## _wp11203

## Source details

**Canonical URL:** [_wp11203](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11203.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11203.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11203.pdf.json)

---

### Introduction and motivation
- International financial market developments can amplify real shocks through:
  - direct impact of capital flows on aggregate demand and domestic financial markets, and
  - multiplication via government access to financing and borrowing costs.
- High public debt countries: negative real or financial shocks can deteriorate creditworthiness and trigger nonlinear second-round effects (high interest rates, credit crunch, recession).
- Policy objective: keep public debt at sufficiently low levels as self-insurance to permit expenditure smoothing and maintain access to financing.
- Paper’s objective: propose a fiscal framework to smooth primary expenditure over the business cycle by providing creditors sufficient assurance that public debt obligations will remain within a preannounced band with high probability.

### The proposal: public debt targeting (overview)
- Core design:
  - Commit to keep public debt within a specified band with high probability.
  - Announce a primary expenditure budget envelope for the next budget cycle such that probabilistic public debt outcomes over the medium term remain within the band.
- Two main public announcements:
  - A trajectory of public debt and its probabilistic band.
  - A primary expenditure budget envelope for the next budget cycle.
- Four intermediate technical parameters (announced and linking the two main announcements):
  - (i) a medium-term horizon in years,
  - (ii) a public debt upper threshold at the end of the medium term horizon,
  - (iii) a notional spending trajectory that maps next years’ primary expenditure announcements to debt trajectories over the medium term,
  - (iv) a probability value attached to debt remaining below the upper threshold.
- Operational recursion:
  - In year t announce S(t) such that if primary spending for the following T years remains at the same level in real per capita terms, public debt would take a level D(t+T) percent of GDP or lower with probability P(t+T).
  - Recompute S(t+1) in year t+1 after observing shocks; budgets adjusted each year.
- Discretionary extension:
  - Spending trajectory can incorporate a pre-announced discretionary component linked to deviations of real GDP growth from population growth.

### Rationale and economic logic
- At sufficiently high public debt levels, marginal deficits and debt accumulation can trigger nonlinear second-round effects: interest rate spikes, foreign exchange instability, capital outflows, asset deflation, credit crunch, and uncertainty about taxation.
- A framework that provides high-probability assurance that public debt will remain on a non-increasing trend acts as an aggregate demand smoothing device.
- The public debt band permits primary balance flexibility and avoids large consolidations in bad states that would undermine credibility.

### Desirable properties of the framework
- Credibility:
  - Anchored by linking government spending to projected debt trajectories; reputational gains if governments abide by announcements.
  - A pre-specified compensating consolidation strategy reduces expectation of unsustainable debt paths after shocks.
- Primary expenditure-smoothing:
  - Debt thresholds allow larger deficits than budgeted if revenues fall while primary expenditure remains at the budget level.
  - Focus on primary spending (excluding interest) is appropriate because primary expenditures are the relevant aggregate for debt sustainability given a fixed tax framework; interest expenditure tends to be largely predetermined.
  - Automatic stabilizers on revenues and interest expenditures are allowed to play in full.
- Flexibility:
  - Not a strict spending rule; the notional spending trajectory is an intermediate technical parameter recalculated recursively.
  - Budgets need contingent allocations for triggered stabilizers (e.g., social insurance increases in bad states).
- Medium-term macroeconomic framework:
  - Requires fiscal projections and probabilistic projections from an explicit, disclosed economic model capable of stochastic forecasts for relevant variables.
  - More vulnerable and volatile economies require more conservative framework parameters.

### Political economy and financial markets context
- Political economy explanations for procyclicality include special interest pressures, common pool problems, and institutional distortions (Talvi and Vegh (2000); Cole and Kehoe (1992); Velazco (1998); Roubini (1991); Grilli, Masciandaro, and Tabellini (1991)).
- Financial markets’ lending behavior interacts with fiscal policy: financing constraints can make interest rates high in low states and low in booms (Gavin et al. (1996)).

### Relation to fiscal rules and comparison with alternatives
- The proposal is framed as a stable framework (four parameters + economic model) rather than fixed numerical targets (Kopits and Symansky (1998) definition of fiscal rules).
- Advantages over structural/cyclically-adjusted balance rules:
  - No need to identify permanent vs. transitory output components.
  - Spending levels determined from projected debt dynamics (debt accumulation identity).
  - Uses a broader set of variables (interest rates, real exchange rates).
  - Directly addresses debt sustainability concerns through commitment to debt thresholds and a debt trajectory band.
  - Can allow constrained cyclical space or state-contingent additional primary expenditure.
- Role as transition: useful as a transition framework toward structural balance rules when public debt and credibility concerns decline.

### Stylized facts for Caribbean sample (empirical motivation)
- High public debt-to-GDP ratios:
  - Median public debt-to-GDP increased from 81 percent to 98 percent between 2008 and 2009.
  - In 2009 public debt was about 4 times the size of total revenues.
  - 2010 estimates show a further increase for almost all countries.
- Procyclical primary expenditure:
  - Median cyclical components of real output, primary expenditures and total revenues show positive co-movement over the last 15 years.
  - Median correlation of primary expenditures with both total revenues and real GDP is about 50 percent.
- Strong cross-country synchronization:
  - Every country’s GDP cycle is positively correlated with that of the US.
  - Cross-country correlation in interest rates, real exchange rates and net capital inflows.
- Financial variables explain substantial share of primary expenditure variation:
  - Net capital inflows, real exchange rate depreciation and interest rates jointly explain ½ of the average forecast-error variance of primary expenditures.
- VAR for variance decomposition: two-lag six-variable VAR with annual observations of output growth rate, a short-term interest rate, growth rate of the real exchange rate, change in net capital inflows, total revenue growth and primary expenditure growth.

### Simulation framework and Monte Carlo procedure
- Debt decomposition into three components:
  - (i) commercial domestic-currency component,
  - (ii) commercial foreign-currency component,
  - (iii) official bilateral/multilateral fraction (usually USD).
- Component evolution equation (textual form): d_t^j evolves according to interest, inflation, growth, and share of primary balance servicing type j (i_t^j, g_t, π_t, α_t^j, b_t).
- Interest rate specification: implicit/average rate approximated by moving average of marginal rate over M years (M = average maturity).
- VAR specification (Xt = AL(L) Xt-1 + B i*t + εt) with Xt = (g_t, i_t, d_t, η_t, m_t, τ_t, s_t)T containing:
  - real output growth,
  - a short-term domestic interest rate,
  - change in bilateral real exchange rate (vs. US dollar),
  - change in real net capital inflows,
  - percentage change in total revenues,
  - percentage change in primary expenditures.
- Monte Carlo draws:
  - At every forecast period draw 2000 random vectors ˆεω using Choleski factorization of Ω.
  - Passive simulations use VAR-fitted primary expenditure equation; active simulations implement expenditure-smoothing fiscal policy with primary expenditures ordered last to minimize contemporaneous feedback.
- Two required inputs for the policy framework simulations:
  1. Starting point of the primary expenditure path (next budget cycle envelope).
  2. Specification of the primary expenditure pattern over the medium term.
- Forward-looking primary expenditure assumption: primary expenditure remains constant in real per-capita terms except for initial consolidation and feedback via responsiveness parameter κ.
- Iterative backward induction: starting primary expenditure level ̃s_{t+1} chosen so that dt+T ≤ d with probability p.
- Interpretation:
  - p = 0.5 yields primary expenditure that stabilizes public debt-to-GDP in expected terms over the medium-term.
  - Higher p implies additional consolidation and a downward expected debt trend.
  - Longer horizon T smooths adjustments; too long a horizon may undermine credibility.

