## CHAPTER 2 ## Source details **Canonical URL:** [CHAPTER 2](https://www.imf.org/-/media/files/publications/weo/2023/october/english/ch2.pdf) ## Other formats - [Markdown version](/-/media/files/publications/weo/2023/october/english/ch2.pdf.md) - [Structured JSON version](/-/media/files/publications/weo/2023/october/english/ch2.pdf.json) --- ### Introduction and motivation - Context: - Inflation reached multidecade highs in 2022 following the COVID-19 pandemic and Russia’s invasion of Ukraine. - Headline inflation is coming down as policy tightening rebalances demand, supply disruptions ease, and commodity prices decline, but core inflation remains elevated. - Professional forecasters expect inflation rates will return closer to central banks’ targets in 2024, with full return to targets only by 2026, on average. - Importance: - Expectations shape consumption, investment, and price- and wage-setting; they therefore influence inflation dynamics and the costs of achieving inflation objectives. ### Main empirical findings and recent patterns in expectations - Cross-agent dynamics: - Near-term (next-12-months) inflation expectations across professional forecasters, households, financial markets, and firms rose sharply in 2022 and broadly concur in timing. - Long-term (five-year-ahead) inflation expectations in the average economy have remained stable, with anchoring metrics indicating well-anchored long-term expectations in most economies. - Each indicator reached two-and-a-half to more than four standard deviations during the recent surge relative to experience since the early 2000s. - Historical persistence: - In historical episodes with persistently rising expectations, it took about three years for inflation and near-term expectations to return to pre-episode levels (median across identified episodes). - Historical sample: 32 episodes (16 AEs and 16 EMEs, 1989:Q4 to 2023:Q1). - Cross-economy contrasts: - For EMEs, the distribution of near-term expectations is wider and skewed up; median long-term expectations for EMEs moved up by a modest 10 basis points. - Anchoring of long-term expectations likely reflects active policy responses. - Key numeric statements (preserved): - Professional forecasters expect return closer to targets in 2024 and fully at targets only by 2026, on average. - Each indicator reached two-and-a-half to more than four standard deviations during the recent surge. - Median long-term inflation expectations for EMEs moved upward by a modest 10 basis points. ### Empirical framework and key estimation results - Framework: - Hybrid price Phillips curve relating current inflation to inflation expectations, lagged inflation, and the output gap. - Baseline uses near-term inflation expectations from professional forecasters; an instrumental variables approach using lags identifies causal impact. - Roles of expectations, lagged inflation, and output gap: - Near-term expectations matter most; long-term expectations have lower predictive power. - A one-standard-deviation increase in near-term expectations is associated with a 0.7 standard deviation increase in current inflation (across professional forecasters, financial markets, and firms). - Baseline associational estimates: a 1 percentage point rise in near-term expectations is associated with a 1.1 percentage point rise in current inflation among advanced economies, and about a 0.8 percentage point rise in emerging market economies (associational coefficients unadjusted for volatility range from 1.1 to 1.4). - Lagged inflation: little explanatory power in AEs; in EMEs, carryover from previous quarter’s inflation is about 0.2 percentage point and statistically significant. - Output gap: statistically significant for both groups; somewhat larger for EMEs. - Causal estimates (instrumental variables): - Accounting for reverse causation and omitted factors reduces effects by about 30 percent relative to associational estimates. - Causal pass-through estimates: - Average advanced economy: a 1 percentage point rise in near-term expectations raises inflation by about 0.8 percentage point. - Average emerging market economy: pass-through about 0.4 percentage point. - Interpretation: EMEs show more backward-looking formation (stronger role for lagged inflation). ### State dependence and high-inflation dynamics - Pass-through increases when inflation is elevated (above economy-specific sample median). - Example baseline: coefficient increases from 0.6 when inflation is low to higher values when inflation is elevated (excerpt notes statistical significance; full higher-value not provided in excerpt). - Empirical implication: the pass-through from expectations to inflation tends to be higher in periods of higher inflation. ### Model of expectations formation, shock propagation, and policy trade-offs - Model structure: - Semistructural dynamic stochastic general equilibrium model with expectational learning (extends Alvarez and Dizioli (2023)): price and wage Phillips curves, IS curve, monetary policy reaction function, heterogeneous agents (backward-looking learners and forward-looking learners), mutual influence between near- and long-term expectations (long-term affects inflation only through near-term). - Estimated shares of backward-looking learners: - Advanced-economy representative: about 20 percent. - Emerging-market representative: about 30 percent. - Difference used in policy intervention illustrations: about 8 percent. - Shock calibrations preserved: - Cost-push shock in impulse responses: increases inflation by 1 percentage point. - Monetary policy shock in impulse responses: increases the policy rate by 100 basis points. - High-inflation initial-condition simulation: model run for eight periods with