## _wp13144

## Source details

**Canonical URL:** [_wp13144](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13144.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13144.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13144.pdf.json)

---

### I. INTRODUCTION — objectives, model, and scenarios
- Objective:
  - Examine implications of scaled-up public investment to boost growth while maintaining fiscal sustainability in a developing country with diminishing oil reserves.
  - Questions: (i) under what conditions can departures from a policy of consumption smoothing be justified; and (ii) what are the implications of such departures on growth and fiscal sustainability?
- Method:
  - Dynamic stochastic general equilibrium (DSGE) model featuring a public investment–growth nexus and explicit debt dynamics.
  - Fiscal policy: rule-based (automatic tax and expenditure responses to economic activity) with allowance for discretionary shocks; tailored to Cameroon’s fiscal and growth challenges.
- Calibration period and assumptions:
  - Model calibrated to Cameroon and simulates three policy scenarios in 2012–32.
  - Oil price forecast range: US$95–US$110 a barrel.
  - Oil reserves assumed to be depleted by 2032.
- Scenarios simulated:
  - “scaling-up with reform”:
    - Public investment increases steadily to 12 percent of GDP by 2032 (from 6.5 percent of GDP in 2012).
    - Investment financed through oil revenue, and additional borrowing if necessary as the oil endowment depletes overtime.
    - Assumes improvement in efficiency of public investment to the median middle-income countries’ level.
  - “conservative”:
    - Public investment remains broadly similar to its 2012 level in terms of GDP.
    - Assumes similar improvements in public investment efficiency as in the “scaling-up with reform” scenario.
    - Expects tighter financial constraints during 2012–32 due to projected decline in oil production and limited fiscal space.
  - “scaling-up no reform”:
    - Public investment path similar to the “scaling-up with reform” scenario but assumes no change in public investment efficiency relative to its historical level.

### Key modeling distinctions and contributions
- Reaction-function approach vs. Permanent Income Hypothesis (PIH):
  - PIH is purely forward-looking with present value of resource wealth as the binding constraint on spending.
  - The model incorporates constraints facing government by looking at past government behavior and includes the need to build fiscal buffers; explicitly tracks a fiscal policy reaction function with explicit public debt dynamics.
- Emphasis on public investment efficiency:
  - Quantifies importance of public capital efficiency in determining growth returns of public infrastructure.
  - Public investment inefficiencies are an important channel for transmission of the “resource curse.”
  - Public investment efficiency parameter in the “scaling-up with reform” scenario is set at 0.54.

### Main simulation findings (summary)
- “Scaling-up with reform” results:
  - Real GDP growth averages 5.1 percent a year in 2012–32, which is 2.3 percentage points higher than the average in 2006–11.
  - Cameroon’s large infrastructural needs could justify deviation from the PIH approach and support gradual scaling up of public investment provided effectiveness is simultaneously enhanced.
- Other scenarios:
  - The “conservative” and “scaling-up no reform” scenarios yield smaller growth gains and worse fiscal outcomes (in terms of the NOPD and debt).
- General conclusion:
  - Efficiency matters for growth; low-income, natural resource-rich countries could benefit from using resource revenue to scale up public investment, but simultaneous improvements in effectiveness are needed to achieve sustainable and higher growth rates.

### II. LITERATURE REVIEW — context and critiques of PIH/BIH
- PIH:
  - Recommends smoothing primary expenditure over time, setting it equal to permanent income.
  - Operational implication: save resource revenue early and run it down as resources deplete; emphasizes stabilization and intergenerational equity.
- BIH (Bird-in-Hand):
  - More conservative than PIH; use resource-related revenue to accumulate financial assets and use only the yield to finance expenditure; Norway as example.
- Critiques and extensions:
  - Both PIH and BIH treat all government spending as consumption and fail to distinguish investment vs. consumption and their growth effects in credit-constrained economies.
  - Uncertainty over determination of permanent income undermines PIH implementation.
  - Optimal use of resource revenue depends on factors beyond permanent income, including returns on public investment and administrative capacity.
  - Administrative capacity and public investment quality can undermine benefits from scaling up public capital spending.

