## sdnea2022001

## Source details

**Canonical URL:** [sdnea2022001](https://www.imf.org/-/media/files/publications/sdn/2022/english/sdnea2022001.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/sdn/2022/english/sdnea2022001.pdf.md)
- [Structured JSON version](/-/media/files/publications/sdn/2022/english/sdnea2022001.pdf.json)

---

### Executive Summary — Overview and Key Findings
- Two years after the onset of the COVID-19 pandemic, employment and other measures of labor market slack have yet to fully recover, yet labor markets are tight in some advanced economies.
- Focus: state of labor markets, drivers of coexisting plentiful vacancies and scarce workers, impacts on wage growth and inequality, and policy implications.

### Executive Summary — Key findings on labor market conditions
- Most labor markets in advanced economies are tighter than pre-COVID-19, particularly in English-speaking countries: Australia, Canada, United Kingdom, United States.
- Tightness evidence:
  - Sharp rise in vacancies and vacancy-to-unemployment ratios.
  - Vacancies rose across all sectors, including contact-intensive, less-teleworkable, and lower-skilled jobs.
- Reduced labor force participation has shrunk the pool of available job seekers.
- Disadvantaged groups lagging:
  - In the median country, low-skilled workers accounted for more than two-thirds of the aggregate employment gap vis-à-vis its pre-COVID-19 trend by late 2021.
  - Older workers—including low-skilled ones—contributed about one-third of that aggregate employment gap.
- Decline in immigration amplified labor shortages among low-pay jobs in some cases.

### Executive Summary — Drivers of worker scarcity despite plentiful vacancies
- Pandemic effects:
  - Health concerns and pension plan valuation gains (amid booming financial markets until late 2021) contributed to older workers’ labor force withdrawal.
  - In the United States, prolonged school closures and scarce childcare opportunities explain some 15 percent of the overall US employment shortfall by the fall of 2021.
- Changing worker preferences:
  - Shift away from low-pay, contact-intensive jobs (“Great Resignation” phenomenon).
  - Voluntary quits rose most in industries with larger shares of contact-intensive, physically strenuous, or less flexible jobs.
- Sectoral and occupational mismatch:
  - By late 2021, rise in mismatch accounted for less than one-tenth of the employment rate gap vis-à-vis pre-COVID-19 levels in the median country.
  - Mismatch spiked at onset of crisis but receded as hard-hit industries recovered.

### Executive Summary — Impact on wages and inequality
- Tightness pushed up wage growth, especially among low-pay jobs, reducing wage inequality in some countries (example: United Kingdom and United States).
  - Tighter labor markets estimated to have raised the annual growth rate of nominal wages in low-pay industries by 4 to 6 percentage points between mid-2020 and late 2021.
- Aggregate impact muted:
  - Estimated at least 1.5 percentage points on annual wage growth in the United Kingdom and the United States between mid-2020 and late 2021.
- Larger inflationary risks emanate from worker demands for compensation for recent price hikes and/or sustained higher inflation expectations.

### Executive Summary — Caveats and scope
- Data limitations: many key issues require granular data available for only a few advanced economies.
- Wage-tightness analysis focused on the United Kingdom and the United States; effects may be different in other advanced economies.

### Executive Summary — Policy implications (summary)
- Further wage increases can attract disadvantaged workers back, but policies should amplify market forces.
- Controlling the pandemic (vaccines, tests, treatments) is paramount.
- Active labor market policies and short-term targeted training can speed up matching.
- Labor laws should facilitate telework.
- Resumption of immigration can ease shortages where relevant.
- Job retention schemes should be mainstreamed for transitory shocks but remain temporary.

### Vacancy-to-Unemployment Ratios — Patterns and explanations
- V/U ratios rose markedly in several advanced economies, "strikingly in Australia, Canada, Luxemburg, Norway, the United Kingdom, and the United States," with Japan, Germany, and Sweden as main exceptions.
- Countries that used job retention schemes saw particularly large increases in tightness (example: the United Kingdom’s Coronavirus Job Retention Scheme "covered almost a third of employment at its peak").
- Sharp rise in tightness correlates positively with the magnitude of fiscal support given to households during the pandemic.
- Increase in tightness:
  - Widely shared across sectors.
  - Slightly stronger in more contact-intensive and less-teleworkable jobs.
  - Particularly large for some low-pay jobs.

### Incomplete employment recovery — levels, cross-country variation, participation
- Employment still recovering to pre-pandemic levels; contact-intensive and less-teleworkable sectors lag but have been catching up since spring 2020.
- Cross-country differences:
  - "Still significant employment shortfalls in Canada, the United Kingdom, or the United States."
  - "Employment is already above pre-pandemic trends in several European countries."
  - Portugal and Sweden "falling significantly behind" pre-pandemic trends.
- In the median country, "the labor force participation rate gap accounts for half of the remaining employment gap vis-à-vis its pre-COVID-19 trend."
- Female participation:
  - By late 2021, female participation generally below a linear projection based on the 2015–19 period, with exceptions such as The Netherlands and Norway.

### Explanations for coexisting tight labor markets and employment slack
- Four broad mechanisms:
  1. Job mismatch (sectoral and occupational mismatch).
  2. Barriers to returning to work (health concerns, childcare, pension gains).
  3. Changing job preferences (toward telework, away from contact-intensive low-pay jobs).
  4. Generous public income support and excess savings — evidence suggests modest role: phasing out of US unemployment supplement had "only a modest effect," excess savings concentrated among higher-income households.

### Sectoral and occupational mismatch — magnitude and comparison to GFC
- Mismatch rose during COVID-19 but generally receded as labor markets recovered.
- Country specifics:
  - Australia: estimated sectoral mismatch "fell below its late 2019 level by the end of 2021."
  - Canada: mismatch "standing 50 percent above its late 2019 level by mid-2021."
  - Japan: mismatch "remained essentially flat."
- Contribution of mismatch to employment shortfalls:
  - Except for Spain, "in all countries mismatch accounted for no more than a fifth of the total employment rate contraction at its peak."
  - By late 2021, mismatch accounted for "about 10 percent in the United Kingdom and the United States, and even less in the median country."
- Compared with the GFC, mismatch rise was smaller during COVID-19.

