Journal Issue
Share
Article

Alternative Fiscal Rules for Norway

Author(s):
Daniel Leigh, and Etibar Jafarov
Published Date:
October 2007
Share
  • ShareShare
Show Summary Details

I. Introduction

Norway’s fiscal position is enviable. Its large oil and gas revenues, as well as the policy of saving these revenues and investing them abroad through the Government Pension Fund -Global (GPF), have allowed Norway to run large budget surpluses and amass large net government assets. In 2006, the budget surplus of the general government was estimated at 25.9 percent of its GDP, and, at the end of 2006, net assets of the general government equaled 150.2 percent of GDP2. Although the government’s net cash flows from petroleum operations are expected to decline gradually, Norway is expected to run large fiscal surpluses for many years to come.

However, Norway faces significant challenges in managing its oil wealth. Spending it, even on investment projects, would risk succumbing to the “Dutch disease,” in which the traded goods sector is damaged by a high real exchange rate. Since 2001, fiscal policy and the disposition of the oil wealth has been governed by fiscal guidelines, including a rule that central government non-oil structural deficit should be 4 percent of the assets of the GPF, the assets of which are invested abroad (Box 1).3 While this policy has so far been effective in limiting Dutch disease effects and insulating the budget from changes in petroleum prices and extraction rates (Jafarov and Moriyama, 2005), the rule implies an expansionary fiscal policy over the next 15 years, as the GPF grows much faster than GDP.

Moreover, in the longer term, Norway faces a significant fiscal challenge related to aging of its population. By 2050, Norway’s population is expected to be considerably older, with the old-age dependency ratio projected to increase from 23 percent in 2005 to 41 percent by 2050.4 Equivalently, the number of people of working age per person over the age of 65 is expected to decline from 4.4 in 2005 to 2.4 in 2050. According to projections in Norway’s 2007 budget, old-age pension spending in percent of GDP will rise by about 10 percentage points over 2005–2050, more than in almost any other advanced economy, reflecting a system that is both generous and maturing. In addition, aging could cause additional spending on health and long-term care of 3.2 percent of GDP (OECD, 2003). Increased participation in the welfare programs also threatens fiscal sustainability.

Box 1.The Government Pension Fund – Global and Fiscal Guidelines

To manage Norway’s oil wealth, the Norwegian authorities established the Government Petroleum Fund (since 2006, called the Government Pension Fund - Global; GPF) in 1990, and adopted fiscal guidelines in 2001 (effective for the 2002 budget). The GPF, which is formally a government account at Norges Bank, receives most of the petroleum revenue and invests it in financial assets abroad. Within the fiscal guidelines, the key rule sets the non-oil structural budget deficit of the central government to the long-run real return on the GPF, assumed to be 4 percent. The guidelines allow temporary deviations from the 4-percent rule over the business cycle and in the event of extraordinary changes in the value of the GPF. The GPF and fiscal guidelines were meant to serve a number of purposes: allow some petroleum revenue to be spent; insulate the budget from changes in petroleum income; preserve assets for use by future generations; and avoid the potential crowding out effects (so-called Dutch disease effects) that rapid spending of oil wealth might bring (Skancke, 2003).

No transfers to the GPF took place until 1995 because of low net oil income and large oil-related investments. Since then, however, assets of the GPF have increased rapidly, as both production and the price of oil picked up while the government’s oil-related investments declined. At end–2006, the market value of the GPF was estimated at Nkr1,784 billion or about 114.1 percent of GDP. The 2007 budget projects that the market value of the fund will reach about 170 percent of GDP in 2009.

The 4-percent rule has been breached every year since its inception, although the deviations from the rule have become smaller. The deviations from the rule in 2002–03, when the size of the GPF shrank because of sharp declines in stock markets and the economy experienced a downturn, could be justified under the fiscal guidelines. The 2007 budget projects the relevant deficit to be very close to the level implied by the 4-percent rule.

GPF Assets

(In percent of GDP)

Sources: Ministry of Finance, 2007 budget; and IMF staff estimates.

In light of these fiscal pressures, this paper assesses the fiscal rule in terms of its medium-term macroeconomic impact and the long-run sustainability of Norway’s public finances. Oil prices are now much higher than had been envisaged when the fiscal rule was adopted. If sustained, this implies a larger fiscal expansion in the next 15 years than anticipated. At the same time, oil wealth accumulated under the 4-percent rule is not likely to be sufficient to cover aging costs over the longer term. Accordingly, the paper analyzes several reform measures, including alternative fiscal rules, that could help resolve these issues, in part by using the IMF’s Global Integrated Monetary Fiscal Model (GIMF) to evaluate the macroeconomic effects of these measures.

Two principal conclusions emerge from the analysis. First, no rule examined here dominates the others. Rather, each involves trade-offs in terms of long-term fiscal sustainability, short-term expansionary impulses, intergenerational wealth transfers, and long-term output gains. Thus, while Norway’s oil wealth is unlikely to be large enough to cover the projected increase in old-age pensions of about 10 percent of GDP under any reasonable rule, alternative rules would require less fiscal consolidation than the 4-percent rule in the long term. Likewise, alternative rules could also yield a less expansionary fiscal stance than the 4-percent rule in the medium term. Second, analysis using GIMF suggests substantial long-run supply-side output gains associated with adopting a rule that stabilizes the GPF as a share of GDP and saves significantly more oil revenue for future generations. These output gains accrue principally because such a rule permits lower taxes in the long run, which stimulates labor supply.

