## 1. Sub-Saharan Africa: Selected Economic Indicators, annual averages, 1980-2005

## Source details

**Canonical URL:** [1. Sub-Saharan Africa: Selected Economic Indicators, annual averages, 1980-2005](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08121.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08121.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08121.pdf.json)

---

### Introduction and theoretical lenses
- Competing views on price determination:
  - Quantity Theory of Money (QTM): fiscal deficits cause inflation via money creation; “inflation is always and everywhere a monetary phenomenon” (Sargent and Wallace, 1981).
  - Fiscal Theory of the Price Level (FTPL): fiscal dominance can arise when fiscal policy is unsustainable and nominal government bonds are net wealth; price level may be determined by fiscal policy (Woodford, Leeper, Sims, Cochrane).
- FTPL mechanism emphasized:
  - Persistent budget deficits raise nominal public debt perceived as higher nominal wealth.
  - Debt deflation (price-level increase eroding real value of public debt) can restore solvency under fiscal dominance.
- Empirical objective: assess plausibility of Ricardian (monetary-dominant) vs. non-Ricardian (fiscal-dominant) regimes across a large sample of SSA countries.

### Data, sample, and scope
- Time span: 1980–2005.
- Datasets:
  - Comprehensive fiscal and monetary outcomes for 22 major SSA countries (with domestic debt).
  - Broader comparisons include up to 42 SSA countries.
- Country grouping (42-country sample):
  - (i) CFA group: 14 CFA countries and the Union of Comoros.
  - (ii) non-CFA-fixed: Botswana, Cape Verde, Lesotho, Namibia, Seychelles, Swaziland (de facto fixed exchange rate throughout period).
  - (iii) non-CFA-floating: 22 countries characterized by independently fixed or floating regimes (many shifted to de facto floating in mid-1990s).
- Exchange rate regime classification sources: IMF de facto classification (1990–2005); Levy-Yeyati and Sturzenegger (2003) de facto classification (1980–1990).

### Stylized facts: inflation, fiscal outcomes, and monetary aggregates (1980–2005)
- Budget balance and inflation (34-country sample, 1980–2004):
  - OLS regression of annual average inflation on average central government balance (including grants) as percent of GDP:
    - Result: significant positive correlation; “a 1 percentage point increase in the surplus-to-GDP ratio is associated with an almost 2 percent reduction in inflation.”
    - R-squared = 0.14.
- Cross-group averages and contrasts:
  - Fixed-exchange-rate groups (CFA and non-CFA fixed):
    - On average lower inflation rates, less seigniorage, higher primary surpluses, more success achieving price stability and single-digit inflation.
  - Non-CFA-floating group:
    - On average double-digit inflation and volatile prices.
    - Average inflation at 26 percent (excluding Angola and DRC which had hyperinflation) — “was four times higher than in the fixed-exchange rate group.”
    - Within group, only Ethiopia, Mauritius, and Rwanda had on average single-digit inflation.
    - Nine countries experienced hyperinflation or average rates above 30 percent (Angola, DRC, Zimbabwe, Ghana, Guinea-Bissau before joining the CFA in 1997, Mozambique, Sierra Leone, Uganda, Zambia).
- Public liabilities-to-GDP (sum of public debt and reserve money):
  - CFA countries: averaged about 125 percent of GDP.
  - Non-CFA floating countries: averaged about 106 percent of GDP.
  - Non-CFA fixed countries: averaged less than 60 percent of GDP.
- Currency composition and domestic debt markets:
  - CFA zone: public debt mostly foreign-currency-denominated; limited domestic debt markets until recently — limited ability to inflate away foreign-currency liabilities.
  - Non-CFA floating countries: more developed domestic government debt markets; average nominal domestic debt ratio rose from 11 percent of GDP in the 1980s to 15 percent in the late 1990s; median rose from 4 percent to 10 percent.
  - Angola, Botswana, DRC, Mozambique, and São Tomé and Príncipe had no domestic government debt markets until 2004.
  - Number of countries with debt-to-GDP ratios exceeding 20 percent rose from three at the beginning of 1980 to nine by 2000.
- Seigniorage and inflation tax:
  - Seigniorage measured as annual change in reserve money as percent of GDP (Fischer, 1981).
  - Seigniorage averaged 1.7 percent of GDP in non-CFA floating countries, almost twice as large as in CFA countries, where the average was below 1 percent of GDP.
  - Inflation tax:
    - CFA fixed countries: average less than 1 percent.
    - Non-CFA fixed countries: averaged about 3 percent.
    - Non-CFA floating countries: very large; even after excluding Angola and the DRC, the average annual rate was about 350 percent (as reported).
    - Examples: Ghana, Sierra Leone, Uganda, and Zambia had very large inflation tax; in Uganda and Zambia it averaged above 2,000. Malawi, Nigeria, and Tanzania: inflation tax rates ranged between 30 and 50 percent where inflation was more moderate but deficits were chronic.

### Model overview and key mechanisms (small open-economy adaptation)
- Model features tailored to SSA:
  - Two-country international setup with single consumption good; one country represents SSA economy receiving grants and facing a domestic borrowing constraint.
  - Explicit borrowing constraints on domestic debt to capture thin capital markets and recourse to seigniorage.
  - No borrowing constraints on foreign-currency–denominated debt (concessional donor/IFI financing).
  - Cash-in-advance (CIA) constraint for households: money required for private consumption and government purchases; government purchases not subject to CIA.
  - Logistic approximation of borrowing-constraint indicator to allow tractable analysis; as γ → ∞ logistic approximates a binding indicator.
- Key theoretical implications:
  - When domestic debt is below the borrowing limit, the present-value government budget constraint reduces to the standard Ricardian form.
  - When debt hits the domestic borrowing constraint:
    - Country can borrow only abroad; domestic borrowing infeasible; the model term associated with the logistic function becomes extreme.
    - To satisfy the present-value constraint, a higher real primary surplus (including seigniorage) is required, or the price level must adjust (domestic debt deflation) to restore solvency.
  - Fiscal-dominant (FD) regimes: surpluses are exogenous and price-level adjustment (inflation deflating nominal debt) may be the equilibrating mechanism.
  - Monetary-dominant (MD) regimes: primary surpluses adjust to liabilities and monetary policy can anchor expectations and determine the price level.