### III. CAMEROON: experience, fiscal rules, and public investment issues
- Historical performance and reserves:
  - Moderate growth with limited improvement in socioeconomic indicators over last 30 years; contraction between 1977 and 1986.
  - Proven oil reserves relatively small; economy more diversified than neighboring oil-producing countries.
- Oil-sector and fiscal statistics (2010):
  - Oil production: 63,800 barrels a day.
  - Oil exports: about half of total exports.
  - Oil revenue: 27 percent of total government revenue.
  - Comparative (2010) Nigeria: over 2.4 million barrels a day; oil exports 96 percent of total exports; oil revenue 70 percent of total government revenue.
- Fiscal framework:
  - CEMAC balanced budget rule for general government (adopted 1996) and debt rule (adopted 2002) limiting debt-to-GDP ratio to 70 percent.
  - Since 2000, Cameroon failed to meet the balanced-budget criterion only once and has met the debt rule since debt relief in 2006.
- Weak link between oil-financed public investment and growth:
  - Explanations: “Dutch disease” effects and poor governance leading to unproductive capital spending.
  - Estimates suggest 54 percent of early oil revenue cannot be accounted for.
  - Controlling for efficiency reduces by more than half the capital stock suggested by cumulative investment in some studies.

### Public investment efficiency and implementation
- Observations:
  - Low public investment efficiency translates into lower public capital stocks.
  - Public investment budgets not fully spent because of implementation problems.
- Causes cited:
  - Low administrative capacity and recurring governance problems.
- Institutional responses:
  - New ministry for public procurement set up in 2012.
  - Preliminary Public Investment Management Index (PIMI) shows Cameroon trailing the SSA average.
  - World Bank PEFA (2007) and CPIA (2010) corroborate weak public investment management.

### Infrastructure bottlenecks and investment needs
- Estimates and potential gains:
  - To reach the infrastructure level of Africa’s middle-income countries, Cameroon would need to invest an additional US$350 million a year (Dominguez-Torres and Foster (2011)).
  - Potential efficiency gains from reforming infrastructure-related utilities estimated at $586 million annually.
  - If such increases in quantity and quality of public investment were sustained over 15 years, Cameroon would enjoy an increase in its long-run per capita GDP growth rate of about 3.3 percentage points.

### Key findings and policy message
- Investment case:
  - A case can be made for Cameroon to invest relatively more of its natural resource income in the near future than the PIH approach would recommend.
- Effects of scaling-up with reforms:
  - Stronger non-oil growth, crowding in of private investment, and reduction in the non-oil primary deficit (NOPD).
- Risks of scaling-up without efficiency gains:
  - Growth gains limited; debt and NOPD outcomes worsen.
- Transparency:
  - Cameroon is not EITI compliant as of February 2013.
- Benchmarks:
  - Historical studies and cross-country evidence cite optimal public investment in SSA between 8 and 11 percent.

### Model framework (structure and assumptions)
- Model type and agents:
  - Standard neoclassical growth framework with two homogeneous goods (oil and non-oil) and four agent types (households, firms, government, rest of the world).
- Households:
  - Ramsey framework; utility depends on private consumption and government consumption; no leisure; infinite labor supply at marginal cost; can borrow/lend at world interest rate.
- Government:
  - Owns oil resources; provides public goods; sets public investment I_g,t and current spending G_t.
  - Fiscal stance governed by NOPD (percent of GDP) and debt ratio d_t.
- Firms:
  - Produce non-oil good using private and public capital; public capital nonexcludable and nonrival; private and public investment complementary.
- Oil:
  - Extracted at a predetermined and declining rate; production follows an AR(1)-type process; oil expected to be depleted by end of 2032.
- Fiscal rules and constraints:
  - NOPD feedback rule includes lagged NOPD, lagged debt objective d_{t-1}, and expected non-oil output at budget decision.
  - Debt dynamics link debt ratio to NOPD, growth, shocks, inflation, interest rate, and oil tax.
  - FIN (fiscal buffer) augments NOPD reaction, capturing discretionary cash deposits and time inconsistency/uncertainty.
- Public capital productivity:
  - Productive efficiency of public capital captured by φ (0 < φ < 1); φ links actual public capital stock to effective productive public capital.
- Oil endowment and depletion:
  - F is proven oil endowment; T is time horizon for depletion (T=21 assuming depletion by 2032 with 2011 base year; note elsewhere T=20 used in a parameter table footnote).

