## sipea2026011

## Source details

**Canonical URL:** [sipea2026011](https://www.imf.org/-/media/files/publications/selected-issues-papers/2026/english/sipea2026011.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/selected-issues-papers/2026/english/sipea2026011.pdf.md)
- [Structured JSON version](/-/media/files/publications/selected-issues-papers/2026/english/sipea2026011.pdf.json)

---

### Overview and key questions
- The Australian labor market displayed “remarkable resilience” post-pandemic despite slowing economic growth, driven by:
  - Persistently strong labor demand.
  - Rapid growth in labor supply (notably strong net migration and rising participation).
  - Contained wage growth despite low unemployment (Wage Phillips curve appearing to flatten or shift leftward).
- Paper focuses on three “labor market paradoxes”:
  - Why persistently low unemployment and high vacancies despite a cooling economy?
  - What explains elevated labor force participation and implications for labor supply?
  - Why has wage growth remained subdued at low unemployment and what does this imply for the NAIRU?

### Major findings (summary)
- Headline indicators likely overstate cyclical tightness
  - Headline unemployment and vacancy rates may have overstated cyclical labor market strength because structural shifts (e.g., large expansion in healthcare) and participation-margin dynamics masked weakening in job finding and retention.
- Elevated labor supply reflects both structural and cyclical forces
  - Structural trends (including strong net migration) and cyclical pressures (notably high cost-of-living and elevated interest rates) induced increased hours, participation, and holding of multiple jobs.
- Wage pressures contained by sectoral concentration and rising participation
  - Robust labor supply plus concentration of employment growth in non-market sectors (where wages are less sensitive to unemployment) reduced wage pressure at low unemployment, temporarily flattening and shifting the wage Phillips curve leftward and yielding temporarily lower NAIRU estimates.
- These forces may reverse as conditions normalize
  - If cyclical drivers retract and private demand picks up, the containment of wages may diminish; monitoring supply and demand drivers is critical for assessing slack and inflationary pressures.

### Paradox 1 — Persistently low unemployment and high vacancies despite cooling growth
- Key statistics and patterns:
  - Unemployment rate: record low of 3.4 percent in July 2022; gradually rose to 4.5 percent since (but remains lower than any time between 2008 and the COVID-19 pandemic).
  - Job vacancy rate: surged late-2020 to mid-2022; although declining since, it remains around half a percentage point higher than the highest levels seen in the pre-pandemic decade.
- Transition dynamics and flows:
  - Job finding rate (unemployment → employment) spiked at end-2021 and has remained above pre-pandemic levels, though gradually declining.
  - Flows from employment → unemployment (job loss) declined after the pandemic surge and have remained persistently lower than pre-pandemic levels.
  - Inflows and outflows between participation and nonparticipation have largely declined since early 2024 to lower than pre-pandemic levels.
  - Subdued flow from nonparticipation → unemployment materially lowered measured unemployment; variance decomposition (Shimer (2012) approach) finds low nonparticipation→unemployment flow rate a main contributor to post-2022 unemployment movements.
  - Counterfactual simulation holding labor force inflow/outflow at pre-2013 levels shows the counterfactual unemployment rate has edged up to pre-pandemic levels (Figure 3 referenced).
- Sectoral composition effects:
  - Non-market sectors (healthcare, education, public administration) had some of the lowest average unemployment rates over 2022Q1-2025Q1.
  - Market sectors exhibited higher unemployment on average; notable high unemployment in retail, administrative services, and accommodation and food services.
  - Labor reallocation toward non-market sectors and sectoral heterogeneity can mask loosening in market sectors.

### Vacancies and labor demand
- Vacancy composition and dynamics:
  - New vacancies have cooled while unfilled/old vacancies continue to keep upward pressure on the vacancy rate (simulations decomposing inflow/outflow of vacancies).
  - High unfilled vacancies may reflect lingering churn effects from the early recovery and/or persistent mismatches between jobs and worker preferences/qualifications.
- Sector concentration of vacancies:
  - High vacancies concentrated in healthcare and social services (COVID effects, NDIS expansion, aging population).
  - High vacancy rates also observed in accommodation and food services and arts and recreation, and in utilities, real estate, and mining.
  - Negative correlation between productivity growth and vacancy rates is robust to year and sector fixed effects.
- Policy-relevant diagnostic:
  - Elevated vacancies no longer mostly reflect new job creation; instead, unfilled positions and mismatches are important. Internet vacancy index and employment intentions point to further loosening underway.

### Paradox 2 — Labor supply and participation
- Aggregate and international context:
  - Australia experienced the largest labor force expansion among peers post-pandemic, driven by strong net migration after borders reopened and a significant rise in labor force participation.
- Cyclical role in participation:
  - Cost-of-living pressures and high interest rates incentivized increased labor supply among debt-constrained households, including multiple job-holding.
- Implication:
  - Above-trend participation growth helped offset upward wage pressures from tight demand.

