## sipea2026046

## Source details

**Canonical URL:** [sipea2026046](https://www.imf.org/-/media/files/publications/selected-issues-papers/2026/english/sipea2026046.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/selected-issues-papers/2026/english/sipea2026046.pdf.md)
- [Structured JSON version](/-/media/files/publications/selected-issues-papers/2026/english/sipea2026046.pdf.json)

---

### Overview
- Paper addresses three key questions:
  - What are the stylized facts about Hong Kong SAR’s labor market? What are structural trends and cyclical patterns in labor force, employment and LFPR (Section B)?
  - What are the main compositional drivers of Hong Kong SAR’s declining LFPR? Specifically, how much of the decline can be attributed to population aging versus falling youth labor force participation (Section C)? How does Hong Kong SAR’s decomposition of LFPR changes compare to peer economies?
  - What factors explain the recent decline in youth participation (Section D)? What are the structural drivers—such as economic structure transformation, higher labor market educational attainment, and technological change—that contribute to the decline of youth LFPR? Does rising re-education among youth reflect adaptation to evolving skill demands amid the emergence of artificial intelligence and other technological disruptions?
- Main synthesis:
  - Demographic trends, youth education choices, and structural labor market shifts shape LFPR outcomes in Hong Kong SAR.
  - Shift-share analysis shows population aging remains the primary driver of the decline in aggregate LFPR, while a recent fall in youth LFPR also contributed, distinguishing Hong Kong SAR from some peer economies.
  - Household surveys and labor market data show continued education is the main driver of the decline in youth LFPR, linked to sectoral structural shifts, evolving labor skill requirements, emergence of AI, and possibly rising labor market competitiveness from higher educational attainment.
  - The paper concludes with policy implications (policy implications are noted but not detailed in the supplied excerpts).

### B. Stylized Facts of Hong Kong SAR’s Labor Market — Structural and Cyclical Trends
- Structural trends (past three decades):
  - Share of older workers (aged 55 and above) rose from 13 percent in 2010 to 25 percent in 2025Q2.
  - Share with post-secondary education (degree or above) increased from 24 percent in 2010 to 36 percent in 2025Q2.
  - Among youth (aged 15–24), post-secondary attainment rose from 25 percent in 2010 to 52 percent in 2025Q2.
  - Note: Labor force includes foreign domestic helpers.
- Cyclical indicators (pandemic to recovery):
  - Unemployment rate declined to below 4 percent following the COVID reopening, down from a peak of 7 percent in 2021Q1.
  - Hourly wage growth rose to above 3 percent year-on-year from a trough of 0.9 percent during the pandemic peak.
  - Unemployment among those with post-secondary qualifications has trended upward over time and recently converged with that of workers with lower educational attainment.
  - Youth unemployment rate has diverged markedly from other age groups over the past decade, with the gap widening in recent years.
- Compositional shifts in LFPR:
  - LFPR for ages 20–24 fell from above 70 percent in 2005 to slightly below 50 percent in recent years.
  - Share of older individuals (aged 65 and above) rose from around 14 percent of the population aged 15 and above in 2005 to over 26 percent of this population.
  - Participation rate among seniors has increased gradually but remains low relative to other age groups.
- Education and occupation structure:
  - Average years of schooling for individuals aged 15-64 increased significantly since the 1990s; Hong Kong SAR experienced a rise exceeding that in most peer advanced economies.
  - Youth participation rates fell across all education groups during 2000-20; for youth with secondary or non-degree post-secondary attainment, LFPR fell by almost 10 percentage points.
  - Post-2000: share of employed workers in manager, professional and associate professional roles has risen since 2005; employment in elementary occupations and clerical support roles has declined.
  - Sectoral employment changes (2025Q4 vs. 2018Q4): employment expanded in professional services and real estate; substantial declines recorded in manufacturing, trade and wholesale, retail, and accommodation and food services.
  - Finance and other professional service sectors account for a growing share of youth employment.