### Forward-looking expenditure-smoothing specification
- Specification: sω = ̃sω − κ (gω − g), ∀ω ∈ t+1,T
  - g is trend growth of real GDP, set equal to average population growth rate.
  - gω is simulated growth rate of real GDP.
  - κ parameter: responsiveness of primary expenditure growth to deviations of growth from trend.
  - ̃st+1 represents initial consolidation (̃st+1 < 0 implies consolidation); ̃sω = g for ω ∈ t+2,T.
- Interpretation of κ:
  - κ = 0: no discretionary impulse; primary expenditure constant in real per-capita terms after initial consolidation.
  - κ > 0: countercyclical impulse (additional expenditure opposite sign to GDP deviations).
  - κ < 0: procyclical impulse.

### St. Lucia illustration (parameter choices and partial result)
- Policymaker parameter choices in example:
  - Horizon T = 5 years.
  - Upper public debt threshold = 70 percent (estimated debt to GDP in 2010 is 80 percent).
  - Probability p = 0.7.
- With this parameter choice, the level of primary expenditure consolidation required in the next budget cycle ranges from 1.7 percentage points of GDP under the pro-cyclical choice, to 3.4 percentage points of GDP under the counter-cyclical choice, with the acyclical impulse lying between these values.

### Monte Carlo simulation results — country-level highlights (2010 baseline; 2011 required consolidation outcomes)
- Jamaica:
  - Estimated debt/GDP in 2010: 139%
  - Estimated primary balance/GDP in 2010: 7.6%
  - Expected debt/GDP in 2015 under "passive": 117%
  - Illustrative debt/GDP threshold in 2015: 100%
  - Required consolidation of primary expenditures in 2011 (percentage of GDP): 50% row: -4.2, -4.0, -4.4; 70% row: -0.7, -0.7, -0.7
  - Note: Jamaica is the only country in the sample that does not show a particular need for further consolidation in 2011 under these assumptions.
- Antigua and Barbuda:
  - Estimated debt/GDP in 2010: 119%
  - Estimated primary balance/GDP in 2010: -1.8%
  - Expected debt/GDP in 2015 under "passive": 160%
  - Illustrative threshold 2015: 120%
  - Required consolidation in 2011: 50% row: 8.3, 9.1, 7.6; 70% row: 10.3, 11.1, 9.4
- The Bahamas:
  - Estimated debt/GDP in 2010: 65%
  - Estimated primary balance/GDP in 2010: -3.9%
  - Expected debt/GDP in 2015 under "passive": 79%
  - Illustrative threshold 2015: 60%
  - Required consolidation in 2011: 50% row: 0.8, 1.1, 0.4; 70% row: 1.9, 2.4, 1.3
- Barbados:
  - Estimated debt/GDP in 2010: 122%
  - Estimated primary balance/GDP in 2010: -1.2%
  - Expected debt/GDP in 2015 under "passive": 137%
  - Illustrative threshold 2015: 130%
  - Required consolidation in 2011: 50% row: 7.8, 7.6, 7.9; 70% row: 10.3, 10.4, 10.2
- Grenada:
  - Estimated debt/GDP in 2010: 116%
  - Estimated primary balance/GDP in 2010: 0.3%
  - Expected debt/GDP in 2015 under "passive": 135%
  - Illustrative threshold 2015: 110%
  - Required consolidation in 2011: 50% row: 1.3, 0.7, 1.4; 70% row: 3.7, 3.2, 3.7
- St. Kitts and Nevis:
  - Estimated debt/GDP in 2010: 184%
  - Estimated primary balance/GDP in 2010: -1.9%
  - Expected debt/GDP in 2015 under "passive": 205%
  - Illustrative threshold 2015: 180%
  - Required consolidation in 2011: 50% row: 11.1, 11.7, 10.5; 70% row: 13.6, 14.3, 13.1
- St. Lucia:
  - Estimated debt/GDP in 2010: 80%
  - Estimated primary balance/GDP in 2010: -3.2%
  - Expected debt/GDP in 2015 under "passive": 87%
  - Illustrative threshold 2015: 70%
  - Required consolidation in 2011: 50% row: 0.7, 1.3, 0.1; 70% row: 2.5, 3.4, 1.7
- St. Vincent and the Grenadines:
  - Estimated debt/GDP in 2010: 91%
  - Estimated primary balance/GDP in 2010: -1.4%
  - Expected debt/GDP in 2015 under "passive": 115%
  - Illustrative threshold 2015: 90%
  - Required consolidation in 2011: 50% row: 6.1, 5.8, 6.4; 70% row: 7.7, 7.5, 7.9
- Summary finding: For Bahamas, Grenada, and St. Lucia a primary expenditure consolidation of around 2 to 3 percentage points of GDP in 2011 together with acyclical/moderately procyclical/countercyclical rules suffices to keep 2015 debt below 2010 levels in 70 percent of simulations. Antigua and Barbuda, Barbados, St. Kitts and Nevis, and St. Vincent and the Grenadines require consolidations in excess of 7 percentage points of GDP in 2011 per the framework.

### Primary expenditure smoothing outcomes (coefficients of variation, average 2011–15)
- Coefficient of Variation (CV) of simulated real growth rates of primary expenditures by policy:
  - Antigua and Barbuda: Historical behavior 25.7; Structural balance rule 17.6; Public Debt Targeting 3.4
  - The Bahamas: Historical behavior 1.5; Structural balance rule 1.3; Public Debt Targeting 2.3
  - Barbados: Historical behavior 3.3; Structural balance rule 1.3; Public Debt Targeting 1.9
  - Grenada: Historical behavior 91.9; Structural balance rule 5.8; Public Debt Targeting 2.7
  - Jamaica: Historical behavior 2.3; Structural balance rule 2.2; Public Debt Targeting 1.7
  - St. Kitts and Nevis: Historical behavior 1.5; Structural balance rule 3.0; Public Debt Targeting 2.7
  - St. Lucia: Historical behavior 3.5; Structural balance rule 2.2; Public Debt Targeting 1.5
  - St. Vincent and the Grenadines: Historical behavior 6.1; Structural balance rule 2.4; Public Debt Targeting 1.8
- Finding: CV of primary expenditure growth under public debt targeting is generally lower than under a structural balance rule for this parameterization (higher only in two out of eight cases).

### Cyclical properties and correlations (select averages)
- Coefficient of Variation of cyclically-adjusted primary balances (average 2011–15):
  - Structural balance rule vs. Public debt targeting averages shown for sample countries (see source text for country-level numbers).
  - Interpretation: For five out of eight cases, structural balance rule shows more variability in cyclically-adjusted primary balances than public debt targeting; revenues and/or trend growth tend to be more stable or higher under debt targeting, lowering CVs.
- Correlations of real growth rates of primary expenditures with GDP and with revenues (Debt Targeting averages):
  - Correlations with GDP:
    - Antigua and Barbuda: 0.28
    - The Bahamas: 0.29
    - Barbados: 0.17
    - Grenada: -0.08
    - Jamaica: 0.44
    - St. Kitts and Nevis: 0.02
    - St. Lucia: -0.43
    - St. Vincent and the Grenadines: -0.16
  - Correlations with revenues:
    - Antigua and Barbuda: 0.70
    - The Bahamas: 0.37
    - Barbados: 0.32
    - Grenada: 0.46
    - Jamaica: 0.56
    - St. Kitts and Nevis: 0.66
    - St. Lucia: 0.77
    - St. Vincent and the Grenadines: 0.32
- Finding: Average correlation between primary expenditure and revenues under public debt targeting in this sample is 52 percent for the parameter choice used; positive revenue-expenditure correlation implies some pro-cyclical adjustment by design to avoid excessive debt accumulation.

