## 1. Dutch Inflation Developments

## Source details

**Canonical URL:** [1. Dutch Inflation Developments](https://www.imf.org/-/media/files/publications/selected-issues-papers/2023/english/sipea2023021.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/selected-issues-papers/2023/english/sipea2023021.pdf.md)
- [Structured JSON version](/-/media/files/publications/selected-issues-papers/2023/english/sipea2023021.pdf.json)

---

### A. Different Measures of Inflation: A Methodological Review
- Headline HICP peaked at 17.1 percent year-on-year in September 2022 and slowed to 11.0 percent year-on-year in December 2022.
- As of December 2022:
  - HICP increased by 11.0 percent year-on-year.
  - CPI grew by 9.6 percent year-on-year.
- Key methodological differences driving the HICP–CPI gap:
  - CPI includes imputed rents for owner-occupied housing (regulated rents grew less), weighing down CPI relative to HICP.
  - The energy weight in the HICP is higher than in the CPI, making HICP more sensitive to energy price changes.
- Factors explaining high energy price growth in Dutch statistics:
  - Greater share of gas in the Dutch consumer price basket relative to other EA countries; most electricity in the Netherlands is produced by fossil fuel-fired powerplants, mainly natural gas.
  - Liberalized electricity market with about 53 suppliers increases transmission of wholesale into retail prices and volatility; no social tariffs.
  - Dutch TTF spot gas prices were somewhat higher than in some other markets during 2022 energy market turbulence.
  - Current consumer energy price calculation relies on variable rate contract pricing, excluding nearly half of households on fixed contracts and thus overestimating energy bills when prices rise sharply.
- Household contract structure (as of end-August 2022):
  - 56 percent of Dutch households had variable rate energy contracts.
  - Among remaining 44 percent on fixed contracts: 11 percent fixed for 1 year, 1 percent for 2 years, 25 percent for 3 years, and 8 percent for 5 years.
- CBS methodological work and preliminary alternative estimates:
  - CBS launched a study in early 2022 using utility company data covering about ¾ of households.
  - Preliminary alternative calculation results (latest data point August 2022):
    - Electricity price growth in range of 24–93 percent year-on-year instead of reported 150 percent.
    - Gas price growth in range of 34–97 percent year-on-year instead of published 170 percent.
    - Implied CPI inflation about 2.4 to 4.5 percentage points lower than published (7.5–9.6 percent instead of published 12 percent).
  - Introduction of a new methodology likely to take effect in mid-2023 for both CPI and HICP.
- Comparison with National Accounts measure:
  - As of 2022Q3: HICP inflation at 14.1 percent year-on-year, average CPI at 12.3 percent, private consumption deflator at 8.9 percent.
  - Private consumption deflator grows more modestly than HICP/CPI because it accounts for changes in households’ consumption and uses quarterly-updated weights.
- Weight-update and consumption-pattern effects:
  - HICP and CPI weights are updated annually, failing to capture within-year consumption shifts; private consumption deflator uses quarterly-updated weights.
  - Energy consumption in the Netherlands adjusted by more than 20 percent in 2022.
- Historical interpretation:
  - Widening gap between HICP and private consumption deflator has coincided with previous economic slowdowns (early 2000s, global financial crisis, European sovereign debt crisis).