### Calibration and estimation
- Data and filtering:
  - Annual Cameroonian economy, 1980–2011.
  - Trend extraction: Christiano and Fitzgerald filter for GDP, consumption, investment, debt, inflation, primary deficit.
- Estimation:
  - Bayesian methods and maximum likelihood; Metropolis-Hastings MCMC with 100,000 iterations and 10,000 burn-in; convergence checks with Brooks and Gelman (1998).
- Priors and parameter settings (preserved exactly as in source):
  - Discount factor β set at 0.96.
  - Intertemporal elasticity of substitution η set at 1.1 (prior).
  - Share of private capital in output γ set at 30 percent (prior).
  - Efficiency index of public capital φ initially set at 0.45.
  - Average tax rates on production: oil ≈ 27 percent; non-oil ≈ 21 percent (five-year averages 2007–11).
  - Average nominal interest rate on public debt r prior mean 0.035; estimated posterior mean 0.0334.
  - AR shocks prior mean autocorrelation 0.80; prior standard errors: 0.5 percent for oil price and production, 0.25 percent for others.
- Selected posterior means (preserved exactly):
  - η posterior mean 2.3261 (MCSE 0.0118).
  - ν (degree of waste of current public spending) posterior mean 0.1924 (MCSE 0.9838).
  - γ posterior mean 0.4507 (MCSE 0.0008).
  - φ posterior mean 0.5389 (MCSE 0.0927).
  - Κ posterior mean 0.2313 (MCSE 0.0018).
  - Δ posterior mean 0.1169 (MCSE 0.0044).
  - θ posterior mean 0.7131 (MCSE 0.0007).
  - ω posterior mean -0.1753 (MCSE 0.0013).
  - σ posterior mean -0.0057 (MCSE 0.2131).
  - τ_o posterior mean 0.3108 (MCSE 0.2731).
  - τ_no posterior mean 0.2412 (MCSE 0.0059).
  - ν (FIN(t), fiscal buffer parameter) posterior mean 0.0661 (MCSE 1.4444).
- MCMC details:
  - Runs of 100,000 iterations; discard first 10,000 as burn-in.

### Policy scenarios and main simulation results (2012–32)
- Common assumptions:
  - Oil prices assumed in the range of US$95–US$110 a barrel.
  - Oil production expected to be depleted by end of 2032.
  - Public investment in scaling-up scenarios increases to 10 percent of GDP by 2032 (from 6.5 percent in 2012).
- Scenario A — “Scaling-Up with Reform”:
  - Public investment path: increases to 10 percent of GDP by 2032.
  - Private investment: 13.1 percent of GDP in 2012; increases to 21 percent of GDP by 2032.
  - Efficiency index φ used: implied φ = 0.54.
  - Outcomes:
    - Higher public investment with improved efficiency crowds in private investment and sustains higher non-oil growth.
    - NOPD declines continuously from an initial large deficit (about 6.5 percent of GDP).
    - Debt-to-GDP rises as oil revenue declines but stabilizes at about 26 percent of GDP by 2029.
    - Uncertainty large for projected oil production and debt-to-GDP; relatively narrow error bands for non-oil GDP growth, NOPD, and non-oil sector growth.
- Scenario B — “Conservative”:
  - Public investment increases moderately: 0.04 percentage point a year on average.
  - φ maintained at a similar level to the scaling-up with reform scenario.
  - Outcomes:
    - More limited spillovers to private sector; lower non-oil growth compared with scaling-up with reform.
    - Debt ratio worsens relative to scaling-up with reform scenario.
    - NOPD is higher than in scaling-up with reform scenario.
- Scenario C — “Scaling-Up No Reform”:
  - Public investment increases to 10 percent of GDP by 2032 but no improvement in governance/effectiveness; φ = 0.45.
  - Outcomes:
    - No crowding-in of private sector.
    - Lower non-oil GDP growth.
    - Higher NOPD and higher ratios of debt to GDP and debt to non-oil GDP relative to reform scenario.
    - At φ = 0.45, long-run non-oil real growth appears bounded below 2.5 percent as public investment reaches 10 percent of GDP.

### Quantitative milestones and notable numeric values
- 3.3 percentage points (estimated increase in long-run per capita GDP growth from sustained quantity and quality investment over 15 years).
- Cameroon is not EITI compliant as of February 2013.
- Optimal public investment in SSA reported between 8 and 11 percent.
- Discount factor β = 0.96.
- Intertemporal elasticity of substitution (prior) = 1.1; posterior η = 2.3261.
- Share of private capital in output γ prior = 0.30000; posterior mean = 0.4507.
- Efficiency index φ: prior 0.45; posterior mean 0.5389; scaling-up with reform implied φ = 0.54; scaling-up no reform φ = 0.45.
- Public investment: 6.5 percent of GDP in 2012; increases to 10 percent of GDP by 2032 in scaling-up scenarios.
- Private investment: 13.1 percent of GDP in 2012; increases to 21 percent of GDP by 2032 under scaling-up with reform.
- Oil prices assumed US$95–US$110 per barrel.
- Oil expected to be depleted by end of 2032.
- Debt-to-GDP stabilizes at about 26 percent of GDP by 2029 under scaling-up with reform.
- Initial large NOPD about 6.5 percent of GDP.
- MCMC: 100,000 iterations, 10,000 burn-in.

