/9 can be directly attributed to the economic cycle. Borrowing is presently being used to finance both current and capital spending, with non-oil tax revenues some 2 percentage points lower, as a proportion of GDP, than at the trough of the previous two recessions in the UK. In this light the recent decisions to restrain public sector pay and announce future tax increases appear to have some justification.

We also examine the long-run trends in the PSBR over the last two decades and attempt to isolate the factors that have led to a generally lower rate of borrowing. If it is reasonable to expect these factors to continue then the recent increase in the PSBR might be viewed as a temporary aberration about a gradual downward trend. Finally we consider the future prospects for the public sector finances using the Government's own projections for its spending and tax plans together with our forecast for the development of the economy more generally. This makes it possible to calculate the likely path for the ratios of both the debt stock and the capital stock to GDP in the medium term on the basis of the announced budgetary policy of the government.

Our analysis suggests that the current fiscal policy is sustainable, in that the forecast suggests that the recent actions announced by the government will tend to stabilise the PSBR at some 2 1/2 per cent of GDP by the late-1990s, with the ratio of general government gross debt to GDP tending to 50 per cent. We also explore the impact on the fiscal position of a failure by the government to implement its announced policies. This shows that the actions that the authorities presently intend to undertake are necessary to prevent the outlook becoming significantly worse.

A comparison of the changes in the components of the PSBR in the prospective recovery with those in earlier periods of economic recovery shows large differences between our forecasts for tax revenues and net capital spending and previous historical experience. Whilst the need to reduce the PSBR to GDP ratio is now as pressing as it was in the mid-1970s, the likely future decline in proceeds from asset sales and in income from the remaining assets of the public sector reinforces the need to correct the present deficit through higher tax revenues.

It is also the case that to the extent that the share of current government expenditure in national income was a problem in the mid-1970s and the early-1980s, it remains so today. Arguments that suggest that the present high level of borrowing arises solely from shortfalls in tax revenue ignore the fact that the secular decline in the share of total government expenditure in the 1980s arose entirely from the reduction in capital accumulation within the public sector rather than from any sustained reduction in current expenditure. If a case is to be made for correcting the present deficit by means of tax changes alone, it requires a more detailed analysis of the separate effects of changes in expenditure and taxation on the wider economy.

Although the present fiscal policy may be a sustainable one, it is not necessarily an optimum one, though an optimum one must be sustainable. There are many different sustainable policies; what is required is a means of choosing between them. In the short term, the present stance of fiscal policy implies that the ratio of government debt to capital is likely to rise sharply, to levels last seen twenty-five years ago, with the government continuing to borrow to finance current expenditure. Whilst such a development might have been appropriate prior to the exploitation of assets such as North Sea oil, it is less so now, given what is known about future trends in the age structure of the population. A very full discussion of these issues can be found in Odling-Smee and Riley (1985).

Preliminaries

As a preliminary to the more detailed description of the development of the public finances to be considered below, it is useful to set down briefly the accounting framework that we use to guide the discussion and also the relationship between debt and deficits.

The public sector borrowing requirement is by definition equal to public sector expenditure less receipts. In discussing the state of the public finances it is useful to classify government activity in five separate categories: current expenditure, capital expenditure, general tax revenue, net income from public sector assets and miscellaneous financial transactions. Within this classification, we have chosen to treat certain activities in a slightly different way to that adopted in the public sector accounts. Current expenditure is equal to total expenditure less capital expenditure and debt interest payments. Net capital expenditure is given by gross fixed capital formation less the proceeds from privatisations, the sales of council houses, land and other buildings.(1) The net income from public sector assets is given by the sum of revenues from the North Sea (tax and royalties), the gross trading surpluses of public corporations and rent receipts, less net payments of property income (largely debt interest). All of these latter categories of income are dependent upon the size and scope of the public sector.

