## wpiea2024047-print-pdf

## Source details

**Canonical URL:** [wpiea2024047-print-pdf](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024047-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024047-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024047-print-pdf.pdf.json)

---

### Context and recent developments
- Europe experienced a high level of economic and geopolitical uncertainty in recent years.
- After more than a decade of subdued inflation, price growth rose in early 2021 mainly because of pandemic-driven supply bottlenecks in the non-energy goods sector.
- By October 2022, energy and food prices had surged because of Russia’s invasion of Ukraine, and inflation had exceeded 10 percent, the highest level in the history of the eurozone, and then started to recede.
- Headline HICP inflation started to moderate at the end of 2022 thanks to lower energy prices, but core inflation peaked later and remained persistent; as of end-2023 many forecasters projected core inflation to remain above target for an extended period.
- Labor costs and firm profits both contributed to momentum of core inflation; there is substantial uncertainty on the direction and magnitude of these forces.

### Key conceptual point: Knightian uncertainty about inflation persistence
- Policymakers faced unusual Knightian uncertainty—structural uncertainty not easily captured with probabilities—about inflation persistence.
- The greater the persistence of inflation, the more forceful monetary policy action is required to curtail it.
- This motivates adopting a robust monetary policy that responds to inflation in a manner robust to alternative but plausible assumptions about persistence.

### Principal findings from the robust-policy analysis
- When inflation persistence is uncertain and underlying wage pressure is robust, central bankers should set policy under the assumption that inflation is more persistent than routine baseline forecasts suggest.
- If inflation persists above target, maintain a tightening bias: the policy rate should go beyond what is predicted under a baseline forecast that assumes routine reaction functions.
- Costs and trade-offs:
  - The costs of not insuring against very persistent inflation outweigh the higher output gap volatility resulting from assuming high persistence when persistence is in fact low.
  - Estimates indicate costs of using a robust monetary policy rule are generally modest, while costs of ignoring the rule can be substantially high in a worst-case scenario.
  - A wait-and-see strategy is not advisable: if policymakers initially assume low persistence but later learn persistence is high, they must tighten much more and more rapidly than if they had used a robust response from the outset; sharp shifts can be disruptive to financial markets and central bank credibility.

### Quantitative and model-based results (illustrative)
- Representative model parameter values used in simulations:
  - 훼 .0625 — Relative weight on output gap stabilization
  - 훽 .995 — Discount factor
  - 휎 1 — CRRA
  - 휉 .85 — Share of fixed prices (Calvo parameter)
  - 휸 .1 — Openness (import share)
- Policymaker belief scenarios and outcomes:
  - Reference (moderate persistence): 휙 ௣ = 0.4
  - Robust (worst-case) assumed persistence: 휙 ௣ = .8
  - Wait-and-see: correction assumed five quarters after shock (the economy runs hot for a year), and assumed completely unanticipated by the private sector.
- Loss/benefit quantification:
  - The robust policy is around 9 percent worse than the policy in a moderate persistence world (when true persistence is moderate).
  - The worst-case error (erring in the opposite direction) produces losses that are 60 percent greater.
- Model impulse-response patterns (illustrative):
  - Under-estimating persistence: inflation peaks higher and takes far longer to return to target; the output gap becomes negative but not sufficiently negative to cool demand.
  - Wait-and-see correction requires much sharper and more costly tightening than the robust policy.

### ECB-specific application (January 2023 WEO baseline / end-2022 perspective)
- Baseline projection (staff, 2023Q1): quarterly average of ECB policy rate would peak at 3.7 percent.
- Robust approach implication:
  - Applying a robust approach would require a peak policy interest rate up to 60 basis points higher than that baseline.
  - Simulation result: policy rate peaks at 4.3 percent instead of 3.7 percent when applying the robust rule to the January 2023 WEO projection.
  - Inflation comes down faster and reaches the ECB’s 2 percent target before the end of 2024 under the robust rule in the simulation.
- Empirical note: after publication the ECB raised its interest rate twice to 4 percent and the WEO updated its terminal rate forecast to that value.
- Estimated policy-rule parameters (Table 3, reported):
  - Historical estimated: 휙 గ 1.68; 휙 ௬ .019; 휙 ୼୷ .15; 휌 ௥ .84
  - Robust value: 휙 గ 1.94; 휙 ௬ .00; 휙 ୼୷ .13; 휌 ௥ .90
- Interpretation:
  - The robust rule implies a stronger response to inflation, weaker response to output, and higher smoothing.
  - The robust path represents an upper bound of policy paths aligned with baseline, central bank judgment add-ons, and priors about backward indexation; actual choices would lie between reference and robust paths depending on credible region for persistence parameters.