### Cross-Economy Comparison (stylized)
- Hong Kong SAR shares low and decreasing youth LFPR with regional peers Korea and Singapore.
  - Korea: youth LFPR relatively stable but significantly below other advanced economies; large share of youth postpone careers to land jobs in large firms or professional jobs.
  - Singapore: downward trend where education-related delays have contributed to falling youth LFPR.
- Comparison finding: Hong Kong SAR stands out relative to most Asia-Pacific peers (Australia, Japan, Korea, New Zealand and Singapore) in the SSA results, with Singapore the closest comparator in terms of overall LFPR decline.

### C. Shift-Share Analysis (SSA) — Quantifying Demographic vs Behavioral Contributions
- Methodology:
  - Decomposition separates change in LFPR into the extensive margin (demographics: cohort population shares held constant) and the intensive margin (behavioral: within-cohort LFPR changes held constant), plus a residual.
  - Population aged 15 and above broken into ten five-year cohorts from 15–19 through 60–64, and a final 65+ cohort.
  - Analysis focuses on medium-term changes: change in each year relative to 5 years earlier (b = 5).
  - Data source: mid-year population statistics and LFPR data from Hong Kong SAR Census and Statistics Department (C&SD).
- Key quantitative results:
  - Aggregate LFPR declined by 3.7 percentage points in 2024 relative to 2020.
  - Of that 3.7 percentage point decline, about 0.22 percentage points was due to behavioral changes (within-cohort LFPR changes).
  - The demographic impact (shift in population shares) accounts for the bulk of the LFPR decline.
  - Cohort-level behavioral findings:
    - Declining LFPR in the 20-24 age cohort was the largest behavioral contributor: it accounted for 0.56 percentage points of the overall 3.71 percentage point decline over 2020-2024.
    - Increased LFPR for workers aged 65 and over offset nearly half the contribution from the behavioral shift among workers aged 20-24.
- Cross-economy SSA insights:
  - For many comparator economies (Australia, Japan, New Zealand), increases in LFPR for workers aged 20-24 partly offset demographic aging.
  - Korea: youth participation declined but its contribution was marginal; overall LFPR rose by 1.2 percentage points in 2024.
  - Singapore: overall LFPR declined by 0.46 percentage points in 2024 relative to 2020; negative contribution from declining participation among young adults was 0.61 percentage points.

### D. Drivers of Declining Youth Labor Force Participation
- Primary proximate driver: delayed entry into the labor force associated with continued education.
  - General Household Survey (C&SD): in 2024, 90 percent of economically inactive persons aged 20–24 cited “attending an educational institution” as the main reason for not participating in the labor market.
  - Comparative figures:
    - 2020: 85.7 percent cited education as main reason.
    - Pre-pandemic average 2000-2019: 86.9 percent.
  - The upward trend in education-related inactivity has been mostly driven by economically inactive females.
  - Growing share of youth enrolled in full-time education supports the conclusion that education-related decisions increasingly shape youth labor supply, particularly among females.
- Structural and technological influences:
  - Extended free education provision (from 9 years to 12 years) and increased post-secondary opportunities raised average years of schooling substantially since the 1990s.
  - Rising demand for high-skill professional occupations and sectoral shift toward finance and professional services increase incentives for youth to pursue further education and possibly postgraduate degrees to adapt to evolving labor market skill demands, including in the context of the emergence of AI.
  - These forces may lead to delays in labor market entry and contribute to lower youth LFPR.

### Rising non-local student enrollments and mechanical LFPR effects
- Rapid increase in the share of non-local student enrollments in Hong Kong SAR’s post-secondary programs has expanded the working-age population with individuals who typically do not participate in the labor market, mechanically lowering the LFPR.
- Text chart indicated post-secondary enrollments growth from 2000 to 2024 (scale shown in thousand), with share components including "Attending educational institutions" among reasons for nonparticipation.