The cash basis on which government accounts are normally presented fails to distinguish adequately between current and capital expenditures. Indeed the failure to consolidate gross capital expenditure with capital divestments through privatisation gives a misleading impression of the extent to which net capital formation by the public sector is changing over time.(2)

An alternative classification of the PSBR would be into expenditure on debt interest and the 'primary deficit': the deficit excluding debt interest. This indicates how the change in the stock of public sector debt (which is is equal to the PSBR absent of revaluations and 'high powered' money creation)is related to the previous level of the debt stock through the effects of debt interest. A debt spiral can arise if a high debt stock leads to high interest payments and an even higher debt stock in the future. The likelihood of this scenario is greater when the rate of interest is greater than the rate of growth of the economy. Under these circumstances the debt income ratio will rise unsustainably as debt interest is rolled up into the debt stock unless there are primary surpluses of sufficient magnitude. In the converse situation where the rate of interest is less than the rate of growth of the economy the debt to income ratio will tend to fall as income growth reduces the relative size of the debt stock unless there are primary deficits of sufficient magnitude.

In general the stock of debt will settle down as a proportion of GDP when the PSBR to GDP ratio settles down. Further simple debt arithmetic indicates that in the steady state the debt to GDP ratio is given by (1+g)/g.PSBR/GDP, where g is the nominal growth rate of the economy. Thus, for example, a constant ratio of the PSBR to GDP of 5 per cent will lead to a debt to GDP ratio of 55 per cent with nominal growth of 10 per cent per annum and to a debt to GDP ratio of 105 per cent when growth is only 5 per cent per annum: the slower the nominal growth rate the longer it takes to reduce the size of outstanding debt in relation to GDP.

The current budgetary position

In order to assess the current fiscal position it is necessary to understand how it has been reached. Chart 1 shows the behaviour of the PSBR to GDP ratio from the beginning of the 1970s to date. The basic pattern displayed here is that of a cycle about a gradual downward trend. The cycle in the public finances reflects the economic cycle, with the peaks in the PSBR being associated with the troughs of economic activity. This is usually thought to be because of the operation of the 'automatic stabilisers'--higher transfers to the unemployed and lower taxes on incomes--over the cycle. This suggests that some part of the current problem is cyclical and therefore likely to correct itself as the economy recovers from recession.

Table 1 reports the changes in the main components of the PSBR/GDP ratio over the five years to the troughs of the three most recent economic cycles. Some 5 percentage points of the 9 percentage point swing in the PSBR/GDP ratio between 1988/9 and 1992/3 can be accounted for by a rise in current public spending relative to GDP. A fall in tax revenue relative to GDP accounts for some 2 1/2 percentage points, with an increase in net capital spending relative to GDP (partly due to a decline in asset sales) and a fall in income from public sector assets relative to GDP both accounting for increases of a little under one percentage point.

Examination of the components of these changes indicates that only a relatively small proportion of the recent deterioration in the PSBR can be attributed to

the economic cycle, particularly once allowance is made for the decline in asset sales and income from assets. On the spending side, two and a half percentage points of the change is due to an increase in current expenditure on goods and services, which partly reflects an increase in the relative pay of many public sector employees. On the tax side, income tax payments as a share of GDP actually rose over the period, reflecting the extent to which 'fiscal drag' raised tax revenue in the late-1980s, in spite of repeated reductions in tax rates. National Insurance contributions fell, but this is partly attributable to the effects of lower rates announced in the 1989 budget as well as to higher unemployment.

Non-oil company tax payments fell by more than other tax revenues over this period but not all of this can be attributed to the effects of the economic cycle. Instead this partly reflects the working through of the changes to the corporation tax system announced in 1984. These changes affected the timing at which capital allowances on new investment could be claimed so that capital is now written off for tax purposes over a longer period of time than was previously the case. This had the effect of temporarily reducing the aggregate quantity of capital allowances claimed in any one year.(3) At the same time the rate of corporation tax was reduced from 52 per cent to 35 per cent and has since been reduced further to 33 per cent. This is a genuine remission and the combined effect of this and the changes to capital allowances is ultimately to reduce the yield from corporation tax.

The main cyclical elements in the change in the PSBR over the period from 1988/9 are some of the change in current grants (which includes unemployment related benefits) and some of the fall in taxation. It is unlikely that this amounts to much more than a quarter of the deterioration in the PSBR, although such an estimate is very uncertain.(4) Even in the case of current grants, the trend in the real level of expenditure was changed by the decision in 1991 to resume the uprating of child benefit payments. Such a factor should be seen as structural rather than cyclical.