### Barriers to returning to work and disadvantaged groups
- Low-skilled and older workers experienced largest employment drops and remain below pre-pandemic levels.
- Low-skilled workers accounted for the majority of aggregate employment gaps in the median country.
- No additional employment stigma from prior work in contact-intensive sectors as those sectors recovered.

### Mothers with Young Children — inactivity rate and contributions
- Definition: share of mothers with young children aged 5 or younger who are not in the labor force; trends estimated over 2015–19.
- Timing: values shown are for 2021Q4 for CAN, ESP, GBR, and USA and 2021Q3 for other countries.
- Key findings:
  - School and daycare closures pushed many mothers with young children out of the labor force in several countries (examples: Canada, Spain, United States).
  - By the fall of 2021, elevated inactivity of mothers with young children remained above pre-COVID trend only in the United States.
    - In the United States, this elevated inactivity accounted for some 15 percent of the overall US employment gap vis-à-vis pre-COVID levels.
  - In other countries the “she-cession” largely dissipated by fall 2021.
  - A resurgence in COVID-19 cases (example: late-2021 Omicron wave) could push mothers with young children out of the labor force again.

### Estimated direct contribution of increased tightness to higher wage inflation (2020Q2–2021Q4)
- Approach: industry-level Wage Phillips Curve for the United Kingdom and the United States; regressions allow interaction with low-pay sector dummy.
- Findings:
  - Tighter labor markets benefited low-paid workers and reduced wage inequality, with a manageable direct impact on inflation.
  - Tightness estimated to have raised annual nominal wage growth in low-pay industries by 4 to 6 percentage points between mid-2020 and late 2021.
  - Estimated direct impact on overall wage inflation: at least 1.5 percentage points in the United Kingdom and the United States between mid-2020 and late 2021.
- Key regression results (selected coefficients preserved as reported):
  - United States (Annex 4 selected):
    - t–12 (lagged wage inflation): -0.400*** (0.036) … -0.402*** (0.029)
    - Tightness t–1: 0.836** (0.319) in Column (1); 0.712*** (0.182) in Column (2)
    - (Tightness * low-pay) t–1: 2.514*** (0.747) in Column (2)
    - 1-year-ahead CPI inflation Consensus Forecast t: 2.268*** (0.390)
    - Labor productivity growth t–1: 0.137** (0.052)
    - Observations: 1,037; R-squared: 0.212–0.221; Estimation sample: 2003Q1–2020Q1
  - United Kingdom (Annex 4 selected):
    - t–12 (lagged wage inflation): -0.420*** (0.027) … -0.430*** (0.027)
    - Tightness t–1: 1.967*** (0.555) in Column (1); 1.616** (0.694) in Column (2)
    - (Tightness * low-pay) t–1: 5.744** (2.671) in Column (2)
    - Labor productivity growth t–1: 0.329** (0.118)
    - Observations: 1,188; R-squared: 0.188–0.192; Estimation sample: 2001Q4–2020Q1

### Boxes and empirical evidence highlights
- Box 1 — Vacancies in Low-Wage Jobs:
  - Indeed vacancy data (Jan 2019–Nov 2021) show vacancies in many low-wage occupations were well above pre-pandemic levels by end-2021.
  - Hiring difficulties cited for construction workers, warehouse workers, truck drivers, retail salespersons.
  - Low-wage occupations defined as occupations comprising one-third of total employment in 2019 in the United Kingdom.
- Box 2 — Receding Immigration and Rising Vacancies in Canada and the United Kingdom:
  - Travel restrictions and immigration declines disproportionately affected low-skill occupations.
  - Negative correlation across low-skill occupations between change in foreign worker share and growth of unfilled vacancies vis-à-vis pre-pandemic levels.
- Box 3 — The Great Resignation:
  - Historically high quit rates in the United States and the United Kingdom since spring 2021.
  - In the United States, quit rates by late 2021 were well beyond what the historical relationship with tightness would suggest; not so in the United Kingdom.
  - Industry patterns of largest quit increases reported (United States and United Kingdom top-five lists preserved in the source).

### United States — data, patterns, and methods
- Primary data: Current Population Survey (CPS), Job Openings and Labor Turnover Survey (JOLTS), Bureau of Labor Statistics, IMF staff calculations.
- Coverage: CPS stocks and flows monthly Jan 2000–Oct 2021; JOLTS vacancies/hires/quits Dec 2000–Nov 2021.
- Observed US patterns:
  - High quit rates and elevated V/U ratios in 2021.
  - Top industries with largest quit increases: retail trade; accommodation and food services; nondurable goods manufacturing; durable goods manufacturing; wholesale trade.
- Mismatch analysis: matching-function framework (Şahin and others 2014); mismatch index displayed as 4-quarter moving average and normalized to 2019Q4.
- Regression specifications for low-skilled employment probability and contact-intensive effects described with timing definitions (e.g., “Peak” defined as 2020Q2 for Canada, Spain, and the United States).

### Google Trends and near real-time indicators (Annex 1)
- Google Trends search intensity used for keywords (“jobs,” “telework,” “part-time jobs,” “remote jobs”) normalized to 2019 average (Index = 100).
- Caveat: two-layer normalization by Google Trends; study normalizes to 2019 to improve cross-country comparability.
- Keyword mappings for AUS, CAN, JPN, NLD, ESP, GBR, USA provided in Annex 2.

