## LONG-TERM COMPETITIVENESS IN NORWAY

## Source details

**Canonical URL:** [LONG-TERM COMPETITIVENESS IN NORWAY](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2013/_cr13273.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2013/_cr13273.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2013/_cr13273.pdf.json)

---

### A. Introduction and structural context
- Concern: high wage costs in Norway driven by robust oil and gas activities over the past decade.
- Measured in a common currency, unit labor costs in Norway are about 40 percent higher than its peers on average.
- Oil sector has increased mainland economy dependence on oil and gas; manufacturing production is supported by engineering industry closely linked to oil and gas.
- Fiscal framework:
  - Fiscal rule established in 2001.
  - Government Pension Fund Global (GPFG) receives all government’s oil and gas-related income.
  - Fiscal rule constrains the non-oil deficit as a share of GPFG assets, not relative to mainland GDP.
  - GPFG assets are growing faster than Norway’s GDP, resulting in a gradual increase in the non-oil deficit as a share of mainland GDP in recent years.
- Statistics Norway estimate: roughly 8 percent of employment in Norway comes from both direct and indirect demand of the petroleum sector.

### B. Trade structure, external position, and tradable sector dynamics
- Export composition (1990–2011):
  - Oil and gas: just below 70 percent of total goods exports in 2011, up from about 50 percent in 1990.
  - Fish exports: stable at around 6 percent of total exports.
  - Manufacturing exports: declined by half to 19 percent in 2011.
- Terms of trade:
  - Terms of trade (including oil and non-oil) improved substantially since 2000 due largely to high oil prices.
  - Mainland terms of trade (excluding oil and gas export prices) have also improved in the past decade, but by much less.
- Tradable sector:
  - The tradable sector is large largely because of the hydrocarbon sector.
  - The non-oil tradable sector is much smaller than peers and has declined over time; it is the smallest among OECD peers.
- Export destinations and balances:
  - Exports largely directed to advanced economies; share to the euro area increased substantially over past decades.
  - Share to emerging and developing economies has been relatively low.
  - Overall merchandise trade surplus increased substantially since early 2000s, driven by offshore trade surplus reflecting high oil prices.
  - Trade balance of traditional goods has been negative but relatively stable at around -7 percent of mainland GDP.

### C. Cost competitiveness, ULCs, and productivity
- REER and price measures:
  - CPI-based REER: fairly stable.
  - ULC-based REER: appreciated substantially over the past decade; Norway recorded one of the highest REER-ULC appreciations among peers and is now significantly above its long-run historical average.
  - Relative price of non-tradables to non-oil tradables increased substantially; Norway had the highest increase among peers.
- Drivers of ULC appreciation:
  - High wage growth and low labor productivity growth: productivity growth much lower in the 2000s than in the 1990s; real wages outpaced productivity while hours worked declined.
  - Unit labor costs grew much more in the 2000s than in the 1990s.
  - Service sector: ULC increased substantially in the 2000s, more than in manufacturing, largely due to lower productivity gains in services.
- Cross-country and sectoral notes:
  - Norway’s ULC growth is significantly higher than peers; productivity growth has been among the lowest.
  - Actual hours worked per employee are much lower than other advanced economies.
  - Removing the oil sector and oil-related productive part of the mainland economy would show even higher ULC growth and even lower labor productivity in traditional manufacturing.

### D. Market share trends and two-tier mainland development
- Aggregate market share:
  - Norway maintained a steady world market share of total exports of goods and services for past few decades, performing better than many peers whose shares declined in the 2000s.
- Non-oil exports and manufacturing:
  - Market share of non-oil exports declined substantially over the past two decades and much faster than other advanced economies.
  - Norway has lost almost 60 percent of its manufacturing share of world exports since 1990.
  - Manufacturing market share stabilized in early 2000s but declined again since the global financial crisis.
- Patterns of specialization:
  - Norway lost market share in manufacturing sectors that are growing in the world market.
  - Two-tier development:
    - Oil-related industries (mechanical engineering tied to oil and gas) are expanding and gaining market share.
    - Traditional manufacturing sectors (e.g., base metals) are shrinking and losing market share.
  - Manufacturing production has been largely driven by oil-related engineering since mid-2000s; other traditional subsectors subdued.
- Wage-share patterns:
  - “Traditional” internationally-exposed industries have sharply rising wage shares, reflecting steep increases in unit labor costs.
  - Internationally-exposed industries related to the booming oil industry have been able to absorb rising unit labor costs; the wage share in these industries has been falling over the last decade.

### E. Key implications for competitiveness and resilience
- Combined factors producing competitiveness losses:
  - High wage growth, low productivity growth, declining hours worked, and growing GPFG relative to GDP have produced substantial appreciation in the ULC-based REER and a loss of competitiveness in the non-oil mainland economy.
- Increasing dependence and Dutch disease risks:
  - Mainland economy increasingly dependent on supplying goods and services to the offshore sector; oil-sector linkages have supported some manufacturing but have propelled wage growth that undermines competitiveness in traditional non-oil sectors.
  - Insulation from Dutch disease appears incomplete despite the fiscal rule and sound oil revenue management.
  - Weakened cost competitiveness is evidenced by appreciation of ULC-based REER and a striking loss of market share for non-oil exports over past decades.
- Resilience concerns:
  - Oil-related parts fuel robust growth but represent a narrow basis for long-run growth.
  - Continued dependence implies potential insufficient capacity to supply employment and income during sustained oil price declines or after extraction peaks.