### Import elasticity of substitution and related model primitives
- Key parameters and primitives reported:
  - 휔 1.095
  - Interest elasticity of demand = 1 + 훾(휎휂 − 1)(2 − 훾) / 휆  .052
  - PC slope = ቂ(ଵِక)(ଵِఉక) క ቃቀఙଵାఠ൅1ቁ 휙 ௫ .75      Habit/demand persistence 휌 .92
  - Cost shock autoregressive coefficient 훿 ௣ .8
- Inflation-prone economy parameterization (Table 2):
  - Price inertia (휄 ௣): Baseline value .0315; Alternative value .0908
  - Wage inertia (휄 ௪): Baseline value .2274; Alternative value .4030
  - Price stickiness (훾): Baseline value .7975; Alternative value .7543
- Implication for CEE economies:
  - Robust approach may yield higher benefits for CEE countries outside the euro: smaller open economies, more frequent/persistent external shocks, less resilient inflation expectations, and higher backward indexation.
  - Financial frictions and currency mismatches may justify additional tools (for example, foreign exchange interventions) alongside interest-rate policy.

### Policy recommendations and strategic guidance
- Presume inflation is persistent until convincing contrary evidence is available.
  - Policy easing should not react prematurely to slowing inflation because temporary factors may affect inflation on the way down as on the way up.
- Maintain a tightening bias when inflation is observed to persist above target.
- Be mindful that protracted policy tightness may lead to imbalances, particularly in the financial sector; policy must balance beating inflation sustainably without causing undue damage.
- Use scenario analysis and communicate decisively:
  - Scenario analysis complements focus on modal outlook and helps illustrate risks and model uncertainty.
  - Decisive communication emphasizing meeting the inflation target can help anchor expectations and amplify benefits of a robust approach.
- Reaction asymmetry:
  - Central banks should respond more to upward surprises to inflation than to downward ones to enable earlier course correction and reduce tail risks.
- Consider alternative tools where appropriate:
  - In economies with shallow financial markets or material currency mismatches, tools besides the interest rate (e.g., foreign-exchange interventions) may be warranted; the case for robust interest-rate policy remains even then.

### Costs of alternative strategies and caveats
- Wait-and-see strategy costs:
  - Requires much sharper later tightening, increases central bank losses relative to the robust policy, and raises tail risks for financial stability.
- ELB considerations:
  - Persistent undershooting of target raises ELB concerns; prompt action can avoid larger rate cuts later and resorting to unconventional tools.
- Modeling caveats:
  - Simulations assume transmission can be stronger than thought but do not provide an explicit account of that aspect of model uncertainty.
  - If the Phillips curve is very flat, the central banker may need to raise rates substantially and for a long time to bring inflation back to target.
- Related literature:
  - Robust-control and max-min strategy literature supports stronger responses to inflation under model uncertainty and desire to avoid worst-case outcomes.

*INTERNATIONAL MONETARY FUND — wpiea2024047-print-pdf*

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

### Executive Summary

### Context and recent developments
- Europe faced a high level of economic and geopolitical uncertainty over the past few years.
- After more than a decade of subdued inflation, the growth rate of prices in Europe started to increase in early 2021 mainly because of pandemic-driven supply bottlenecks in the non-energy goods sector.
- By October 2022, energy and food prices had surged because of Russia’s invasion of Ukraine, and inflation had exceeded 10 percent, the highest level in the history of the eurozone, and then started to recede.
- Headline HICP inflation started to moderate at the end of 2022 thanks to lower energy prices, but core inflation peaked later and has been persistent; its decline has been more modest through 2023 and as of end-2023, it was projected by many forecasters to remain above target for an extended period.
- Labor costs and firm profits have both contributed to the momentum of core inflation, and there is substantial uncertainty on the direction and magnitude of these two forces.
- These developments are not easily compared with historical episodes since the start of the currency union, raising the possibility of a regime shift in inflation and greater uncertainty about the persistence of inflation (Fuhrer, 2010).