### Policy recommendations (detailed)
- Active labor market policies:
  - Short-term training programs financed or subsidized by public employment agencies; target professions concentrating shortages and where skills can be acquired quickly.
  - Larger returns if targeted at low-skilled youth.
- Labor laws and telework:
  - Adapt regulations to facilitate telework/remote work access, balancing working hours flexibility, definition of office hours, and cost of home-work resources.
  - Enhance telework access especially for lower-skilled workers who favor working from home for preferences or health reasons.
- Fight the pandemic to bring disadvantaged workers back:
  - vaccines, tests, and treatments remain essential.
  - support for schools and daycare to stay open; consider greater public childcare spending and subsidies, and paid family leave.
  - reduce gender-based tax and regulatory barriers to female labor force participation as part of a broader package.
- Immigration:
  - Resume or tailor immigration to ease shortages where appropriate; tailor entry criteria to existing labor shortages where systems are employment- and skill-based.
  - Provide access to training programs for immigrants who became non- or under-employed during the pandemic.
- Job retention schemes:
  - Mainstream job retention schemes after major temporary shocks to damp initial job losses and labor force withdrawals.
  - Ensure support is temporary to avoid impeding reallocation when shocks prove persistent.

*Staff Discussion Note: Labor Market Tightness in Advanced Economies (sdnea2022001).*

### Executive Summary ......................................................................................................

### Executive Summary

### Overview
- Two years after the onset of the COVID-19 pandemic, employment and other measures of labor market slack have yet to fully recover, yet labor markets are tight in some advanced economies.
- The note focuses on a selected group of countries and addresses: the state of labor markets, drivers of coexisting plentiful vacancies and scarce workers, impacts on wage growth and inequality, and policy implications.

### Key findings on labor market conditions
- Most labor markets in advanced economies are tighter than they were prior to COVID-19, particularly in English-speaking countries: Australia, Canada, United Kingdom, United States.
- Tightness is visible in a sharp rise in vacancies and vacancy-to-unemployment ratios.
- Vacancies have risen across all sectors, including contact-intensive, less-teleworkable, and lower-skilled jobs that were hit hard by the pandemic; fears that COVID-19 might permanently destroy these jobs, including via automation, have not yet materialized.
- Tight labor markets partly reflect reduced labor force participation, which has shrunk the pool of available job seekers and made it harder to fill vacancies.
- Disadvantaged groups—including the low-skilled, older workers, and women with young children, depending on the countries—have yet to fully return to the labor market.
  - In the median country, low-skilled workers accounted for more than two-thirds of the aggregate employment gap vis-à-vis its pre-COVID-19 trend by late 2021.
  - Older workers—including low-skilled ones—contributed about one-third of that aggregate employment gap.
- In some cases, a decline in immigration also amplified labor shortages among low-pay jobs.

### Drivers of worker scarcity despite plentiful vacancies
- The pandemic itself partly explains why some workers have not fully come back:
  - Health concerns and, in some cases, pension plan valuation gains amid booming financial markets until late 2021 contributed to older workers’ labor force withdrawal.
  - In the United States, prolonged school closures and scarce childcare opportunities have kept women with young children home, explaining some 15 percent of the overall US employment shortfall by the fall of 2021.
- Signs of changing worker preferences:
  - Some workers have shifted away from low-pay, contact-intensive jobs (a phenomenon sometimes labeled the “Great Resignation”).
  - A growing share of workers with contact-intensive jobs have been moving into other jobs or left the labor force altogether, despite rising contact-intensive job vacancies.
  - Voluntary quits have risen most in industries with larger shares of contact-intensive, physically strenuous, or less flexible jobs, producing pockets of labor shortages.
- Sectoral and occupational mismatch played a role but less than after the 2008–09 global financial crisis (GFC):
  - By late 2021, the rise in mismatch accounted for less than one-tenth of the employment rate gap vis-à-vis pre-COVID-19 levels in the median country.
  - In most countries, mismatch spiked at onset of the crisis but receded as hard-hit industries recovered and rehired previously laid-off workers.

### Impact on wages and inequality
- Tightness has pushed up wage growth, especially among low-pay jobs, helping reduce wage inequality in some countries (for example, the United Kingdom and the United States).
  - Wages are more responsive to tightness in low-pay industries; these industries have also seen larger increases in tightness.
  - Tighter labor markets are estimated to have raised the annual growth rate of nominal wages in low-pay industries by 4 to 6 percentage points between mid-2020 and late 2021.
- The overall impact of increased tightness on economy-wide wage inflation has been more muted so far:
  - The analysis estimates at least 1.5 percentage points on annual wage growth in the United Kingdom and the United States between mid-2020 and late 2021, partly because the small overall share of low-pay jobs in firms’ total labor costs limits aggregate effects.
- Larger inflationary risks would stem from worker demands for compensation for recent price hikes and/or sustained higher inflation expectations than from tight labor markets per se.

### Caveats on evidence and scope
- Data limitations: many key issues require granular data available for only a few advanced economies.
- The impact of labor market tightness on wage growth is explored for the United Kingdom and the United States; effects might be different—possibly weaker—in other advanced economies (for example, continental Europe) with more stringent labor regulations, higher minimum wages, and more centralized collective bargaining.

### Policy implications and recommendations
- Further wage increases can help by attracting more disadvantaged workers back into the labor force, but policies need to amplify market forces.
- Controlling the pandemic (vaccines, tests, treatments) is paramount to enable seniors, workers in contact-intensive industries, and women with young children (in the United States) to rejoin the labor force fully.
- Active labor market policies:
  - Short-term training programs targeted at professions that concentrate shortages can help speed up matching.
  - Well-designed programs could help detached (and employed) lower-skilled workers build skills required for new fast-growing occupations or traditional occupations with acute shortages.
- Labor laws and regulations should facilitate telework to accommodate shifting worker preferences.
- Resumption of immigration can help “grease the wheels” of the labor market where reduced immigration has amplified shortages.
- Mainstreaming job retention schemes after major temporary shocks can help keep disadvantaged workers attached to the labor force; job retention support should remain temporary so as not to impede eventual worker reallocation after persistent shocks.