### VII. CONCLUSION AND POLICY IMPLICATIONS
- Core findings:
  - Both the size and effectiveness of public investment financed through oil revenue matter for increasing growth in Cameroon.
  - Effect operates through efficiency in increasing public capital stocks; complementarity with private investment is key.
  - Results conditional on efficiency "closer to that observed in the average middle-income country."
  - Scenario without reforms tends to exert a dampening effect on growth and deteriorates fiscal performance over time.
- Policy implications and recommendations:
  - Plan a sustainable medium-term fiscal strategy that accommodates a gradual scaling up of public investment financed through oil revenue.
  - If existing fiscal buffers do not suffice, policymakers could borrow to finance investment, provided policy reforms effectively enhance investment quality.
  - Structural reforms to enhance public investment efficiency and fiscal performance—examples: enhanced tax base, better tax collection, including user fees.
  - Fiscal anchors that target a medium-term debt ratio and NOPD could incorporate flexibility to respond to uncertainty.
  - Maintain a comfortable level of government reserves as a buffer against exogenous shocks given oil price volatility.
  - Conclusions rest on the assumption that issues relating to public investment (budgetary process, size, and effectiveness) are addressed swiftly.

### APPENDIX: methodological notes and selected indicators
- Appendix: first-order conditions derived via Lagrange method for intra- and intertemporal optimization; notation includes C_t, K_t, and λ_t.
- Selected empirical indicators (Cameroon, 1980–2010) covered:
  - Real Non–Oil GDP Growth; Real Oil GDP Growth; Non–Oil Primary Deficit (Percent of nonoil GDP); Public Debt (Log, millions of US$); Public Investment (Log, billions of CFA F, 2000=100); Constant GDP (Log, billions of CFA F, 2000=100); Non–Oil Inflation; Oil Inflation (US$ per barrel).
  - Data sources: International Financial Statistics 2012; authors' calculations and estimations.

*Source: IMF working paper content unit _wp13144 (excerpts provided).*

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

### _wp13144 - References .............................................................................................................

### I. INTRODUCTION — objectives, model, and scenarios
- Objective: examine implications of scaled-up public investment to boost growth while maintaining fiscal sustainability in a developing country with diminishing oil reserves; questions: (i) under what conditions can departures from a policy of consumption smoothing be justified; and (ii) what are the implications of such departures on growth and fiscal sustainability?
- Method: dynamic stochastic general equilibrium (DSGE) model featuring a public investment–growth nexus and explicit debt dynamics; fiscal policy is rule-based (automatic tax and expenditure responses to economic activity) but allows discretionary shocks; model tailored to Cameroon’s fiscal and growth challenges with particular attention to the evolution of public debt and the fiscal policy reaction function.
- Calibration period and assumptions: model calibrated to Cameroon and simulates three policy scenarios in 2012–32; during this period, the oil price is forecast to be between US$95–110 a barrel and oil reserves are assumed to be depleted by 2032.
- Scenarios simulated:
  - “scaling-up with reform”:
    - public investment increases steadily to 12 percent of GDP by 2032 (from 6.5 percent of GDP in 2012).
    - investment financed through oil revenue, and additional borrowing if necessary (as the oil endowment depletes overtime).
    - assumes improvement in efficiency of public investment relative to Cameroon’s historical average; efficiency raised to the median middle-income countries’ level; structural impediments lifted so a greater share of public investment translates into effective capital stock.
  - “conservative”:
    - public investment remains broadly similar to its 2012 level in terms of GDP.
    - assumes similar improvements in the efficiency of public investment as in the “scaling-up with reform” scenario.
    - expects tighter financial constraints during 2012–32 due to projected decline in oil production, limited fiscal space, limited possibility to tap international financial market, and prospective financial difficulties associated with declining oil revenue.
  - “scaling-up no reform”:
    - public investment path similar to the “scaling-up with reform” scenario but assumes no change in public investment efficiency relative to its historical level (i.e., lower efficiency than in the other two scenarios).

### Key modeling distinctions and contributions
- Reaction-function approach vs. Permanent Income Hypothesis (PIH):
  - PIH is purely forward-looking with present value of resource wealth as the binding constraint on spending.
  - Model’s reaction-function approach incorporates constraints facing government by looking at past government behavior and includes the need to build fiscal buffers; explicitly tracks a fiscal policy reaction function with explicit public debt dynamics.
- Emphasis on public investment efficiency:
  - quantifies importance of public capital efficiency in determining growth returns of public infrastructure.
  - public investment inefficiencies are an important channel for transmission of the “resource curse.”
  - public investment efficiency parameter in the “scaling-up with reform” scenario is set at 0.54.

