## 1. The Baseline Scenario

## Source details

**Canonical URL:** [1. The Baseline Scenario](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13138.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13138.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13138.pdf.json)

---

### Context and purpose
- Papua New Guinea (PNG) faces a sizable natural resource boom driven by a major Liquefied Natural Gas (LNG) project; construction peaks in 2012–13 with export receipts and dividend payments beginning approximately in 2015.
- Most LNG revenue inflows accrue to the government; policy choices concern whether to save or spend these resources, composition of spending (tradables vs non-tradables), timing of spending, and how central bank reserve policy and monetary policy respond.
- The chapter uses a model-based analysis to investigate fiscal and monetary policy options and likely macroeconomic impacts.

### Key macroeconomic channels of concern
- Three channels by which a commodity boom may harm development:
  - Unproductive use of natural resource revenue inflows: government boosts consumption without lasting productive effect, crowding out saving or productive public investment.
  - Loss in external competitiveness (Dutch Disease): government consumption raises non-tradable demand and causes real appreciation that reduces tradable sector profitability.
  - Tradable sector’s special role: shrinkage of tradables can lower long-run growth if tradables are the main source of productivity growth.
- This chapter focuses mainly on the second channel (real exchange rate and tradable vs non-tradable sectors) and the role of monetary policy; other issues are noted for future work.

### Model framework (structure and agents) — overview
- Sectors: non-resource tradable (T), non-tradable (N), natural resource/mining (O).
- Households: two types — savers (fraction f) and hand-to-mouth (fraction 1−f).
- Firms: sectoral production with public capital externality (elasticity α_G).
- Natural resource sector: production profile and foreign-financed investment exogenous; resource output exported.
- Government: collects tax receipts and dividends; can spend (consumption or investment), deposit revenues (central bank or offshore SWF).
- Central bank: manages foreign exchange inflows (reserves) and can sterilize via issuing debt; may target an interest rate or money growth rate.
- Key calibration and steady-state inputs (selected):
  - Elasticity of public investment: 0.1 at the 1 year horizon to 0.25 at the 10 year horizon.
  - Labor income share: 0.7.
  - Home bias: 0.3.
  - Learning-by-doing parameter: set relatively low.

### Policy questions framed for scenarios
- Should the government spend or save the revenue inflows?
- If spending, what composition (tradables vs non-tradables) and timing (front-loaded vs smoothed)?
- If saving, what saving vehicle is preferable (central bank reserves vs offshore SWF)?
- Should the central bank accumulate reserves and, if so, should it sterilize?

### Preview of central findings (from scenario simulations)
- Fiscal spending on non-tradables is the principal driver of macroeconomic effects.
- Front-loading expenditures can generate significant macroeconomic volatility, especially when the share of spending on non-tradables is high and there are spending bottlenecks.
- Monetary policy cannot fully undo counter-productive fiscal policy without cost.
- Saving via reserve accumulation will crowd-out private sector investment (when sterilized).
- Saving in an investment fund (SWF) can substantially mitigate adverse macroeconomic effects, at the cost of lower initial domestic development gains.

---

### A. Spending — baseline and mechanism

- Baseline parameterization (Baseline Scenario):
  - Government spends 90 percent of inflows on consumption and 10 percent on investment.
  - Of consumed portion, 70 percent is spent on non-tradables (ν_g = 0.7).
  - Central bank sells entirety of foreign exchange inflows on the forex market and is interest-rate targeting.

- Mechanisms and short-run dynamics (oil price shock → surge in oil revenues; government spends entirety):
  - Demand for non-tradables rises; monopolistic competition causes short-run increases in non-tradable prices and production.
  - Factors reallocate: labor migrates from tradables to non-tradables and investment shifts toward non-tradables.
  - Relative price movements: non-tradable prices increase relative to tradables → real exchange rate appreciates.
  - Tradable sector profitability declines → outward migration of capital and labor from tradables.
  - Current account deficit widens because:
    - Real appreciation makes tradables relatively cheaper → import demand increases.
    - Shrinking domestic tradable production raises the import share of demand.
  - Inflation dynamics: persistent exchange rate appreciation → overall inflation falls → moderate loosening of monetary policy (short-run decrease in real interest rate).

- Evaluation focus:
  - Emphasis on macroeconomic volatility (deviations from baseline) rather than medium-term output levels, because medium-term output depends on sustainability of public capital accumulation.

### B. Inter-sectoral spending profile (tradables vs non-tradables)
- Alternative parameterization:
  - Government devotes 80 percent of spending to tradables (ν_g = 0.2); other assumptions per baseline.

- Results and implications:
  - Smaller increase in non-tradable demand → smaller rises in non-tradable prices and output → smaller migration of labor and capital to non-tradables.
  - Tradable sector retains larger demand share and profitability.
  - Current account still widens, now via direct government spending on imports.
  - Overall inflationary pressure reported as identical to baseline because a smaller nominal appreciation offsets the smaller increase in non-tradable prices.
- Key inference:
  - Resource inflows per se do not cause macroeconomic volatility; government spending on non-tradables does, because non-tradables supply is limited domestically while tradables can be met via imports.