### Key conceptual point: Knightian uncertainty about inflation persistence
- Policymakers faced an unusual degree of Knightian uncertainty—structural uncertainty that is not easily captured with probabilities—about the persistence of inflation.
- Other uncertainties (natural rate of unemployment, neutral interest rate, strength of transmission) matter, but uncertainty about inflation persistence is arguably most important because:
  - The greater the persistence of inflation, the more forceful monetary policy action needs to be to curtail it.
- This motivates adopting a robust approach to monetary policy that responds to inflation in a manner robust to alternative but plausible assumptions about persistence (Qvigstad, 2006).

### Principal findings from the robust-policy analysis
- When facing uncertain inflation persistence and robust underlying wage pressure (as in much of the post-pandemic period), central bankers should set the policy interest rate under the assumption that inflation is more persistent than routine baseline forecasts suggest.
- If inflation persists above target, this argues for maintaining a tightening bias—i.e., the policy rate should go beyond what is predicted under a baseline forecast in which policy is assumed to follow the usual reaction function.
- The costs of not insuring against the worst-case scenario with very persistent inflation outweigh the higher output gap volatility that results from assuming high persistence when persistence is in fact low.
- Estimates indicate the cost of using a robust monetary policy rule are generally modest, while the costs of ignoring the rule can be substantially high in a worst-case scenario.
- A wait-and-see strategy to learn the inflation regime is not advisable: if policymakers initially assume low persistence but later learn persistence is high, they would need to tighten much more and more rapidly than if they had used a robust response from the outset; such sharp policy shifts can be disruptive to financial markets and challenging for central bank communications and credibility.

### Policy recommendations and strategic guidance
- Presume inflation is persistent until convincing contrary evidence is available.
  - Policy easing should not react prematurely to slowing inflation because temporary factors may affect inflation on the way down as on the way up.
- Maintain a tightening bias when inflation is observed to persist above target.
- Be mindful that protracted policy tightness may lead to imbalances, particularly in the financial sector; policy must balance beating inflation sustainably without causing undue damage.
- Cautious judgements on inflation persistence are essential for a central bank like the ECB to sustainably achieve its inflation objective; such judgements should be based on progress in bringing down core inflation and evolving evidence on structural changes in the economy.

*INTERNATIONAL MONETARY FUND — Executive Summary (from wpiea2024047-print-pdf)*

### Introduction

### wpiea2024047-print-pdf - Introduction

### Central problem and motivation
- Central bankers face uncertainty about the path of inflation and appropriate policy rates.  
- Supply disturbances (higher energy prices, supply-chain constraints, shifts in labor supply) push inflation and output growth in opposite directions, complicating policy trade-offs and sometimes calling for a partial “see through” of such shocks.  
- Uncertainty about the persistence of inflation—and about how persistence may change through shifts in wage- and price-setting—critically affects policy choices and required policy “force.”  
- Informal definition: “inflation persistence” denotes the tendency of the rate of change of prices (or wages) to remain constant in the absence of an economic ‘force’ acting upon it (Fuhrer, 2010). The more persistent is inflation, the greater the force (for example, from policy actions) required to reduce it.

### Consequences of misjudging persistence
- Under-estimating inflation persistence:
  - May lead to applying too little policy “force,” producing inflation that stays higher for longer and eventually requires higher rates and a deeper slowdown to tame.
  - Produces serially correlated inflation-forecast errors, undermining credibility.
- Persistent undershooting of the target:
  - Raises concerns about the effective lower bound (ELB) and the uncertainty surrounding it.
  - A prompt reaction can avoid larger rate cuts later and the need to utilize other monetary policy tools (Tillman, 2020).
- The paper argues that the costs of underestimating persistence are higher than the costs of overestimating it, motivating a “better safe than sorry” or robust policy philosophy.

### Robust policy recommendation (qualitative)
- When persistence is highly uncertain, central banks should:
  - React forcefully to inflation.
  - Keep a tightening bias to bring down inflation.
  - Act as if inflation persistence is high because the cost of underestimating persistence exceeds the cost of overestimating it.
- Practical implication (application to ECB, based on January 2023 WEO baseline):
  - Staff projected a baseline where the quarterly average of the ECB policy rate would peak at 3.7 percent (as of 2023Q1).
  - Applying a robust approach would have required a peak policy interest rate up to 60 basis points higher than that baseline.
  - Since then, the ECB has already hiked its deposit facility rate (DFR) to meet about half that distance.