### Public debt outcomes by 2015 (selected entries: Expected, Std. Dev., CV)
- Antigua and Barbuda:
  - Historical behavior Expected 160.4, Std. Dev. 42.3, CV 0.26
  - Structural balance rule Expected 120.9, Std. Dev. 18.0, CV 0.15
  - Public debt targeting Expected 114.3, Std. Dev. 11.0, CV 0.10
- The Bahamas:
  - Historical behavior Expected 79.4, Std. Dev. 12.5, CV 0.16
  - Structural balance rule Expected 64.3, Std. Dev. 6.7, CV 0.10
  - Public debt targeting Expected 60.9, Std. Dev. 5.3, CV 0.09
- Barbados:
  - Historical behavior Expected 136.8, Std. Dev. 15.9, CV 0.12
  - Structural balance rule Expected 125.6, Std. Dev. 12.8, CV 0.10
  - Public debt targeting Expected 125.3, Std. Dev. 10.8, CV 0.09
- Grenada:
  - Historical behavior Expected 134.6, Std. Dev. 37.9, CV 0.28
  - Structural balance rule Expected 108.9, Std. Dev. 7.9, CV 0.07
  - Public debt targeting Expected 111.3, Std. Dev. 6.4, CV 0.06
- Jamaica:
  - Historical behavior Expected 116.9, Std. Dev. 26.0, CV 0.22
  - Structural balance rule Expected 139.1, Std. Dev. 11.3, CV 0.08
  - Public debt targeting Expected 110.9, Std. Dev. 5.4, CV 0.05
- St. Kitts and Nevis:
  - Historical behavior Expected 205.3, Std. Dev. 19.9, CV 0.10
  - Structural balance rule Expected 174.7, Std. Dev. 12.7, CV 0.07
  - Public debt targeting Expected 170.7, Std. Dev. 19.7, CV 0.12
- St. Lucia:
  - Historical behavior Expected 87.1, Std. Dev. 21.1, CV 0.24
  - Structural balance rule Expected 84.0, Std. Dev. 8.4, CV 0.10
  - Public debt targeting Expected 79.8, Std. Dev. 5.7, CV 0.07
- St. Vincent and the Grenadines:
  - Historical behavior Expected 115.0, Std. Dev. 15.7, CV 0.14
  - Structural balance rule Expected 105.2, Std. Dev. 6.5, CV 0.06
  - Public debt targeting Expected 94.9, Std. Dev. 4.3, CV 0.05
- Finding: By 2015, volatility of debt-to-GDP under public debt targeting is generally less than ½ of that under historical behavior and on average about ¾ of the volatility under a structural balance rule.

### Sovereign debt dynamics (Appendix I): key equations and definitions
- Aggregate debt-to-GDP decomposition: d_t = d_t^d + e_t d_t^f + e_t d_t^o. (A.1)
- Domestic-currency debt dynamics (real terms): d_t^d = (1 + i_t^d)/(1 + π_t)/(1 + g_t) d_{t-1}^d + α_t ϕ_t. (A.2)
- Foreign-currency debt dynamics (real terms): d_t^f = (1 + η_t) (1 + i_t^f) (1 + g_t)^{-1} (1 + π_t)^{-1} d_{t-1}^f + β_t ϕ_t. (A.3)
- Official debt dynamics (real-ratio form): d_t^o = (1 + η_t) (1 + i_t^o) (1 + g_t)^{-1} (1 + π_t)^{-1} d_{t-1}^o (1 − α_t − β_t) ϕ_t. (A.4)
- Interest-rate and maturity definitions:
  - Effective interest rates: i_t^d = (1/h^d) Σ_{s=1}^{h^d} z_{t,s}^d, i_t^f = (1/h^f) Σ_{s=1}^{h^f} z_{t,s}^f, i_t^o = (1/h^o) Σ_{s=1}^{h^o} z_{t,s}^o.
  - Commercial foreign one-period rate decomposition: z_{t,1}^f = i_t^* + κ_t^f.
- Allocation of primary balance across debt components (when unavailable): α_t = d_{t-1}^d / d_{t-1}, β_t = d_{t-1}^f / d_{t-1}.

### Other empirical observations, limitations and prerequisites
- Debt creation outside central government: observed changes in central government debt often exceed reported central government deficit plus valuation changes — suggesting important debt creation outside central government (e.g., debt takeovers, other public sector liabilities).
- Limitations and prerequisites:
  - Requires strong political and societal commitment; cannot substitute for lack of political will.
  - Needs developed fiscal institutions: budget processes, fiscal accounting and auditing, public financial management.
  - Framework should cover consolidated government finances or be as aggregated as possible to capture off-central-government debt creation.
  - Budgets need provisions for unforeseen spending (natural disasters); consider a specific counter-cyclical component for natural shocks.
  - Complexity: four intermediate parameters; recommend independent institutional assessment (e.g., Fiscal Council) to evaluate consistency and communicate to public.
  - Institutional arrangements should avoid arbitrary changes in horizon, debt threshold, and probability across cycles.

### Conclusions and policy implications
- Public debt targeting trades off smoothing primary expenditures with avoiding sharp increases in public debt that threaten creditworthiness.
- For the Caribbean sample and chosen parameters:
  - Public debt targeting can significantly reduce primary expenditure procyclicality and reduce volatility of public debt relative to historical behavior.
  - Public debt targeting achieves more stabilization of primary expenditures than a structural balance rule in these simulations, despite designed pro-cyclical components, because accounting for indirect (non-Keynesian) effects reduces overall macro volatility.
- Operational advantages vs. structural balance rule:
  - Eliminates transparency issues of distinguishing permanent vs. transitory shocks.
  - Anchors spending to projected debt dynamics using a broader set of variables (interest rates, real exchange rates).
  - Can be designed to allow constrained cyclical space for discretionary spending or state-contingent additional primary expenditure components.
- Role as transitional framework:
  - Can be used as a transition toward a structural balance rule once public debt and credibility concerns diminish.
- Institutional complement:
  - An independent, empowered fiscal council or equivalent institution could enhance credibility, assess consistency, and inform the public.

*Source: _wp11203 - IMF working paper content provided.*

### 1.  Correlation Coefficients between Country-Specific Financial ....................................13

### 1.  Correlation Coefficients between Country-Specific Financial ....................................13

### Introduction and motivation
- International financial market developments can be a source of instability and can amplify real shocks through:
  - direct impact of capital flows on aggregate demand and domestic financial markets, and
  - multiplication via government access to financing and borrowing costs.
- In high public debt countries, negative real or financial shocks can deteriorate creditworthiness and trigger a second round of effects: high interest rates, credit crunch, recession.
- Keeping public debt at sufficiently low levels is framed as a form of self-insurance: it permits expenditure smoothing over the business cycle by maintaining access to financing without jeopardizing creditor perceptions.
- The paper proposes a fiscal framework to smooth primary expenditure over the business cycle by providing creditors sufficient assurance that public debt obligations will remain within a preannounced band with high probability.