*Staff Discussion Note: Labor Market Tightness in Advanced Economies (Executive Summary).*

### 1.  Vacancy-to-Unemployment Ratios

### sdnea2022001 - 1.  Vacancy-to-Unemployment Ratios

### Labor market tightness: vacancy-to-unemployment ratios and patterns
- Vacancy-to-unemployment (V/U) ratios rose markedly in several advanced economies, "strikingly in Australia, Canada, Luxemburg, Norway, the United Kingdom, and the United States," with Japan, Germany, and Sweden as main exceptions.
- Countries that protected existing jobs through extensive job retention schemes saw particularly large increases in tightness (example: the United Kingdom’s Coronavirus Job Retention Scheme "covered almost a third of employment at its peak").
- Other countries relied more on extended unemployment benefits (example: the United States), although "the Paycheck Protection Program also saved jobs" there.
- The sharp rise in labor-market tightness correlates positively with the magnitude of fiscal support given to households during the pandemic.
- Within countries, the increase in tightness:
  - Has been widely shared across sectors.
  - Has been "slightly stronger in those with more contact-intensive and less-teleworkable jobs."
  - Has been "particularly large for some low-pay jobs."

### Incomplete employment recovery: levels, cross-country variation, and participation
- Employment across all sectors remains in the process of recovering to pre-pandemic levels; hard-hit contact-intensive and less-teleworkable sectors are still lagging but have been catching up since the spring 2020 trough.
- Cross-country differences:
  - "Still significant employment shortfalls in Canada, the United Kingdom, or the United States."
  - "Employment is already above pre-pandemic trends in several European countries."
  - Some countries (example: Portugal and Sweden) "falling significantly behind" their pre-pandemic trends.
- Regional pattern: Employment fell less in Europe in the early phase of the crisis, in part due to extensive reliance on job retention schemes.
- Labor force participation shortfalls have accompanied employment gaps in many countries (example: the United Kingdom and the United States), though tight labor markets are bringing workers back in some other countries (example: Norway and The Netherlands).
- In the median country in the sample, "the labor force participation rate gap accounts for half of the remaining employment gap vis-à-vis its pre-COVID-19 trend."
- Female participation and employment:
  - Employment has not fully recovered for women in the United States and "in fact in most other advanced economies as well once factoring in that female participation and employment rates would have likely increased in the absence of COVID-19."
  - By late 2021, female participation was "still generally below a linear projection based on the 2015–19 period," with exceptions such as The Netherlands and Norway.

### Explanations for coexisting tight labor markets and employment slack
- Three broad types of explanations are highlighted:
  1. Job mismatch (sectoral and occupational mismatch).
  2. Barriers to returning to work.
  3. Changing job preferences among workers.
- A fourth factor considered in policy discussions is the role of generous public income support and excess savings during the pandemic on non-employed workers’ willingness to seek and take up jobs; evidence summarized:
  - Phasing out of the US federal unemployment insurance supplement in the second half of 2021 had "only a modest effect on getting people back to work" (Coombs and others 2021).
  - Excess savings were temporary and "tended to be concentrated among higher-income households," making them unlikely to durably suppress labor supply, particularly among low-income workers.
  - Excess savings may have enabled some workers to take time off job search to find better opportunities, but are unlikely to explain a persistent labor supply shortfall among low-income workers.

### Sectoral and occupational mismatch: evidence and magnitude
- COVID-19 could increase mismatch through structural transformation (some industries shrink while others expand or can telework).
- Empirical findings:
  - Sectoral mismatch rose during the COVID-19 crisis in several countries but "generally receded as labor markets recovered."
  - In Australia, Spain, the United Kingdom, and the United States, mismatch "rose markedly throughout 2020" consistent with heterogeneous sectoral impacts.
  - As hard-hit industries recovered and laid-off workers transitioned to new jobs, mismatch generally receded.
  - Australia’s estimated sectoral mismatch "fell below its late 2019 level by the end of 2021."
  - Canada’s mismatch "has been rising gradually since late 2020, standing 50 percent above its late 2019 level by mid-2021."
  - In Japan mismatch "remained essentially flat."
- Contribution of mismatch to employment shortfalls:
  - With the exception of Spain, "in all countries mismatch accounted for no more than a fifth of the total employment rate contraction at its peak."
  - By late 2021, mismatch accounted for "about 10 percent in the United Kingdom and the United States, and even less in the median country."
- Comparison with the GFC:
  - The estimated rise in mismatch during COVID-19 has been smaller than during the GFC.
  - The GFC triggered long-lasting contractions and reallocation away from industries such as construction and manufacturing, whereas pandemic-hit industries recovered sharply and job retention schemes reduced total displaced workers.

### Barriers to returning to work and disadvantaged groups
- Disadvantaged groups remain scarred and are shrinking the pool of available job seekers, contributing to labor market tightness.
- Demographic impacts:
  - Low-skilled and older workers experienced the sharpest employment rate drops during the crisis and "are still showing sizeable gaps vis-à-vis pre-pandemic levels."
  - Low-skilled workers were "hit particularly hard at the onset of the COVID-19 pandemic," amplified in some countries by over-representation in contact-intensive sectors.
  - Although employment rates for low-skilled workers have risen steadily, they remain below pre-COVID-19 levels.
- The analysis notes that there is "no longer any additional employment stigma from being previously employed in contact-intensive sectors," partly because those sectors have been recovering.

*Source: STAFF DISCUSSION NOTES Labor Market Tightness in Advanced Economies, INTERNATIONAL MONETARY FUND*

### 4.  Mothers with Young Children: Inactivity Rate

### 4.  Mothers with Young Children: Inactivity Rate

### Definition and measurement
- Inactivity rate of mothers with young children is defined as the share of mothers with young children aged 5 or younger who are not in the labor force (inactive), while trends are estimated over 2015–19.
- Values shown in the underlying panels are for 2021Q4 for CAN, ESP, GBR, and USA and 2021Q3 for the other countries.
- Sources: Current Population Survey, Job Openings and Labor Turnover Survey (USA); Encuesta de Población Activa; Instituto Nacional de Estadística (ESP); Eurostat (EUR); Indeed (CAN); International Labour Organization (CAN, USA); Labour Force Survey, Office of National Statistics (GBR); Organisation for Economic Co-operation and Development (CAN, GBR, USA); Statistics Canada; and IMF staff calculations.