### Econometric strategy and identification of regimes
- Four complementary empirical approaches (nonstructural VAR and cointegration):
  1. Dynamic relationship between public liabilities and primary surpluses (CCD methodology): test how fiscal authorities respond to ensure solvency.
     - Identification: under MD, a positive shock to current surplus leads to a fall in future liabilities; under FD, surpluses are exogenous and future liabilities may be unresponsive or rise.
     - Surplus persistence criterion: positive autocorrelation up to at least 5 lags → persistent.
  2. Interaction between fiscal and monetary variables (pass-through analysis): variance decomposition in a VAR ordered nominal domestic debt growth → reserve money growth → real output gap → inflation.
     - If inflation forecast error largely explained by nominal debt growth shocks → evidence consistent with FTPL/fiscal dominance (wealth effect of nominal debt).
     - If largely explained by money growth shocks → evidence of monetary accommodation / QTM channel.
  3. Active monetary authority test (policy responsiveness): VAR identification ordering real output gap → reserve money growth (or discount rate) → inflation.
     - Identification criterion: money growth should respond negatively (or discount rate positively) to a positive inflation innovation to indicate active monetary policy (monetary dominance).
     - If money growth increases (or discount rate decreases) after an inflation shock, monetary authority is passive/accommodating.
  4. Johansen cointegration analysis (Ricardian equivalence test): cointegration between lagged outstanding public liabilities and primary surplus (including grants and seigniorage) in percent of GDP supports Ricardian equivalence (CI(1,1) stationary linear combination).
- VAR estimation specifics:
  - Stationarity tested by augmented Dickey-Fuller (ADF) with constant and trend (five lags starting point).
  - Lag length chosen by likelihood ratio, Akaike and Schwarz criteria.
  - VARs estimated in levels and first differences; stability via roots of characteristic polynomial.
  - Impulse responses and variance decompositions via Cholesky decomposition; Monte Carlo bootstrap with 1,000 iterations for standard deviation bands.
- Panel VAR (PVAR) estimated for the entire SSA panel and three subgroups (pooling and group-mean estimators considered).
- Sample for single-country VARs: unbalanced panel of 22 SSA countries with at least 17 and at most 25 years of observations (1980–2005). For domestic debt pass-through, 18 SSA countries with at least 16 years (excluding Botswana, Mali, Senegal, Togo).

### Main empirical findings
- Regime identification (22-country sample, CCD approach; Table 4 summary):
  - 9 countries had an identified regime throughout the sample:
    - FD regimes (4): Botswana, Burundi, Tanzania, Zimbabwe.
    - MD regimes (5): Cameroon, Kenya, Nigeria, Rwanda, South Africa.
  - Remaining 13 countries: no regime identified (12 had negatively correlated surpluses; Zambia had positively correlated surpluses but inconclusive impulse responses).
- Ricardian equivalence (Johansen cointegration results):
  - Results robust to Haug (1991) except Kenya and South Africa where MD regimes had no cointegrating relationship.
  - Zimbabwe (FD) had one cointegrating relationship: fiscal authorities matched growing liabilities via increases in reserve money/seigniorage rather than primary surplus.
- Pass-through (variance decomposition; Table 5 averages for ten periods):
  - Inflation variability mostly explained by nominal domestic debt growth in 5 countries: Burundi, Lesotho, Mauritius, South Africa, Zambia.
  - Inflation variability mostly explained by money growth in 6 countries: Cameroon, Kenya, Malawi, Sierra Leone, Tanzania, Zimbabwe.
  - Both determinants important in 5 countries: Ethiopia, Nigeria, Seychelles, Swaziland, Uganda.
  - Selected country shares (examples):
    - Burundi: 18% of inflation variability due to nominal domestic public debt; 5% due to nominal reserve money growth.
    - Cameroon: 29% (debt); 40% (money).
    - Zimbabwe: 28% (debt); 49% (money).
    - Average, whole sample: 19% (debt) and 20% (money).
  - Ghana and Rwanda: both variables have very low explanatory power for inflation variability (large standard errors).
- Active monetary authority tests (robustness on FD identification; Table 6):
  - For CCD-identified FD countries (Botswana, Burundi, Tanzania, Zimbabwe), tests generally reject that monetary authorities tightened reserve money or raised discount rates in response to inflationary shocks—except Botswana where results differ.
- Stylized facts consistent with empirical results:
  - Fixed-exchange regimes (including CFA) associated with lower inflation, lower seigniorage, and more stable price outcomes.
  - Non-CFA fixed countries with domestic-currency nominal sovereign debt experienced higher average inflation than CFA countries.
  - Nominal public liabilities (money or nominal debt) matter for price stability; wealth effects of nominal debt can increase inflation variability, consistent with fiscal-dominant mechanisms.

### Policy implications and recommendations
- Monetary–fiscal coordination is crucial for price stability, especially where public financial management is weak and monetary institutions are inadequate.
- Exchange rate regime choice affects inflation and fiscal–monetary interactions:
  - Fixed-exchange-rate arrangements can help achieve lower and more stable inflation but limit ability to inflate away foreign-currency–denominated liabilities.
  - In countries with underdeveloped domestic debt markets, seigniorage and inflation tax remain important fiscal financing risks.
- Identifying whether a country operates under fiscal or monetary dominance informs appropriate institutional and policy responses:
  - Under FD, restoring solvency may require fiscal consolidation (higher primary surpluses including seigniorage), external borrowing, or price-level adjustment—each with different economic costs.
  - Under MD, strengthening monetary anchors and ensuring primary-surplus responsiveness to liabilities supports price-level determinacy.
- Reduce reliance on implicit fiscal revenues (seigniorage and inflation tax) by developing domestic debt markets, improving tax systems, and strengthening fiscal institutions to limit fiscal dominance risks.