It is interesting to compare the changes since 1988 with those that occurred in earlier periods. One point that emerges from such an exercise is that there is comparatively little that is common in the different periods considered, reinforcing the worries about the accuracy of any attempt to adjust the PSBR for the effects of the cycle. Table 1 shows a similar increase in the PSBR/GDP ratio in the five years to 1975/6 but a fall in the PSBR/GDP ratio in the five years to 1982/3, even though the recession in the early-1980s was as severe as the most recent one. The main difference in the components of the public accounts appears to be on the tax side, with total tax revenues rising as a proportion of GDP in both previous episodes. This suggests that on both occasions the authorities had taken early action to prevent the deficit becoming too large.

One factor that is likely to have stopped this from occurring this time is the election in April 1992. Some TABULAR DATA OMITTED evidence for this is borne out by the simple regression equation shown in Box A. This relates the PSBR, adjusted to remove the effects of oil revenues and asset sales, to a measure of the business cycle, the indebtedness of the public sector, inflation and a variable that takes account of the electoral cycle. The equation suggests that the adjusted PSBR tends to be about one and a half percentage points higher than otherwise in election years and the years after elections. It also provides some quantification of the typical effect of the economic cycle on the PSBR. This indicates that about two and a half percentage points of the change in the PSBR/GDP ratio from 1988/9 to 1992/3 is accounted for by the cycle, an estimate broadly consistent with that obtained from the earlier consideration of the changes in the individual components of the PSBR.

Overall, the analysis of recent events confirms that temporary (albeit protracted) events such as recessions and elections can have a significant influence on the public finances. However the quantitative importance of these events in explaining the deterioration in the public finances from 1988/9 to 1992/3 is relatively small (since 1987/8 was also an election year). There appears to be a much larger deterioration that must be attributed to other causes. It is therefore of some importance to investigate what has caused the reversal in the apparent trend decline in the PSBR to GDP ratio shown in Chart 1. Table 2 lists the broad components of the PSBR for each fiscal year from 1970/1 to date and also shows how we foresee these developing in the forecast period. The developments over the forecast are discussed in more detail below.

The evolution of the PSBR and its causes

The most important components of the PSBR to GDP ratio shown in Table 2 are the balance between current public sector spending and tax revenue, net capital spending and net income from assets. These separate series are plotted in Chart 2. This suggests that much of the explanation for the trend decline in the PSBR is attributable to a similar decline in net capital spending. Gross capital spending by the public sector has fallen steadily as a share of GDP since the early-1970s, from a peak level of 9 1/4 per cent in 1975/6 to 3 1/3 per cent in 1992/3, with public corporations gross investment falling by 3/4 per cent of GDP to 3 3/4 per cent, and general government investment declining to 2.6 per cent of GDP from 4 3/4 per cent in 1975/6 (although it has been relatively constant since 1982/3). The effects of this downward trend have been accentuated by the sales of public TABULAR DATA OMITTED sector assets, beginning with council house sales in the mid-1970s and continuing with the privatisation programme that began in earnest in 1984. At their peak in 1988/9, the resulting revenue was equivalent to 2 3/4 per cent of GDP. By 1992/3 this had declined to 1 3/4 per cent.

Chart 3 compares the PSBR to GDP ratio with the net capital spending to GDP ratio. This reveals both the strong correlation between the two series and that the PSBR has been less than the borrowing required to finance net capital spending in every year considered until 1991/2. This indicates that the public sector has generally respected the idea that borrowing not be used to finance current spending, the so-called 'golden rule'.

In fact, current spending exceeded tax revenue (on our definitions) from 1975 to 1985. To a large extent any deficit of current spending over tax revenue was offset in the past by the contribution of net income from assets, ensuring that total borrowing remained below the borrowing required to finance net investment. At present the degree to which current expenditure exceeds tax revenues is larger as a proportion of GDP than at any time over the period considered. The main components of public sector expenditure and tax revenue are shown in Tables 3 and 4 respectively.

