## _wp06193

## Source details

**Canonical URL:** [_wp06193](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06193.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06193.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06193.pdf.json)

---

### I. Introduction: problem statement and modelling approach
- Major oil-producing countries exhibit overall fiscal surpluses and non-oil deficits when petroleum prices are high; if reserves decline, maintaining non-oil deficits may become impossible.
- Abundant, but exhaustible, oil revenues create political pressure to increase current spending and form spending habits, complicating later fiscal consolidation.
- Existing models based on Friedman’s (1957) permanent income hypothesis (PIH) recommend an abrupt, one-time fiscal consolidation to a constant expenditure path equal to the annuity value of oil wealth and non-oil revenue; political realities often preclude this approach.
- The paper proposes a quantitative model that:
  - Incorporates short-run social and political adjustment costs (“habits”) linking past consumption of public goods and services to the speed of fiscal adjustment.
  - Allows for different interest rates on sovereign debt and financial assets.
  - Seeks to estimate Gabon’s permanently sustainable non-oil primary deficit and the optimal adjustment path toward that level.
- Framing: the optimal path is defined as the one a social planner would choose.

### II. Background: institutional context and Gabon’s experience
- Emerging oil exporters are creating institutional arrangements (fiscal reserves and rules over use of oil revenue) to limit fiscal discretion over oil revenue.
- Many recent oil-fund arrangements treat all income from oil production as different from other revenue streams (examples noted: Azerbaijan, Mauretania, São Tomé and Príncipe, Timor-Leste).
- Gabon’s historical pattern:
  - Governments treated oil revenues analogously to tax receipts, engaging in large-scale projects with high recurrent costs; in periods of falling oil prices, these recurrent costs exceeded budget resources and required dramatic fiscal adjustments.
  - Resulting legacy: stalled socioeconomic development, substandard and ill-maintained public infrastructure, a high degree of public indebtedness, and an uncompetitive non-oil sector.
- Institutional reforms in Gabon:
  - Post-1998 institutional changes included establishment of a Fonds pour les générations futures (FGF) with corresponding fiscal rules, though a number of these rules have not been implemented.
  - In early 2006, the Government of Gabon (2006) approved a Growth and Poverty-Reduction Strategy Paper (PRSP) acknowledging related challenges and embedding socioeconomic development strategy into a macro-fiscal framework.

### III. Key empirical and analytical elements
- Figures and tables referenced in the source include:
  - Figure 1: Overall Fiscal Balance and International Oil Price, 1980–2005.
  - Table 1: Emerging Oil Exporters: Limits to Fiscal Discretion over Oil Revenues.
  - Box 1: Gabon’s Fund for Future Generations.
  - Other figures and tables: Oil Production Profile and World Oil Prices, 2005–45; Optimal Adjustment Path Under Baseline Assumptions, 2000–45; Sensitivity Analysis on Habit Strength and the Optimal Adjustment Path, 2000–45; Introducing a Debt/Asset Interest Rate Spread, 2000–45; Oil Production and Socioeconomic Development; Baseline Assumptions; Sensitivity Analysis.
- Specific FGF rules reported (Gabon, 1998):
  - Budgeted oil revenues into Fonds pour les générations futures (FGF):
    - 10% if FGF < CFAF 500 billion,
    - 25% of FGF revenue if FGF > CFAF 500 billion.
  - Windfall oil revenues:
    - 50% if FGF < CFAF 500 billion,
    - 100% if FGF > CFAF 500 billion.
  - Note: these rules were not implemented.