### F. Household balance sheets, vulnerabilities, and distributional analysis
- Headline household vulnerabilities:
  - Household debt: about 200 percent of disposable income in 2011, among the highest in the OECD.
  - Real house prices increased by nearly 30 percent from their lowest level in 2008; real house prices rose continuously since 2009.
  - Price-to-income and price-to-rent ratios increased by 14 percent and 23 percent, respectively, from their lowest levels in 2008.
- House price valuation estimates (2013Q1 staff estimates):
  - Three measures used: (i) IMF VEA econometric model; (ii) deviation of price-to-rent ratio from historical average; (iii) deviation of price-to-income ratio from historical average.
  - Average estimate across the three measures in 2013Q1 is a little over 40 percent.
  - Excluding the price-to-rent ratio reduces the average estimate to about 30 percent.
  - 2011 Article IV concluded house prices might be overvalued by 15–20 percent (IMF VEA model).
- Factors contributing to house price increases:
  - Demand: high wage/income growth tied to high oil prices; population growth about 1 percent on average for the past decade; average annual immigration growth around 8 percent; low interest rates since 2000.
  - Supply: constraints, especially in the Oslo area, due to land-use regulations and minimum unit size.
  - Institutional: preferential tax treatment of owner-occupied properties (full deductibility of interest payments and lower effective tax rates); lower risk weights of residential mortgages relative to corporate loans.
- Aggregate buffers and composition (2011):
  - Household net worth a little over 200 percent of disposable income in 2011.
  - Adjusting tax-valued non-financial assets upward (tax valuations believed to understate market values by about one third) raises net worth to 340 percent of disposable income, but this would still be the second lowest among peers.
  - Household asset composition:
    - Non-financial assets: about 49 percent of total household assets.
    - Pensions and insurance: 20 percent.
    - Currency and deposits: 17 percent.
    - Shares and other equities: 20 percent.
    - Securities other than shares: 7 percent.
    - Other financial assets: 7 percent.
- Liquidity and likely behavior:
  - Share of liquid household assets is relatively small; if non-financial assets are revalued, illiquid assets rise to nearly 80 percent of household financial assets.
  - Liquid assets are smaller than the sum of mostly-floating-rate mortgages and other household financial liabilities, with the gap growing steadily over time.
  - A house price correction is more likely to induce households to reduce consumption rather than default, given full recourse mortgage lending and historically low mortgage losses.
- Distributional vulnerabilities (microdata 1987–2010; focus 2005 onward):
  - Household debts largely aligned with income levels; debt-to-income ratios do not differ considerably across income groups, particularly income deciles 4 and above.
  - Household debt rose since 2005 in all income groups.
  - Net worth in higher income groups (deciles 8 and 9) is near-zero; these groups have the highest debt-to-income ratios.
  - Non-financial assets (housing) are almost evenly distributed across income groups.
  - Financial assets are generally lower than total liabilities, with the gap negative for most income groups.
  - Share of households with debt more than 500 percent of disposable income rose from 4 percent in mid-1990s to 11 percent in 2010.
- Loan characteristics and policy measures:
  - New loans exceeding the FSA-recommended 85 percent LTV cap fell to 17 percent in 2012.
  - Share of interest-only loans among new loans declined from 23 percent in 2011 to 17 percent in 2012.
  - Saving rates are rising recently, suggesting increased household caution.
- Financial-stability conclusion:
  - Elevated household debt poses a significant challenge: a house price correction could significantly reduce household consumption through wealth effects.
  - Banks’ mortgage losses may not rise much, but lower consumption could impair business activity and increase bank losses on enterprise loans.
  - Ongoing strengthening of capital requirements for banks will help mitigate some risks; rigorous monitoring on residential mortgage lending is needed.
  - The FSA tightened the LTV cap from 90 percent to 85 percent in December 2011.

### G. Long-run fiscal framework, GPFG characteristics, and scenario analysis
- Resource and GPFG metrics:
  - Oil production accounts for over one fifth of total GDP, one quarter each of government revenues and total investment, and over fifty percent of total exports.
  - Norway was the world’s third largest exporter of gas and the seventh largest exporter of oil in 2011.
  - Total assets of the GPFG were $713 billion at end-2012.
  - Current target portfolio: 60 percent equity, 35 percent fixed-income, and 5 percent real estate.
  - Management costs: 7-10 basis points annually.
- Fiscal rule design:
  - 2001 guidelines: “Petroleum income should be phased into the economy on par with the developments in expected returns on the Government Pension Fund Global. Considerable emphasis must be put on stabilizing the economy.”
  - Expected real return estimated at 4 percent for establishing trend fiscal transfer.
  - Rule permits cyclical deviations but constrains structural non-oil deficit as share of GPFG assets, not mainland GDP.
  - Because GPFG assets grew faster than the economy, the rule allowed a steadily rising non-oil deficit as share of mainland GDP; when GPFG assets start to decline relative to mainland GDP (expected in about two decades), transfers would decline as a share of mainland GDP.
- Baseline projection assumptions:
  - Authorities’ projections for oil production volumes.
  - A 4 percent real rate of return.
  - Average fiscal transfer equal to 4 percent (trend real rate of return).
  - Oil price projections based on WEO and futures, constant in real terms after 2022.
  - Population growth declines from 1.3 percent to 0.2 percent by 2100.
  - Real GDP growth: staff projections through 2015 and authorities’ projections thereafter (3.3 percent in 2015 declining to 1.8 percent in 2060-2100).
- Baseline projection outcomes:
  - GPFG peaks at about 265 percent of mainland GDP by 2030 and declines thereafter.
  - Non-oil deficit rises sharply as a share of mainland GDP over the next decade and a half, peaking at just below 11 percent of mainland GDP by 2030.
  - Fiscal transfers decline thereafter to just below 4 percent of mainland GDP by 2100.
- Sensitivity and uncertainty:
  - Small changes in key assumptions can have very large effects due to compounding.
  - Shorter time horizons matter for managing crowding out pressure on non-oil tradable sectors in the next two decades.