TABULAR DATA OMITTED

Net income from assets has been declining since its peak in 1982 and is put at close to zero in 1992/3. Around half of this is due to the decline in oil revenues. The remainder of the decline either reflects the sale of many public assets for less than their true market value or the use of the proceeds from the reduction in net capital expenditure to finance net current expenditure rather than to repay debt. In principle, the proceeds raised from the sale of such assets should be sufficient so as to reduce net debt by such a degree as to offset exactly the resulting loss of revenue with lower debt interest payments.(5)

The effects on the public sector balance sheet

The above analysis of the components of the PSBR suggests that the decline in the PSBR to GDP ratio over the period under consideration has been associated with a similar decline in net capital spending by the public sector relative to GDP. Indeed, it could be argued that the fall in the borrowing requirement was warranted by the decline in the public sector capital stock relative to GDP. It is of some importance to assess what effects these changes have had on the balance sheet of the public sector.

TABULAR DATA OMITTED

This presents some well known difficulties in first agreeing what ought to be included in the balance sheet and then determining how it should be measured. These two issues are very closely connected since problems of measurement often dictate what is included in the balance sheet. For example the depletion of oil reserves has provided revenue for the public sector and there ought perhaps to be a corresponding change in the value of oil reserves in the balance sheet of the public sector. Similarly, those who have contributed to the National Insurance Scheme can expect to receive benefits when they are sick, unemployed and when they retire. It is in principle possible to calculate the present value of the future obligations of the public sector, with the resulting liability added to the balance sheet.

The public sector balance sheet produced by the Central Statistical Office includes neither of these items. Instead it includes financial assets and liabilities and tangible assets. Even the measurement of tangible assets is not straightforward. In particular the CSO has not produced estimates of the tangible assets in the balance sheet since 1987 (although it is expected to produce up to date estimates in this year's Blue Book) and therefore does not include a period when the capital stock of the public sector is likely to have been significantly depleted as a consequence of the decline in net capital spending. In order to provide up to date estimates of the net worth of the public sector we use estimates of its net capital stock. This is measured on a replacement cost basis rather than at market values and is typically higher than market valuations. It may therefore overestimate the true value of the public sector's tangible assets. It has the merit of providing a consistent estimate of the trends in net capital assets, the most important factor for current purposes where interest is focussed on how the balance sheet is changing over time.(6)

Table 5 shows the key items in the balance sheet as they have developed over the period from 1957. The stock of tangible assets owned by the public sector has fallen as a proportion of GDP since the mid-1970s. This is fully consistent with the decline in net capital spending by the public sector described above. The table also reports alternative measures of indebtedness. The ratio of general government debt to GDP is the debt stock criterion TABULAR DATA OMITTED proposed in the Maastricht treaty as a means to assess whether a country is ready to join a monetary union. Net public sector financial liabilities (also known as net financial debt) is the widest definition of public sector debt.

The table shows a trend decline in debt in relation to both GDP and public sector tangible assets. Both of these are measures of the sustainability of fiscal policy. The first--the solvency ratio--is the most widely used measure, but the second--the gearing ratio--is also important because it shows how the public sector's debt is changing in relation to the capital stock which it may be seen as financing. The fact that the solvency and gearing ratios have both fallen over the period since 1957 suggests that more of the financing of the capital stock has come from current rather than future taxes. This is probably appropriate given that the inherited debt stock in 1957 had largely arisen as a consequence of borrowing during wartime.

The ratio of the stock of net financial debt to the stock of public owned capital is now around 45 per cent and has remained close to this ratio since the early-1970s. The fact that this is less than than 100 per cent suggests that future taxpayers are expected to contribute less to the cost of existing capital than it is worth. However, in the absence of a fully adumbrated balance sheet for the public sector, it would be unwise to attach too much significance to the measured level of the gearing ratio, although changes in the ratio over time will still provide an indication of the sustainability of fiscal policy.

One implication of the debt to capital ratio being less than 100 per cent is that its level can remain unchanged when capital assets are sold even if the resulting repayment of debt is less than the reduction in the value of the capital stock. Thus a privatisation can still reduce the gearing ratio even if the sale of capital is not at the full market value. It should be noted that this differs from the simultaneous effects on the flows of net income from assets; if these are to be unaffected, debt needs to be reduced so that the resulting fall in interest payments offsets the loss of capital income.

