## 1. Introduction (and related analyzed sections)

## Source details

**Canonical URL:** [1. Introduction (and related analyzed sections)](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022221-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022221-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022221-print-pdf.pdf.json)

---

### Recent developments and motivation
- Inflation in 2021 rose to levels that had not been seen in almost 40 years in many economies.
- Labor demand surged with the recovery while labor supply was slow to respond, contributing to tighter labor markets and increased wage pressures; average nominal wages (per worker) rose and the unemployment rate fell from the second half of 2020 across economy groups.
- Observers worry about a potential wage-price spiral: rising inflation and tight labor markets prompting nominal wage increases that catch up to or exceed inflation (Blanchard, 2022).
- Empirical concerns:
  - Vacancy and quit rates having substantial predictive power for wage inflation (Domash and Summers, 2022).
  - Firms and business leaders reporting greater attention to aggregate inflation measures and incorporating them into pricing decisions (Schwartzman and Waddell, 2022).
  - The possibility that the responsiveness of non-union wages to inflation increases with higher inflation (Holland, 1988).

### Definition and empirical strategy
- Definition adopted: a wage-price spiral episode is one where at least three out of four successive quarters saw accelerating consumer prices and nominal wages.
- Applying this definition:
  - 79 episodes identified in the baseline database.
  - 100 episodes identified when using a narrower wage concept covering a longer time period.
- Analytical approach:
  - Illustrate behavior of key macroeconomic indicators around these episodes.
  - Decompose wage dynamics through the lens of a wage Phillips curve to link wage growth to inflation and labor market dynamics and to assess deviations from relationships observed during normal times.

### Key empirical findings (summary)
- The great majority of identified episodes are not followed by a sustained acceleration in wages and prices; only a few exceptions exist.
- Inflation and nominal wage growth tended to stabilize in the quarters following episodes, leaving real wage growth broadly unchanged.
- Although nominal wage growth deviates from established wage Phillips curve relationships during the initial acceleration stage, it eventually stabilizes to levels consistent with observed inflation and unemployment.
- Wage-price spiraling dynamics appear to have short lives.
- Historical analogs:
  - After episodes similar to recent developments—characterized by accelerating inflation, increasing nominal wages, falling real wages, and falling unemployment—past patterns show declining inflation while nominal wage growth increased, allowing real wages to catch up.
  - Decompositions via the wage Phillips curve indicate that labor market tightening resulting from real wage falls is a significant driver of sustained—but eventually stabilizing—nominal wage growth after such episodes.
- Implication: acceleration of nominal wages should not be taken as definitive evidence that a sustained wage-price spiral is taking hold; history indicates nominal wages can accelerate while inflation recedes from high levels.

### Data and measurement (summary)
- Dataset coverage:
  - 38 advanced economies in the compiled dataset.
  - Period covered: 1960:Q1 – 2021:Q4.
  - In some analyses the sample size drops to 31 economies because data for all variables is required.
- Main variables and measurement details:
  - Price level: Consumer price index; quarterly, 1960:Q1 – 2021:Q4; N = 36.
  - Energy price index: Period average end-use energy price index; quarterly, 1960:Q1 – 2021:Q4; N = 38.
  - Nominal wage (average per person, local currency, index): quarterly, 1960:Q1 – 2021:Q4; N = 31.
  - Nominal wage (manufacturing, avg hourly earnings, local currency, index): quarterly, 1960:Q1 – 2021:Q4; N = 29.
  - Unemployment rate: quarterly, 1960:Q1 – 2021:Q4; N = 38.
  - Real GDP (output-side, Chained PPP, mil. 2017 USD): annually, 1950 – 2019; N = 31.
  - Productivity (TFP level at current PPP, USA=1): annually, 1950 – 2019; N = 31.
- Data construction and harmonization:
  - Quarterly series seasonally adjusted using X-13ARIMA-SEATS if not already adjusted.
  - When combining sources to extend coverage, OECD data prioritized; extensions use growth rates from other sources (typically ILO, Haver, or IMF) to splice series forward and backward.
  - Inflation measured as year-on-year growth of the consumer price index at quarterly frequency.
  - Nominal wages measured as average earnings per worker in local currency units; primary source OECD wage rate indicator, extended with Haver Analytics where needed; all earnings data are in local currency and annualized.
  - Unemployment measured as unemployed divided by active labor force; primary source OECD, extended with ILO, Haver Analytics and IFS.
  - Productivity for the wage Phillips curve uses real GDP per worker from Penn World Table 10.0 with a five-year moving average as a proxy for slow-moving productivity trends.