### H. GPFG returns, alternative scenarios, and fiscal-transfer implications
- Historical returns and comparisons:
  - Lifetime (1997–2012) real rate of return on the GPFG: 3.2 percent.
  - Difference between assumed (4 percent) and historical: 0.8 percent.
  - Long-run global benchmark (60% equities / 40% bonds) real return for 1900–2012: 3.7 percent.
  - World average real return on equity 1900–2012: 5 percent; global equity-bond outperformance since 1900: 1.8 percent.
  - European rates 1900–2012: equities 4.2 percent and bonds 0.8 percent.
- Risk that future returns may be lower:
  - Possibility that GPFG may not maintain historical average returns if asset-price cycles align with oil-price peaks or if longer-run rates fall below recent decades.
  - Some analysts argue today’s low interest rates are the best estimate for expected real return over the next two decades.
- Effects of varying real rates on fiscal transfers:
  - Baseline real rate: 4 percent.
  - Historical-average real rate: 3.2 percent.
  - Lower-rate scenario used: 2.5 percent.
  - Near-term (until about 2030): differences in fiscal transfer across these rates are less than one or two percentage points of mainland GDP between highest and lowest projections.
  - Long run (end of century): lower real rates reduce fiscal transfer relative to baseline by 1.7 and 2.5 percentage points of mainland GDP (for intermediate and low return scenarios).
- Effect of varying transfer-rate (spending rule) examples:
  - 3.0 percent of GPFG: lower fiscal transfer as share of mainland GDP than baseline until 2054, but higher thereafter.
  - 2.0 percent of GPFG: lower fiscal transfer in percent of mainland GDP in the coming years which increases to about 8 percent of mainland GDP in 2050-2060 and remains roughly in that range through 2100.
- Impact of a permanent 40 percent decline in real oil prices:
  - Affects short- and medium-term fiscal transfers by somewhat (2 percentage points) but less so the long-run fiscal transfer.
  - Proportional impact on GPFG assets is slightly more than half of the percentage change in oil prices because 44 percent of total oil resources had been extracted as of 2012.
  - Impact on GPFG assets and fiscal transfer largely phased in by mid-century (nearly 90 percent of remaining oil extracted by then).
- Combined scenario definitions and selected outcomes (Table 3.1 highlights):
  - Baseline: Rate of return 4.0; Rate of fiscal transfer (%MGDP) 4.0; Fiscal transfer 2020 (%MGDP) 9.7; Fiscal transfer 2060 (%MGDP) 8.1; Fiscal transfer 2100 (%MGDP) 3.8; GPFG peak year 2029; GPFG assets 2020 (%MGDP) 243; GPFG assets 2060 (%MGDP) 203; GPFG assets 2100 (%MGDP) 95.
  - Constant 4% transfer: Rate of return 3.2; Rate of fiscal transfer 4.0; Fiscal transfer 2020 9.3; Fiscal transfer 2060 6.2; Fiscal transfer 2100 2.2; GPFG peak year 2027; GPFG assets 2020 233; GPFG assets 2060 155; GPFG assets 2100 54.
  - Low return: Rate of return 2.5; Rate of fiscal transfer 4.0; Fiscal transfer 2020 9.0; Fiscal transfer 2060 5.0; Fiscal transfer 2100 1.3; GPFG peak year 2025; GPFG assets 2020 224; GPFG assets 2060 124; GPFG assets 2100 33.
  - Conservative: Rate of return 4.0; Rate of fiscal transfer 3.0; Fiscal transfer 2020 7.7; Fiscal transfer 2060 8.6; Fiscal transfer 2100 6.0; GPFG peak year 2043; GPFG assets 2020 255; GPFG assets 2060 287; GPFG assets 2100 200.
  - Prudent: Rate of return 3.2; Rate of fiscal transfer 3.0; Fiscal transfer 2020 7.3; Fiscal transfer 2060 6.5; Fiscal transfer 2100 3.4; GPFG peak year 2030; GPFG assets 2020 245; GPFG assets 2060 218; GPFG assets 2100 112.
  - Low transfer: Rate of return 3.2; Rate of fiscal transfer 2.0; Fiscal transfer 2020 5.1; Fiscal transfer 2060 6.2; Fiscal transfer 2100 4.7; GPFG peak year 2046; GPFG assets 2020 257; GPFG assets 2060 310; GPFG assets 2100 235.
  - Low oil price (40% lower), high return: Rate of return 4.0; Rate of fiscal transfer 4.0; Fiscal transfer 2020 8.3; Fiscal transfer 2060 6.1; Fiscal transfer 2100 2.9; GPFG peak year 2027; GPFG assets 2020 208; GPFG assets 2060 153; GPFG assets 2100 72.
  - Low oil price (40% lower), historical return: Rate of return 3.2; Rate of fiscal transfer 4.0; Fiscal transfer 2020 7.9; Fiscal transfer 2060 4.6; Fiscal transfer 2100 1.6; GPFG peak year 2013; GPFG assets 2020 199; GPFG assets 2060 115; GPFG assets 2100 40.

### I. Aging, pensions, and interactions with fiscal transfers
- Demographics and pensions:
  - Old-age ratio (65+/20–64) projected to nearly double from 2013 to 2060.
  - Statistics Norway estimate: pension reform would gradually raise participation rate by 8 percent by 2060.
  - Disability pensions in Norway are 1.6 percentage points of GDP higher than the average of higher-income OECD comparators in 2013.
- Policy interaction:
  - Even with pension reforms and reductions in disability pensions, pension expenditure would remain higher than fiscal transfers from GPFG unless additional reforms are implemented (e.g., 8 percent increase in participation rate and reducing disability pensions to OECD average).