### Costs of a “wait-and-see” strategy
- A “wait-and-see” approach—initially acting as if persistence is low and correcting later—can be very costly:
  - Would require a much sharper monetary policy tightening later than the robust approach.
  - Increases central bank losses relative to the robust policy.
  - Raises tail risks for financial stability by large effects on borrowers with variable-rate debt and potential impairment of financial market functioning.
- Policy implication: central bank should respond more to upward surprises to inflation than to downward ones to enable earlier course correction.

### Literature and theoretical support
- Robust control literature (Hansen and Sargent, 2008, 2011) supports more aggressive responses under fundamental uncertainty to minimize likelihood and impact of worst-case scenarios.
- Prior monetary policy work (Walsh 2003, 2004; Levin and Williams 2003; Orphanides and Williams 2007) argues for erring on the side of more backward-looking models or greater inertia and stronger responses to inflation under model uncertainty.

### Sources of uncertainty emphasized
- Uncertainty about inflation persistence (survey evidence shows wide dispersion in perceived persistence among SPF respondents up to end-2022).  
  - Perceived persistence measured as correlation between revisions to 12-month and 24-month ahead inflation forecasts on successive dates; wide cross-forecaster dispersion (10th–90th percentiles) as of end-2022.
- Labor market uncertainties:
  - Transmission of labor market tightness to wage growth (uncertainty heightened by structural shifts such as work-from-home preferences and hours preferences).
  - Link between wage growth and price inflation depends on wage-setting institutions (collective bargaining, cost-of-living clauses, public-sector benchmarking) which can change endogenously in a regime switch.
- Profitability regime uncertainty:
  - The passthrough from wages to prices depends on firm profits. Evidence: when firm profits are high, wage shocks have a very small effect on prices while price shocks increase wages persistently; when profits are low, wage shocks raise prices but price shocks do not raise wages.
- Uncertainty about transmission strength and Phillips-curve slope:
  - Debate renewed after the 2020-21 Covid-19 shock; a steeper Phillips curve implies stronger transmission of output-gap affecting policies to inflation, potentially making inflation come down faster for a given stance but increasing sacrifice ratios for the “last mile.”
- Measurement uncertainty in policy stance:
  - Real policy rate (r) vs. natural rate r* measurement error: ex-ante vs. ex-post real rates can differ substantially; r* is unobservable and affected by many drivers (real and financial, short- and long-term).

### Role of communication
- During high uncertainty, decisive communication and emphasis on meeting the inflation target can help anchor expectations.
- Explicitly adopting a robust approach to policy could amplify the benefits of communication.

### Illustrative model and approach
- Model framework: open economy New Keynesian model as in Clarida, Galí, and Gertler (2001) (CGG), augmented by:
  - Intrinsic inertia in prices and demand controlled by 휙௣ and 휙௫.
  - Behavioral discounting in the open-economy Phillips curve governed by 훿௣.
- Focus: cost-push disturbance c in the Phillips curve follows an AR(1) with coefficient 휌; main interest in consequences of misjudging intrinsic persistence 휙.
- Policy: central bank minimizes period loss ℒᵗ = (πᵗ)² + α (xᵗ)² under constraints (PC), (IS), (RER) and uses a “timeless perspective” (TP) optimal-control solution that yields a targeting rule (TR) dependent on the central bank’s belief about structural persistence 휙ᵇˡᶠ.
- Three cases of policymaker belief vs. true persistence:
  - (i) Correct beliefs: 휙ᵇˡᶠ = 휙ᵗʳᵘᵉ leads to textbook optimal outcomes.
  - (ii) True persistence high but policymakers believe moderate: 휙ᵗʳᵘᵉ > 휙ᵇˡᶠ — produces the “worst” outcomes (large losses).
  - (iii) Policymakers act as if persistence is high when it is moderate: 휙ᵇˡᶠ > 휙ᵗʳᵘᵉ — plausibly the best robust outcome.

