Fiscal policy: what should the targets be?
The current fiscal framework
The central government budget process was tightened in the mid-1990s, at a time of extensive fiscal consolidation, following the economic crisis of the early 1990s, which had led to increased government spending, large deficits and rising debt. The new approach introduced a medium-term fiscal framework (previously budgeting was only for a single year) and a top-down budgeting process. In 1997, multiple-year expenditure ceilings were adopted, and in 2000 a budget surplus target and balanced-budget requirement for local governments were implemented (Box 2.1).
What has the fiscal framework achieved?
The net lending targets and spending ceilings have been met. Net lending has averaged 1.4% of GDP from 2000 to 2007 (Table 2.1). The indicator showing the average value of net lending since 2000 has exceeded 1% since 2006, and the seven-year moving average indicator has been close to or above 1% since 2003. Estimated structural budget balances show that, unlike in many other OECD countries, discretionary policy has acted counter-cyclically (Ahrend, et al., 2006). The balanced-budget rule for local governments has not been met every year since its inception, with the local government sector as a whole running deficits between 2001 and 2003. However, the rule that local government deficits should be corrected within three years appears to have been met at the aggregate level. The expenditure ceilings have not been breached but the budget margin (the difference between the ceiling and actual spending) has been eroded by additional outlays after the expenditure ceiling was fixed. Moreover, tax expenditures have been used to reduce the risk of the expenditure ceilings being breached. Also, borrowing from the National Debt Office to finance investment spending has been used to avoid breaching the expenditure ceilings since this defers parliamentary appropriation (limited by the expenditure ceiling) from the time that the investment takes place until the time the debt is amortised (Hansson Brusewitz and Lindh, 2005). (1)
The improvement in net lending has been achieved through expenditure restraint rather than higher taxes. The share of public spending in GDP is still the highest in the OECD, for a variety of reasons. A large proportion of individuals' smoothing of their income over their lifetime is carried out through the public sector and a significant share of private services are provided by the public sector. In addition, transfers are higher than in many other countries, although they are generally taxed like earned income. Taking account of taxation of benefits and mandatory private social spending, Sweden's net publicly-mandated social spending is some 7 percentage points of GDP less than gross public social expenditure, and lower than in France and Germany. The trend decline in spending has largely been driven by reduced interest expenditure, in line with lower debt, and a reduction in transfers, particularly in relation to housing, support for the unemployed and labour market measures (Figure 2.1). In recent years, lower transfers have reflected a fall in the number of people excluded from the labour market due to sickness and disability, which to some extent may have been cyclical. Government investment spending as a share of GDP has fallen since the 1970s but is not evidently low by international standards (see below, Box 2.2).
Box 2.1. Key elements of Sweden's fiscal policy framework The main fiscal target is a general government surplus of 1% of GDP over the business cycle (Swedish government, 2008a). * It is motivated by the following goals: * Public finances should be sustainable in the long term. * Fiscal resources should be distributed evenly across generations. * Economic efficiency should be promoted through a predictable development of taxes and expenditures. Three indicators are used by the government to assess performance against the surplus target: * Average net lending from the year 2000, the year in which the target was first applied, combined with information about the estimated output gap over the same period. * Average net lending over a seven-year period, including the current year, the next three and the previous three, combined with information about the estimated output gap over the same period (introduced in the 2007 Spring Fiscal Policy Bill). * The structural balance, i.e. cyclically-adjusted net lending adjusted for one-off and temporary effects. The expenditure ceilings are designed to support the surplus target and prevent temporary revenue from being used to finance permanent expenditure. They take into account economic developments to avoid pro-cyclical fiscal policy and include a budget margin to cope with uncertainties. The aggregate expenditure ceilings are broken down into 26 expenditure areas, covering central government expenditures and the spending of the income pension system, but do not cover interest or local government expenditures. When first introduced, the expenditure ceilings were intended to be set three years in advance. However, for a number of years, the ceilings were only set two years in advance. The three-year-ahead ceiling was re-instated in 2007. The balanced-budget requirement for local governments applies to the aggregate position of all local governments and calls for any deficits to be offset by surpluses within three years. In 2007, the government established a Fiscal Policy Council (Finanspolitiska Radet), to monitor the fulfilment of fiscal targets. It's role includes examining the clarity of budget documents and quality of the data on which they are based, and whether developments are in line with long-term sustainable growth and high employment. The establishment of the Council was also intended to foster public debate on economic policy. The government has begun work on a comprehensive review of the fiscal framework aimed at increasing its clarity and strengthening it to face future challenges. * The surplus target was initially 2% of GDP but was reduced to 1% when the premium pension system, introduced in the reforms of the late 1990s, was reclassified to the private sector in 2007.
A key motivation for the general fiscal policy framework was to strengthen the fiscal position following the economic crisis of the early 1990s. Swedish general government gross debt has fallen from a peak of almost 85% of GDP in 1996 to about 47% of GDP in 2007.
[FIGURE 2.1 OMITTED]
Net debt had deteriorated from a small net asset position in the early 1990s to 27% of GDP in 1996 but the government had restored a net asset position of 20% of GDP by 2007 (Table 2.2). (2) The framework appears to have been successful in helping to restore local government finances, since local government net financial position has also improved. In addition, the stock of physical assets, which are not included in the net assets measure, amounts to some 40 to 50% of GDP. The overall improvement in the government's balance sheet has not been driven solely by net lending: more than half of the increase in net financial assets since the mid-1990s can be ascribed to GDP growth and valuation effects (Swedish Fiscal Policy Council, 2008). Sweden is one of only a few OECD countries that has financial assets in excess of liabilities. In comparison to other OECD countries, it has also achieved one of the largest improvements in net financial assets over the last decade (Figure 2.2).
The government has introduced measures in the 2009 Budget Bill that will reduce net lending by almost 1% of GDP. These measures cover three broad areas: investment in infrastructure and research, lower income taxes, and increased spending on pensioners and psychiatric care. Cyclically-adjusted net lending would be reduced to around 2% of GDP in 2009 although slower GDP growth may bring headline net lending close to balance (Chapter 1).