### Paradox 3 — Wages and drivers
- Wage outcomes:
  - Wage growth remained contained relative to historical relationships with unemployment; Wage Phillips curve appears flatter or shifted left.
  - Real wage growth over 2020-2024 was lower than in many peers and below the OECD average, though real wage growth picked up in late 2024 and the first half of 2025 (cross-country data only available annually through 2024).
- Contributing factors:
  - Concentration of employment growth in non-market sectors where wages are less sensitive to unemployment.
  - Elevated labor force participation increasing labor supply.
  - Sectoral productivity differences affecting wage dynamics and job desirability.
- Implication for NAIRU:
  - These effects likely produced temporarily lower estimates of the NAIRU; these could be temporary if cyclical elements reverse.

### Participation trends and driving mechanisms (section highlights)
- Long-run trend:
  - Labor force participation (LFP) has been growing at a rate of 0.1 percentage points per year in recent decades.
- Demographic and policy contributors:
  - Higher female LFP, level shift up post pandemic; LFP for mothers of young children improved sharply by 2023, potentially reflecting July 2023 reforms to childcare subsidies and flexible work arrangements.
  - Rising LFP of older workers and workers with long-term health conditions contributed to higher labor supply over the past decade.
- Role of migration:
  - Migrants have, on average, higher LFP due to migration system focus on work visas; strong inward migration flows over several decades supported long-term growth in LFP.
  - Pandemic-related drop in net migration in 2020-2021 contributed to a temporary decline in LFP; return of migrant workers after borders reopened contributed to a quick recovery in 2021-2022.
- Monetary tightening and indebted households (evidence from Das et al., forthcoming):
  - Australia entered a monetary tightening cycle in May 2022 as inflation rose above the target band.
  - Individuals in the most indebted quintile experienced a 1 percentage point rise in their employment probability compared to those in the lowest indebtedness quintile following the onset of monetary tightening (statistically significant at the 1 percent level).
  - Quintile 5 (most indebted) shows:
    - a 1 percentage point rise in the number of jobs they hold,
    - a 3.5 percentage point rise in total labor earnings.
  - Effects larger for individuals who were not full-time workers prior to tightening; effect roughly doubles when restricting sample to those not full-time beforehand.
  - Larger increases for individuals who were not primary earners in 2021 and those without children; parents of young children showed a muted response but increased labor supply after 2023 childcare subsidy changes.
- Magnitude of cyclical contribution to rising LFP:
  - Preliminary quantitative bounds: the squeeze on household balance sheets may have contributed between 0.1 and 0.4 percentage points to the approximately 2 percentage point increase in LFP observed since July 2021.
    - Lower bound assumes 50 percent of households are mortgage holders, 1/5th of which see an increase in labor supply.
    - Upper bound also includes those who are non-mortgage holders.
  - Regression evidence indicates interest rates are a much stronger predictor of labor supply in the fifth quintile than inflation.
- Possible future dynamics:
  - LFP may decline as disinflation becomes more entrenched and monetary easing proceeds; if labor demand does not soften similarly, a decline in participation could increase measured labor market tightness.

### Wage Phillips Curve (WPC) empirical results and decomposition
- Aggregate WPC findings:
  - Negative real wage gap over 2022-2023 indicates real wages were below levels implied by long-run productivity trends.
  - Short-run sensitivities:
    - A one percentage point decrease in unemployment below its trend is associated with an increase in wage inflation of 0.3-0.4 percentage points.
    - Participation: For every one percentage point increase in participation above its long-run trend, wage growth is 0.5 percentage points slower.
    - Productivity catch-up: For every one percentage point that real wages are below the level implied by productivity, nominal wage inflation is pushed up by just under 0.1 percentage points per quarter (statistically significant).
    - Lagged wage growth and expected inflation are both significant drivers of wage growth.
    - Vacancies-to-unemployment ratio associated with upward wage pressures, though not always statistically significant.
- Regression-based decomposition (2016–2025):
  - Unemployment gap contributed positively to wage growth from 2022Q3 through mid-2024.
  - Participation contribution: positive early in the decade when participation was restricted, then negative once participation rose above trend.
  - Residual term positive in 2021 and early 2022, suggesting cost-of-living pressures strengthened worker wage demands even as inflation expectations remained anchored.
- Sector-level heterogeneity:
  - Employment growth over Feb 2020–Feb 2025 concentrated in healthcare and social assistance (largest), followed by construction, public administration, education, and professional services.
  - Sector-level WPCs corrected for productivity catch-up show varying sensitivities:
    - Strong negative correlation between unemployment gaps and wage growth in sectors with moderate/high share of individual agreements: administrative, financial, professional services, retail trade.
    - Non-significant or positive relationships in sectors with low share of individual agreements: education, healthcare, public administration.
    - Among high-employment-growth sectors, healthcare, construction, public administration, and education show no significant short-term WPC relationship; professional services is an exception with significant negative relationship.
  - Productivity and wage gaps by sector:
    - Productivity growth since 2020: highest in information media and telecommunications; sharpest declines in mining and healthcare.
    - Real wage gaps opened across sectors, largest in mining, narrowest in accommodation and food services; recent narrowing only in accommodation and food services, healthcare, and retail trade.
- Resulting assessment of NAIRU:
  - Latest IMF assessment: NAIRU at 4.3 percent as of 2025Q3, with a range of 4.1 to 4.5 percent.
  - Likely temporary decline in NAIRU driven by elevated labor supply and sectoral composition; medium-term NAIRU likely to return closer to pre-pandemic levels if cyclical supply support reverses.