The rest of the paper is organized as follows. Chapter II presents long-run projections for Norway’s oil revenues and compares Norway’s age-related spending pressures with those of other advanced industrial countries. Chapter III assesses fiscal sustainability under the existing 4-percent fiscal rule. Chapter IV assesses fiscal sustainability under three alternative fiscal rules, while undertaking a number of sensitivity tests. Chapter V assesses the macroeconomic consequences of adopting the alternative rules using GIMF. Finally, Chapter VI compares the four rules.

II. Declining Oil Revenue and Age-Related Spending Pressures

A. Oil production and revenue

Norway’s petroleum reserves and production are significant. The country started oil production in the North Sea in 1971, and is now the tenth largest producer of oil worldwide and the fifth largest exporter. Although oil production has recently started declining, rising gas production has offset this; currently, Norway is the third largest exporter of gas (Figure 1). The production of oil and gas (together) is expected to peak in 2008 and gradually decline thereafter, halving by 2030 (Figure 2).5

Figure 1.Norway: Production, Exports, and Reserves of Gas and Oil

Sources: BP Statistical Review of World Energy June 2007; International Energy Annual, 2005; and Energy Information Administration (USA).

Figure 2.Norway: Oil Production and the General Government’s Oil Revenues,1995-2030

Source: Ministry of Finance, 2007 budget.

Reflecting increasing oil production and high oil prices, the Norwegian government’s revenues from petroleum operations have surged.6 For example, in 2000-06, on average, oil revenue was about 18.1 percent of GDP. As a result, the general government budget surplus (including the return on the GPF) averaged to 17.2 percent, whereas the non-oil budget (excluding the return on the GPF) was in deficit of 2.8 percent of GDP. In the 2007 budget, revenues from petroleum activities are expected to be about 23.1 percent of GDP, and the general government budget surplus is projected at 24.7 percent.

However, Norway’s oil revenues are expected to decline over time (Figure 2). From 2007 to 2030, the Norwegian authorities project oil production to decline by more than 40 percent. In addition, they expect real oil prices to fall by more than 40 percent from 2007 to 2015 and remain unchanged after 2015. In this paper, we assume the declining trend in production to continue beyond 2030.

The 4-percent rule, together with high oil prices, implies a rapid increase in the non-oil primary deficit (NOPD) in the next ten years. With a growing GPF, the 4-percent rule always meant some structural expansion, but fiscal impulses were expected to be small when the rule was introduced. For example, the 2001 budget projections for 2008–10, which assumed oil prices of Nkr185, implied an increase in the non-oil budget deficit of the central government of ¼ percent each year. However, the 2007 budget, which assumes oil prices of Nkr323-357 implies that the deficit will increase by some ¾ percent of GDP each year in the same period (Table 1). If this expansion were implemented through higher spending, real spending of the central government could rise by more than 4 percent a year. For comparison, the 2007 budget projects a 2¾ percent increase in real spending.

The Fiscal Rule and Actual Non-oil Budget Budget Deficits of the Central Government

(Billions of Nkr at 2007 prices)

Sources: Ministry of Finance, 2001 and 2007 budgets; and IMF staff estimates

Table 1.Central Government Fiscal Position Under Different Oil Prices, 2002-10(In percent of GDP; unless otherwise specified)
Projections
Based on the 4-percent rule
200220032004200520062007200820092010
Based on the 2007 National Budget projections
Structural non-oil balance-3.2-3.6-3.9-3.6-4.0-4.4-5.2-6.0-6.6
4 percent of GPF assets-2.0-1.9-2.5-2.8-3.6-4.4-5.2-6.0-6.6
Non-oil expenditures46.445.144.543.542.143.143.844.845.7
Increase in real terms

(applying GDP deflator)
13.0-1.53.12.01.32.84.44.53.9
Government Pension Fund (GPF) assets

 (in percent of GDP)
49.466.474.696.1114.1135.6153.8170.6
Oil price assumption (in Norwegian krones)197.9204.4254.1343.5411.8390.0357.0334.6323.0
Higher oil price scenario (20 percent more than in the 2007 budget projections)
Structural non-oil balance-3.2-3.6-3.9-3.6-4.0-4.4-5.4-6.3-7.0
4 percent of GPF assets-2.0-1.9-2.5-2.8-3.6-4.4-5.4-6.3-7.0
Non-oil expenditures46.445.144.543.542.143.143.945.146.1
Increase in real terms

(applying GDP deflator)
13.0-1.53.12.01.32.84.94.84.2
Government Pension Fund (GPF) assets

 (in percent of GDP)
49.466.474.696.1114.1140.1162.3182.3
Oil price assumption (in Norwegian krones)197.9204.4254.1343.5411.8468.0427.0398.6383.0
Based on WEO projections
Structural non-oil balance (WEO)-3.2-3.6-3.9-3.6-4.0-4.4-5.0-5.8-6.5
4 percent of GPF assets-2.0-1.9-2.5-2.8-3.6-4.4-5.0-5.8-6.5
Non-oil expenditures46.445.144.543.542.143.143.644.645.5
Increase in real terms

(applying GDP deflator)
13.0-1.53.12.01.32.84.04.44.1
Government Pension Fund (GPF) assets

 (in percent of GDP)
49.466.474.696.1114.1133.6151.0168.7185.2
Oil price assumption (in Norwegian krones)197.9204.4254.1343.5411.8384.1416.2422.3427.8
Sources: Ministry of Finance; and IMF staff estimates.
Sources: Ministry of Finance; and IMF staff estimates.