### J. Conclusions and policy recommendations (key points)
- The percentage of GPFG assets transferred to the budget each year is the key parameter affecting the size of the fiscal transfer over the next two decades and central to addressing competitiveness and Dutch disease concerns.
- Over longer horizons, the realized real rate of return and oil prices become increasingly important for intergenerational equity and aligning fiscal transfers with aging-related costs.
- Reasons to consider a fiscal transfer rate below 4 percent of GPFG assets:
  - The real rate of return might align with history (around 3.2 percent) rather than 4 percent; if long-run real rate of return is below 4 percent, this will only be evident after a long lag.
  - Stabilizing the economy to prevent Dutch disease may require a lower rate of spending to lessen crowding out of “exposed industries.”
  - A slower rate of spending could better align expected fiscal transfers with expected long-run aging-related costs that peak after oil production declines.
- Continued emphasis on rigorous monitoring of residential mortgage lending, tightening capital requirements for banks, and considering structural reforms to boost labor productivity and non-oil tradable sectors are implied by the analysis.

*Italic: IMF staff compilation from document content.*

### REFERENCES ____________________________________________________________________________________33

### LONG-TERM COMPETITIVENESS IN NORWAY

### A. Introduction
- Concern: high wage costs in Norway driven by robust oil and gas activities over the past decade.
- Measured in a common currency, unit labor costs in Norway are about 40 percent higher than its peers on average.
- Oil sector success has created competitiveness challenges by increasing mainland economy dependence on oil and gas: manufacturing production supported by engineering industry closely linked to oil and gas.
- Fiscal framework:
  - Fiscal rule established in 2001.
  - Government Pension Fund Global (GPFG) receives all government’s oil and gas-related income.
  - Fiscal rule constrains the non-oil deficit as a share of GPFG assets, not relative to mainland GDP.
  - GPFG assets are growing faster than Norway’s GDP, resulting in a gradual increase in the non-oil deficit as a share of mainland GDP in recent years.
- Statistics Norway’s estimate: roughly 8 percent of employment in Norway comes from both direct and indirect demand of the petroleum sector.

### B. Trade Structure — composition and external position
- Export composition (1990–2011):
  - Oil and gas: just below 70 percent of total goods exports in 2011, up from about 50 percent in 1990.
  - Fish exports: stable at around 6 percent of total exports.
  - Manufacturing exports: declined by half to 19 percent in 2011.
- Terms of trade:
  - Terms of trade (including oil and non-oil) improved substantially since 2000 due largely to high oil prices.
  - Mainland terms of trade (excluding oil and gas export prices) have also improved in the past decade, but by much less.
- Tradable sector:
  - The tradable sector is large in Norway largely because of the hydrocarbon sector.
  - The non-oil tradable sector in Norway is much smaller than peers and has declined over time; it is the smallest among OECD peers.
- Export destinations:
  - Norway’s exports are largely directed to advanced economies; the share of exports to the euro area increased substantially over past decades.
  - Share of exports to emerging and developing economies has been relatively low.
  - Emerging markets are increasingly important because their demand affects oil prices and thus Norway’s overall export performance.
- Trade balances:
  - Overall merchandise trade surplus increased substantially since early 2000s, driven by offshore trade surplus reflecting high oil prices.
  - The trade balance of traditional goods has been negative but relatively stable at around -7 percent of mainland GDP.

### C. Cost Competitiveness Measures and Labor Productivity
- REER measures:
  - CPI-based REER: fairly stable (index references: Figure 1.8, Figure 1.9).
  - ULC-based REER: appreciated substantially over the past decade; Norway recorded one of the highest REER-ULC appreciations among peers and is now significantly above its long-run historical average (Figure 1.11).
- Relative prices:
  - Relative price of non-tradables to non-oil tradables increased substantially; Norway had the highest increase among peers (Figure 1.12).
  - Inclusion of the oil sector changes the picture because of rising oil prices since early 2000s.
- Drivers of ULC appreciation:
  - High wage growth and low labor productivity growth: productivity growth much lower in the 2000s than in the 1990s; real wages outpaced productivity while hours worked declined (Figure 1.13).
  - Unit labor costs grew much more in the 2000s than in the 1990s (Figure 1.14).
  - Service sector: ULC increased substantially in the 2000s, more than in manufacturing, largely due to lower productivity gains in services (Figure 1.14).
- Cross-country comparisons:
  - Norway’s ULC growth is significantly higher than peers; productivity growth has been among the lowest (Figure 1.15).
  - Actual hours worked per employee are much lower than other advanced economies (Figure 1.16).
- Sectoral note:
  - If the oil sector and oil-related productive part of the mainland economy are removed, Norway’s ULC growth would be even higher and labor productivity of traditional manufacturing would be even lower.