### Paper outline
- Section 2: presents the data.
- Section 3: identifies past episodes of wage-price spirals and analyzes subsequent macroeconomic dynamics.
- Section 4: analyzes wage dynamics after those episodes through a wage Phillips curve decomposition.
- Section 5: discusses episodes most similar to current macroeconomic conditions to assess wage-price spiral risks.
- Section 6: concludes.

### Historical experiences (selected findings from Section 3)
- Identification rule: an episode where both price and nominal wage inflation (measured year-over-year) increase successively for at least three out of four consecutive quarters.
- If criteria hold several times within three years, only the first episode is selected.
- Period 0 is defined as the first period where the criteria that define a wage-price spiral are met.
- Episodes identified:
  - Applying the definition to the sample with aggregate nominal wages identifies 79 episodes.
    - First identified episode: 1973.
    - Last identified episode: 2017.
  - Using the narrower manufacturing-wages concept identifies 100 episodes.
- Trends in prevalence:
  - Episodes with price and wage accelerations have become less prevalent since the 1970s, clearest when using the manufacturing-wage concept.
- Macroeconomic dynamics around episodes:
  - Pre-period behavior: both consumer price inflation and nominal wage growth increase before period 0 (by construction).
  - Post-period average dynamics:
    - Inflation and nominal wage growth on average tended to stabilize in the quarters following the wage-price spiral.
    - Real wage growth was broadly unchanged on average after episodes.
    - The unemployment rate tended to edge down slightly following episodes.
- Notable extreme episode example:
  - United States 1973:Q3 (spurred by the first OPEC oil embargo):
    - Price inflation surged for five additional quarters before beginning to fall in 1975.
    - Nominal wage growth did not increase during that episode, leading to a decline in real wage growth.

### Episodes similar to the COVID-19 shock (selected findings from Section 5)
- Selection criteria for similarity to COVID-19: at least three out of four consecutive quarters with (i) increasing year-on-year inflation, (ii) positive nominal wage growth, (iii) negative real wage growth, and (iv) flat or falling unemployment.
- Applying those criteria to the dataset with aggregate nominal wages yields 22 episodes.
- Average post-episode macroeconomic developments for these 22 episodes:
  - Inflation tended to decline.
  - Nominal wage growth tended to increase.
  - Real wages started increasing again.
  - The unemployment rate tended to fall.
  - These episodes were followed by a higher increase in wage growth than in the wider set of episodes, but wage growth eventually stabilized rather than continuing to accelerate indefinitely.
- Heterogeneity and notable case:
  - United States, 1979:Q2 — inflation rose rapidly for four quarters after the episode before starting to decline; unemployment rose more than in other identified episodes; aggressive monetary tightening (the Volcker disinflation) underlay outcomes; nominal wage growth was relatively flat, producing an early decline in real wages.
- Decomposition of nominal wage growth (wage Phillips curve):
  - Using pooled coefficients from column (5) of Table 4.1, nominal wage growth is decomposed across the 22 episodes.
  - Findings:
    - In the periods up to period zero, wage growth lagged that expected by inflation and unemployment gap movements alone — consistent with falling real wages.
    - A negative and decreasing "other" component is visible up to period zero.
    - After period zero, nominal wage growth starts to catch up and by two-years after the episodes nominal wage growth is broadly consistent with higher inflation and tighter labor markets.
    - The shrinking "other" component near the end of the episode window illustrates this catch-up.
    - The positive contribution from the unemployment gap indicates tighter labor markets helped the catch-up; this role is greater than in the wider set of episodes documented in Section 4.
    - The initial fall in real wages appears to have led to tighter labor markets that co-existed with decelerating inflation.
  - Note: A full decomposition of wage growth around the 2021 wage-price episode following the pandemic cannot yet be done at the same horizon. Analysis up to 2021Q4 suggests rising inflation and labor market tightening have acted to increase nominal wage growth, while other factors have kept nominal wage growth contained.
- Robustness: conditioning on energy prices
  - Replacing headline inflation with increasing year-on-year energy inflation and applying analogous selection criteria yields 13 episodes.
  - The 13 episodes exhibit similar average post-episode dynamics:
    - Nominal wage growth tended to increase.
    - Headline and energy inflation tended to decline.
    - Real wages began to increase again.
    - Unemployment tended to fall.
  - Extending the sample backward using manufacturing wages and extrapolated energy prices for economies outside the United States yields patterns robust to the extension.
- Implications and conclusion:
  - Of the 79 episodes identified with accelerating prices and wages (back to the 1960s), only a minority saw further acceleration after eight quarters.
  - Sustained wage-price acceleration is even harder to find when conditioning on episodes similar to today (where real wages have fallen).
  - In those cases, nominal wages tended to catch up to inflation to partially recover real wage losses, and growth rates stabilized at a higher level than before the initial acceleration.
  - Wage growth rates eventually became consistent with inflation and labor market tightness; this did not typically lead to persistent acceleration dynamics that can be characterized as a wage-price spiral.
  - Key takeaway: An acceleration of nominal wages should not necessarily be interpreted as evidence that a wage-price spiral is taking hold; history shows nominal wages can accelerate while inflation recedes from high levels.