### C. Inter-temporal spending profile (front-loaded vs smoothed)
- Experiment:
  - Compare front-loaded fiscal spending (spend all revenues as received) versus smoothing (spread spending over time).
  - Modified oil price shock case: more temporary — more than 90 percent of the oil price shock dissipates now after three years (versus approximately ten years in prior specification).

- Smoothing implementation:
  - Send initial fiscal surpluses into government deposit with central bank and draw down later to finance deficits.
  - Central bank sells foreign exchange matching government spending financed out of additional oil revenues to keep money supply broadly unchanged.
  - Smooth spending spread evenly over five years.

- Quantified trade-offs and illustrative results:
  - Front-loaded spending produces strong short-run macro pressures and significant macroeconomic volatility.
  - Smoothing reduces volatility and yields more gradual adjustments.
  - Example peak real exchange rate outcomes reported:
    - Real exchange rate appreciation peaks at about 6.5percent under front-loaded spending.
    - Peaks at only about 2.5 percent under smooth spending.
  - Front-loading costs include reallocation, unemployment and retraining costs, and possible permanent losses in tradable sector output if firms fail to restart.
- Policy conclusion:
  - Front-loading spending should be avoided, especially where bottlenecks to effective public spending exist.

- Absorptive capacity costs (Box 1):
  - Rapid spending can create public financial management and implementation bottlenecks; inefficiency of investment modeled as:
    - Effective public investment ̃G^I_t = [1 − b (G^I_t / G^I)^2] G^I_t, b ≥ 0.
  - Higher b reduces effective investment; illustrative comparisons show greater ineffectiveness when spending is rapid.

### D. Monetary policy experiments
- Baseline central bank action: sells all foreign exchange inflows, which contributes to nominal appreciation and hence real appreciation.
- Alternative: Reserve accumulation with sterilization:
  - Central bank retains 70 percent of foreign exchange acquired through resource inflows as foreign reserves (ω_o = 0.7).
  - To sterilize the monetary impact, central bank issues public debt; interest-rate targeting tightened with φ_π = 5.
  - Oil price shock specification reverts to baseline persistent case.

- Results and trade-offs:
  - Reserve accumulation mitigates nominal exchange rate appreciation and limits real appreciation; tradable sector remains more competitive and its production declines much less than under baseline.
  - Reserve accumulation does not significantly reduce demand for non-tradables generated by front-loaded fiscal spending.
  - Sterilization implies monetary tightening → real interest rates increase dramatically → crowding-out of private sector demand (investment and consumption fall).
  - Private sector crowding-out frees supply capacity to meet government demand, but at the cost of reduced private consumption and investment.
  - Core message: monetary policy can protect tradables from appreciation, but not costlessly — sterilization imposes substantial real-economy costs.

---

### E. Saving — reserve accumulation versus Sovereign Wealth Fund (SWF)

- Reserve accumulation (central bank) — mechanism and limits:
  - If the central bank accumulates reserves without sterilization: domestic money supply grows → inflation → real appreciation → return to negative effects of full spending.
  - If sterilized: sterilization crowds out private sector investment.
  - Net assessment: reserve accumulation is at best a partial remedy; sterilization shifts costs to private sector.

- Sovereign Wealth Fund (SWF) design in model and outcomes:
  - Government saves a proportion φ_f of resource revenue inflows in an offshore SWF that accrues interest at an exogenous rate; dividends from SWF are transfers.
  - Government saves enough to raise the fund to a permanently higher level in detrended equilibrium and divides remainder between consumption and investment per baseline (φ_gc = 0.9, φ_gi = 0.1).
  - Macroeconomic consequences:
    - Intertemporal smoothing of spending significantly reduces macroeconomic volatility (similar to the smooth spending scenario).
    - Less front-loaded spending reduces demand pressure on non-tradables and lowers volatility.
    - SWF transforms a temporary revenue surge into a permanent, smaller revenue stream via interest earnings; fiscal spending sustainable and public capital stock does not decay as under the baseline.
    - Monetary policy challenges reduced: with most inflows saved offshore, little rationale to accumulate reserves or to pursue very tight sterilizing monetary policy; fiscal and monetary objectives become more consistent.
  - Net assessment:
    - Smoothing via an SWF can support sustainable fiscal strategies that build human and physical capital while avoiding Dutch Disease pressures, though initial domestic development gains are smaller because part of the inflows are saved.

- Current account and exchange rate distinctions (key mechanics):
  - In both baseline and higher government import scenarios, the current account deficit widens by the same amount matching oil revenue inflows (full absorption).
  - Two-phase absorption:
    1. Government spends the revenue inflows.
    2. Fiscal spending converts into a larger non-oil current account deficit — either via direct government imports (high government import share) or via real appreciation lowering tradable prices and raising import demand.
  - Under tight monetary policy that blocks nominal appreciation, the non-oil current account deficit cannot rise; economy-wide absorption cannot increase, so higher government spending crowds out private sector demand.
  - The nominal exchange rate is set in forex market (central bank reserve management); the real exchange rate is determined by both fiscal and monetary policy. Preventing nominal appreciation alone does not prevent real appreciation unless monetary policy sterilizes reserve effects.