### Quantitative and illustrative results (model-based)
- Losses rise steeply with true inflation persistence when the central bank’s reference model assumes moderate persistence (example reference 휙ᵇˡᶠ = .4 while the range of possibilities considered is 휙 ∈ { .4, .8 } ). Relative losses are highest when true persistence lies at the upper end of the range policymakers entertain.
- Model impulse responses (illustrative):
  - When the central bank underestimates persistence, inflation peaks higher and takes far longer to return to target; the output gap becomes negative but not negative enough to cool demand sufficiently.
  - A “wait-and-see” correction path requires much sharper and more costly tightening than the robust policy.
- ECB application (January 2023 WEO baseline, vantage point of end-2022 / early-2023):
  - Baseline projection (staff, 2023Q1): quarterly average of ECB policy rate would peak at 3.7 percent.
  - Robust approach would require a peak up to 60 basis points higher than the baseline.
  - Empirical note: ECB has since hiked its deposit facility rate (DFR) to meet about half that distance.

### Model parameter values used in illustrative simulations (Table 1)
- 훼 .0625 — Relative weight on output gap stabilization
- 훽 .995 — Discount factor
- 휎 1 — CRRA
- 휉 .85 — Share of fixed prices (Calvo parameter)
- 휸 .1 — Openness (import share)
- Additional modeling notes:
  - The central bank places a relatively high weight on output-gap stabilization; openness parameter implies a relatively closed economy (similar to euro area or United States).
  - The targeting rule depends on the central bank’s belief about structural persistence (휙ᵇˡᶠ) and mismatches between belief and true 휙 generate welfare losses measured by the loss function.

*Source: Introduction, wpiea2024047-print-pdf*

### 1.5 Import elasticity of substitution

### 1.5 Import elasticity of substitution

### Key parameters and model primitives
- 휔 1.095
- Interest elasticity of demand = 1 + 훾(휎휂 − 1)(2 − 훾) / 휆  .052
- PC slope = ቂ(ଵِక)(ଵِఉక) క ቃቀఙଵାఠ൅1ቁ 휙 ௫ .75      Habit/demand persistence 휌 .92
- Cost shock autoregressive coefficient 훿 ௣ .8
- Behavioral (“Gabaix”) discount factor, Phillips Curve (symbol listed without numeric value in source)

### Inflation persistence and stabilization challenges
- Higher degrees of inflation persistence pose greater stabilization challenges; the extreme case 휙 ௣ = 1 and 훿 ௣ = 0 corresponds to an ‘accelerationist’ Phillips curve (change of inflation depends on slack).
- Lower persistence can make apparently better-than-optimal outcomes possible because a looser Phillips curve constraint allows incorrect policy to perform relatively well in that circumstance.

### Policy regimes examined
- Reference (moderate persistence): 휙 ௣ = 0.4
- Robust (worst-case) assumed persistence: 휙 ௣ = .8
- Wait-and-see: central bank corrects course when true persistence is revealed, assumed to occur five quarters after the shock (the economy runs hot for a year), and assumed completely unanticipated by the private sector.

### Dynamics under wait-and-see and robust rules
- Wait-and-see:
  - When persistence is high but initially unrecognized, the central bank raises the policy rate sharply upon recognition.
  - Immediate dampening of demand and output gap, but built-up inflation inertia cannot be quickly reversed; return to target is delayed and requires a longer period of tight policy.
- Robust rule (Eq. (TR) with 휙 ௣ ୮ୡୠ = .8):
  - Produces lower peak inflation and quicker return to target compared to the reference case when true persistence is moderate, at the cost of a wider output gap.
  - Nominal interest rate is initially raised more under the robust rule but falls back more quickly.

### Quantified costs and benefits of robustness
- Comparing outcomes via Eq. (LOSS):
  - The robust policy is around 9 percent worse than the policy in a moderate persistence world.
  - The worst-case error (erring in the opposite direction) produces losses that are 60 percent greater.
- Figure 7 (described):
  - Unconditional loss plotted relative to optimal policy when 휙 ௣ = 0.4.
  - Cost function is asymmetric: costs of being excessively dovish (acting as though persistence is lower than the truth) are lower than costs of being aggressive (acting as though persistence is higher than the truth).
  - Deviating from optimal policy is costly, but the costs of acting tough (robustness) are dwarfed by the cost of pursuing a baseline policy when persistence turns out to be high.