[FIGURE 2.2 OMITTED]
What should guide the choice of future fiscal targets?
While the fiscal targets have helped restore public finance sustainability, they have been criticised on a number of points. The surplus target was set to achieve fiscal consolidation but the level of 1% of GDP is largely arbitrary. The expenditure ceilings have been met, but they have also been circumvented by the use of tax expenditures and borrowing to finance investment, and they are not clearly linked to the surplus target (Swedish National Audit Office, 2008; European Commission, 2008a; Swedish Fiscal Policy Council, 2008). These problems, and the approach of the demographic transition that the surplus target was intended to facilitate, has led the government to initiate a review of the fiscal framework. This was announced in the Budget Bill for 2007 and the outcomes of this review will be reported by 2010.
The ultimate goal of fiscal policy targets should be to ensure sustainability, meaning that current tax and spending policy settings can remain unchanged in the future without generating an explosive path for government debt, which could ultimately lead to higher interest rates and cause macroeconomic as well as financial market instability. The standard approach to fiscal sustainability rests on the idea that government debt should be stable as a proportion of GDP at some finite horizon (often 50 years). A strict approach to ensuring sustainability would be to set the target to maintain the current financial position. A net lending surplus of about 1% per year would be sufficient to stabilise net assets at their current level of around 20% of GDP, given the government's projections of GDP growth and interest rates. However, this approach hinges on an essentially arbitrary level of net assets. Also, setting a fixed target to maintain net financial assets, even if expressed as an average over the economic cycle, constrains the use of pre-funding or tax-smoothing as a long-term policy tool, implying perhaps temporary changes in tax rates or spending. Also, this approach does not ensure a particular level of intergenerational distribution of resources, which is currently stated as an ambition in the government's fiscal policy statements. Finally, it has been argued that the factors that can affect net financial assets which are not included in net lending (the so-called stock-flow adjustment) could be manipulated to help achieve fiscal targets (Buti et al., 2006).
A more rigorous forward-looking approach would be to set short-run fiscal targets with reference to the existing stock of assets and liabilities, the expected path of revenues and expenditures, and fiscal sustainability calculations based on these projections. This could incorporate an assessment of the uncertainty around long-term projections and an analysis of the risks to the budget resulting from the composition of the balance sheet (for example, the likelihood of default on government-provided loans) and other transactions (for example, government guarantees). The expected paths of balance sheet items and revenues and expenditures are important, in addition to net present values. The inter-temporal budget constraint may be satisfied (the net present value of future budget surpluses may be enough to cover existing net debt), but with a trajectory that involves sharply increasing debt initially followed by a strong fiscal improvement. Such a trajectory would leave the budget more susceptible to the effects of a negative economic shock than a sustainable path with a smoother trajectory for net debt.
Fiscal targets should also take into account the fact that parts of the general government balance sheet will evolve over time according to factors that are exogenous to fiscal policy. For example, higher returns on assets in the pension system would likely reduce future government spending related to the basic age pension. Similarly, higher investment returns on state-owned enterprises would reduce the need to raise taxes to finance a given path of expenditure. However, in both cases, if the higher return was achieved through a more risky investment strategy, the budget may also be more exposed to occasional losses. In addition, a decision to invest in physical capital may increase the required ongoing investment to maintain the asset as it depreciates (Box 2.2). Loans to the private sector increase government assets and interest income but the principal is not captured by net lending. Publication of a full balance sheet with a reconciliation of changes in net worth due to net lending and other factors would make it easier to assess the extent to which a net lending target would need to be adjusted to take these factors into account. Full accrual balance sheets are currently published in the United Kingdom, New Zealand and Australia.
The government also sees a role for fiscal policy in promoting an equitable distribution of resources across generations. To monitor achievement of this goal, generational accounts should be produced on the same basis (i.e., with the same assumptions and methodologies) as fiscal sustainability calculations. Generational accounts can be used to assess the distribution of public sector activities by mapping the taxes paid and public benefits received across age cohorts. They do come with a number of caveats, however. They ultimately rely on the lifecycle consumption model and require assumptions about the extent of intergenerational altruism. They may not be as familiar to policymakers, potentially increasing the risk of mis-interpretation. They treat government payments made today on par with payments to be made in the future, which may be considered uncertain due to the possibility of policy change (Auerbach, 2008). Expenditures that cannot be specifically linked to a particular generation are assumed to benefit all generations equally. This means that expenditures whose costs and benefits do not occur in the same period may be misallocated in the generational accounts (Heijdra and van der Ploeg, 2002). This is a risk with capital investment, where the expenditure is incurred in one period but the benefits are spread out over time. Finally, generational accounts do not capture non-tangible assets such as environmental amenity, so cannot give a complete picture of intergenerational redistribution (Productivity Commission, 2005). Despite these drawbacks, they can usefully inform target setting and assessment of policy achievement.