### The proposal: public debt targeting (overview)
- Core idea:
  - Commit to keep public debt within a specified band with high probability.
  - Announce a primary expenditure budget envelope for the next budget cycle such that probabilistic public debt outcomes over the medium term remain within the band.
- Two main public announcements:
  - A trajectory of public debt and its probabilistic band.
  - A primary expenditure budget envelope for the next budget cycle.
- Four intermediate technical parameters (announced and linking the two main announcements):
  - (i) a medium-term horizon in years,
  - (ii) a public debt upper threshold at the end of the medium term horizon,
  - (iii) a notional spending trajectory that maps next years’ primary expenditure announcements to debt trajectories over the medium term,
  - (iv) a probability value attached to debt remaining below the upper threshold.
- Operational recursion:
  - In year t, the government announces S(t) (primary spending budget envelope) such that if primary spending for the following T years remains at the same level in real per capita terms, public debt would take a level D(t+T) percent of GDP or lower with probability P(t+T).
  - In year t+1 the government announces S(t+1) such that after T years public debt takes at most D(t+T+1) with probability P(t+T+1), and so on.
  - Budgets are adjusted each year after observing shocks and their impact on deficits and public debt.
- Discretionary extension:
  - The spending trajectory can incorporate a pre-announced degree of discretionary expenditure, e.g., an impulse (procyclical or countercyclical) tied to the difference between real GDP growth and population growth.

### Rationale and economic logic
- At sufficiently high public debt levels, marginal deficits and debt accumulation can trigger nonlinear second-round effects: interest rate spikes, foreign exchange instability, capital outflows, asset deflation, credit crunch, and uncertainty about taxation.
- These non-Keynesian effects amplify economic cycles; therefore, a framework that provides high-probability assurance that public debt will remain on a non-increasing trend acts as an aggregate demand smoothing device.
- The public debt band is critical because it allows primary balance flexibility and avoids the need for large fiscal consolidations in bad states that would undermine credibility.

### Desirable properties of the framework
1. Credibility
  - Commitment to a public debt band is anchored on a framework linking government spending to projected debt trajectories.
  - Negative shocks that increase deficits and debt relative to budget projections are less likely to trigger expectations of unsustainable debt paths because a compensating consolidation strategy is pre-specified.
  - Budget announcements acknowledge possible shocks and clarify the mechanism by which public debt will be brought back on track.
  - Reputation builds over time as governments abide by announcements, improving access to financing in bad times.
  - Setting primary spending on a smooth trajectory reduces the need for unrealistically large primary spending consolidations in bad states.

2. Primary expenditure-smoothing
  - Debt thresholds set according to medium-term paths permit larger deficits than budgeted if revenues turn out lower while primary expenditure remains at the budget level.
  - Focus on primary spending (excluding interest expenditure) is appropriate because:
    - For a given tax framework, primary expenditures are the relevant aggregate for debt sustainability.
    - Interest expenditure tends to be largely predetermined; excluding it accommodates shocks to interest costs.
  - Expenditure smoothing reduces direct demand fluctuations and can reduce non-Keynesian amplification when creditworthiness deteriorates.
  - Automatic stabilizers on revenues and interest expenditures are allowed to play in full under the framework.

3. Flexibility
  - The framework is not a spending rule; the notional spending trajectory is an intermediate technical parameter rather than a strict commitment to keep primary spending constant in real per capita terms.
  - Primary spending is recalculated recursively each year.
  - Budgets would need sufficient contingent allocations in case primary expenditure stabilizers are triggered (e.g., unexpected increases in social insurance demand in bad states).

4. Medium-term macroeconomic framework
  - Public debt targeting requires fiscal projections over the medium term and probabilistic projections based on an explicit, disclosed economic model.
  - The model must be capable of providing stochastic forecasts for variables relevant to probabilistic public debt projections.
  - Economies that are more vulnerable and volatile require more conservative framework parameters (e.g., a higher primary surplus target or a higher ...).

### Empirical application and illustrative findings
- Empirical simulations for a sample of Caribbean countries indicate:
  - Allowing debt to move within a relatively narrow band permits large improvements in primary expenditure smoothing over the business cycle.
  - In some cases, smoothing under the proposed framework outperforms a structural balance rule.
- The framework is tested in economies that are typically highly indebted, very open, undiversified, and subject to significant real and financial shocks.

### Relation to fiscal procyclicality literature (selected empirical and theoretical points)
- Empirical evidence highlights pervasive procyclical fiscal policy in middle and low income economies:
  - Talvi and Vegh (2000): cyclical components of output and government consumption are positively correlated in every region except G7; same for fiscal revenues.
  - Braun (2001): in a panel of 54 countries, real output growth leads to decline in government expenditure/GDP ratio in OECD but relation is statistically insignificant in developing countries; the increase in the primary surplus from higher GDP growth is three times higher in OECD than in developing countries.
  - Kaminsky, Reinhart and Vegh (2004): cyclical component of government expenditure and output are positively correlated in non-OECD countries across measures and filters; procyclicality stronger in tranquil times.
  - Balassone and Kumar (2007): discretionary fiscal policy has been procyclical in good times in both industrial and developing countries and has adversely affected growth and fiscal sustainability.
- Theoretical perspectives (Box 2) categorize explanations of procyclical fiscal policy into groups opposing normative prescriptions (Keynesian countercyclical, Barro/Lucas tax-smoothing).

*Source: _wp11203 - 1.  Correlation Coefficients between Country-Specific Financial ....................................13*

### 1. Political economy explanations. To this group belong theories based on special interest pressures for

### _wp11203 - 1. Political economy explanations. To this group belong theories based on special interest pressures for

### Political economy explanations
- Theories included:
  - Talvi and Vegh (2000): assume a cost function of running primary fiscal surpluses; interpret this as existence of special interest groups that pressure the government to spend in wasteful allocations.
  - Cole and Kehoe (1992): present a common pool problem explanation for procyclical fiscal policy; static model of public expenditure in a context of special interests’ pressures for spending; expenditure is higher when income is high.
  - Velazco (1998): dynamic version of the special interests model with debt accumulation; public debt grows continuously until it reaches an upper bound, indicating special interests pressures can generate excessive government spending and unsustainable fiscal finances.
  - Roubini (1991) and Grilli, Masciandaro, and Tabellini (1991): analyze institutional distortions affecting the decision horizon framework of government officials.

### Financial markets’ lending behavior along the business cycle
- Key idea:
  - Importance of incorporating financial markets considerations when government finances may be unsustainable.
- Illustrative mechanism:
  - Gavin et al. (1996): explanation of procyclical fiscal policy based on financing constraints; interest rates offered to governments tend to be high in low states of the cycle and low during booms.

