## wp18188

## Source details

**Canonical URL:** [wp18188](https://www.imf.org/-/media/files/publications/wp/2018/wp18188.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18188.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18188.pdf.json)

---

### Euro area inflation has been below the European Central Bank’s (ECB) inflation objective since 2013, adjusting very slowly to tightening labor market conditions

- Recent macroeconomic and inflation facts:
  - Unemployment rate declined from the peak of 12.1 percent in 2013 to 8.5 percent by Q2 2018.
  - More than 7.5 million jobs were created (period covering recovery after 2013).
  - As of June 2018, the less volatile core inflation measures were still below 1.5 percent.
  - Core inflation persistence: over 1999Q1 through 2017Q4, the simple correlation of core inflation with its first own lag is about 0.8 and with the fifth lag about 0.3.
  - Nominal effective exchange rate (NEER) of the euro appreciated by 2.1 percent (y/y) over 2017, and further appreciated by about 3.2 percent in 2018Q1 compared with the 2017 average.
  - Oil price and non-oil import prices had a downturn between 2014 and 2016.

- Puzzling inflation chronology:
  - 2000Q1–2011Q4: well-established Phillips curve (negative relationship between inflation and domestic slack).
  - From 2012 onward:
    - “Missing disinflation” over the two years starting early 2012: inflation reacted slowly to higher unemployment.
    - Followed by “missing inflation”: inflation stable despite large declines in unemployment.
  - Service inflation declined more than goods inflation after the global financial crisis, suggesting dominant domestic factors.

- Hypotheses and global factors considered:
  - Proposed mechanisms for weakened domestic Phillips curve include globalization of inflation, cheaper imports, and synchronization of business cycles via trade and global value chains.
  - Global variables explicitly considered: foreign output gap, foreign inflation, non-oil import prices, oil prices, and NEER growth.
  - Exchange rate appreciation and oil/import price movements can delay upward inflation adjustment.

- Empirical approach:
  - Focus: euro area-wide core inflation (headline excluding energy and unprocessed food), measured as annualized Q/Q core inflation.
  - Baseline model: New-Keynesian style expectation-augmented Phillips curve with:
    - Forward-looking inflation expectations (current five-year ahead survey inflation expectations).
    - Economic slack (unemployment gap).
    - Lagged inflation.
    - Vector of global variables (non-oil import prices, oil prices inflation, foreign inflation, foreign output gap, NEER growth).
  - Lags: up to eight lags for NEER and up to four lags for other variables.
  - Estimation sample: 1999Q1–2015Q4.
  - Model selection: General-to-Specific (GTS) from nine GUMs; out-of-sample forecast comparison period: 2016Q1–2017Q4.

### Main empirical findings on domestic vs global drivers

- Preferred specification:
  - Domestic Phillips curve including lagged inflation, long-term inflation expectations, and domestic slack (unemployment gap).

- Dominance of domestic factors:
  - Decomposition shows domestic factors are economically more meaningful than global factors in explaining recent inflation dynamics.
  - Global factors do not help explain the recent low inflation episode.

- GTS selection and forecast performance:
  - Exchange rate, non-oil import prices, and foreign inflation selected in some reduced models; oil prices and foreign output gap are not selected.
  - Some global-augmented Phillips curves improve out-of-sample RMSE versus the domestic benchmark; specifically three specifications with comparable or better out-of-sample performance:
    - Exchange rate alone;
    - Exchange rate combined with non-oil import prices;
    - Exchange rate combined with foreign inflation.
  - Preferred global Phillips curve: non-oil import prices + exchange rate + domestic variables — better out-of-sample performance in some tests.

- Limitations of global-augmented models:
  - Adding global factors increases in-sample goodness-of-fit only marginally (and sometimes decreases it).
  - Some global-augmented models fail diagnostic tests (residual autocorrelation, ARCH, normality, heteroskedasticity, general misspecification); the domestic Phillips curve passes diagnostics.
  - When heteroskedasticity present, robust standard errors mostly leave coefficient significance unchanged, with limited exceptions.

### Interpretation: persistence and labor market slack

- Role of persistence:
  - High persistence of euro area core inflation central to explaining “flat” Phillips curve: persistence implies longer transition after negative shocks, delaying upward adjustment as activity improves.
  - Inflation process in the euro area is more backward-looking than in the United States.
  - Possible sources of persistence include sluggish transmission from labor market developments to prices (wage dynamics and price-setting specificities).