Box 2.2. Is a golden rule necessary? A risk with a strict rule-based fiscal framework with a net lending target and expenditure ceilings is that these rules may discourage physical capital investment. Such expenditures may be politically easier to defer in order to meet a net lending or expenditure target, because they are discrete and often one-off. It can be argued that government investment is different from other types of spending because the benefits are spread further over time and investment increases private sector output (Fatas, 2005). Hence, a number of countries (the United Kingdom, Germany) have adopted a "golden rule", where budget deficits are allowed but only up to the amount of capital spending (that is, debt issuance can finance investment). Such a rule provides for the costs and benefits of an investment to be spread over a similar period, rather than the costs being borne upfront and the benefits accruing later. Over a long time span, government investment as a share of GDP has fallen, but it has been stable since the mid-1990s, suggesting that the fiscal framework introduced in the late 1990s did not induce a fall in public investment (although it is possible that the investment share of GDP would have increased in the absence of the rule). Estimates of the stock of public capital suggest that Swedish public capital may be low by international standards (Kamps, 2004). Excluding investment from the fiscal target could make fiscal policy more volatile and loosen control over the aggregate budget position. If capital spending is politically easy to defer in order to meet a target, it may not have the same public value as other types of spending so loosening the budget constraint for capital spending and not other spending may not be consistent with public preferences. The definition of investment (expenditures that result in the acquisition of physical capital) most likely would exclude expenditures providing long-term benefits, such as education or research and development. In addition, a golden rule may create incentives to artificially reclassify certain types of spending as investment in order to bypass the budget constraint. In addition, a golden rule might create a bias in favour of purchasing physical capital over other forms of capital service provision. Government can either buy capital outright, affecting net lending upfront, or it can contract a private enterprise to construct the capital and pay a rental fee or service charge as the infrastructure is used by the public. Such an arrangement would be classified as consumption for the purposes of net lending. A golden rule loosens the budget constraint for purchasing but not for renting capital. Reporting a balance sheet including physical assets would improve coverage of the government asset position and could prompt further analysis of the role of government-owned assets and how they are expected to develop over time. The use of an accrual-accounting flow measure for the budget target, incorporating depreciation as an expense, would replace the upfront cost of investment by the depreciation of the asset over time in the fiscal target. This would facilitate the choice between purchasing and renting capital and create an explicit link between the fiscal flow target and government net worth. However, using an accrual measure for the budget flow target raises a number of complex issues which would need to be carefully considered, such as the cash management implications of budget allocations for accruing expenses like depreciation and civil servant pension liabilities (OECD, 2007a).
Adopting an approach to setting fiscal policy with reference to projected long-term developments would be more complex than the current surplus target. In particular, it could be argued that the government could manipulate the underlying modelling and assumptions. This may weaken the discipline imparted by the fiscal targets. To guard against this, the Fiscal Policy Council should be required to formally verify the assumptions and methodologies used to determine the targets. (3) The targets could be re-affirmed or adjusted at regular intervals, based on updated long-term projections and refinement of the estimation methodologies, to allow for early action if the impact of demographic or other changes turned out to be larger than expected.
A number of OECD countries, particularly those which have recently recorded budget surpluses, have adopted policy targets derived from projections of future fiscal developments. In New Zealand, the government is required by law to produce a long-term fiscal outlook every four years and a fiscal strategy report each year. The latter report demonstrates how fiscal targets are linked to long-term developments. The budget is prepared on an accrual basis, and the target is for operating surpluses (accrued revenues must exceed accrued expenses, regardless of the actual cash flow) over the economic cycle consistent with increasing net worth. The exact level of the surplus is determined by what is required to achieve the long-run objectives (New Zealand government, 2008). In Denmark, the budget target is for structural surpluses of 3/4-1 3/4 per cent of GDP per year to 2010 and then the budget should at least be balanced from 2011 to 2015. These targets were set with reference to the expected future development of public finances, to ensure fiscal sustainability measured using a variant of the inter-temporal budget constraint (OECD, 2008a). The sustainability indicator is also used to assess the long-term implications of major policy measures. Such an approach could be usefully applied in Sweden, building on the long-term projections already presented each year in the Budget Bills (Swedish National Audit Office, 2007b). (4)
The link between the expenditure ceilings and the net lending target should be clarified. With a target for net lending tied to the development of the balance sheet, the purpose of an expenditure ceiling would be to constrain the overall size of government. If tax smoothing is pursued, the expenditure ceiling becomes the control variable to ensure that the net lending target is met given the anticipated path of revenue. The expenditure ceiling has been an important fiscal control mechanism in recent years and has been met every year since its introduction. The level of ceiling is currently set to allow for cyclical fluctuations in spending, with the three-year-ahead ceiling set to be 3% above the estimate of ceiling-restricted expenditure. However, in the past, the budget margins have been fully used during upswings, forcing cuts in expenditures during downturns. This could be addressed by increasing the size of the budget margins, although larger margins might be interpreted as additional scope for permanent spending increases. Alternatively, consideration could be given to excluding cyclically-sensitive expenditures from the expenditure ceiling to prevent them from limiting the operation of the automatic stabilisers. Consideration could also be given to introducing a formal assessment of new tax expenditures as part of the budget decision-making process, in order to prevent their use as a means to bypass the expenditure ceilings (Swedish National Audit Office, 2007a). This would ensure that tax expenditures at least receive the same scrutiny as direct expenditures in the budget process. Similarly, arrangements for borrowing from the National Debt Office to finance capital investment should also be eliminated. It is therefore welcome that the 2009 Budget states that government investment should be financed from general revenue, like other expenditures, rather than by separate borrowing. Part of the 2008 surplus (0.75% of GDP) will be allocated to repayment of loans previously taken out by the National Rail Administration and Swedish Road Administration. Loan financing will not be considered for new investment projects unless the costs are to be recovered by user charges (Swedish government, 2008a).
The local-government balanced-budget rule should be amended to reduce the risk of pro-cyclical local fiscal policy. The balanced budget rule for local governments is a minimum requirement and is often interpreted by local governments as requiring a small surplus. However, if a deficit does emerge the requirement that the financial position is to be restored within three years raises the possibility that local governments may need to tighten fiscal policy during economic downturns. Local governments should have the incentive to save during economic upswings to avoid having to raise taxes during slowdowns, as happened in the early 2000s. The 2005 Survey recommended reforming the balanced-budget requirement so that revenue for the purposes of the target is calculated with reference to the average taxable income over a number of years, creating a form of over-the-cycle target (OECD, 2005).
Consideration could also be given to formalising periodic reviews of the overall allocation of spending, to supplement reviews of the effectiveness of particular programmes. The Swedish annual budget process is incremental, considering new policy proposals although often with a requirement that new spending is offset with savings elsewhere in the budget. While it would be resource-intensive, multi-year reviews could free up resources from low-priority areas to ensure overall spending restraint and comply with the fiscal targets. The Canadian government has recently introduced a four-year spending review cycle with a view to freeing resources for reallocation to other priorities, although to date this process has not yielded significant savings (OECD, 2008b). The UK government conducted comprehensive spending reviews in 1998 and 2007, which examined all government programmes from a zero-base perspective, helping to reallocate spending towards higher priorities, in health and education in particular Bi-annual spending reviews were conducted in the interim setting out three-year spending plans for departmental expenditures. However, actual real spending has exceeded that planned in all spending reviews since 1998 and a considerable fiscal deficit has emerged (OECD, 2007b). Thus, spending reviews do not automatically deliver spending control; they are not a substitute for fiscal discipline.