inflation 2 percentage points above target. - Fiscal consolidation intervention: fiscal spending cut by 1 percent of GDP for two years. - Key model findings: - Cost-push shocks produce more persistent inflation under heterogeneous expectations than under rational expectations because backward-looking learners make expectations stickier. - Monetary policy is initially less effective at influencing inflation with heterogeneous agents because backward-looking learners do not internalize policy impacts on future marginal costs; policy can affect expectations mainly via output gap effects. - Sacrifice ratio (percent of output forgone to achieve a 1 percentage point faster reduction in inflation over three years) is larger in heterogeneous-agents model than in rational-expectations model; EMEs tend to have higher sacrifice ratios than AEs. - In a high-inflation environment, sacrifice ratios worsen slightly due to endogenous inflation de-anchoring by backward-looking learners. ### Policy interventions, timelines, and welfare trade-offs - Illustrative timeline outcomes under heterogeneous-agents model (baseline objective equally weights output gap and inflation; central bank knows expectations formation and future cost-push path): - Assumed cost-push shock: raises inflation 2 percentage points above target initially; estimated half-life of 14 quarters. - Baseline (equal weights): bring inflation back to target in about four years. - Double weight on inflation: about three years. - Central bank cares only about inflation: about two years, but lower welfare if society values output gap and inflation equally. - Only forward-looking learners: optimal to bring inflation back to target in about three years. - Less persistent shock (half-life 6.5 quarters): monetary policy could bring inflation back to target in less than four years; sometimes optimal to wait about two years. - Welfare comparison: - For identical welfare function, social welfare is about 20 percent higher with rational than with heterogeneous expectations. - Role of communications and frameworks: - Improvements in monetary policy frameworks and communications that increase the share of forward-looking learners (illustrative increase of about 8 percent) make policy more effective through expectations, lowering inflation faster with smaller output costs (softer landing). - Tighter cyclical policies (fiscal consolidation or monetary tightening) lower inflation more quickly but at larger output costs. - Practical implementation: improving frameworks and communications is powerful in the model but not a silver bullet; such interventions are complementary to conventional monetary policy. ### Policy implications and recommendations - Central banks should: - Understand expectations formation processes in their economies and tailor communications accordingly. - Invest in data collection and monitoring of expectations across agents, especially near-term expectations. - Use clearer, simpler, and more regular messaging targeting appropriate audiences (three Es: explanation, engagement, education). - Reinforce central bank independence, transparency, and operational effectiveness to raise the share of forward-looking learners. - Measurement and technology: - Technological improvements enable alternative measures of expectations (example: text-based analysis of firms’ earnings calls) to broaden and speed monitoring. - Fiscal–monetary interactions: - Prudent fiscal policy remains important: fiscal imprudence (high public debt) is associated with higher inflation expectations, particularly in EMEs; stronger monetary policy frameworks reduce this sensitivity. ### Illustrative empirical boxes and model simulations (key quantified outcomes preserved) - Box 2.1 (firms’ attention): - More attentive firms decrease their inflation expectations by about 1 percent of one standard deviation more than the average after four quarters; this corresponds to an amplification of about one-fourth to the sector’s average negative response. - Box 2.2 (fiscal imprudence and frameworks): - Uses the IAPOC index across 13 AEs and 37 EMDEs; improving frameworks reduces sensitivity of expectations to public debt. - Box 2.3 (euro area energy relief simulation): - Fiscal relief measures lowered euro area inflation by 0.9 percentage point in 2022 and by 0.5 percentage point in 2023 in the baseline where agents understand subsidies are temporary. - Measures increase core inflation expectations by 0.7 percentage point over 2023–24 in the baseline. - Alternative misperception scenario: impact in 2022 increases from –0.9 to –1.1 percentage points; 2023 impact increases from –0.5 to –0.6 percentage point. - Figure panels reference years 2022 23 24 25 and vertical scale tick labels: –2.0, –1.5, –1.0, –0.5, 0.0, 0.5, 1.0, 1.5, 2.0. ### Caveats, data limits, and interpretation - Data and identification: - Data limitations constrain cross-agent comparisons; analysis focuses on mean expectations (typically professional forecasters) for broad coverage. - Causal Phillips curve estimates rely on instrumental variables assumptions; results are largely robust to timing variations but should be treated as associational if assumptions fail. - Structural breaks could limit empirical inference; model incorporates a limited form of structural change through learning but is not exhaustive. - Model limitations: - Mapping from improvements in frameworks and communications to changes in the share of learner types is stylized and illustrative; other structural interventions (education, fiscal frameworks, governance) could affect expectations formation. *Source: WORLD ECONOMIC OUTLOOK: Navigating Global Divergences — CHAPTER 2, International Monetary Fund | October 2023 (IMF staff calculations).