### Sectoral structural shifts and evolving labor skill requirements
- Ongoing structural shifts: relocation and expansion of traditional sectors (retail and consumer services) to neighboring GBA cities (e.g., Shenzhen) and rising employment share in high-value sectors including financial and professional services.
- Youth employment trends:
  - Total youth employment has fallen steadily since the late 2000s and accelerated since the pandemic (noted outbreak during 2020 to 2023).
  - Contraction concentrated in retail, accommodation, and food services—historically largest employers of youth.
  - Youth employment in finance and professional-related sectors has remained relatively stable but small in scale.
  - Occupational distribution shifted away from lower-skill jobs: shrinking shares of craft and related workers, services and sales workers, and clerical support workers.
- Implication: Increasing importance of skills upgrading and continued education as pathways to labor market entry amid transition to higher-value activities.

### Increasing labor market competition and rising educational attainment
- Continued education is the main reason for delayed labor force entry (General Household Survey).
- Educational attainment of employed workers has been rising, particularly within the youth cohort (share of employed workers with post-secondary educational attainment rising over 2008–2024).
- Cohort-based regression setup:
  - Model indexes five-year age groups ranging from 15–19 to 45-49.
  - Cohorts covered: born between 1976 and 2004 (seven 5-year age groups covering individuals aged 15 to 49).
  - Controls included: business cycle (output gap, contemporaneous and two lags), fertility (total fertility rate—for females aged 20 to 39), and educational attainment measured as the share of employed persons aged 25 and above who have ever attained a post-secondary degree.
  - For each gender, a system of seemingly unrelated regressions for seven age groups estimated with identical cohort effects constraint.
  - Output gap estimated by the Hong Kong Monetary Authority (HKMA).
- Regression findings:
  - A one-percentage-point increase in the share of workers aged 25 and above that have ever attained a post-secondary degree is associated with the decline in the relative likelihood of entering the labor force for the 20-25 age group by about 5 basis points for both males and females.
  - Gender differences: impact differs between genders in the 25-30 age group, possibly reflecting household duties and fertility-related considerations among women.
- Counterfactual exercise:
  - Holding educational attainment at its 2010–2021 average from 2022 onward implies that LFPR among the 20-24 age group during 2022–2024 would have been, on average, 5.7 percentage points higher for both males and females.
- Note: Educational attainment is included only in equations for age groups 15-19, 20-24 and 25-29.

### Emergence of AI and exposure heterogeneity
- AI exposure varies across occupations and demographic groups with implications for displacement risk:
  - Younger workers slightly more concentrated in high-exposure, low-complementarity (HELC) occupations, particularly routine service roles—implying elevated displacement risks as AI adoption progresses.
  - Older workers more likely employed in occupations with higher task complementarity.
  - Women show a slightly higher concentration in HELC occupations, suggesting uneven AI-related adjustment costs across genders.
- Occupational classification and methodology:
  - Occupations classified by AI exposure and task complementarity following a two-step mapping (measures from Felten, Raj, and Seamans (2021); complementarity indices from Pizzinelli et al. (2023)).
  - Occupations decomposed into HEHC (high exposure–high complementarity), HELC (high exposure–low complementarity), and LE (low exposure).
  - Mapping uses ISCO one-digit categories and U.S.-based occupational task data due to lack of granular local occupation-level data—acknowledged potential measurement bias.
- Aggregate and cross-economy comparisons:
  - Hong Kong SAR’s employment share in HEHC occupations broadly comparable to some regional peers, but behind AI frontier economies such as Singapore.
- Implications of AI adoption:
  - Potential channels increasing downside risks to youth employment and delayed labor market entry: lower vacancy creation for routine entry-level occupations, heightened job competition, downward pressure on starting salaries, increased reliance on short-hour or temporary contracts in HELC occupations.
  - Frictions to occupational mobility as transitions from HELC to high-complementarity roles require substantial changes in task content and skills.