### D. Market Share and Divergent Trends within the Mainland Economy
- Aggregate market share:
  - Norway has maintained a steady world market share of total exports of goods and services for past few decades (Figure 1.17A), performing better than many peers whose shares declined in the 2000s.
- Non-oil exports:
  - Norway’s market share of non-oil exports has declined substantially over the past two decades and much faster than other advanced economies (Figure 1.17B).
- Manufacturing:
  - Norway has lost almost 60 percent of its manufacturing share of world exports since 1990, among the largest drops compared with peers (Figure 1.17C).
  - Manufacturing market share stabilized in early 2000s but declined again since the global financial crisis.
- Patterns of specialization and two-tier development:
  - Sectoral specialization: among Norway’s manufacturing specializations, Norway lost market share in sectors that are growing in the world market (quadrant II in Figure 1.18).
  - Two-tier development in manufacturing (Figures 1.19–1.20):
    - Oil-related industries (mechanical engineering sectors tied to oil and gas) are expanding and gaining market share.
    - Traditional manufacturing sectors (e.g., base metals) are shrinking and losing market share.
    - Manufacturing production has been largely driven by the oil-related engineering industry since mid-2000s; production in other traditional manufacturing subsectors has been subdued.
- Wage shares (Figure 1.21):
  - The document presents wage share patterns across sectors (internationally exposed industries, traditional; sheltered sector; internationally exposed industries related to oil sector) but specific numeric breakdowns are provided in figures.

### Key implications from the supplied analysis
- High wage growth, low productivity growth, declining hours worked, and growing GPFG relative to GDP have combined to produce a substantial appreciation in the ULC-based REER and a loss of competitiveness in the non-oil mainland economy.
- The mainland economy is increasingly dependent on supplying goods and services to the offshore sector; oil-sector linkages have supported some manufacturing activity but have also propelled wage growth that undermines competitiveness in more traditional non-oil sectors.
- Norway’s export performance is strong at the aggregate level because of oil and gas, but the non-oil tradable sector is small and declining, with substantial market-share losses in non-oil manufacturing since 1990.

*Source: July 19, 2013 — LONG-TERM COMPETITIVENESS IN NORWAY (chapter text and figures).*

### 20.      The split between the oil-related and non-

### _cr13273 - 20.      The split between the oil-related and non-

### Split between oil-related and non-oil mainland economy
- “Traditional” internationally-exposed (tradable) industries have sharply rising wage shares in recent years, reflecting a steep increase in unit labor costs.
- Internationally-exposed industries related to the booming oil industry have been able to absorb rising unit labor costs; the wage share in these industries has been falling over the last decade.

### Conclusion on structural implications
- Norway is losing ground relative to its peers in more traditional non-oil sectors.
- The economy is becoming more dependent on oil and gas activities; more productive parts of the mainland economy, such as engineering, are closely tied to the hydrocarbon sector.
- Insulation from Dutch disease appears incomplete despite the fiscal rule and sound management of oil revenue.
- Oil and gas production affects the mainland economy via indirect channels, including higher wage pressures in traditional non-oil sectors where labor productivity growth has been lagging.
- Weakened cost competitiveness is evidenced by appreciation of ULC-based REER and a striking loss of market share for non-oil exports over past decades.
- Increasing pressures on non-oil parts raise concerns about resilience if oil prices fall substantially:
  - Oil-related parts fuel robust growth but represent a narrow basis for long-run growth.
  - Continued dependence on oil and gas activities implies potential insufficient capacity to supply employment and income during sustained oil price declines or after extraction peaks.

### Household balance sheets — headline vulnerabilities
- Norway’s household debt stood at about 200 percent of disposable income in 2011, among the highest in the OECD.
- Real house prices increased by nearly 30 percent from their lowest level in 2008; real house prices bounced back and rose continuously since 2009.
- Price-to-income and price-to-rent ratios increased by 14 percent and 23 percent, respectively, from their lowest levels in 2008.
- House prices and household debt in Norway continue to rise relative to disposable income, contrasting with corrections elsewhere.

### Are house prices overvalued? (valuation gap estimates and robustness)
- 2011 Norway Article IV concluded house prices might be overvalued by 15–20 percent (based on IMF VEA econometric model).
- New 2013Q1 staff estimates use three measures: (i) IMF VEA econometric model; (ii) deviation of the price-to-rent ratio from its historical average; (iii) deviation of the price-to-income ratio from its historical average.
  - The average estimate across these three measures in 2013Q1 is a little over 40 percent.
  - Excluding the price-to-rent ratio reduces the average estimate to about 30 percent.
  - Estimates change very little when using different historical-average windows (e.g., 1990 to present).
- Caveats and uncertainties:
  - Price-to-rent-based measures may overstate gaps if rent series understates rental price increases (rent series largely captures existing rental contracts moving at CPI inflation due to regulations).
  - Long-term historical averages may not capture “fair” values if earlier periods were highly regulated.
  - Models relying on potentially volatile “fundamentals” (wage/income growth, immigration) may show smaller overvaluations; these fundamentals could reverse if oil prices drop significantly.

### Factors contributing to house price increases (Box 2.1 summary)
- Demand factors:
  - High wage/income growth tied to high oil prices.
  - Population growth: about 1 percent on average for the past decade; average annual immigration growth at around 8 percent.
  - Low interest rate: lending rates have gone down substantially since 2000 and remained low.
- Supply factors:
  - Supply constraints, especially in the Oslo area, due to land-use regulations and minimum unit size; housing supply falling behind increasing number of households.
- Institutional factors:
  - Preferential tax treatment of owner-occupied properties (full deductibility of interest payments on mortgages and lower effective tax rates).
  - Lower risk weights of residential mortgages relative to corporate loans, incentivizing bank lending to households.