---

### F. Conclusions, limitations, and research agenda

- Key conclusions and policy messages:
  - Fiscal spending on non-tradables is the primary driver of Dutch Disease and macroeconomic volatility.
  - Front-loading expenditures can create substantial macroeconomic volatility, especially with high non-tradable spending shares and absorptive capacity bottlenecks.
  - Monetary policy cannot eliminate counter-productive fiscal policy effects without real economic costs (notably private sector crowding-out when sterilizing).
  - Reserve accumulation as a saving mechanism leads to crowding-out when sterilized and is only a partial remedy.
  - Saving in an offshore investment fund (SWF) can significantly mitigate adverse macroeconomic effects, though it reduces immediate domestic spending and development gains.

- Model limitations and caveats:
  - Assumes frictionless labor mobility; does not model realistic labor migration and skill acquisition delays important in PNG.
  - Lacks a commercial banking sector; in practice, government deposits at commercial banks affect domestic money supply and the Bank of PNG used Central Bank Bills (CBBs) to sterilize liquidity, a process not captured fully here.
  - Model complexity limits ease of adding further institutional detail; results intended for qualitative, not quantitative, forecasting due to parameter uncertainty and data limitations.
  - Complementarity with other tools: complements richer-sector CGE models by focusing on short- to medium-term macro adjustments; work underway to develop a smaller New Keynesian FPAS model for PNG estimation and forecasting.

*Source: IMF working paper chapter "1. The Baseline Scenario" (excerpt from _wp13138).*

### 1. The Baseline Scenario......................................................................................18

### 1. The Baseline Scenario

### Context and purpose
- Papua New Guinea (PNG) is experiencing a sizable natural resource boom driven by a major Liquefied Natural Gas (LNG) project.
- The construction phase peaks in 2012–13, with subsequent earnings from export receipts and dividend payments beginning approximately in 2015.
- Most LNG revenue inflows will accrue to the government; policy choices revolve around whether to save or spend these resources, and the composition and timing of spending.
- The paper investigates the interplay of fiscal and monetary policy options regarding LNG revenue inflows and the likely macroeconomic impact using a model-based analysis.

### Key macroeconomic channels of concern
- Three macroeconomic channels through which a commodity boom may harm economic development:
  - Unproductive use of natural resource revenue inflows: government boosts consumption without lasting productive effect, at the expense of saving or productive public investment.
  - Loss in external competitiveness: increase in government consumption leads to a real appreciation that reduces profitability of the tradable sector (Dutch Disease effects).
  - Tradable sector’s special role: if the tradable sector is the main source of productivity growth, its shrinkage can lower long-term growth prospects.
- The paper focuses mostly on the second channel (real exchange rate and tradable vs non-tradable sectors) and the role of monetary policy; issues related to public investment (first channel) and the tradable–growth linkage (third channel) are noted as covered elsewhere or for future work.

### Model framework (structure and agents)
- Sectors: non-resource tradable sector, non-tradable sector, natural resource sector.
- Agents and institutions: households (savers and hand-to-mouth consumers), firms, government, central bank.
- Households choose consumption and leisure; producers in tradable and non-tradable sectors choose labor and capital inputs.
- Natural resource sector: simpler specification with production profile depending on exogenous foreign investment.
- Government: collects tax receipts and dividends; choices include spending (consumption or investment) or depositing revenues.
  - Deposits can be at the central bank or in a sovereign wealth fund held offshore.
- Central bank: decides how much of foreign exchange inflows to accumulate as reserves (influencing the exchange rate) and whether to sell or buy government securities (influencing interest rate and money growth).
  - Central bank may target an interest rate or a money growth rate.
- Assumptions about LNG revenue flows: revenues are split between foreign investor and government; foreign investor’s share is repatriated and not available to the PNG economy; government collection of taxes and dividends are the only inflow of LNG revenues into the PNG economy in the model.

### Policy questions framed for scenarios
- Should the government spend or save the revenue inflows?
- If it spends the inflows, what should it spend it on? Tradables or non-tradables?
- If it spends the inflows, when should it spend? Now or later? Do the existence of spending bottlenecks affect this decision?
- If it saves, what is the best way of saving?
- Should the central bank accumulate reserves or not? If they do, should they sterilize these reserves?

### Preview of central findings from scenario simulations
- Fiscal spending on non-tradables is what drives the macroeconomic effects.
- Front-loading of expenditures can lead to macroeconomic volatility, especially if the share of spending on non-tradables is high and there are bottlenecks in the economy.
- Monetary policy cannot undo the effect of counter-productive fiscal policy without cost.
- Saving through reserve accumulation will crowd-out private sector investment.
- Saving in an investment fund can considerably mitigate the adverse macroeconomic effects (at the cost of lower initial development gains as part of the revenue inflows are saved and not invested into the economy).

### Modeling lineage and application
- Model roots: framework initially developed for researching macroeconomic implications of scaling up aid inflows (United Nations MDG Africa Steering Group project).
- Expanded to include a natural resource sector and augmented to investigate Dutch Disease effects and public investment efficacy.
- A version of this model has been used previously to explore public investment for natural resource-rich countries; here it is used to examine fiscal spending transmission mechanisms with focus on real exchange rate and monetary policy options.