The discussion of the longer-term trends in the components of the PSBR and the arithmetic of debt accumulation points to some rather unpalatable conclusions. First, the reduction in the ratio of the PSBR to GDP observed since the 1970s has come about largely as a reduction in net capital spending. This is partly due to the effects of sales of assets but also reflects a decline in gross capital spending by the public sector. There is a limit to how far this process can continue: ultimately the stock of marketable public sector assets will run out. Indeed the recent decline in investment might have to be reversed in the future if replacement investment has been neglected. Second, and a direct consequence of the first point, the net income from assets (including oil) has declined steadily since the mid-1980s. At present this makes a zero contribution to the PSBR but, in the absence of any adjustments to the stance of fiscal policy prevailing in 1992/3, might be expected to make a negative contribution in the near future if interest payments on an enlarged debt stock outstrip income from other assets. Third, unlike in the 1970s, the rate of growth of nominal income is likely to be quite subdued in the 1990s so that the debt stock will not be diminished in relation to GDP by growth in the latter.

Changes in prospect

These arguments suggest that it will now be more difficult than in the past to reduce the PSBR to GDP ratio as temporary influences on the public finances disappear. If the PSBR is to be reduced then it will have to come about as a result of greater control over current spending and tax revenue.

Our latest forecast for the development of the UK economy in the short and medium term is described elsewhere in the Review. The forecast shows a modest recovery over the next two years, with growth around 3 per cent in 1994 and 2 1/2 per cent in 1995. Inflation rises somewhat from its present level, with consumer prices rising by 4-5 per cent in each of the next two years. Unemployment is projected to average 2.8 million next year and 2 3/4 million in 1995. These developments of themselves affect the fiscal position of the public sector. In addition, the forecast builds in the various tax and spending plans announced by the government in the March Budget this year and assumes that additional net tax rises worth some |pounds~2 1/2 billion will be announced in the November Budget.

The outlook for the various components of the PSBR contingent on the implementation of these plans and other aspects of the forecast is reported in Tables 2, 3 and 4. We expect the PSBR to GDP ratio to reach a peak of 7 per cent in the current fiscal year before declining to a level of about 2 1/2 per cent by the end of the decade. In broad terms this turnaround comes about as a consequence of a fall in the ratio of current spending to GDP of 2 1/2 percentage points, a rise in tax revenue of a similar amount and an offsetting rise in net capital spending relative to GDP of close to 2 percentage points. This latter change largely reflects the projected decline in proceeds from

privatisation. The PSBR is projected to exceed net capital spending and thereby break the 'golden rule' until 1996/7.

The change in current expenditure is largely accounted for by changes, in expenditure on goods and services and current grants. These settle down at levels below those of the current year, but above those prevailing in the late-1980s. On the tax side, the rise in revenues is due to similar increases, relative to GDP, in income taxes, non-oil company taxes and indirect taxes. None of the resulting tax shares appears implausible when judged against previous historical experience.

The overall impact of these developments on the balance sheet of the public sector is set out in Table 5. The net debt to GDP ratio is expected to rise throughout the period, due to a combination of a run of relatively high deficits and relatively slow growth in nominal income. The debt arithmetic suggests that a constant PSBR to GDP ratio of 2 1/2 per cent will, with a 6 per cent growth rate in nominal income, lead to a ratio of net financial debt to GDP of 44 per cent. This is similar to the level reached by the debt to GDP ratio at the end of the decade in the forecast. The general government gross debt stock rises to 48.6 per cent of GDP by this time, well within the guidelines set out in the Maastricht Treaty. Thus, in answer to the question that this note is attempting to address, it does appear that the future path of fiscal policy as set out in the present plans of the government is sustainable.

However, it is important to consider the development of the public sector capital stock over this period. Table 6 indicates that this is expected to increase in line with the anticipated increase in net capital spending by the public sector. However it does not increase by as much as the predicted rise in the ratio of net financial debt to GDP, suggesting that the public sector gearing ratio will rise. This is in line with the projected contravention of the 'golden rule'. This is illustrated in Chart 4 which shows the gearing ratio from the early-1970s to the end of the forecast period.

This sharp rise in the ratio of public sector debt to capital over the forecast period suggests one of two things; either the public sector is prepared to allow more of its capital to be financed by debt than in the recent past or fiscal policy can be expected to be adjusted further to maintain the past relationship between public sector debt and capital. Both possibilities present difficulties for the government. The policy changes that are already built into the forecast suggest that fiscal policy is set to be quite tough throughout the forecast period and a further tightening may not be welcomed. But, at the same time there is less justification than in earlier periods for allowing the gearing ratio to rise. The unrecorded parts of the public sector's balance sheet that are nevertheless known about, such as the value of oil reserves and the pension liabilities to the currently working population, are probably changing in such a way as to justify a reduction in the gearing ratio. In this sense, there may be a case for further fiscal tightening.