### Conclusion: key findings
- Sustained decline in Hong Kong SAR’s LFPR driven by:
  - Decline in youth LFPR (education-driven delayed entry).
  - Population aging—the rising weight of older cohorts with structurally low LFPR.
- Education is the predominant reason for delayed labor-force entry among youth, consistent with:
  - Sectoral shifts toward higher-skill activities.
  - Emergence of AI.
  - Intensifying labor-market competition.
- Continued pursuit of further education reduces youth labor supply but can enhance human capital formation and long-term productivity and potential growth.

### Policy recommendations
- Population aging and female labor force participation:
  - Promote longer and healthier working lives through incentives to raise effective retirement ages.
  - Offer flexible and phased retirement options.
  - Implement targeted mid-career reskilling.
  - Expand capacity of overall elderly care services and childcare infrastructure to relieve family carers’ stress.
  - Leverage the Northern Metropolis development to attract younger skilled workers by integrating housing, employment, and innovation ecosystems.
  - Expand access to affordable childcare and more flexible work arrangements to support sustained female labor market attachment during prime working years.
- Skill mismatch:
  - Education and training systems should emphasize market-relevant competencies as demand shifts toward digital, green, professional, and advanced services skills.
  - Provide targeted support for vulnerable youth, particularly low-income and low-skilled groups, to mitigate scarring effects and ease school-to-work transitions.
- AI risk:
  - Accelerate acquisition of AI-complementary skills.
  - Strengthen occupational mobility.
  - Reinforce social protection systems to support workers during transition periods.
  - Enhance unemployment insurance and active labor market policies to improve resilience and enable efficient labor reallocation.

*Source: IMF staff analysis in sipea2026046 (People’s Republic of China—Hong Kong SAR).*

### 3.      This Selected Issues Paper addresses the following three sets of key questions with a

### 3.      This Selected Issues Paper addresses the following three sets of key questions with a focus on youth labor force participation, while recognizing population aging as the main driver of the decline in LFPR: 

### Overview
- Paper addresses three key questions:
  - What are the stylized facts about Hong Kong SAR’s labor market? What are structural trends and cyclical patterns in labor force, employment and LFPR (Section B)?
  - What are the main compositional drivers of Hong Kong SAR’s declining LFPR? Specifically, how much of the decline can be attributed to population aging versus falling youth labor force participation (Section C)? How does Hong Kong SAR’s decomposition of LFPR changes compare to peer economies?
  - What factors explain the recent decline in youth participation (Section D)? What are the structural drivers—such as economic structure transformation, higher labor market educational attainment, and technological change—that contribute to the decline of youth LFPR? Does rising re-education among youth reflect adaptation to evolving skill demands amid the emergence of artificial intelligence and other technological disruptions?
- Main synthesis:
  - Demographic trends, youth education choices, and structural labor market shifts shape LFPR outcomes in Hong Kong SAR.
  - Shift-share analysis shows population aging remains the primary driver of the decline in aggregate LFPR, while a recent fall in youth LFPR also contributed, distinguishing Hong Kong SAR from some peer economies.
  - Household surveys and labor market data show continued education is the main driver of the decline in youth LFPR, linked to sectoral structural shifts, evolving labor skill requirements, emergence of AI, and possibly rising labor market competitiveness from higher educational attainment.
  - The paper concludes with policy implications (policy implications are noted but not detailed in the supplied excerpts).