### Key findings on mothers with young children
- School and daycare closures at the onset of the COVID-19 recession likely pushed many mothers with young children out of the labor force in several countries (examples cited include Canada, Spain, and the United States).
- By the fall of 2021, the elevated inactivity rate of mothers with young children remained above the pre-COVID trend only in the United States.
  - In the United States, this elevated inactivity among mothers with young children accounted for some 15 percent of the overall US employment gap vis-à-vis pre-COVID levels.
- In other countries where mothers with young children struggled initially, the “she-cession” had largely dissipated by the fall of 2021.
- A resurgence in COVID-19 cases (for example, the late-2021 Omicron wave) could push mothers with young children out of the labor force again.

### Context and related heterogeneity
- The initial labor force withdrawal of mothers with young children is one element of broader heterogeneous labor market effects from COVID-19 documented in the chapter, which also highlights:
  - Large contributions of low-skilled workers and older workers to aggregate employment gaps.
  - Differential recovery dynamics across countries and demographic groups.
- Panel timing and definitions matter for cross-country comparisons: the chapter notes differences in the quarters reported across countries (2021Q4 versus 2021Q3).

*Source: IMF staff discussion notes, “Labor Market Tightness in Advanced Economies,” chapter 4: Mothers with Young Children: Inactivity Rate.*

### 1.  Estimated Direct Contribution of Increased

### Estimated Direct Contribution of Increased Tightness to Higher Wage Inflation during COVID-19

### Estimated Direct Contribution to Wage Inflation (2020Q2–2021Q4)
- The analysis reports the estimated percentage point change in nominal hourly (year-over-year) wage inflation between 2020 Q2 and 2021Q4 attributable to the increase in labor market tightness over the same period.
- Regression specifications:
  - The first bar is from a specification that does not allow for different slope coefficients between low-pay and other industries.
  - All other bars allow for interaction effects between tightness and industry groups.
  - The last bar reports the payroll-weighted impact and is derived as the weighted average of the direct impact of tightness on wage growth for low-pay and other industries, respectively, using income shares as weights.
- Sources used: Current Population Survey, Job Openings and Labor Turnover Survey, Bureau of Labor Statistics, Federal Reserve Bank of Atlanta (USA); and IMF staff calculations.

### Wage Growth by Average Wage Quartile (12-month moving average)
- Panel 2 plots 12-month moving averages of monthly hourly wage growth from the Wage Growth Tracker of the Federal Reserve Bank of Atlanta for two categories:
  - the median worker of the average wage distribution (green line)
  - the median worker of the lowest 25th percentile of the average wage distribution (blue line)
- The average wage distribution is computed based on the average of hourly wages in months t and t – 12.

### Key Findings on Labor Markets, Wages, and Inflation
- Tighter labor markets have benefited low-paid workers and reduced wage inequality, with a manageable direct impact on inflation.
- More important for inflation dynamics going forward:
  - the extent to which recent inflation feeds into wage growth
  - the extent to which higher wages are then passed onto prices
- Price inflation has largely or (more than) fully outpaced wage increases so far in several advanced economies, even where labor markets are tightest; this poses a material risk.
- Demands for compensation for recent price hikes and/or sustained higher inflation expectations among workers could trigger larger second-round effects and potentially wage-price spirals.
- Central banks should continue to signal strong determination to alleviate second-round effects from recent high inflation.
- There is high uncertainty; wage growth and inflation could be higher than expected.

### Policy Recommendations (Labor Market and Structural Policies)
- Use active labor market policies to:
  - bring more people back into the labor force
  - fill available vacancies
  - help workers obtain preferred jobs
- Targeted short-term training programs:
  - financed or subsidized by public employment agencies
  - targeted at professions that concentrate shortages and where required skills can be realistically acquired over a short timeframe
  - returns may be larger if targeting low-skilled youth
- Adapt labor laws and regulations to facilitate telework/remote work access:
  - strike the right balance between working hours flexibility, definition of office hours, and cost of home-work resources
  - enhance telework access especially for lower-skilled workers who favor working from home due to preferences or health reasons
- Fight the pandemic to bring disadvantaged workers back (older workers, women with young children):
  - vaccines, tests, and treatments remain essential
  - support for schools and daycare to stay open to help mothers of young children re-enter the labor force
  - consider greater public childcare spending and subsidies, and paid family leave
  - as part of a broader package, reduce gender-based tax and regulatory barriers to female labor force participation
- Resume or tailor immigration to ease shortages where appropriate:
  - effects vary across countries depending on pre- and during-pandemic immigration declines and the match between unfilled jobs and immigrants’ profiles
  - countries with predominantly employment- and skill-based immigration systems may wish to tailor entry criteria to existing labor shortages
  - provide access to training programs for immigrants who became non- or under-employed during the pandemic
- Mainstream job retention schemes for future transitory shocks:
  - can damp initial job losses, labor force withdrawals, and subsequent increases in tightness
  - ensure job retention support is temporary to avoid impeding reallocation when shocks prove persistent
  - currently relevant as countries transition away from such schemes two years into the pandemic