### Do households have sufficient buffers?
- Aggregate picture:
  - Household net worth is a little over 200 percent of disposable income in 2011.
  - When adjusting tax-valued non-financial assets (tax valuations believed to understate market values by about one third), household net worth would increase to 340 percent of disposable income, but would still be the second lowest among peers.
- Composition of household assets (2011):
  - Non-financial assets: about 49 percent of total household assets.
  - Pensions and insurance: 20 percent.
  - Currency and deposits: 17 percent.
  - Shares and other equities: 20 percent.
  - Securities other than shares: 7 percent.
  - Other financial assets: 7 percent.
- Liquid buffers:
  - The share of liquid household assets is relatively small.
  - If non-financial assets are revalued to market prices (same correction as above), illiquid assets rise to nearly 80 percent of household financial assets.
  - Liquid assets are smaller than the sum of mostly-floating-rate mortgages and other household financial liabilities, with the gap growing steadily over time.
- Likely household response to shocks:
  - A house price correction is more likely to induce households to reduce consumption rather than default, given full recourse mortgage lending and historically low mortgage losses.
  - Limited buffers imply households would need to increase savings or cut consumption to continue mortgage payments; construction-sector impacts could lower growth and increase bank losses.

### Distributional vulnerabilities (micro data findings)
- Microdata (Norges Bank / Statistics Norway, sample 1987–2010; analysis focused on 2005 onward) shows:
  - Household debts are largely aligned with income levels, but debt-to-income ratios do not differ considerably across income groups, particularly income deciles 4 and above.
  - Exception: income decile 1 is heterogeneous (pensioners with large financial assets, students, immigrants), making interpretation difficult.
  - Household debt rose since 2005 in all income groups.
  - Net worth in higher income groups (deciles 8 and 9) is near-zero; these groups also have the highest debt-to-income ratios.
  - Non-financial assets (housing) are almost evenly distributed across income groups.
  - Financial assets are generally lower than total liabilities, with the gap negative for most income groups.
- Implication:
  - Debt burdens are relatively evenly distributed across income groups in Norway.
  - Most income groups (except the very lowest decile) have relatively limited buffers against income loss or house price reversals, implying significant potential impacts on consumption in the event of a house price correction.

*Source: _cr13273 - 20.      The split between the oil-related and non-*

### 13.      The number of households that are highly

### _cr13273 - 13.      The number of households that are highly

### Household indebtedness and liquidity buffers
- The share of households whose debt is more than 500 percent of disposable income rose from 4 percent in mid-1990s to 11 percent in 2010.
- The tendency toward high indebtedness is prevalent in almost all income groups.
- The staff adopted the gap between financial assets and financial liabilities as an alternative measure for “liquidity buffers” because the breakdown of liquid and illiquid assets was not available to the staff.
- These highly indebted households would be even more vulnerable to a house price correction and other negative shocks.

### Distribution of loan-to-value (LTV) and recent loan characteristics
- New loans exceeding the FSA-recommended 85 percent LTV cap fell to 17 percent in 2012.
- The share of interest-only loans among new loans declined from 23 percent in 2011 to 17 percent in 2012.
- Saving rates are rising recently, suggesting increased household caution.

### Conclusion on household vulnerability and financial stability (Section E)
- Norwegian households have limited financial buffers in the event of a house price correction or other negative shocks; this holds regardless of household income levels.
- Elevated household debt poses a significant challenge:
  - A house price correction could significantly reduce household consumption through wealth effects.
  - Banks’ losses related to residential mortgages may not increase much, but financial stability could be undermined if lower consumption impairs business activity and thus pushes up bank losses associated with enterprise loans.
- Ongoing strengthening of capital requirements for banks will help mitigate some risks, but rigorous monitoring needs to continue on residential mortgage lending.
- Note: The FSA tightened the LTV cap by lowering it from 90 percent to 85 percent in December 2011.

### Long-run fiscal challenges: overview and purpose of chapter
- The chapter considers long-term implications of Norway’s fiscal rule under different scenarios by varying:
  - real rates of return;
  - the share of the Government Pension Fund Global (GPFG) assets transferred to the budget each year; and
  - oil prices.
- Findings summarized:
  - A lower real rate of return has large implications for the long-run value of the sovereign wealth fund.
  - The rate of fiscal transfer is the key parameter that would mitigate or exacerbate “Dutch disease” pressures and the increasing non-oil fiscal deficit.
  - Oil prices have limited impact on the rate of transfer unless there are sustained deviations from the baseline over the next couple of decades; after that they become less important as most of the oil will have already been extracted.
  - Combined with long-run aging-related costs, the scenarios indicate that further pension reforms are needed to address long-run aging-related costs.

### Resource wealth, GPFG design, and governance (Section B)
- Oil-related metrics and importance:
  - Oil production now accounts for over one fifth of total GDP, one quarter each of government revenues and total investment, and over fifty percent of total exports.
  - Norway was the world’s third largest exporter of gas and the seventh largest exporter of oil in 2011.
- GPFG characteristics:
  - Total assets of the GPFG were $713 billion at end-2012.
  - Current target portfolio: 60 percent equity, 35 percent fixed-income, and 5 percent real estate.
  - Management costs are low at 7-10 basis points annually.
- Governance:
  - The GPFG is owned by the government (Ministry of Finance) and managed by Norges Bank (NBIM).
  - The Ministry sets investment strategy and broad allocations; NBIM is charged with achieving the highest possible return subject to limits and governance, environmental, and social concerns.
  - An independent Council on Ethics advises the Ministry on exclusions.

### The fiscal rule: design and practice
- 2001 fiscal guidelines: “Petroleum income should be phased into the economy on par with the developments in expected returns on the Government Pension Fund Global. Considerable emphasis must be put on stabilizing the economy.”
- The expected real return has been estimated at 4 percent for purposes of establishing the trend fiscal transfer.
- The rule permits spending more than the expected return in a cyclical downturn and spending below the expected return when capacity utilization is high.
- The fiscal rule constrains the structural non-oil deficit as a share of GPFG assets, but not as a share of mainland GDP.
- Because GPFG assets have grown much more rapidly than the economy, the rule has allowed a steadily rising non-oil deficit as a share of mainland GDP; when GPFG assets start to decline relative to mainland GDP (expected in about two decades), transfers would decline as a share of mainland GDP.