### B. Stylized Facts of Hong Kong SAR’s Labor Market — Structural and Cyclical Trends
- Structural trends (past three decades):
  - Share of older workers (aged 55 and above) rose from 13 percent in 2010 to 25 percent in 2025Q2.
  - Share with post-secondary education (degree or above) increased from 24 percent in 2010 to 36 percent in 2025Q2.
  - Among youth (aged 15–24), post-secondary attainment rose from 25 percent in 2010 to 52 percent in 2025Q2.
  - Note: Labor force includes foreign domestic helpers.
- Cyclical indicators (pandemic to recovery):
  - Unemployment rate declined to below 4 percent following the COVID reopening, down from a peak of 7 percent in 2021Q1.
  - Hourly wage growth rose to above 3 percent year-on-year from a trough of 0.9 percent during the pandemic peak.
  - Unemployment among those with post-secondary qualifications has trended upward over time and recently converged with that of workers with lower educational attainment.
  - Youth unemployment rate has diverged markedly from other age groups over the past decade, with the gap widening in recent years.
- Compositional shifts in LFPR:
  - LFPR for ages 20–24 fell from above 70 percent in 2005 to slightly below 50 percent in recent years.
  - Share of older individuals (aged 65 and above) rose from around 14 percent of the population aged 15 and above in 2005 to over 26 percent of this population.
  - Participation rate among seniors has increased gradually but remains low relative to other age groups.
- Education and occupation structure:
  - Average years of schooling for individuals aged 15-64 increased significantly since the 1990s; Hong Kong SAR experienced a rise exceeding that in most peer advanced economies (Text Figure 4, panel 1 referenced).
  - Youth participation rates fell across all education groups during 2000-20; for youth with secondary or non-degree post-secondary attainment, LFPR fell by almost 10 percentage points.
  - Post-2000: share of employed workers in manager, professional and associate professional roles has risen since 2005; employment in elementary occupations and clerical support roles has declined.
  - Sectoral employment changes (2025Q4 vs. 2018Q4): employment expanded in professional services and real estate; substantial declines recorded in manufacturing, trade and wholesale, retail, and accommodation and food services.
  - Finance and other professional service sectors account for a growing share of youth employment.

### Cross-Economy Comparison (stylized)
- Hong Kong SAR shares low and decreasing youth LFPR with regional peers Korea and Singapore.
  - Korea: youth LFPR relatively stable but significantly below other advanced economies; large share of youth postpone careers to land jobs in large firms or professional jobs.
  - Singapore: downward trend where education-related delays have contributed to falling youth LFPR.
- Comparison finding: Hong Kong SAR stands out relative to most Asia-Pacific peers (Australia, Japan, Korea, New Zealand and Singapore) in the SSA results, with Singapore the closest comparator in terms of overall LFPR decline.

### C. Shift-Share Analysis (SSA) — Quantifying Demographic vs Behavioral Contributions
- Methodology:
  - Decomposition separates change in LFPR into the extensive margin (demographics: cohort population shares held constant) and the intensive margin (behavioral: within-cohort LFPR changes held constant), plus a residual.
  - Population aged 15 and above broken into ten five-year cohorts from 15–19 through 60–64, and a final 65+ cohort.
  - Analysis focuses on medium-term changes: change in each year relative to 5 years earlier (b = 5).
  - Data source: mid-year population statistics and LFPR data from Hong Kong SAR Census and Statistics Department (C&SD).
- Key quantitative results:
  - Aggregate LFPR declined by 3.7 percentage points in 2024 relative to 2020.
  - Of that 3.7 percentage point decline, about 0.22 percentage points was due to behavioral changes (within-cohort LFPR changes).
  - The demographic impact (shift in population shares) accounts for the bulk of the LFPR decline.
  - Cohort-level behavioral findings:
    - Declining LFPR in the 20-24 age cohort was the largest behavioral contributor: it accounted for 0.56 percentage points of the overall 3.71 percentage point decline over 2020-2024.
    - Increased LFPR for workers aged 65 and over offset nearly half the contribution from the behavioral shift among workers aged 20-24.
- Cross-economy SSA insights:
  - For many comparator economies (Australia, Japan, New Zealand), increases in LFPR for workers aged 20-24 partly offset demographic aging.
  - Korea: youth participation declined but its contribution was marginal; overall LFPR rose by 1.2 percentage points in 2024.
  - Singapore: overall LFPR declined by 0.46 percentage points in 2024 relative to 2020; negative contribution from declining participation among young adults was 0.61 percentage points.