- Labor market slack measurement:
  - Broader indicators (including underemployed and marginally attached) show larger “non-employment” than headline unemployment.
  - Evidence mixed on whether low inflation and weak wage growth reflect unmeasured cyclical slack; additional non-employment appears mainly structural in the euro area.

### The domestic Phillips curve therefore appears more robust than the global

- Main findings:
  - Domestic factors dominate global factors in explaining recent euro area inflation dynamics.
  - Preferred Phillips curve is the purely domestic specification augmented with lagged inflation and forward-looking inflation expectations, which “solves the inflation puzzle” for the euro area.

- Key empirical estimates and statistics:
  - Estimation sample: 1999Q1–2015Q4.
  - Coefficient on lagged core inflation: mostly stable at 0.6 across specifications.
  - Slope of inflation vs slack:
    - Unemployment rate specification slope: -0.13 (vs. -0.14 for the unemployment gap).
  - Goodness of fit: model fit drops from 0.95 to 0.65 if long-term inflation expectations are dropped.
  - In-sample sigma (one domestic specification): 0.36.
  - Reported R2 for a domestic model: 0.95.
  - Out-of-sample RMSE (2016Q1–2017Q4) for a domestic model: 0.46.

- Robustness checks:
  - Slack measures tested: unemployment gap (baseline), unemployment rate, non-employment gap, IMF and OECD output gaps.
    - Negative relationship between inflation and slack remains significant across measures.
    - Result not driven by underlying estimate of natural rate of unemployment.
  - Expectations variants:
    - Long-term (five-year) SPF expectations favored; robust to one- or two-year ahead expectations.
    - Forward-looking expectations statistically significant but smaller than backward-looking component.
  - Model comparisons: purely backward-looking or models excluding long-term expectations or unemployment gap perform worse on misspecification tests and out-of-sample performance (2016Q1–2017Q4).

- Comparison with the United States:
  - Euro area: largely backward-looking inflation process (lagged inflation significant and persistent).
  - U.S.: over same period (1999Q1–2015Q4) lagged core inflation not statistically significant; long-term inflation expectations dominate with coefficient not statistically different from unity.
  - Global factors have larger role in U.S. (e.g., non-oil import price growth larger coefficient).
  - Estimated slope on unemployment gap broadly similar between U.S. and euro area.

- Dynamics and projections:
  - Core inflation driven down from 2012 largely due to high unemployment; negative shocks dissipate slowly because of persistence.
  - In-sample residuals not consistently biased.
  - Tightening labor markets and closing output gaps should increasingly contribute to reflation as past negative shocks fade.
  - Increase in long-term inflation expectations along current trend could aid convergence of core inflation to the ECB objective.

- Explanations for limited role of global factors:
  - Domestic output gap closely follows global output gap; domestic slack can render global slack variables redundant.
  - Professional forecasters’ expectations may embed global inflation information; correlation between 5Y ahead SPF expectations and foreign inflation is about 0.3.
  - Exchange rate pass-through limited due to strong international competition, higher import content of exports, shift to sectors with low pass-through, and increased euro invoicing.
  - Energy prices have larger direct effect on headline than on core inflation; indirect/second-round impacts on core inflation are small (ECB (2010a) estimated about 0.02).

- Possible drivers of stickier inflation in the euro area:
  - More backward-looking inflation expectations and higher share of SMEs (about 60 percent of total value added vs. 50 percent in the U.S.) relying more on past inflation.
  - Labor market rigidities:
    - Average duration of wage agreements: two years in the euro area vs. about one year in the U.S.
    - Wage indexation prevalence: around 35 percent of wage contracts in the euro area vs. around 50 percent in the U.S.
  - Pricing power and barriers to price adjustment influence frequency of price changes.

- Role and anchoring of expectations:
  - Persistence need not imply unanchored expectations; backward-looking expectations can coexist with structural factors causing persistence.
  - Forward-looking expectations remain statistically significant and their removal worsens fit materially.

### Inflation expectations by professional forecasters and policy implications

- Inflation expectations measures:
  - Long-term SPF expectations now back to 1.86 percent (after bottoming at 1.77 percent in Q1 2014).
  - Market-based expectations bottomed at around 1.3 percent in mid-2016 and are now close to 1.8 percent.
  - Sequence of deflationary shocks could de-anchor expectations by reducing perceived inflation target.