* ### Introduction ### Introduction ### Context and motivation - In the wake of the shocks of the COVID-19 pandemic and Russia’s invasion of Ukraine, inflation around the world reached multidecade highs in 2022, well above central bank targets, particularly in advanced economies. - As policy tightening gradually rebalances aggregate demand toward potential output, supply chain disruptions have eased, and commodity prices have declined, headline inflation is coming down, but underlying price pressures (as captured by core inflation) remain elevated. - Professional forecasters expect inflation rates will return closer to central banks’ targets in 2024, with a shift in their median deviation toward zero and a sharp narrowing of the distribution. However, they also expect that, given the current contractionary stance and anticipated policy action going forward, rates will be fully back at targets only by 2026, on average. - Expectations matter because consumption and investment decisions as well as price- and wage-setting processes partly reflect households’ and firms’ expectations about the future pace of price changes; expectations can therefore shape inflation dynamics and the costs of achieving inflation objectives. ### Questions addressed - How have inflation expectations across different agents and at alternative horizons behaved before and after the pandemic across economies? Are there signs of inflation expectations deanchoring since 2021? Or do the rapid interest rate hikes over 2022 appear to have contained risks? - How important are expectations in explaining inflation dynamics, particularly since the COVID-19 shock? Does the prevailing level of inflation (high or low) affect the explanatory power of inflation expectations? - How do expectations affect monetary policy effectiveness, and how does policy affect expectations? How does the expectations formation process affect the trade-offs that monetary policymakers face to bring inflation rates back to their targets? ### Main findings (empirical and model-based) - Across economic agents, movements in near-term (next-12-months) inflation expectations broadly concur, showing a sharp rise in 2022. Survey-based measures of expectations of professional forecasters and households, financial-market-implied expectations, and a newly constructed measure of firms’ expectations (based on the text analysis of firms’ earnings calls) fluctuate differently, but around a common trend. - Despite the sharp increase in inflation over 2022 across many economies, long-term (five-year-ahead) inflation expectations in the average economy have remained stable. According to multiple metrics—including inflation target deviations, expectations’ variability, and expectations’ disagreement—long-term expectations have remained well anchored in most economies. - Historical episodes characterized by initial periods of persistently rising expectations suggest that expectations come down only slowly. In these cases, it took about three years for inflation and near-term expectations to return to their pre-episode levels. Notably, real policy rates were lower and are now higher, on average, compared with those in past episodes, suggesting that monetary tightening since 2022 has been unusually sharp. - Near-term expectations are critical to understanding inflation dynamics and explain a growing share of inflation since 2022. Using a novel causal identification strategy to estimate Phillips curves, the chapter finds a strong role for inflation expectations in the group of advanced economies. In emerging market economies, lagged inflation is also important, suggesting a greater role for more backward-looking learners. - There are signs that the pass-through from inflation expectations to inflation tends to be higher in periods of higher inflation, such as those experienced recently worldwide. - The properties of the expectations formation process have a strong impact on the effectiveness of monetary policy. A newly developed dynamic stochastic general equilibrium model with a mix of forward- and backward-looking agents that learn demonstrates that the output costs of monetary tightening rise with the share of backward-looking learners in the economy or with the prevailing level of inflation. - The analysis shows that both inflation expectations and inflation would decline modestly more quickly with improvements in monetary policy frameworks and communication—such as simpler and more regular messaging and better targeting of audiences—that boost the share of forward-looking learners in the economy. However, such measures may take time or be more difficult to implement than tighter cyclical policies, which come with much higher costs in terms of slowing growth. - If central banks were to focus solely on bringing inflation down quickly, they would tighten even further and reduce the time required to bring inflation rates back to targets by two years, but at the cost of a sharper economic slowdown. When policymakers account for trade-offs among inflation close to target, output at potential, and smooth policy rate paths (helping manage financial stability concerns), a scenario for a representative advanced economy facing today’s inflation circumstances suggests that it is likely to take about three to four years for inflation and expectations to converge back to the central bank’s target. ### Policy implications and recommendations - Central banks benefit from having clear understandings of the expectations formation processes at work in their economies and tailoring their communications strategies accordingly, in parallel with structural reforms to reinforce central bank independence and transparency. - Managing expectations better could