### Public debt targeting and structural balance rules
- Definition and framing:
  - Kopits and Symansky (1998): fiscal rules as numerical constraints set on the budget on a permanent basis. The paper’s proposal is framed as a stable framework (choices of four intermediate medium-term parameters and the economic model) rather than fixed numerical targets.
  - Under public debt targeting, primary expenditure and debt levels change in every budget cycle as the budget envelope for the following year is recalculated after observing shocks’ impact on public debt.
- Comparison with alternative rules:
  - Numerical constraints (e.g., Maastricht Treaty) are simple and potentially useful under strong political economy pressures, but may not avoid procyclical spending or excessive debt accumulation.
  - Structural or cyclically-adjusted balance rules:
    - Advantages: account for transitory impact of business cycle fluctuations; if targets set appropriately, consistent with debt sustainability and a nonprocyclical stance.
    - Limitations: (a) difficulty discriminating permanent vs. transitory changes in revenues and expenditures (transparency and credibility risks), (b) insufficiency of GDP to assess full state of the cycle, (c) do not allow space for discretionary fiscal policy.
    - Refinements can mitigate limitations (ex-post compensation mechanisms, broader set of variables, inclusion of discretionary expenditure attached to over-the-cycle compensatory mechanism).
- Advantages of public debt targeting (per paper):
  - No need to identify permanent and transitory output components.
  - Spending levels determined based on projected public debt dynamics, anchored around the debt accumulation identity.
  - Forces use of a broader set of variables for debt projections (interest rates, real exchange rates) important for assessing the business cycle.
  - Tackles impact of debt sustainability concerns on the business cycle directly via commitment to public debt thresholds and announcement of a debt trajectory band.
  - Can be designed to allow cyclical space for discretionary primary spending, including state-contingent additional primary expenditure components.
- Role as transition framework:
  - Public debt targeting seen as a promising transition towards adoption of structural balance rules when public debt is high and marginal deficits and public debt accumulation can impact sovereign spreads.
  - Use of a debt band balances expenditure smoothing with need to avoid sharp increases in public debt.
  - Simple deterministic debt rules may lack credibility under tail risks and could force strongly procyclical fiscal stances.

### Application to Caribbean countries — Stylized facts
- Rationale: Caribbean economies are candidates for the proposal due to high public debt levels and procyclical primary expenditure; empirical evidence links financial developments to this behavior.
- Stylized facts (empirical indicators):
  - High public debt-to-GDP ratios:
    - Between 2008 and 2009 the median public debt-to-GDP ratio increased from 81 percent to 98 percent.
    - In 2009 public debt showed a magnitude of about 4 times the size of total revenues.
    - For almost all countries the 2010 estimates show a further increase in the public debt-to-GDP ratio.
  - Procyclical primary expenditure:
    - Median cyclical components of real output, primary expenditures and total revenues show positive co-movement over the last 15 years.
    - The cyclical components’ median correlation of primary expenditures with both total revenues and real GDP is about 50 percent.
  - Strong cross-country synchronization of GDP, government revenues and primary expenditures:
    - Virtually all countries show a positive correlation of the real output, total revenue and primary expenditure cycles with the cross-country median fluctuations.
  - Strong cross-country correlation of domestic financial and foreign exchange markets, possibly corresponding to the state of the US business cycle:
    - Every country’s GDP cycle is positively correlated with that of the US.
    - Evidence of cross-country correlation in interest rates, real exchange rates and net capital inflows.
  - Financial variables explain a substantial share of the variation in primary expenditures:
    - Net capital inflows, real exchange rate depreciation and interest rates jointly explain ½ of the average forecast-error variance of primary expenditures.
    - Note on methodology: individual-country forecast variance decompositions are derived from a two-lag six-variable VAR featuring annual observations of: output growth rate, a short-term interest rate, the growth rate of the real exchange rate, the change in net capital inflows, total revenue growth and primary expenditure growth.

### Simulation framework (public debt accumulation)
- Debt decomposition:
  - Public debt-to-GDP decomposed into:
    - (i) commercial component denominated in the domestic currency;
    - (ii) commercial component denominated in foreign currency;
    - (iii) official bilateral/multilateral fraction of public debt (usually denominated in US dollars).
- Dynamic equation for component d_t^j:
  - d_t^j evolves according to:
    - d_t^j = (1 + i_t^j) / (1 + π_t) * (1 + g_t)^{-1} * d_{t-1}^j + α_t^j * b_t
      (presented in the text as an evolution equation where i_t^j, g_t, π_t, α_t^j, and b_t are defined).
- Variable definitions (as given in the source):
  - i_t^j: interest rate paid on a government debt of type j.
  - g_t: real GDP growth rate.
  - π_t: annual inflation rate.
  - α_t^j: share of primary balance servicing public debt of type j.
  - b_t: primary balance-to-GDP ratio (difference between total revenues and primary (non-interest) expenditures divided by GDP).
  - Total debt-to-GDP ratio for a year: d_t = sum_j d_t^j.

*Source: _wp11203 - 1. Political economy explanations. To this group belong theories based on special interest pressures for (IMF working paper content provided).*

### Appendix A provides a full description of how we have computed and decomposed public

### _wp11203 - Appendix A provides a full description of how we have computed and decomposed public

### Debt decomposition and interest-rate specification
- Decomposition of public debt-to-GDP in the simulations allows a more realistic evaluation of the sensitivity of debt indicators to interest rate and exchange rate shocks than maintaining assumptions of fixed shares.
- The interest rate specification used in (1) is the implicit or average rate for the debt category (not the marginal rate).
- Implicit/average rates are approximated by taking the moving average of the marginal interest rate in year t and the previous M-1 years, where M is the average maturity of the debt category measured in years.

### Model and estimation framework
- Inputs for debt simulations according to (1) and (2) are obtained from stochastic simulations that capture volatility, correlation and persistence of required inputs and other business-cycle variables.
- Simulations are produced by fitting historical data to a VAR model and using the coefficients to produce a large number of stochastic projections.
- VAR specification (3):
  - Xt = AL(L) Xt-1 + B i*t + εt
  - Xt = (gt, it, dt, ηt, mt, τt, st)T is a 6×1 vector of endogenous variables containing:
    - real output growth,
    - a short-term market-determined domestic interest rate,
    - the change in the bilateral real exchange rate (versus the US dollar),
    - the change in real net capital inflows,
    - the percentage change in total revenues,
    - the percentage change in primary expenditures.
  - i*t is the US federal funds rate.
  - AL(L) is a polynomial in the lag operator of degree ranging from 1 to 4, depending on the country estimation output.
  - B is a vector of coefficients; εt ~ N(0, Ω).
- Rationale:
  - Employing growth rates helps deal with stationary variables and avoids explicit assumptions on trends.
  - Main differences relative to earlier VAR-based literature:
    - explicit introduction of financial variables (net capital inflows and the foreign interest rate) relevant for Caribbean countries,
    - distinct treatment of total revenues and primary expenditures so that the primary balance is computed as a second step from simulated growth rates of revenues and expenditures.
- Caveat:
  - The simulation results are subject to the Lucas critique; VAR-based policy simulations can be biased if behavior changes with policy. A general equilibrium model can address this limitation.

### Monte Carlo simulation procedure
- Use the fitted VAR model to simulate the effect of expenditure-smoothing fiscal policies (active).
- For comparison, simulate primary expenditure using the VAR-fitted equation for primary expenditures (passive).
- Passive simulations provide an agnostic diagnosis of prevailing public debt sustainability conditions.
- Simulation draws:
  - Generate a sequence of random vectors ˆεt+1,...,ˆεT such that ∀ω ∈ t+1,T, ˆεω = W υω, where υω ~ N(0,1) and W is the Choleski factorization of Ω = W'W.
  - At every forecast period draw 2000 random vectors ˆεω while the VAR produces joint dynamic responses of all elements in X.
- Ordering sensitivity:
  - Passive simulations are not sensitive to variable ordering because results are shaped by Ω, which is unique.
  - Active simulations are sensitive to ordering; setting primary expenditures last minimizes contemporaneous (within-year) feedback from primary expenditures to the rest of the economy.