### Policy implications and recommendations
- Monitoring and measurement:
  - Use high-frequency and flow-based labor market metrics (job finding/retention, participation flows, internet vacancy indices, employment intentions) to detect early easing in cyclical conditions masked by headline indicators.
  - Use alternative measures of slack—adjusted unemployment or vacancies measures, employment intentions—to capture changes more timely.
- Structural reforms to improve matching and reduce unfilled vacancies:
  - Facilitate labor mobility.
  - Continue to adjust migration policies in response to sectoral labor shortages.
  - Reforms to improve job quality and re-skill workers for sectors with persistent vacancies.
- Recent policy steps viewed as constructive:
  - Restrictions on non-competes for low-skill jobs.
  - Prioritization and streamlining for some healthcare workers under immigration policies.
- Fiscal considerations:
  - Strength in non-market sectors (partly driven by COVID-related gaps and NDIS expansion) is likely to fade as those gaps are filled and fiscal consolidation continues; policymakers should account for this in projections.
- Monetary policy guidance:
  - Carefully track wages and links with participation, productivity, and labor demand to appropriately target contained price pressures and full employment going forward.

### Annex — counterfactual methods (summary)
- Counterfactual unemployment (three-state continuous-time Markov chain across {E, N, U}):
  - Construct monthly transition probability matrix P_t; compute infinitesimal generator Q_t as matrix logarithm of P_t.
  - Build counterfactual generator Q_t tilde by holding specific transition rates at period averages; obtain counterfactual monthly transition matrix P_t tilde via matrix exponential; simulate shares forward and compute counterfactual unemployment as unemployed share / (unemployed + employed shares).
  - Variance decomposition: simulate counterfactual unemployment holding all but one transition rate constant; contribution measured by Cov(u_t tilde, u_t) / Var(u_t).
  - Decomposition findings for two periods (selected contributions):
    - 2020-2021 transition contributions: UE 0.29, EU 0.27, NU 0.20, UN 0.08, NE 0.04, EN 0.01.
    - Since 2022 transition contributions: UE 0.40, NU 0.34, EU 0.26, UN 0.04, NE -0.02, EN -0.14.
- Counterfactual vacancy rates:
  - Using Revelio Labs monthly data on removed and new vacancies, outflow rate φ_t = Removed Vacancies / Active Vacancies, Lag; inflow rate i_t = New Vacancies / Employment.
  - Steady-state vacancy rate approximation: v_i = i / (i + φ̅) with φ fixed at its average; v_φ = i̅ / (i̅ + φ) with i fixed at its average.

### Annex II — Monetary Policy and Labor Supply (data and empirical strategy)
- Data:
  - Administrative data covering all 9.3 million Australian households, including 3.3 million mortgage-holding households.
  - Matches household characteristics from the 2021 Census (including mortgage debt service) with monthly administrative data on income and employment.
- Empirical strategy:
  - Estimates dynamic response of labor supply by indebtedness using repeated cross-section local regressions for individual i in month t:
    - Y_i,t − Y_i,Jul21 = α_i,t + ∑_g β_g,t * I(G=g) + X'_i,t γ_t + u_i,t
    - Outcomes: {Labour income, employment (1/0), number of jobs (if >0)}.
    - I(G=g) indicates DSR group g; first (lowest) DSR quintile omitted as reference.
    - Controls X_i,t include individual and HH income, age, age^2, gender, education.
  - Specification allows estimation of dynamic labor supply responses by indebtedness group and outcome.
- Controls and robustness:
  - Results robust to additional controls: household income in 2020 and 2019, detailed geographic location, number of bedrooms and people in the home, number of children, tenure (>1 or >5 years).
  - Qualitatively unaffected by dropping States with longer lockdowns.