*Source: IMF working paper chapter "1. The Baseline Scenario".*

### Chapter V concludes.

### _wp13138 - Chapter V concludes.

### The model setup
- Small open New Keynesian economy with a natural resource sector.
- Two household types: savers and hand-to-mouth (fraction f are savers, 1−f are hand-to-mouth).
- Three production sectors: non-resource tradable (T), non-tradable (N), and natural resource/mining (O).
- Non-traded sector: monopolistic competition. Tradable sector: perfect competition.
- Natural resource sector: production profile and prices essentially exogenous; resource output assumed solely for exports.
- Model solved and simulated using Dynare.

### Households
- Savers maximize expected utility:
  - Objective E_t Σ_{t=0}^∞ β^t [u^a(c^a_t,m^a_t) − κ^a (1+ψ^a)^{-1} (l^a_t)^{1+ψ^a}].
  - Budget constraint: (1+τ^c_t) c^a_t + m^a_t + b^a_t + s_t b^{*a}_t + (v/2) s_t (b^{*a}_t − b^{*a})^2 = (1−τ^l_t) w_t l^a_t + m^a_{t−1}/π_t + i_{t−1} b^a_{t−1}/π_t + s_t i^* b^{*a}_{t−1}/π^*_t + Ω^T_t + Ω^N_t + s_t rm^* + z_t.
  - Utility u^a(c^a_t,m^a_t) = (1/(1−σ)) { [ θ^a (c^a_t)^{(η−1)/η} + (1−θ^a) (m^a_t)^{(η−1)/η} ]^{η/(η−1)} }^{1−σ}.
  - Composite consumption: c^a_t = [ φ^{1/χ} (c^{aN}_t)^{(χ−1)/χ} + (1−φ)^{1/χ} (c^{aT}_t)^{(χ−1)/χ} ]^{χ/(χ−1)}.
  - Consumption price index: P_t = [ φ (P^N_t)^{1−χ} + (1−φ) (P^T_t)^{1−χ} ]^{1/(1−χ)}.
  - Law of one price: P^T_t = S_t P^* (P^* constant). Real exchange rate s_t ≡ S_t P^*/P_t.
  - Savers hold real money m_t, domestic bonds b^a_t, foreign assets b^{*a}_t; portfolio adjustment cost v/2 s_t (b^{*a}_t − b^{*a})^2 with steady-state b^{*a}.
  - Labor aggregation across sectors with elasticity of substitution ρ and steady-state share δ:
    l^a_t = [ δ^{−1/ρ} (l^{aN}_t)^{1+1/ρ} + (1−δ)^{−1/ρ} (l^{aT}_t)^{1+1/ρ} ]^{ρ/(1+ρ)}.
  - Real wage index: w_t = [ δ (w^N_t)^{1+ρ} + (1−δ) (w^T_t)^{1+ρ} ]^{1/(1+ρ)}.

- Hand-to-mouth households:
  - Same intratemporal utility form but intratemporal optimization only (myopic), budget constraint:
    (1+τ^c_t) c^h_t + m^h_t = (1−τ^l_t) w_t l^h_t + m^h_{t−1}/π_t + s_t rm^* + z_t.

### Firms and production structure
- Three sectors: non-traded (N), traded non-resource (T), natural resource (O).
- Natural resource sector uses capital-intensive technology; labor abstracted from resource production.

### Non-traded good sector
- Firm i production: y^N_{it} = z^N (k^N_{it−1})^{1−α_N} (l^N_{it})^{α_N} (K^G_{t−1})^{α_G}.
  - Public capital K^G_t enters with output elasticity α_G; constant returns to scale in private inputs and increasing returns with public capital.
- Private capital law of motion:
  n k^N_{it} = (1−δ_N) k^N_{it−1} + [ 1 − κ_N/2 ( i^N_{it} / i^N_{it−1} − 1 )^2 ] i^N_{it}, with κ_N ≥ 0 (investment adjustment cost).
- Firm maximizes net present-value profit weighted by savers' marginal utility λ^a_t, with distortions parameter ι and corporate tax τ^k_t. After-tax profit Ω^N_t = (1−τ^k_t) ∫ ̃Ω^N_{it} di, where ̃Ω^N_{it} defined in text and includes price adjustment cost term with parameter ς.

### Traded good sector and learning-by-doing
- Traded sector production: y^T_{it} = z^T_t (k^T_{it−1})^{1−α_T} (l^T_{it})^{α_T} (K^G_{t−1})^{α_G}.
- Productivity law: ln z^T_t = ρ_{zT} ln z^T_{t−1} + d ln y^T_{t−1} (learning-by-doing externality).
  - Note: a shrinkage of the tradable sector due to real appreciation from resource inflows reduces productivity growth in tradables; the learning parameter set relatively small in this paper.
- Capital law of motion analogous to non-traded sector with κ_T adjustment cost.
- Firm profits and after-tax profit Ω^T_t defined analogously; traded sector profits include scale factor s_t.

