## _sdn1311

## Source details

**Canonical URL:** [_sdn1311](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2013/_sdn1311.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2013/_sdn1311.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2013/_sdn1311.pdf.json)

---

### Overview
- International macro policy coordination has been rare; successful episodes occurred when the world economy seemed on the brink of collapse (examples cited: the 1978 Bonn Summit, the 1985 Plaza Agreement, the 1987 Louvre Accord, the 1987 stock market crash coordinated interest rate cuts and liquidity provision, and the 2008 global financial crisis coordinated fiscal expansions).
- In ordinary times policymaking tends to be national rather than multilateral despite theoretical reasons and systemic stresses arguing for coordination.
- The case for coordination is grounded in standard welfare economics: policymaking involves tradeoffs across targets, and absent coordination externalities from cross-border spillovers imply Pareto-inefficient outcomes.

### Key findings on why coordination is uncommon
- Three principal impediments:
  - Policymakers often do not think in terms of trade-offs across objectives; parties fixate on single macroeconomic goals and ignore other objectives.
  - Disagreement about the economic situation and cross-border transmission effects (“model uncertainty” or deliberate “model disagreements”) raises potential gains from coordination while making agreements harder to reach and sustain.
  - Asymmetries in country size, where gains from coordination may accrue to countries too small to be included in any agreement.
- Noncooperative equilibrium modeled as Nash equilibrium where authorities maximize national welfare taking other countries’ policies as given; this equilibrium is generally Pareto-inefficient because coordinated perturbations can yield first-order gains to the home country and only second-order losses abroad.
- Coordination is fragile because individual countries have first-order incentives to “cheat”; absent international sanctions, cooperation must be sustained by the implicit threat of future refusal to coordinate.

### Cross-border spillovers and welfare implications
- Policy channels and tradeoffs:
  - Monetary easing: domestic tradeoff (boost output vs. greater inflation/financial stability risks); foreign effects include positive demand-for-exports channel and negative exchange-rate appreciation channel; net foreign impact can be negative.
- Noncooperative bias:
  - In noncooperative equilibrium excessive stimulus can occur because countries ignore negative externalities; coordinated internalization can yield first-order gains to each party when both move toward cooperative equilibrium.
- Magnitude of gains:
  - Gains from coordination are not huge but are measurable and “very similar to the estimated gains from global trade liberalization.”

### Proposals to enhance prospects for cooperation
- Proposal 1: A neutral assessor
  - Rationale: Uncertainty and disagreements impede cooperation; a neutral, credible, impartial assessor can bridge divergent views.
  - Role: Present analyses of alternative strategies and resulting tradeoffs (not necessarily propose policies), enabling countries to judge quid pro quos underpinning coordination.
  - Advantage: IMF bilateral surveillance can underscore tradeoffs across goals and help identify coordinated policy packages that are welfare superior.
- Proposal 2: Two guideposts to limit egregious negative spillovers
  - Purpose: Provide safeguards when coordination is impossible or fails to account for spillovers on “small” countries.
  - Scope: Establish guideposts limiting the most egregious negative spillovers through countries’ current account and capital account, respectively.
  - Design: Specifics to be decided by the international community; logic is to press countries to abjure policies with large negative cross-border spillovers even at some domestic cost.

### Relationship to existing IMF processes
- Both proposals build on existing processes:
  - IMF surveillance aims for objective analysis and “ruthless truth-telling” to overcome biases in country self-assessments.
  - The Integrated Surveillance Decision urges countries to consider policies that engender less adverse outward spillovers while achieving domestic objectives.
  - Proposed guideposts build on the Integrated Surveillance Decision and would press countries to reject policies with large negative cross-border spillovers (through trade or financial flows) even if there is some domestic cost.

### Structure of the full paper
- Section II: Theory of international policy coordination and overview of reasons episodes of coordination are rare.
- Section III: Survey of evidence on cross-border spillovers and policy transmission effects.
- Section IV: How uncertainty raises both the gains from, and obstacles to, successful coordination.
- Section V: Ways of reducing obstacles to coordination.
- Section VI: Conclusion.

### Box 1 — The Theory of International Policy Coordination (framework and implications)
- Framework and equilibrium concepts:
  - Two symmetric countries, targets (y1, y2) affected by domestic and foreign policies:
    - y1 = α11 m1 + β12 m2
    - y2 = α22 m2 + β21 m1
    - α are domestic multipliers; β are transmission multipliers.
  - Nash (noncooperative) equilibrium: MRS = MRT using domestic multipliers; formally [ (∂^2 v / ∂y1 ∂y2) / (∂^2 v / ∂y1^2) ] = - (α12 / α11 ).
  - Cooperative (global planner) equilibrium: planner maximizing v_c = 0.5 v_home + 0.5 v_foreign sets MRS = MRT using domestic plus transmission multipliers; formally [ (∂^2 v / ∂y1 ∂y2) / (∂^2 v / ∂y1^2) ] = - (α12 + β12) / (α11 + β11 ).
  - Analogy: Nash ~ autarky; coordination ~ free trade.
- Monetary policy example (preserved comparative statics):
  - Targets: output y and inflation π with instrument m:
    - y = α11 m + β12 m* + ε
    - π = α2 m
    - Objective v(y,π) = (y^2) - (1/2) ω π^2; α11 > 0, α2 > 0, β12 may be <0 or >0.
  - Comparative statics:
    - Monetary policy too expansionary in noncooperative equilibrium (m_NC > m_C) if β1 < 0 (negative transmission).
    - Monetary policy insufficiently expansionary in noncooperative equilibrium (m_NC < m_C) if β1 > 0 (positive transmission).
  - Interpretation:
    - β1 < 0: cooperation implies less stimulus, lower output, higher welfare due to lower inflation/financial-stability risk.
    - β1 > 0: cooperation implies greater stimulus, higher output, with financial-stability risk.