### Box 1 — Gabon’s Fund for Future Generations (FGF): law, implementation, outcomes, and implications
- The law:
  - Law No. 9/98 (mid-1998) established a Fonds pour les générations futures (FGF) and fiscal rules to guarantee accumulation of fiscal reserves.
  - Law No. 9/98 defined a minimum FGF capital of CFAF 500 billion (Art. 2), off-limit to the government even in exceptional circumstances or as a source of collateral (Art. 3).
  - The assets were to be invested in stocks of companies with international reputation and low-risk financial instruments (Art. 6).
  - Only during a transition period up until end-1998, the budget was to contribute an exceptional endowment, which the regional central bank was to manage in a special account (Art. 14).
- Implementation and outcomes:
  - Transitional stipulations continued to be applied after 1998: FGF funds were held by the Banque des Etats de l’Afrique centrale (BEAC) and remunerated in line with other “special accounts.”
  - Reported BEAC special-account remunerations increased from 1.6 percent at end-2005 to 1.9 percent in March 2006 and—to 2.2 percent in June 2006—by creating a special category of fiscal reserves for future generations (BEAC Decision No. 09/GR/2006).
  - Contributions by the budget to the FGF were sporadic, and—in violation of the FGF law—capital was withdrawn in 2003.
  - By end-2005, the budget contributed CFAF 84 billion to the FGF, representing 8.8 percent of the amount that would have been saved had the law been respected fully.
  - During the same period, the authorities significantly reduced their stock of public debt, including arrears. Given that Gabon’s debt carries an interest rate far in excess of the FGF interest rate, the government’s realized net savings were probably a more rational strategy of reserve management, even though they were more erratic than what the FGF law implied and represented only about 60 percent of the implicit total asset accumulation vis-à-vis the oil fund.
- Key statistics and fiscal mechanics (as reported in the box):
  - Minimum FGF capital mandated by law: CFAF 500 billion.
  - Budget contribution to FGF by end-2005: CFAF 84 billion (8.8 percent of legally implied savings).
  - BEAC special-account remuneration rates reported: 1.6 percent (end-2005); 1.9 percent (March 2006); 2.2 percent (June 2006).
  - Government’s realized net asset accumulation relative to the implicit total oil-fund accumulation: about 60 percent.
  - Capital withdrawal from the FGF occurred in 2003 (in violation of the FGF law).
- Analysis and policy implications:
  - The FGF law aimed to insulate a minimum fiscal reserve (CFAF 500 billion) from political pressures and restrict permissible investments to low-risk assets and reputable stocks.
  - Actual practices diverged from the law through sporadic budget contributions and a 2003 withdrawal, reducing the effectiveness of the statutory savings rule.
  - The authorities’ strategy of using oil receipts to reduce high-interest public debt delivered realized net savings that may have been more economically rational than strictly adhering to the FGF law, given the FGF’s lower interest earnings; however, this strategy was more erratic and did not achieve the legally envisaged accumulation.
  - Low returns on FGF holdings (as of June 2006, a mere 2.15 percent nominal rate is noted elsewhere in the chapter) suggest that reforms to raise returns—potentially by redesigning the FGF toward instruments and governance arrangements akin to higher-performing sovereign wealth funds—would be needed to improve fiscal reserve management.

### Baseline assumptions and key parameters
- Effective oil tax take z = 35.9 percent of oil GDP.
- Effective non-oil tax take τ = 23.9 percent of non-oil GDP.
- Total primary expenditure in 2005 = 36.4 percent of non-oil GDP in 2005.
- Total public debt in 2005 = 95.2 percent of non-oil GDP in 2005.
- Sum of all future oil revenue = 711.0 percent of non-oil GDP in 2005.
- Real interest rate r = 3.0 percent.
- Real non-oil growth rate γ = 2.0 percent.
- Habit strength α = 0.7.
- Note: "These are held constant at the actual values realized in 2005."