### Natural resource sector
- World resource price relative to foreign CPI: p^{O*}_t; relative to domestic CPI: s_t p^{O*}_t.
- Resource price AR(1): ˆp^{O*}_t = ρ_{po} ˆp^{O*}_{t−1} + ε^{po}_t, ε^{po}_t ∼ i.i.d. N(0, σ^2_{po}).
- Resource investment financed by FDI; FDI process: ˆi^{f*}_t = ρ_f ˆi^{f*}_{t−1} + ε^f_t, ε^f_t ∼ i.i.d. N(0, σ^2_f).
- Resource investment in domestic consumption prices: I^O_t = s_t i^{f*}_t.
- Resource capital law: n K^O_t = (1−δ_O) K^O_{t−1} + I^O_t.
- Resource output CES technology: Y^O_t = z^O_t [ α_O^{1/ξ} (K^O_{t−1})^{(ξ−1)/ξ} + (1−α_O)^{1/ξ} (s_t X^*_t)^{(ξ−1)/ξ} ]^{ξ/(ξ−1)} with ξ elasticity of substitution.
- Resource productivity AR(1): ˆz^O_t = ρ_{zo} ˆz^O_{t−1} + ε^{zo}_t, ε^{zo}_t ∼ i.i.d. N(0, σ^2_{zo}).
- Government royalties τ^o_t levied on production quantity; resource sector net profit in dollars:
  Ω^{O*}_t = (1−τ^o_t) p^{O*}_t Y^O_t − s_t X^*_t.
- Total resource revenues collected each period:
  T^O_t = s_t [ τ^o p^{O*}_t y^O_t + ι_{div} Ω^{O*}_t + (i^* − 1) F^*_ {t−1} / π^* ].
- Nominal GDP in domestic consumption prices:
  Y_t = p^N_t Y^N_t + s_t Y^T_t + s_t p^{O*}_t Y^O_t.

### Fiscal policy and sovereign wealth fund (SWF)
- Government flow budget constraint (nominal):
  T_t + B_t − (D^G_t − D^G_{t−1}/π_t) + s_t A^* + s_t (F^*_{t−1}/π^*) = p^g_t G_t + Z_t + (i_{t−1} − 1) B^c_t / π_t + B_{t−1} / π_t + s_t F^*_t.
- Total receipts T_t = τ^c_t C_t + τ^l_t w_t L_t + τ^k_t (Ω^N_t + Ω^T_t) + T^O_t.
- Public investment time-to-build: if X^GI_t is authorized investment with spending rates g_0,...,g_{k−1}, then executed public investment G^I_t = Σ_{j=0}^{k−1} g_j X^GI_{t−j}, Σ g_j = 1.
- Absorptive capacity/delivery inefficiency:
  log ̃G^I_t / ̃G^I = b × log G^I_{t−1} / G^I, 0 ≤ b ≤ 1 (b governs severity; b > 0 when G^I_t < G^I).
- Public capital law: K^G_t = (1−δ_G) K^G_{t−1} + ϵ ̃G^I_t, with inefficiency ϵ < 1.
- Government consumption and investment composition rules for an unsaved windfall (T^O_t − T^O):
  - First stage: fraction φ_f of windfall saved in SWF; SWF law:
    F^*_t = (1 − 1/π^*) F^* + 1/π^* F^*_{t−1} + 1/s_t [ φ_f (T^O_t − T^O) ].
  - Second stage: unsaved share (1 − φ_f) allocated by fixed shares φ_{gc}, φ_{gi}:
    G^C_t = G^C + (1/p^G_t) (1 − φ_f) φ_{gc} (T^O_t − T^O).
    G^I_t = G^I + (1/p^G_t) (1 − φ_f) φ_{gi} (T^O_t − T^O).
- All other fiscal variables (tax rates, transfers, debt) held at steady-state levels in this specification.

### Reserve and monetary policy
- Central bank balance sheet identity:
  m_t − m_{t−1}/π_t + D^G_t − D^G_{t−1}/π_t = B^{cb}_t − B^{cb}_{t−1}/π_t + s_t (R^*_t − R^*_{t−1}/π^*).
- Reserve policy:
  R^*_t / R^* = (R^*_{t−1} / R^*)^{ρ_R} (T^{O*}_t / T^{O*})^{ω_o} (π^S_t / π^S)^{−ω_s}, where ω_o indicates share of oil windfall and nominal exchange rate flexibility; ω = 0 indicates a flexible exchange rate.
- Monetary policy (Taylor-type rule):
  i_t / i = (i_{t−1} / i)^{ρ_i} [ (Y_t / Y)^{φ_y} (π_t / π)^{φ_π} (μ_t / μ)^{φ_μ} ]^{(1−ρ_i)}, where μ_t = nM_t / M_{t−1} / π_t is nominal money growth.

### Aggregation, identities, and external accounts
- Aggregation across household types: X_t = f x^a_t + (1−f) x^h_t for x ∈ {c, c^N, c^T, m, l, rm^*, b^*, b^c}.
- Total government bonds: B_t = B^c_t + B^{cb}_t.
- Total demand for non-traded goods:
  D^N_t = φ [ C^N_t + I^N_t + I^T_t + ς/2 (π^N_{it}/π^N_{it−1} − 1)^2 p^N_t y^N_t ] + ν (p^G_t)^χ G_t.
- Market clearing for non-traded goods: Y^N_t = (p^N_t)^{−χ} D^N_t.
- Current account deficits:
  CA^d_t = [ C_t + I_t + p^G_t G_t + ζ/2 (π^N_t / π^N_{t−1} − 1)^2 p^N_t y^N_t ] − p^N_t y^N_t − s_t y^T_t − s_t p^{O*} y^O_t − s_t (i^* − 1) B^*_ {t−1} + F^*_ {t−1} / π^* − s_t rm^*.
- Balance of payments condition specified in full in text including portfolio adjustment cost v/2 (B^*_t − B^*)^2 and FDI and dividend terms.