### D. Drivers of Declining Youth Labor Force Participation
- Primary proximate driver: delayed entry into the labor force associated with continued education.
  - General Household Survey (C&SD): in 2024, 90 percent of economically inactive persons aged 20–24 cited “attending an educational institution” as the main reason for not participating in the labor market.
  - Comparative figures:
    - 2020: 85.7 percent cited education as main reason.
    - Pre-pandemic average 2000-2019: 86.9 percent.
  - The upward trend in education-related inactivity has been mostly driven by economically inactive females.
  - Growing share of youth enrolled in full-time education supports the conclusion that education-related decisions increasingly shape youth labor supply, particularly among females.
- Structural and technological influences:
  - Extended free education provision (from 9 years to 12 years) and increased post-secondary opportunities raised average years of schooling substantially since the 1990s.
  - Rising demand for high-skill professional occupations and sectoral shift toward finance and professional services increase incentives for youth to pursue further education and possibly postgraduate degrees to adapt to evolving labor market skill demands, including in the context of the emergence of AI.
  - These forces may lead to delays in labor market entry and contribute to lower youth LFPR.

*Source: Excerpts from the Selected Issues Paper (sipea2026046) provided in the source content.*

### 18.      Besides the incentive for further education,

### 18.      Besides the incentive for further education,

### Rising non-local student enrollments and mechanical LFPR effects
- Rapid increase in the share of non-local student enrollments in Hong Kong SAR’s post-secondary programs has expanded the working-age population with individuals who typically do not participate in the labor market, mechanically lowering the LFPR.
- Text chart (post-secondary enrollments) indicated growth from 2000 to 2024 (scale shown in thousand), with share components including "Attending educational institutions" among reasons for nonparticipation.

### Sectoral structural shifts and evolving labor skill requirements
- Ongoing structural shifts: relocation and expansion of traditional sectors (retail and consumer services) to neighboring GBA cities (e.g., Shenzhen) and rising employment share in high-value sectors including financial and professional services.
- Youth employment trends:
  - Total youth employment has fallen steadily since the late 2000s and accelerated since the pandemic (noted outbreak during 2020 to 2023).
  - Contraction concentrated in retail, accommodation, and food services—historically largest employers of youth.
  - Youth employment in finance and professional-related sectors has remained relatively stable but small in scale.
  - Occupational distribution shifted away from lower-skill jobs: shrinking shares of craft and related workers, services and sales workers, and clerical support workers.
- Implication: Increasing importance of skills upgrading and continued education as pathways to labor market entry amid transition to higher-value activities.

### Increasing labor market competition and rising educational attainment
- Continued education is the main reason for delayed labor force entry (General Household Survey).
- Educational attainment of employed workers has been rising, particularly within the youth cohort (share of employed workers with post-secondary educational attainment rising over 2008–2024 in text chart).
- Cohort-based regression setup:
  - Model indexes five-year age groups ranging from 15–19 to 45-49.
  - Cohorts covered: born between 1976 and 2004 (seven 5-year age groups covering individuals aged 15 to 49).
  - Controls included: business cycle (output gap, contemporaneous and two lags), fertility (total fertility rate—for females aged 20 to 39), and educational attainment measured as the share of employed persons aged 25 and above who have ever attained a post-secondary degree.
  - For each gender, a system of seemingly unrelated regressions for seven age groups estimated with identical cohort effects constraint.
  - Output gap estimated by the Hong Kong Monetary Authority (HKMA).
- Regression findings:
  - A one-percentage-point increase in the share of workers aged 25 and above that have ever attained a post-secondary degree is associated with the decline in the relative likelihood of entering the labor force for the 20-25 age group by about 5 basis points for both males and females.
  - Gender differences: impact differs between genders in the 25-30 age group, possibly reflecting household duties and fertility-related considerations among women.
- Counterfactual exercise:
  - Holding educational attainment at its 2010–2021 average from 2022 onward implies that LFPR among the 20-24 age group during 2022–2024 would have been, on average, 5.7 percentage points higher for both males and females.
- Note: Educational attainment is included only in equations for age groups 15-19, 20-24 and 25-29.