### Box 1 — Why coordination gains may be limited or episodic (six reasons)
- 1) Policymakers may believe spillovers are too small to offset coordination costs; historical models varied but growing linkages likely raised transmission multipliers.
- 2) Policymakers must face real trade-offs (fewer instruments than targets) for coordination to help; post-crisis usable instruments are diminished and attention to additional targets has increased.
- 3) Coordination can provide quid pro quos to overcome domestic constraints if trust that foreign quid pro quos will be delivered.
- 4) Nature of shocks matters: gains require economy sufficiently off desired path and policy effectiveness.
- 5) Distribution of gains may be asymmetric; many gains may accrue to small non-participating countries.
- 6) Uncertainty about state of economy or policy effects both raises gains from coordination and complicates agreement formation and sustainment; strategic misrepresentation impedes cooperation.

### Box 1 — Empirical notes on transmission multipliers and gains
- Early literature found modest gains (example: Oudiz and Sachs (1984) estimate utility equivalent of one-half percentage point of GNP in each of the next few years from coordinated expansion).
- Accounting for uncertainty can materially increase estimated gains (Ghosh and Masson (1988) find incorporating uncertainty roughly doubles estimated gains).
- Evolution of transmission multipliers:
  - Early averages suggested transmission multipliers ~ one tenth the size of domestic multipliers; in absolute value often one-third to one-half the size of domestic multipliers.
  - Recent evidence suggests transmission multipliers have grown and are now about one half the size of domestic multipliers.
  - For the United States, recent estimates (IMF, 2013b) suggest fiscal policy transmission multipliers as high as 60 percent of domestic multipliers.
  - Monetary policy transmission multipliers are about 40 percent as large as domestic multipliers.
  - Financial linkages explain about one third of cross-country variation in multipliers; trade linkages explain about 10 percent (or more for fiscal shocks).
- Crisis vs quiet times and unconventional policies:
  - Transmission multipliers may be larger in crisis periods.
  - Effects of quantitative easing episodes are heterogeneous across episodes and regions.
  - Simulations indicate large adverse spillovers from failures in important country policies (example: failure of internal rebalancing in China could cause spillovers on the order of 1–2 percent of world GDP).
  - A repricing of Japan’s sovereign debt or U.S. sovereign debt with yields rising by 200 basis points could cause large global losses.

### Box 3 — Cross-Border Transmission Effects: Regression evidence (Real GDP Growth and Trading Partner Growth, 1980-2011)
- Sample and methodology:
  - Sample: 29 advanced economies and 53 emerging markets over 1980–2011.
  - Dependent variable: real GDP growth rate (in percent). Advanced and EME growth is export-share weighted avg. of real GDP growth rates of top three advanced and EME export partners, respectively. Constant included. Outliers (bottom and top percentile) excluded. Clustered standard errors at the country level.
- Selected coefficient estimates (lagged partner growth or lagged own growth as reported in source):
  - Advanced growth (lagged): 0.340** (0.161); 0.264** (0.124); 0.152 (0.172); 0.062 (0.121); 0.492** (0.198); 0.346** (0.157)
  - EME growth (lagged): 0.761*** (0.198); 0.343*** (0.096); 0.241* (0.130); 0.267*** (0.089); 0.856*** (0.200); 0.367*** (0.132)
  - Real GDP growth (lagged): 0.442*** (0.047); 0.499*** (0.050); 0.396*** (0.057)
- Sample statistics:
  - Observations: 2,182; 2,182; 826; 826; 1,356; 1,356
  - R-squared: 0.182; 0.343; 0.405; 0.528; 0.183; 0.308
  - No. of countries: 82; 82; 29; 29; 53; 53
- Interpretation:
  - Regression results suggest appreciable cross-border correlations: domestic output growth is correlated with lagged output growth in both advanced and emerging market partners, even controlling for the country’s own lagged output growth.
  - Supports view that policy spillovers can be substantial.

### Box 3 and Box 4 implications on uncertainty and coordination
- Uncertainty about the state of the economy and policy effects (long/uncertain lags; real-financial linkages) is a serious impediment to policymaking and coordination.
- Multiplier uncertainty (nonzero variances of policy multipliers) can give rise to gains from coordination even when none exist absent uncertainty:
  - Quoted formulation retained from source: "when there is no multiplier uncertainty, 22 0, αβ σσ = = the cooperative and noncooperative policies coincide so there are no gains from cooperation. Conversely, starting from a situation in which there are no gains from coordination, multiplier uncertainty (either 2 0 α σ > or 2 0 β σ > ) will itself give rise to gains from coordination (additive uncertainty 2 (0) ε σ > is irrelevant for the incentive to coordinate)."
  - Uncertainty about transmission multipliers (β) tends to increase gains from coordination; uncertainty about domestic multipliers (α) tends to reduce gains.
- Uncertainty complicates bargaining and sustainment of cooperative agreements due to incentives to misrepresent spillovers and unobservable beliefs.