### Box 1 — Vacancies in Low-Wage Jobs
- Observation: Low-skilled workers remain less likely to be employed than pre-crisis in many advanced economies, but job vacancies in low-wage occupations sharply recovered in several countries and were well above pre-pandemic levels by end-2021.
- Hiring difficulties reported for specific professions: construction workers, warehouse workers, truck drivers, retail salespersons.
- Data and methodology:
  - Vacancy data from Indeed for Australia, Canada, the United Kingdom, and the United States, starting January 2019.
  - Several millions of job posts per country between January 2019 and November 2021 (approximately 100 million for the United States).
  - Postings categorized to 4-digit ISCO-08 via matching algorithms; average number of new job posts in the first four weeks of each month computed for each occupation.
- Figure 1.1 details:
  - grey areas report the 10th–90th percentile range of the distribution of vacancies by occupation
  - the dashed black line reports the 90th percentile of the distribution for low-wage occupations (defined as occupations with the lowest hourly wages, comprising one-third of total employment in 2019 in the United Kingdom)
  - individual series plotted: shop sales assistants, construction laborers, warehouse workers, truck drivers
  - red vertical lines mark months when the COVID-19 pandemic approximately reached a peak in new daily deaths
- Findings:
  - vacancies for almost all occupations fell sharply during the first COVID-19 wave then began steep recovery from mid-2020
  - recovery was broad-based, with lower percentiles returning to pre-pandemic levels by 2021
  - low-wage occupations experienced a larger fraction with steep rises in vacancies compared to medium- and high-paying ones
  - construction laborers, warehouse workers, and truck drivers were often in the top 10 percent of occupations with the largest vacancy increases relative to pre-pandemic levels
  - vacancies for shop sales assistants were consistently in the bottom quarter of the distribution in Canada, the United Kingdom, and the United States, reflecting likely structural shifts (lower in-person demand, growth in e-commerce)
- Annex 2 documents matching methodology and definition of low-wage occupations.
- Box prepared by Carlo Pizzinelli and Alessandra Sozzi.

### Box 2 — Receding Immigration and Rising Vacancies in Canada and the United Kingdom
- Context: Travel restrictions and changes to immigration policies during the pandemic limited worker mobility, potentially amplifying labor market tightness among low-pay jobs in Canada and the United Kingdom.
- Key cross-country differences:
  - United Kingdom: foreign workers concentrated in occupations with very high and very low shares of college workers.
  - Canada: relationship between skills and immigration is flat; several low-skill occupations are predominantly covered by domestic workers.
- Empirical patterns:
  - the fall in immigration during the pandemic affected predominantly low-skill occupations
  - occupations with fewer university graduates saw falling shares of foreign workers between 2019 and 2021, while occupations with more university-educated workers saw rising shares of immigrants
  - negative correlation across low-skill occupations between change in foreign worker share and growth of unfilled vacancies vis-à-vis pre-pandemic levels; positive correlation for high-skill jobs
  - suggests reduction in immigration contributed to low-pay job shortages
- Figures and notes:
  - Figure 2.1 plots shares of college workers and foreign workers in each occupation in 2019 for Canada and the United Kingdom; solid lines report quadratic polynomial fits
  - Figure 2.2 plots change in share of foreign workers against percent growth in vacancies since 2019; change in foreign worker share computed on annual averages using first three quarters of 2021; growth rate of vacancies computed using 2021Q3 against average of 2019
  - for each country, occupations with the lowest share of university-educated workers in 2019 comprising 50 percent of employment are classified as low-skill
- Box prepared by Longji Li and Carlo Pizzinelli.
- Annex 5 provides data source and occupational category details.

### Box 3 — The Great Resignation
- Phenomenon: Historically high job quit rates in the United States and the United Kingdom since spring 2021, coinciding with sharply increased labor market tightness.
- Interpretation possibilities:
  - growing job dissatisfaction leading workers to leave unsatisfactory jobs (remuneration, flexibility, safety, purpose)
  - workers taking advantage of hot labor market to seize better opportunities — quits could reflect a well-functioning labor market enabling upward mobility
- Empirical relationship:
  - quit rates typically rise with labor market tightness because workers find better jobs more easily
  - by late 2021, quit rates in the United States were well beyond what the historical relationship with tightness would suggest; not so in the United Kingdom
  - “Excess quits” may indicate strengthening worker bargaining power and increased reservation wages
- Industry patterns in quit increases (relative to 2019):
  - United States top five industries with largest quit increases: (1) retail trade, (2) accommodation and food services, (3) nondurable goods manufacturing, (4) durable goods manufacturing, (5) wholesale trade
  - United Kingdom top five industries: (1) accommodation and food services, (2) real estate, (3) transportation and storage, (4) wholesale and retail trade, (5) construction
  - quit rates rose least in less contact-intensive and/or high-pay industries (examples cited: finance and insurance; educational services; public administration and defense; mining and quarrying)
- Figures:
  - Figure 3.1 shows employment quit rate (percent) and V/U ratio (level) for United States and United Kingdom
  - Figure 3.2 compares historical relationship between tightness and job quits

*Staff Discussion Notes — Labor Market Tightness in Advanced Economies; INTERNATIONAL MONETARY FUND*

### 1.  United States

### sdnea2022001 - 1.  United States

### Data sources and sample coverage
- Aggregate and sectoral indicators draw on: Current Population Survey (CPS), Job Openings and Labor Turnover Survey (JOLTS), Bureau of Labor Statistics (USA), and IMF staff calculations.
- Latest observation for descriptive charts in Sections I and II is 2021Q3 for most countries in the broader note.
- For the United States specifically:
  - CPS: monthly nationally representative survey; used to compute stocks (employed, unemployed, not in labor force) and flows at monthly frequency between January 2000 and October 2021.
  - JOLTS: used for vacancies (levels), hires (levels), and quit rates for aggregate US economy and 17 industries (NAICS) between December 2000 and November 2021.
- Prepared by Carlo Pizzinelli and Ippei Shibata.

### Job quits, labor market tightness, and industry patterns
- Visualizations (Figure 3.3) compare aggregate job quit rates (y-axis) with aggregate labor market tightness measured by vacancy-to-unemployment ratios (x-axis) for the United States at monthly frequency; a “Linear” trend shows the historical linear relationship for the pre-pandemic period.
- Figure 3.3 reports the top five industries with the biggest rise in job quits during the pandemic as percentage point change relative to 2019 average; for the United States the left panel shows monthly industry-level changes between the 2019 average and December 2021.