### Scenario analysis and baseline projections (Section C)
- Baseline assumptions:
  - Authorities’ projections for oil production volumes.
  - A 4 percent real rate of return.
  - An average fiscal transfer equal to the assumed trend real rate of return of 4 percent.
  - Oil price projections based on WEO price projections and futures markets, remaining constant in real terms after 2022.
  - Population growth declines from 1.3 percent to 0.2 percent by 2100.
  - Real GDP growth: staff projections through 2015 and authorities’ projections thereafter (3.3 percent in 2015 declining to 1.8 percent in 2060-2100).
- Baseline projections:
  - The GPFG peaks at about 265 percent of mainland GDP by 2030 and declines thereafter.
  - The non-oil deficit rises sharply as a share of mainland GDP over the next decade and a half, peaking at just below 11 percent of mainland GDP by 2030.
  - Fiscal transfers decline thereafter, to just below 4 percent of mainland GDP by 2100.
- Implications:
  - The fiscal rule would preserve the real value of oil wealth in perpetuity if fiscal transfers equal the ex post real return.
  - Maintaining the GPFG relative to a growing real per capita GDP would be a regressive transfer from less wealthy current generations to wealthier future ones.
  - The uneven pattern of fiscal stimulus could have significant implications for competitiveness and preservation of a tradable goods and services sector outside oil.
  - Long-term fiscal outlook is subject to considerable uncertainty: small changes in key assumptions can have very large effects due to compounding.
  - Shorter time horizons matter for managing crowding out pressure on non-oil tradable sectors in the next two decades.

### Alternative scenarios considered
- Types of scenarios compared:
  - Baseline (authorities’ assumptions).
  - Alternative scenarios varying: real rate of return (in line with or lower than historical experience rather than the assumed 4 percent), a more conservative approach to fiscal transfers than implied by an average of 4 percent of GPFG assets per year, and lower oil prices individually.
  - Combined scenarios where multiple parameters vary simultaneously.
  - Scenarios comparing projections of pension expenditures and fiscal transfers.
- Historical returns context:
  - Earned 3.2 total real rate of return since 1997 (Figure 3.5).
  - Figure 3.5 shows evolution of real returns of GPFG with total return, equity, and fixed income series (data for 2013 are as of March 31).

*Italic: IMF staff compilation from document content.*

### 17.      The assumed 4 percent real rate of return for the GPFG is high compared with its past

### _cr13273 - 17.      The assumed 4 percent real rate of return for the GPFG is high compared with its past

### Returns versus the 4 percent assumption
- Lifetime (1997–2012) real rate of return on the GPFG: 3.2 percent.
- Difference between assumed and historical: 0.8 percent (4 percent minus 3.2 percent) — noted as having substantial consequences for the long-run value of the GPFG and reducing the fiscal transfers it can finance.
- Long-run global benchmark (60 percent equities / 40 percent bonds) real rate of return for the past 113 years (1900–2012): 3.7 percent.
- World average real return on equity from 1900 to 2012: 5 percent; global equity-bond outperformance since 1900: 1.8 percent.
- European rates of return: equities 4.2 percent and bonds 0.8 percent (1900–2012).

### Risk that future returns may be lower than historical
- Possibility that GPFG may not maintain historical average returns if:
  - Asset price cycles align with oil prices and production, resulting in a large share of investments being made when markets are high.
  - Longer run rates of return fall below those experienced in recent decades.
- Some analysts (McKinsey (2010), Siegel (2012), Dimson et al. (2013)) argue that today’s low interest rates are the best estimate for the expected real return over the next two decades and represent a new equilibrium.

### Effect of varying real rates of return on fiscal transfers
- Baseline scenario real rate of return: 4 percent.
- Historical average scenario real rate of return: 3.2 percent.
- Lower rate scenario used in comparison: 2.5 percent.
- Near-term (until about 2030): differences in fiscal transfer across these rates are less than one or two percentage points of mainland GDP between highest and lowest projections.
- Long run (end of century): lower real rates reduce fiscal transfer relative to baseline by 1.7 and 2.5 percentage points of mainland GDP (for intermediate and low return scenarios as described in the text).

### A lower rate of fiscal transfer (spending rule effects)
- Transfer rate examples:
  - 3.0 percent of GPFG: translates into lower fiscal transfer as share of mainland GDP than baseline until 2054, but higher thereafter.
  - 2.0 percent of GPFG: translates into lower fiscal transfer in percent of mainland GDP in the coming years which increases to about 8 percent of mainland GDP in 2050-2060 and remains roughly in that range through 2100.
- The percentage of GPFG assets transferred to the budget each year has a direct and immediate impact on the non-oil deficit.

### Impact of a permanent oil price decline
- Permanent, 40 percent decline in real oil prices:
  - Affects short- and medium-term fiscal transfers by somewhat (2 percentage points) but less so the long run fiscal transfer.
  - Proportional impact on GPFG assets is slightly more than half of the percentage change in oil prices because 44 percent of the total oil resources had been extracted as of 2012.
  - Impact on GPFG assets and fiscal transfer is largely phased in by mid-century, by which time nearly 90 percent of the remaining oil resources will have been extracted.