### Box 5 — Uncertainty and the (Un)Sustainability of Cooperation
- Repeated-game sustainment:
  - The “folk theorem” implies threats of future noncooperation can sustain coordination if punishment is long enough and discounting is low.
  - Triggering punishment under uncertainty requires setting a trigger based on deviations of macro variables from forecasts (notation: z and ˆz; punishment triggered if ˆ||zz  in source).
  - Trade-offs in setting triggers:
    - Too tight ⇒ punishment too often.
    - Too loose ⇒ scope for cheating.
    - Greater uncertainty (larger variance of ) flattens the density and raises probability that random realizations trigger punishment, increasing likelihood cooperation breaks down.
  - A neutral assessor can reduce uncertainty, allow less stringent triggers, and reduce breakdowns driven by random shocks.
- Participation and excluded countries:
  - Coordination among subsets internalizes spillovers among them only; excluded countries individually lack mass to matter to big players, though collectively significant.
  - Small countries may disagree among themselves, complicating Pareto-improving policies.
- Role and design of neutral assessor:
  - Functions: assess spillovers, identify tradeoffs, subject assessments to scrutiny, highlight mutually beneficial policy trades.
  - Neutrality is relative; assessor must avoid systematic bias and recognize that coordinated moves often require quid pro quos.
  - Historical shortcomings: confounding policies needed to reach global optimum with those to reach Nash; insufficient recognition of pros and cons across dimensions (e.g., growth vs. financial stability).
- Country guideposts as substitute:
  - Guideposts aim to limit the most harmful outward spillovers in trade flows and financial flows.
  - Limiting outward spillovers can be domestically costly (examples: curtailing cross-border lending in one’s currency; correcting undervaluation).
  - Guideposts should not be so stringent as to prevent adoption and should build on IMF surveillance fundamentals.

### Box 6 — Historical perspective and policy instruments
- Historical coordination episodes: interwar conferences (1920 Brussels, 1922 Genoa), Bretton Woods, 1977/78 London and Bonn Summits, 1985 Plaza Agreement, 1987 Louvre Accord, coordinated central bank action after October 1987 crash, G-20 fiscal coordination after the global financial crisis.
- Two recent coordination efforts with limited success:
  - Mid-2000s multilateral consultation on global imbalances: identified policy packages but implementation “fell short of the intentions”; process lacked ownership and became a “blame game.”
  - G-20 Mutual Assessment Process (MAP): IMF as secretariat; limited evidence that large countries adjusted policies in response to peer pressure; MAP shifted toward structural reforms and lacked an effective broker to identify mutually beneficial trades.
- Proposed mechanisms to strengthen coordination:
  - Neutral assessor to bridge divergent views, present analyses of alternative policy strategies and tradeoffs, and highlight mutually beneficial policy packages.
  - Guideposts for conduct in the international monetary system:
    - First guidepost: prevent currency misalignments using Fund surveillance tools (External Balance Assessment/External Stability Report) to identify exchange rates and external balances consistent with fundamentals and appropriate policies; urge corrective actions addressing monetary policy, FX intervention, fiscal policy, and structural reforms.
    - Second guidepost: limit cross-border financial-stability risks by shining light on exporting financial-stability risks, developing tools to assess credit/asset bubble risks and contributions of cross-border flows, and encouraging reciprocity (source countries raising costs/reducing quantity of risky carry trade lending; recipient countries adopting prudential policies and, when necessary, capital inflow controls).
- Key conclusions:
  - Less coordination in practice than theory because policymakers seldom think across objectives, disagree on spillovers, and face global asymmetry in gains.
  - In crisis times coordination can emerge spontaneously; in intermediate periods worthwhile gains may be unrealized.
  - Neutral assessor and guideposts can complement existing surveillance to press countries to avoid policies with large negative cross-border spillovers even at some domestic cost.

*Source: EXECUTIVE SUMMARY and Boxes 1, 3, 5, and 6, _sdn1311*

### Executive Summary ......................................................................................................

### EXECUTIVE SUMMARY

### Overview
- International macro policy coordination has been rare, with notable successful cases when the world economy seemed on the brink of collapse (examples cited: the 1978 Bonn Summit, the 1985 Plaza Agreement, the 1987 Louvre Accord, the 1987 stock market crash coordinated interest rate cuts and liquidity provision, and the 2008 global financial crisis coordinated fiscal expansions).
- In more normal times, policymaking tends to take a national rather than multilateral perspective despite strong theoretical arguments and systemic stresses.
- The case for coordination is grounded in standard welfare economics: policymaking involves tradeoffs across targets, and absent coordination externalities from cross-border spillovers imply Pareto-inefficient outcomes.

### Key findings on why coordination is uncommon
- Three principal impediments to coordination are identified:
  - Policymakers often do not think in terms of trade-offs across objectives; discussions fail when parties fixate on one macroeconomic goal and ignore other objectives.
  - Disagreement about the economic situation and cross-border transmission effects of policies—referred to as “model uncertainty” or deliberate “model disagreements”—which both raise potential gains from coordination and make agreements harder to reach and sustain.
  - Asymmetries in country size, where a significant portion of gains from coordination may accrue to countries too small to be included in any agreement.
- The noncooperative equilibrium is modeled as a Nash equilibrium where authorities maximize national welfare taking other countries’ policies as given; this equilibrium is generally Pareto-inefficient because a coordinated perturbation can yield first-order gains to the home country and only second-order losses abroad.
- Coordination can be inherently fragile because individual countries have a first-order incentive to “cheat” on agreed policies, and absent international sanctions, cooperation must be sustained by the implicit threat of future refusal to coordinate.