Are current policies sustainable given future fiscal pressures?
The government's long-term fiscal projections indicate that the general government budget balance will strengthen to about 2.5% of GDP in 2011 (the last year of budget forecasts) and is then projected to deteriorate progressively to 1% of GDP by 2015. Beyond that horizon, net lending is driven by demographic changes, deteriorating to deficits of around 1.5% of GDP in the 2040s before recovering slightly to a deficit of 1.2% of GDP in 2050 (Figure 2.3). Accordingly, the net financial asset position is projected to improve as a share of GDP until about 2020, and then to deteriorate, to net debt of 3.3% of GDP by 2050.
Based on a number of technical assumptions, (5) including that of an unchanged retirement age despite the foreseeable increase in longevity, the government projects transfers and consumption combined to decline by about 1.5 percentage points of GDP between 2008 and 2050. The fiscal gap opens up because of a technical adjustment--a transfer of 3.3% of GDP to households--phased in between 2011 and 2015. This adjustment is introduced in order to reduce the surplus from its level in 2011 to the target of 1% of GDP by 2015, after which the projections are based on demographics. This technical adjustment changes with the projection for the surplus at the end of the budget forecast horizon. For example, in the 2008 Spring Fiscal Policy Bill, the surplus in 2011 was estimated to be 4.1% of GDP, and the technical adjustment required to reduce the surplus to 1% in 2015 was therefore 5% of GDP per year (Swedish government, 2008b).
[FIGURE 2.3 OMITTED]
Demographically-driven pressures are mainly in elderly care. The government projects consumption related to elderly care to rise by about 2 percentage points of GDP between 2008 and 2050. Consumption of medical services is expected to rise by about one-half a percentage point and transfers related to health are expected to fall by about one-half a percentage point of GDP, based on a fixed standard of care. Pension expenditure is projected to be broadly stable as a share of GDP until 2040, falling thereafter, due largely to the pension reforms of the late 1990s, which reduced pension entitlements and introduced the self-balancing income pension system. Through the universal guarantee pension, the government retains the risk that retirement income from the new income pension system is inadequate. Education spending is projected to decline by about 1 percentage point of GDP between 2008 and 2050 due to demographics.
OECD projections of growth in health and long-term care spending based on demographics alone are lower than those of the government. They take into account an adjustment for healthy ageing, and an income effect for age care which assumes that the income elasticity of demand for aged care services is zero, so income growth drives down long-term care spending as a share of GDP. Even projections that take into account non-demographic factors, combined with measures to contain the growth in costs (the cost-containment scenario in Figure 2.4), suggest that Sweden faces the lowest increase in health and long-term care spending in the OECD, in part due to a relatively small increase in the old-age dependency ratio. However, expectations for standards of care will almost certainly rise, and much larger increases in spending are projected in the absence of cost-containment measures--public spending on health and long-term care could rise by more than 4 percentage points of GDP between 2005 and 2050 relative to the impact of demography alone (OECD, 2006).
Projections produced by the European Commission, which cover pensions, education and labour market measures in addition to health and age care, are more pessimistic than the government's. That said, even these projections show that Swedish public finances face a smaller demographic challenge than in many other European countries. These estimates take account of reduced care needs due to improved health accompanying increased life expectancy. However, the government expects larger falls in both pension and unemployment benefit spending.
[FIGURE 2.4 OMITTED]
Using internationally comparable figures, Sweden is one of a handful of European countries which are considered to have sustainable fiscal positions. This is based on an indicator which shows how much the budget balance would need to change each year in order to satisfy the inter-temporal budget constraint. A negative value of the indicator suggests that the fiscal stance can be permanently loosened and fiscal policy would still be sustainable (European Commission, 2004). In Sweden's case, the sustainability indicator value is -1.2; that is, the primary budget surplus (total budget surplus less net interest) could be permanently reduced by 1.2% of GDP and the fiscal position would still be sustainable. This result is achieved because the starting surplus and initial debt position are sufficient to offset the impact of the fiscal deterioration associated with demographic change (Figure 2.5). (6)
[FIGURE 2.5 OMITTED]
The government's calculations of fiscal sustainability show a less positive picture, but still point to long-term sustainability. The central scenario yields a sustainability indicator value of -0.1. The calculations are based on the European Commission indicator of sustainability but more optimistic projections for long-term spending, as outlined above. In addition, the projections cover a longer time horizon, which allows them to take on board a projected improvement in the fiscal position beyond 2050. However, as noted above, the government's projections also incorporate an ad hoc technical adjustment of 3.3% of GDP, designed to bring net lending to the 1% surplus target by 2015. (7) Without this adjustment, surpluses would deteriorate from around 4% in 2015 until the early 2030s, but even at the height of the impact of demographic change, the budget would most likely still be in surplus. Calculations made prior to the fiscal measures announced in the 2009 Budget indicated that, without the technical adjustment, the fiscal stance could be permanently loosened from its 2007 level by nearly 4 percentage points and still be considered sustainable, assuming no other adjustment to the projections (Swedish Fiscal Policy Council, 2008). There is clearly a need to refine the approach to moving from the current path of net lending to the 1% target (Swedish Fiscal Policy Council, 2008; Swedish National Audit Office, 2008).
At present, there is no formal reporting of whether fiscal policy is achieving the government's intergenerational aspirations. Generational accounts have been produced for Sweden, however, they are based on different assumptions from the government's. (8) These accounts show that all generations born between 1930 and 2009 are to receive a net transfer from the government. This comes about because government debt has been accumulated since 1930, and the share of debt to GDP is posited to be the same at the end of the projection period in 2110 as in 2006. Thus, the accounts support for the idea that current generations are receiving benefits that are being financed by future generations (Pettersson et al., 2007).
How should public spending growth be tackled?