### Fiscal policy framework inputs and procedure
- Two required inputs to compute simulations for the proposed fiscal policy framework:
  1. Starting point of the primary expenditure path (the resulting level represents the primary expenditure envelope for the next budget cycle).
  2. Specification of the primary expenditure pattern going forward to compute public debt projections over the medium term.
- Forward-looking primary expenditure assumption:
  - Primary expenditure remains constant in real per-capita terms (see equation (4) below) except for the initial consolidation and feedback via responsiveness parameter κ.
- Implementation across the Monte Carlo draws:
  - For each of the 2000 random draws, the debt accumulation identity is applied to produce a distribution of public debt outcomes at each projected horizon.
  - The starting level of primary expenditure is pinned down iteratively by backward induction after the policymaker chooses desired debt upper threshold d, horizon length T, and probability p.
  - Denote ̃st+1 the primary expenditure level consistent with dt+T ≤ d with probability p.
- Interpretation of p:
  - Setting p = 0.5 yields the level of primary expenditure that stabilizes public debt to GDP in expected terms over the medium-term.
  - Higher p implies additional primary expenditure consolidation and a downward expected trend in public debt; a higher probability is more prudent.
- Role of horizon T:
  - Longer T smooths the primary expenditure path because adjustments are averaged over more years.
  - Too long a horizon could undermine credibility.

### Forward-looking expenditure-smoothing specification
- Specification (4):
  - sω = ̃sω-κ(gω - g), ∀ω ∈ t+1,T
  - g is a “trend” growth rate of real GDP, set equal to the average population growth rate.
  - gω is the simulated growth rate of real GDP.
  - κ is a parameter representing the responsiveness of the primary expenditure growth rate to deviations of real GDP growth from trend growth.
  - ̃st+1 represents the consolidation (if ̃st+1 < 0) of primary expenditures at time t+1, while ̃sω = g, ∀ω ∈ t+2,T.
- Interpretation of κ:
  - κ = 0 corresponds to no discretionary expenditure impulse: primary expenditure remains constant in real per-capita terms after the initial consolidation in year t+1. For κ = 0, the primary expenditure level at t+1 is set so that if primary expenditure remains constant in real per-capita terms for T years, public debt will remain below chosen upper threshold d with probability p in year t+T.
  - Positive κ represents a countercyclical fiscal policy impulse (additional primary expenditure growth of opposite sign to real per-capita GDP growth).
  - Negative κ represents a procyclical fiscal policy impulse.

### Example: St. Lucia illustration
- Figure 6 (described) illustrates the budget-setting implications based on the described simulation for St. Lucia:
  - Left: historical observations.
  - Right: one random draw from the 2000 stochastic simulations where a simulated random shock negatively affects real GDP growth and the passive fiscal policy shows procyclical behavior observed historically.
  - Three alternative active expenditure-smoothing paths shown based on equation (4) for κ ∈ { -0.5; 0; 0.5 }.
- Policymaker parameter choices in the example:
  - Horizon T = 5 years.
  - Upper public debt threshold = 70 percent (estimated debt to GDP in 2010 is 80 percent).
  - Probability p = 0.7.
  - This parameter choice ensures debt will remain on a declining path with relatively high assurance by construction.
- Result reported in the source (partial):
  - With this parameter choice, the level of primary expenditure consolidation required in the next budget cycle ranges from 1.7 percentage points of GDP under the pro-cyclical choice, to
  - (text ends here in the source; no further numeric continuation is provided)

*Source: _wp11203 - Appendix A provides a full description of how we have computed and decomposed public*

### 3.4 percentage points of GDP under the counter-cyclical choice, with the acyclical impulse

### _wp11203 - 3.4 percentage points of GDP under the counter-cyclical choice, with the acyclical impulse

### Illustration and simulation setup
- Figure 6 illustrates three primary expenditure patterns (Passive, Acyclical, Countercyclical) using St. Lucia data.  
- Simulations: Monte Carlo with 2,000 draws; medium-term horizon of five years (2011–2015).  
- Policymaker announcements: medium-term debt trajectory with probabilistic bands and next year’s primary expenditure envelope.  
- Four technical parameters linking announcements: (i) illustrative upper debt-to-GDP threshold, (ii) horizon in number of budget cycles, (iii) probability attached to remaining below threshold, (iv) notional primary expenditure trajectory through the horizon.  
- Comparators: (a) historical behavior from unrestricted VAR; (b) structural primary balance rule targeting the same public debt-to-GDP ratio (with output gap elasticities of revenues and primary expenditures of 1 and 0, respectively). GDP gaps estimated using Hodrick-Prescott filter (lambda = 100).

### Main simulation results (Table 2 highlights and country-specific adjustments)
- Jamaica:
  - Estimated debt/GDP in 2010: 139%
  - Estimated primary balance/GDP in 2010: 7.6%
  - Expected debt/GDP in 2015 under "passive" behaviour: 117%
  - Illustrative debt/GDP threshold in 2015: 100%
  - Probability debt/GDP < threshold: 50% and 70% rows yield:
    - Required consolidation of primary expenditures in 2011 (in percentage of GDP): 50% row: -4.2, -4.0, -4.4; 70% row: -0.7, -0.7, -0.7
  - Note: Jamaica is the only country in the sample that does not show a particular need for further consolidation in 2011 under these assumptions.
- Antigua and Barbuda:
  - Estimated debt/GDP in 2010: 119%
  - Estimated primary balance/GDP in 2010: -1.8%
  - Expected debt/GDP in 2015 under "passive": 160%
  - Illustrative threshold 2015: 120%
  - Required consolidation in 2011: 50% row: 8.3, 9.1, 7.6; 70% row: 10.3, 11.1, 9.4
- The Bahamas:
  - Estimated debt/GDP in 2010: 65%
  - Estimated primary balance/GDP in 2010: -3.9%
  - Expected debt/GDP in 2015 under "passive": 79%
  - Illustrative threshold 2015: 60%
  - Required consolidation in 2011: 50% row: 0.8, 1.1, 0.4; 70% row: 1.9, 2.4, 1.3
- Barbados:
  - Estimated debt/GDP in 2010: 122%
  - Estimated primary balance/GDP in 2010: -1.2%
  - Expected debt/GDP in 2015 under "passive": 137%
  - Illustrative threshold 2015: 130%
  - Required consolidation in 2011: 50% row: 7.8, 7.6, 7.9; 70% row: 10.3, 10.4, 10.2
- Grenada:
  - Estimated debt/GDP in 2010: 116%
  - Estimated primary balance/GDP in 2010: 0.3%
  - Expected debt/GDP in 2015 under "passive": 135%
  - Illustrative threshold 2015: 110%
  - Required consolidation in 2011: 50% row: 1.3, 0.7, 1.4; 70% row: 3.7, 3.2, 3.7
- St. Kitts and Nevis:
  - Estimated debt/GDP in 2010: 184%
  - Estimated primary balance/GDP in 2010: -1.9%
  - Expected debt/GDP in 2015 under "passive": 205%
  - Illustrative threshold 2015: 180%
  - Required consolidation in 2011: 50% row: 11.1, 11.7, 10.5; 70% row: 13.6, 14.3, 13.1
- St. Lucia:
  - Estimated debt/GDP in 2010: 80%
  - Estimated primary balance/GDP in 2010: -3.2%
  - Expected debt/GDP in 2015 under "passive": 87%
  - Illustrative threshold 2015: 70%
  - Required consolidation in 2011: 50% row: 0.7, 1.3, 0.1; 70% row: 2.5, 3.4, 1.7
- St. Vincent and the Grenadines:
  - Estimated debt/GDP in 2010: 91%
  - Estimated primary balance/GDP in 2010: -1.4%
  - Expected debt/GDP in 2015 under "passive": 115%
  - Illustrative threshold 2015: 90%
  - Required consolidation in 2011: 50% row: 6.1, 5.8, 6.4; 70% row: 7.7, 7.5, 7.9
- Summary: For Bahamas, Grenada, and St. Lucia a primary expenditure consolidation of around 2 to 3 percentage points of GDP in 2011 together with acyclical/moderately procyclical/countercyclical rules suffices to keep 2015 debt below 2010 levels in 70 percent of simulations. Antigua and Barbuda, Barbados, St. Kitts and Nevis, and St. Vincent and the Grenadines require consolidations in excess of 7 percentage points of GDP in 2011 per the framework.