### Calibration (overview and key numeric instances)
- Parameters partitioned into technology, preference, and national accounts steady-state values.
- Labor income share set to 0.7.
- Elasticity of output with respect to public capital calibrated based on regression results cited (Arslanalp et al. (2010)); further calibration details follow Berg et al. (2010) (omitted here).

*Source: _wp13138 - Chapter V concludes.*

### 0.1 at the 1 year horizon to 0.25 at the 10 year horizon. Here, the calibration uses an elasticity

### IV. POLICY SIMULATIONS

### Model calibration and steady state
- Elasticity used for public investment: 0.1 at the 1 year horizon to 0.25 at the 10 year horizon.
- Rationale for elasticity choice:
  - Intended forecast horizon of the model is approximately one to five years.
  - Capacity constraints in PNG regarding planning and implementation of public investment argue for a choice of a lower-than-average elasticity.
- Learning-by-doing parameters set to a relatively low value to capture long-run Dutch Disease effects; more detailed analysis deferred to future work.
- Capital account closed via high portfolio adjustment costs reflecting Bank of PNG staff assessment that portfolio flows are not very sensitive to interest rate differentials.
- Preference and structural parameters:
  - Home bias: 0.3 (computed from national accounts data reflecting a small share of the non-tradable sector in total output).
  - Substitution elasticities: taken from Berg et al. (2010) and Berg et al. (2012b); Frisch elasticity for hand-to-mouth consumers calibrated to minimize labor supply response to income variations.
  - Money demand elasticity calibrated to prevent interest rate fluctuations from becoming unrealistically large.
- Steady state values:
  - Calculated using GDP data from the National Statistics Office of PNG until 2006.
  - Absence of published GDP data since 2006: use Ministry of Treasury official estimates thereafter.
  - For 2011 data currently unavailable, the nominal growth rates from 2010 were applied.
  - Steady state values computed as the average over 2000–06 to represent “normal” conditions prior to the commodity price boom in the second half of the 2000’s.

### A. Spending — scenarios and key mechanisms
- Central question: should the government spend or save the resource revenues? This section analyzes scenarios where the government spends the entirety of the revenues.
- Government consumption focus: macroeconomic effects driven largely by government spending, particularly on non-tradables; public investment has different medium-term effects by building public capital.

Subsection: 1. The Baseline Scenario
- Baseline parameterization:
  - Government spends 90 percent of inflows on consumption and 10 percent on investment.
  - Of consumed portion, 70 percent is spent on non-tradables (ν_g = 0.7).
  - Central bank policy: sells entirety of foreign exchange inflows on the forex market and is interest-rate targeting.
- Mechanisms and results (oil price shock):
  - Increase in oil prices → surge in oil revenues → government spending rises in line with inflows (government spends entirety).
  - Demand for non-tradables increases → monopolistic competition causes short-run increases in non-tradable prices and production.
  - Production increase requires factors of production → labor migration from tradables and increased investment in non-tradables.
  - Non-tradable prices increase relative to tradable prices → real exchange rate appreciates (in absence of intervention).
  - Real appreciation reduces profitability in tradable sector → outward migration of capital and labor.
  - Current account deficit widens because:
    - Real appreciation makes tradables relatively cheaper → import demand increases.
    - Shrinking domestic tradable sector increases import share of demand.
  - Persistent exchange rate appreciation → overall inflation falls → moderate loosening of monetary policy (decrease in real interest rate) in short run.
- Evaluation metric:
  - Focus on macroeconomic volatility (deviations of macro variables from baseline/zero line) rather than medium-term output levels, because medium-term output depends on sustainability of public capital accumulation.
  - Lower volatility deemed preferable; measure qualitatively via impulse response deviation magnitude and frequency.

Subsection: 2. Inter-sectoral Spending Profile
- Alternative: government devotes 80 percent of spending to tradables (ν_g = 0.2); other assumptions per baseline.
- Results and implications:
  - Smaller increase in non-tradable demand → smaller increases in non-tradable prices and output → smaller migration of labor and capital to non-tradables.
  - Tradable sector retains larger share of demand, experiences smaller price drop, retains profitability.
  - Current account still widens, now via direct government spending on imports.
  - Overall inflationary pressure identical to baseline because smaller nominal appreciation offsets smaller increase in non-tradable prices.
- Key inference: Resource inflows per se do not cause macroeconomic volatility; government spending on non-tradables does, because supply of non-tradables is limited domestically whereas tradables can be met via imports.