*Source: INTERNATIONAL MONETARY FUND — "1. Introduction", Sections 3 and 5 from the supplied content.*

### 1. Introduction ........................................................................................................

### 1. Introduction

### Recent developments and motivation
- Inflation in 2021 rose to levels that had not been seen in almost 40 years in many economies (Figure 1.1, panel A).
- Labor demand surged with the recovery while labor supply was slow to respond, contributing to tighter labor markets and increased wage pressures; average nominal wages (per worker) rose (Figure 1.1, panel B) and the unemployment rate fell from the second half of 2020 across economy groups.
- Observers worry about a potential wage-price spiral: rising inflation and tight labor markets prompting nominal wage increases that catch up to or exceed inflation (Blanchard, 2022).
- Empirical concerns include:
  - Vacancy and quit rates having substantial predictive power for wage inflation (Domash and Summers, 2022).
  - Firms and business leaders reporting greater attention to aggregate inflation measures and incorporating them into pricing decisions (Schwartzman and Waddell, 2022).
  - The possibility that the responsiveness of non-union wages to inflation increases with higher inflation (Holland, 1988).

### Definition and empirical strategy
- Definition adopted: a wage-price spiral episode is one where at least three out of four successive quarters saw accelerating consumer prices and nominal wages.
- Applying this definition:
  - 79 episodes identified in the baseline database.
  - 100 episodes identified when using a narrower wage concept covering a longer time period.
- Analytical approach:
  - Illustrate behavior of key macroeconomic indicators around these episodes.
  - Decompose wage dynamics through the lens of a wage Phillips curve to link wage growth to inflation and labor market dynamics and to assess deviations from relationships observed during normal times.

### Key empirical findings
- The great majority of identified episodes are not followed by a sustained acceleration in wages and prices; only a few exceptions exist.
- Inflation and nominal wage growth tended to stabilize in the quarters following episodes, leaving real wage growth broadly unchanged.
- Although nominal wage growth deviates from established wage Phillips curve relationships during the initial acceleration stage, it eventually stabilizes to levels consistent with observed inflation and unemployment.
- Wage-price spiraling dynamics appear to have short lives.
- Historical analogs:
  - After episodes similar to recent developments—characterized by accelerating inflation, increasing nominal wages, falling real wages, and falling unemployment—past patterns show declining inflation while nominal wage growth increased, allowing real wages to catch up.
  - Decompositions via the wage Phillips curve indicate that labor market tightening resulting from real wage falls is a significant driver of sustained—but eventually stabilizing—nominal wage growth after such episodes.
- Implication: acceleration of nominal wages should not be taken as definitive evidence that a sustained wage-price spiral is taking hold; history indicates nominal wages can accelerate while inflation recedes from high levels.