### B. The Role of Domestic and Global Factors: An Empirical Assessment
- Empirical approach:
  - Phillips curve estimated for a panel of 17 EA economies and separately for countries with comprehensive data over 2000Q1–2022Q2.
  - Rolling window panels estimated to assess changes over time.
  - Estimates used to compute implied contributions to inflation dynamics; specification details in Annex I.
- Main empirical findings:
  - Expectations and lagged inflation:
    - Lagged inflation and 3-year-ahead inflation expectations are statistically significant and have the largest estimated coefficients in both panel and Netherlands-specific regressions, for headline and core inflation.
  - Domestic cyclical conditions:
    - Three slack proxies used in separate regressions: unemployment gap, output gap, and a combined slack measure (first principal component of seven variables).
    - At EA level, unemployment gap, output gap, and combined slack are statistically significant for headline and core inflation.
    - In Netherlands-specific regressions:
      - Unemployment gap is statistically significant for core inflation; output gap and principal components measure are not significant (with one exception).
      - Magnitude: a one percentage point increase in the unemployment gap is associated with a 0.35 percentage point decline in core inflation in the Netherlands-specific baseline; about 0.2 in the panel setting.
  - External factors and commodity prices:
    - A general external price pressures measure is not statistically significant in most specifications.
    - Commodity prices—energy and food—are positive and statistically significant determinants of inflation for both the EA and the Netherlands.
    - Rolling window panel analysis indicates pass-through from energy to core inflation has increased substantially in the EA, with a higher coefficient on lagged inflation, suggesting EA inflation has become more backward-looking and persistent.
  - Contributions to Dutch inflation:
    - Dynamic simulations of the Netherlands-specific Phillips curve indicate global commodity prices, specifically food and energy, have been key drivers of Dutch headline inflation since 2021.
    - Energy prices, and to a lesser extent food prices, have also been key to core inflation acceleration since mid-2021.
- Empirical model diagnostics and key coefficients:
  - Phillips curve analysis covers 17 Euro Area economies and quarterly data 2000Q1–2022Q2.
  - The first principal component used for a domestic slack measure explains about 30 percent of total variance.
  - Explanatory variables account for about 66 (65) percent of the variation of headline (core) OLS regressions with country fixed effects; median regressions yield pseudo R-squared of 0.38 (0.32) for headline (core).
  - Estimated coefficients on three-year-ahead inflation expectations range between 0.4 and 0.6 for headline, and 0.2 and 0.4 for core inflation regressions (statistically significant in most specifications).
  - A one percentage point increase in the output gap is associated with an increase in the headline (core) inflation rate by about 1.1 (0.9) percentage points under the baseline specification.
  - One percentage point increase in the unemployment gap is associated with a 0.2 percentage point decline in headline/core inflation under the baseline specification.
  - Commodity prices matter: energy price coefficients are positive and statistically significant; food price effects exhibit longer lags (food included contemporaneously and with four lags in the baseline).

### C. Wholesale Energy Price Pass-Through
- Energy pass-through and cross-country heterogeneity:
  - Increasing contribution of gas and electricity prices to consumer inflation has produced greater cross-country heterogeneity in the euro area.
  - Pass-through from wholesale to consumer prices depends on:
    - the energy mix in electricity generation,
    - the price-setting mechanism,
    - price composition (e.g., whether tariffs are variable or pre-fixed).
  - ECB (2022) finding: if tariffs are set at a variable price, consumer prices adjust faster; pre-fixed tariffs imply slower adjustment.
- Why the Netherlands experienced higher inflation than EA peers:
  - Three main factors:
    1. CPI/HICP methodology consideration of variable-rate contracts introduces bias in inflation estimations.
    2. Demand side: the share of gas and electricity items in the final consumer basket is among the highest in the euro area (only surpassed by Slovakia).
    3. Supply side: high use of gas in electricity generation in the Netherlands makes electricity prices more sensitive to gas price changes.
  - EU wholesale electricity market uses marginal pricing: a higher share of gas in generation lowers the threshold for gas to determine final electricity prices, increasing elasticity of Dutch electricity prices to gas markets.
- Composition of household bills and taxation:
  - Average household bill components: commodity price (including transportation and distribution), VAT, and other taxes.
  - In the Netherlands, "other taxes" include the ODE (Sustainable Energy Surcharge, Opslag Duurzame Energie) and the energy tax. Electricity bills also reflect an energy tax refund or energy tax credit per electricity connection.
  - Key tax levels and changes:
    - Energy tax on gas: 0.26 EUR/m3 in the first half of 2022, up from 0.16 EUR/m3 in 2018.
    - Energy tax on electricity: reduced to 0.037 EUR/kWh in 2022 from 0.10 EUR/kWh in 2018. (Note: energy tax on electricity was markedly lowered to 0.037 EUR/kWh in January 2022 from previous 0.094 EUR/kWh.)
    - Energy tax credit per electricity connection: 681.6 EUR in 2022, up from 308 EUR in 2018.
- Macroeconomic transmission, persistence, and distributional impacts:
  - Empirical pass-through timing:
    - Energy prices take about "6–8 months" to pass-through to core items.
    - Lead-lag structure and persistence estimates suggest effects of high energy prices could last up to "18 months".
  - Core inflation developments:
    - Netherlands core inflation: 8.4 percent in December 2022 (2.3 percent in December 2021).
    - Euro area core inflation: 6.6 percent in December 2022 (2.7 percent in December 2021).
    - Contributions to Dutch core inflation between December 2021 and December 2022:
      - Processed food: 14.1 percent in December (from 2.4 percent in December 2021).
      - Non-energy industrial goods (NEIG): 8.8 percent in December (from 3.4 percent in December 2021).
      - Services: 5.8 percent in December (from 1.3 percent in December 2021).
  - Cost-of-living risks:
    - Food price acceleration and persistently high energy prices worsen cost-of-living pressures, disproportionately affecting low-income households due to lower substitution capacity, higher marginal propensity to consume, and lower liquid assets (ECB, 2022).
- Monetary policy, base effects, and persistence risks:
  - Quantitative carry-over effects from 2022 to 2023:
    - The 2022 carry-over effect will account for "3.7 percentage points" in the average core inflation rate in 2023.
    - On a monthly basis, this effect will keep year-on-year core inflation above "2 percent" until August 2023, absent a strong adjustment of core prices.
  - Policy implication: monetary policy tightening should at some point diminish inflationary pressures on core inflation, reducing persistence and pass-through.
- Wages, profit margins, and the risk of a wage–price spiral:
  - Wage dynamics:
    - Contractual wage growth: "5.0 percent" in December 2022, up from "1.0 percent" one year earlier.
    - Average wage (reflecting multi-year agreements): "3.7 percent" year-on-year.
    - Minimum wage: increase of "10.3 percent" in early 2023.
  - Labor income and wage share:
    - Since mid-2021, the Dutch wage share has declined, largely driven by loss of labor income purchasing capacity.
    - No inflation-linked clauses in the Dutch regulatory framework for wages, pensions, and social benefits.
  - Profit margins and pass-through capacity:
    - The translation of higher wages into higher inflation depends on corporate profit margins and firms’ capacity/willingness to absorb costs versus passing them through.
    - Rising input costs (energy) and wage growth amid a deteriorating macroeconomic outlook could compress profits and pose risks to corporate liquidity and solvency if pass-through occurs.