As outlined above, future fiscal pressures are likely to derive from two main sources: the impact of demography on age care and health spending, which is already high in comparison to other OECD countries, and the impact on public spending of technological change and increasing service demands. There are several policy approaches to dealing with these fiscal pressures. Tax rates could be raised as budget pressures arise to ensure that the budget remains balanced. However, the impact of globalisation on tax bases and the incentive effects of high labour taxes are likely to put downward pressure on tax rates. While Sweden was an early mover in reducing corporate income tax rates, there has recently been further pressure for further tax cuts, culminating in the government announcing a reduction in the corporate income tax rate in the 2009 Budget. High marginal tax rates on labour are likely a contributing factor behind relatively low average hours worked. These issues are taken up in more detail in Chapter 3. The alternatives to higher taxes are pre-funding, measures to increase the growth of tax bases and measures to address the expected growth in spending or reduce spending in other areas to create room for additional demands. These are discussed below.
The current fiscal target, and projected surpluses in excess of the target, implies that fiscal policy is currently pre-funding, or saving, to meet future fiscal pressures. Pre-funding is appropriate from an efficiency perspective when it results in smoothing tax rates through time. That is, a temporary period of surpluses can finance a subsequent temporary period of deficits while maintaining constant tax rates throughout. This is efficient since the deadweight costs of taxation are posited to be more than proportional to the tax rate--stable tax rates produce lower deadweight losses than near-term falls followed by future increases in tax rates. Pre-funding is not appropriate for dealing with permanent increases in spending, such as those associated with higher service demands or future technological change. The gains from tax smoothing compared to a balanced-budget strategy may be small, possibly less than 1% of GDP, but could be significantly larger, up to over 5% of GDP, if the initial surpluses are invested in assets earning a higher rate of return than government debt (Davis and Fabling, 2002) (9). The benefits are also larger the greater the expected increase in government spending and the higher the initial tax rates.
However, there are reasons--notably political economy ones--to be cautious about the extent of pre-funding. Pre-funding might create the perception that the government has excess resources in the short term, creating incentives for additional public spending and reducing incentives to address spending growth. This would exacerbate the fiscal problem pre-funding was designed to help alleviate (Pinfield, 1998). If pre-funding involves asset accumulation, these problems can be partly mitigated by specifically linking the assets to already committed future expenditures and ensuring a strong governance framework for the management of the assets. This "expenditure creep" may not be a major problem in Sweden's case since the fiscal surpluses are being applied to reducing gross debt and accumulating assets in the income pension system, which has clear rules for how and when the assets can be used. Also, clearly explaining the rationale for surpluses in the near term to help meet longer-term fiscal pressures should reduce the demands for higher spending in the short run. However, the higher the degree of pre-funding in the short term, the larger the risk that it would lead to increased spending pressures.
The intergenerational equity implications of pre-funding also need to be further investigated. Scenario analysis conducted with the generational accounts shows that replacing the 1% surplus target by a balanced-budget rule results in generations born before 1960 receiving higher net transfers from the government and generations born after 1960 receiving lower net transfers--sharply lower in the case of those born since 1990. This supports the view that a strategy of near-term surpluses should be maintained (Pettersen et al., 2007). However, it does not show the intergenerational effects of surpluses of more than 1% in the coming years or higher spending in the long term. There would be more transparency if generational accounts were produced with the same projections and assumptions as sustainability analysis. From an ex post perspective, too little pre-funding implies that current generations receive benefits paid for by future generations, while too much pre-funding generates the opposite effect. This calls for regular review of estimates of the long-term fiscal trends and regular re-assessment of the required extent of pre-funding.
The sustainability analysis outlined earlier suggests that the modest easing in the fiscal stance in the 2009 Budget is unlikely to jeopardise long-term sustainability even though it will reduce pre-funding (Chapter 1). Indeed, the strong fiscal framework and improved fiscal position means that Sweden is much better prepared than many OECD countries to deal with the current downturn. If the economic situation deteriorates further in the wake of the global financial crisis and additional fiscal stimulus is deemed necessary, it should be designed to avoid negative long-term fiscal implications. As recognised in the Budget documentation, fiscal balances often worsen more during downturns than anticipated using standard cyclical-adjustment techniques. Any new measures should either be temporary or aimed at boosting the economy's growth potential. The high marginal tax rates and the associated deadweight costs (discussed in Chapter 3) suggest that income tax reduction should continue to have priority. Some areas of public spending have been linked to higher national income, e.g. education, childcare, research and development and infrastructure (Alonzo at al., 2005) and so the 2009 Budget's focus on infrastructure and research is welcome. However, not all of the new spending measures in the Budget will be growth enhancing and these measures reduce the scope for further tax cuts. In addition, what ultimately matters is not the level of spending but the outcomes achieved and, in some cases, such benefits could be reaped without additional funding, through efficiency measures or increased private contributions towards currently publicly-funded services (see below).
Increasing labour utilisation
Further increases in labour utilization would expand the tax base and reduce public social spending. (10) This has been the government's focus in recent years. The Fiscal Policy Council estimates that larger-than-anticipated effects of the recent labour market policies could allow for a permanent relaxation of the fiscal stance of about half a per cent of GDP (Swedish Fiscal Policy Council, 2008). There is, of course, a natural limit to the expansion of employment, since ultimately individuals' decisions about whether to work or not are based on their preferences with respect to leisure and the consumption of goods purchased with the returns from supplying labour. That said, the current level of hours worked per person and the extent of exclusion from the labour market, described in Chapter 1, suggest that there is a long way to go in raising labour supply before these natural limits would be binding.
Increasing the statutory retirement age may reduce expenditure on the basic pension and generate additional tax revenue. Raising the average retirement age would reduce the longevity risk borne by the government in the income pension system (the risk that the income pension system will not deliver "sufficiently" high retirement income due to longer life). Finland, Germany, and Portugal have linked pension benefit levels to life expectancy. In Denmark, a decision has been taken to index the retirement age to life expectancy, although this will not apply until 2024 (Denmark faces a much larger problem with early retirement than Sweden, as documented in OECD, 2008a). The Swedish income pension system already entails some incentives to extend working life, since additional years of work raise the replacement rate in retirement. Also, the system allows pensioners to access part of their savings while still working. This eases the transition from work to retirement and in principle could extend people's working lives. The Fiscal Policy Council estimates that gains from increasing the retirement age could be enough to permanently reduce the budget surplus by around 1% of GDP (Swedish Fiscal Policy Council, 2008).