### Data and measurement (summary)
- Dataset coverage:
  - 38 advanced economies in the compiled dataset.
  - Period covered: 1960:Q1 – 2021:Q4.
  - In some analyses (e.g., Section 4) the sample size drops to 31 economies because data for all variables is required.
- Main variables and measurement details (see Table 2.1):
  - Price level: Consumer price index; quarterly, 1960:Q1 – 2021:Q4; N = 36.
  - Energy price index: Period average end-use energy price index; quarterly, 1960:Q1 – 2021:Q4; N = 38.
  - Nominal wage (average per person, local currency, index): quarterly, 1960:Q1 – 2021:Q4; N = 31.
  - Nominal wage (manufacturing, avg hourly earnings, local currency, index): quarterly, 1960:Q1 – 2021:Q4; N = 29.
  - Unemployment rate: quarterly, 1960:Q1 – 2021:Q4; N = 38.
  - Real GDP (output-side, Chained PPP, mil. 2017 USD): annually, 1950 – 2019; N = 31.
  - Productivity (TFP level at current PPP, USA=1): annually, 1950 – 2019; N = 31.
- Data construction and harmonization:
  - Quarterly series seasonally adjusted using X-13ARIMA-SEATS if not already adjusted.
  - When combining sources to extend coverage, OECD data prioritized; extensions use growth rates from other sources (typically ILO, Haver, or IMF) to splice series forward and backward.
  - Inflation measured as year-on-year growth of the consumer price index at quarterly frequency.
  - Nominal wages measured as average earnings per worker in local currency units; primary source OECD wage rate indicator, extended with Haver Analytics where needed; all earnings data are in local currency and annualized.
  - Unemployment measured as unemployed divided by active labor force; primary source OECD, extended with ILO, Haver Analytics and IFS.
  - Productivity for the wage Phillips curve uses real GDP per worker from Penn World Table 10.0 with a five-year moving average as a proxy for slow-moving productivity trends.

### Paper outline
- Section 2: presents the data.
- Section 3: identifies past episodes of wage-price spirals and analyzes subsequent macroeconomic dynamics.
- Section 4: analyzes wage dynamics after those episodes through a wage Phillips curve decomposition.
- Section 5: discusses episodes most similar to current macroeconomic conditions to assess wage-price spiral risks.
- Section 6: concludes.

*Source: INTERNATIONAL MONETARY FUND — "1. Introduction" from the supplied content.*

### 3. Historical Experiences

### 3. Historical Experiences

### Identification of wage-price spiral episodes
- Definition used: an episode where both price and nominal wage inflation (measured year-over-year) increase successively for at least three out of four consecutive quarters.
- If criteria hold several times within three years, only the first episode is selected.
- Period 0 is defined as the first period where the criteria that define a wage-price spiral are met.

### Data, measurement choices, and sample coverage
- Primary wage concept: aggregate average wages (or earnings) per worker from Haver Analytics; preferred over per-hour wages to allow broader country and time coverage.
- Alternative (narrower) concept: average hourly earnings in manufacturing (OECD) used for robustness and to extend coverage into the 1970s.
- Data vetting: each series was individually verified to avoid abnormal spikes or sudden level changes when combining sources.
- Sample availability varies across exercises; annex tables list economies and sources by quarter.
- Typical data start dates: for most economies data series start in the 1980s or 1990s; manufacturing-wages sample allows inclusion of the 1970s for a wider set of economies.

### Episodes identified
- Applying the definition to the sample with aggregate nominal wages identifies 79 episodes.
  - First identified episode: 1973.
  - Last identified episode: 2017.
- Using the narrower manufacturing-wages concept identifies 100 episodes.
- Table 3.2 lists the 79 episodes (selected entries include: United States 1973:Q3, United States 1978:Q4, United States 1987:Q3, United States 1996:Q4, United States 2000:Q3, United States 2017:Q3; other country-quarter episodes are listed in Table 3.2).

### Trends in prevalence
- Episodes with price and wage accelerations have become less prevalent since the 1970s.
- This decline is clearest using the narrower manufacturing-wage concept, given its longer time coverage (Figure 3.1, panel B).

### Macroeconomic dynamics around episodes
- Pre-period behavior: both consumer price inflation and nominal wage growth increase before period 0 (by construction).
- Post-period average dynamics (Figure 3.2, panels A and B):
  - Inflation and nominal wage growth on average tended to stabilize in the quarters following the wage-price spiral.
  - Real wage growth was broadly unchanged on average after episodes.
  - The unemployment rate tended to edge down slightly following episodes.
  - Patterns are robust when using the longer manufacturing-wage sample.

### Notable extreme episode example
- United States 1973:Q3 (spurred by the first OPEC oil embargo):
  - Price inflation surged for five additional quarters before beginning to fall in 1975.
  - Nominal wage growth did not increase during that episode, leading to a decline in real wage growth.