### GDP Deflator, Income Distribution, and Productivity
- Demand-side and price developments to 2022Q3:
  - Purchasing prices accumulated an increase of "6.8 percent" until 2022Q3 since energy prices accelerated in 2021Q2.
  - Final consumer prices increased by "9 percent" in that period, reflecting pass-through from wholesale to retail energy prices and post-pandemic recovery effects.
  - Private consumption slowdown in 2022 moderated the impact on the GDP deflator; investment recovery from 2022Q2 and capital investment price acceleration contributed to a rising GDP deflator.
  - As a net energy importer, rising energy prices worsened terms-of-trade, eroding income available to remunerate domestic factors.
- Income approach considerations:
  - GDP deflator can be interpreted as a combination of remuneration to capital (profits) and labor (wages), plus the public sector (indirect taxes net of subsidies).
  - Profit measures:
    - Profit share: ratio of gross operating surplus (or gross value added) to GDP—sensitive to interactions between gross operating surplus and labor costs.
    - Profit mark-up proxy: difference between gross value added and remuneration of employees—gross value added excludes intermediate consumption (relevant amid high energy costs).
  - Methodological distinctions:
    - Gross value added excludes intermediate consumption (energy), making profit-mark-up particularly relevant in the current energy-cost-driven episode.
    - Profit share captures interactions with the public sector and sectoral income flows; pandemic income support affected profit-share performance.
  - Outlook:
    - With pandemic income support phased out, sustained high energy prices, and economic slowdown, the profit mark-up may deteriorate further via declining value added, accelerating input costs, and rising wages.
- Labor productivity, wages, and unit labor costs (ULC):
  - ULC decompose into wages and labor productivity; wage performance raises ULC while labor productivity growth reduces ULC.
  - Latest information related to 2022Q3 confirms increasing weight of wages on labor costs and lack of productivity growth in a context of economic slowdown.
  - Labor productivity is a key determinant of the impact of increasing wages on prices; productivity improvements can offset wage-driven ULC increases.
- Historical interaction of profits, wages, and domestic prices:
  - Profits in the Netherlands historically have been more stable than in the EA, providing a buffer against short-term cyclical fluctuations in ULC.
  - Past episodes show corporates absorbed shocks (2008) or recovered profits without adding inflationary pressure (2011–12), aided by productivity and wage dynamics.
- Role of government support since 2020:
  - Pandemic public support (2020–21) channeled notably through subsidies to corporates, reducing immediate inflationary impact of rising labor costs and preserving profits.
  - End of remaining pandemic-income support played a key role in 2022, driven by the increase of indirect taxes and the end of subsidies.
- Sectoral adaptation and productivity importance:
  - Different sectors face heterogeneous exposure to high energy prices; ability to adapt and improve productivity is key to preserving employment and firm viability.