### Considerations for Central and Eastern European (CEE) economies
- Robust approach may yield higher benefits for CEE countries that do not use the euro:
  - Operate in small open economy environment with more frequent and persistent external shocks.
  - Inflation expectations are less resilient to de-anchoring.
  - Backward indexation in wage and price setting can be significantly higher due to weaker central bank credibility and a shallower Phillips curve.
- Existence of financial frictions and higher risk of inflation de-anchoring could justify using tools besides the interest rate (e.g., foreign exchange interventions) where financial markets are shallow and currency mismatches material.
- Even if other tools are used, the case for a robust approach remains.

### Robust policy analysis for the ECB (end-2022 / early-2023 perspective)
- Model and data:
  - Medium-scale DSGE model estimated on euro area data (1980-2016), featuring nominal and real distortions, financing and investment frictions; referred to Laureys, Meeks, and Wanengkirtyo (LMW, 2021).
  - Two germane features: price and wage stickiness (Calvo) and intrinsic price and wage inertia (partial indexation).
- Monetary policy rule (SR):
  - 푟 ௧ = 휌 ௥ 푟 ௧ِଵ + (1 − 휌 ௥)(휙 గ 휋 ௧ + 휙 ௬ 푦 ௧ + 휙 ୼୷ Δ푦 ௧) + 휀 ௧ ௥
  - Coefficients of the rule were estimated by LMW; fiscal policy passive.

### Steps in the robust policy analysis
- Step 1: Recover fundamental shocks (TFP, investment-specific tech, household preferences, autonomous spending incl. government, price mark-ups, wage mark-ups) using baseline (LMW modal parameters) and January 2023 WEO projections.
- Step 2: Hold shocks and policy rule constant; generate alternative economic structures by allowing price stickiness and intrinsic price/wage inertia to deviate toward an “inflation-prone” economy (parameter values from 90% HPD bounds).
- Step 3: With the alternative structure and shocks, search for the simple rule minimizing sum of squared deviations: inflation from 2% target, output from efficient level, and changes in nominal interest rates. Rule parameters constrained to [1,5] for inflation, [0,5] for output gap, and [0,.99] for adjustment speed.
- Step 4: Construct the robust policy rule as the rule that performs best in the inflation-prone economy; hold fundamentals constant and set structural parameters to estimated modes except the policy-rule parameters obtained in Step 3. Resulting endogenous paths are the robust outcomes.

### Calibrating the “inflation-prone economy”
- The parameterization raises wage and price inertia and steepens the price Phillips curve; values chosen from within the estimated 90 percent HPD interval.
- Table 2 (Parameterizing the Inflation-Prone Economy):
  - Price inertia (휄 ௣): Baseline value .0315; Alternative value .0908
  - Wage inertia (휄 ௪): Baseline value .2274; Alternative value .4030
  - Price stickiness (훾): Baseline value .7975; Alternative value .7543
- Interpretation:
  - Baseline price inertia modal estimate is very close to zero; upper bound of 90% HPD only slightly higher.
  - Wage inertia is of moderate size with lower precision.
  - Difference in 훾 translates into average duration of price fixity almost one quarter less in the inflation-prone economy (4 quarters rather than 5).

### Results for the euro area (simulation based on January 2023 WEO Update)
- Baseline projection (starts 2023Q1):
  - January 2023 WEO forecast: interest rates peaking at about 3.7 percent in 2024Q1.
  - Market pricing at that time peaked at 3.3 percent.
  - In the simulation, inflation falls in 2023 but remains marginally above target even in 2025.
- Inflation-prone scenario:
  - Outcomes deteriorate notably: elevated inflation path necessitates much tighter policy under the estimated rule, resulting in a sharp and prolonged recession.
  - The policy rule that performs best in this scenario is reported in Table 3.
- Table 3 (Policy Rule Parameters):
  - Estimated historical value: 휙 గ 1.68; 휙 ௬ .019; 휙 ୼୷ .15; 휌 ௥ .84
  - Robust value: 휙 గ 1.94; 휙 ௬ .00; 휙 ୼୷ .13; 휌 ௥ .90
- Key quantitative outcomes:
  - Applying the robust rule to the January 2023 WEO projection:
    - Policy rate peaks at 4.3 percent instead of 3.7 percent.
    - Inflation comes down faster and reaches the ECB’s target of 2 percent before the end of 2024.
    - Thereafter, rates are eased starting in 2024 because inflation comes close to the 2 percent target rather than plateauing.
  - Note: After publication, the ECB raised its interest rate twice to 4 percent and the WEO updated its terminal rate forecast to that value.
- Interpretation:
  - The robust rule features a stronger response to inflation, weaker overall response to output, and a higher degree of smoothing.
  - Robust policy reflects an implicit commitment to tighter future policy when inflation is above target.
  - The robust path should be viewed as the upper bound of a set of interest rate paths aligned with: (i) baseline forecast, (ii) central bank judgment add-ons, (iii) central bank’s aversion to uncertainty about backward indexation, and (iv) tightness of priors about backward indexation.
  - Central bankers with narrower credible regions for persistence parameters would choose a path between the reference and robust paths (shaded region in Figure 8).
  - Policy advice should be a range of options guided by a tightening bias; when persistence of inflation is highly uncertain, central bankers should tighten more (and possibly for longer) than indicated by the baseline forecast, and react more aggressively to positive inflation surprises than to negative ones.