### Cross-border spillovers and welfare implications
- Policies such as monetary easing have domestic tradeoffs (e.g., boosting output vs. greater inflation or financial stability risks) and two effects on foreign countries (a positive effect via demand for exports and a negative effect via exchange rate appreciation); the net foreign impact can be negative.
- In a noncooperative equilibrium, excessive stimulus can occur because countries ignore negative externalities; coordinated internalization can yield first-order gains to each party when both parties move toward the cooperative equilibrium.
- Gains from coordination are not huge but are measurable and are noted to be "very similar to the estimated gains from global trade liberalization."

### Proposals to enhance prospects for cooperation
- Proposal 1: A neutral assessor
  - Rationale: Uncertainty and disagreements are genuine impediments; a neutral, credible, and impartial assessor can help bridge divergent views of national policymakers.
  - Role: The assessor would present analyses of alternative strategies and the resulting tradeoffs, not necessarily propose policies, enabling countries or groups to judge reasonable quid pro quos that underpin coordination.
  - Advantage: The IMF’s bilateral surveillance may position it to underscore that macro objectives involve tradeoffs across goals, aiding countries in identifying coordinated policy packages that are welfare superior.
- Proposal 2: Two guideposts to limit egregious negative spillovers
  - Purpose: To buttress international coordination and provide safeguards when coordination is impossible or fails to adequately account for spillovers on “small” countries.
  - Scope: Establish guideposts limiting the most egregious negative spillovers through countries’ current account and capital account, respectively.
  - Design: The specifics are for the international community to decide; the logic is to press countries to abjure policies with large negative cross-border spillovers even at some domestic cost.

### Relationship to existing IMF processes
- Both proposals build on existing processes:
  - The IMF’s surveillance aims for objective analysis and “ruthless truth-telling” to overcome biases in country self-assessments of domestic and cross-border effects.
  - The Integrated Surveillance Decision recently adopted by the IMF’s membership urges countries to consider policies that engender less adverse outward spillovers while still achieving domestic objectives.
  - The proposed guideposts build on the Integrated Surveillance Decision and would press countries to reject policies with large negative cross-border spillovers (through trade or financial flows) even if there is some domestic cost.

### Structure of the full paper (sections)
- Section II: Theory of international policy coordination and overview of reasons episodes of coordination are rare.
- Section III: Survey of evidence on cross-border spillovers and policy transmission effects.
- Section IV: How uncertainty raises both the gains from, and obstacles to, successful coordination.
- Section V: Ways of reducing obstacles to coordination.
- Section VI: Conclusion.

*Source: EXECUTIVE SUMMARY, _sdn1311 - Executive Summary*

### Box 1. The Theory of International Policy Coordination

### Box 1. The Theory of International Policy Coordination

### Framework and equilibrium concepts
- Policymakers in two symmetric countries have an objective function defined over two targets, (y1, y2), affected by domestic and foreign policies:
  - y1 = α11 m1 + β12 m2
  - y2 = α22 m2 + β21 m1
  - where α are domestic multipliers, and β are transmission multipliers.
- Nash (noncooperative) equilibrium condition:
  - The marginal rate of substitution (MRS) between the two targets is set equal to the marginal rate of transformation (MRT) achievable by the home country’s instrument (i.e., MRS = MRT using domestic multipliers).
  - Formally at Nash: (∂^2 v / ∂m1 ∂y1) = 0 leading to [ (∂^2 v / ∂y1 ∂y2) / (∂^2 v / ∂y1^2) ] = - (α12 / α11 ).
- Cooperative (global planner) equilibrium:
  - A planner maximizing a weighted average v_c = 0.5 v_home + 0.5 v_foreign sets the MRS equal to the MRT achievable through coordinated policies (domestic plus transmission multipliers).
  - Formally: (∂^2 v_c / ∂m1 ∂y1) = 0 implying [ (∂^2 v / ∂y1 ∂y2) / (∂^2 v / ∂y1^2) ] = - (α12 + β12) / (α11 + β11 ).

- Analogy: Nash equilibrium ~ autarky (MRS = MRT implied by domestic multipliers). Coordination ~ free trade (MRS = MRT implied by domestic plus foreign transmission multipliers).

### Monetary policy example and implications
- Suppose instrument is monetary policy m and targets are output y and inflation π:
  - y = α11 m + β12 m* + ε
  - π = α2 m
  - Parameter signs/assumptions: α11 > 0, α2 > 0, β12 may be <0 or >0, ε is shock; objective v(y,π) = (y^2) - (1/2) ω π^2.
- Nash policies:
  - m_N^* = m_N = α11 / [ α11 + α2 ω + β12 ? ]  (as in source: *2 11112 /[   ()] NN mm α ε α α β ω α = = = + +)
- Cooperative policies:
  - m_C^* = m_C = α11 / [ α11 + α2 ω + β12 ? ]  (as in source: *22 11112 ()/[()  ] CC mm α β ε α β ω α = = = + + +)
- Key comparative statics preserved from source phrasing:
  - Monetary policy will be too expansionary in the noncooperative equilibrium (m_NC > m_C) if β1 < 0 (policy is negatively transmitted).
  - Monetary policy will be insufficiently expansionary in the noncooperative equilibrium (m_NC < m_C) if β1 > 0 (policy is positively transmitted).
- Interpretation:
  - Negative transmission (β1 < 0): cooperation entails less monetary stimulus and lower output but higher welfare due to lower inflation/financial stability risk.
  - Positive transmission (β1 > 0): cooperation entails greater stimulus and higher output (with associated financial stability risk), and countries gain from higher output.

