## Causes and Implications of Elevated Inflation in Cyprus — Section 1

## Source details

**Canonical URL:** [Causes and Implications of Elevated Inflation in Cyprus — Section 1](https://www.imf.org/-/media/files/publications/selected-issues-papers/2023/english/sipea2023042.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/selected-issues-papers/2023/english/sipea2023042.pdf.md)
- [Structured JSON version](/-/media/files/publications/selected-issues-papers/2023/english/sipea2023042.pdf.json)

---

### A. Introduction
- Paper completed on May 9, 2023; also published as IMF Country Report No 23/193.
- Context:
  - Inflation in Cyprus surged in 2022, driven initially by imported prices and amplified by tourism demand as the economy reopened.
  - The repercussions of the Russian invasion of Ukraine caused a sharp increase in oil prices that further accelerated inflation.

### B. Recent Inflation Trends
- Broad characterization:
  - Inflation was strongly affected by sectoral and external shocks and became more broad-based over time due to second-round effects.
  - Cyprus’s small size and import dependence amplified the impact relative to the rest of the euro area.
- Key quantitative findings:
  - Headline inflation peaked at over 10 percent in July 2022.
  - Energy prices contributed 3 percentage points and transport pricing contributed 4 percentage points to the trough-to-peak increase of 13 percentage points from July 2020.
  - Tourism-related prices (hotels and restaurants) contributed 3 percentage points to the trough-to-peak increase in headline inflation.
- Measures of domestic inflation:
  - Core inflation (excluding food and energy) increased before declining slightly more recently.
  - A Low-Import-Intensity (LIMI) inflation indicator follows core inflation closely; LIMI excluding restaurants and accommodation is more persistent and reflects domestic second-round effects.

### C. Inflation Drivers: Phillips Curve Perspective
- Model specification and data notes:
  - Augmented Phillips Curve includes lagged inflation, one-year ahead inflation expectations, unemployment gap (or output gap), external price pressures, and global energy and food prices interacted with CPI weights.
  - Specification follows IMF 2019 adaptations; one-year ahead expectations used due to data constraints in Cyprus.
- Main attributions and comparisons:
  - Pass-through of external price pressures and energy prices to core and headline inflation is higher in Cyprus than in other euro area members.
  - Phillips Curve attributes a large contribution to inflation from external price pressures and energy prices during the first three quarters of 2022.
  - Inflation expectations increased throughout 2022 and contributed considerably to inflation in the last two quarters of 2022.
  - The unemployment gap turned positive in 2022 but contributed only a small amount to overall inflation in the Phillips Curve estimates (flat response).
  - Sectoral demand/supply factors in tourism are reflected in a large residual: large and negative during 2021 deflation, and large and positive in the first three quarters of 2022; the residual is larger in Cyprus than elsewhere, suggesting sectoral pressures.

### D. Wage Pass-Through
- Methodology:
  - Local projection method with oil-price-driven inflation shocks serving as an instrument (following Baba and Lee (2022)); see Annex II for details.
- Quantitative pass-through estimates:
  - A 1 percent increase in inflation raises private sector wages by 1.4 percent.
  - A 1 percent increase in inflation raises public sector wages by 2.3 percent.
- Comparative and structural context:
  - These pass-through estimates are considerably higher than in most advanced European economies.
  - Higher union density may contribute to higher pass-through in Cyprus (consistent with Baba and Lee (2022)).

### E. Inflation Outlook
- Summary expectation:
  - Elevated inflation is expected to persist and moderate only slowly.
- Drivers of persistence:
  - Phillips Curve estimates point to slow moderation, largely reflecting elevated inflation expectations.
  - Survey of Professional Forecasters showed 1-year inflation expectations surging in Q3 and Q4 of 2022 and an expectation that inflation will still exceed 4 percent at the end of 2023.
  - Expectations for longer horizons (3-, 5-, and 10-year) remain well anchored.
- Labor market and second-round effects:
  - Wage pressures expected to intensify due to historical pass-through and labor market tightness.
  - Vacancy-to-unemployment ratios:
    - thirty vacancies per one hundred unemployed in Q1 2019,
    - more than ninety in Q1 2022,
    - sixty in Q4 (implying persistent tightness).
  - Shortages are most severe in tourism-related occupations; vacancy increases are broad based.
- Profits and producer expectations:
  - Corporate profits have remained above their pre-Covid level and higher than in the rest of the euro area.
  - Producer inflation expectations plateaued at exceptionally high levels in services, retail sales, and construction, indicating potential continued pricing power and limited margin compression.