### Limitations and suggestions for further research
- Limited public debt data for many SSA countries reduces statistical power.
- CCD and nonstructural VAR approaches cannot identify within-sample regime switching (alternation between FD and MD); regime-switching VAR techniques would be appropriate for time-varying identification.
- Nonstructural VARs cannot distinguish capacity versus willingness to honor fiscal obligations; structural dynamic general equilibrium assumptions could aid identification.
- Further work should analyze heterogeneity by inflation level (hyperinflation vs. moderate inflation) and consider additional institutional and financial-depth variables.

*Source: _wp08121 - 1. Sub-Saharan Africa: Selected Economic Indicators, annual averages, 1980-2005*

### 1. Sub-Saharan Africa: Selected Economic Indicators, annual averages, 1980-2005 .............25

### 1. Sub-Saharan Africa: Selected Economic Indicators, annual averages, 1980-2005

### Introduction: Competing views on fiscal vs. monetary determination of prices
- Classical QTM (Sargent and Wallace, 1981): fiscal deficits cause inflation because governments resort to money creation; "inflation is always and everywhere a monetary phenomenon."
- Fiscal Theory of the Price Level (FTPL) (Woodford, Leeper, Sims, Cochrane): fiscal dominance (non-Ricardian regimes) can arise when fiscal policy is not sustainable and government bonds are considered net wealth; in such regimes fiscal policy, not monetary policy, determines the price level.
- Mechanism in FTPL: persistent budget deficits increase nominal public debt and are perceived as increased nominal wealth; debt deflation (an increase in prices that erodes the real value of public debt) may be the adjustment channel to restore solvency.
- Empirical aim: assess empirical plausibility of Ricardian and non-Ricardian regimes for a large sample of Sub-Saharan African (SSA) countries.

### Data, scope, and objectives
- Dataset: comprehensive database of fiscal and monetary outcomes for 22 major SSA countries, including domestic debt; broader datasets include up to 42 SSA countries for some comparisons (1980–2005).
- Key objectives:
  - Provide evidence on interaction between fiscal and monetary policy across SSA countries with different monetary and exchange rate regimes.
  - Propose an econometric procedure to identify fiscal- or monetary-dominant regimes.
  - Highlight policy implications for price stability where coordination is hampered by poor public financial management and inadequate monetary institutions.

### Stylized facts: Inflation, fiscal outcomes, and monetary aggregates (1980–2005)
- Sample and grouping:
  - 42 SSA countries divided into three groups:
    - (i) 14 CFA countries and the Union of Comoros (CFA group);
    - (ii) "non-CFA-fixed countries": Botswana, Cape Verde, Lesotho, Namibia, Seychelles, Swaziland (de facto fixed exchange rate throughout period);
    - (iii) "non-CFA floating countries": 22 countries characterized by independently fixed or floating exchange rate regimes (many shifted to de facto floating in mid-1990s).
  - Exchange rate regime classification: IMF de facto classification for 1990–2005; Levy-Yeyati and Sturzenegger (2003) de facto classification for 1980–1990.
- Budget balance and inflation (34-country sample, 1980–2004):
  - Simple OLS regression: annual average of inflation (annual percentage change in CPI) regressed on annual average of central government balance (including grants) as percent of GDP.
  - Result: significant positive correlation; "a 1 percentage point increase in the surplus-to-GDP ratio is associated with an almost 2 percent reduction in inflation."
  - R-squared = 0.14, indicating other factors play a role.
- Cross-group comparisons (averages reported in Table 1 and time series in Figures 2–4):
  - Countries with fixed exchange rate arrangements (CFA and non-CFA fixed):
    - On average lower inflation rates.
    - Less seigniorage.
    - Higher primary surpluses.
    - More success in achieving price stability and single-digit inflation.
  - Non-CFA floating countries:
    - On average double-digit inflation and volatile prices.
    - Average inflation at 26 percent (excluding Angola and DRC which had hyperinflation) — "was four times higher than in the fixed-exchange rate group."
    - Within this group, only 3 countries had on average single-digit inflation: Ethiopia, Mauritius, and Rwanda.
    - Nine countries experienced either hyperinflation (Angola, DRC, Zimbabwe) or average rates above 30 percent: Ghana, Guinea-Bissau before joining the CFA in 1997, Mozambique, Sierra Leone, Uganda, Zambia.
- Public liabilities-to-GDP ratios (sum of public debt and reserve money):
  - CFA countries: averaged about 125 percent of GDP.
  - Non-CFA floating countries: averaged about 106 percent of GDP.
  - Non-CFA fixed countries: averaged less than 60 percent of GDP.
- Currency composition of public debt:
  - CFA zone: public debt mostly foreign-currency-denominated (limited domestic debt markets until recently); thus CFA countries unable to inflate away real value of public liabilities (indexed to foreign currencies with low inflation).
  - Non-CFA floating countries: more developed domestic government debt markets; average nominal domestic debt ratio increased from 11 percent of GDP in the 1980s to 15 percent in the late 1990s; median increased from 4 percent to 10 percent.
  - Within non-CFA floating group, Angola, Botswana, DRC, Mozambique, and São Tomé and Príncipe had no domestic government debt markets until 2004.
  - Number of countries with debt-to-GDP ratios exceeding 20 percent rose from three at the beginning of 1980 to nine by 2000.
- Seigniorage (annual change in reserve money as a percent of GDP):
  - Seigniorage is an implicit form of taxation to finance deficits; measurement follows Fischer (1981).
  - Experience of seigniorage in SSA is mixed (Table 1). [Note: the source text breaks off mid-sentence regarding seigniorage in non-CFA floating countries.]

### Empirical strategy and model overview (as presented)
- The paper:
  - Discusses theories explaining fiscal-dominant regimes and why lack of coordination threatens price stability.
  - Presents a simple model emphasizing the role of fiscal policy in determining prices in a small open economy facing borrowing constraints (motivation: thin capital markets, inefficient tax systems in SSA increase recourse to seigniorage).
  - Describes econometric methodologies to test model predictions and to differentiate monetary vs. fiscal dominance.
  - Reports results and policy implications for achieving and maintaining price stability in the SSA region.