require investing more in data collection and monitoring of expectations, including across different agents. - Technological improvements mean that alternative methods of measuring expectations—such as the text-based analysis of firms’ earnings calls pioneered here—may make broader and timelier monitoring more feasible. - Improvements in monetary policy frameworks and communications that increase the share of forward-looking learners can modestly speed the decline of inflation and expectations, but these measures are illustrative and their mapping to changes in the share of learner types is stylized. ### Caveats and limitations - Data limitations constrain the empirical analysis of inflation expectations across exercises and cross-agent comparisons. To ensure the broadest sample coverage, the chapter takes a macroeconomic perspective and focuses on mean expectations, typically among professional forecasters, rather than the distribution or behavior of individual-level expectations. - The causal interpretation of the Phillips curve estimates is conditional on the assumptions of the instrumental variables estimation strategy based on lags. Findings are largely robust to varying the timing of the instruments, but if the underlying assumptions do not hold, the estimates should be interpreted as associational. - Structural breaks in the behavior of the economy could limit the informativeness of empirical and historical analyses. State dependence in the Phillips curve analysis addresses one possible form of break. The model-based analysis incorporates a limited form of structural change through learning but is not exhaustive. - The model-based analysis findings on the impact of improved monetary policy frameworks and communications on expectations and inflation are illustrative. The mapping from an increase in the share of forward- compared with backward-looking agents in the economy to monetary policy framework and communications improvements is stylized and other institutional or structural interventions (for example, educational attainment, fiscal frameworks, governance) could also be associated with changes in expectations formation. ### Approach and chapter structure - The chapter presents patterns in inflation expectations, focusing on the postpandemic recovery, and compares them with observed patterns after historical episodes in which expectations rose over an extended period. - It uses a novel identification approach to study the channel from expectations to inflation and how well recent inflation dynamics can be explained by expectations. - A model-based analysis with a mix of forward- and backward-looking learning agents examines how the expectations formation process may influence the conduct of monetary policy and vice versa. - The final section suggests potential policy actions in light of the chapter’s findings. ### Recent patterns in inflation expectations (summary) - Indicators of near-term inflation expectations across professional forecasters, financial markets, households, and firms (the latter constructed via text analysis of firms’ earnings calls) show broadly similar dynamics since 2017 for selected economies, agreeing on an inflation upswing from 2021, peaking in 2022, and a subsequent downswing. - Each indicator reached two-and-a-half to more than four standard deviations during the recent surge, highlighting the extraordinary size of the rise in inflation expectations compared with the experience since the early 2000s. - Different agents’ expectations have distinct properties: households’ expectations appear noisier and may lead or lag other agents; financial-market-implied expectations are continuously available but entangled with risk-premium fluctuations; firms’ near-term expectations tended to mark the upper bound of the cross-agent range during the recent surge; professional forecasters’ expectations convey more signal but may suffer from herding and strategic behavior. - For comparability across agents, expectations were transformed into z-scores for selected exercises; analyses mostly use professional forecasters’ expectations because of their broadest coverage across economies, time, and horizon. *Source: ch2 - Introduction (chapter text).* ### 1. United States ### ch2 - 1. United States ### Inflation expectations: recent patterns and cross-economy comparisons - Near-term inflation expectations shot up rapidly from 2022 but are now reverting. - Long-term inflation expectations have moved only marginally and within a narrowing range. - For emerging market economies (EMEs), the distribution of near-term inflation expectations is wider and skewed to the upside. - Median long-term inflation expectations for EMEs have moved upward by a modest 10 basis points. - Multiple anchoring metrics (average absolute deviations from target, variability over time, and disagreement across individuals) indicate long-term inflation expectations have stayed anchored despite recent rises in inflation. - The anchoring of long-term expectations likely reflects, in part, the active response of policymakers to dampen price pressures. *Historical context and persistence* - After past episodes in which near- and long-term inflation expectations rose persistently for a year or more, headline inflation and near-term expectations typically took about three years to revert to pre-episode levels (median across identified episodes). - Historical sample: 32 episodes identified, with 16 from AEs and 16 from EMEs (sample spans 1989:Q4 to 2023:Q1). - Recent paths for real policy rates and long-term inflation expectations differ from historical medians: real policy rates in 2022 were well below earlier comparative paths but are now well above the historical median; long-term