### Rolling Estimation, Stability of Phillips Curve, and Implications
- Rolling estimation approach and findings:
  - Panel-data Phillips curve estimated on a rolling basis over 16-quarter windows for 17 economies.
  - Recent rolling-window changes:
    - Estimated coefficient on lagged headline (core) inflation has significantly increased in the last 3 rolling window periods, indicating inflation has become more backward-looking.
    - Pass-through of energy prices to core inflation has increased substantially.
    - The coefficient on the unemployment gap suggests the Phillips curve has become steeper over the last few years.
  - Caveat: few data points underlie these changes, so long-lasting structural shifts are premature to conclude.
- Implications:
  - Greater pass-through of commodity prices into core inflation increases susceptibility to negative supply shocks.
  - A larger effect of past inflation on current inflation suggests potential difficulties in reducing inflation going forward.
- Contribution calculation:
  - Contributions computed for each regression-quarter using dynamic simulations accounting for persistence.
  - Contribution formula used: CCPD_x = beta_D_x * eP_D_x + beta1 * CCPD_{t-1,x}, where coefficients are country-specific and beta1 is coefficient on lagged inflation.

### Conclusions (selected findings)
- 2022 headline inflation in the Netherlands reached unprecedented levels since the record began, largely driven by high energy price growth and increasingly broad-based price pressures.
- Factors contributing to higher energy price growth in the Netherlands include: a greater share of gas in the Dutch energy mix, liberalized energy markets, and the current method of energy inflation calculation.
- Phillips curve analysis indicates:
  - past and future inflation expectations are the main determinants of consumer prices;
  - domestic and external factors are statistically significant but less dominant.
- Empirical results underscore the importance of energy and food commodity prices, lagged inflation, and inflation expectations in explaining recent Dutch inflation dynamics.

*Source: sipea2023021 (IMF staff chapter text as provided).*

### 1. Dutch Inflation Developments  ______________________________________________________ 3

### 1. Dutch Inflation Developments

### A. Different Measures of Inflation: A Methodological Review
- Headline HICP peaked at 17.1 percent year-on-year in September 2022 and slowed to 11.0 percent year-on-year in December 2022.
- As of December 2022:
  - HICP increased by 11.0 percent year-on-year.
  - CPI grew by 9.6 percent year-on-year.
- Key methodological differences driving the HICP–CPI gap:
  - CPI includes imputed rents for owner-occupied housing (regulated rents grew less), weighing down CPI relative to HICP.
  - The energy weight in the HICP is higher than in the CPI, making HICP more sensitive to energy price changes.
- Factors explaining high energy price growth in Dutch statistics:
  - Greater share of gas in the Dutch consumer price basket relative to other EA countries; most electricity in the Netherlands is produced by fossil fuel-fired powerplants, mainly natural gas.
  - Liberalized electricity market with about 53 suppliers increases transmission of wholesale into retail prices and volatility; no social tariffs.
  - Dutch TTF spot gas prices were somewhat higher than in some other markets during 2022 energy market turbulence.
  - Current consumer energy price calculation relies on variable rate contract pricing, excluding nearly half of households on fixed contracts and thus overestimating energy bills when prices rise sharply.
- Household contract structure (as of end-August 2022, footnote detail included in source):
  - 56 percent of Dutch households had variable rate energy contracts.
  - Among remaining 44 percent on fixed contracts: 11 percent fixed for 1 year, 1 percent for 2 years, 25 percent for 3 years, and 8 percent for 5 years.
- CBS methodological work and preliminary alternative estimates:
  - CBS launched a study in early 2022 using utility company data covering about ¾ of households.
  - Preliminary alternative calculation results (latest data point August 2022):
    - Electricity price growth in range of 24–93 percent year-on-year instead of reported 150 percent.
    - Gas price growth in range of 34–97 percent year-on-year instead of published 170 percent.
    - Implied CPI inflation about 2.4 to 4.5 percentage points lower than published (7.5–9.6 percent instead of published 12 percent).
  - Introduction of a new methodology likely to take effect in mid-2023 for both CPI and HICP.
- Comparison with National Accounts measure:
  - As of 2022Q3: HICP inflation at 14.1 percent year-on-year, average CPI at 12.3 percent, private consumption deflator at 8.9 percent.
  - Private consumption deflator grows more modestly than HICP/CPI because it accounts for changes in households’ consumption and uses quarterly-updated weights.
- Weight-update and consumption-pattern effects:
  - HICP and CPI weights are updated annually, failing to capture within-year consumption shifts; private consumption deflator uses quarterly-updated weights.
  - Energy consumption in the Netherlands adjusted by more than 20 percent in 2022.
- Historical interpretation:
  - Widening gap between HICP and private consumption deflator has coincided with previous economic slowdowns (early 2000s, global financial crisis, European sovereign debt crisis).