### Key empirical findings (summarized from text)
- Positive and economically meaningful correlation between fiscal balances and inflation: "a 1 percentage point increase in the surplus-to-GDP ratio is associated with an almost 2 percent reduction in inflation."
- Fixed-exchange-rate regimes (including CFA) are associated with:
  - Lower average inflation.
  - Lower seigniorage.
  - Higher primary surpluses.
  - Lower public liabilities-to-GDP (for non-CFA fixed group).
- Floating regimes associated with:
  - Higher average inflation (average 26 percent excluding Angola and DRC).
  - Higher variability and episodes of hyperinflation in some countries.
  - More extensive domestic debt accumulation over time.
- Institutional and market structure factors matter:
  - Underdeveloped domestic debt markets and thin capital markets in SSA increase reliance on seigniorage.
  - Currency denomination of debt affects ability to "inflate away" liabilities.

### Policy implications highlighted
- Coordination between monetary and fiscal authorities is crucial for price stability, especially where:
  - Public financial management is weak.
  - Monetary institutions are inadequate.
- Choice of exchange rate regime affects inflation outcomes and fiscal-monetary interactions:
  - Fixed-exchange-rate arrangements can help achieve lower and more stable inflation but may limit ability to inflate away foreign-currency-denominated liabilities.
  - In countries with developing domestic debt markets, inflationary financing via seigniorage remains an important risk.
- Identifying whether a country operates under fiscal or monetary dominance can inform appropriate policy mix and institutional reforms to secure price stability.

*Source: _wp08121 - 1. Sub-Saharan Africa: Selected Economic Indicators, annual averages, 1980-2005*

### 1.7 percent of GDP, almost twice as large as in CFA countries, where the average was below

### _wp08121 - 1.7 percent of GDP, almost twice as large as in CFA countries, where the average was below

### Seigniorage and inflation tax: empirical summary for SSA
- Seigniorage in the sample averaged 1.7 percent of GDP in non-CFA floating countries, almost twice as large as in CFA countries, where the average was below 1 percent of GDP.
- For some countries, seigniorage was an important component of financing policies: Angola, DRC, Ghana, Guinea-Bissau, Sao Tomé and Príncipe, Sierra Leone, Zambia, and most recently Zimbabwe.
- For fixed-rate countries (both CFA and non-CFA), seigniorage was much lower, with exceptions including Guinea-Bissau between 1997 and 2005 and Cape Verde.

### Inflation tax and its cross-country variation
- The inflation tax (the reduction in the real value of the stock of base money due to inflation) often provides a more striking picture than seigniorage.
- Table 1 (as described) reports:
  - Inflation tax was minimal in CFA fixed countries: average less than 1 percent.
  - Inflation tax in non-CFA fixed countries: averaged about 3 percent.
  - Inflation tax in non-CFA floating countries: very large; even after excluding Angola and the DRC, the average annual rate was about 350 percent.
- Country-specific high inflation/inflation-tax examples:
  - Ghana, Sierra Leone, Uganda, and Zambia: inflation tax was very large; in Uganda and Zambia it averaged above 2,000.
  - Malawi, Nigeria, and Tanzania: inflation tax rates ranged between 30 and 50 percent where inflation was more moderate but budget deficits were chronic.

### Preliminary empirical conclusions on inflation and fiscal determinants in SSA
- Countries with chronic budget deficits and high nominal public liabilities seem to have been more prone to fiscal-dominant regimes and high inflation.
- Countries that had no systematic recourse to implicit fiscal revenues (seigniorage and inflation tax) had lower inflation. Most low-inflation countries anchored monetary and fiscal policies either to a currency union with a fixed exchange rate (the CFA zone) or to a basket of currencies.
- Countries with multiple anchors or changes in monetary and exchange rate arrangements experienced much more difficulty credibly committing to low inflation and fiscal discipline.

### Theoretical background: fiscal dominance and price determination
- Discretionary policy can produce “dynamic inconsistency problems” leading to excessively high inflation (Barro and Gordon, 1983; Kydland and Prescott, 1977).
- Two main theoretical lenses:
  - Quantity Theory of Money (QTM):
    - Sustained inflation requires sustained increases in the money supply; inflation is interpreted as always and everywhere a monetary phenomenon.
    - Government intertemporal budget equation is treated as a solvency constraint; price level taken as given.
  - Fiscal Theory of the Price Level (FTPL):
    - The government intertemporal budget equation is an equilibrium condition that includes determination of the price level.
    - If primary surpluses are weakly correlated or uncorrelated with public liabilities, prices will adjust to ensure fiscal solvency and a fiscally dominant (FD) regime may prevail even without accommodating monetary policy.
    - If primary surpluses adjust promptly to limit growth of public liabilities, a monetary dominant (MD) regime prevails and monetary policy becomes the nominal anchor.

### Debates and critiques of FTPL
- Buiter (2002): FTPL may confuse budget constraints with equilibrium conditions and can lead to inconsistencies (e.g., overdetermined equilibria or negative price levels).
- Niepelt (2004): FTPL inconsistent with rational expectations equilibrium where asset holdings reflect optimal household choices; non-Ricardian policy feasibility relies on “not well founded” assumption of non-zero initial nominal government liabilities.
- Cochrane (2000) and Sims (1999): defend FTPL interpretation that price level is determined by valuation of nominal government debt as a claim to primary surpluses.

### Modeling fiscal dominance for a SSA small open economy
- To adapt FTPL insights to SSA, the authors relax some FTPL assumptions and incorporate features relevant to SSA:
  - Explicit inclusion of borrowing constraints on capital markets to explain recourse to implicit budgetary financing (seigniorage, inflation tax).
  - Two-country international setup with a single consumption good; one country represents a SSA economy receiving grants and facing a borrowing constraint on domestic debt.
  - No borrowing constraints on foreign currency–denominated debt (reflecting concessional donor and IFI financing).
- The model is based on Canzoneri, Cumby, and Diba (CCD) and draws testable predictions to identify fiscal requirements compatible with a fiscally dominant regime.