B. Aging-related Spending

Norway faces a fiscal challenge related to aging of its population. Specifically, Norway’s old-age pensions are expected to increase sharply over the next several decades, reflecting rising longevity, the retirement of the baby boom generation, and easy access to early retirement (Box 2).

Increasing Pension Spending and Decreasing Oil Revenue

(Percent of GDP)

Sources: Ministry of Finance, 2007 budget; and IMF staff estimates.

Box 2.Decomposition of Changes in Old-age Pension Spending (Based on OECD, 2003)

From 2000 to 2050, age-related spending is expected to increase by 13.4 percentage points of GDP. Of this, 8 percentage points are due to old-age pensions, 3.2 percentage points are due to increases in health care and long-term care, and 1.6 percentage points are due to early retirement programs. Demographics account for about one-third of the 8 percent increase in pension spending. The rest is due mainly to the full phasing-in of benefits, in part related to increases in female labor participation.

Projections of Age-related Spending in OECD Countries, 2000-2050 1/
Total Age-related

Spending
Old-age Pensions“Early-retirement”

Programmes
Health Care and Long-

term Care
Child/Family Benefits

and Education
level 2000

(1)
change

2000-50

(2)
level 2000

(3)
change

2000-50

(4)
level 2000

(5)
change

2000-50

(6)
level 2000

(7)
change

2000-50

(8)
level 2000

(9)
change

2000-50

(10)
(Levels in percent of GDP, changes in percentage points)
Australia16.75.63.01.60.90.26.86.26.1-2.3
Austria 2/[10.4][2.3]9.52.2[5.1][3.1]
Belgium22.15.28.83.31.10.16.23.06.0-1.3
Canada17.98.75.15.86.34.26.4-1.3
Czech Republic23.16.97.86.81.8-0.77.52.06.0-1.2
Denmark 3/29.35.76.12.74.00.26.62.76.30.0
Finland19.48.58.14.83.1-0.18.13.8
France 4/[18.0][6.4]12.13.9[6.9][2.5]
Germany[17.5][8.1]11.85.0[5.7][3.1]
Hungary 5/7.11.66.01.21.20.3
Italy[19.7][1.9]14.2-0.3[5.5][2.1]
Japan13.73.07.90.65.82.4
Korea3.18.52.18.00.30.00.70.5
Netherlands 6/19.19.95.24.81.20.47.24.85.40.0
New Zealand18.78.44.85.76.74.07.2-1.3
Norway17.913.44.98.02.41.65.23.25.50.5
Poland 5/12.2-2.610.8-2.51.4-0.1
Spain[15.6][10.5]9.48.0[6.2][2.5]
Sweden29.03.29.21.61.9-0.48.13.29.8-1.2
United Kingdom15.60.24.3-0.75.61.75.7-0.9
United States11.25.54.41.80.20.32.64.43.9-1.0
Average of countries

above7/
21.25.87.43.41.60.25.93.16.2-0.9
Portugal 8/15.64.38.04.52.5-0.4

Data for health care shown in parentheses are drawn from EPC (2001). They are the results of an EC exercise using a common methodology for all countries. The projections are based on the same macroeconomic assumptions as in OECD(2001) Table 3.1. These health and long-term care projections assume that costs per capita rise in line with productivity/wages. They do not allow for technological change or other non-age-related factors.

Total pension spending for Austria includes other age-related spending which does not fall within the definitions in 3-10. This represents 0.9 percent of GDP in 2000 and rises by 0.1 percentage point in the period of 2050.

Total for Denmark includes other age-related spending not classifiable under the other headings. This represents 6.3 percent of GDP in 2000 and increases by 0.2 percentage point for 2000 to 2050.

For France, the latest available year is 2040.

Total includes old-age pensions spending and “early-retirement” programmes only.

“Early-retirement” programmes only include spending on persons 55+.

Sum of column averages. OECD average excludes countries where information is not available and Portugal where the data are less comparable than for other countries.

Portugal provided an estimate for total age-related spending but did not provided expenditure for all the spending components.

Source: Adapted from Table 2 on p. 35 in “Policies For An Ageing Society: Recent Measures And Areas For Further Reform,” OECD, Economics Department Working Papers No.369. Paris: OECD

Data for health care shown in parentheses are drawn from EPC (2001). They are the results of an EC exercise using a common methodology for all countries. The projections are based on the same macroeconomic assumptions as in OECD(2001) Table 3.1. These health and long-term care projections assume that costs per capita rise in line with productivity/wages. They do not allow for technological change or other non-age-related factors.

Total pension spending for Austria includes other age-related spending which does not fall within the definitions in 3-10. This represents 0.9 percent of GDP in 2000 and rises by 0.1 percentage point in the period of 2050.

Total for Denmark includes other age-related spending not classifiable under the other headings. This represents 6.3 percent of GDP in 2000 and increases by 0.2 percentage point for 2000 to 2050.

For France, the latest available year is 2040.

Total includes old-age pensions spending and “early-retirement” programmes only.

“Early-retirement” programmes only include spending on persons 55+.

Sum of column averages. OECD average excludes countries where information is not available and Portugal where the data are less comparable than for other countries.

Portugal provided an estimate for total age-related spending but did not provided expenditure for all the spending components.