Subsection: 3. Inter-temporal Spending Profile
- Compare front-loaded fiscal spending vs smooth spending:
  - Modified oil price shock: more temporary — more than 90 percent of the oil price shock dissipates now after three years (versus approximately ten years under previous specification).
- Front-loaded scenario:
  - Government spends all oil revenues as they come in → large ramp-up of expenditures during price surge, rapid reduction as prices fall.
- Smooth spending alternative:
  - Spending of temporary rise spread evenly over five years; requires saving part of temporary oil revenues and drawing them down later.
  - Implementation in model: transfer initial fiscal surpluses into government deposit with central bank, draw down later to finance deficits.
  - Foreign exchange management modified: central bank sells foreign exchange matching amount of government spending financed out of additional oil revenues → buildup of government deposits matched with buildup of foreign exchange reserves → overall money supply broadly unchanged.
  - Smoothing implemented both fiscally and on foreign exchange sales; nominal exchange rate made somewhat sticky.
- Results and trade-offs:
  - Smooth spending → more sustained, gradual adjustment; front-loaded spending → strong short-run macro pressures and significant macroeconomic volatility.
  - Quantified example:
    - Real exchange rate appreciation peaks at about 6.5percent under front-loaded spending; peaks at only about 2.5 percent under smooth spending.
  - Front-loading costs:
    - Macroeconomic volatility: larger swings in inflation, nominal interest rates, and output.
    - Economic costs not captured by the model: massive reallocation of labor → periods of unemployment and retraining costs; possible permanent output losses in tradable sector as some firms fail to restart.
  - Conclusion: Front-loading spending should be avoided because of macroeconomic volatility and economic costs; costs would be more pronounced if bottlenecks to effective public spending exist.

Box 1: Absorptive Capacity Costs
- Rapid spending can create bottlenecks in public financial management and implementation; higher volume of public spending increases inefficiency of investment.
- Modeling absorptive capacity costs:
  - Effective public investment ̃GIt = [1 − b (GIt / GI)^2] GIt, b ≥ 0.
  - Greater b → smaller ratio ̃GIt / GIt → less effective public investment.
- Illustrative results compare small b (solid line) vs large b (dashed line); results are illustrative and parametric choices are ad-hoc.

Subsection: 4. Monetary Policy
- Question: Can monetary policy mitigate Dutch Disease effects (real appreciation and tradable sector decline) from rapid government spending on non-tradables?
- Baseline: central bank sells all foreign exchange inflows → key factor for nominal appreciation and hence real appreciation.
- Alternative: Reserve accumulation with sterilization:
  - Central bank retains 70 percent of foreign exchange acquired through resource inflows as foreign reserves (ω_o = 0.7).
  - To offset monetary expansion, the central bank issues public debt to sterilize the reserve accumulation.
  - Interest rate targeting regime made stricter in response to inflation: φ_π = 5.
  - Oil price shock specification reverts to baseline persistent case.
- Results:
  - Reserve accumulation mitigates nominal exchange rate appreciation and keeps real appreciation limited.
  - Tradable sector remains largely competitive; production drops but by far less than under baseline.
  - From fiscal perspective, reserve accumulation does not significantly affect demand for non-tradables; supply response to front-loaded spending still generated.
  - How non-tradable demand is met shifts:
    - Baseline: met via factor migration facilitated by real appreciation.
    - Reserve accumulation scenario: met via direct government spending on imports is blocked; central bank reserve management blocks first channel and government spending profile may block second.
  - Sterilization implies monetary tightening → real interest rates increase dramatically → crowding-out of private sector demand.
    - Private sector demand falls → frees up supply capacity to meet expanding government demand.
    - Side effects: reduction in private consumption → households increase labor supply → larger short-run GDP increase and significant rise in non-tradable production.
  - Trade-off: monetary policy can protect tradable sector from real appreciation but does so at the cost of private sector crowding-out; no costless monetary neutralization of front-loaded fiscal effects.

### B. Saving — introduction to the saving scenarios
- Motivation: Given macroeconomic volatility from spending (especially front-loaded), examine scenarios where the government saves a portion of resource inflows.
- Saving will be analyzed with distinctions such as reserve accumulation versus investment in a sovereign wealth fund (further scenarios follow beyond the supplied text).

*Source: Excerpt from IMF working paper (policy simulation section).*

### 1.  Reserve Accumulation

### _wp13138 - 1.  Reserve Accumulation

### Reserve accumulation: mechanism and macroeconomic effects
- Reserve accumulation by the central bank can be interpreted as a form of saving when the government spends resource earnings and the central bank holds the remainder.
- If the central bank accumulates reserves without sterilization:
  - Growth in the domestic money supply and ensuing increase in inflation will cause a real appreciation.
  - This returns the economy to the negative macroeconomic effects witnessed under full spending.
- If the central bank sterilizes the reserve accumulation:
  - Sterilization will eventually crowd out private sector investment.
- Net assessment:
  - Reserve accumulation as a saving mechanism is at best a partial remedy to the macroeconomic volatility produced by full spending.