### F. Policy Recommendations
- Fiscal stance:
  - Fiscal policies should help contain price pressures while protecting vulnerable households.
  - 2023 budget spending plans are characterized as sufficiently tight to help contain inflation pressures from aggregate demand.
- Automatic Cost-of-Living Allowance (CoLA):
  - Current CoLA is set to half of previous year inflation and covers the public sector and unionized employees in the private sector (about ⅓ of the total).
  - Any upwards revision to the automatic CoLA would risk sustaining high inflation and weakening the economy structurally by:
    - Increasing a relatively high wage bill and reducing fiscal space for the public sector;
    - Making fiscal policy more pro-cyclical;
    - Not accounting for productivity developments in the private sector, thereby reducing adjustment capacity to adverse shocks and weakening resilience and competitiveness;
    - Deepening labor market duality since CoLA mostly benefits public sector and collectively bargained private sector employees.
- Social protection:
  - Fiscal policy should protect vulnerable households disproportionately impacted by the cost-of-living crisis (see Box 1).

### Distributional Impact of Inflation (Box 1)
- Differential impact by income:
  - Lower-income households spend a higher share on transport and food—both categories with larger price increases in 2022.
  - Inflation for the fifth quintile was over 2 ppt higher than for the first quintile in December 2022.
- Real-income and saving-consumption differences:
  - Considering different saving-consumption ratios in December 2022:
    - Inflation reduced real incomes close to 11 percent for households in the fifth income quintile.
    - Inflation reduced real incomes by around 6 percent for those in the first quintile.
- Other balance-sheet considerations:
  - Poorer households’ assets tend to be less protected from inflation.
  - The Fisher channel (inflation redistributing from lenders to borrowers) may benefit poorer households.

*Source: Causes and Implications of Elevated Inflation in Cyprus, IMF Selected Issues Paper SIP/2023/042 (Section 1).*

### sipea2023042 - Section 2

### Phillips Curve: model specification and estimation approach
- Model estimated:
  - πi,t = β1 πi,t−1 + β2 πi,t e + β3 u i,t + β4 Energy i,t + β5 Food i,t + β4 Food i,t−1 + β5 External i,t−1 + FE i,t + ε i,t
- Variable definitions:
  - πi,t: q-o-q annualized inflation rate (either headline or core) in country i in quarter t.
  - πi,t e: 1-year ahead inflation expectations.
  - u i,t: unemployment gap (deviation from HP filtered unemployment rate).
  - Energy i,t and Food i,t: energy and food price inflation, in domestic currency and weighted by share in domestic CPI baskets.
  - External i,t−1: lagged external price pressures (changes in producer prices in trading partners and exchange rate fluctuations).
  - FE i,t: country fixed effects.
  - ε i,t: error term.
- Estimation choices and constraints:
  - Sum of coefficients on past and expected inflation rates constrained to be equal to one (as in IMF (2022) and Baba et al. (2023)).
  - Use of 1-year inflation expectations rather than 3-year due to data constraints in Cyprus.
  - Include only one lag of food prices rather than four because food prices play a smaller role in Cyprus than in other countries.
  - For Cyprus an alternative specification replaces the unemployment gap with the output gap (based on staff estimates).
- Sample and data sources:
  - 29 European countries: BEL, BGR, CHE, CYP, CZE, DEU, DNK, ESP, EST, FIN, FRA, GBR, GRC, HRV, HUN, IRL, ITA, LTU, LVA, NLD, NOR, POL, PRT, ROU, RUS, SVK, SVN, SWE, and TUR.
  - Data sources: IMF’s World Economic Outlook database, Eurostat, and Consensus Economics.
  - Estimation period: 2000Q1 to 2022Q4.
- Note on presentation:
  - Table 1 compares Cyprus results to three different panel regressions for both headline and core inflation: whole sample, seven emerging European countries, and twenty-five advanced European countries.
  - Figure 2 plots coefficients on external price pressures against coefficients on energy prices (for headline inflation) across countries, highlighting high pass-through for Cyprus relative to peers.