Source: Adapted from Table 2 on p. 35 in “Policies For An Ageing Society: Recent Measures And Areas For Further Reform,” OECD, Economics Department Working Papers No.369. Paris: OECD

Statistics Norway’s latest population projections suggest that from 2005 to 2050 the life expectancy for men and women will grow by about 7.4 and 6 years, respectively. The number of persons over age 67 (the official age of retirement) will remain broadly stable until the end of this decade, but will then grow substantially. On the other hand, the fertility rate has fallen over the past few decades and is expected to remain at the current level of 1.8. As a result, the old-age dependency ratio, defined as the ratio of the number of aged persons (defined here as persons above age 64) to the number of working-age persons (20–64 year olds), is expected to increase from 24.7 percent in 2005 to 45.4 percent in 2060. The dependency ratio for very old people (above age 79) will more than double during 2010–50, with the sharpest increases taking place after 2025, when the baby boom generation retires.

Norway: Ratios of Old People to Working-age Population

Sources: Statistics Norway; and IMF staff estimates.

Norway’s demographic outlook is not worse than that of many other advanced industrial countries. For example, Norway’s old-age population growth is close to the G7 average based on projections published by the United Nations (2007). The dependency ratio in Norway is also expected to evolve in line with the G7 average (Figure 3).

Figure 3.Old-age Demographics in Norway and G-7 Countries

Source: United Nations (2007)

1/ The old-age dependency ratio is defined as the ratio of the population aged 65 or over to the population aged 15-64.

In addition to purely demographic factors, the continued maturation of the earnings-related pay-as-you-go pension system will be an important factor contributing to the sharp increases in pension spending. Most of this maturation will take place in the next two decades. The Norwegian social security system is currently immature because it was introduced only in 1967, and 40 years of service are required to receive a full pension. Thus, people born in 1940 and reaching retirement age in 2007 will be the first cohort to qualify for maximum benefits. Only beyond 2030, when most pensioners become qualified for maximum benefits, will the ratio of average pension benefits to the wage level reach its steady state (Fredriksen and Stolen, 2005). Female labor participation, growing since the 1970s, also contributes, since increasing participation of women has so far boosted fiscal revenues more than spending, which has resulted in declines in pension spending in percent of GDP. However, as these cohorts of women retire, pension spending will increase rapidly.

Regarding the generosity of the pension system in Norway, by European standards, replacement rates are not particularly high, and the statutory retirement age of 67 is high (Figure 4). However, easy access to early retirement, disability benefits, and sick leave, together with high tax rates on labor income, have taken their toll on public finances and labor supply. In particular, the effective retirement age, especially for men, has been declining while the number of disability cases and days lost owing to sickness have been increasing (Bellone and Bibbee, 2006).

Sources: OECD Tax Database, data as of year 2005; Ministry of Finance: The 2007 National Budget; Statistics Norway.

1/ The all-in (top marginal) tax rate, calculated as the additional central and sub-central government personal income tax, plus employee social security contribution, resulting from a unit increase in gross wage earnings. The all-in rate takes account of the same aspects as the combined rate, but does in addition include employee social security contributions and if they are deductible in central government taxes etc.

Figure 4.Generosity of the Pension System in OECD Countries

Note: 1/ According to the MOF, in 2005 the total effective age of retirement in Norway was 59.

2/ Official and effective age for men in the U.S. were the same.

Source: OECD, Society at a Glance: OECD Social Indicators, 2005 Edition and Pensions at a Glance: Public Policies Across OECD Countries, 2005 Edition.

In March 2007, the Norwegian parliament agreed to reforms that would reduce old-age pension spending. Key cost-cutting measures are (i) basing benefits on lifetime earnings, instead of the best 20 years, as now; (ii) adjusting benefits for life expectancy; (iii) indexing benefits to the simple average of wages and prices, rather than to wages, as now (except for a minimum pension, to be indexed to wages); and (iv) making individual benefits actuarially neutral, with the replacement rate depending on retirement age and a flexible retirement age as early as 62. These reforms could reduce future pension spending by about 3 percent of GDP. Furthermore, the government has been negotiating reforms that could reduce participation in the welfare programs with the social partners.

III. Fiscal Sustainability Under the 4-Percent Fiscal rule

This chapter assesses the long-run sustainability of Norway’s public finances under the 4-percent fiscal rule. The simulations assume that the non-oil primary deficit (NOPD) is set equal to 4 percent (the real return) of GPF assets in each year, petroleum revenue evolves according to the 2007 national budget (Figure 2), and real GDP grows by 2¼ percent a year, in line with the authorities’ projections of potential GDP growth. Sensitivity tests illustrate how alternative assumptions regarding oil revenue, the real interest rate, and GDP affect the results.

In percent of mainland GDP, the 4-percent rule under current projections implies a hump-shaped non-oil budget deficit and GFP assets. Specifically, GPF assets will peak at 240 percent of GDP in 2022 and decline thereafter. Similarly, the NOPD is projected to increase from 5.2 percent of GDP in 2006 to a peak of 9½ percent of GDP in 2023, and to decline thereafter (Figure 5). Note that the increase is sharper than the decline, reflecting accumulation of large oil revenues over the next 15 years.

Figure 5.Fiscal Position of the General Government Under the 4-percent Rule

(In percent of GDP)

Source: IMF staff estimates. Until 2030, the baseline scenario is based on the 2007 budget projections of oil prices and revenues. Thereafter, oil production is assumed to decline gradually. The upper/lower band corresponds to 20 percent higher/lower oil prices and 50 basis points higher/lower yield on government assets.