- Policy implications and recommendation:
  - Convergence of core inflation toward the ECB’s medium-term objective is likely to be gradual.
  - Empirical analysis indicates low inflation outcomes largely result from inertia and high persistence, not disinflation elsewhere or global factors.
  - This supports continued strongly accommodative monetary policy:
    - Accommodative policy needed to support higher inflation via stronger demand and anchored expectations.
    - Premature tightening could quickly undo slow price adjustment with long-lived consequences.
  - ECB guidance:
    - Interest rates should remain low well beyond the end of net asset purchases.
    - Strong forward guidance should not be changed until actual inflation performance or a firm assessment shows decisive improvement.
    - Backward-looking inflation process reinforces relevance of forward guidance and the need to look through temporary upward shifts in inflation.
    - Forward guidance contributes to easy financial conditions and supports inflation through real activity.
    - Forward-looking expectations are significant in the domestic Phillips curve and guide future inflation.

- Appendix numerical highlights (selected):
  - Figure A.1 note: out-of-sample root-mean square errors of the global and domestic Phillips curves are respectively 0.46 and 0.39 (period 2016Q1–2017Q4).
  - Table A.1 (Domestic Phillips Curve Estimates — Slack Variants):
    - Lagged core inflation (first lag) coefficients: 0.646, 0.614, 0.617, 0.619, 0.672.
    - Second row coefficients: 0.327, 0.953, 0.327, 0.335, 0.269.
    - Third row coefficients: ‐0.141, ‐0.127, 0.103, 0.088, ‐0.333.
    - sigma (in‐sample): 0.36, 0.36, 0.35, 0.35, 0.37.
    - R2: 0.95, 0.95, 0.96, 0.96, 0.95.
    - RMSE (out of sample 2016Q1‐2017Q4): 0.46, 0.46, 0.43, 0.45, 0.75.
    - Estimation sample: 1999(1) ‐ 2015(4) (all columns).
    - Tests: heteroskedasticity test not passed in some variants.
  - Table A.2 (Domestic Phillips Curve Estimates — Expectations Variants):
    - Lagged core inflation (first lag) coefficients: 0.646, 0.516, 0.603, 0.661, 0.595.
    - Second row coefficients: 0.327, 0.485, 0.385, 0.314, 0.381.
    - Third row coefficients: ‐0.141, ‐0.153, ‐0.082, ‐0.138, ‐0.089.
    - sigma (in‐sample): 0.36, 0.36, 0.36, 0.37, 0.36.
    - R2: 0.95, 0.96, 0.96, 0.95, 0.96.
    - RMSE (out of sample 2016Q1‐2017Q4): 0.46, 0.48, 0.38, 0.45, 0.38.
    - Estimation samples reported (e.g., 1999(1) ‐ 2015(4); 2000(3) ‐ 2015(4) for one variant).
    - Notes: variants differ by measures of forward-looking expectations (five-year, two-year, one-year SPF; five-year and one-year consensus). Significance: *** p-value<0.01, ** p-value<0.05, * p-value<0.10.

*Source: wp18188.*

### 1.      Euro area inflation has been below the European Central Bank’s (ECB) inflation

### Euro area inflation has been below the European Central Bank’s (ECB) inflation objective since 2013, adjusting very slowly to tightening labor market conditions

### Recent macroeconomic and inflation facts
- Unemployment rate declined from the peak of 12.1 percent in 2013 to 8.5 percent by Q2 2018.
- More than 7.5 million jobs were created (period covering recovery after 2013).
- As of June 2018, the less volatile core inflation measures were still below 1.5 percent.
- Core inflation is highly persistent: over 1999Q1 through 2017Q4, the simple correlation of core inflation with its first own lag is about 0.8 and with the fifth lag about 0.3.
- Nominal effective exchange rate (NEER) of the euro appreciated by 2.1 percent (y/y) over 2017, and further appreciated by about 3.2 percent in 2018Q1 compared with the 2017 average.
- Oil price and non-oil import prices had a downturn between 2014 and 2016.