### Main simulation results
- Starting point: 2005 non-oil primary deficit = 12.5 percent of non-oil GDP.
- First major result:
  - The 2005 non-oil primary deficit is not sustainable; under baseline assumptions the permanently sustainable non-oil primary deficit = 5.5 percent of non-oil GDP.
  - Sensitivity: if total reserves increase by 25 percent relative to baseline, sustainable deficit rises to 7.0 percent of non-oil GDP.
  - Sensitivity: if oil tax take rises by 10 percentage points to 46 percent, sustainable non-oil primary deficit = 7.4 percent of non-oil GDP.
- Second major result (optimal adjustment timing):
  - Under baseline parameters, non-oil deficit declines by 4.6 percentage points from 12.5 to 7.9 percent of non-oil GDP during 2005–08 (more than 65 percent of total adjustment).
  - By 2010, non-oil primary deficit = 6.7 percent of non-oil GDP (implying 83 percent of required adjustment completed).
  - Primary surpluses occur during oil period to pay off debt and accumulate financial assets; permanent sustainable level converges to 5.5 percent of GDP in 2036 (year when oil revenue assumed to dry up).
  - Adjustment path depends on habit strength; alternative paths vary with α within the empirical range.
- Third major result (risk and precautionary motive):
  - A risk-averse policymaker would prefer faster adjustment and lower long-run deficit.
  - If oil price reverts to US$30/bbl, permanently sustainable level = 4.8 percent of non-oil GDP.
  - If effective oil tax take declines by 10 percentage points to 26 percent, sustainable deficit = 3.6 percent of non-oil GDP.
  - Real interest rate is the most critical assumption with highest downside risks; precautionary savings argument implies front-loading fiscal adjustment.
  - Note: uncertainty and risk aversion are not formally modeled in this paper.

### Sensitivity analysis (selected results from Table 4; Permanently Sustainable Non-Oil Primary Deficit, Percent of non-oil GDP)
- Baseline long-run oil price scenarios:
  - US$30/bbl = 4.8
  - US$57/bbl = 5.5
  - US$84/bbl = 7.6
- Oil reserves:
  - Higher oil reserves (125 percent) = 5.7 / 7.0 / 9.8 (US$30/57/84).
  - Lower oil reserves (75 percent) = 3.8 / 4.0 / 5.2.
- Effective oil tax take z:
  - Higher z = 45.9 percent of oil GDP → 6.5 / 7.4 / 10.0.
  - Lower z = 25.9 percent of oil GDP → 3.1 / 3.6 / 5.1.
- Non-oil tax take τ:
  - Higher τ = 33.9 percent of non-oil GDP → 5.0 / 5.7 / 7.8.
  - Lower τ = 13.9 percent of non-oil GDP → 4.6 / 5.3 / 7.3.
- Real interest rate r:
  - Higher r = 3.5 percent → 6.9 / 7.9 / 10.8.
  - Lower r = 2.5 percent → 2.5 / 2.9 / 4.0.
- Real non-oil growth rate γ:
  - Higher γ = 2.5 percent → 2.5 / 2.9 / 4.0.
  - Lower γ = 1.5 percent → 6.9 / 7.9 / 10.8.
- Habit strength α:
  - No habits α = 0.05 → 5.0 / 5.7 / 7.7.
  - Stronger habits α = 0.8 → 4.7 / 5.4 / 7.5.
  - Weaker habits α = 0.6 → 4.9 / 5.6 / 7.6.

### Extensions and model implications
- Introducing an interest-rate spread between sovereign debt and oil-fund assets:
  - Creates incentive to run a smaller non-oil deficit in the short run and to repay debt sooner.
  - With a 50 basis point spread the optimal path includes more rapid repayment of debt, earlier net asset accumulation, and a higher permanently sustainable fiscal deficit.
- Redefining objective to guarantee stability in real per capita expenditure:
  - Adding population growth lowers net real interest rate (r minus γ and population growth) and lowers the optimal sustainable primary deficit; implies faster adjustment is required.
- Incorporating public investment:
  - If government expenditure includes productive public investment, the initial fiscal deficit could be higher as investment raises future productive capacity.
  - Barnett and Ossowski condition (equation (13)) implies governments should invest in projects that pay for themselves; with tax rate 23.9 percent and interest rate 3 percent, rate of return on public investment would have to exceed 12.6 percent given the simulations’ assumptions.
- Portfolio decision implications:
  - Trade-off between financial assets and physical/social assets; different rates of return on debt, financial, physical, and social investments would affect optimal policy.