Raising public sector efficiency
The efficiency of public service delivery could be increased through greater use of contracting out and increasing the degree of user choice. Such mechanisms have been shown to lower the cost of service delivery, improve outcomes (for example in education), and increase the flexibility and tailoring of services to consumer needs (Lundsgaard, 2003). Sweden ranks around the middle of a sample of OECD countries for the extent of user choice and competition in local government service provision and at the top in terms of private provision and contracting out (Figure 2.6).
[FIGURE 2.6 OMITTED]
The Swedish Competition Authority has argued that competition in public procurement in Sweden appears to be moving in the wrong direction. In a growing number of instances, procurement entities are receiving fewer tenders. Requirements that government agencies post public notification of contracts which have a value exceeding particular thresholds are being disregarded. In addition, there is a tendency to use the lowest price as a selection criterion, which may adversely affect incentives to improve service content (Swedish Competition Authority, 2007). There may be reluctance on the part of municipalities and state institutions to change the way they operate to facilitate more competitive tendering (OECD, 2007c). Recently, a proposal has been put forward to expand user choice in social services and health care. This would involve setting a contract with a fixed price and fixed quality standards and allowing the beneficiary to choose amongst alternative providers (SOU, 2008). However, the proposal does not require municipalities and county councils to use such a system. Although mandatory action might be seen as impacting on the autonomy of local governments, it may be necessary in order to bring about desired changes in the implementation of public procurement. Mandation would need to be accompanied by effective sanctions for non-compliance (OECD, 2007c).
Shifting the balance between public and private funding
Requiring individuals to finance more of the consumption elements that are currently publicly provided would reduce pressure on the spending side of the budget. Currently, some 80% of all government redistribution is over the lifetime of the same individual (Pettersson and Pettersson, 2007). Increasing private provision of some services that are currently publicly provided, but are not pure public goods, may reduce over-consumption and help contain costs. User fees are also a complement to choice and contestability in public service delivery, in the sense that better-tailored and higher-quality public service delivery may simply create more demand for these services. Price signals are needed to keep demand under control. At present, user fees at the local government level, where many public services like sub-tertiary education, health and medical care are delivered, are used only moderately compared to other OECD countries (Figure 2.7).
[FIGURE 2.7 OMITTED]
There are a number of areas where further user charging could be considered. Public spending on long-term care is high by international standards and is the area of spending that is set to rise most due to demographic change. More private contributions towards elements of age-care expenditures, such as home care, may be warranted. Currently, around 4% of municipalities' costs for health and social care for the elderly are financed by charges. In 2007, the highest charge for home aged-care services was SEK 1 612 (around 175 [euro]) per month. Around 22% of all people granted home care received between 10 and 25 hours per month and 39% received between I and 9 hours a month. The public sector share of total health expenditure is the fourth highest in the OECD and fees play a relatively limited role in financing health and medical care. Within certain limits, county councils are free to decide how much patients must pay but a high-cost protection scheme exists, ensuring that no patient ever needs to pay more than a total of SEK 900 (around 100 [euro]) over a 12-month period. Greater private contributions to health and age-care funding could take the form of higher user charges, reducing the need for pre-funding, or introducing a form of individual accounts, as in the pension system, aimed partly at pre-funding, as outlined in the 2005 Survey (OECD, 2005). Public spending on tertiary education as a per cent of GDP is fourth-highest in the OECD, but private funding of tertiary education is very low by international standards suggesting scope to raise education funding with greater private contributions, as discussed in Chapter 4.
The fiscal policy framework in Sweden has helped lower public debt and smooth the economic cycle. The fiscal surplus will be reduced by the current economic downturn but the cyclically-adjusted surplus is expected to remain above the target of 1% of GDP in the next few years. This is likely to lead to pressure for permanent increases in spending. The question of how quickly the surplus should be returned to target is intertwined with the issue of what the target should be. A way forward is to develop a full government balance sheet and consistent fiscal sustainability and intergenerational equity calculations. These could then be used to guide the setting of the fiscal targets over a number of years. Using these calculations more actively and publicly to explain the benefits of various policy options would help ensure that fiscal policy retains a long-run focus (Box 2.3).
Box 2.3. Summary of fiscal policy recommendations * Fiscal sustainability should be emphasized as a prime goal. Transparent assessment of sustainability should be used more in discussing the long-term effects of proposed policies. * Fiscal targets should be set with reference to the existing stock of assets and liabilities, and the future path of spending and revenues. The government's balance sheet should be presented in budget reporting. * To better appreciate the implications of fiscal policy for intergenerational distribution, generational accounts should be produced and published using the same data and assumptions as the fiscal sustainability calculations that are already produced. * In the process of establishing a new set of fiscal targets, the Fiscal Policy Council should be required to formally verify the assumptions and methodologies used. * Periodic reviews of the overall allocation of spending could be considered to assess how new spending pressures arising from greater service demands or technological change could be financed either via private spending or via savings on existing programmes. * Consideration should be given to formally linking the retirement age to life expectancy. * User choice and contestability in publicly-funded services should be expanded. * User charges and other forms of private funding should be given a bigger role in many parts of the services that are publicly funded today in order to limit excessive use and free up resources for other priorities.
Afonso, A., W. Ebert, L. Schuknecht and M. Thone (2005), "Quality of Public Finances and Growth", ECB Working Paper No. 438.
Ahrend, R. P. Catte, and R. Price (2006), "Interactions Between Monetary and Fiscal Policy: How Monetary Conditions Affect Fiscal Consolidation", OECD Economics Department Working Paper No. 521.
Auerbach, A. (2008), "Long-term Objectives for Government Debt", Paper prepared for the Fiscal Policy and Labour Market Reforms conference organised by the Swedish Fiscal Policy Council, Stockholm, 29 January.