expectations have been unusually stable coming into the recent high inflation regime. ### Empirical framework: hybrid price Phillips curve and identification - Framework relates current inflation to inflation expectations, lagged inflation, and the output gap. - Baseline specification uses near-term inflation expectations from professional forecasters for broader coverage. - An instrumental variables approach using lags of near-term inflation expectations and the output gap is used to identify the causal impact of expectations on inflation, addressing reverse causation and omitted factors. ### Key empirical findings: roles of expectations, lagged inflation, and output gap - Near-term expectations matter most: - Long-term expectations have lower predictive power than near-term measures. - A one-standard-deviation increase in near-term expectations is associated with a 0.7 standard deviation increase in current inflation (when considered across professional forecasters, financial markets, and firms’ near-term measures). - Households’ near-term expectations have a coefficient between near- and long-term measures of other agents. - Baseline (associational) estimates: - Estimated relationship suggests a 1 percentage point rise in near-term expectations is associated with a 1.1 percentage point rise in current inflation among advanced economies, and about a 0.8 percentage point rise in emerging market economies (associational coefficients unadjusted for volatility range from 1.1 to 1.4). - Lagged inflation and output gap: - Lagged inflation has little explanatory power in advanced economies (slightly negative and not statistically different from zero). - In EMEs, carryover from the previous quarter’s inflation is about 0.2 percentage point and is statistically significant. - The output gap has a statistically significant relationship with current inflation for both groups and is somewhat larger for EMEs. ### Causal estimates and magnitude of the expectations channel - Accounting for reverse causation and omitted factors reduces estimated effects of near-term expectations on current inflation by about 30 percent relative to associational estimates. - Causal pass-through estimates: - For the average advanced economy, a 1 percentage point rise in near-term expectations raises inflation by about 0.8 percentage point. - For the average emerging market economy, the pass-through is about 0.4 percentage point. - Interpretation: - Differences in magnitudes and the stronger role of lagged inflation in EMEs suggest expectations formation in EMEs tends to be more backward looking compared with AEs. ### Decomposition: contributions to recent inflation dynamics - For the average advanced economy: - Factors other than expectations and lagged inflation (including common global factors: COVID-19 disruptions, commodity price swings, global supply chain issues; economy-specific energy price effects; and the output gap) initially drove most of the increase in inflation over 2021–22. - Near-term inflation expectations have become a large and growing contributor in the most recent quarters. - Lagged inflation played a small role. - For the average emerging market economy: - Other factors also drove the peak in inflation in 2022. - Expectations played a significant but smaller role than in advanced economies. - Lagged inflation explained almost half of the average rise in inflation since 2020:Q1. ### State dependence: higher inflation environment increases pass-through - Estimates indicate the pass-through from inflation expectations to current inflation is higher when inflation is elevated (above its economy-specific sample median). - Example baseline: the coefficient increases from 0.6 when inflation is low (below its economy-specific sample median) to higher values when inflation is elevated (full higher-value not provided in the excerpt). *Sources: IMF staff calculations; Consensus Economics; Central bank websites; Haver Analytics; European Commission; NL Analytics; S&P Capital IQ.* ### 0.9 when inflation is high and statistically significant ### ch2 - 0.9 when inflation is high and statistically significant ### Expectations formation and model structure - Analysis extends the Alvarez and Dizioli (2023) semistructural dynamic stochastic general equilibrium model with expectational learning to include: - Price and wage Phillips curves, an IS curve, and a monetary policy reaction function. - A mix of heterogeneous agents: backward-looking learners and forward-looking learners (rational expectations). - Two additional features: heterogenous agents and mutual influence between near-term and long-term expectations (long-term expectations affect inflation only through near-term expectations). - Estimated shares of backward-looking learners: - Advanced-economy representative: about 20 percent. - Emerging-market representative: about 30 percent. - Alternative comparison: a model with only forward-looking learners (rational expectations). ### How expectations affect shock propagation and monetary transmission - Cost-push shock (examples: surprise rise in energy and commodity prices, supply-chain disruption): - Inflation increases persistently more under heterogeneous expectations than under rational expectations. - Backward-looking learners make inflation expectations more sensitive and stickier, assuming higher current inflation implies persistently higher future inflation. - Forward-looking learners view the cost-push shock as transitory and adjust expectations less. - Monetary policy effectiveness: - With heterogeneous agents, monetary policy is initially less effective at influencing inflation. - Reason: backward-looking learners do not account for the impact of monetary policy on future