### The Sovereign Wealth Fund (SWF): design and outcomes
- SWF operation as specified:
  - The government saves a proportion of resource revenue inflows in an offshore SWF that accrues interest at an exogenous rate.
  - Dividend payments from the SWF are made in the form of transfers.
  - The government saves enough to enable the fund to reach a permanently higher level in the detrended equilibrium and divides the remainder between consumption and investment as per the baseline (φ_gc = 0.9, φ_gi = 0.1) and φ_f = nπ∗−1 i∗−1).
  - All other assumptions follow the baseline scenario.
- Macroeconomic consequences of the SWF:
  - Intertemporal smoothing of spending results in much reduced macroeconomic volatility, similar to the smoothing in Scenario 3.
  - Less overall front-loaded spending reduces demand on non-tradables and further lowers macroeconomic volatility (as in Scenario 2).
  - The SWF effectively transforms a temporary increase in oil revenues into a permanent, smaller revenue stream fed by interest earnings on SWF balances.
  - Fiscal spending under the SWF scenario is sustainable; the public capital stock will not wither away as it does under the baseline.
  - Monetary policy challenges are much reduced:
    - With most oil revenue inflows saved offshore and limited macroeconomic pressure from moderate increases in spending, there is little rationale to accumulate the remaining foreign exchange inflows as reserves or to pursue a very tight monetary policy.
    - Monetary and fiscal objectives are consistent, allowing fiscal expansion without triggering a monetary policy response that crowds out the private sector.
- Net assessment:
  - Smoothing fiscal spending through an SWF can support a fiscal strategy that builds human and physical capital sustainably, calibrates spending to avoid Dutch Disease effects, and avoids macroeconomic pressures that would jeopardize monetary policy objectives.
- Institutional/design note:
  - The SWF design follows the bird-in-hand rule and creates a unit root in the accumulation process.

### Further issues: current account dynamics and exchange rate distinctions
- Current account (absorption) dynamics:
  - In both the baseline scenario and the scenario with higher government imports, the current account deficit widens by the same amount; this amount matches the oil revenue inflows precisely (full absorption occurs in both cases).
  - Two-phase absorption process:
    1. Government fully spends the revenue inflows (occurs in both cases).
    2. Fiscal spending must be converted into an increase of the non-oil current account deficit (i.e., all spending ultimately converts into imports).
       - With a high government import share, conversion happens directly through government imports.
       - In the baseline, conversion happens via real appreciation making imports cheaper and a decline in domestic tradable production raising import demand.
  - In Scenario 4 (tight monetary policy blocking nominal appreciation), monetary policy blocks both channels; the non-oil current account deficit cannot rise:
    - Under the baseline, increased non-oil current account deficit allows economy-wide absorption (consumption and investment) to increase so government spending need not crowd out private sector demand.
    - Under tight monetary policy, economy-wide absorption cannot increase, so higher government consumption and investment must reduce private sector consumption and investment (private sector crowding-out).
- Real versus nominal exchange rate — policy roles:
  - The nominal exchange rate is set in the forex market and is controlled by the central bank’s reserve management policy.
  - The real exchange rate is influenced by both fiscal and monetary policy:
    - Expansionary fiscal policy tends to lead to real appreciation that can arise through a nominal appreciation or through inflation.
    - Preventing nominal appreciation (via reserve accumulation) does not by itself prevent real appreciation.
    - To prevent real appreciation, monetary policy must sterilize the impact of reserve buildup on the money supply — a tightening of the monetary stance.
  - Reserve management policy alone does not determine the path of the real exchange rate; the combination of reserve management policy and monetary policy (sterilization) determines the real exchange rate path.

### Conclusions and policy messages
- Key conclusions from the scenario simulations:
  - Fiscal spending on non-tradables is what drives the macroeconomic effects.
  - Front-loading of expenditures can lead to macroeconomic volatility, especially if the share of spending on non-tradables is high and there are bottlenecks in the economy.
  - Monetary policy cannot undo the effect of counter-productive fiscal policy without cost.
  - Saving through reserve accumulation will crowd-out private sector investment.
  - Saving in an investment fund (SWF) can considerably mitigate adverse macroeconomic effects (at the cost of lower initial development gains because part of the revenue inflows are saved and not invested into the domestic economy).

### Model limitations and research agenda
- Shortcomings and caveats in the model:
  - Assumes frictionless labor mobility; does not realistically model labor migration and skill acquisition delays (important in PNG given low education and skill levels).
  - Lacks a commercial banking sector; in PNG, most government deposits are held in commercial banks rather than the central bank:
    - Empirically, government deposits at commercial banks have had a significant impact on the domestic money supply.
    - The Bank of PNG issued significant volumes of Central Bank Bills (CBBs) to sterilize excess liquidity, contradicting the model’s prediction that sterilization should be unnecessary if resource revenues are saved.
  - Given model complexity, adding further sectors or institutional detail could reduce interpretability.
  - The model should be used for qualitative analysis only, not quantitative forecasts, due to parameter calibration uncertainty and lack of required data for estimation.
- Fit within PNG modeling toolkit and research priorities:
  - Complements existing CGE models (Bank of PNG and Department for National Planning and Monitoring) that have richer sectoral structure but medium- to long-term horizons; this model provides short- to medium-term adjustment analysis.
  - Ongoing research to develop a smaller New Keynesian FPAS model that can be estimated for PNG to produce quantitative and qualitative forecasts; the models will be complementary.
  - Research agenda goals:
    - Use the present model to investigate other policy questions (e.g., Dutch Disease effects with a large value of the learning-by-doing parameter).
    - Complete and operationalize the FPAS model as a complement to the financial programming framework for PNG.

*Source: _wp13138 - 1.  Reserve Accumulation, IMF working paper chapter content provided.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13138.pdf_