### Concluding remarks and policy recommendations
- Three main conclusions:
  - Gabon’s current non-oil primary deficit is not sustainable: permanently sustainable non-oil primary deficit estimated at 5.5 percent of non-oil GDP vs actual 12.5 percent in 2005.
  - Habit formation implies spreading adjustment over three to five years rather than a single abrupt cut.
  - An interest rate spread incentivizes front-loading adjustment to pay off net debt sooner; uncertainty gives a risk-averse policymaker a precautionary motive to front-load adjustment.
- Recommended elements of an economic reform program for Gabon:
  - (i) a phased adjustment of the primary non-oil balance to a permanently sustainable level;
  - (ii) reforms to the management of FGF assets;
  - (iii) a rapid repayment of public debt;
  - (iv) structural reforms aimed at improving the design and quality of public investments.
- Future research directions:
  - Relax assumption that government expenditure is only consumption and has no effect on productivity growth.
  - Model different rates of return on government debt and on financial, physical, and social investment.
  - Endogenize allocation between physical and financial investments in a finite-horizon optimization maximizing national wealth, explicitly modeling the interaction between investment quality and future growth.

*Source: IMF working paper content unit titled "35.9   percent of oil GDP" (excerpt).*

### References..............................................................................................................

### _wp06193 - References............................................................................................................................... 25

### I. Introduction: problem statement and modelling approach
- Major oil-producing countries exhibit overall fiscal surpluses and non-oil deficits when petroleum prices are high; if reserves decline, maintaining non-oil deficits may become impossible.
- Abundant, but exhaustible, oil revenues create political pressure to increase current spending and form spending habits, complicating later fiscal consolidation.
- Existing models based on Friedman’s (1957) permanent income hypothesis (PIH) recommend an abrupt, one-time fiscal consolidation to a constant expenditure path equal to the annuity value of oil wealth and non-oil revenue; political realities often preclude this approach.
- The paper proposes a quantitative model that:
  - Incorporates short-run social and political adjustment costs (“habits”) linking past consumption of public goods and services to the speed of fiscal adjustment.
  - Allows for different interest rates on sovereign debt and financial assets.
  - Seeks to estimate Gabon’s permanently sustainable non-oil primary deficit and the optimal adjustment path toward that level.
- Framing: the optimal path is defined as the one a social planner would choose.

### II. Background: institutional context and Gabon’s experience
- Emerging oil exporters are creating institutional arrangements (fiscal reserves and rules over use of oil revenue) to limit fiscal discretion over oil revenue (Table 1).
- Many recent oil-fund arrangements treat all income from oil production as different from other revenue streams (examples noted: Azerbaijan, Mauretania, São Tomé and Príncipe, Timor-Leste).
- Gabon’s historical pattern:
  - Governments treated oil revenues analogously to tax receipts, engaging in large-scale projects with high recurrent costs; in periods of falling oil prices, these recurrent costs exceeded budget resources and required dramatic fiscal adjustments.
  - Resulting legacy: stalled socioeconomic development, substandard and ill-maintained public infrastructure, a high degree of public indebtedness, and an uncompetitive non-oil sector.
- Institutional reforms in Gabon:
  - Post-1998 institutional changes included establishment of a Fonds pour les générations futures (FGF) with corresponding fiscal rules, though a number of these rules have not been implemented (Box 1).
  - In early 2006, the Government of Gabon (2006) approved a Growth and Poverty-Reduction Strategy Paper (PRSP) acknowledging related challenges and embedding socioeconomic development strategy into a macro-fiscal framework.