### Microdata sources and country-specific notes (US context)
- US analysis uses CPS and JOLTS as primary micro and vacancy data sources.
- Industry classification: 17 industries based on NAICS for vacancy and quit rate series.
- Timing coverage for US series: stocks and flows monthly Jan 2000–Oct 2021; vacancies/hires/quits Dec 2000–Nov 2021.

### Sectoral classification and definitions
- Sectors are classified into “More Contact-Intensive” and “More Teleworkable” based on ISIC Rev.4 mapping (Annex 1 Table 1.2). Examples:
  - Trade (Wholesale and retail trade; repair of motor vehicles and motorcycles) — Trade — More Contact-Intensive.
  - Transport (Transportation and storage) — Transport — More Contact-Intensive.
  - Acc./Food (Accommodation and food service activities) — Acc./Food — More Contact-Intensive.
  - Info./Con. (Information and communication) — marked as More Teleworkable.
  - Fin./Ins. (Financial and insurance activities) — marked as More Teleworkable.
  - Educ. (Education) and Health (Human health and social work activities) marked as both More Contact-Intensive and More Teleworkable where indicated.
- Aggregated sector groupings (examples shown): Min./Ener. (B, D & E), Utilities (D & E), Trade/Acc. (G, H & I), Pro. Serv. (M & N), Edu./Heal. (O, P & Q), Arts/Serv. (R & S).

### Mismatch analysis (method and measures)
- Approach follows the matching-function framework by Şahin and others (2014), assuming matching function h_it = φ m(v_it, u_it) = φ v_it^η u_it^(1−η), with η ∈ (0,1) and φ the matching efficiency.
- Mismatch M_t defined as percent shortfall in total hires relative to the optimal allocation that equalizes vacancies-to-unemployment ratios across sectors:
  - M_t = 1 − (h_t / h_t^* ) = 1 − [φ / φ^*] Σ ( (v_it / v_t)^η (u_it / u_t)^(1−η) ).
- The mismatch index is displayed as a 4-quarter moving average and normalized to its value in the fourth quarter of 2019 for cross-country comparison.
- Counterfactual series constructed for job finding rate f_t^*, employment rate e_t^*, and unemployment rate u_t^* by iterating from historical e_0 and u_0 using historical separation rates and:
  - f_t^* = f_t / (1 − M_t)^(1/(1−η))   (expressed in the note as f_t^* = f_t (1−M_t)^(−1/(1−η)) in the functional form shown).

### Barriers to returning to work — empirical specifications and samples
- Low-skilled employment probability regressions use microdata from Canada, Spain, the United Kingdom, and the United States; sample restricted to low-skilled workers (below college degree).
- Basic linear probability model:
  - y_it = α + γ D_post + ε_it,
    - where y_it = 1 if employed, D_post = 1 if observation is in post period (peak or latest).
  - “Peak” defined as 2020Q2 for Canada, Spain, and the United States; 2020Q4 for the United Kingdom.
  - “Latest” quarter: 2020Q3 for Spain and the United Kingdom; 2020Q4 for Canada and the United States.
  - Pre-COVID quarter defined as same quarter one year before “peak” and two years before “latest” to avoid seasonality.
  - Standard errors clustered at period (quarterly for Spain and the United Kingdom; monthly for Canada and the United States).
  - Figure 5 panel 1 plots γ × 100 as the percentage point change in employment probability for low-skilled workers between pre-COVID19 and post periods.
- Additional specification to capture contact-intensive sector effect:
  - y_it = α + β ConInt_it + θ D_post + γ (D_post × ConInt_it) + ε_it,
    - ConInt_it = 1 if worker in contact-intensive industry.
  - “Low-Skilled & Contact-Intensive” in Figure 5 panel 2 plots γ × 100 as the additional impact on employment probability for low-skilled workers in contact-intensive sectors.

### Observed demographic patterns: inactivity of elderly and mothers with young children
- Using labor force survey data from Canada, Spain, the United Kingdom, and the United States:
  - Inactivity (out of labor force) rates plotted for women with young children (children age five or younger) and for elderly (age 55–74).
  - Rates calculated as share of subgroup population.
  - Trend shown as linear trend based on 2015–19 data.

### Shifting worker preferences and transitions
- Transitions away from contact-intensive jobs:
  - Data from Australia, Canada, Spain, the United Kingdom, and the United States are used; industry-to-industry transitions requiring industry info for both t−1 and t are computed for Australia, the United Kingdom, and the United States.
  - Share of transitions from contact-intensive jobs to other jobs = (EE flows where worker moves from contact-intensive industry in t−1 to a different industry in t) / (total EE flows that change industries).
  - Share of transitions from contact-intensive jobs to out of the labor force = (separations from contact-intensive sector to out of labor force) / (all separations to out of labor force).
  - Statistics computed for 2015–19, 2020, and 2021 averages; percentage point differences computed across periods.
- Vacancy-to-employment ratios in contact-intensive sectors:
  - Computed as total vacancies in contact-intensive industries divided by total employment in contact-intensive industries for 2020 and 2021.
  - Percentage point deviation calculated relative to the 2015–19 average.

*Source: IMF staff compilation.*

### Annex 1 Table 1.1. The 2020 value for Canada is missing due to the unavailability of sectoral vacancy data for

### sdnea2022001 - Annex 1 Table 1.1. The 2020 value for Canada is missing due to the unavailability of sectoral vacancy data for

### Google Trends: Search intensity for job and telework
- Purpose: Use Google Trends as a near real-time indicator of worker preference shifts and job-search behavior (examples cited: Woloszko (2020); Borup and Montes Schütte (2020)).
- Data characteristics:
  - Google Trends provides search volume intensity on a 0–100 scale (number of searches for specific keywords over total searches).
  - Data selectable by keyword, frequency (month/week/day), and geography; accessible directly or via APIs such as PyTrends.
- Keywords used:
  - Search intensities collected for “jobs,” “telework,” “part-time jobs,” and “remote jobs” in local language for each country (see Annex 2 Table 1.1).
  - Examples: “jobs” used for Australia, Canada, the United Kingdom, and the United States; “仕事” for Japan; “baan” for The Netherlands; “empleo” for Spain.
- Normalization:
  - Each keyword’s search intensity normalized to its 2019 average (Index = 100 in 2019 average) to compare countries and recent trends.
  - Panels plotted: (1) Part-Time Job Search Intensity (Index = 100 in 2019 average); (2) Remote Job Search Intensity (Index = 100 in 2019 average).
- Caveat on interpretability:
  - Google Trends applies two normalization layers: (1) rescale searches for a keyword over total website search volume; (2) rescale based on each keyword’s historical trend and label its peak as 100.
  - The note normalizes to 2019 average to mitigate these effects for cross-country and temporal comparisons.