*Source: IMF working paper text excerpt provided in the content unit*

### 4.5 percent of the GDP during the sample period (1980–2005).

### _wp08121 - 4.5 percent of the GDP during the sample period (1980–2005).

### Model setup and household optimization
- Sample period reference: 1980–2005 and 4.5 percent of the GDP during the sample period (1980–2005).
- Households (domestic and foreign) maximize expected utility:
  - Objective: U = ∑_{s=t}^{∞} β^{s−t} u(c_{1,s}, c_{2,s}) (equation (1)), where β is the intertemporal discount rate and c_{j,s} is consumption of country j = 1, 2 goods at date s.
  - Goods are perishable and each individual is endowed with y_{j,s} units of country j goods at date s, with market equilibrium given by c_{j,s} + g_{j,s} = y_{j,s} (equation (2)), where g_{j,s} are government purchases following a regular exogenous stochastic process.
- Cash-in-advance (CIA) constraint:
  - Government purchases are not subject to CIA.
  - Household CIA constraint in country j (binding in equilibrium when nominal interest rates are positive): M_{j,s} = p_{j,s} c_{j,s} + p_{j,s} g_{j,s} (equation (5)).
  - Household budget constraint after CIA is given by a comprehensive nominal accounting expression (equation (6)) that includes domestic and foreign bond holdings, taxes τ_j, grants A, money balances, and endowments.

### Government budget, borrowing constraint, and market conditions
- Government instruments and flows:
  - Governments impose lump sum taxes τ_j, issue currency M_j, borrow via domestic and foreign-currency sovereign bonds (B_j and B^{*}_j), and receive/donate grants A used in the same period.
  - Flow budget constraint for government 1 given by a detailed nominal identity (equation (3)) involving i_{j,s}, e_s (units of currency 1 per unit of currency 2), and p_{j,s}.
- Domestic borrowing constraint for country 1:
  - Domestic bond issuance limited by B; B_{1,s} expressed with an indicator function BBI_{s;1} to represent the constraint (equation (4)).
  - Indicator BBI_{s;1} equals 1 if B_{1,s} ≤ B and 0 otherwise.
  - For computational convenience, the indicator is approximated by a logistic transition function ll_{s;BB,γ} = 1/(1 + exp(−γ(...))) (equation (7)); as γ → ∞ the logistic approximates the indicator, making the constraint practically binding when B_{1,s} > B.

### First-order conditions, PPP, IRP, and risk premium
- Purchasing power parity (PPP):
  - First-order condition with respect to consumption yields p_{1,t} = p_{2,t} (equation (8)); equation (8) represents PPP.
- Euler equations for bond holdings:
  - First-order conditions with respect to B_{1,s+1} and B_{2,s+1} yield Euler conditions (equations (9) and (10)) involving the intertemporal rate of substitution α_t ≡ β u'(c_{t+1})/u'(c_t), expectations, and the logistic approximation ll terms.
- Interest rate parity (IRP) with borrowing-constraint-induced premium:
  - Combining Euler equations produces an IRP condition (equation (11)) that contains an extra term due to the borrowing constraint: the term involving ll and l' (the logistic and its derivative).
  - Interpretation:
    - If B_{1,t+1} ≤ B, the term is approximately 1 and IRP resembles standard uncovered interest parity.
    - If E_t{B_{1,t+1}} > B, the term tends to infinity, causing the country-1 nominal interest rate i_t to explode and making domestic borrowing infeasible.
- Additional FOC:
  - First-order condition with respect to foreign-denominated bond *B_{1,1+s} yields i_{r,t} = i_{t} (equation (12)).

### Present value government budget constraint and implications of hitting the borrowing limit
- Real transformation and key definitions used to express the budget constraint:
  - Real primary balance (including grants): d_{t} ≡ (p_{1,t} / p_{1,t}) (g_{1,t} + τ_{1,t} − A_{1,t}) — (text defines d implicitly as a real primary balance including grants).
  - Real transfers from central bank (seigniorage): θ_{t} ≡ (i_{1,t} + i_{1,t+1}) p_{1,t} / M_{1,t} (definition in text).
  - γ_t = y_{t+1}/y_t is the ratio of real GDP between t and t+1.
- Present-value condition:
  - Government flow constraint combined with equilibrium conditions yields an expression for country 1 (equation (13)) that depends on l and ll terms, α_t, expectations of future variables, and real primary balances and seigniorage.
  - Solving forward with a transversality condition (equation (14)) produces the present-value government budget constraint for the SSA country (equation (15)): the real value of public liabilities over GDP equals the present discounted value of future real primary surplus and seigniorage over GDP minus the discounted sum of real future money and domestic debt over GDP for periods when the borrowing constraint is binding.
- When debt is below the borrowing constraint (B_{1,n} < B):
  - The logistic-derived term is approximately 1 and (15) reduces to the standard government budget constraint (comparable to country 2).
- When debt equals the borrowing constraint (B_{1,n} = B):
  - Country 1 can only borrow abroad; l_n is close to zero, and the logistic term is negative.
  - To satisfy the present-value constraint (Ricardian equivalence), real primary surplus and seigniorage over GDP must be higher than for country 2; i.e., hitting the constraint requires larger fiscal effort (higher primary surplus including seigniorage).
  - Policy implication: when debt hits the domestic borrowing limit the government must increase primary surplus (including seigniorage) or otherwise restore solvency via external borrowing or price level adjustment.

### Fiscal-dominant vs. monetary-dominant regimes — theoretical implications
- Regime definitions and mechanisms:
  - Fiscal-dominant (FD) regimes:
    - Primary surpluses (including seigniorage) are determined by an arbitrary process unrelated to liabilities.
    - Fiscal policy sustainability requires "domestic debt deflation": an increase in the price level that erodes the real value of domestic public debt and real financial wealth until market equilibrium is restored.
  - Monetary-dominant (MD) regimes:
    - Primary surpluses (including seigniorage) are closely related to the level of liabilities and are adjusted to satisfy the government budget constraint (15).
    - Monetary authorities can anchor expectations and determine the price level.
- Result: the model shows how fiscal and monetary dominance are determined in an open economy facing borrowing constraints and through the intertemporal government budget constraint.