### Combined scenarios (definitions and comparative outcomes)
- Scenario definitions:
  - Baseline: real rate of return = 4 percent and fiscal transfer = 4 percent.
  - Conservative: fiscal transfer = 3 percent; rate of return = 4 percent.
  - Constant 4% transfer: fiscal transfer = 4 percent; rate of return = 3.2 percent (historical).
  - Prudent: fiscal transfer = 3 percent; rate of return = 3.2 percent.
  - Low transfer: fiscal transfer = 2 percent; rate of return = 3.2 percent.
- Selected results from Table 3.1 (scenario summary):
  - Baseline: Rate of return (percent) 4.0; Rate of fiscal transfer (%MGDP) 4.0; Fiscal transfer 2020 (%MGDP) 9.7; Fiscal transfer 2060 (%MGDP) 8.1; Fiscal transfer 2100 (%MGDP) 3.8; GPFG peak year 2029; GPFG assets 2020 (%MGDP) 243; GPFG assets 2060 (%MGDP) 203; GPFG assets 2100 (%MGDP) 95.
  - Constant 4% transfer scenario: Rate of return (percent) 3.2; Rate of fiscal transfer (%MGDP) 4.0; Fiscal transfer 2020 (%MGDP) 9.3; Fiscal transfer 2060 (%MGDP) 6.2; Fiscal transfer 2100 (%MGDP) 2.2; GPFG peak year 2027; GPFG assets 2020 (%MGDP) 233; GPFG assets 2060 (%MGDP) 155; GPFG assets 2100 (%MGDP) 54.
  - Low return: Rate of return (percent) 2.5; Rate of fiscal transfer (%MGDP) 4.0; Fiscal transfer 2020 (%MGDP) 9.0; Fiscal transfer 2060 (%MGDP) 5.0; Fiscal transfer 2100 (%MGDP) 1.3; GPFG peak year 2025; GPFG assets 2020 (%MGDP) 224; GPFG assets 2060 (%MGDP) 124; GPFG assets 2100 (%MGDP) 33.
  - Conservative scenario: Rate of return (percent) 4.0; Rate of fiscal transfer (%MGDP) 3.0; Fiscal transfer 2020 (%MGDP) 7.7; Fiscal transfer 2060 (%MGDP) 8.6; Fiscal transfer 2100 (%MGDP) 6.0; GPFG peak year 2043; GPFG assets 2020 (%MGDP) 255; GPFG assets 2060 (%MGDP) 287; GPFG assets 2100 (%MGDP) 200.
  - Prudent scenario: Rate of return (percent) 3.2; Rate of fiscal transfer (%MGDP) 3.0; Fiscal transfer 2020 (%MGDP) 7.3; Fiscal transfer 2060 (%MGDP) 6.5; Fiscal transfer 2100 (%MGDP) 3.4; GPFG peak year 2030; GPFG assets 2020 (%MGDP) 245; GPFG assets 2060 (%MGDP) 218; GPFG assets 2100 (%MGDP) 112.
  - Low transfer scenario: Rate of return (percent) 3.2; Rate of fiscal transfer (%MGDP) 2.0; Fiscal transfer 2020 (%MGDP) 5.1; Fiscal transfer 2060 (%MGDP) 6.2; Fiscal transfer 2100 (%MGDP) 4.7; GPFG peak year 2046; GPFG assets 2020 (%MGDP) 257; GPFG assets 2060 (%MGDP) 310; GPFG assets 2100 (%MGDP) 235.
  - Low oil price (40% lower), high return: Rate of return (percent) 4.0; Rate of fiscal transfer (%MGDP) 4.0; Fiscal transfer 2020 (%MGDP) 8.3; Fiscal transfer 2060 (%MGDP) 6.1; Fiscal transfer 2100 (%MGDP) 2.9; GPFG peak year 2027; GPFG assets 2020 (%MGDP) 208; GPFG assets 2060 (%MGDP) 153; GPFG assets 2100 (%MGDP) 72.
  - Low oil price (40% lower), historical return: Rate of return (percent) 3.2; Rate of fiscal transfer (%MGDP) 4.0; Fiscal transfer 2020 (%MGDP) 7.9; Fiscal transfer 2060 (%MGDP) 4.6; Fiscal transfer 2100 (%MGDP) 1.6; GPFG peak year 2013; GPFG assets 2020 (%MGDP) 199; GPFG assets 2060 (%MGDP) 115; GPFG assets 2100 (%MGDP) 40.

### Interactions with aging and pension costs
- Old-age ratio projection: ratio of elderly (65+) to working-age population (20–64) is projected to nearly double from 2013 to 2060.
- Statistics Norway estimate: the pension reform would gradually raise the participation rate by 8 percent by 2060.
- Disability pensions: Norway’s disability pensions are 1.6 percentage points of GDP higher than the average of higher-income OECD comparators in 2013.
- Even with pension reforms and reductions in disability pensions, pension expenditure would remain higher than fiscal transfers from GPFG unless additional reforms are implemented (e.g., 8 percent increase in participation rate and reducing disability pensions to the OECD average).

### Conclusions and policy recommendations (key points)
- The amount of the annual fiscal transfer as a share of GPFG assets is the key parameter affecting the size of the fiscal transfer over the next two decades and is central to addressing competitiveness and Dutch disease concerns.
- Over longer horizons, the realized real rate of return and oil prices become increasingly important for intergenerational equity and aligning fiscal transfers with aging-related costs.
- Reasons to consider a fiscal transfer rate below 4 percent of GPFG assets:
  - The real rate of return might align with history (around 3.2 percent) rather than 4 percent; if long-run real rate of return is below 4 percent, this will only be evident after a long lag.
  - Stabilizing the economy to prevent Dutch disease may require a lower rate of spending to lessen crowding out of “exposed industries.”
  - A slower rate of spending could better align expected fiscal transfers with expected long-run aging-related costs that peak after oil production declines.

*Source: _cr13273 - 17.      The assumed 4 percent real rate of return for the GPFG is high compared with its past*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2013/_cr13273.pdf_