### Puzzling inflation dynamics and chronology
- From 2000Q1 to 2011Q4 there appeared to be a well-established Phillips curve (negative relationship between inflation and domestic slack).
- From 2012 onward an inflation puzzle emerged:
  - A “missing disinflation” episode over the two years starting early 2012, with inflation reacting slowly to higher unemployment rates.
  - Followed by a “missing inflation” period, with inflation remarkably stable despite considerable declines in the unemployment rate.
- Service inflation declined by more than goods inflation after the global financial crisis, suggesting domestic factors played a dominant role.

### Hypotheses and global factors considered
- Proposed mechanisms for a weakened domestic Phillips curve include globalization of inflation: low global inflation via cheaper imports; synchronization of foreign and domestic business cycles through trade and global value chains.
- Global variables explicitly considered in the analysis: foreign output gap, foreign inflation, non-oil import prices, oil prices, and NEER growth (nominal effective exchange rate).
- Other external factors noted: exchange rate appreciation and oil/import price movements can delay upward adjustment of inflationary pressures.

### Empirical approach and model specification
- Focus: euro area-wide core inflation (headline excluding energy and unprocessed food), measured as annualized Q/Q core inflation.
- Baseline model: New-Keynesian style expectation-augmented Phillips curve where core inflation is a function of:
  - Forward-looking inflation expectations (current five-year ahead survey inflation expectations).
  - A measure of economic slack (unemployment gap: unemployment rate less estimate of natural rate).
  - Lagged inflation (to capture persistence).
  - Vector of global variables (non-oil import prices, oil prices inflation, foreign inflation, foreign output gap, NEER growth).
- The general specification allows up to eight lags for NEER and up to four lags for other variables.
- Estimation sample for GTS and main exercises: 1999Q1–2015Q4.
- Model selection: General-to-Specific (GTS) methodology used to select relevant global variables from initial unrestricted models (five GUMs with single global variable each, plus four GUMs adding exchange rate to one additional global variable — nine GUMs total). Out-of-sample forecast comparison (2016Q1–2017Q4) used to assess predictive usefulness.

### Main empirical findings
- Preferred specification is a domestic Phillips curve including:
  - Lagged inflation,
  - Long-term inflation expectations,
  - Domestic slack (unemployment gap).
- Domestic factors dominate global factors in explaining recent inflation dynamics:
  - Decomposition of inflation drivers shows domestic factors are economically more meaningful than global factors.
  - Global factors do not help explain the recent low inflation episode.
- GTS selection results:
  - Exchange rate, non-oil import prices, and foreign inflation appear useful (selected in some reduced models).
  - Oil prices and foreign output gap are not selected in the reduced models.
- Forecast performance:
  - Some global-augmented Phillips curves improve out-of-sample RMSE relative to the domestic benchmark; specifically, three specifications had comparable or better out-of-sample performance than the domestic Phillips curve:
    - Exchange rate alone,
    - Exchange rate combined with non-oil import prices,
    - Exchange rate combined with foreign inflation.
  - The preferred global Phillips curve (non-oil import prices + exchange rate + domestic variables) presents better out-of-sample forecast performance in some tests.
- Limitations of global-augmented models:
  - Adding global factors increases in-sample goodness-of-fit only marginally (and sometimes decreases it).
  - Some global-augmented models fail standard diagnostic tests (e.g., residual autocorrelation, autoregressive conditional heteroskedasticity, normality, heteroskedasticity, general misspecification), while the domestic Phillips curve passes these diagnostics.
  - When global models do not pass heteroskedasticity tests, robust standard errors mostly leave coefficient significance unchanged, with limited exceptions noted.

### Interpretation and role of persistence
- High persistence of euro area core inflation is central to explaining the “flat” observed Phillips curve: persistence implies longer transition after negative inflation shocks, delaying upward adjustment even as activity improves.
- The inflation process in the euro area is more backward-looking than in the United States.
- Possible sources of persistence (not exhaustively analyzed here) include sluggish transmission from labor market developments to prices, linked to wage dynamics and price-setting specificities.

### Labor market slack measurement and its implications
- Broader labor market indicators (incorporating underemployed and marginally attached workers) show a larger degree of “non-employment” than the headline unemployment rate.
- Evidence is mixed on whether low inflation and weak wage growth reflect unmeasured labor market slack; in the euro area the additional non-employment appears mainly structural, so may not represent cyclical slack.