The broad picture of declining GPF as a share of GDP under the 4-percent rule is robust to alternative oil revenue, interest rate, and growth assumptions. For example, in the long term, 20 percent higher/lower oil revenues and 50 basis points higher/lower yield on government assets would change the NOPD path, but the broad picture of an initial rise followed by a gradual decline remains (Figure 5).

Given the projected increase in pension spending, sticking to the 4-percent rule implies sharp cuts in non-pension spending or sharp increases (in percent of GDP) in taxes in the long term. On current projections, in 2060, income from the GPF would cover only about 2 percentage points of the projected 10 percentage point of GDP increase in pension spending. Accordingly, the 4-percent rule would require about an 8 percentage point cut in non-pension spending or the same size increase in taxes in the same period. More fiscal tightening would be needed in the longer term as the income from the GPF in percent of GDP declines over time.

Alternatively, in the absence of fiscal tightening after 2022, the rule would not be met, and the fiscal position would become unsustainable. In particular, if the NOPD were to remain at its peak of 9.5 percent of GDP after 2022, instead of declining in line with the 4-percent rule, net government financial assets would be on an unstable path. As Figure 6 illustrates, under this scenario, net financial liabilities would eventually explode.

Figure 6.Fiscal Position of the General Government: No Fiscal Tightening After 2022 1/

(In percent of GDP)

Sources: Ministry of Finance; and IMF staff estimates.

1/Baseline scenario is based on the 2007 budget projections of oil prices and revenues. Non-oil budget revenues and non-pension spending are assumed to remain unchanged from the 2007 levels. The upper and lower bands correspond to 20 percent higher and 20 percent lower oil prices, respectively.

2/ Includes oil revenues and interest on net assets.

These conclusions are broadly consistent with those of other studies. The government’s 2007 budget projections suggest that under the 4-percent rule GPF assets are unlikely to become large enough to fully fund the expected rise in public spending associated with population aging. These projections include simulations suggesting that meeting the 4-percent rule would require significant financial tightening even under higher-than-assumed oil prices and higher labor-force participation rates (Figure 7). The IMF (2005), Heide and others (2006), and the OECD’s latest country survey draw similar conclusions.

Figure 7.Norway: Old-Age Pension Liabilities and Need for Financial Tightening

Sources: Ministry of Finance, 2007 budget; and IMF staff etsimates.

IV. Alternative Fiscal Rules

This chapter assesses fiscal implications of three alternative fiscal rules, each of which preserve the GPF as a proportion of GDP, compared with the 4-percent rule that preserves the real value of the GPF. The first alternative rule targets a constant permanently sustainable NOPD in percent of GDP, analogous to Friedman’s (1957) Permanent Income Hypothesis (PIH). The second alternative rule is similar to the 4-percent rule, except that it limits the NOPD to the growth-adjusted return on the GPF, rather than to the full 4 percent return. The third alternative rule targets a level for GPF assets of 250 percent of GDP, which is close to the projected peak under the 4-percent rule. The choice of these rules is motivated by the objective of avoiding either a sharp fiscal consolidation or an increase in public debt in the future when, as expected, age-related expenditures increase.7 All the simulations presented in this chapter rely on the same assumptions regarding government revenue, growth and interest rates as those in Chapter III.

The Permanent Income Rule

The PIH implies that the government does not spend out of current income, but out of permanent income or total wealth. In its simplest form, the government’s permanent income is the annuity value of its net wealth, defined as the sum of its net assets and the discounted present value of future expected petroleum revenues. Here, the problem is formulated in terms of GDP, through using the interest-rate growth differential, rather than the real return on the GPF. This approach, by construction, ensures that the government accumulates sufficient financial assets to sustain a constant fiscal deficit as a share of GDP once oil reserves are depleted.

Formally, the permanent income rule implies setting the NOPD according to:

where r denotes the real interest rate; γ real GDP growth; GPF the value of GPF assets; and oil government oil revenue in percent of GDP. The equation thus involves computing the present discounted value of all future oil revenue using the growth-adjusted interest rate as the discount factor.8 The IMF has recommended this approach to setting targets for the NOPD to oil-producing countries such as Gabon.9

The permanently sustainable NOPD (PSNOPD) is estimated at 6.0 percent of GDP (Figure 8). Compared to the 4-percent rule, the non-oil deficit is smaller in the coming several years (that is, the large run-up in the deficit is muted), in exchange for more assets being available in the long run. Accordingly, in the long term, income from the oil wealth under the permanent income rule covers more of the projected increase in pension spending. However, given the large size of the pension increase, fiscal sustainability under the permanent income rule would still require significant cuts (about 8 percent of GDP) in non-pension spending or increases in taxes. Under the rule, GPF assets are expected to increase to 350 percent of GDP by 2080, and to remain constant thereafter.

Figure 8.Fiscal Position of the General Government Under the 4-percent and Permanent Income Rules

(In percent of GDP)

Source: IMF staff estimates. Until 2030, the baseline scenario is based on the 2007 budget projections of oil prices and revenues. Thereafter, oil production is assumed to decline gradually. The upper/lower band corresponds to 20 percent higher/lower oil prices and 50 basis points higher/lower yield on government assets.