### Main simulation findings (summary)
- “Scaling-up with reform” results:
  - real GDP growth averages 5.1 percent a year in 2012–32, which is 2.3 percentage points higher than the average in 2006–11.
  - suggests that Cameroon’s large infrastructural needs could justify deviation from the PIH approach and support gradual scaling up of public investment provided effectiveness is simultaneously enhanced.
- Other scenarios:
  - the “conservative” and “scaling-up no reform” scenarios yield smaller growth gains and worse fiscal outcomes (in terms of the NOPD and debt).
- General conclusion: efficiency matters for growth; low-income, natural resource-rich countries could benefit from using resource revenue to scale up public investment, but efforts are needed to enhance effectiveness of public investment simultaneously to achieve sustainable and higher growth rates.

### II. LITERATURE REVIEW — context and critiques of PIH/BIH
- PIH (Permanent Income Hypothesis):
  - recommends smoothing primary expenditure over time, setting it equal to permanent income.
  - operational implication: save resource revenue in early years and run it down as resources deplete; policy prescriptions emphasize stabilization (avoid procyclical policies) and intergenerational equity.
- BIH (Bird-in-Hand):
  - more conservative than PIH; requires using resource-related revenue to accumulate financial assets and use only the yield to finance expenditure.
  - provides more predictable expenditure flows and stronger stabilization; Norway cited as prominent BIH example.
- Critiques and extensions:
  - Both PIH and BIH treat all government spending as consumption and fail to distinguish investment vs. consumption and their growth effects in credit-constrained economies.
  - Uncertainty over determination of permanent income undermines implementation of PIH.
  - Recent literature argues for integrating return on public investment and fiscal sustainability; optimal use of resource revenue depends on factors beyond permanent income, including spread between current and long-run resource price, capital scarcity, external indebtedness, returns on public investment, and administrative capacity.
  - Administrative capacity and public investment quality can undermine benefits from scaling up public capital spending; adjustment costs and governance constraints can make redistribution preferable in some cases.

### III. CAMEROON: experience, fiscal rules, and public investment issues
- Historical performance and reserves:
  - Cameroon experienced moderate growth with limited improvement in socioeconomic indicators over last 30 years; economic activity contracted sharply between 1977 and 1986.
  - Cameroon’s proven oil reserves are relatively small and the economy is more diversified than neighboring oil-producing countries.
- Oil-sector and fiscal statistics:
  - In 2010, Cameroon’s oil production amounted to 63,800 barrels a day.
  - In 2010, oil exports were about half of total exports, while oil revenue was 27 percent of total government revenue.
  - By comparison (2010), Nigeria produced over 2.4 million barrels a day; oil exports were 96 percent of total exports; and oil revenue was 70 percent of total government revenue.
- Fiscal framework and historical use of oil revenue:
  - oil revenue was saved abroad in early years of production and may have been partly used to weather the 1985–88 economic crisis.
  - public investment tracked oil sector revenue relatively closely in 1980–2011; since 1991, oil revenue has been consistently higher than capital expenditure, implying oil revenue was also used to finance current spending or reserve building.
  - Cameroon, as a CEMAC member, is subject to an explicit balanced budget rule for general government (adopted in 1996) and a debt rule (adopted in 2002) limiting debt-to-GDP ratio to 70 percent.
  - Since 2000, Cameroon failed to meet the balanced-budget criterion only once and has met the debt rule since debt relief in 2006.
- Weak link between oil-financed public investment and growth:
  - explanations include “Dutch disease” effects and poor governance leading to unproductive capital spending.
  - estimates suggest a relatively large amount of early oil revenue (54 percent) cannot be accounted for.
  - public investment inefficiency is highlighted as directly relevant: controlling for efficiency reduces by more than half the capital stock suggested by cumulative investment in some studies; inefficient investment and weak institutions can explain why investment fails to spur growth in oil-producing CFA-zone countries.

*Source: _wp13144 - References (excerpts from the PDF chapter/section provided).*

### 3.5 percent, compared to 10 percent in the Central African Economic and Monetary

### _wp13144 - 3.5 percent, compared to 10 percent in the Central African Economic and Monetary

### Public investment efficiency and implementation
- Public investment in Cameroon shows low efficiency, translating into lower public capital stocks.
- Public investment budgets were also not being fully spent because of implementation problems.
- Causes cited include low administrative capacity and recurring governance problems.
- A new ministry for public procurement was set up in 2012 in response to these issues.
- Preliminary results from the Public Investment Management Index (PIMI) show Cameroon trailing the SSA average.
- The World Bank’s 2007 Public Expenditure and Financial Accountability (PEFA) evaluation and the 2010 Country Policy and Institutional Assessment (CPIA) corroborate the finding of weak public investment management.