### Why coordination gains may be limited or episodic (six reasons)
- First: Policymakers may believe spillovers are too small to offset coordination costs.
  - Historical multi-country models in the 1980s incorporated appreciable cross-border transmission effects; estimates differed markedly in size and sign. Growing trade and financial linkages since the 1980s likely raised transmission multipliers.
  - Conclusion in box: small spillovers are not a plausible reason for episodic coordination.
- Second: Policymakers must face real policy trade-offs (fewer instruments than targets) for coordination to help.
  - If policymakers ignore some targets (e.g., care only about output), perceived welfare gains from coordination vanish.
  - Post-crisis, usable instruments are diminished (zero lower bound for monetary policy; high public debt and political paralysis for fiscal policy) and attention to additional targets (financial stability) has increased, raising the cost of myopia.
- Third: Coordination is moving from Nash to cooperative policies; if initial policies are not at Nash, gains from moving to globally optimal policies can be larger.
  - Coordination can provide a quid pro quo that helps overcome domestic constraints (e.g., political paralysis), but only if policymakers trust that foreign quid pro quo will be delivered.
- Fourth: The nature of shocks matters.
  - Gains from coordination require the economy to be sufficiently off its desired path and policy to be able to make an appreciable difference. If economy is near equilibrium or policy is ineffective, gains are limited.
  - Estimates suggest gains are not huge but not negligible—similar in magnitude to gains from multilateral trade liberalization.
- Fifth: Distribution of gains may be asymmetric.
  - Most welfare gains from coordination may accrue to small countries (possibly non-parties), while economically important countries may see little incentive to coordinate.
  - Asymmetric gains/losses suggest the possible need for “rules of the road” to constrain policies that impose adverse cross-border spillovers on small countries.
- Sixth: Uncertainty about the state of the economy or policy effects can both raise and complicate coordination gains.
  - Uncertainty about cross-border effects increases the volatility that acts like a negative spillover; thus gains from coordination increase with uncertainty in spillovers.
  - But disagreement over models and the size or sign of spillovers makes negotiation and sustaining cooperative agreements harder.
  - Strategic misrepresentation (each party overstating domestic needs or understating transmission effects) can impede agreement formation and maintenance.

### What models and empirical evidence say (Box 2 highlights)
- Early literature (1980s) found modest gains from coordination (e.g., Oudiz and Sachs (1984) estimate utility equivalent of one-half percentage point of GNP in each of the next few years from coordinated expansion).
  - Reasons for small estimated gains:
    - Cross-border multipliers in those models were relatively small and varied in sign across models.
    - Some studies reverse-engineer preferences assuming observed policies are Nash; if Nash outcomes implied low weight on unemployment, coordination gains appear small.
    - If shocks are temporary and policy can only shift timing of output losses, gains are limited to smoothing benefits.
- Model uncertainty matters:
  - Accounting for uncertainty can materially increase estimated gains (Ghosh and Masson (1988) find incorporating uncertainty roughly doubles estimated gains).
- Empirical evolution of transmission multipliers:
  - Early averaging across models suggested transmission multipliers ~ one tenth the size of domestic multipliers; in absolute value, transmission multipliers often were one-third to one-half the size of domestic multipliers.
  - More recent evidence suggests transmission multipliers have grown and are now about one half the size of domestic multipliers, reflecting deeper trade and financial integration.
  - Transmission effects are larger for large economies, during downturns, and for closely interconnected countries.
  - For the United States, recent estimates (IMF, 2013b) suggest fiscal policy transmission multipliers as high as 60 percent of domestic multipliers, with larger multipliers for Latin America and Europe, and smaller ones for Asian economies.
  - Monetary policy transmission multipliers are found to be about 40 percent as large as domestic multipliers, with largest effects for Latin American countries.
  - Financial linkages explain about one third of cross-country variation in multipliers; trade linkages explain about 10 percent (or more for fiscal shocks).
- Crisis vs quiet times and unconventional policies:
  - Transmission multipliers may be larger in crisis periods.
  - Quantitative easing episodes show heterogeneous external effects: some episodes produced higher foreign output, lower foreign bond yields, higher equity prices, and currency appreciations; other episodes (e.g., Bank of Japan) were associated with falls in foreign equity prices and yen depreciation transmitting negative effects to trade partners.
  - Capital flow responses to quantitative easing have varied over time (outflows in early episodes, inflows in later episodes, “taper” talk producing outflows), with regional variation (Asia and Latin America distinct from Europe).
- Simulations of exit from unconventional monetary policy yield mixed results:
  - If exit happens with good domestic growth news, transmission tends to be positive (growth effects dominate).
  - If exit reflects rising domestic financial risks, transmission can cause sizable global downdrafts.
  - Simulations highlight large adverse spillovers from failures in important country policies (e.g., failure of internal rebalancing in China could cause spillovers on the order of 1–2 percent of world GDP).
  - A repricing of Japan’s sovereign debt or U.S. sovereign debt with yields rising by 200 basis points could cause large global losses.