### Econometric methodology — identification of fiscal or monetary dominance
- Overview: use nonstructural VAR and cointegration analysis to provide evidence on monetary vs. fiscal determinants of inflation. Four complementary approaches are developed:
  1. Dynamic relationship between public liabilities and primary surpluses: test how fiscal authorities respond to ensure public-sector solvency.
  2. Interaction between fiscal and monetary variables: estimate relative importance of wealth effects of domestic public debt and money growth on inflation.
  3. Relationship among inflation, aggregate demand, and monetary policy instruments: test monetary authority responses to inflation shocks.
  4. Johansen cointegration analysis: test for deviations from Ricardian equivalence to help identify fiscal-dominant regimes.
- VAR and estimation details:
  - Stationarity tested via augmented Dickey-Fuller (ADF) unit root test with a constant and trend (five lags as starting point) and general-to-specific methodology.
  - Joint stationarity and cointegration assessed using Johansen cointegration analysis.
  - Lag length chosen by likelihood ratio, Akaike and Schwarz information criteria.
  - VARs estimated in levels and first differences; stability checked via roots of characteristic polynomial (stable if all roots inside the unit circle).
  - Normality of residuals tested via Urzua (1997) multivariate extension of Jarque-Bera; homoskedasticity tested with an extension of White’s heteroskedasticity test.
  - Final decision: estimate a standard VAR including a constant and two lags for each country.
  - Impulse responses and variance decompositions computed using Cholesky decomposition; standard deviation bands derived via Monte Carlo bootstrap with 1,000 iterations.
- Panel VAR (PVAR):
  - Estimated for the entire SSA panel and three subgroups (Table 1 referenced in source).
  - Estimation strategy based on Kireyev (2000): pool data, test panel unit roots and lag selection, then estimate PVAR.
  - Group mean estimator also used (relaxes common coefficient hypothesis); results similar to pooled estimation.
- Public liabilities and primary surpluses approach (CCD methodology):
  - Identify regimes by impulse-response analysis of future total public liabilities to a shock in current surpluses (d_{t+θ}).
  - Surplus persistence notion:
    - Surplus with positive autocorrelation up to at least 5 lags = positive and persistent; otherwise considered negatively autocorrelated.
  - Identification criteria (summarized in Table 2 in source):
    - Under MD: a positive shock to current surplus leads to a fall in future liabilities (negative relationship between current surpluses and future liabilities).
    - Under FD: surpluses are exogenous; future liabilities may be unresponsive or increase after a surplus shock (consistent with FTPL where price level adjusts).
    - Ambiguous/non-identified cases arise under certain combinations of responses or when surpluses are negatively autocorrelated.
- Domestic public debt vs. money growth (pass-through analysis):
  - Empirical aim: determine whether inflation variability is better explained by nominal domestic debt growth or money growth, controlling for aggregate demand (proxied by real output gap).
  - VAR causal ordering: nominal domestic debt growth → growth rate of reserve money → real output gap → inflation rate (ensures contemporaneous response only by inflation to fiscal/monetary shocks).
  - Variance decomposition for inflation isolates percentage of forecast error of inflation explained by shocks to nominal debt growth or money growth.
    - If forecast error largely explained by nominal debt growth shocks → evidence consistent with FTPL / fiscal dominance (wealth effects of nominal debt growth).
    - If forecast error largely explained by money growth shocks → monetary accommodation (debt monetization) channel consistent with QTM and associated with FD regime.
  - Caution: increases in domestic debt may be driven by output collapses from exogenous shocks (e.g., social unrest or civil wars in SSA) that lower taxation and raise domestic debt; this complicates interpretation of wealth effects and warrants in-depth single-country analyses.
- Active monetary authority test (approach 3):
  - A two-step procedure is implemented to verify whether monetary authorities actively mitigated inflation shocks during the sample period (methodological details continue beyond provided excerpt).

*Source: _wp08121 - 4.5 percent of the GDP during the sample period (1980–2005).*

### 1.      Run a VAR with the following causal ordering: real output gap (proxy for aggregate

### _wp08121 - 1.      Run a VAR with the following causal ordering: real output gap (proxy for aggregate

### VAR specification and identification
- Run a VAR with the following causal ordering: real output gap (proxy for aggregate demand) → inflation → reserve money growth (or the discount rate when available).
- This ordering ensures that the monetary instrument (either reserve money or the discount rate) is the only variable responding directly to inflationary shocks.

### Key detail
- 33

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08121.pdf*

### 2.      Compute the impulse responses to estimate how the monetary instrument responds to

### 2.      Compute the impulse responses to estimate how the monetary instrument responds to an innovation in inflation

### Methodology: Impulse responses and identification of monetary policy responsiveness
- Compute impulse responses from a VAR model ordered: Output gap → Reserve money growth → Inflation (Table 3).
- Identification criterion: money growth rates should respond negatively (or, equivalently, discount rates respond positively) to a positive innovation in current inflation to indicate an active monetary policy (Table 3, Criterium 1).
- If money growth increases (or discount rate decreases) after a positive inflation shock, monetary authority is passive (accommodating fiscal shocks) and the country cannot be characterized as following a MD regime (text).
- VARs are also estimated including time fixed effects to control for other factors (footnote 31).
- A VAR with discount rate instead of nominal reserve growth provides similar results (Table 6, footnote 1).
- Monetary policy responsiveness identification relies on the sign of the impulse response functions over the analyzed horizon (7 periods) (Table 4 notes).

### Ancillary tests and complementary approaches
- Ricardian equivalence (FD test): check cointegration between lagged outstanding public liabilities and primary surplus (including grants and seigniorage) in percent of GDP (corresponding to equations (14) and (15) of the model). Cointegration of order CI (1, 1) implies existence of a stationary linear combination and supports Ricardian equivalence (section on Active fiscal authority: the Ricardian Equivalence, or FD test).
- CCD approach (public liabilities → primary surplus VAR) provides signs of responses and surplus autocorrelation to identify FD, MD, or not identified regimes (Tables 2 and 4).
- Third approach: domestic public debt pass-through analysis via VAR decomposition (nominal domestic public debt → nominal reserve money growth → real output gap → inflation) for countries with sufficient domestic debt data (Table 5, VAR model description).
- Robustness: include tests of monetary authority activity (Table 6) as a robustness check on FD identifications.