### Infrastructure bottlenecks and investment needs
- Lack of adequate infrastructure remains a key bottleneck to Cameroon’s economic growth.
- Dominguez-Torres and Foster (2011) estimate that to reach the infrastructure level of Africa’s middle-income countries, Cameroon would need to invest an additional US$350 million a year.
- The same authors estimate potential efficiency gains from reforming infrastructure-related utilities at $586 million annually.
- Taken together, if such an increase in both the quantity and quality of public investment were to be sustained over 15 years, Cameroon would enjoy an increase in its long-run per capita GDP growth rate of about

*Source: _wp13144 - 3.5 percent, compared to 10 percent in the Central African Economic and Monetary*

### 3.3 percentage points. A case can thus be made for Cameroon to invest relatively more of its

### _wp13144 - 3.3 percentage points. A case can thus be made for Cameroon to invest relatively more of its

### Key findings and policy message
- A case can be made for Cameroon to invest relatively more of its natural resource income in the near future than the PIH approach would recommend (statement opening the content).
- Public investment scaling-up combined with reforms (higher efficiency of public capital) yields stronger non-oil growth, crowds in private investment, and helps reduce the non-oil primary deficit (NOPD).
- If public investment is increased without improving investment efficiency, growth gains are limited and debt and NOPD outcomes worsen.
- Cameroon is not EITI compliant as of February 2013 (transparency has improved but compliance absent).
- Historical studies and cross-country evidence cited: optimal public investment in SSA between 8 and 11 percent (Fosu, Getachew, and Ziesemer (2011)).

### Model framework (structure and assumptions)
- Model type: standard neoclassical growth framework with two homogeneous goods (oil and non-oil) and four agent types (households, firms, government, rest of the world).
- Households: Ramsey framework; utility depends on private consumption and government consumption; no leisure; infinite labor supply at marginal cost; can borrow/lend at world interest rate.
- Government: owns oil resources; provides public goods; sets public investment I_g,t and current spending G_t; fiscal stance governed by NOPD (percent of GDP) and debt ratio d_t.
- Firms: produce non-oil good using private and public capital; public capital is nonexcludable and nonrival; private and public investment complementary.
- Oil: extracted at a predetermined and declining rate; oil production follows an AR(1)-type process (equation (11)); oil expected to be depleted by end of 2032.
- Fiscal rules and constraints:
  - NOPD feedback rule (equation (15)) includes lagged NOPD, lagged debt objective d_{t-1}, and expected non-oil output at budget decision.
  - Debt dynamics captured by equation (16), linking debt ratio to NOPD, growth, shocks, inflation, interest rate, and oil tax.
  - FIN (fiscal buffer) augments the NOPD reaction, capturing discretionary cash deposits and time inconsistency/uncertainty in budget process (equation (18)).
- Productive efficiency of public capital captured by φ (0 < φ < 1); φ links actual public capital stock to effective productive public capital.
- Oil endowment and depletion: F is proven oil endowment; T is time horizon for depletion (T=21 assuming depletion by 2032 with 2011 base year; elsewhere a note indicates T=20 years was used in a parameter table footnote).