The result that the NOPD and GPF assets are higher under the permanent income rule in the long run than under the 4-percent rule is robust to alternative oil revenue and interest rate and growth projections. However, the levels of permanently sustainable deficit levels are quite sensitive to changes in these assumptions. For example, 20 percent higher/lower oil revenues and 50 basis points higher/lower yield on government assets would increase/reduce the NOPD-to-GDP ratio by more than 2 percentage points (Figure 8).

A significant practical drawback to the PIH rule is its forward-looking nature. Permanent income is not observable, but must be estimated using projections for petroleum revenue, interest rates, and economic growth into the far future. Thus, if the world oil price changes, then the effect on permanent income will have to be estimated, and this would depend on, among other things, the degree to which the price change is expected to be permanent. Fiscal policy would have to adjust to the corresponding PSNOPD. By contrast, the 4-percent rule is backward looking and avoids such problems.

A Growth-adjusted Rule

Another alternative fiscal rule would stabilize the GPF in terms of GDP, rather than in real terms. This rule can be thought of as a variant of the 4-percent rule, in that the “real return” that is spent is adjusted for economic growth. Under the assumptions of Chapter III, this growth-adjusted return is 4 percent less real GDP growth (2¼ percent), or 1¾ percent.

Under the growth-adjusted rule, the NOPD would be smaller than that under the 4-percent rule in the near term, but larger in the long term. The permanently sustainable NOPD under the growth-adjusted rule is calculated at 7.0 percent of GDP, with the NOPD gradually expanding toward that level over 70 years (Figure 9). Note that this deficit is somewhat larger than the 6.0 percent under the permanent income rule, because more of the oil wealth is saved under the growth-adjusted rule: GPF assets increase from about 115 percent of GDP to a constant 410 percent of GDP by 2080. This is, however, an artifact of the immediate adjustment path chosen to reach the steady state GPF-to-GDP ratio; a slower adjustment would result in more short-term spending and a lower steady-state NOPD (see below Section An Asset Targeting Rule).

Figure 9.Fiscal Position of the General Government Under the 4-percent and Growth-adjusted Income Rules

(In percent of GDP)

Source: IMF staff estimates. Until 2030, the baseline scenario is based on the 2007 budget projections of oil prices and revenues. Thereafter, oil production is assumed to decline gradually. The upper/lower band corresponds to 20 percent higher/lower oil prices and 50 basis points higher/lower yield on government assets.

An Asset-targeting Rule

A potential disadvantage of both the PIH and growth-adjusted rules is the very large size of the GPF, which may be difficult to justify. An alternative would be to target a “reasonable” long-term asset level. For purposes of illustration, a steady-state GPF of 250 percent of GDP—close to the peak projected under the 4-percent rule—is simulated. A number of deficit paths could achieve this outcome, but here it is assumed that the deficit follows the 4-percent rule until the asset target is achieved, then switches to the adjusted-growth rule. Under the asset-targeting rule, the long-term NOPD would be about 4.4 percent of GDP (Figure 10). This is less than under the 4-percent rule in 2022–2080, but greater thereafter.

Figure 10.Fiscal Position of the General Government Under the Asset-targeting Rule

(In percent of GDP)

Source: IMF staff estimates. Until 2030, the baseline scenario is based on the 2007 budget projections of oil prices and revenues. Thereafter, oil production is assumed to decline gradually. The upper/lower band corresponds to 20 percent higher/lower oil prices and 50 basis points higher/lower yield on government assets.

The asset-targeting rule, however, also incorporates a forward-looking component: the target itself. Consequently, changes to oil prices, for example, would have to be assessed to determine their effect on achieving the rule. Also, the NOPD (in the simulation) or the path of GPF assets could be quite sensitive to oil price shocks, implying a procyclical fiscal policy.

The path for the NOPD and GPF assets under the target rule can also be interpreted as a growth-adjusted rule, but with a different (slower) transition from the 4-percent rule. Under this interpretation, there is no forward-looking aspect, and no need to adjust fiscal policy in the wake of oil price changes. However, such changes would, as under the growth-adjusted rule, result in a different steady-state asset-to-GDP ratio, and a different adjustment to compensate for rising aging costs.

V. The Macroeconomic Consequences of the Growth-Adjusted Rule

This chapter focuses on macroeconomic consequences of the growth-adjusted rule of the previous chapter. The analysis uses GIMF, a general equilibrium model developed at the IMF to examine monetary and fiscal policy issues in a multi-country setting. The model includes the following features (which, among other things, renders it non-Ricardian): overlapping generations of consumers with finite horizons, distortionary taxation, and liquidity constrained consumers who do not have access to financial markets and thus have to vary their consumption one-for-one with after-tax labor income. As such, the model is well equipped to analyze fiscal policy issues that involve permanent changes in government assets or debt. The model includes a large menu of fiscal policy tools, including labor income taxes, VAT, corporate income taxes, government consumption, and productive infrastructure expenditures. The model also includes a number of nominal and real rigidities, and a central bank that manipulates interest rates to achieve an inflation target of 2.5 percent a year.10 For the purposes of this paper, the underlying parameters of the model are calibrated to fit the key features of Norway’s economy. For example, the real return on government assets is calibrated at 4 percent a year. The model also contains a second region, the rest of the world, where long-run productivity growth is 2.25 percent a year. Kumhof and Laxton (2007) provide a detailed presentation of the model, and apply it to study the effects of fiscal deficits in the United States.