### Data: sample, variables, and sources
- Sample: unbalanced panel of 22 SSA countries with at least 17 and at most 25 years of observations for each country from 1980 through 2005 (section V.A).
- 22 countries: Botswana, Burundi, Cameroon, Ethiopia, Ghana, Kenya, Lesotho, Malawi, Mali, Mauritius, Nigeria, Rwanda, Senegal, Seychelles, Sierra Leone, South Africa, Swaziland, Tanzania, Togo, Uganda, Zambia, and Zimbabwe (footnote 34).
- For the third approach (domestic public debt pass-through), VAR decomposition computed for 18 SSA countries with at least 16 years of observation (excluding Botswana, Mali, Senegal, and Togo) (section V.A).
- Main data sources:
  - Grants and primary and total public surplus: IMF African Department database.
  - Output, price indexes, and monetary aggregates: IMF World Economic Outlook (WEO).
  - Discount rate and foreign, domestic and total public debt: IMF International Financial Statistics (IFS).
  - Additional public debt sources: Christensen (2004), IMF country desks data, World Bank African Indicators (WBAI), World Bank Development Indicators (WDI).
- De facto exchange rate regime classification: compiled since 1997 from “IMF Annual Report on Exchange Arrangements and Exchange Restrictions,” retroactively to 1980 using Bubula and Ötker-Robe (2002), Levy-Yeyati and Sturzenegger (2002), and Reinhart and Rogoff (2003) (section V.A).

### Single-country analysis: identification framework and sample coverage
- VAR ordering used for CCD approach: public liabilities → primary surplus (including seigniorage) (Table 4 caption).
- Variables expressed in real terms as percent of GDP; surplus is primary surplus, including grants and seigniorage; liabilities include public debt and reserve money (Table 2 notes, Table 4 notes).
- A plus sign indicates positive impulse response over the time horizon analyzed (7 periods); a ─/+ sign indicates impulse responses switched sign during the horizon (Table 4 notes).
- A plus sign indicates positive autocorrelation of at least 5 periods; a minus sign indicates otherwise (Table 4 notes).

### Main empirical findings (Tables 4, 5, and 6; summarized results)
- Regime identification (Table 4 summary, sample of 22 countries):
  - Of 22 SSA countries, 9 have an identified regime throughout the sample period:
    - 4 estimated to have followed a FD regime: Botswana, Burundi, Tanzania, and Zimbabwe.
    - 5 estimated to have followed a MD regime: Cameroon, Kenya, Nigeria, Rwanda, and South Africa.
  - For the remaining 13 countries, no regime was identified:
    - 12 cases had negatively correlated surpluses.
    - 1 case (Zambia) had positively correlated surpluses but inconclusive impulse-response analysis.
- Ricardian equivalence (Johansen cointegration between public liabilities and primary surplus):
  - Results robust to Haug (1991) cointegration test except for Kenya and South Africa where MD regimes had no cointegrating relationship.
  - Zimbabwe (FD regime) had one cointegrating relationship, implying fiscal authorities matched growing liabilities by increasing money reserve and seigniorage rather than primary surplus (section V.B).
- Pass-through analysis on inflation variability (Table 5; average variance decomposition for ten periods):
  - Inflation variability could be mostly explained by:
    - Domestic debt growth in 5 countries: Burundi, Lesotho, Mauritius, South Africa, and Zambia.
    - Money growth in 6 countries: Cameroon, Kenya, Malawi, Sierra Leone, Tanzania, and Zimbabwe.
    - Both determinants in 5 countries: Ethiopia, Nigeria, Seychelles, Swaziland, and Uganda.
  - Selected country percentages (Table 5):
    - Burundi: Inflation variability due to nominal domestic public debt 18%; due to nominal growth in money reserves 5%.
    - Cameroon: 29% (debt); 40% (money).
    - Zimbabwe: 28% (debt); 49% (money).
    - Inflation variability (average, whole sample): 19% (debt) and 20% (money).
  - For Ghana and Rwanda both variables have very low explanatory power for inflation variability (standard errors quite large) (Table 5, footnote 38).
- Active monetary authority tests (Table 6; robustness check for FD identifications):
  - For countries identified as FD by CCD (Botswana, Burundi, Tanzania, and Zimbabwe), tests generally reject that monetary authorities acted consistently with a MD regime by tightening reserve money or increasing discount rate in response to an inflationary shock—except for Botswana where results differ (section V.B).
  - Table 6 reports signs of responses of nominal reserve money growth to a shock in inflation and response of inflation to a shock in nominal reserve money growth for each country (Table 6).

### Interpretation and broader insights
- Nominal public liabilities, whether reflected in money growth or nominal public debt, matter for price stability; wealth effects of changes in nominal public debt can increase inflation variability, consistent with the fiscal theory of price determination (Conclusion, VI).
- Differences in relative importance of monetary and fiscal sources of inflation correspond to differences in exchange rate regimes:
  - Contribution of money growth to inflation in the CFA monetary union is far less relevant than in floating exchange rate regimes (Conclusion).
  - Restrictions on issuing nominal public (domestic) debt in CFA countries during the period may have helped keep inflation low (Conclusion).
  - Non-CFA fixed exchange rate countries with domestic-currency nominal sovereign debt (e.g., Botswana, Lesotho, Swaziland) experienced higher average inflation than CFA area countries (Conclusion).
- Fiscal-dominant regimes can arise even where monetary policy is independent and not accommodative (e.g., CFA), highlighting the importance of monetary–fiscal coordination (Conclusion).
- Inflation variability can be associated with changes in nominal public debt even in MD or Ricardian regimes, implying nominal debt variability can be detrimental to price volatility (section V.B).