### Annex I. Table 1 — Selected regression coefficient sequences
- Unemployment gap coefficients (reporting order as in table):
  - -0.184***, -0.366***, -0.145***, -0.081, 0.113*, -0.302***, -0.620***, -0.189***, -0.232, 0.045
  - Standard errors: (0.042), (0.135), (0.040), (0.183), (0.057), (0.058), (0.188), (0.048), (0.179), (0.055)
- Lag of inflation coefficients:
  - 0.532***, 0.525***, 0.540***, 0.286**, 0.067, 0.420***, 0.353***, 0.529***, 0.102, 0.279**
  - Standard errors: (0.076), (0.127), (0.042), (0.112), (0.116), (0.047), (0.082), (0.036), (0.115), (0.110)
- Inflation expectations 1 year ahead coefficients:
  - 0.468***, 0.475***, 0.460***, 0.714***, 0.933***, 0.580***, 0.647***, 0.471***, 0.898***, 0.721***
  - Standard errors: (0.076), (0.127), (0.042), (0.112), (0.116), (0.047), (0.082), (0.036), (0.115), (0.110)
- Lag of external price pressure coefficients:
  - 0.032***, 0.061***, 0.017***, 0.073*, 0.140***, 0.043***, 0.067***, 0.020***, 0.138***, 0.072*
  - Standard errors: (0.007), (0.017), (0.004), (0.040), (0.048), (0.010), (0.024), (0.007), (0.049), (0.041)
- Food price inflation coefficients:
  - 0.066***, 0.105**, 0.030***, -0.051, -0.004, 0.100***, 0.146***, 0.050***, 0.008, -0.054
  - Standard errors: (0.021), (0.043), (0.009), (0.060), (0.055), (0.028), (0.052), (0.013), (0.054), (0.061)
- Lagged food price inflation coefficients:
  - 0.043***, 0.034, 0.035***, -0.056, -0.046, 0.041**, 0.049, 0.017, -0.050, -0.055
  - Standard errors: (0.014), (0.026), (0.010), (0.069), (0.064), (0.020), (0.038), (0.014), (0.065), (0.068)
- Energy price inflation coefficients:
  - 0.018*, 0.032, 0.023***, 0.082, 0.309***, 0.160***, 0.154***, 0.177***, 0.322***, 0.076
  - Standard errors: (0.010), (0.021), (0.007), (0.059), (0.056), (0.013), (0.028), (0.010), (0.057), (0.057)
- Observations (per column): 2,594; 505; 2,089; 747; 42,692; 587; 2,105; 747; 4
- Country FE indicators (per column): Yes, Yes, Yes, No, No, Yes, Yes, Yes, No, No
- Notes: Robust standard errors in parentheses. Significance: *** p<0.01, ** p<0.05, * p<0.1.
- Table distinguishes between Core inflation and Headline inflation regressions.

### External price pressure and energy price dynamics (figures referenced)
- Figures illustrate:
  - Time series of External price pressure and Energy price inflation (Percent, year-on-year) across 2005Q1 to 2022Q3, showing large swings in external and energy price measures.
  - Time series of 1-year inflation expectation and Unemployment gap (Percent, year-on-year) for the same period.
  - Figure 2 plots coefficients on external price pressures versus coefficients on energy prices (headline inflation), highlighting high pass-through of energy and external prices to inflation among European countries.
- Sources for figures: IMF International Financial Statistics and IMF staff calculations.

### Annex II. Estimation of wage response to inflation shocks
- Estimation approach:
  - Quarterly local projection model following Section 5.1. (Transmission via CPI inflation) in Baba and Lee (2022).
  - For each horizon h = 1,...,H estimate:
    - ∆h w t+h−1 = αh + ∑ βi(h) ∆w t−i i−1 + ∑ γi(h) ∆cpi t−i i−1 + ∑ δi(h) ∆y t−i i−1 + ε t,h
  - ∆h w t+h−1: cumulative growth in wages in h quarters from t−1.
  - ∆w t−i and ∆cpi t−i: quarterly wage growth and CPI inflation with lag of i-quarter.
  - ∆y t−i: set of control variables.
  - Oil price shocks with lags of four quarters used as instruments for CPI inflation to obtain wage response to inflation induced by oil shocks.
- Estimation details:
  - Estimation period: 2000Q1 to 2019Q4 (excludes pandemic effects).
  - Models estimated separately for overall wages, public wages, and private sector wages.
  - Regressions include four lags for AR terms, lagged CPI inflation, changes in unemployment rates, and changes in NEER.
  - Shaded areas in impulse response plots represent the 68 and 90 percent confidence bands.
- Findings (as summarized in figures):
  - Impulse responses for baseline (all wages) and public wages to inflation shocks instrumented by oil price shocks are presented (see figures referenced).
- Data source for wage estimation: Estimation follows Baba and Lee (2022) with additional data from Cystat.