*Source: Box 1. The Theory of International Policy Coordination, from the provided IMF PDF content.*

### Box 3. Cross-Border Transmission Effects—Some Evidence

### Box 3. Cross-Border Transmission Effects—Some Evidence

### Evidence from regressions (Table 1: Real GDP Growth and Trading Partner Growth, 1980-2011)
- Sample: 29 advanced economies and 53 emerging markets over 1980–2011.
- Dependent variable: real GDP growth rate (in percent). Advanced and EME growth is export-share weighted avg. of real GDP growth rates of top three advanced and EME export partners, respectively. Constant included in all specifications. Outliers (i.e., growth rates in the bottom and top percentile of the distribution) are excluded. Clustered standard errors at the country level reported in parentheses. ***, **, and * indicate statistical significance at the 1, 5 and 10 percent levels, respectively.

- Advanced growth (lagged):
  - 0.340** (0.161)
  - 0.264** (0.124)
  - 0.152 (0.172)
  - 0.062 (0.121)
  - 0.492** (0.198)
  - 0.346** (0.157)

- EME growth (lagged):
  - 0.761*** (0.198)
  - 0.343*** (0.096)
  - 0.241* (0.130)
  - 0.267*** (0.089)
  - 0.856*** (0.200)
  - 0.367*** (0.132)

- Real GDP growth (lagged):
  - 0.442*** (0.047)
  - 0.499*** (0.050)
  - 0.396*** (0.057)

- Observations:
  - 2,182; 2,182; 826; 826; 1,356; 1,356

- R-squared:
  - 0.182; 0.343; 0.405; 0.528; 0.183; 0.308

- No. of countries:
  - 82; 82; 29; 29; 53; 53

### Interpretation and links to the literature
- The regression results suggest appreciable cross-border correlations: across the full sample, output growth is correlated with lagged output growth in both advanced and emerging market economies, even controlling for the country’s own lagged output growth.
- These findings are consistent with literature finding significant cross-border correlations, especially when there are strong trade and financial linkages.
- Selected literature evidence cited:
  - Kose and others (2008) use a global dynamic factor model and find convergence in business cycle fluctuations both within advanced economies and emerging market economies, but decoupling between these two groups.
  - Cesa-Bianchi and others (2012) using a GVAR find that the impact of Chinese shocks on Latin America has increased dramatically since the mid-1990s and has come to dominate that of the United States, with much of the strength coming from indirect effects via third countries.
  - Additional references listed: Heathcote and Perri (2004), Stock and Watson (2003), Kose and others (2003), Bordo and Helbling (2004), Baxter and Kouparitsas (2005), Kalemli-Ozcan and others (2013).

### Implications for coordination under uncertainty (summary of Sections IV and Box 4)
- Uncertainty about the state of the economy and about policy effects (long and uncertain lags; real-financial linkages) is a serious impediment to effective policymaking and to international policy coordination.
- Uncertainty about transmission effects can itself be a negative cross-border spillover because risk-averse policymakers dislike resulting volatility; by increasing the magnitude of spillovers, such uncertainty raises the gains from international policy coordination.
- Box 4: analytical illustration of how uncertainty affects gains from coordination:
  - Starting from a case with no gains from coordination in the absence of uncertainty, multiplier uncertainty (nonzero variances of policy multipliers) will give rise to gains from coordination.
  - Quoted formulation from the source: "when there is no multiplier uncertainty, 22 0, αβ σσ = = the cooperative and noncooperative policies coincide so there are no gains from cooperation. Conversely, starting from a situation in which there are no gains from coordination, multiplier uncertainty (either 2 0 α σ > or 2 0 β σ > ) will itself give rise to gains from coordination (additive uncertainty 2 (0) ε σ > is irrelevant for the incentive to coordinate)."
  - The Box 4 analysis also shows that uncertainty about transmission multipliers (β) tends to increase gains from coordination, whereas uncertainty about domestic multipliers (α) tends to reduce gains from coordination.
- Uncertainty complicates bargaining and sustainment of cooperative agreements:
  - Disagreement about models and unobservable beliefs create incentives to misrepresent policy spillovers, which can make negotiating and sustaining agreements difficult or impossible even when coordination would be Pareto-improving.
  - In repeated interactions, sustaining cooperation requires sufficiently long horizons, low discounting of the future, and credible punishment strategies; uncertainty tightens triggers for punishment and can make cooperation break down more often even absent deliberate cheating.
  - The possibility of deliberate disagreements about the state of the economy or spillovers helps explain episodic and sporadic international policy coordination.

### Policy-relevant takeaways
- Empirical evidence points to nontrivial cross-border correlations in output growth—both advanced and EME partner growth matter for domestic growth—supporting the view that policy spillovers can be substantial.
- Uncertainty about transmission magnifies the potential gains from international policy coordination, particularly when transmission multipliers are more uncertain than domestic multipliers.
- Practical obstacles to coordination include unobservable beliefs, incentives to misrepresent spillovers during bargaining, and fragility of agreements in the face of shocks; these factors argue for institutional mechanisms (e.g., neutral assessment, guidelines) to bridge differing perspectives and to guide policies when spillovers affect parties not included in coordination exercises.
- Policymakers should recognize trade-offs (e.g., closing output gaps versus inflationary or financial-stability risks) and consider combinations of macro objectives that may be superior for domestic welfare and essential for successful international coordination.