### Calibration and estimation
- Data: annual Cameroonian economy, 1980–2011.
- Trend extraction: Christiano and Fitzgerald filter for GDP, consumption, investment, debt, inflation, primary deficit.
- Estimation: Bayesian methods and maximum likelihood; Metropolis-Hastings MCMC with 100,000 iterations and 10,000 burn-in; convergence checks with Brooks and Gelman (1998).
- Priors and parameter settings (selected, preserved exactly as in source):
  - Discount factor β set at 0.96.
  - Intertemporal elasticity of substitution η set at 1.1 (following Takizawa, Gardner, and Ueda (2004)).
  - Share of private capital in output γ set at 30 percent (and Bayesian/estimated means reported elsewhere).
  - Efficiency index of public capital φ initially set at 0.45 (Hulten (1996) baseline).
  - Average tax rates on production: oil ≈ 27 percent; non-oil ≈ 21 percent (five-year averages 2007–11).
  - Average nominal interest rate on public debt r set around 0.035 (prior mean) and estimated posterior mean 0.0334.
  - AR shocks prior mean autocorrelation set at 0.80; prior standard errors: 0.5 percent for oil price and production, 0.25 percent for others.
- Posterior estimation specifics (selected posterior means from Table 2, preserved exactly):
  - η (Elasticity of intertemporal substitution) posterior mean 2.3261 (MCSE 0.0118).
  - ν (Degree of waste of current public spending) posterior mean 0.1924 (MCSE 0.9838).
  - γ (Share of private capital in output) posterior mean 0.4507 (MCSE 0.0008).
  - φ (Efficiency index of public capital) posterior mean 0.5389 (MCSE 0.0927).
  - Κ (Unexplained determinant of oil production) posterior mean 0.2313 (MCSE 0.0018).
  - Δ (Depreciation of capital stock) posterior mean 0.1169 (MCSE 0.0044).
  - θ (Initial deficit effect) posterior mean 0.7131 (MCSE 0.0007).
  - ω (Initial debt effect) posterior mean -0.1753 (MCSE 0.0013).
  - σ (Cyclical effect) posterior mean -0.0057 (MCSE 0.2131).
  - τ_o (Tax on oil production) posterior mean 0.3108 (MCSE 0.2731).
  - τ_no (Tax on non-oil production) posterior mean 0.2412 (MCSE 0.0059).
  - ν (FIN(t), fiscal buffer parameter) posterior mean 0.0661 (MCSE 1.4444).
- MCMC details: runs of 100,000 iterations, discard first 10,000 as burn-in; rejection rate and diagnostics applied.

### Policy scenarios and main simulation results (2012–32)
- Common modeling assumptions across scenarios:
  - Oil prices assumed in the range of US$95–US$110 a barrel (aligned with IMF World Economic Outlook projections).
  - Oil production expected to be depleted by end of 2032.
  - Public investment in scaling-up scenarios increases to 10 percent of GDP by 2032 (from 6.5 percent in 2012).
- Scenario A — “Scaling-Up with Reform”:
  - Public investment path: increases gradually to 10 percent of GDP by 2032.
  - Private investment: increases from 13.1 percent of GDP in 2012 to 21 percent of GDP by 2032.
  - Efficiency index φ used: implied efficiency index φ = 0.54 (Bayesian estimate).
  - Outcomes:
    - Higher public investment with improved efficiency crowds in private investment and sustains higher non-oil growth.
    - NOPD declines continuously from an initial large deficit (about 6.5 percent of GDP).
    - Debt-to-GDP rises as oil revenue declines but stabilizes at about 26 percent of GDP by 2029.
    - Model predictions for 2012 broadly consistent with IMF World Economic Outlook (2012) data; narrow confidence intervals on many variables.
    - Uncertainty large for projected oil production and debt-to-GDP; relatively narrow error bands for non-oil GDP growth, NOPD, and non-oil sector growth.
- Scenario B — “Conservative”:
  - Public investment increases moderately: 0.04 percentage point a year on average (slightly higher than historical average).
  - φ maintained at a similar level to the scaling-up with reform scenario.
  - Outcomes:
    - More limited spillovers to private sector; lower non-oil growth compared with scaling-up with reform.
    - Debt ratio worsens relative to the scaling-up with reform scenario.
    - NOPD is higher than in the scaling-up with reform scenario.
- Scenario C — “Scaling-Up No Reform”:
  - Public investment increases to 10 percent of GDP by 2032 but no improvement in governance/effectiveness; φ = 0.45.
  - No crowding-in of private sector.
  - Outcomes:
    - Lower efficiency index associated with lower non-oil GDP growth.
    - Higher NOPD and higher ratios of debt to GDP and debt to non-oil GDP relative to reform scenario.
    - At φ = 0.45, long-run non-oil real growth appears bounded below 2.5 percent as public investment reaches 10 percent of GDP.

### Quantitative milestones and notable numeric values (preserved exactly)
- 3.3 percentage points (phrase opening the content).
- Cameroon is not EITI compliant as of February 2013.
- Optimal public investment in SSA range reported: between 8 and 11 percent.
- Discount factor β = 0.96.
- Intertemporal elasticity of substitution = 1.1 (prior); posterior η = 2.3261.
- Share of private capital in output γ prior = 0.30000; posterior mean = 0.4507.
- Efficiency index φ: initial prior 0.45; Bayesian posterior mean 0.5389; implied φ used in scaling-up with reform = 0.54; scaling-up no reform φ = 0.45.
- Public investment: 6.5 percent of GDP in 2012; increases to 10 percent of GDP by 2032 in scaling-up scenarios.
- Private investment: 13.1 percent of GDP in 2012; increases to 21 percent of GDP by 2032 under scaling-up with reform.
- Oil prices assumed in range US$95–US$110 per barrel.
- Oil expected to be depleted by end of 2032.
- Debt-to-GDP stabilizes at about 26 percent of GDP by 2029 under scaling-up with reform.
- Initial large NOPD about 6.5 percent of GDP.
- MCMC: 100,000 iterations, 10,000 burn-in.
- Posterior parameter values (selected) reproduced exactly in "Calibration and estimation" section above.