*Source: sipea2023042 - Section 2 (selected section content).*

### Section 1

### Causes and Implications of Elevated Inflation in Cyprus — Section 1

### A. Introduction
- Paper completed on May 9, 2023; also published as IMF Country Report No 23/193.
- Context:
  - Inflation in Cyprus surged in 2022, driven initially by imported prices and amplified by tourism demand as the economy reopened.
  - The repercussions of the Russian invasion of Ukraine caused a sharp increase in oil prices that further accelerated inflation.

### B. Recent Inflation Trends
- Broad characterization:
  - Inflation was strongly affected by sectoral and external shocks and became more broad-based over time due to second-round effects.
  - Cyprus’s small size and import dependence amplified the impact relative to the rest of the euro area.
- Key quantitative findings:
  - Headline inflation peaked at over 10 percent in July 2022.
  - Energy prices contributed 3 percentage points and transport pricing contributed 4 percentage points to the trough-to-peak increase of 13 percentage points from July 2020.
  - Tourism-related prices (hotels and restaurants) contributed 3 percentage points to the trough-to-peak increase in headline inflation.
- Measures of domestic inflation:
  - Core inflation (excluding food and energy) increased before declining slightly more recently.
  - A Low-Import-Intensity (LIMI) inflation indicator follows core inflation closely; LIMI excluding restaurants and accommodation is more persistent and reflects domestic second-round effects.

### C. Inflation Drivers: Phillips Curve Perspective
- Model specification and data notes:
  - Augmented Phillips Curve includes lagged inflation, one-year ahead inflation expectations, unemployment gap (or output gap), external price pressures, and global energy and food prices interacted with CPI weights.
  - Specification follows IMF 2019 adaptations; one-year ahead expectations used due to data constraints in Cyprus.
- Main attributions and comparisons:
  - Pass-through of external price pressures and energy prices to core and headline inflation is higher in Cyprus than in other euro area members.
  - Phillips Curve attributes a large contribution to inflation from external price pressures and energy prices during the first three quarters of 2022.
  - Inflation expectations increased throughout 2022 and contributed considerably to inflation in the last two quarters of 2022.
  - The unemployment gap turned positive in 2022 but contributed only a small amount to overall inflation in the Phillips Curve estimates (flat response).
  - Sectoral demand/supply factors in tourism are reflected in a large residual: large and negative during 2021 deflation, and large and positive in the first three quarters of 2022; the residual is larger in Cyprus than elsewhere, suggesting sectoral pressures.

### D. Wage Pass-Through
- Methodology:
  - Local projection method with oil-price-driven inflation shocks serving as an instrument (following Baba and Lee (2022)); see Annex II for details.
- Quantitative pass-through estimates:
  - A 1 percent increase in inflation raises private sector wages by 1.4 percent.
  - A 1 percent increase in inflation raises public sector wages by 2.3 percent.
- Comparative and structural context:
  - These pass-through estimates are considerably higher than in most advanced European economies.
  - Higher union density may contribute to higher pass-through in Cyprus (consistent with Baba and Lee (2022)).

### E. Inflation Outlook
- Summary expectation:
  - Elevated inflation is expected to persist and moderate only slowly.
- Drivers of persistence:
  - Phillips Curve estimates point to slow moderation, largely reflecting elevated inflation expectations.
  - Survey of Professional Forecasters showed 1-year inflation expectations surging in Q3 and Q4 of 2022 and an expectation that inflation will still exceed 4 percent at the end of 2023.
  - Expectations for longer horizons (3-, 5-, and 10-year) remain well anchored.
- Labor market and second-round effects:
  - Wage pressures expected to intensify due to historical pass-through and labor market tightness.
  - Vacancy-to-unemployment ratios:
    - thirty vacancies per one hundred unemployed in Q1 2019,
    - more than ninety in Q1 2022,
    - sixty in Q4 (implying persistent tightness).
  - Shortages are most severe in tourism-related occupations; vacancy increases are broad based.
- Profits and producer expectations:
  - Corporate profits have remained above their pre-Covid level and higher than in the rest of the euro area.
  - Producer inflation expectations plateaued at exceptionally high levels in services, retail sales, and construction, indicating potential continued pricing power and limited margin compression.