*Italic: Source — wpiea2024047-print-pdf, section 1.5 Import elasticity of substitution.*

### Conclusion

### Conclusion

### Policy recommendation under uncertainty
- When the persistence of underlying inflation and wage dynamics is uncertain and inflation is too high, the best policy response may imply interest rate paths which are higher and/or more frontloaded than what would be implied by a routine, baseline policy reaction function.
- If inflation indicates a tendency to persistently dip below target, a prompt and decisive action by the central bank would prevent a prolonged period of inflation undershooting.
- The risk of hitting the ELB reinforces the case for a robust approach; a “wait-and-see” strategy incurs additional costs, including the risk of being forced to cut rates to the ELB and resort to unconventional monetary policy tools.
- The paper illustrates substantial costs associated with erroneous assumptions about persistence and adoption of a "wait-and-see" approach.
- Embracing a robust approach, even if inflation ultimately proves less persistent, results in only marginal deviations from the outcome under an ideal policy rule in a world devoid of uncertainty.

### ECB application and numerical scenarios (January 2023 WEO baseline)
- The paper constructs and analyzes various policy rate paths tailored to the European Central Bank (ECB) based on the January 2023 WEO forecast.
- Baseline scenario:
  - Baseline interest rate path for the January 2023 WEO: interest rates peaked at 3.7 percent.
  - In that baseline, inflation begins to recede but remains above the target in 2025.
- Persistent inflation scenario:
  - Should inflation prove more persistent, policy must be tightened more aggressively, potentially leading to a severe and protracted economic downturn.
- Robust-rule scenario:
  - The robust rule prescribes a quicker ascent in interest rates, culminating in a higher peak of up to 4.3 percent in the simulation based on the January 2023 WEO.
  - This proactive approach aims to steer inflation back to target levels by early 2025 rather than allowing it to plateau.
- Other considerations:
  - A more expansionary fiscal policy would underscore the merits of adopting a robust stance in monetary policy.

### Uncertainty sources and communication
- Central bankers must grapple with additional sources of uncertainty; incorporating scenario analysis into policy deliberations is a valuable complement to the conventional focus on the modal outlook (as highlighted by Bordo, Levin, and Levy (2020)).
- Scenario analysis can illustrate risks to the outlook and additional dimensions of model uncertainty, and be useful tools to inform monetary policy strategy and communication in periods of high uncertainty.
- The pace of monetary policy transmission remains crucial; depending on agents’ discounting behavior, transmission may either strengthen or weaken.
  - Cognitive discounting (Gabaix (2020)) potentially diminishes the potency of monetary policy.
  - A strong communication strategy to signal the central bank’s preference for decisive action (such as the robust approach) could offset some of the impact of cognitive discounting and backward indexation.
- If the slope of the Phillips curve is very flat, the central banker needs to raise rates a lot to create a large output gap before it can have a meaningful effect on inflation; under such conditions, monetary policy may need to tighten significantly and for a long time to bring inflation back to target—but the opposite could also be true.

### Modeling caveats
- The paper’s simulations for the robust path of monetary policy interest rates assume that transmission can be stronger than thought, but the paper does not provide an explicit account of that aspect of model uncertainty.
- Related literature and similar findings:
  - Tillmann (2011) reaches a similar conclusion that central banks applying a max-min strategy (i.e., wishing to avoid a worst case) would respond more strongly to inflation than under a baseline model.

*Source: wpiea2024047-print-pdf - Conclusion*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024047-print-pdf.pdf_