*Source: wp18188.*

### 18.      The domestic Phillips curve therefore appears more robust than the global

### The domestic Phillips curve therefore appears more robust than the global

### Main findings
- Domestic factors dominate global factors in explaining recent euro area inflation dynamics.
- The preferred Phillips curve for the euro area is the purely domestic specification (focus of the analysis).
- The estimated domestic Phillips curve "solves the inflation puzzle" for the euro area when augmented with lagged inflation and forward-looking inflation expectations.

### Key empirical estimates and statistics
- Estimation sample: 1999Q1–2015Q4.
- Coefficient on lagged core inflation: mostly stable at 0.6 across specifications.
- Slope of inflation vs slack:
  - Unemployment rate specification slope: -0.13 (vs. -0.14 for the unemployment gap).
- Goodness of fit: model fit drops from 0.95 to 0.65 if long-term inflation expectations are dropped.
- In-sample sigma (one domestic specification): 0.36; reported R2 for a domestic model: 0.95; out-of-sample RMSE (2016Q1–2017Q4) reported for a domestic model: 0.46.

### Robustness checks and alternative measures
- Slack measures tested: unemployment gap (baseline), unemployment rate, non-employment gap (including underemployed and marginally attached), and alternative output gaps (IMF and OECD).
  - Negative relationship between inflation and slack remains significant across these measures.
  - Result not driven by the underlying estimate of the natural rate of unemployment.
- Forward-looking inflation expectations tested: long-term and short-term expectations from ECB’s Survey of Professional Forecasters and Consensus Forecasts.
  - Long-term expectations (five years ahead) favored in main estimates; results robust to one- or two-year ahead expectations.
  - Forward-looking expectations remain statistically significant but smaller in magnitude than the backward-looking component.
- Model comparisons: purely backward-looking models or models excluding long-term expectations or unemployment gap perform worse on misspecification tests and out-of-sample performance (2016Q1–2017Q4).

### Comparison with the United States
- Euro area: inflation process is largely backward-looking (lagged inflation significant and persistent).
- U.S.: when estimated over the same period (1999Q1–2015Q4) with the same methodology, lagged core inflation does not appear statistically significant; long-term inflation expectations dominate realized inflation with an estimated coefficient not statistically different from unity.
- Global factors play a larger role in the U.S. than in the euro area (e.g., non-oil import price growth has a larger coefficient in U.S. Phillips curve).
- The estimated slope on the unemployment gap is broadly similar in magnitude between the U.S. and the euro area.

### Dynamics and projections
- Euro area core inflation was driven down from 2012 onwards largely due to high unemployment; negative shocks take time to dissipate because of inflation persistence.
- The model’s in-sample residuals are not consistently biased, indicating no systematic over- or under-estimation.
- Looking forward, tightening labor markets and closing output gaps should increasingly contribute to reflation as past negative shocks fade.
- A potential increase in long-term inflation expectations along the current trend could contribute to eventual convergence of core inflation toward the ECB’s objective.

### Explanations for limited role of global factors
- Domestic output gap closely follows global output gap; including domestic slack can render global slack variables redundant.
- Professional forecasters’ inflation expectations may incorporate global inflation information; correlation between 5Y ahead SPF expectations and foreign inflation is about 0.3.
- Exchange rate pass-through to prices is limited in the euro area due to strong international competition, higher import content of exports, a shift to sectors with low pass-through, and increased euro invoicing.
- Energy prices have a larger direct effect on headline inflation than on core inflation; indirect/second-round impacts on core inflation are small (ECB (2010a) estimated about 0.02).

### Possible drivers of stickier inflation in the euro area (compared with the U.S.)
- More backward-looking inflation expectations embedded in agents’ price- and wage-setting, potentially linked to a higher share of SMEs (about 60 percent of total value added vs. 50 percent in the U.S.) that may rely more on past inflation.
- Euro area labor market rigidities:
  - Average duration of wage agreements: two years in the euro area vs. about one year in the U.S.
  - Wage indexation prevalence: around 35 percent of wage contracts in the euro area vs. around 50 percent in the U.S.
  - U.S. wages are reset more frequently, contributing to less persistence.
- Pricing power and barriers to price adjustment (long-term customer relationships, product market regulation, degree of competition) can affect frequency of price changes and thus inflation stickiness.

### Role and anchoring of inflation expectations
- Inflation persistence need not imply expectations are unanchored; backward-looking expectations may coexist with other structural factors (labor market rigidities, pricing power) driving persistence.
- Forward-looking expectations still matter: they are statistically significant in models and their removal materially worsens fit.