An alternative option under discussion at present is the possibility of greater private sector participation in the financing of public sector capital. Examples include the leasing of capital assets from the private sector and joint ventures where the public and private sector share the cost and the risks from a project. However it is by no means clear that these types of schemes amount to more than creative accounting. Initially the PSBR will be lower because of the smaller contribution made by the public sector to the costs of particular projects. However in later years, the PSBR will be higher than otherwise because of increased payments to lessors or because the return from the project is shared with the private sector. In stock terms, the amount of public sector debt and public sector capital will both be lower as a consequence of these types of schemes. This does not present a solution to the problem of how to to reduce debt relative to capital.(7)

Table 6 compares the changes in the components of the PSBR/GDP ratio in the recovery from 1993/4 with those in the recoveries from the troughs of the recessions in TABULAR DATA OMITTED 1975/6 and 1982/3. In each case we consider the change over the ensuing four year period. The anticipated change in the PSBR to GDP ratio in the recovery from 1993/4 to 1996/7 is of a similar size to that achieved over the four years from 1975/6 to 1978/9 and greater than that achieved in the period from 1982/3. In part this reflects the unusual decline in the PSBR/GDP ratio during the recession of the early-1980s. The change in public sector current spending relative to GDP of 1 1/2 percentage points is fully in line with that achieved before. However there is a large difference in the anticipated change in tax revenue from the changes observed in earlier periods, with tax revenues, particularly from income tax, expected to rise as a share of GDP this time.

Other items affecting the PSBR also appear less propitious than previously. In particular net capital spending is expected to rise as a proportion of GDP this time as a consequence of lower asset sales. Net income from assets is however expected to make a positive contribution to public sector revenue as a consequence of increased oil revenue following the changes to the structure of oil taxation announced in the Budget.

It is of some interest to consider the extent to which the actions announced in the 1992 Autumn Statement and the March Budget, along with the further action we assume forthcoming in the November Budget, have improved the prospective fiscal position of the public sector. The decisions to restrain public sector pay and announce future tax increases both appear justified given the need to restrain the current expenditure of the public sector and raise tax revenues. There are a number of other factors which can be expected to influence the fiscal position in the medium term. Chief among these is our (implicit) assumption that benefit payments and personal tax allowances will continue to be uprated in line with prices rather than earnings, although this may not be a realistic long-term policy since it implies a continual erosion in the real living standards of benefit recipients.

To obtain an estimate of the importance of these assumptions, the present forecast was re-run with each being relaxed in turn. The effect of each change on the fiscal position of the public sector will not be the same as the ex-ante size of the change since alterations in tax rates can be expected to induce substantial second-round effects on the economy, as the comparative model properties paper elsewhere in this Review illustrates. The results are shown in Charts 5, 6 and 7, which show the PSBR/GDP, debt/GDP and debt capital ratios in three cases, the main forecast, the forecast without the announced tax increases and the forecast without the tax rises and with uprating in line with earnings rather than prices. Additional summary information is provided in Table 7.

On the basis of the trends in public debt and the present expenditure plans of the government, it is possible to suggest that fiscal solvency could be achieved in the medium term without the tax rises announced for next year. However the PSBR/GDP ratio stabilises at around 3 3/4 percentage points, some 1 1/4 percentage points above the projection in the main forecast. The ratio of gross general government debt to GDP by 1999 is correspondingly higher at 53 1/4 per cent, around 5 percentage points above that in the central forecast. Table 7 indicates that the gearing ratio also begins to stabilise in this variant, although the level in 1999 is some 8 percentage points above that achieved with the tax rises in place.