*Source: "1. Labor Market Conditions in Australia" (selected issues paper excerpt).*

### 1. Labor Market Conditions in Australia_________________________________________________ 4

### 1. Labor Market Conditions in Australia

### Overview and key questions
- The Australian labor market has displayed “remarkable resilience” post-pandemic despite slowing economic growth, driven by:
  - Persistently strong labor demand.
  - Rapid growth in labor supply (notably strong net migration and rising participation).
  - Contained wage growth despite low unemployment (Wage Phillips curve appearing to flatten or shift leftward).
- The paper focuses on three “labor market paradoxes”:
  - Why persistently low unemployment and high vacancies despite a cooling economy?
  - What explains elevated labor force participation and implications for labor supply?
  - Why has wage growth remained subdued at low unemployment and what does this imply for the NAIRU?

### Major findings (summarized)
- Headline indicators likely overstate cyclical tightness
  - Headline unemployment and vacancy rates may have overstated cyclical labor market strength because structural shifts (e.g., large expansion in healthcare) and participation-margin dynamics masked weakening in job finding and retention.
- Elevated labor supply reflects both structural and cyclical forces
  - Structural trends (including strong net migration) and cyclical pressures (notably high cost-of-living and elevated interest rates) induced increased hours, participation, and holding of multiple jobs.
- Wage pressures have been contained by sectoral concentration and rising participation
  - Robust labor supply plus concentration of employment growth in non-market sectors (where wages are less sensitive to unemployment) reduced wage pressure at low unemployment, temporarily flattening and shifting the wage Phillips curve leftward and yielding temporarily lower NAIRU estimates.
- These forces may reverse as conditions normalize
  - If cyclical drivers retract and private demand picks up, the containment of wages may diminish; monitoring supply and demand drivers is critical for assessing slack and inflationary pressures.

### Paradox 1 — Persistently low unemployment and high vacancies despite cooling growth
- Key statistics and patterns:
  - Unemployment rate: record low of 3.4 percent in July 2022; gradually rose to 4.5 percent since (but remains lower than any time between 2008 and the COVID-19 pandemic).
  - Job vacancy rate: surged late-2020 to mid-2022; although declining since, it remains around half a percentage point higher than the highest levels seen in the pre-pandemic decade.
- Transition dynamics:
  - Job finding rate (unemployment → employment) spiked at end-2021 and has remained above pre-pandemic levels, though gradually declining.
  - Flows from employment → unemployment (job loss) declined after the pandemic surge and have remained persistently lower than pre-pandemic levels (strong job retention).
  - Inflows and outflows between participation and nonparticipation have largely declined since early 2024 to lower than pre-pandemic levels.
  - The subdued flow from nonparticipation → unemployment materially lowered measured unemployment; variance decomposition (Shimer (2012) approach) finds low nonparticipation→unemployment flow rate a main contributor to post-2022 unemployment movements.
  - Counterfactual simulation holding labor force inflow/outflow at pre-2013 levels shows the counterfactual unemployment rate has edged up to pre-pandemic levels (Figure 3).
- Sectoral composition effects:
  - Non-market sectors (healthcare, education, public administration) had some of the lowest average unemployment rates over 2022Q1-2025Q1.
  - Market sectors exhibited higher unemployment on average; notable high unemployment in retail, administrative services, and accommodation and food services.
  - Labor reallocation toward non-market sectors and sectoral heterogeneity can mask loosening in market sectors.

### Vacancies and labor demand
- Vacancy composition and dynamics:
  - New vacancies have cooled while unfilled/old vacancies continue to keep upward pressure on the vacancy rate (simulations decomposing inflow/outflow of vacancies).
  - High unfilled vacancies may reflect lingering churn effects from the early recovery and/or persistent mismatches between jobs and worker preferences/qualifications.
- Sector concentration of vacancies:
  - High vacancies concentrated in healthcare and social services (COVID effects, NDIS expansion, aging population).
  - High vacancy rates also observed in accommodation and food services and arts and recreation (consistent with peer AEs), and in utilities, real estate, and mining.
  - Negative correlation between productivity growth and vacancy rates is robust to year and sector fixed effects (interpretations include productivity affecting wage growth and job desirability).
- Policy-relevant diagnostic:
  - Elevated vacancies no longer mostly reflect new job creation; instead, unfilled positions and mismatches are important. Internet vacancy index and employment intentions point to further loosening underway.

### Paradox 2 — Labor supply and participation
- Australia experienced the largest labor force expansion among peers post-pandemic, driven by:
  - Strong net migration after borders reopened.
  - A significant rise in labor force participation (some peers saw declines).
- Cyclical role in participation:
  - Cost-of-living pressures and high interest rates incentivized increased labor supply among debt-constrained households, including multiple job-holding.
- Implication:
  - Above-trend participation growth helped offset upward wage pressures from tight demand.