### B. The Role of Domestic and Global Factors: An Empirical Assessment
- Empirical approach:
  - Phillips curve estimated for a panel of 17 EA economies and separately for countries with comprehensive data over 2000Q1–2022Q2.
  - Rolling window panels estimated to assess changes over time.
  - Estimates used to compute implied contributions to inflation dynamics; specification details in Annex I.
- Main empirical findings:
  - Past and future inflation expectations play a key role in determining consumer price growth:
    - Lagged inflation and 3-year-ahead inflation expectations are statistically significant and have the largest estimated coefficients in both panel and Netherlands-specific regressions, for headline and core inflation.
  - Domestic cyclical conditions:
    - Three slack proxies used in separate regressions: unemployment gap, output gap, and a combined slack measure (first principal component of seven variables).
    - At EA level, unemployment gap, output gap, and combined slack are statistically significant for headline and core inflation.
    - In Netherlands-specific regressions:
      - Unemployment gap is statistically significant for core inflation; output gap and principal components measure are not significant (with one exception noted).
      - Magnitude: a one percentage point increase in the unemployment gap is associated with a 0.35 percentage point decline in core inflation in the Netherlands-specific baseline; about 0.2 in the panel setting.
  - External factors and commodity prices:
    - A general external price pressures measure is not statistically significant in most specifications.
    - Commodity prices—energy and food—are positive and statistically significant determinants of inflation for both the EA and the Netherlands.
    - Rolling window panel analysis indicates pass-through from energy to core inflation has increased substantially in the EA, with a higher coefficient on lagged inflation, suggesting EA inflation has become more backward-looking and persistent.
  - Contributions to Dutch inflation:
    - Dynamic simulations of the Netherlands-specific Phillips curve indicate global commodity prices, specifically food and energy, have been key drivers of Dutch headline inflation since 2021.
    - Energy prices, and to a lesser extent food prices, have also been key to core inflation acceleration since mid-2021.

### C. Wholesale Energy Price Pass-Through
- (Section heading present in source; detailed content on wholesale-to-retail pass-through begins after the provided excerpt.)

*Source: INTERNATIONAL MONETARY FUND.*

### 14.      The increasing role of gas and electricity price contribution to consumer’s inflation has

### 14.      The increasing role of gas and electricity price contribution to consumer’s inflation has

### Energy pass-through and cross-country heterogeneity
- Evidence shows increasing contribution of gas and electricity prices to consumer inflation has produced greater cross-country heterogeneity in the euro area.
- Estimates are based on a panel-data setting with country fixed effects and a rolling window of 16 quarters ending in the quarter indicated on the x-axis (rolling regression coefficients for headline and core regressions; 10th–90th percentile confidence intervals shown).
- Pass-through from wholesale to consumer prices depends on:
  - the energy mix in electricity generation,
  - the price-setting mechanism,
  - price composition (e.g., whether tariffs are variable or pre-fixed).
- ECB (2022) finding: if tariffs are set at a variable price, consumer prices adjust faster; pre-fixed tariffs imply slower adjustment.

### Why the Netherlands experienced higher inflation than EA peers
- Three main factors explaining higher inflation in the Netherlands:
  1. CPI/HICP methodology consideration of variable-rate contracts introduces bias in inflation estimations.
  2. Demand side: the share of gas and electricity items in the final consumer basket is among the highest in the euro area (only surpassed by Slovakia).
  3. Supply side: high use of gas in electricity generation in the Netherlands makes electricity prices more sensitive to gas price changes.
- EU wholesale electricity market uses marginal pricing: the last producer’s (marginal) price sets the market price for all producers. Because gas often bids at higher prices than renewables, a higher share of gas in generation lowers the threshold for gas to determine final electricity prices, increasing the elasticity of Dutch electricity prices to gas markets.