### III. Key empirical and analytical elements
- Figure and table references in the source:
  - Figure 1: Overall Fiscal Balance and International Oil Price, 1980–2005 (shows boom-and-bust cycles for Gabon and oil-exporting countries in sub-Saharan Africa; nominal crude oil prices on right-hand scale).
  - Table 1: Emerging Oil Exporters: Limits to Fiscal Discretion over Oil Revenues.
  - Box 1: Gabon’s Fund for Future Generations.
  - Other figures and tables in the paper (listed): Oil Production Profile and World Oil Prices, 2005–45; Optimal Adjustment Path Under Baseline Assumptions, 2000–45; Sensitivity Analysis on Habit Strength and the Optimal Adjustment Path, 2000–45; Introducing a Debt/Asset Interest Rate Spread, 2000–45; Oil Production and Socioeconomic Development; Baseline Assumptions; Sensitivity Analysis.
- Specific FGF rules as reported in Table 1 (Gabon, 1998):
  - Budgeted oil revenues into Fonds pour les générations futures (FGF):
    - 10% if FGF < CFAF 500 billion,
    - 25% of FGF revenue if FGF > CFAF 500 billion.
  - Windfall oil revenues:
    - 50% if FGF < CFAF 500 billion,
    - 100% if FGF > CFAF 500 billion.
  - Note: these rules were not implemented.

### IV. Main conclusions and policy implications (as stated)
- Three main conclusions:
  - First, Gabon’s current fiscal policy stance cannot be maintained.
  - Second, due to habit formation, the optimal policy spreads the bulk of the adjustment over three to five years, rather than requiring a single correction that standard permanent income models would prescribe.
  - Third, the interest rate differential between sovereign debt and financial assets creates an incentive for front-loading adjustment and paying off debt sooner than in the absence of the spread.
- Additional policy considerations:
  - Gradual adjustment is politically preferable because it reduces short-run pain and increases acceptability.
  - Front-loading adjustment is further supported by precautionary considerations stemming from uncertainty regarding future economic conditions.
  - A clear fiscal anchor—established at the start of oil production—helps distinguish sound, forward-looking fiscal policies from short-sighted ones and can prevent large, sudden contractions when oil revenues are exhausted.
  - Institutional arrangements (fiscal reserves, rules over oil revenue use) aim to limit fiscal discretion and prevent formation of difficult-to-reverse spending habits, but implementation remains challenging.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06193.pdf*

### Box 1.   Gabon’s Fund for Future Generations

### Box 1.   Gabon’s Fund for Future Generations

### The law
- Law No. 9/98 (mid-1998) established a Fonds pour les générations futures (FGF) and fiscal rules to guarantee accumulation of fiscal reserves.
- Law No. 9/98 defined a minimum FGF capital of CFAF 500 billion (Art. 2), off-limit to the government even in exceptional circumstances or as a source of collateral (Art. 3).
- The assets were to be invested in stocks of companies with international reputation and low-risk financial instruments (Art. 6).
- Only during a transition period up until end-1998, the budget was to contribute an exceptional endowment, which the regional central bank was to manage in a special account (Art. 14).

### Implementation and outcomes
- Transitional stipulations continued to be applied after 1998: FGF funds were held by the Banque des Etats de l’Afrique centrale (BEAC) and remunerated in line with other “special accounts.”
- Reported BEAC special-account remunerations increased from 1.6 percent at end-2005 to 1.9 percent in March 2006 and—to 2.2 percent in June 2006—by creating a special category of fiscal reserves for future generations (BEAC Decision No. 09/GR/2006).
- Contributions by the budget to the FGF were sporadic, and—in violation of the FGF law—capital was withdrawn in 2003.
- By end-2005, the budget contributed CFAF 84 billion to the FGF, representing 8.8 percent of the amount that would have been saved had the law been respected fully.
- During the same period, the authorities significantly reduced their stock of public debt, including arrears. Given that Gabon’s debt carries an interest rate far in excess of the FGF interest rate, the government’s realized net savings were probably a more rational strategy of reserve management, even though they were more erratic than what the FGF law implied and represented only about 60 percent of the implicit total asset accumulation vis-à-vis the oil fund.