*Source: Box 3 and related text from the supplied IMF content unit.*

### Box 5. Uncertainty and the (Un)Sustainability of Cooperation

### Box 5. Uncertainty and the (Un)Sustainability of Cooperation

### Coordination under uncertainty and repeated games
- Coordinated equilibrium: as long as one party sticks to the agreement, the other can do even better by reneging.
- In dynamic settings, the “folk theorem” of repeated games implies that the threat of future noncooperation (a “punishment period”) can sustain coordination if the punishment is long enough and parties do not discount the future too heavily.
- Triggering punishment under uncertainty:
  - Policies are generally observable, but the information and forecasts on which they are based typically are not.
  - Reversion must be triggered on the basis of macroeconomic outcomes being sufficiently different from what would have been expected had parties designed policies on their truthful beliefs.
  - Notation as presented in source: Let z be the macro variable, ˆz its forecast, then the punishment is triggered if ˆ||zz , where  is the trigger level.
  - Trade-off in setting trigger:
    - Too tight a trigger ⇒ punishment imposed too often.
    - Too loose a trigger ⇒ scope for cheating.
    - The benefit of cheating must be weighed against the increased probability of “getting caught” (triggering the punishment).
    - For a given benefit of cheating and cost of being caught, there is a minimum increase in likelihood of triggering the punishment that makes cheating not worthwhile.
    - The increase in probability is the derivative of the distribution function—that is, the density of . Ensuring incentive compatibility amounts to setting the trigger to achieve a certain minimum height of the density function.
  - Even with an incentive-compatible trigger, random realizations of  can trigger punishment and lead to breakdowns of cooperation even though neither party cheated. The probability of this is denoted by the area marked A in the source.
  - Greater uncertainty about policy effects is equivalent to a larger variance of —a flattening of the density function and a larger area under the curve (B). Hence, greater uncertainty increases the likelihood that cooperation will break down.
  - A neutral assessor that provides unbiased assessments about the state of the economy and the effects of policies may reduce uncertainty, allowing for a less stringent trigger and fewer breakdowns due to random shocks.

### Policy spillovers, model uncertainty, and episodic coordination
- Policy spillovers:
  - Spillovers are large, have grown larger as real and financial integration has progressed, and are particularly meaningful during turbulent periods when economic variables are considerably off their desired paths.
  - Different players perceive spillovers/transmissions from actual/prospective policies differently—the divergent perspectives on unconventional monetary policies and on policies to accelerate internal and external rebalancing are cited as prime examples.
- Multidimensional nature of spillovers:
  - Output transmission, financial flows, and prices are all channels.
  - Model uncertainty gives scope for disagreement on the size and even the sign of spillovers.
  - Such disagreements help explain the episodic nature of coordination.

### Participation, excluded countries, and limits to coordination
- Most countries do not participate in coordination exercises:
  - Coordination among a subset of countries will not converge to a global optimum but at best to an optimum that internalizes spillovers among that subset only.
  - Excluded countries individually lack sufficient mass in goods or financial markets to matter to big players, even though collectively they constitute a significant part of the global economy.
  - Small countries may disagree among themselves on the nature of spillovers (one subgroup preferring more use of an instrument, another preferring less), making it hard to find Pareto-improving policies.
  - Results:
    - Coordinated policies may be some way from the global optimum (optimal for big players only); or
    - Coordination may not occur because average spillovers are small even though bilateral spillovers may be large (possibly one-directional).

### Role, design, and limits of a neutral third-party “assessor”
- Main suggestion to address uncertainty and spillovers: a neutral third-party assessor to scrutinize country assessments of inward and outward spillovers and assess alternative policy packages or trades acceptable to principals while increasing global welfare.
- Intended functions of the assessor:
  - Assess policy spillovers, identify tradeoffs, and subject assessments to scrutiny by all parties.
  - Use results of bilateral surveillance and policy tradeoffs identified therein to highlight merits of alternative national policy packages.
  - Assess the extent to which mutually beneficial policy trades exist.
- Neutrality and credibility considerations:
  - Neutrality is relative; an assessor need only be more neutral than individual participants to raise global efficiency.
  - Credibility is undermined if assessor’s assessments suggest systematic bias (e.g., always identifying policy changes as yielding welfare gains at both national and global levels).
  - The assessor should avoid implying that countries always fail to exploit available welfare gains; instead, recognize that policy changes toward the coordinated outcome are often contrary to a single nation’s interest and require quid pro quos.
- Historical performance and possible shortcomings:
  - The assessor role has existed in past coordination attempts, but may not have been performed optimally.
  - Two identified issues:
    1. Confounding policy changes needed to reach the global optimum with those needed to reach the Nash:
       - If countries do not accept they are off their Nash but accept a global problem, asserting policies are unambiguously in national interest may undercut acceptance.
       - Often policy changes to reach coordinated outcome are contrary to a single nation’s interest without foreign quid pro quo.
    2. Failure to recognize pros and cons of policy actions consistently:
       - Assessments have at times emphasized growth effects while insufficiently acknowledging external and financial stability risks (or latent risks) in other countries or the country undertaking the policy.
       - Assessments need to acknowledge the multidimensional aspects of spillovers at all times to avoid appearing biased as the balance of risks shifts.

### Proposed substitute for broader coordination: country guideposts
- Expanding participation in coordination agreements is seen as impractical; heterogeneity in larger groupings (like the G-20) may hamper effectiveness.
- Proposed substitute: the international community agrees to abide by a set of guideposts for each country (see Ostry and others (2012) in source).
- Purpose of guideposts: limit the most potentially harmful outward spillovers in two key areas:
  - Trade flows; and
  - Financial flows.
- Recognized tradeoffs and domestic costs:
  - Limiting harmful outward spillovers will sometimes be costly domestically.
    - Example: Curtailing lending in one’s currency to unhedged borrowers abroad can reduce profitability of domestic banks and growth while reducing recipient-country financial-stability risks.
    - Example: Curtailing a policy of undervaluation that spurs domestic growth may force undesirable external adjustments in other countries and be costly for the home country.
  - Correcting policies that violate guideposts may impose costs on violators, similar to moving from domestic Nash toward global optimum.
- Design constraint:
  - Guideposts should not be so stringent that adoption is unlikely.
  - Stick to areas already fundamentals of IMF surveillance.