### Composition of household bills and taxation
- Average household bill components: commodity price (including transportation and distribution), VAT, and other taxes.
- In the Netherlands, "other taxes" include the ODE (Sustainable Energy Surcharge, Opslag Duurzame Energie) and the energy tax. Electricity bills also reflect an energy tax refund or energy tax credit per electricity connection.
- Key tax levels and changes:
  - Energy tax on gas: 0.26 EUR/m3 in the first half of 2022, up from 0.16 EUR/m3 in 2018.
  - Energy tax on electricity: reduced to 0.037 EUR/kWh in 2022 from 0.10 EUR/kWh in 2018. (A note indicates the energy tax on electricity was markedly lowered to 0.037 EUR/kWh in January 2022 from previous 0.094 EUR/kWh.)
  - Energy tax credit per electricity connection: 681.6 EUR in 2022, up from 308 EUR in 2018.

### Macroeconomic transmission, persistence, and distributional impacts
- Empirical pass-through timing:
  - Energy prices take about "6–8 months" to pass-through to core items.
  - Lead-lag structure and persistence estimates suggest effects of high energy prices could last up to "18 months".
- Core inflation developments (Netherlands vs. euro area):
  - Netherlands core inflation: 8.4 percent in December 2022 (2.3 percent in December 2021).
  - Euro area core inflation: 6.6 percent in December 2022 (2.7 percent in December 2021).
  - Contributions to Dutch core inflation between December 2021 and December 2022:
    - Processed food: 14.1 percent in December (from 2.4 percent in December 2021).
    - Non-energy industrial goods (NEIG): 8.8 percent in December (from 3.4 percent in December 2021).
    - Services: 5.8 percent in December (from 1.3 percent in December 2021).
- Cost-of-living risks:
  - Food price acceleration and persistently high energy prices worsen cost-of-living pressures, disproportionately affecting low-income households due to lower substitution capacity, higher marginal propensity to consume, and lower liquid assets (ECB, 2022).

### Monetary policy, base effects, and persistence risks
- Quantitative carry-over effects from 2022 to 2023:
  - The 2022 carry-over effect will account for "3.7 percentage points" in the average core inflation rate in 2023.
  - On a monthly basis, this effect will keep year-on-year core inflation above "2 percent" until August 2023, absent a strong adjustment of core prices.
- Policy implication: monetary policy tightening should at some point diminish inflationary pressures on core inflation, reducing persistence and pass-through.

### Wages, profit margins, and the risk of a wage–price spiral
- Wages–prices interaction is time-varying and depends on the economy’s stage; ECB research finds wage pass-through to prices is lower in low-inflation periods than in high-inflation periods.
- Current Dutch wage dynamics:
  - Contractual wage growth: "5.0 percent" in December 2022, up from "1.0 percent" one year earlier.
  - Average wage (reflecting multi-year agreements): "3.7 percent" year-on-year.
  - Minimum wage: increase of "10.3 percent" in early 2023.
- Labor income and wage share:
  - Since mid-2021, the Dutch wage share has declined, largely driven by loss of labor income purchasing capacity.
  - No inflation-linked clauses in the Dutch regulatory framework for wages, pensions, and social benefits, providing flexibility in negotiations among social partners.
- Profit margins and pass-through capacity:
  - The extent to which higher wages translate into higher inflation depends on corporate profit margins and the capacity/willingness of firms to absorb costs versus passing them through to prices.
  - Rising input costs (energy) and wage growth amid a deteriorating macroeconomic outlook could compress profits and pose risks to corporate liquidity and solvency if pass-through occurs.

### GDP deflator and income distribution between labor and capital
- Demand-side decomposition (2022 developments):
  - Purchasing prices accumulated an increase of "6.8 percent" until 2022Q3 since energy prices accelerated in 2021Q2.
  - Final consumer prices increased by "9 percent" in that period, reflecting pass-through from wholesale to retail energy prices and post-pandemic recovery effects.
  - Private consumption slowdown in 2022 moderated the impact on the GDP deflator; investment recovery from 2022Q2 and capital investment price acceleration contributed to a rising GDP deflator.
  - As a net energy importer, rising energy prices worsened terms-of-trade, eroding income available to remunerate domestic factors.
- Income approach considerations:
  - GDP deflator can be interpreted as a combination of remuneration to capital (profits) and labor (wages), plus the public sector (indirect taxes net of subsidies).
  - Profit measures:
    - Profit share: ratio of gross operating surplus (or gross value added) to GDP—sensitive to interactions between gross operating surplus and labor costs.
    - Profit mark-up proxy: difference between gross value added and remuneration of employees—gross value added excludes intermediate consumption (relevant amid high energy costs).
  - Methodological differences between profit-share and profit-mark-up indicators yield complementary analytical insights:
    - Gross value added excludes intermediate consumption (energy), making profit-mark-up particularly relevant in the current energy-cost-driven episode.
    - Profit share captures interactions with the public sector and sectoral income flows; pandemic income support affected profit-share performance.
  - Going forward, with pandemic income support phased out, sustained high energy prices, and economic slowdown, the profit mark-up may deteriorate further via declining value added, accelerating input costs, and rising wages.