Table 7. Public sector finances in the medium term

all figures per cent

Case A. The main forecast


Case B. The main forecast without future tax rises

Case C. Case B plus uprating with wages rather than prices

 1992/3 1995/6 1997/8 1999/2000

PSBR/GDP ratio
Case A. 6.0 3.9 2.8 2.5
Case B. 6.0 5.4 4.1 3.8
Case C. 6.0 5.9 5.2 5.0

Debt/GDP ratio
Case A. 40.4 47.7 47.8 48.6
Case B. 40.4 50.3 51.7 53.3
Case C. 40.4 51.5 53.8 56.9

Debt/Capital Stock Ratio
Case A. 43.0 59.0 57.0 55.0
Case B. 43.0 62.0 63.0 63.0
Case C. 43.0 64.0 67.0 70.0

The second variant illustrates that the fiscal gains from uprating allowances by prices are equally important in maintaining the prospective future solvency of the public sector. Whilst the PSBR/GDP ratio again stabilises by 1999, albeit at the higher level of 5 percentage points, it would be much harder to claim that the public sector was solvent, with little sign of any end to the trend in the gearing ratio. Moreover, the public sector would be in a position from which it could be extremely difficult to adjust policy in the face of an unexpected shock. A constant PSBR/GDP ratio of 5 per cent will eventually generate a steady state debt/GDP ratio of 88 per cent (given nominal growth of 6 per cent), implying the existence of a sizable primary surplus to ensure that solvency is maintained.

Conclusions

The current fiscal position appears serious when judged by the current PSBR to GDP ratio of 7 per cent. Indeed, the position would be unsustainable if recent trends in spending and taxation were maintained, because then the accumulation of debt and the corresponding interest liabilities would lead to an eventual debt explosion. Even if the present level of the PSBR/GDP ratio could be kept constant (by a primary surplus to offset debt interest) and the long-term growth of nominal income settled down at the rate of 6 per cent as in our forecast then the debt to GDP ratio would approach 125 per cent of GDP. However this is not the outcome we expect. Instead we envisage that the policies that the Government has set in hand will lead to a much lower level of the PSBR to GDP ratio of about 2 1/2 per cent of GDP which will lead to a stabilisation of the ratio of net financial debt to GDP at about 45 per cent. The corresponding figure for the general government gross debt to GDP ratio (the criterion used in the Maastricht treaty) is about 50 per cent.

Even so, the amount of public sector debt is expected to increase in relation to the stock of publicly owned capital. Whilst the debt to GDP ratio is likely to return to a similar level to that pertaining in the late-1970s and early-1980s, the public sector capital stock is likely to be at a much lower level. Since the public sector debt to capital ratio has been fairly constant since the beginning of the 1970s it is not clear what justifies such an increase.

Although an increase might have been appropriate prior to the exploitation of assets such as North Sea oil, it is less so now, given what is known about future trends in the age structure of the population. Ultimately it is not possible to state an optimum level for the net worth of the public sector; in contrast to private companies the government can allow debt to exceed its assets by simply raising the burden faced by future taxpayers. Nonetheless, reforming the present basis of the public sector accounts so as to include a complete balance sheet of known assets and liabilities would provide a better basis on which to judge the present stance of fiscal policy.

The future improvement in the public finances that we expect is shown to come about in a fairly smooth manner without dramatic cuts in spending or substantial increases in taxation. Current spending and tax revenue are expected to return to quite normal levels in relation to GDP, with net capital spending by the public sector rising a little from its current level. However this pattern is not the familiar one that follows the recovery from recessions. In particular, it has been the case in the last two recoveries that income tax receipts have fallen in the recovery rather than increasing as we expect them to do now. Thus the Government will need to restore control over the public finances by raising taxes at a time when previous governments have been able to be slightly more relaxed.

REFERENCES

Davies, S. (1990), 'Fiscal developments and the role of the cycle', Treasury Bulletin, Winter, 1990-91.

Odling-Smee, J. and Riley, C. (1985), 'Approaches to the PSBR', National Institute Economic Review, No. 113, August.

NOTES

(1) Some privatisations such as council house sales are recorded as negative fixed capital formation in the public accounts. The proceeds raised from the sale of former public corporations are shown separately as 'cash expenditure on company securities'.

(2) There are a number of other possible definitions of net capital expenditure. The public expenditure supplement to the Autumn Statement provides figures for 'public sector asset creation'. This is equivalent to our net capital expenditure plus capital grants and defence investment spending (which we include in current expenditure), less the proceeds from privatisations. In 1992/3 public sector asset creation is estimated at 5 per cent of GDP. In principle, investment expenditures could be wider still, including other current expenditures such as education expenditure, see Odling-Smee and Riley (1985).