### Primary expenditure smoothing (Table 3)
- Coefficient of Variation (CV) of simulated real growth rates of primary expenditures (average 2011–15) by policy:
  - Antigua and Barbuda: Historical behavior 25.7; Structural balance rule 17.6; Public Debt Targeting 3.4
  - The Bahamas: Historical behavior 1.5; Structural balance rule 1.3; Public Debt Targeting 2.3
  - Barbados: Historical behavior 3.3; Structural balance rule 1.3; Public Debt Targeting 1.9
  - Grenada: Historical behavior 91.9; Structural balance rule 5.8; Public Debt Targeting 2.7
  - Jamaica: Historical behavior 2.3; Structural balance rule 2.2; Public Debt Targeting 1.7
  - St. Kitts and Nevis: Historical behavior 1.5; Structural balance rule 3.0; Public Debt Targeting 2.7
  - St. Lucia: Historical behavior 3.5; Structural balance rule 2.2; Public Debt Targeting 1.5
  - St. Vincent and the Grenadines: Historical behavior 6.1; Structural balance rule 2.4; Public Debt Targeting 1.8
- Finding: The CV of primary expenditure growth under public debt targeting is generally lower than under a structural balance rule for this parameterization (higher only in two out of eight cases, and not significantly).

### Cyclical properties of primary expenditures (Tables 4 and 5)
- Coefficient of Variation of cyclically-adjusted primary balances (average 2011–15) — select averages:
  - Antigua and Barbuda: Structural balance rule average 0.69; Public debt targeting average 0.50
  - The Bahamas: Structural balance rule average 3.00; Public debt targeting average 1.09
  - Barbados: Structural balance rule average 1.51; Public debt targeting average 3.23
  - Grenada: Structural balance rule average 0.25; Public debt targeting average 0.29
  - Jamaica: Structural balance rule average 0.46; Public debt targeting average 0.23
  - St. Kitts and Nevis: Structural balance rule average 0.58; Public debt targeting average 2.10
  - St. Lucia: Structural balance rule average 4.94; Public debt targeting average 0.85
  - St. Vincent and the Grenadines: Structural balance rule average 1.43; Public debt targeting average 0.85
- Interpretation: For five out of eight cases, the structural balance rule shows more variability in cyclically-adjusted primary balances (implying more shock-absorbing capacity) compared to public debt targeting; however, combined with Table 3 results, revenues and/or trend growth rates tend to be more stable or higher under debt targeting, lowering CVs.
- Correlations of real growth rates of primary expenditures with GDP and with revenues (Table 5, averages):
  - Correlations with GDP (Debt Targeting averages):
    - Antigua and Barbuda: 0.28
    - The Bahamas: 0.29
    - Barbados: 0.17
    - Grenada: -0.08
    - Jamaica: 0.44
    - St. Kitts and Nevis: 0.02
    - St. Lucia: -0.43
    - St. Vincent and the Grenadines: -0.16
  - Correlations with revenues (Debt Targeting averages):
    - Antigua and Barbuda: 0.70
    - The Bahamas: 0.37
    - Barbados: 0.32
    - Grenada: 0.46
    - Jamaica: 0.56
    - St. Kitts and Nevis: 0.66
    - St. Lucia: 0.77
    - St. Vincent and the Grenadines: 0.32
- Finding: Average correlation between primary expenditure and revenues under public debt targeting in this sample is 52 percent for the parameter choice used. A positive revenue-expenditure correlation implies some pro-cyclical adjustment by design to avoid excessive debt accumulation.

### Public debt smoothing (Table 6 and fan charts)
- By 2015, volatility of debt-to-GDP ratio under public debt targeting is:
  - Generally less than ½ of that under historical behavior.
  - On average about ¾ of the volatility under a structural balance rule (coefficient of variation comparison).
- Table 6 selected entries (Expected, Std. Dev., CV by 2015):
  - Antigua and Barbuda:
    - Historical behavior Expected 160.4, Std. Dev. 42.3, CV 0.26
    - Structural balance rule Expected 120.9, Std. Dev. 18.0, CV 0.15
    - Public debt targeting Expected 114.3, Std. Dev. 11.0, CV 0.10
  - The Bahamas:
    - Historical behavior Expected 79.4, Std. Dev. 12.5, CV 0.16
    - Structural balance rule Expected 64.3, Std. Dev. 6.7, CV 0.10
    - Public debt targeting Expected 60.9, Std. Dev. 5.3, CV 0.09
  - Barbados:
    - Historical behavior Expected 136.8, Std. Dev. 15.9, CV 0.12
    - Structural balance rule Expected 125.6, Std. Dev. 12.8, CV 0.10
    - Public debt targeting Expected 125.3, Std. Dev. 10.8, CV 0.09
  - Grenada:
    - Historical behavior Expected 134.6, Std. Dev. 37.9, CV 0.28
    - Structural balance rule Expected 108.9, Std. Dev. 7.9, CV 0.07
    - Public debt targeting Expected 111.3, Std. Dev. 6.4, CV 0.06
  - Jamaica:
    - Historical behavior Expected 116.9, Std. Dev. 26.0, CV 0.22
    - Structural balance rule Expected 139.1, Std. Dev. 11.3, CV 0.08
    - Public debt targeting Expected 110.9, Std. Dev. 5.4, CV 0.05
  - St. Kitts and Nevis:
    - Historical behavior Expected 205.3, Std. Dev. 19.9, CV 0.10
    - Structural balance rule Expected 174.7, Std. Dev. 12.7, CV 0.07
    - Public debt targeting Expected 170.7, Std. Dev. 19.7, CV 0.12
  - St. Lucia:
    - Historical behavior Expected 87.1, Std. Dev. 21.1, CV 0.24
    - Structural balance rule Expected 84.0, Std. Dev. 8.4, CV 0.10
    - Public debt targeting Expected 79.8, Std. Dev. 5.7, CV 0.07
  - St. Vincent and the Grenadines:
    - Historical behavior Expected 115.0, Std. Dev. 15.7, CV 0.14
    - Structural balance rule Expected 105.2, Std. Dev. 6.5, CV 0.06
    - Public debt targeting Expected 94.9, Std. Dev. 4.3, CV 0.05
- Fan charts (Figures 7 and 8) show projected public debt trajectories and dispersion under passive vs. acyclical policy with country-specific initial consolidations; Figure 8 illustrates public debt dispersion when public debt targeting is applied recursively over budget cycles.

### Other empirical observations
- Figure 9 (Debt creation outside central government) shows that for all sample Caribbean countries observed changes in central government debt have been significantly larger than reported central government deficit plus valuation changes — suggesting important other sources of debt creation (e.g., debt takeovers or liabilities from other public sector bodies) beyond measurement issues like pre-financing or government deposits.