The objective of the analysis is to compare the evolution of key macroeconomic variables under the growth-adjusted rule with economic performance under the existing 4-percent rule. As discussed in Chapter IV, the 4-percent fiscal rule involves a fiscal expansion (as measured by the NOPD) in the near term and a fiscal contraction in the long run, with GPF assets peaking in 2021 before gradually declining as a share of GDP. In contrast, under the growth-adjusted rule, the NOPD is smaller in the near term, but a larger accumulation of GPF assets occurs. Therefore, in comparison with the 4-percent rule scenario, the growth-adjusted rule scenario involves a smaller NOPD in the near term, but a larger, permanently sustainable NOPD in the long run. The simulations focus on the macroeconomic consequences of these alternative paths for GPF assets and the NOPD.

While there are a large number of possible ways to design the composition of fiscal adjustment in the near term, and the expansion in the long run under the growth-adjusted rule, the discussion focuses on the following scenario:

  • Under the 4-percent rule, during the first 60 years, the NOPD is larger than under the growth-adjusted rule, and this expansion is implemented by an increase in government spending (Figure 11). In the long run, the NOPD is declining relative to GDP, and fiscal sustainability is achieved by increasing labor income taxes (Figure 11).

  • The growth-adjusted rule reverses these responses: compared to the baseline of the 4-percent rule, government spending is lower in the short term, and taxes are lower in the long run.

Figure 11.Government’s Fiscal Position Under the 4-percent and Growth-adjusted Rules

Source: GIMF simulations.

The following main results emerge from the analysis:

  • In the near term, compared with the 4-percent rule scenario, the tighter fiscal position under the growth-adjusted rule requires lower government spending, implying lower aggregate demand, and lower inflation (Figures 12 and 13).

  • However, the near-term effects of the fiscal consolidation on aggregated demand are in part offset by two factors. First, the central bank responds to the decline in inflation by reducing real interest rates. This monetary expansion stimulates consumption, investment, and labor effort (Figure 12), resulting in a lower value of the krone (compared to the 4-percent rule), which stimulates net exports. Second, the reduction in government consumption is associated with a non-Keynesian increase in private consumption. In particular, households anticipate that the persistent decline in government consumption will enable reductions in labor income taxes in the future, and respond to this increase in their permanent disposable income by increasing consumption today.11

  • In the long term, supply-side output gains are associated with the growth-adjusted rule, because the larger stock of GPF assets permits lower labor income taxes (compared to the 4-percent rule), raising labor supply, and enabling higher levels of private consumption.

Figure 12.Adopting the Growth-adjusted Rule: Economic Activity

(Deviation from the baseline 4-percent rule scenario)

Source: GIMF simulations.

Note: Baseline scenario corresponds to the current 4-percent fiscal rule.

Figure 13.Adopting the Growth-adjusted Rule: Inflation and Monetary Policy

(Deviation from baseline 4-percent rule scenario)

Source: GIMF simulations.

Note: Baseline scenario corresponds to the current 4-percent fiscal rule.

One gauge of whether the long-run gains of adopting the growth-adjusted rule warrant the near-term aggregate demand moderation is the discounted value of real consumption. As mentioned, while a formal welfare analysis of the growth-adjusted rule, which would account for changes in both household consumption and leisure, is beyond the scope of this paper, it is possible to evaluate the present discounted value of the additional consumption from the model simulation described above. As Table 2 suggests, the present discounted value of adopting the growth-adjusted rule, in terms of the additional private consumption obtained, is positive for real discount rates up to 4.3 percent a year. At a real discount rate of 4 percent, there is little difference between the 4-percent rule and the growth-adjusted rule.

Table 2:Present Discounted Value of Future Consumption(Deviation from Baseline)
Discount rate

(percent per year)
Value
11647
2389
3108
416
5-19
Note: baseline corresponds to current 4-percent fiscal rule scenario.Value reaches zero at discount rate of 4.3 percent.
Note: baseline corresponds to current 4-percent fiscal rule scenario.Value reaches zero at discount rate of 4.3 percent.

VI. Comparison of the Four rules

The 4-percent rule and three alternative rules considered here involve trade-offs in terms of covering the expected increase in pension spending, long-term fiscal sustainability, short-term expansionary impulses, intergenerational wealth transfers, and long-term output gains. As can be seen from Figure 14, in terms of covering the expected pension increase in the long term, the growth-adjusted rule that saves more of the oil wealth than other rules comes the closest to covering the pension increase. Moreover, this rule provides the least stimulus to the economy in the short-term term and performs the best in terms of long-term growth (Table 3).12 However, this rule also involves a much larger transfer of oil wealth to future generations. Performance under the PIH rule is close to that under the growth-adjusted rule, but fiscal policy under this rule is very sensitive to assumptions of long-term oil prices, growth rate, and interest rate. The asset-targeting rule avoids the much larger transfers, but it performs worse than the growth-adjusted and PIH rules in terms of long-term growth and covering the pension increase in the long term. Moreover, this rule could be procyclical. The 4-percent rule is procyclical in the short-term and performs the worst in terms of long-term growth as well as covering the pension increase in the long term.

Figure 14:Will the Non-oil Primary Deficit Cover the Pensions Spending Increase?

Table 3.Fiscal Rules: Summary
4% rulePIH ruleGA ruleAsset rule
Short-run level of NOPD/GDP4-9%6%2-6%4-9%
Long-run NOPD/GDP0%6%7%4%
Long-run GPF/GDP0%350%410%250%
NOPD/pension spending increase in 21508%55%64%40%
Sensitive to long-run forecastsNoYesNoNo
Largest simulated long-run output gain 1/4th2nd1st3rd

Output gain based assumption that increases in NOPD rely on increases in government consumption, and reductions in NOPD rely on increases in payroll taxes.