(3) See Table 11.8 of Inland Revenue Statistics, 1993.

(4) Similar conclusions would emerge if 1990/1, a year in which the public accounts were in balance, is taken as the basis for comparison with 1992/3.

(5) It is possible to envisage situations where privatisation could result in higher net income from public sector assets if the privatised assets are ones from which the government did not obtain any revenue.

(6) With the rapid rate of inflation of most tangible asset prices in the latter half of the 1980s, revaluations are likely to have exceeded the losses due to sales and privatisations. In the 1990s, however, asset prices have declined, accentuating the losses from asset sales. These changes may affect the extent to which assets valued on a replacement cost basis are good approximations to those assets valued at market prices. There is, however, no particular reason to expect this fluctuation in prices to be repeated during the rest of the decade.

(7) A more extensive discussion of these issues is contained in the Institute for Fiscal Studies Green Budget, 1993.

Box A. The PSBR and the Economic and Electoral Cycles

In order to guide our assessment of the effects of the economic cycle on the public finances, we estimated a regression equation which explains the behaviour of an adjusted PSBR to GDP ratio in terms of a measure of the cycle and some other key variables. A related approach is adopted in Davies (1990).

It is important to recognise that the overall effects of the cycle on the PSBR depend to a large extent on whether the authorities are prepared to allow the automatic stabilisers to operate fully or whether they take discretionary action to offset their effect. The equation presented here indicates the typical response of the PSBR to GDP ratio over the cycle and hence includes the effects of discretionary changes in taxation. The equation also allows for the effects of elections which might be expected to increase public borrowing and net financial liabilities of the public sector which will tend to have a negative effect on borrowing. Finally, we also include a term in inflation which might be expected to raise borrowing as it acts to reduce the real value of the outstanding debt stock. The equation is:

|Mathematical Expression Omitted~

(White heteroscedasticity corrected t statistics in parentheses) Sample: 1971/72-1992/93. |R.sup.2~ = 0.75 DW = 1.52 LM(1) = 1.07 RESET(1) = 0.003 NORM(2) = 4.58 HET (1)= 0.11

The variables are defined as follows: APSBR is the PSBR excluding the revenue from oil and asset sales, CYCLE is capacity utilisation in manufacturing, DEBT is the net financial liabilities of the public sector, INFLATION is the rate of growth in the consumers' expenditure deflator and ELECTION is an average of DUM and DUM(-1) where DUM is equal to unity when an election occurs within the financial year and zero otherwise. (Hence DUM is equal to unity in 1973, 1974, 1979, 1983, 1987 and 1992.) The choice of dependent variable was guided by the empirical evidence, but to the extent that the estimated equation accurately reflects the authorities' preferences, it suggests that oil revenues and asset sales have been used to reduce the PSBR rather than to finance tax cuts or expenditure increases. The contribution of these various influences to the annual change in the adjusted PSBR is shown below for the period from 1980/81:

TABULAR DATA OMITTED

This shows that the largest influences on the adjusted PSBR to GDP ratio are the electoral and business cycles. The effect of the debt stock on the PSBR although significant, is not quantitatively important and inflation only has a large effect when it is itself high.

The fit of the equation over the past few years is not particularly good with a string of four positive residuals: most (4.9 percentage points) of the 7.8 percentage point change in the adjusted PSBR to GDP ratio since 1988/89 is unexplained by the equation. This bears out the discussion in the text that the deterioration in the public finances over this period is not easily explained in terms of the economic cycle.

It is interesting to note that the largest error over this period is in explaining the change in the adjusted PSBR to GDP ratio between 1988/9 and 1989/90. On the basis of past experience, the state of the economic cycle and the absence of an election, it would have been reasonable to have expected a reduction in the adjusted PSBR to GDP ratio of 2 1/2 percentage points rather than the 0.6 percentage point increase that actually occurred. It is not immediately obvious to what this error can be attributed except perhaps that it would have been natural to anticipate a tightening of fiscal policy at this stage of the electoral cycle which did not occur and has not been made up since.

It is relatively easy to explain the subsequent deterioration in the public finances since 1990/91 in terms of the economic and electoral cycles. The difficulty is in explaining why the deterioration began from a higher level of the PSBR to GDP ratio than past experience would have indicated.