### Paradox 3 — Wages and drivers
- Wage outcomes:
  - Wage growth remained contained relative to historical relationships with unemployment; the Wage Phillips curve appears flatter or shifted left.
  - Real wage growth over 2020-2024 was lower than in many peers and below the OECD average, though the source notes real wage growth picked up in late 2024 and the first half of 2025 (cross-country data only available annually through 2024).
- Contributing factors to subdued wages:
  - Concentration of employment growth in non-market sectors where wages are less sensitive to unemployment.
  - Elevated labor force participation increasing labor supply.
  - Sectoral productivity differences affecting wage dynamics and job desirability.
- Implication for NAIRU:
  - The combination of effects likely produced temporarily lower estimates of the NAIRU; these could be temporary if the cyclical elements reverse.

### Policy implications and recommendations (from discussion)
- Monitoring and measurement:
  - Use high-frequency and flow-based labor market metrics (job finding/retention, participation flows, internet vacancy indices, employment intentions) to detect early easing in cyclical conditions masked by headline indicators.
- Structural reforms to improve matching and reduce unfilled vacancies:
  - Facilitate labor mobility.
  - Continue to adjust migration policies in response to sectoral labor shortages.
  - Reforms that reduce mismatches could include measures that improve job quality and re-skill workers for sectors with persistent vacancies.
- Recent policy steps noted as constructive:
  - Restrictions on non-competes for low-skill jobs.
  - Prioritization and streamlining for some healthcare workers under immigration policies.
- Fiscal considerations:
  - Strength in non-market sectors (partly driven by COVID-related gaps and NDIS expansion) is likely to fade as those gaps are filled and fiscal consolidation continues; policymakers should account for this in projections.

*Source: "1. Labor Market Conditions in Australia" (selected issues paper excerpt).*

### 13.      Labor force participation is on a long-run upward trend which has accelerated in

### 13.      Labor force participation is on a long-run upward trend which has accelerated in recent years, reflecting both structural and cyclical developments.

### Structural trends and long-run evolution of LFP
- Long-term trend: labor force participation (LFP) has been growing at a rate of 0.1 percentage points per year in recent decades.
- Female participation:
  - Higher female LFP has played a key role supporting the rise in aggregate LFP.
  - Female LFP in Australia saw a level shift up in the post pandemic period.
  - LFP for mothers of young children improved sharply by 2023 — potentially reflecting July 2023 reforms to childcare subsidies and the growing prevalence of flexible work arrangements.
- Older workers and health conditions:
  - Rising LFP rates of older workers (consistent with longer retirement periods due to rising life expectancy) and of workers with long-term health conditions have contributed to higher labor supply over the past decade.

### Role of migration
- Migrants have, on average, higher labor force participation than domestic citizens due to the migration system’s focus on work visas to meet skill shortages.
- Strong inward migration flows over several decades supported long-term growth in LFP.
- Pandemic impacts:
  - Cross-border movement restrictions during the pandemic produced temporary volatility in net migration.
  - The sharp drop in net migration in 2020-2021 contributed to a temporary decline in LFP.
  - The return of migrant workers after borders reopened contributed to a quick recovery of LFP in 2021-2022; this impact subsided as net migration normalized.

### Monetary tightening, cost-of-living, and cyclical effects on labor supply
- Timeline:
  - Australia entered a monetary tightening cycle in May 2022 as inflation rose above the target band.
- Mechanism:
  - Higher mortgage payments and higher cost-of-living squeezed household balance sheets.
  - Credit-constrained households may have increased labor supply to afford mortgage payments and to cope with higher inflation-driven expenses.
- Evidence summary (Das et al., forthcoming; methodology in Annex II):
  - Individuals in the most indebted quintile experienced a 1 percentage point rise in their employment probability compared to those in the lowest indebtedness quintile following the onset of monetary tightening. This effect is statistically significant at the 1 percent level and consistent throughout the sample period.
  - Quintiles 2 through 4 did not exhibit a similar change.
  - The fifth (most indebted) quintile shows:
    - a 1 percentage point rise in the number of jobs they hold,
    - a 3.5 percentage point rise in total labor earnings.
  - These changes indicate labor supply expansion along both the extensive and intensive margins.
- Heterogeneity:
  - Effects larger for individuals who were not full-time workers prior to tightening — the effect on labor supply roughly doubles when restricting the sample to those not full-time beforehand.
  - Larger increases for individuals who were not primary earners in 2021 and those without children.
  - Parents of young children showed a muted response, but following the 2023 increase in the generosity of the government’s childcare subsidy there is evidence that individuals with young children increased their labor supply, nearly catching up with those with older children.
  - The increase in labor supply corresponded with individuals’ expectations of rate increases.

### Magnitude of cyclical contribution to rising LFP
- Attribution challenge: It is difficult to definitively allocate labor supply increases between higher interest rates and higher cost-of-living since both occurred simultaneously and squeezed household balance sheets.
- Preliminary quantitative bounds:
  - The squeeze on household balance sheets may have contributed between 0.1 and 0.4 percentage points to the approximately 2 percentage point increase in LFP observed since July 2021.
    - The lower bound assumes 50 percent of households are mortgage holders, 1/5th of which see an increase in labor supply.
    - The upper bound also includes those who are non-mortgage holders.
- Regression evidence in Das et al. indicates interest rates are a much stronger predictor of labor supply in the fifth quintile than inflation.