*Italicized source: IMF working paper content unit _wp13144 (extracted text provided).*

### 2032. This scenario also shows that higher public debt appears to exert a dampening effect on

### _wp13144 - 2032. This scenario also shows that higher public debt appears to exert a dampening effect on

### VII. CONCLUSION AND POLICY IMPLICATIONS
- Model and setup:
  - Small-scale DSGE model applied to Cameroon with features: (i) exhaustible and volatile oil revenue; (ii) the impact of investment on non-oil growth; and (iii) consideration of explicit debt dynamics and a fiscal reaction function.
- Core findings:
  - Both the size and effectiveness of public investment financed through oil revenue matter for increasing growth in Cameroon.
  - The effect of public investment on growth operates through its efficiency in increasing public capital stocks.
  - Complementarity with private investment appears to be the key transmission channel.
  - Results are conditional on an assumed level of efficiency "closer to that observed in the average middle-income country."
  - The scenario without reforms tends to exert a dampening effect on growth and deteriorates fiscal performance over time.
  - The scenario run to illustrate the risk of lack of progress in governance indicates that poor governance may make the use of natural resource revenue insufficient to boost growth and provide better social indicators. (Footnote 38)
- Policy-relevant environment described:
  - Severe infrastructure bottlenecks, limited fiscal buffers, declining oil reserves, and uncertain access to financing.
- Policy implications and recommendations:
  - Plan a sustainable medium-term fiscal strategy that accommodates a gradual scaling up of public investment financed through oil revenue.
  - If existing fiscal buffers do not suffice, policymakers could borrow to finance investment, provided that policy reforms are effective for enhancing investment quality.
  - Structural reforms should enhance public investment efficiency and fiscal performance; examples given include enhanced tax base and better tax collection, including user fees.
  - Fiscal anchors that target a medium-term debt ratio and NOPD could incorporate sufficient flexibility to respond to uncertainty.
  - Maintain a comfortable level of government reserves as a buffer against exogenous shocks, given volatility in oil prices.
  - The framework is designed to ensure a sustainable level of non-oil production to anchor medium- and long-term sustainability, consistent with IMF policy advice for low-income, resource-rich countries.
  - These conclusions rest on the crucial assumption that issues relating to public investment (its budgetary process and its size) and its effectiveness are addressed swiftly to secure growth dividends.

### APPENDIX: FIRST ORDER CONDITIONS (methodological notes)
- Approach:
  - Making the relevant substitutions in equation (2), and maximizing the objective function described in (1), yields intra- and intertemporal conditions for private consumption given by first-order conditions in equations (A.1.)–(A.3.).
  - Notation: C_t, K_t, and the Lagrange operator λ_t (or shadow price) solve the representative problem in (1) and satisfy the first order conditions.
  - The productivity of public capital is given by (text continues in source).
- Methodological remark:
  - Focus is on first-order conditions for optimum solutions based on the Lagrange method; the knowledge of the value is not necessary in solving the dynamic optimization problem. (Footnote 39)

### Selected empirical and descriptive indicators (high-level)
- Time series and indicators covered for Cameroon, 1980–2010 (as presented in Figure 5 and continued):
  - Real Non–Oil GDP Growth (Log constant, 2000 = 100): Trend, Level, Cyclical component.
  - Real Oil GDP Growth (Log constant, 2000 = 100): Trend, Level, Cyclical component.
  - Non–Oil Primary Deficit (Percent of nonoil GDP): Trend, Level, Cyclical component.
  - Public Debt (Log of debt level, millions of US$): Trend, Level, Cyclical component.
  - Public Investment (Log of billions of CFA F, 2000=100): Trend, Level, Cyclical component.
  - Constant GDP (Log of billions of CFA F, 2000=100): Trend, Log level, Cyclical component.
  - Non–Oil Inflation (Log of Index, 2000=100): Trend, Level, Cyclical component (right scale).
  - Oil Inflation (Log of price per barrel, US$): Trend, Level, Cyclical component.
- Data sources noted:
  - International Financial Statistics 2012; and authors' calculations and estimations.

*Italic: IMF working paper content from _wp13144 (selected sections provided).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13144.pdf_