### F. Policy Recommendations
- Fiscal stance:
  - Fiscal policies should help contain price pressures while protecting vulnerable households.
  - 2023 budget spending plans are characterized as sufficiently tight to help contain inflation pressures from aggregate demand.
- Automatic Cost-of-Living Allowance (CoLA):
  - Current CoLA is set to half of previous year inflation and covers the public sector and unionized employees in the private sector (about ⅓ of the total).
  - Any upwards revision to the automatic CoLA would risk sustaining high inflation and weakening the economy structurally by:
    - Increasing a relatively high wage bill and reducing fiscal space for the public sector;
    - Making fiscal policy more pro-cyclical;
    - Not accounting for productivity developments in the private sector, thereby reducing adjustment capacity to adverse shocks and weakening resilience and competitiveness;
    - Deepening labor market duality since CoLA mostly benefits public sector and collectively bargained private sector employees.
- Social protection:
  - Fiscal policy should protect vulnerable households disproportionately impacted by the cost-of-living crisis (see Box 1).

### Distributional Impact of Inflation (Box 1)
- Differential impact by income:
  - Lower-income households spend a higher share on transport and food—both categories with larger price increases in 2022.
  - Inflation for the fifth quintile was over 2 ppt higher than for the first quintile in December 2022.
- Real-income and saving-consumption differences:
  - Considering different saving-consumption ratios in December 2022:
    - Inflation reduced real incomes close to 11 percent for households in the fifth income quintile.
    - Inflation reduced real incomes by around 6 percent for those in the first quintile.
- Other balance-sheet considerations:
  - Poorer households’ assets tend to be less protected from inflation.
  - The Fisher channel (inflation redistributing from lenders to borrowers) may benefit poorer households.

*Source: Causes and Implications of Elevated Inflation in Cyprus, IMF Selected Issues Paper SIP/2023/042 (Section 1).*

### Section 2

### sipea2023042 - Section 2

### Phillips Curve: model specification and estimation approach
- Model estimated:
  πi,t = β1 πi,t−1 + β2 πi,t e + β3 u i,t + β4 Energy i,t + β5 Food i,t + β4 Food i,t−1 + β5 External i,t−1 + FE i,t + ε i,t
- Variable definitions:
  - πi,t: q-o-q annualized inflation rate (either headline or core) in country i in quarter t.
  - πi,t e: 1-year ahead inflation expectations.
  - u i,t: unemployment gap (deviation from HP filtered unemployment rate).
  - Energy i,t and Food i,t: energy and food price inflation, in domestic currency and weighted by share in domestic CPI baskets.
  - External i,t−1: lagged external price pressures (changes in producer prices in trading partners and exchange rate fluctuations).
  - FE i,t: country fixed effects.
  - ε i,t: error term.
- Estimation choices and constraints:
  - Sum of coefficients on past and expected inflation rates constrained to be equal to one (as in IMF (2022) and Baba et al. (2023)).
  - Use of 1-year inflation expectations rather than 3-year due to data constraints in Cyprus.
  - Include only one lag of food prices rather than four because food prices play a smaller role in Cyprus than in other countries.
  - For Cyprus an alternative specification replaces the unemployment gap with the output gap (based on staff estimates).
- Sample and data sources:
  - 29 European countries: BEL, BGR, CHE, CYP, CZE, DEU, DNK, ESP, EST, FIN, FRA, GBR, GRC, HRV, HUN, IRL, ITA, LTU, LVA, NLD, NOR, POL, PRT, ROU, RUS, SVK, SVN, SWE, and TUR.
  - Data sources: IMF’s World Economic Outlook database, Eurostat, and Consensus Economics.
  - Estimation period: 2000Q1 to 2022Q4.
- Note on presentation:
  - Table 1 compares Cyprus results to three different panel regressions for both headline and core inflation: whole sample, seven emerging European countries, and twenty-five advanced European countries.
  - Figure 2 plots coefficients on external price pressures against coefficients on energy prices (for headline inflation) across countries, highlighting high pass-through for Cyprus relative to peers.