Output gain based assumption that increases in NOPD rely on increases in government consumption, and reductions in NOPD rely on increases in payroll taxes.

In addition, the choice of rule often depends on political considerations. In particular, accumulating 250-400 percent of GDP in financial assets could be very difficult to justify politically. Moreover, adopting a rule that appears to excessively favor future generations over current generations could meet political resistance. Finally, any attempt to change the current 4-percent rule that is well understood and accepted by most political parties could open a “Pandora’s box,” with the risk of no consensus on a new rule emerging.

References

    BarnettSteven and RolandoOssowski2003Operational Aspects of Fiscal Policy in Oil-Producing Countries” in Fiscal Policy Formulation and Implementation in Oil-Producing Countriesed. by JeffreyDavisRolandoOssowski and AnnaliseFedelino (Washington: International Monetary Fund) pp. 45-81.

    BelloneBenoit and AlexandraBibbee2006The Ageing Challenge in Norway: Ensuring a Sustainable Pension and Welfare SystemOECD Economics Department Working PaperNo. 480.

    CarcilloStephaneDanielLeigh and MauricioVillafuerte2007Natural-Resource Depletion, Habit Formation, Income Convergence, and Sustainable Fiscal Policy: the Case of CongoforthcomingIMF Working Paper.

    FredriksenDennis and Nils MartinStølen2005Effects of Demographic Development, Labour Supply and Pension Reforms On the Future Pension BurdenStatistics Norway Discussion Papers No. 418April 2005.

    HeideKim MasseyErlingHolmøyIngeborg FoldøySolli and BirgerStrøm2006A Welfare State Funded by Nature and OPECStatistics Norway Discussion Papers No. 464July 2006.

    International Monetary Fund2005IMF Staff Country Report No. 05/196 of May13 2005 (Washington: International Monetary Fund).

    JafarovEtibar and KenjiMoriyama2005The Norwegian Government Petroleum Fund and the Dutch Disease” in IMF Staff Country Report No. 05/197 (Norway—Selected Issues) of May13 2005 Chapter III (Washington: International Monetary Fund).

    KumhofMichael and DouglasLaxton2006A Party Without a Hangover? On the Effects of U.S. Government Deficits” (unpublishedInternational Monetary Fund).

    LeighDaniel and Jan-PeterOlters2006Natural Resource Depletion, Habit Formation, and Sustainable Fiscal Policy: Lessons from GabonIMF Working Paper No. 06/193 (Washington: International Monetary Fund).

    OECD2003Policies For An Ageing Society: Recent Measures And Areas For Further ReformEconomics Department Working Papers No.369 (ECO/WKP(2003)23).

    OECD2006Economic Survey of Norway 2007.

    United Nations2007Population Projectionshttp://esa.un.org/unpp/index.asp?panel=2

    TersmanGunnar1991Oil. National Wealth, and Current and Future Consumption PossibilitiesIMF Working Paper No. 91/60 (Washington:International Monetary Fund).

European Department and Fiscal Affairs Department, respectively. The authors would like to thank Steven Barnett, Dennis Botman, Mark De Broeck, Robert Ford, Manmohan Kumar, Rolando Ossowski, and participants in a seminar at Norges Bank for helpful comments.

Unless otherwise specified, GDP in this paper refers to mainland GDP, which is all domestic production except from exploration of crude oil and natural gas, services activities incidental to oil and gas, and transport via pipelines; and ocean transport.

Norway has been one of the first oil-producing countries measuring its fiscal policy stance based on non-oil budget balances. See Barnett and Ossowski (2003) on why this approach is more appropriate for countries with exhaustible resources.

The old-age dependency ratio is defined as the ratio of the population aged 65 or over to the population aged 15-64. Population projections come from the United Nations (2007).

Hereafter, oil and gas revenues/production will be called oil revenues/production.

The state receives revenues from oil enterprises through taxes (ordinary corporate income tax at 28 percent; special tax rate for oil producers at 50 percent of income; and the green gas emission (CO2) tax), royalties, fees, its direct financial interest in the petroleum sector (SDFI), and dividends from state shares of Statoil and Norsk Hydro (see IMF 2001).

The alternative rules are not necessarily meant to be welfare optimizing. This paper does not analyze inter-generational equity impact of these alternative rules. Heide and others (2006) argue that higher pre-funding of future spending favors future generations, who would be better off even without such redistribution because of economic growth.

For the derivation of Equation (1), and its application to a number of oil producing countries, see, for example, Barnett and Ossowski (2003), Leigh and Olters (2006), and Carcillo, Leigh, and Villafuerte (2007). Tersman (1991) applies a similar framework for Norway.

See IMF (2006) for recommendations to Gabon made in the context of the 2006 Article IV consultations.

In particular, monetary policy follows a forward-looking reaction function that targets the one-year ahead forecast of domestic inflation, and contains an interest rate inertia component in line with the monetary policy literature.

Note, however, that households without access to financial markets do not increase consumption in response to the future expected reduction in the tax burden.

As discussed in the previous chapter, output gains under the growth-adjusted rule accrue mainly because this rule involves the least tax burden, which stimulates labor supply. However, this does not necessarily mean that the growth-adjusted rule is welfare optimizing.

Other Resources Citing This Publication