*Source: IMF staff estimates and calculations, chapter content provided in the source PDF.*

### 31.      Labor productivity is also a key determinant of the impact of increasing wages into

### sipea2023021 - 31.      Labor productivity is also a key determinant of the impact of increasing wages into

### Labor productivity, wages, and unit labor costs (ULC)
- Unit labor costs (ULC) decompose into wages and labor productivity; wage performance raises ULC while labor productivity growth reduces ULC, potentially offsetting wage effects on corporate costs.
- A relevant pick-up in wages could be offset by a similar increase in productivity, reducing the final impact of ULC on prices.
- Latest information related to 2022Q3 confirms:
  - increasing weight of wages on labor costs; and
  - lack of productivity growth in a context of economic slowdown.

### Historical interaction of profits, wages, and domestic prices
- Profits in the Netherlands historically have been more stable than in the Euro Area (EA), providing a buffer against short-term cyclical fluctuations in ULC.
- Growth of domestic prices in the Netherlands has been more intense and volatile than in the EA over the last 20 years.
- Past episodes:
  - 2008 global financial crisis: sudden demand shock was absorbed by corporate profits, avoiding pass-through of rising wages and productivity loss to final prices.
  - 2011–12 sovereign debt crisis aftermath: Dutch corporate profits recovered faster and stronger than in the EA without adding inflationary pressures, supported by more moderate wage growth and stronger productivity gains than the EA.
  - 2017 until 2020: stronger economic performance translated into more intense recovery of wages and corporate profits, increasing domestic prices during strong growth.

### Role of government support since 2020
- Unprecedented public support during the pandemic in 2020–21 was channeled notably through subsidies to corporates, shown as a strong negative contribution of public sector income to overall domestic costs.
- This support:
  - allowed corporates to partially pass-through higher ULC to final consumers while preserving profits; and
  - reduced the immediate inflationary impact of rising labor costs.
- The effect was more visible in the Netherlands than in the EA, driven by larger public support and stronger wage growth.
- The end of remaining pandemic-income support played a key role in 2022, driven by the increase of indirect taxes and the end of subsidies.

### Current dynamics, inflation persistence, and wage-price spiral risk
- In mid-2021 inflation accelerated with gradual re-opening of European economies; fiscal support also affected income distribution between wages and profits and the pass-through of wages into labor costs.
- Key empirical finding on energy pass-through:
  - It takes about 6–8 months for energy prices to pass-through to core items, increasing the likelihood of inflation persistence and adding downside risks for domestic demand.
- Despite accelerating contractual wages in 2022, risks for a wage-price spiral remain limited at an aggregate level because:
  - wage growth remains well below inflation despite a tight labor market;
  - wage share accumulates five-quarter in a row on a decreasing trend;
  - no inflation-linked clauses in the Dutch regulatory framework; and
  - flexibility to negotiate among social partners has historically allowed healthy interaction between wages and corporate profits.
- However, heterogeneity at the micro-level:
  - some firms/sectors may be compromised by high energy prices and slower demand; capacity to absorb higher input costs varies across firms and sectors.
  - loss of purchasing power for households will be heterogeneous among income quintiles, with unclear aggregate demand effects.

### Sectoral adaptation and productivity importance
- Different sectors are affected differently by high energy prices; ability to adapt to the new energy environment and improve productivity is key to:
  - preserving employment through competitive wages and firms’ own profits; and
  - maintaining viability for vulnerable firms and sectors.

### Conclusions (selected findings)
- 2022 headline inflation in the Netherlands reached unprecedented levels since the record began, largely driven by high energy price growth and increasingly broad-based price pressures.
- Factors contributing to higher energy price growth in the Netherlands include: a greater share of gas in the Dutch energy mix, liberalized energy markets, and the current method of energy inflation calculation.
- Phillips curve analysis indicates:
  - past and future inflation expectations are the main determinants of consumer prices;
  - domestic and external factors are statistically significant but less dominant.
- Empirical model results and robustness:
  - Phillips curve analysis covers 17 Euro Area economies and quarterly data 2000Q1–2022Q2.
  - The first principal component used for a domestic slack measure explains about 30 percent of total variance.
  - Explanatory variables account for about 66 (65) percent of the variation of headline (core) OLS regressions with country fixed effects; median regressions yield pseudo R-squared of 0.38 (0.32) for headline (core).
  - Estimated coefficients on three-year-ahead inflation expectations range between 0.4 and 0.6 for headline, and 0.2 and 0.4 for core inflation regressions (statistically significant in most specifications).
  - A one percentage point increase in the output gap is associated with an increase in the headline (core) inflation rate by about 1.1 (0.9) percentage points under the baseline specification.
  - One percentage point increase in the unemployment gap is associated with a 0.2 percentage point decline in headline/core inflation under the baseline specification.
  - Commodity prices matter: energy price coefficients are positive and statistically significant; food price effects exhibit longer lags (food included contemporaneously and with four lags in the baseline).