*Source: Box 5. Uncertainty and the (Un)Sustainability of Cooperation, _sdn1311*

### Box 6. International Policy Coordination in Historical Perspective

### Box 6. International Policy Coordination in Historical Perspective

### Historical attempts at coordination
- Early interwar conferences in Brussels in 1920, and in Genoa in 1922.
- Bretton Woods: sought to codify “rules of the road” to limit beggar-thy-neighbor policies.
- 1977/78  London and Bonn Summits: coordinated efforts during the stagflationary period after the first oil price shock.
- 1985 Plaza Agreement and 1987 Louvre Accord: focused on coordinated foreign exchange intervention.
- G-7 central banks coordinated interest rate cuts and liquidity provision after the stock market crash in October 1987.
- G-20 coordinated fiscal expansion in the aftermath of the global financial crisis.

### Two recent episodes illustrating coordination difficulties
- Multilateral consultation on global imbalances (established in the mid-2000s)
  - Aim: multilateral surveillance to resolve global imbalances while maintaining robust global growth; facilitate action-oriented debate and policy actions by participants to reduce imbalances.
  - Outcome: consultations identified policy packages for participants but implementation “fell short of the intentions.”
  - Contributing factors:
    - Process lacked ownership because it “did not come from the participants themselves.”
    - Exercise became a “blame game” with participants preferring to blame others.
    - Despite recognition of risks from growing imbalances, participants were not seized by urgency; policies were not materially altered.
    - The Great Recession that followed reduced the urgency of dealing with global imbalances.

- G-20 Mutual Assessment Process (MAP) after the global financial crisis
  - MAP: IMF asked to undertake a mutual assessment of policies, with IMF acting as secretariat and peer scrutiny of policy frameworks.
  - Evidence to date does not suggest that any of the large countries have made significant adjustments in response to peer pressure under the MAP (Faruqee and Srinivasan, 2012).
  - Incentives for collective action waning due to distance from crisis peak, political-economy factors, and a multi-speed global recovery.
  - MAP shifted focus toward structural reforms rather than macro policies—reflecting areas with stronger political resistance and where peer monitoring is lighter.
  - Absence of an effective broker in the MAP to identify mutually beneficial policy trades based on a shared model contributed to its limited success.

### Proposed mechanisms to strengthen coordination
- Role of a neutral assessor
  - Purpose: bridge divergent views of national policymakers; must be perceived as impartial.
  - Function: present analyses of alternative policy strategies and resulting tradeoffs; highlight policy packages that make each party better off.
  - Emphasis: coordination as mutually beneficial trades, not “making concessions”; assessor would not necessarily propose policies but enable recognition that alternative packages could better achieve domestic mandates.

- Guideposts for conduct in the international monetary system
  - Goal: limit negative spillovers via current and capital accounts when coordination cannot be achieved.
  - First guidepost: prevent currency misalignments
    - Use Fund surveillance tools—including the External Balance Assessment/External Stability Report—to identify exchange rates and external balances consistent with fundamentals and appropriate policies.
    - Urge countries to address deviations and identify contributing policies: monetary policy; foreign exchange market intervention policy; fiscal policy; structural reforms affecting demand composition between tradables and nontradables.
    - International community would back reductions in the most salient policy distortions and place weight behind Fund staff assessments.
  - Second guidepost: limit cross-border financial-stability risks
    - Shine light on exporting financial-stability risks and contributing policies (e.g., loose monetary policy, lax prudential regulation).
    - Develop tools building on analytical/empirical work assessing risk of credit/asset bubbles/booms/cycles and the contribution of cross-border flows.
    - Encourage reciprocity: source countries may raise the cost and reduce quantity of risky carry trade lending; recipient countries adopt prudential policies and, when necessary, capital inflow controls.
    - Rationale: convexity of policy costs implies acting at both ends of transactions can be more efficient than acting only at one end.

### Key findings and challenges
- Less coordination in practice than in theory for three main reasons:
  - Policymakers seldom think in terms of trade-offs across objectives; they fixate on specific objectives without recognizing longer-term implications.
  - Countries disagree on the nature and size of spillovers and how policy packages could deliver better spillovers and credible policy trades.
  - The global economy is highly asymmetric: gains from coordination may be small and diffuse, making it difficult to arrange policy trades.
- Timing and incentives:
  - In normal times, benefits from coordination may be too small to overcome hurdles.
  - In crisis times, coordination can emerge almost spontaneously.
  - In intermediate periods (neither crisis nor normalcy), there may be worthwhile gains from coordination that are not being realized.

### Conclusions and policy implications
- Neutral assessor could tip countries toward greater cooperation by providing impartial analysis of alternative policy packages and tradeoffs.
- Guideposts would act as safeguards and complements to coordination efforts, pressing countries to avoid policies with large negative cross-border spillovers even if there are some domestic costs.
- Both proposals build on existing bilateral and multilateral Fund surveillance and the Integrated Surveillance Decision, which encourages consideration of policies that engender less adverse outward spillovers while achieving domestic objectives.

*Source: Box 6. International Policy Coordination in Historical Perspective (excerpt).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2013/_sdn1311.pdf_