### Key statistics and fiscal mechanics (as reported in the box)
- Minimum FGF capital mandated by law: CFAF 500 billion.
- Budget contribution to FGF by end-2005: CFAF 84 billion (8.8 percent of legally implied savings).
- BEAC special-account remuneration rates reported: 1.6 percent (end-2005); 1.9 percent (March 2006); 2.2 percent (June 2006).
- Government’s realized net asset accumulation relative to the implicit total oil-fund accumulation: about 60 percent.
- Capital withdrawal from the FGF occurred in 2003 (in violation of the FGF law).

### Analysis and policy implications (from the box)
- Legal design: The FGF law aimed to insulate a minimum fiscal reserve (CFAF 500 billion) from political pressures by making it off-limits even in exceptional circumstances and restricting permissible investments to low-risk assets and reputable stocks.
- Implementation gap: Actual practices diverged from the law through sporadic budget contributions and a 2003 withdrawal, reducing the effectiveness of the statutory savings rule.
- Trade-off in reserve management: The authorities’ strategy of using oil receipts to reduce high-interest public debt delivered realized net savings that may have been more economically rational than strictly adhering to the FGF law, given the FGF’s lower interest earnings; however, this strategy was more erratic and did not achieve the legally envisaged accumulation.
- Institutional challenge: Low returns on FGF holdings (as of June 2006, a mere 2.15 percent nominal rate is noted elsewhere in the chapter) suggest that reforms to raise returns—potentially by redesigning the FGF toward instruments and governance arrangements akin to higher-performing sovereign wealth funds—would be needed to improve fiscal reserve management.

*Source: Gabonese authorities; and Fund staff estimates. (Box 1, "Gabon’s Fund for Future Generations")*

### 35.9   percent of oil GDP

### 35.9   percent of oil GDP

### Baseline assumptions and key parameters
- Effective oil tax take z = 35.9 percent of oil GDP.
- Effective non-oil tax take τ = 23.9 percent of non-oil GDP.
- Total primary expenditure in 2005 = 36.4 percent of non-oil GDP in 2005.
- Total public debt in 2005 = 95.2 percent of non-oil GDP in 2005.
- Sum of all future oil revenue = 711.0 percent of non-oil GDP in 2005.
- Real interest rate r = 3.0 percent.
- Real non-oil growth rate γ = 2.0 percent.
- Habit strength α = 0.7.
- Note: "These are held constant at the actual values realized in 2005."

### Main simulation results
- Starting point: 2005 non-oil primary deficit = 12.5 percent of non-oil GDP.
- First major result:
  - The 2005 non-oil primary deficit is not sustainable; under baseline assumptions the permanently sustainable non-oil primary deficit = 5.5 percent of non-oil GDP.
  - Sensitivity: if total reserves increase by 25 percent relative to baseline, sustainable deficit rises to 7.0 percent of non-oil GDP.
  - Sensitivity: if oil tax take rises by 10 percentage points to 46 percent, sustainable non-oil primary deficit = 7.4 percent of non-oil GDP.
- Second major result (optimal adjustment timing):
  - Under baseline parameters, non-oil deficit declines by 4.6 percentage points from 12.5 to 7.9 percent of non-oil GDP during 2005–08 (more than 65 percent of total adjustment).
  - By 2010, non-oil primary deficit = 6.7 percent of non-oil GDP (implying 83 percent of required adjustment completed).
  - Primary surpluses occur during oil period to pay off debt and accumulate financial assets; permanent sustainable level converges to 5.5 percent of GDP in 2036 (year when oil revenue assumed to dry up).
  - Adjustment path depends on habit strength; Figure 4 shows alternative paths for α values within empirical range.
- Third major result (risk and precautionary motive):
  - A risk-averse policymaker would prefer faster adjustment and lower long-run deficit.
  - If oil price reverts to US$30/bbl, permanently sustainable level = 4.8 percent of non-oil GDP.
  - If effective oil tax take declines by 10 percentage points to 26 percent, sustainable deficit = 3.6 percent of non-oil GDP.
  - Real interest rate is the most critical assumption with highest downside risks; precautionary savings argument implies front-loading fiscal adjustment.
  - Note: uncertainty and risk aversion are not formally modeled in this paper.