marginal costs; monetary policy can only influence expectations via direct effects on the output gap. - Model impulse-response calibration notes: - Cost-push shock examined increases inflation by 1 percentage point. - Monetary policy shock examined increases the policy rate by 100 basis points. ### Sacrifice ratio and state dependence - Sacrifice ratio definition used: the percentage of output forgone to achieve a 1 percentage point faster reduction in the inflation rate over a three-year period. - Key findings on the sacrifice ratio: - The sacrifice ratio is larger in the heterogeneous-agents model than in the rational-expectations model for both economy groups. - Emerging-market economies tend to have higher sacrifice ratios than advanced economies (reflecting higher shares of backward-looking learners). - In a high-inflation environment, sacrifice ratios worsen slightly because backward-looking learners behave as though inflation will be permanently higher, producing slight endogenous inflation de-anchoring. - High-inflation initial-condition simulation: - To simulate a high-inflation environment, the model is run for eight periods with inflation 2 percentage points above target to establish initial conditions. ### Monetary policy frameworks, communications, and interventions - Empirical associations: - The pass-through from inflation expectations to current inflation is higher when prevailing inflation is higher; the difference by prevailing inflation level is larger for advanced economies. - Weaker monetary policy frameworks (lower independence, transparency, communications) are associated with a higher share of economies in which forecast rationality of mean inflation expectations is rejected. - As monetary policy frameworks improve, deviations of near-term inflation expectations (or realized inflation rates) from targets are smaller. - Improving the share of forward-looking learners: - Improvements in monetary policy frameworks (central bank independence, transparency, communications, operational effectiveness) can increase agents’ attention to and understanding of policy and raise the share of forward-looking learners. - Example institutional actions cited (illustrative in the chapter): changes in target presentation, preset meeting calendars, and clearer primary objectives. - The model quantifies a stylized intervention that reduces the share of backward-looking learners by the measured difference between emerging-market and advanced representative economies (difference in share of backward-looking learners is about 8 percent). - Modelled policy interventions and assumed calibrations: - Monetary policy framework and communications improvements: assume share of forward-looking learners increases by about 8 percent relative to baseline. - Fiscal consolidation: assume fiscal spending is cut by 1 percent of GDP for two years and monetary policy does not offset the fiscal effort. - Monetary tightening: assume an initial 100 basis points rise in the policy rate on impact that then declines endogenously. - Policy outcomes in the model: - Improvements in monetary policy frameworks and communications that boost the share of forward-looking learners make monetary policy more effective via expectations, lowering inflation faster with smaller output costs (a softer landing). - Tighter cyclical policies (fiscal consolidation or monetary tightening) also lower inflation and inflation expectations but at larger output costs because they work partly through lowering aggregate demand initially. - Implementation challenges: improvements in frameworks and communications are powerful in the model but are not silver bullets; timely and effective implementation can be difficult, so such interventions are complementary to conventional monetary policy actions. ### Key illustrative numeric calibrations and scenario parameters preserved from the chapter - Estimated shares of backward-looking learners: about 20 percent (advanced), about 30 percent (emerging market). - Difference in share of backward-looking learners used in policy intervention illustrations: about 8 percent. - Cost-push shock magnitude in impulse responses: increases inflation by 1 percentage point. - Monetary policy shock magnitude in impulse responses: increases the policy rate by 100 basis points. - Fiscal consolidation intervention: fiscal spending cut by 1 percent of GDP for two years. - High-inflation initial-condition simulation: model run for eight periods with inflation 2 percentage points above target. - Sacrifice ratio: defined as percent of output forgone to achieve a 1 percentage point faster reduction in inflation over three years (chapter references a mode of seven in Tetlow (2022) across models as comparable context). *Source: WORLD ECONOMIC OUTLOOK: Navigating Global Divergences — CHAPTER 2, International Monetary Fund | October 2023 (IMF staff calculations).