Blochliger, H. (2008), "Market Mechanisms in Public Service Provision", OECD Economics Department Working Paper No. 626.
Buff, M., J. Nogueira, and A. Turini (2006), "From Deficits to Debt and Back: Political Incentives Under Numerical Fiscal Rules", Fiscal Indicators, Proceedings of the conference of the Banca D'Italia Workshop on Public Finance, 30 March-1 April 2006.
Davis, N. and R. Fabling (2002), "Population Ageing and the Efficiency of Fiscal Policy in New Zealand", New Zealand Treasury Working Paper No. 02/11.
Eichengreen, B., R. Hausmann, and J. Von Hagen (1999), "Reforming Budgetary Institutions in Latin America: The Case for a National Fiscal Council", Open Economies Review, No. 10.
European Commission (2004), "The Long-Term Sustainability of Public Finances in the European Union", European Economy No. 4/2006, Brussels.
European Commission (2008a), "Sweden: Macro Fiscal Assessment", ECFIN/REP50205/08-EN, Brussels.
European Commission (2008b), Public Finances in the EMU 2008, Brussels.
Fatas, A. (2005), "Is There a Case for Sophisticated Balanced-Budget Rules?", OECD Economics Department Working Paper No. 466.
Hansson Brusewitz, U. and Y. Lindh (2005), "Expenditure Ceilings and Fiscal Policy: Swedish Experiences", Public Expenditure, Proceedings of the conference of the Banca D'Italia Workshop on Public Finance, 31 March-2 April 2005.
Heijdra, B. and F. van der Ploeg (2002), Foundations of Modern Macroeconomics, Oxford University Press, Oxford.
Kamps, C. (2004), "New Estimates of Government Net Capital Stocks for 22 OECD Countries 1960-2001", IMF Working Paper No. 04/67.
Lundsgaard, J. (2003), "Competition and Efficiency in Publicly Funded Services", OECD Economic Studies No. 35, 2002/2.
New Zealand Government (2008), Budget 2008: Fiscal Strategy Report, 22 May.
Norwegian Ministry of Finance (2001), "Guidelines for Economic Policy", Report No. 29 to the Storting.
OECD (2005), OECD Economic Surveys: Sweden, Paris.
OECD (2006), "Projecting OECD Health and Long-term Care Expenditures: What are the Main Drivers?", OECD Economics Department Working Paper No. 477.
OECD (2007a), "Performance Budgeting and Accrual Budgeting: Decision Rules or Analytic Tools?," Paper prepared for the Working Party on Senior Budget Officials, 11 January, Paris.
OECD (2007b), OECD Economic Surveys: United Kingdom, Paris.
OECD (2007c), Sweden: Achieving Results for Sustained Growth, OECD Reviews of Regulatory Reform.
OECD (2008a), OECDEconomic Surveys: Denmark, Paris.
OECD (2008b), OECD Economic Surveys: Canada, Paris.
Pettersson, T., T. Pettersson, and A. Westerberg (2007), "Intergenerational Distribution In a Growing Welfare State", Paper presented at the 1st General Conference of the International Microsimulation Association in Vienna, 20-22 August.
Pettersson, T. and T. Pettersson (2007), "Lifetime Redistribution Through Taxes, Transfers and Non-Cash Benefits", in Harding, A. and A. Gupta (eds.), Modeling Our Future - Population Ageing, Social Security and Taxation, Elsevier.
Productivity Commission (2005), "Economic Implications of an Ageing Australia", Productivity Commission Research Report, 24 March.
Pinfield, C. (1998), "Tax Smoothing and Expenditure Creep", New Zealand Treasury Working Paper No. 98/9.
SOU (2008), Lag om valfrihet (Law about choice), Swedish Government Official Reports, No. 2008:15.
Swedish Fiscal Policy Council (2008), Svensk Finanspolitik (Swedish Fiscal Policy), Stockholm.
Swedish Competition Authority (2007), Competition in Sweden 2007, Stockholm.
Swedish Government (2007), Sweden's Economy: From the Budget Bill for 2008, Stockholm.
Swedish Government (2008a), Budget Bill 2009, Stockholm.
Swedish Government (2008b), Spring Fiscal Policy Bill 2008, Stockholm.
Swedish National Audit Office (2007a), "The government's Preparation and Statement of Tax Expenditure", RIR 2007:3, Stockholm.
Swedish National Audit Office (2007b), "The government's Analysis of the Long-Term Sustainability of Fiscal Policy", RIR 2007:21, Stockholm.
Swedish National Audit Office (2008), "Tillampningen av det finanspolitiska ramverket--Regeringens redovisning i 2008 ars ekonomiska varproposition" (Adjustment of the Fiscal Policy Framework--the government's reporting in the Spring Fiscal Policy Bill 2008), RIR No. 2008:15, Stockholm.
Wyplosz, C. (2002), "Fiscal Policy: Institutions us. Rules", HEI Working Paper No. 03/2002, Graduate Institute of International Studies, Geneva.
(1.) It should be noted that borrowing for investment through the National Debt Office is permitted by the budget laws.
(2.) Since the early 1970s, the government has held 50 to 70% of GDP in financial assets, including a large state-owned enterprise sector and assets in the social security system. Following the pensions reforms of the late 1990s, the assets in the social security sector mainly relate to the income pension system.
(3.) An alternative approach would be to require an independent authority to provide the government with a path of net lending which ensured fiscal sustainability and information on the implications for intergenerational equity. The government would be expected to implement a fiscal policy that achieved this path of net lending, but with a tax and spending mix of its choice (Eichengreen et al., 1999; Wyplosz, 2002). This model has not been implemented in any OECD country and would imply a considerable strengthening of the role of the Fiscal Policy Council
(4.) In Norway, which is a somewhat special case, the policy framework aims to spread the benefits of future oil extraction evenly over generations. The main fiscal target is for the central government non-oil structural budget deficit to equal 4% of the asset value of the Government Pension Fund (subject to some specific exceptions). Budget documentation reports structural fiscal balances including and excluding oil revenue and generational accounts are updated each year (Norwegian Ministry of Finance, 2001).