### Implications and policy considerations for participation monitoring
- Both structural and cyclical factors explain rising LFP; monitoring participation remains critical for policymakers given its role in labor market outcomes.
- Recent rising labor supply met strong labor demand in the post-pandemic period, supporting high employment growth.
- Possible future dynamics:
  - LFP may decline as disinflation becomes more entrenched and monetary easing proceeds.
  - If labor demand does not soften similarly, a decline in participation could increase measured labor market tightness.
- Policy implications:
  - Monetary policymakers should closely monitor participation developments and implications for labor market outcomes given the dual mandate.
  - Over the longer term, structural changes (aging populations, immigration policies, participation of older workers) will continue to shape LFP.
  - Structural policies should be calibrated to ensure labor supply shortages are met, to avoid wage pressures or negative impacts on growth.

### Paradox: Subdued wage growth despite tight labor markets (context and key findings)
- Recent pattern:
  - Tight labor markets and high inflation in the post-pandemic period did not produce as large a rise in wages (not adjusted for productivity) as historically expected.
  - In 2022 and 2023, vacancies rose sharply and businesses reported difficulty finding and retaining workers, but real wage growth was negative and at its lowest levels in almost two decades.
  - In 2024, real wage growth ticked up only slightly into positive territory.
- Wage Phillips Curve (WPC) dynamics:
  - The wage Phillips curve has appeared to shift left or flatten since 2021: unemployment well below historical levels, but nominal wage growth no higher than in the past.
- Candidate explanations for muted wage growth:
  - Productivity:
    - Measured as GDP per hour worked, labor productivity rose over 2020-early 2022, then declined sharply in the following years, only stabilizing in 2025.
    - Declining labor productivity can exert downward pressure on wages; rapid rise in unit labor costs suggests wage pressures from labor markets may have been offset by waning productivity, though productivity alone does not fully explain muted wage growth in 2022–early 2023.
  - Labor force participation shifts:
    - Increased LFP can reduce upward pressure on wages at the same level of unemployment (shifting WPC left).
    - The rise in LFP above its long-term trend likely decreased wage pressure in recent years.
  - Sectoral concentration of labor demand:
    - Concentration of labor demand in sectors where wages are less responsive to tightness (potentially non-market sectors or sectors with lower preponderance of individual agreement wages) could flatten aggregate wage responsiveness.
    - Employment concentrated in sectors with lower productivity and slower productivity growth could also contain wage growth.
- Empirical approach to drivers:
  - A wage Phillips curve augmented with an error correction term for deviations of wages from the long-run relationship with productivity was estimated (two-step Engle and Granger approach; methodology in Annex III).
  - Analysis performed at national and sector levels; WPI measure of wages used in the regressions (AENA measures suffered non-stationarity and autocorrelation and were not included).

### Results from aggregate Wage Phillips Curve analysis (key quantitative findings)
- Wage gaps:
  - A negative real wage gap over 2022-2023 indicates real wages were below levels implied by long-run productivity trends; soft productivity growth alone cannot account for low wage growth in the post-pandemic period.
- Short-run regression sensitivities:
  - Labor market tightness:
    - On average, a one percentage point decrease in unemployment below its trend is associated with an increase in wage inflation of 0.3-0.4 percentage points (linear relationship assumed).
    - Increase in the vacancies-to-unemployment ratio is associated with upward wage pressures, though results are not statistically significant.
    - Sensitivity of wages to unemployment was stronger in the pre-pandemic period than over the full sample, suggesting reduced sensitivity in recent years.
  - Productivity catch-up:
    - For every one percentage point that real wages are below the level implied by productivity, nominal wage inflation is pushed up by just under 0.1 percentage points per quarter (statistically significant).
  - Participation:
    - For every one percentage point increase in participation above its long-run trend, wage growth is 0.5 percentage points slower.
    - Participation was not significant in the pre-2020 sample alone, suggesting the association between wages and participation may have strengthened since 2020.
  - Other factors:
    - Lagged wage growth and expected inflation are both significant drivers of wage growth.
    - Short-run changes in import price growth and in labor productivity growth (outside the long-run catch-up effect) are not statistically significant.