*Source: wpiea2022221-print-pdf - 3. Historical Experiences*

### 5. A look at Episodes Similar to the COVID-19

### 5. A look at Episodes Similar to the COVID-19 Shock

### Selection and sample
- Episodes selected required that at least three out of four consecutive quarters satisfy: (i) increasing year-on-year inflation, (ii) positive nominal wage growth, (iii) negative real wage growth, and (iv) flat or falling unemployment.  
- Applying those criteria to the dataset with aggregate nominal wages yields 22 episodes (Table 5.1).  
- The 22 episodes are identified using aggregate wages per worker; conclusions are robust to using hourly manufacturing wages and to conditioning on acceleration in energy prices rather than the CPI headline.

### Average post-episode macroeconomic developments
- After these episodes:
  - Inflation tended to decline.  
  - Nominal wage growth tended to increase.  
  - Real wages started increasing again.  
  - The unemployment rate tended to fall.  
- Overall, these episodes were followed by a higher increase in wage growth than in the wider set of episodes, but wage growth eventually stabilized rather than continuing to accelerate indefinitely.

### Heterogeneity and notable case
- There is heterogeneity across episodes. Notable example:
  - United States, 1979:Q2 — inflation rose rapidly for four quarters after the episode before starting to decline; unemployment rose more than in other identified episodes.  
  - Underlying these outcomes was an aggressive monetary tightening that began around the inflation peak (the Volcker disinflation).  
  - Nominal wage growth was relatively flat during this period, producing an early decline in real wages; as inflation fell, the deterioration in real wages decreased.

### Decomposition of nominal wage growth (wage Phillips curve)
- Using the estimated wage Phillips curve (described in Section 4) and pooled coefficients from column (5) of Table 4.1, nominal wage growth is decomposed across the 22 episodes.  
- Findings from the decomposition (Figure 5.2):
  - In the periods up to period zero (the window where wage-price acceleration is identified), wage growth lagged that expected by inflation and unemployment gap movements alone — consistent with falling real wages and similar to post-COVID-19 dynamics.  
  - This is visible in a negative and decreasing "other" component.  
  - After period zero, nominal wage growth starts to catch up and by two-years after the episodes nominal wage growth is broadly consistent with higher inflation and tighter labor markets.  
  - The shrinking "other" component near the end of the episode window illustrates this catch-up.  
  - The positive contribution from the unemployment gap indicates tighter labor markets helped the catch-up; this role is greater than in the wider set of episodes documented in Section 4.  
  - The initial fall in real wages appears to have led to tighter labor markets that co-existed with decelerating inflation.  
- Note: A full decomposition of wage growth around the 2021 wage-price episode following the pandemic cannot yet be done at the same horizon. Analysis up to 2021Q4 suggests rising inflation and labor market tightening have acted to increase nominal wage growth, while other factors have kept nominal wage growth contained.

### Robustness: conditioning on energy prices
- Replacing headline inflation with increasing year-on-year energy inflation (using energy component of consumer prices from the International Energy Agency and US BLS) and applying analogous selection criteria yields 13 episodes (Table B.1).  
- The 13 episodes exhibit similar average post-episode dynamics:
  - Nominal wage growth tended to increase.  
  - Headline and energy inflation tended to decline.  
  - Real wages began to increase again.  
  - Unemployment tended to fall.  
- Extending the sample backward using manufacturing wages and extrapolated energy prices for economies outside the United States yields patterns robust to the extension.

### Implications and conclusion
- Of the 79 episodes identified with accelerating prices and wages (back to the 1960s), only a minority saw further acceleration after eight quarters.  
- Sustained wage-price acceleration is even harder to find when conditioning on episodes similar to today (where real wages have fallen). In those cases:
  - Nominal wages tended to catch up to inflation to partially recover real wage losses, and growth rates stabilized at a higher level than before the initial acceleration.  
  - Wage growth rates eventually became consistent with inflation and labor market tightness; this did not typically lead to persistent acceleration dynamics that can be characterized as a wage-price spiral.  
- Key takeaway: An acceleration of nominal wages should not necessarily be interpreted as evidence that a wage-price spiral is taking hold; history shows nominal wages can accelerate while inflation recedes from high levels.

*Source: IMF Working Paper No. WP/2022/221 — Section 5, "A look at Episodes Similar to the COVID-19 Shock".*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022221-print-pdf.pdf_