### Annex — Phillips curve specification and estimation overview
- Baseline Phillips curve: standard New Keynesian Phillips curve augmented with global price variables, constrained so sum of coefficients on past inflation and inflation expectation equals one in benchmark specification; alternative specifications relax this constraint (OLS, median).
- Key variables and data sources (sample and frequency):
  - Country coverage: 17 Euro Area economies (Luxembourg and Malta excluded due to limited inflation expectations data).
  - Quarterly frequency, 2000Q1–2022Q2.
  - Core/headline CPI, energy and food weights, three-year-ahead inflation expectations, world food and energy price indices, producer price index, GDP deflator, real and nominal GDP, unemployment and participation rates, shares of part-time, temporary, self-employed, hours worked per employee.
- Estimation results summary:
  - Lagged inflation significant → backward-looking inflation dynamics.
  - Inflation expectations significant → forward-looking component in price-setting.
  - Domestic cyclical conditions (output gap, unemployment gap, economic slack) statistically significant but modest in magnitude.
  - External price pressure measure largely not significant; commodity (energy and food) prices are important determinants.

*Source: sipea2023021 (IMF staff chapter text as provided).*

### 10. Country-specific coefficients are used to calculate contributions to inflation dynamics from

### 10. Country-specific coefficients are used to calculate contributions to inflation dynamics from

### Calculation of contributions to inflation dynamics
- Contributions are computed for each regression in every quarter using dynamic simulations to account for the persistence of inflation.
- The contribution of each explanatory variable 푒 to inflation dynamics in country 푖 at time 푒 is denoted 퐶퐶푃푃,퐷퐷푥푥 and calculated as:
  - 퐶퐶푃푃,퐷퐷푥푥 = 훽훽퐷퐷푥푥�푒푃푃,퐷퐷 + 훽훽1� 퐶퐶푃푃,퐷퐷−1푥푥
  - 훽훽퐷퐷푥푥� is the corresponding coefficient on variable 푒 obtained from country-specific Phillips curve regressions.
  - 훽훽1� is the coefficient on lagged inflation.

### Stability of the Phillips Curve estimates
- Rolling estimation approach:
  - The panel-data Phillips curve is estimated on a rolling basis.
  - Panel data cover 17 economies over 16 quarters.
  - Estimated coefficients are reported with their corresponding 90th percentile confidence intervals (see Annex Figure 1).
- Purpose:
  - To examine possible structural shifts and change in estimated coefficients over time.

### Rolling regression results and key findings
- Changes observed in recent rolling-window periods:
  - The estimated coefficient on the lagged headline (core) inflation has significantly increased in the last 3 rolling window periods, indicating inflation has become more backward-looking.
  - The pass-through of energy prices has increased substantially to core inflation.
  - The coefficient on the unemployment gap suggests the Phillips curve has become steeper over the last few years.
- Caveats:
  - Given few data points that underlie the change in estimated coefficients, it is premature to suggest any long-lasting changes in the structural relationships.
- Implications of observed changes:
  - Greater pass-through of commodity prices into core inflation increases susceptibility to negative supply shocks.
  - A larger effect of past inflation on current inflation suggests potential difficulties in reducing inflation going forward.

### Annex Figure note
- Annex I. Figure 1 presents Phillips Curve rolling regression results:
  - Solid lines indicate rolling regression coefficients of the headline (left) or core (right) regressions.
  - Dashed lines show their 10th–90th percentile confidence intervals.
  - Estimates are based on the baseline constrained regressions with country fixed effects and cover a rolling window of 16 quarters ending in the quarter indicated on the x axis.

*Source: IMF staff estimates and calculations (Annex I. Figure 1).*

---


_Source: https://www.imf.org/-/media/files/publications/selected-issues-papers/2023/english/sipea2023021.pdf_