### Annex I. Table 1 — Selected regression coefficient sequences (reporting order as in table)
- Unemployment gap:
  - Coefficients: -0.184***  -0.366***  -0.145***  -0.081  0.113*  -0.302***  -0.620***  -0.189*** -0.232  0.045
  - Standard errors (in parentheses): (0.042) (0.135) (0.040) (0.183) (0.057) (0.058) (0.188) (0.048) (0.179) (0.055)
- Lag of inflation:
  - Coefficients: 0.532*** 0.525*** 0.540*** 0.286** 0.067 0.420*** 0.353*** 0.529*** 0.102 0.279**
  - Standard errors: (0.076) (0.127) (0.042) (0.112) (0.116) (0.047) (0.082) (0.036) (0.115) (0.110)
- Inflation expectations 1 year ahead:
  - Coefficients: 0.468*** 0.475*** 0.460*** 0.714*** 0.933*** 0.580*** 0.647*** 0.471*** 0.898*** 0.721***
  - Standard errors: (0.076) (0.127) (0.042) (0.112) (0.116) (0.047) (0.082) (0.036) (0.115) (0.110)
- Lag of external price pressure:
  - Coefficients: 0.032*** 0.061*** 0.017*** 0.073* 0.140*** 0.043*** 0.067*** 0.020*** 0.138*** 0.072*
  - Standard errors: (0.007) (0.017) (0.004) (0.040) (0.048) (0.010) (0.024) (0.007) (0.049) (0.041)
- Food price inflation:
  - Coefficients: 0.066*** 0.105** 0.030*** -0.051 -0.004 0.100*** 0.146*** 0.050*** 0.008 -0.054
  - Standard errors: (0.021) (0.043) (0.009) (0.060) (0.055) (0.028) (0.052) (0.013) (0.054) (0.061)
- Lagged food price inflation:
  - Coefficients: 0.043*** 0.034 0.035*** -0.056 -0.046 0.041** 0.049 0.017 -0.050 -0.055
  - Standard errors: (0.014) (0.026) (0.010) (0.069) (0.064) (0.020) (0.038) (0.014) (0.065) (0.068)
- Energy price inflation:
  - Coefficients: 0.018* 0.032 0.023*** 0.082 0.309*** 0.160*** 0.154*** 0.177*** 0.322*** 0.076
  - Standard errors: (0.010) (0.021) (0.007) (0.059) (0.056) (0.013) (0.028) (0.010) (0.057) (0.057)
- Observations (per column): 2,594 505 2,089 747 42,692 587 2,105 747 4
- Country FE indicators (per column): Yes Yes Yes No No Yes Yes Yes No No
- Notes: Robust standard errors in parentheses. Significance: *** p<0.01, ** p<0.05, * p<0.1.
- Table distinguishes between Core inflation and Headline inflation regressions.

### External price pressure and energy price dynamics (figures referenced)
- Annex I. Figure 1 and Figure 2 illustrate:
  - Time series of External price pressure and Energy price inflation (Percent, year-on-year) across 2005Q1 to 2022Q3, showing large swings in external and energy price measures.
  - Time series of 1-year inflation expectation and Unemployment gap (Percent, year-on-year) for the same period.
  - Figure 2 plots coefficients on external price pressures versus coefficients on energy prices (headline inflation), highlighting high pass-through of energy and external prices to inflation among European countries.
- Sources for figures: IMF International Financial Statistics and IMF staff calculations.

### Annex II. Estimation of wage response to inflation shocks
- Estimation approach:
  - Quarterly local projection model following Section 5.1. (Transmission via CPI inflation) in Baba and Lee (2022).
  - For each horizon h = 1,...,H estimate:
    ∆h w t+h−1 = αh + ∑ βi(h) ∆w t−i i−1 + ∑ γi(h) ∆cpi t−i i−1 + ∑ δi(h) ∆y t−i i−1 + ε t,h
  - ∆h w t+h−1: cumulative growth in wages in h quarters from t−1.
  - ∆w t−i and ∆cpi t−i: quarterly wage growth and CPI inflation with lag of i-quarter.
  - ∆y t−i: set of control variables.
  - Oil price shocks with lags of four quarters used as instruments for CPI inflation to obtain wage response to inflation induced by oil shocks.
- Estimation details:
  - Estimation period: 2000Q1 to 2019Q4 (excludes pandemic effects).
  - Models estimated separately for overall wages, public wages, and private sector wages.
  - Regressions include four lags for AR terms, lagged CPI inflation, changes in unemployment rates, and changes in NEER.
  - Shaded areas in impulse response plots represent the 68 and 90 percent confidence bands.
- Findings (as summarized in figures):
  - Annex II. Figure 1 shows impulse responses for baseline (all wages) and public wages to inflation shocks instrumented by oil price shocks.
- Data source for wage estimation: Estimation follows Baba and Lee (2022) with additional data from Cystat.

*Source: sipea2023042 - Section 2 (selected section content).*

---


_Source: https://www.imf.org/-/media/files/publications/selected-issues-papers/2023/english/sipea2023042.pdf_