* ### CHAPTER 2 ### CHAPTER 2 ### Monetary policy faces inflation–output trade-offs: model setup and timelines - Baseline objective: central bank minimizes a welfare loss function that equally weights output gap and inflation deviations and includes interest rate smoothing. The central bank is assumed to know the expectations formation process and have full information on the path of future cost-push shocks. - Assumed cost-push shock in illustrative exercise: raises inflation 2 percentage points above target initially; shock has an estimated half-life of 14 quarters. - Timeline outcomes under heterogeneous agents’ model: - Baseline (equal weights on output gap and inflation): bring inflation back to target in about four years. - Double weight on inflation in objective: bring inflation back to target in about three years. - Central bank cares only about inflation (no weight on output gap): bring inflation back to target in two years, but this choice entails lower welfare if society values output gap and inflation deviations equally. - If there were only forward-looking learners in the economy: optimal to bring inflation back to target in about three years. - Less persistent shock scenario (half-life reduced to 6.5 quarters): monetary policy could bring inflation back to target in less than four years; even so, optimal wait about two years in some scenarios. - Welfare comparison: - For an identical welfare function, social welfare is about 20 percent higher with rational than with heterogeneous expectations, reflecting enhanced policy effectiveness and lower endogenous persistence of shocks. - Faster disinflation paths (placing greater weight on inflation) may come with welfare costs unless shocks are less persistent. ### Role of expectations formation and implications for policy effectiveness - Presence of backward-looking learners: - Larger share of backward-looking learners makes mean expectations more persistent and can get stuck at a higher level when inflation is higher for a sustained period. - This stickiness reduces the potency of monetary policy and increases the sacrifice ratio (output forgone) compared with purely forward-looking expectations. - Policy implication: - In the presence of a persistent cost-push shock and partially backward-looking expectations, it may be optimal to use a more extended timeline (up to four years under baseline) to bring inflation back to target when central banks equally weigh inflation and output gap deviations. - Disregarding output gap effects and tightening more aggressively can shorten the timeline (to two years) but at the cost of lower output. ### Empirical findings and broad conclusions on expectations and inflation dynamics - Near-term inflation expectations rose sharply across professional forecasters, financial markets, households, and firms amid the pandemic recovery and the 2022 cost-push shocks; long-term expectations have remained broadly stable on average, with no signs of de-anchoring. - Historical episodes with jointly rising near- and long-term expectations: about three years on average for inflation and near-term expectations to return to pre-episode levels, with wide variability across episodes. - Estimated hybrid Phillips curve: - Near-term inflation expectations play a more prominent role in explaining current inflation than long-term expectations. - For the average advanced economy, drivers of inflation have shifted from underlying cost-push shocks toward inflation expectations. - For the average emerging market economy, expectations play a smaller role than lagged inflation but still a significant one. - Pass-through of expectations to inflation increases when inflation is already elevated. ### Recommendations on data, frameworks, and communication - Improved data on expectations: - Close monitoring and enhanced collection of information on expectations across economic agents, particularly near-term expectations, which appear more important for current inflation dynamics. - Technological advances enable cost-effective, timely extraction of expectations (example: firms’ inflation expectations from text analysis of firms’ earnings calls). - Monetary policy frameworks and communication: - Enhancing central bank independence and transparency and improving communication strategies can boost the share of forward-looking learners and thereby the effectiveness of monetary policy. - Communication guidance (three Es): explanation, engagement, and education. Focus on audience segmentation, simple and repeated messages, investing in financial literacy, emphasizing goals rather than instruments, and targeting messages to the conjuncture. - Exposure to news improves precision of perceptions and expectations, increases confidence, and lowers dispersion of beliefs. ### Box 2.1 — Firms’ inflation expectations, attention, and monetary policy effectiveness - New firm-level index: firm-level index of near-term inflation expectations based on text analysis of firms’ earnings calls; parallel construction of an Earnings-Calls-based Firm Attention to the Central Bank index (ECFACB) measuring intensity of discussion related to the Federal Reserve. - Empirical approach: - Dynamic responses estimated using local projections to assess effect of a monetary policy shock on a firm’s inflation expectations conditional on the firm’s attentiveness to monetary policy. - Specification includes an interaction between a US monetary policy shock measure and an attention index, firm and time fixed effects, and firm-level controls (sales growth, leverage, employment, total assets, and share of current assets in total assets). Standard errors are two-way clustered by firms and time. - Shocks scaled to have unit standard deviation; attentiveness by firm is de-meaned by sectoral average attentiveness. - Key quantified findings: - More attentive firms decrease their inflation expectations by about 1 percent of one standard deviation more than the average after four quarters. - This corresponds to an amplification of about one-fourth to the sector’s average negative response. - Implication: Monetary policy is more effective when agents pay attention to monetary policy and understand central bank decisions. ### Box 2.2 — Fiscal imprudence and inflation expectations: interactions with monetary policy frameworks - Fiscal policy and expectations: - Fiscal imprudence—high levels of public debt to GDP—is associated with uncertainty and can influence inflation expectations by eroding perceptions of monetary policy credibility and independence. - Empirical study: - Uses the IAPOC index (Independence and Accountability; Policy and Operational Strategy; Communications) to capture soundness of monetary policy frameworks across 13 advanced economies and 37 emerging market and developing economies; data updated to 2021. - Fixed-effects panel regression of mean inflation expectations on interaction of IAPOC index score and debt to GDP, controlling for economy-specific characteristics and time-invariant effects. - Key findings: - Higher public debt is associated with expectations of higher inflation, conditional on a given level of monetary policy framework (effect evident in emerging market and developing economies; advanced economies do not show this differential sensitivity). - Impact is larger when focusing on stock of public debt in foreign currency and exacerbated when fiscal deficits are persistent. - Improving monetary policy frameworks (IAPOC index distribution shifted right between 2007 and 2021 in emerging market and developing economies) reduces sensitivity of inflation expectations to level and composition of public debt and to persistent deficits. - Policy implication: Strong monetary policy frameworks can ease difficulties posed by higher public debt for managing inflation expectations, but prudent fiscal policy remains key to prevent fiscal dominance. ### Box 2.3 — Fiscal relief for the 2022 energy shock: model simulation for euro area - Simulation setup: - IMF’s Flexible System of Global Models used to simulate impacts on expected and realized inflation of announced energy relief measures (price subsidies and caps) in the euro area. - Assumes the sharp upward shock to energy prices in 2022 is temporary and unwinds; includes indirect effects of energy prices on core inflation through the supply chain. - Quantified model results (baseline where agents understand subsidies are temporary): - Fiscal relief measures lowered euro area inflation by 0.9 percentage point in 2022. - Fiscal relief measures lowered euro area inflation by half a percentage point in 2023. - Measures smooth the inflation impact of the energy shock over time, leading to a rise in inflation over 2024–25 relative to the no-measures scenario and preventing an undershoot as subsidies expire and the energy shock unwinds. - Net neutral effect on core inflation expectations in 2022 but increase expectations by 0.7 percentage point over 2023–24. - Alternative scenario (agents misperceive and think subsidies last one year longer than announced): - Expectations fall more in 2022; firms lower prices by more in 2022 because they expect core inflation to be lower in 2023. - Fall in inflation expectations increases impact of fiscal policy on inflation from –0.9 to –1.1 percentage points in 2022. - Impact on 2023 increases from –0.5 to –0.6 percentage point. - Once agents correct the misperception, inflation and expectations bounce back, highlighting the expectations channel. - Caveat: Effectiveness and desirability of energy relief measures depend on many factors beyond the box (impact on energy markets, resource misallocation, fiscal sustainability, and policy design details). *International Monetary Fund | October 2023* ### 1. Channels ### 1. Channels ### Overview - Analysis focuses on the marginal impacts on inflation of announced fiscal relief measures for energy, using the IMF’s Flexible System of Global Models. - Visual panels reference years: 2022, 23, 24, 25 and vertical scales from –2.0 to 2.0 (with tick marks at –1.5, –1.0, –0.5, 0.0, 0.5, 1.0, 1.5 in the figure annotation). ### Panel findings (figure note) - Panel 1: Marginal impacts on inflation of announced fiscal relief measures for energy. - Blue bars show the direct effects of measures (subsidies, tax cuts, or price caps on consumer energy prices). - Red bars show the indirect effects from changes in aggregate demand, supply chain costs, and core inflation expectations. - Panel 2 (Headline): The baseline assumes fiscal relief measures last in 2022 as originally announced. - The alternative assumes that households misperceive and expect measures will last longer, but then in 2023 they realize their error and adjust to the announced path. - Panel 3 (Core Inflation Expectations): documents movements in core inflation expectations across 2022, 23, 24, 25 on the same –2.0 to 2.0 scale. ### Interpretation and channels of transmission - Direct channel: subsidies, tax cuts, and price caps reduce consumer energy prices (blue bars), mechanically lowering headline inflation in the period measures are in effect. - Indirect channels (red bars): - Changes in aggregate demand resulting from fiscal relief can raise inflationary pressures. - Supply chain costs may be affected, transmitting to broader price levels. - Core inflation expectations can be influenced by the perceived persistence of relief measures, affecting wage-setting and price-setting behavior. ### Scenario comparison and expectations effects - Baseline scenario: measures are perceived to last only in 2022 (as announced), implying limited persistent effect on inflation beyond their active period. - Alternative scenario: households initially misperceive measures as lasting longer, boosting demand and expectations; in 2023 households adjust to the announced path, producing a correction in expectations and associated inflation dynamics. ### Box reference - Box 2.3 title: Energy Subsidies, Inflation, and Expectations: Unpacking Euro Area Measures ### Key numeric labels and timeline points preserved from the figure notes - Years shown: 2022 23 24 25 - Vertical scale tick labels: –2.0, –1.5, –1.0, –0.5, 0.0, 0.5, 1.0, 1.5, 2.0 *Source: Dao and others (2023); and IMF staff calculations.* --- _Source: https://www.imf.org/-/media/files/publications/weo/2023/october/english/ch2.pdf_