(5.) The old-age dependency ratio is posited to rise from 30% in 2007 to 44% in 2050 while the youth-dependency ratio remains stable at around 40%. Average working hours, the employment rate and the unemployment rate are constant in groups defined by sex, age and country of birth, so population structure change drives potential employment. Tax receipts evolve with constant tax rates for each tax base. Tax bases grow according to the underlying economic assumptions and the implicit tax rate is calculated as the ratio of tax revenue to the tax base applying in 2011. Public consumption expenditure is based on age-and sex-disaggregated unit costs for childcare, education, medical care, elderly care and labour market measures. Volume projections are based on the population projections, implying that public service provision is carried out with the same staffing ratio as at present. Other public consumption elements (general administration, judiciary, defence, etc.) follow the change in total population. Productivity growth in the public sector is assumed to be zero, but public sector wages grow in line with private sector wages (which in turn are driven by private sector productivity). Public transfers are assumed to grow in line with wages, even where this is not the case under current policy (Swedish government, 2008a).
(6.) Technically, a negative figure for the sustainability indicator suggests that policy is unsustainable in the sense that it would eventually lead to explosive asset accumulation.
(7.) The strength of the surpluses in the period to 2011 may partly reflect the fact that not all expenditure programmes are automatically indexed to prices or wages. The surpluses (and hence the required technical adjustment) may be reduced if there are decisions to increase the fixed nominal value of such programmes.
(8.) These results were obtained assuming a 1% of GDP surplus through 2015. Thereafter, spending was to be driven by demographic change, but the budget deficit was limited to 2% of GDP. This was achieved by allowing taxes to vary. The calculations are not directly comparable to the sustainability calculations presented in the budget documents since they do not take into account recent above-target surpluses and incorporate a limit on budget deficits.
(9.) This is based on a simulation using historical rates of return on various asset classes. It is possible that reduction in government debt and government purchases of private financial assets could change these rates of return, which would affect the calculated gains from tax smoothing.
(10.) Higher productivity growth would also increase the growth of tax bases, but would add to growth in public spending, since some public benefits (like old-age pensions) are linked to growth in average wages.
Table 2.1. Fiscal balance and expenditure: have the targets been met? As per cent of GDP 2000 2001 2002 2003 2004 Net lending 3.7 1.7 -1.4 -1.2 0.6 Central government 4.0 7.3 -1.9 -1.9 -0.5 Local government 0.2 -0.2 -0.5 -0.2 0.2 Pension system -0.4 -5.4 1.0 1.0 0.9 Net lending average from 2000 to given year 3.7 2.7 1.3 0.7 0.7 7-year moving average of net lending 0.9 0.9 Cyclically-adjusted net lending 3.0 1.8 -0.9 -0.6 0.7 Structural balance 1.9 2.2 -0.7 -0.4 0.4 OECD cyclically-adjusted net lending 2.4 1.0 -1.7 -0.8 0.9 Expenditure ceiling 34.0 34.0 33.5 32.7 32.7 Actual spending subject to ceiling 33.8 33.8 33.5 32.6 32.6 Budget margin 0.2 0.2 0.0 0.1 0.1 2008 2009 2005 2006 2007 est. est. Net lending 2.0 2.2 3.5 2.8 1.1 Central government 0.4 0.8 2.1 1.6 0.5 Local government 0.6 0.3 0.3 0.3 0.1 Pension system 1.0 1.0 1.1 0.8 0.4 Net lending average from 2000 to given year 0.9 1.1 1.4 1.5 1.5 7-year moving average of net lending 1.0 1.4 1.8 2.1 Cyclically-adjusted net lending 1.7 1.3 2.8 3.2 2.1 Structural balance 1.1 0.8 2.1 2.8 1.9 OECD cyclically-adjusted net lending 2.4 1.7 3.0 3.2 2.2 Expenditure ceiling 31.8 31.3 30.5 30.0 30.0 Actual spending subject to ceiling 31.6 30.9 29.6 29.4 28.8 Budget margin 0.2 0.4 0.9 0.5 1.1 Note: The 7-year moving average is calculated using the given year, the three years prior and the three following years. Cyclically-adjusted net lending is calculated using net lending and the "adjustment for the economic cycle" from Table 20 of Annex 2 to the 2009 Budget Bill and the equivalent table in the 2008 Budget Bill for figures relating to 2000 and 2001. Source: Swedish government (2007, 2008a), OECD Economic Outlook No. 84 database. Table 2.2. General government balance sheet Consolidated, per cent of GDP 1996 Assets Liabilities General government Net financial position -26.6 Total financial instruments 72.4 99.0 Debt 45.3 90.3 Equity 20.6 Other 6.5 8.7 of which: Central government Net financial position -1.5 Total financial instruments 27.8 89.4 Debt 9.2 84.2 Equity 15.6 Other 3.1 5.1 Local government Net financial position -0.9 Total financial instruments 8.7 9.6 Debt 4.2 6.1 Equity 1.2 Other 3.3 3.5 Social security Net financial position 35.8 funds Total financial instruments 35.9 0.0 Debt 31.9 0.0 Equity 3.8 Other 0.1 0.0 2007 Assets Liabilities General government Net financial position 19.9 Total financial instruments 70.5 50.6 Debt 29.2 45.8 Equity 36.4 Other 4.9 4.9 of which: Central government Net financial position -10.3 Total financial instruments 32.3 42.5 Debt 11.9 40.0 Equity 17.7 Other 2.6 2.6 Local government Net financial position 1.2 Total financial instruments 8.8 7.6 Debt 5.9 5.6 Equity 1.3 Other 1.6 2.0 Social security Net financial position 29.0 funds Total financial instruments 29.5 0.5 Debt 11.3 0.2 Equity 17.4 Other 0.7 0.3 Source: Statistics Sweden Financial Accounts and National Accounts.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Chapter 2|
|Publication:||OECD Economic Surveys - Sweden|
|Date:||Dec 1, 2008|
|Previous Article:||Key challenges for the Swedish economy.|
|Next Article:||Taxation and growth: what direction should Sweden take?|