### Emergence of AI and exposure heterogeneity
- AI exposure varies across occupations and demographic groups with implications for displacement risk:
  - Younger workers slightly more concentrated in high-exposure, low-complementarity (HELC) occupations, particularly routine service roles—implying elevated displacement risks as AI adoption progresses.
  - Older workers more likely employed in occupations with higher task complementarity.
  - Women show a slightly higher concentration in HELC occupations, suggesting uneven AI-related adjustment costs across genders.
- Occupational classification and methodology:
  - Occupations classified by AI exposure and task complementarity following a two-step mapping (measures from Felten, Raj, and Seamans (2021); complementarity indices from Pizzinelli et al. (2023)).
  - Occupations decomposed into HEHC (high exposure–high complementarity), HELC (high exposure–low complementarity), and LE (low exposure).
  - Mapping uses ISCO one-digit categories and U.S.-based occupational task data due to lack of granular local occupation-level data—acknowledged potential measurement bias.
- Aggregate and cross-economy comparisons:
  - Hong Kong SAR’s employment share in HEHC occupations broadly comparable to some regional peers, but behind AI frontier economies such as Singapore.
- Implications of AI adoption:
  - Potential channels increasing downside risks to youth employment and delayed labor market entry: lower vacancy creation for routine entry-level occupations, heightened job competition, downward pressure on starting salaries, increased reliance on short-hour or temporary contracts in HELC occupations.
  - Frictions to occupational mobility as transitions from HELC to high-complementarity roles require substantial changes in task content and skills.

### Conclusion: key findings
- Sustained decline in Hong Kong SAR’s LFPR driven by:
  - Decline in youth LFPR (education-driven delayed entry).
  - Population aging—the rising weight of older cohorts with structurally low LFPR.
- Education is the predominant reason for delayed labor-force entry among youth, consistent with:
  - Sectoral shifts toward higher-skill activities.
  - Emergence of AI.
  - Intensifying labor-market competition.
- Continued pursuit of further education reduces youth labor supply but can enhance human capital formation and long-term productivity and potential growth.

### Policy recommendations
- Population aging and female labor force participation:
  - Promote longer and healthier working lives through incentives to raise effective retirement ages.
  - Offer flexible and phased retirement options.
  - Implement targeted mid-career reskilling.
  - Expand capacity of overall elderly care services and childcare infrastructure to relieve family carers’ stress.
  - Leverage the Northern Metropolis development to attract younger skilled workers by integrating housing, employment, and innovation ecosystems.
  - Expand access to affordable childcare and more flexible work arrangements to support sustained female labor market attachment during prime working years.
- Skill mismatch:
  - Education and training systems should emphasize market-relevant competencies as demand shifts toward digital, green, professional, and advanced services skills.
  - Provide targeted support for vulnerable youth, particularly low-income and low-skilled groups, to mitigate scarring effects and ease school-to-work transitions.
- AI risk:
  - Accelerate acquisition of AI-complementary skills.
  - Strengthen occupational mobility.
  - Reinforce social protection systems to support workers during transition periods.
  - Enhance unemployment insurance and active labor market policies to improve resilience and enable efficient labor reallocation.

*Source: IMF staff analysis in sipea2026046 (People’s Republic of China—Hong Kong SAR).*

---


_Source: https://www.imf.org/-/media/files/publications/selected-issues-papers/2026/english/sipea2026046.pdf_