### Limitations and prerequisites for implementation
- Political economy:
  - Requires sufficiently strong political and societal commitment to fiscal prudence; framework cannot substitute for lack of political will.
  - Effectiveness depends on capacity to contain special interests, distributional conflicts, and political pressures.
- Institutional and technical preconditions:
  - Need sufficiently developed fiscal institutions (budget processes, fiscal accounting and auditing, public financial management).
  - Framework should cover consolidated government finances or as-aggregated-as-possible definitions to capture debt creation outside central government.
  - Budgets need provisions for unforeseen spending (e.g., natural disasters); consideration for a specific counter-cyclical component for natural shocks may be warranted.
- Complexity and transparency:
  - Framework has some technical complexity (four intermediate parameters). Main public announcements are simple: debt trajectory with probability bands and next year’s primary expenditure envelope.
  - Independent institutional assessment (e.g., a Fiscal Council) recommended to evaluate consistency and communicate to the public.
- Stability of parameters:
  - Institutional arrangements should avoid arbitrary changes in horizon, debt threshold level, and probability across budget cycles.

### Conclusions and policy implications
- The public debt targeting framework asks policymakers to trade off smoothing primary expenditures with avoiding sharp variations in public debt that threaten creditworthiness.  
- For the Caribbean sample and chosen parameters:
  - Public debt targeting can significantly reduce primary expenditure procyclicality and reduce volatility of public debt relative to historical behavior.
  - Public debt targeting achieves more stabilization of primary expenditures than a structural balance rule in these simulations, despite its designed pro-cyclical components, because accounting for indirect (non-Keynesian) effects reduces overall macro volatility.
- Operational advantages vs. structural balance rule:
  - Eliminates transparency issues of distinguishing permanent vs. transitory shocks.
  - Anchors spending to projected debt dynamics via the debt accumulation identity and uses a broader set of variables (interest rates, real exchange rates).
  - Can be designed to allow a constrained cyclical space for discretionary spending or state-contingent additional primary expenditure components.
- Role as transitional framework:
  - Public debt targeting can be used as a transitional framework toward a structural balance rule once public debt and credibility concerns diminish.
- Institutional complement:
  - An independent, empowered fiscal council or equivalent institution could enhance credibility, assess consistency, and inform the public.

*Source: Country authorities’ and authors’ estimates and projections (excerpted from the provided IMF working paper content).*

### APPENDIX I. SOVEREIGN DEBT DYNAMICS

### APPENDIX I. SOVEREIGN DEBT DYNAMICS

### Debt-to-GDP decomposition
- The stock of debt at the end of period t, denoted by D_t, equals the sum of:
  - domestic-currency debt, B_t^d;
  - foreign-currency debt, B_t^f, converted at the nominal exchange rate e_t (units of domestic currency per unit of foreign currency);
  - official debt, B_t^o.
- In ratio-to-GDP terms (dividing by nominal GDP Y_t):  
  d_t = d_t^d + e_t d_t^f + e_t d_t^o.  (A.1)
- Definitions used in ratios:  
  d_t^d = B_t^d / Y_t, d_t^f = B_t^f / Y_t, d_t^o = B_t^o / Y_t.

### Domestic-currency-denominated debt dynamics (flow budget identity)
- Government flow budget identity for domestic debt:  
  B_t^d = (1 + i_t^d) B_{t-1}^d + α_t P_t,  
  where i_t^d is the interest rate on local-currency debt, P_t is the primary balance, and α_t ∈ (0,1) is the fraction of primary balance used to service domestic debt exposure.
- Primary balance definition: P_t = T_t − G_t, where T_t is revenue and G_t is non-interest public spending.
- In ratio-to-GDP form (dividing by Y_t):  
  d_t^d = (1 + i_t^d)/(1 + y_t) d_{t-1}^d + α_t π_t,  
  where y_t is the nominal growth rate of GDP and π_t is the ratio of primary balance to GDP.  (text immediately following shows variant in real terms)
- Expressed using real growth rate g_t deflating by inflation π_t:  
  d_t^d = (1 + i_t^d)/(1 + π_t)/(1 + g_t) d_{t-1}^d + α_t ϕ_t.  (A.2)  

(Note: the source denotes ݕ_t as growth rate of nominal GDP and ܾ_t as ratio of primary balance to GDP; the text converts to real growth g_t by deflating with inflation π_t.)

### Foreign-currency-denominated debt dynamics
- First-order difference equation for foreign-currency debt:  
  B_t^f = (1 + i_t^f) B_{t-1}^f + β_t P_t,  
  where i_t^f is the nominal interest rate on foreign-currency bonds and β_t ∈ (0,1) is the fraction of primary balance used to service foreign commercial debt exposure.
- Multiplying by e_t and dividing by Y_t yields:  
  e_t B_t^f / Y_t = (e_t/e_{t-1}) (Y_{t-1}/Y_t) (e_{t-1} B_{t-1}^f / Y_{t-1}) (1 + i_t^f) + (e_t P_t / Y_t) β_t.
- In ratio form with notation for nominal exchange rate depreciation ω_t and nominal growth y_t:  
  d_t^f = (1 + ω_t)(1 + i_t^f)/(1 + y_t) d_{t-1}^f + β_t ϕ_t.  
- Deflated into real terms using real growth rate g_t, real exchange rate depreciation η_t, and real interest rate on foreign commercial debt (1 + i_t^f)/(1 + π_t):  
  d_t^f = (1 + η_t) (1 + i_t^f) (1 + g_t)^{-1} (1 + π_t)^{-1} d_{t-1}^f + β_t ϕ_t.  (A.3)

### Official debt dynamics
- The dynamics of official debt-to-GDP ratio is analogous to the foreign-currency debt equation but typically with a lower interest rate i_t^o charged by bilateral or multilateral lenders.
- In real-ratio form given in the source:  
  d_t^o = (1 + η_t) (1 + i_t^o) (1 + g_t)^{-1} (1 + π_t)^{-1} d_{t-1}^o (1 − α_t − β_t) ϕ_t.  (A.4)
  (The source indicates the interest rate applied by official lenders, i_t^o, is typically lower than the market rate.)

### Interest rates, spreads, and maturities
- Average maturities of debt classes: h^d, h^f, h^o for domestic, foreign, and official debt respectively.
- One-period government bond rates for each class: z_{t,1}^d, z_{t,1}^f, z_{t,1}^o.
- Effective interest rates computed as averages across maturities:  
  i_t^d = (1/h^d) Σ_{s=1}^{h^d} z_{t,s}^d,  
  i_t^f = (1/h^f) Σ_{s=1}^{h^f} z_{t,s}^f,  
  i_t^o = (1/h^o) Σ_{s=1}^{h^o} z_{t,s}^o.
- Nominal one-period commercial foreign-currency rate z_{t,1}^f decomposed as:  
  z_{t,1}^f = i_t^* + κ_t^f,  
  where i_t^* is an international risk-free rate and κ_t^f is the country’s sovereign risk spread.
- Similarly for official one-period rate: z_{t,1}^o = i_t^* + κ_t^o, with κ_t^o typically lower than κ_t^f.

### Allocation of primary balance across debt components
- When information on the fraction of primary balance servicing each component is not available, the source assumes proportional allocation by existing debt shares:  
  α_t = d_{t-1}^d / d_{t-1},  β_t = d_{t-1}^f / d_{t-1}.
- Summing equations (A.2) through (A.4) yields the aggregate debt-to-GDP identity in (A.1).

*Source: APPENDIX I. SOVEREIGN DEBT DYNAMICS (from the provided IMF content unit).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11203.pdf_