### Sensitivity analysis (selected results from Table 4; Permanently Sustainable Non-Oil Primary Deficit, Percent of non-oil GDP)
- Baseline long-run oil price scenarios: US$30/bbl = 4.8; US$57/bbl = 5.5; US$84/bbl = 7.6.
- Oil reserves:
  - Higher oil reserves (125 percent) = 5.7 / 7.0 / 9.8 (US$30/57/84).
  - Lower oil reserves (75 percent) = 3.8 / 4.0 / 5.2.
- Effective oil tax take z:
  - Higher z = 45.9 percent of oil GDP → 6.5 / 7.4 / 10.0.
  - Lower z = 25.9 percent of oil GDP → 3.1 / 3.6 / 5.1.
- Non-oil tax take τ:
  - Higher τ = 33.9 percent of non-oil GDP → 5.0 / 5.7 / 7.8.
  - Lower τ = 13.9 percent of non-oil GDP → 4.6 / 5.3 / 7.3.
- Real interest rate r:
  - Higher r = 3.5 percent → 6.9 / 7.9 / 10.8.
  - Lower r = 2.5 percent → 2.5 / 2.9 / 4.0.
- Real non-oil growth rate γ:
  - Higher γ = 2.5 percent → 2.5 / 2.9 / 4.0.
  - Lower γ = 1.5 percent → 6.9 / 7.9 / 10.8.
- Habit strength α:
  - No habits α = 0.05 → 5.0 / 5.7 / 7.7.
  - Stronger habits α = 0.8 → 4.7 / 5.4 / 7.5.
  - Weaker habits α = 0.6 → 4.9 / 5.6 / 7.6.

### Extensions and model implications
- Introducing an interest-rate spread between sovereign debt and oil-fund assets:
  - Creates incentive to run a smaller non-oil deficit in the short run and to repay debt sooner.
  - With a 50 basis point spread the optimal path includes more rapid repayment of debt, earlier net asset accumulation, and a higher permanently sustainable fiscal deficit.
- Redefining objective to guarantee stability in real per capita expenditure:
  - Adding population growth lowers net real interest rate (r minus γ and population growth) and lowers the optimal sustainable primary deficit; implies faster adjustment is required.
- Incorporating public investment:
  - If government expenditure includes productive public investment, the initial fiscal deficit could be higher as investment raises future productive capacity.
  - Barnett and Ossowski condition (equation (13)) implies governments should invest in projects that pay for themselves; with tax rate 23.9 percent and interest rate 3 percent, rate of return on public investment would have to exceed 12.6 percent given the simulations’ assumptions.
- Portfolio decision implications:
  - Trade-off between financial assets and physical/social assets; different rates of return on debt, financial, physical, and social investments would affect optimal policy.

### Concluding remarks and policy recommendations
- Three main conclusions:
  - Gabon’s current non-oil primary deficit is not sustainable: permanently sustainable non-oil primary deficit estimated at 5.5 percent of non-oil GDP vs actual 12.5 percent in 2005.
  - Habit formation implies spreading adjustment over three to five years rather than a single abrupt cut.
  - An interest rate spread incentivizes front-loading adjustment to pay off net debt sooner; uncertainty gives a risk-averse policymaker a precautionary motive to front-load adjustment.
- Recommended elements of an economic reform program for Gabon:
  - (i) a phased adjustment of the primary non-oil balance to a permanently sustainable level;
  - (ii) reforms to the management of FGF assets;
  - (iii) a rapid repayment of public debt;
  - (iv) structural reforms aimed at improving the design and quality of public investments.
- Future research directions:
  - Relax assumption that government expenditure is only consumption and has no effect on productivity growth.
  - Model different rates of return on government debt and on financial, physical, and social investment.
  - Endogenize allocation between physical and financial investments in a finite-horizon optimization maximizing national wealth, explicitly modeling the interaction between investment quality and future growth.

*Source: IMF working paper content unit titled "35.9   percent of oil GDP" (excerpt).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06193.pdf_