*Source: IMF staff analysis as presented in the provided content unit.*

### 26.      Both cyclical and structural factors

### Both cyclical and structural factors

### Decomposition of wage growth drivers (2016–2025)
- Regression-based decomposition findings:
  - The unemployment gap contributed positively to wage growth from 2022Q3 through mid-2024, consistent with tight labor market conditions.
  - Inflation expectations have historically played a significant role in wage growth formation, but because they remained anchored even as the economy faced significant price pressures, their contribution to wage growth did not increase substantially.
  - A positive contribution of the residual term to wage growth in 2021 and early 2022 suggests rising cost of living pressures may have strengthened worker demands for higher wages even as inflation expectations remained anchored.
  - Participation (deviation from long term trends, capturing cyclical/short-term components) contributed positively to wage growth early in the decade when labor supply was restricted by border closures; once participation began to increase more rapidly above its long-term trend the contribution became negative.
  - Gradual wage catch-up to the level implied by productivity trends has contributed positively to wage growth since 2022. Even though productivity was declining over much of this period, real wages initially dropped even further below levels implied by the long-run co-integrating relationship with productivity.

### Sector-level wage Phillips curve (WPC) analysis — key regression results
- Purpose: determine sensitivity of wages to cyclical and structural factors across sectors given differing wage setting methods (individual arrangements vs award wages or collective agreements).
- Selected coefficient estimates (sector-aggregated WPC specifications reported):
  - Lagged wage inflation (sum of coefficients): 0.764***, 0.760***, 0.812***, 0.796***, 0.769***, 0.790***, 0.790***, 0.715*** (robust SEs shown in table).
  - Inflation expectations (1 yr-ahead): 0.236***, 0.240***, 0.188***, 0.204***, 0.231***, 0.210***, 0.210***, 0.285***.
  - Catch-up to productivity trends (ECM term): -0.089***, -0.088***, -0.080***, -0.066**, -0.074***, -0.092***, -0.126***, -0.126***.
  - Unemployment (deviation from trend): -0.337**, -0.357**, -0.365***, -0.407**, -0.523** (selected columns).
  - Vacancies to unemployment: 0.280, 1.269, 0.824*** (selected columns).
  - Participation gap: -0.523*, -0.398, -0.002, 0.035, 0.168 (selected columns).
  - Labor productivity growth and import price growth coefficients reported (varied signs and significance across specifications).
  - Observations across regressions: 107, 107, 107, 101, 98, 99, 86, 81.
  - Periods: Full sample; variants exclude 2020-21; pre-2020 specifications also shown.
  - Robust standard errors reported in parentheses; significance: *** p<0.01, ** p<0.05, * p<0.1.

### Employment concentration and sectoral implications
- Employment growth over Feb 2020–Feb 2025 concentrated in:
  - Healthcare and social assistance (largest), followed by construction, public administration, education, and professional services.
- Implications:
  - If wages in high-employment-growth sectors are less responsive to labor market tightness (higher use of collective agreements and award wages, lower share of individual agreements), aggregate wage sensitivity to labor market conditions is reduced.
  - This can produce a short-term reduction in the NAIRU (lower unemployment not translating as quickly into higher wages and higher prices), though not expected to persist in the medium term.

### Wage-productivity gaps by sector
- Productivity growth since 2020 (change in GVA per hours worked) varied across sectors:
  - Highest growth: information media and telecommunications.
  - Sharpest declines: mining and healthcare.
- Error-correction WPC results indicate:
  - Real wages across sectors fell below levels implied by the long-run relationship with productivity early in the decade and have not fully recovered.
  - Negative real wage gaps opened across sectors, largest in mining, narrowest in accommodation and food services.
  - In recent quarters wage gaps narrowed only in accommodation and food services, healthcare, and retail trade.
- Aggregate reduction in the real wage gap may reflect employment shifts toward sectors with smaller gaps, not only closing of sector-level gaps.

### Heterogeneity in wage sensitivity across sectors
- Sector-level WPCs (corrected for productivity catch-up) show varying sensitivities of wage growth to unemployment gaps:
  - Strong and significant negative correlation between unemployment gaps and wage growth in sectors with a moderate/high share of individual agreements: administrative, financial, professional services, retail trade.
  - Non-significant or positive relationships in sectors with low share of individual agreements: education, healthcare, public administration.
- Among sectors with highest employment growth since 2020, healthcare, construction, public administration, and education show no significant short-term WPC relationship, implying flatter WPCs.
- Professional services was an exception among high-growth sectors, showing a significant negative relationship between unemployment gaps and wages.
- Conclusion: Concentration of labor demand in sectors with flatter WPCs helps explain contained aggregate wage growth despite tight aggregate labor market conditions.