*Source: wp18188 - 18.      The domestic Phillips curve therefore appears more robust than the global*

### 33.      In the case of the euro area, inflation expectations by professional forecasters

### 33.      In the case of the euro area, inflation expectations by professional forecasters

### Inflation expectations and market-based measures
- Long-term inflation expectations from the survey of professional forecasters have been relatively stable over time and are now back to 1.86 percent (after reaching a bottom of 1.77 percent in Q1 2014).
- Market-based inflation expectations, which also incorporate a compensation for inflation and liquidity risk premia and tend to be more volatile, bottomed at around 1.3 percent in mid-2016 and are now back to close to 1.8 percent.
- A sequence of deflationary shocks could contribute to de-anchoring inflation expectations through a reduction of the perceived inflation target (Busetti et al., 2017).
- Conclusion from observed measures: even if inflation expectations had declined over the recent past, they seem to be recovering gradually towards the ECB’s inflation objective of below but close to 2 percent.

### Policy implications and conclusion (paragraphs 34–36)
- Convergence of core inflation towards the ECB’s medium-term objective is likely to be gradual.
- Empirical analysis indicates low inflation outcomes have been largely the result of inertia and high persistence in the inflation process, rather than disinflation elsewhere or other global factors.
- This supports continued strongly accommodative monetary policy:
  - Accommodative policy is needed to support higher inflation through both stronger demand and anchored inflation expectations.
  - A premature tightening could quickly undo the slow process of price adjustment, with long-lived consequences.
- The ECB is correct to emphasize that interest rates will remain low well beyond the end of net asset purchases.
  - This strong forward guidance should not be changed until the performance of actual inflation supports it or there is a firm assessment that the inflation outlook has improved decisively.
  - The backward-looking nature of the inflation process reinforces the relevance of forward guidance by making it imperative to look through temporary upward shifts in inflation.
  - Forward guidance contributes to easy financial conditions and supports inflation through real activity.
  - Forward-looking expectations, although not preponderant, are still a significant factor of the domestic Phillips curve and guide future inflation.

### Appendix findings (Figures and Phillips-curve estimates)
- Figure A.1 note: Inflation forecasts based on the global and domestic Phillips curves presented in Table 1 (respectively second and last columns). The out-of-sample root-mean sqarue errors of the models are respectively 0.46 and 0.39 (on the period 2016Q1–2017Q4).
- Table A.1. Domestic Phillips Curve Estimates (Slack Variants) — key numeric entries (as presented):
  - Lagged core inflation (first lag) coefficients across columns: 0.646 0.614 0.617 0.619 0.672
  - Second row coefficients: 0.327 0.953 0.327 0.335 0.269
  - Third row coefficients: ‐0.141 ‐0.127 0.103 0.088 ‐0.333
  - sigma (in‐sample): 0.36 0.36 0.35 0.35 0.37
  - R2: 0.95 0.95 0.96 0.96 0.95
  - RMSE (out of sample 2016Q1‐2017Q4): 0.46 0.46 0.43 0.45 0.75
  - Estimation sample: 1999(1) ‐ 2015(4) (all columns)
  - Tests: heteroskedasticity test not passed in some variants as noted
- Table A.2. Domestic Phillips Curve Estimates (Expectations Variants) — key numeric entries (as presented):
  - Lagged core inflation (first lag) coefficients across columns: 0.646 0.516 0.603 0.661 0.595
  - Second row coefficients: 0.327 0.485 0.385 0.314 0.381
  - Third row coefficients: ‐0.141 ‐0.153 ‐0.082 ‐0.138 ‐0.089
  - sigma (in‐sample): 0.36 0.36 0.36 0.37 0.36
  - R2: 0.95 0.96 0.96 0.95 0.96
  - RMSE (out of sample 2016Q1‐2017Q4): 0.46 0.48 0.38 0.45 0.38
  - Estimation samples as reported (e.g., 1999(1) ‐ 2015(4); 2000(3) ‐ 2015(4) for one variant)
  - Notes: variants differ by measures of forward-looking expectations (five-year, two-year, one-year SPF expectations; five-year and one-year consensus expectations). Significance notation: *** p-value<0.01, ** p-value<0.05, * p-value<0.10.

*wp18188 - 33.      In the case of the euro area, inflation expectations by professional forecasters*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18188.pdf_