### Limitations and suggestions for future work
- Limited sample size of public debt for SSA reduces statistical power of the econometric tests (footnote 37).
- CCD approach limitation: cannot identify predominant regime if FD and MD alternate during the sample period; regime-switching VAR techniques would be appropriate but are beyond the paper’s scope (Table 4 notes).
- Nonstructural VARs cannot distinguish between capacity and willingness to honor fiscal obligations; alternative strategy: use dynamic general equilibrium model assumptions to identify fiscal-dominant price determination (Conclusion).
- Study does not allow for regime switching between MD and FD; applying VAR techniques that allow regime identification over time would be useful (Conclusion).
- The study does not distinguish among countries by average inflation level (very high/hyperinflation vs low/moderate); other factors (output and price volatility, institutional capacity, financial depth) were not analyzed but may matter (Conclusion).

### Stylized facts and summary classification (1980–2005)
- Countries characterized by MD throughout 1980–2005: Cameroon, Kenya, Nigeria, Rwanda, and South Africa (Conclusion summary).
- Countries characterized by FD throughout 1980–2005: Botswana, Burundi, Tanzania, and Zimbabwe (Conclusion summary).
- For remaining countries, evidence is less conclusive to infer a dominant regime throughout the sample period (Conclusion summary).

*Source: _wp08121 - 2.      Compute the impulse responses to estimate how the monetary instrument responds to an innovation in inflation (PDF chapter/section).*

### References

### _wp08121 - References

### Major themes covered by the references
- Fiscal theory of the price level and fiscal–monetary interactions
  - Barro, 1974; Barro and Gordon, 1983
  - Buiter, 2002; Buiter, 2007
  - Canzoneri, Cumby, and Diba, 1998; Canzoneri, Cumby, and Diba, 2001
  - Christiano and Fitzgerald, 2000
  - Cochrane, 1998; Cochrane, 2000
  - Daniel, 2001
  - Dupor, 2000
  - Fialho and Portugal, 2005
  - Fisher, 1981
  - Leeper, 1991; Leeper and Troy, 2006
  - Loyo, 1998; Loyo, 1999; Loyo, 2000
  - Niepelt, 2004
  - Sala, 2004
  - Sargent and Wallace, 1981
  - Sims, 1994; Sims, 1999
  - Woodford, 1994; Woodford, 1998

- Exchange rate regimes, currency crises, and "original sin"
  - Bubula and Ötker-Robe, 2002
  - Corsetti and Mackowiak, 2005
  - Khan, 2005
  - Levy-Yeyati and Sturzenegger, 2003
  - Reinhart and Rogoff, 2002
  - Tornell and Velasco, 2000

- Public finance, sovereign borrowing, domestic debt markets, and seigniorage
  - Bayoumi, Goldstein, and Woglom, 2005
  - Burnside, Eichenbaum, and Rebelo, 2003
  - Christensen, 2004
  - Grilli, Masciandaro, and Tabellini, 1991
  - Haug, 1991
  - Kremers, 1987
  - Ribeiro, Beetsma, and Schabert, 2007
  - Rogoff, 1985
  - Tanner and Ramos, 2002
  - Trehan and Walsh, 1991
  - Wheaton, 2000
  - Zoli, 2005
  - Buiter, 2007 (Seigniorage)

- Empirical methods, macroeconomic dynamics, and country studies
  - Kireyev, 2000 (Panel VAR; Arab World)
  - Loungani and Swagel, 2001 (Sources of Inflation in Developing Countries)
  - Rand and Tarp, 2002 (Business Cycles in Developing Countries)
  - Saxegaard, 2006 (Excess Liquidity; Sub-Saharan Africa)
  - Khan, 2005 (Bond Market Development in Sub-Saharan Africa)
  - Trehan and Walsh, 1991 (Testing Intertemporal Budget Constraints)
  - Urzúa, 1997 (Omnibus Tests for Multivariate Normality)
  - Sala, 2004 (Identifying Restrictions and Empirical Evidence)

### Representative publication details and numeric citations preserved
- Journal articles and volumes:
  - Barro, 1974. Journal of Political Economy, Vol. 82(6), pages 1095–1117.
  - Canzoneri, Cumby, and Diba, 2001. American Economic Review, Vol. 91, No. 5, pp. 1221–38.
  - Dupor, 2000. Journal of Monetary Economics, Vol. 45, pp. 613–30.
  - Tornell and Velasco, 2000. Journal of Monetary Economics, Vol. 45, pp. 399–436.
  - Trehan and Walsh, 1991. Journal of Money, Credit and Banking, May, Vol. 23(2), pp. 206–23.
  - Rand and Tarp, 2002. World Development, Vol. 30, Issue 12, pp. 2071–88.

- Working papers and reports:
  - Bubula and Ötker-Robe, 2002. IMF Working Paper 02/155.
  - Christensen, 2004. IMF Working Paper 04/46.
  - Kireyev, 2000. IMF Working Paper 00/54.
  - Saxegaard, 2006. IMF Working Paper 06/115.
  - Burnside, Eichenbaum, and Rebelo, 2003. NBER Working Paper No. W9786.
  - Buiter, 2007. NBER Working Paper No. W12919.
  - Reinhart and Rogoff, 2002. NBER Working Paper No. W8963.

- Specific numeric identifiers and dates preserved:
  - IMF Working Paper 02/155; IMF Working Paper 04/46; IMF Working Paper 00/54; IMF Working Paper 06/115.
  - NBER Working Paper No. W9786; NBER Working Paper No. W12919; NBER Working Paper No. W8963.
  - Journal volumes and page ranges are preserved exactly as cited above.

### Inferred topical focus for the citing work (based only on listed references)
- The references collectively indicate emphasis on:
  - Theoretical and empirical dimensions of the fiscal theory of the price level.
  - Interactions between fiscal policy and monetary policy, including issues of fiscal dominance, seigniorage, and price-level determinacy.
  - How exchange rate regimes influence fiscal discipline and the incidence of currency crises, including literature on "original sin" and de facto classifications of exchange rate regimes.
  - Sovereign debt structure, domestic debt market development (including Sub-Saharan Africa), and testing fiscal sustainability.
  - Empirical methods applied to developing-country macro dynamics, including panel VARs, tests for cointegration, and multivariate normality.

*Source: _wp08121 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08121.pdf_