### Conclusions on drivers of labor market and wage dynamics
- Factors keeping wage pressures contained despite low unemployment:
  - Elevated labor supply.
  - Concentration of employment growth in sectors with flatter wage Phillips curves (non-market sectors).
  - Contained inflation expectations mitigating cost-of-living impacts.
  - Above-trend labor force participation and low productivity growth (with low productivity growth contributing more early in the decade).
- Temporary nature:
  - These factors are consistent with a temporary shift in the Wage Phillips curve and a temporarily lower NAIRU.
  - The latest IMF assessment: NAIRU at 4.3 percent as of 2025Q3, with a range of 4.1 to 4.5 percent.
- Risks of re-emerging wage pressures:
  - If above-trend labor force participation (partly reflecting cost-of-living pressures and tight monetary policies) stabilizes or declines as economic conditions normalize, labor supply could tighten and increase wage pressure at the same unemployment levels.
  - Employment growth shifting from non-market to market sectors (where wages are more sensitive) and lagged rises in collective agreement and award wages could raise wage pressures.
  - Medium-term NAIRU likely to return closer to pre-pandemic levels.

### Policy implications and recommendations
- Monitor labor market beyond headline figures:
  - Use alternative measures of slack—adjusted unemployment or vacancies measures, employment intentions—to capture changes in labor market conditions more timely or at higher frequency.
  - Disentangle cyclical vs structural drivers of participation and vacancies.
- To address persistent unmet labor demand:
  - Consider refining migration frameworks or improving labor market mobility.
  - Note recent reforms around non-competes and efforts to streamline and prioritize migration processes for some healthcare workers are viewed as welcome steps.
- Track labor force participation carefully:
  - Structural shifts raising participation include strong net migration, higher female participation, and increased participation by older and less healthy workers.
  - Cyclical forces (cost-of-living, elevated interest payments) also boosted participation through second jobs and higher labor supply.
  - Participation from constrained households could unwind as conditions normalize, tightening labor supply.
- Monetary policy guidance:
  - Carefully track wages and links with participation, productivity, and labor demand to appropriately target contained price pressures and full employment going forward.

### Annex — counterfactual methods (summary)
- Counterfactual unemployment (three-state continuous-time Markov chain across {E, N, U}):
  - Construct monthly transition probability matrix P_t; compute infinitesimal generator Q_t as matrix logarithm of P_t.
  - Build counterfactual generator Q_t tilde by holding specific transition rates at period averages; obtain counterfactual monthly transition matrix P_t tilde via matrix exponential; simulate shares forward and compute counterfactual unemployment as unemployed share / (unemployed + employed shares).
  - Variance decomposition: simulate counterfactual unemployment holding all but one transition rate constant; contribution measured by Cov(u_t tilde, u_t) / Var(u_t).
  - Decomposition findings for two periods:
    - 2020-2021 transition contributions (selected): UE 0.29, EU 0.27, NU 0.20, UN 0.08, NE 0.04, EN 0.01.
    - Since 2022 transition contributions (selected): UE 0.40, NU 0.34, EU 0.26, UN 0.04, NE -0.02, EN -0.14.
- Counterfactual vacancy rates:
  - Using Revelio Labs monthly data on removed and new vacancies, outflow rate φ_t = Removed Vacancies / Active Vacancies, Lag; inflow rate i_t = New Vacancies / Employment.
  - Steady-state vacancy rate approximation: v_i = i / (i + φ̅) with φ fixed at its average; v_φ = i̅ / (i̅ + φ) with i fixed at its average.

*Source: sipea2026011 - 26.      Both cyclical and structural factors*

### Annex II. Monetary Policy and Labor Supply

### Annex II. Monetary Policy and Labor Supply

### Data
- Uses administrative data from Australia covering all 9.3 million Australian households, including 3.3 million mortgage-holding households.
- Matches household characteristics from the 2021 Census (including mortgage debt service) with monthly administrative data on income and employment.
- The data set enables quantification of the magnitude of the average household response to the rate hike and exploration of heterogeneity in responses across households with different characteristics.

### Empirical strategy
- To study the dynamic response of labor supply by level of indebtedness, a local regression is estimated in repeated cross sections for individual i in month t:
  - Y_i,t − Y_i,Jul21 = α_i,t + ∑_g β_g,t * I(G=g) + X'_i,t γ_t + u_i,t
  - Where Y_i,t − Y_i,Jul21 is the change in outcome of interest since July 2021 ∈ {Labour income, employment (1/0), number of jobs (if >0)}.
  - I(G=g) is an indicator if in DSR group g; the first (lowest) DSR quintile is omitted and serves as the reference group.
  - X_i,t is a vector of controls, including individual and HH income, age, age^2, gender, education.
- The specification allows estimation of dynamic labor supply responses by indebtedness group (DSR quintiles) and outcome.

### Controls and robustness
- Results are robust to using a large suite of additional controls, including:
  - household income in 2020 and 2019,
  - detailed geographic location,
  - number of bedrooms and people in the home,
  - number of children,
  - whether the household has been in that home for more than 1 or 5 years.
- Results are qualitatively unaffected by dropping States which had longer lockdowns.

*Source: Annex II. Monetary Policy and Labor Supply, sipea2026011*

---


_Source: https://www.imf.org/-/media/files/publications/selected-issues-papers/2026/english/sipea2026011.pdf_