### Annex 2: Keywords for jobs and telework (selected mapping)
- Countries and keywords as presented in Annex 2 Table 1.1 (columns shown as AUS, CAN, JPN, NLD, ESP, GBR, USA):
  - Jobs: AUS "jobs"; CAN "jobs"; JPN "仕事"; NLD "baan"; ESP "empleo"; GBR "jobs"; USA "jobs"
  - Telework: AUS "telework"; CAN "telework"; JPN "テレワーク"; NLD "telewerken"; ESP "teletrabajo"; GBR "telework"; USA "telework"
  - Part-time jobs: AUS "part-time jobs"; CAN "part-time jobs"; JPN "アルバイト"; NLD "parttime banen"; ESP "trabajo a tiempo parcial"; GBR "part-time jobs"; USA "part-time jobs"
  - Remote jobs: AUS "remote jobs"; CAN "remote jobs"; JPN "リモートワーク"; NLD "werken op afstand"; ESP "trabajo remoto"; GBR "remote jobs"; USA "remote jobs"

### Annex 3: Wage Phillips Curve (WPC) conceptual framework
- Model specification (industry-level WPC estimated separately for the United Kingdom and the United States):
  - π_{i,t+s} = c + β π_{i,t–1+s} + γ θ_{i,t–1} + δ π_{t}^e + ζ prod_{t–1} + η_i + ε_{i,t}
  - Definitions:
    - π_{i,t+s}: nominal hourly (yoy) wage inflation in industry i
    - θ_{i,t–1}: labor market tightness in industry i (preferred measure: V/U ratio)
    - π_{t}^e: economy-wide inflation expectations
    - prod_{t–1}: aggregate labor productivity growth (output per hour)
    - η_i: industry fixed effects
- Robustness checks:
  - Inflation expectations proxies: (1) one-year-ahead and five-years-ahead Consensus Economics forecasts on CPI inflation; (2) past realized price inflation (four-quarter moving average of headline and core CPI); (3) linear combinations of forward- and backward-looking measures.
  - Use of (pseudo) unemployment rates at the industry level as alternative slack measure; results comparable.
  - Estimation methods robustness: ordinary least squares and Arellano-Bond for lagged dependent variable bias.
- Non-linearity:
  - Some evidence of mild non-linearity: tightness coefficient (γ) tends to be larger at higher tightness levels, but not large enough to materially affect main quantitative estimates.

### Heterogeneous impact: low-pay vs other sectors
- Extended model allows tightness impact to differ for low-pay sectors:
  - π_{i,t+s} = c + β π_{i,t–1+s} + γ_1 θ_{i,t–1} + γ_2 lp_i * θ_{i,t–1} + δ π_{t}^e + ζ prod_{t–1} + η_i + ε_{i,t}
  - lp_i: dummy for low-pay sectors (defined as sectors where employed individuals ranked among the lowest 25 to 30 percent of average wages in 2019).
- Low-pay sector definitions:
  - United States: accommodation and food services; retail trade; arts and entertainment.
  - United Kingdom: accommodation and food services; wholesale and retail and repair of vehicles services; arts and entertainment; administrative and support services.
- Empirical findings:
  - Higher tightness associated with higher wage growth.
  - Association relatively stronger for low-pay sectors.
  - Results robust when using pseudo unemployment rate and alternative estimation methods.
  - Statistical significance of tightness coefficient robust over full sample (United States: early 2000s–2021Q4; United Kingdom: early 2000s–2021Q3), with some evidence of strengthened impact in the United States during the pandemic.

### Annex 4: Wage Phillips Curve estimation — United States (selected coefficients and statistics)
- Dependent variable: Nominal hourly (percent, year-over-year) wage inflation
- Key coefficient estimates (columns (1)–(4)):
  - t–12 (lagged wage inflation): -0.400*** (0.036); -0.407*** (0.029); -0.397*** (0.035); -0.402*** (0.029)
  - Tightness t–1: 0.836** (0.319) in Column (1); 0.712*** (0.182) in Column (2)
  - (Tightness * low-pay) t–1: 2.514*** (0.747) in Column (2)
  - 1-year-ahead CPI inflation Consensus Forecast t: 2.268*** (0.390); 2.163*** (0.386); 1.828*** (0.595); 1.808*** (0.432)
  - Labor productivity growth t–1: 0.137** (0.052); 0.137*** (0.040); 0.141** (0.052); 0.139*** (0.041)
  - Unemployment rate (%) t–1: -0.206** (0.080) in Column (3); -0.165*** (0.052) in Column (4)
  - (Unemployment rate(%) * low-pay) t–1: -0.259** (0.113) in Column (4)
- Constant terms:
  - Column (1): -2.407*** (0.814)
  - Column (2): -2.303*** (0.813)
  - Column (3): 0.324 (1.706)
  - Column (4): 0.502 (1.102)
- Sample and fit:
  - Observations: 1,037 in each column
  - R-squared: 0.212 (Col 1); 0.221 (Col 2); 0.209 (Col 3); 0.213 (Col 4)
  - Sector fixed effects: YES
  - Estimation sample: 2003Q1–2020Q1
- Notes:
  - Tightness calculated as V/U ratio.
  - Labor productivity = output per hour for all employed persons in nonfarm business sector.
  - Robust standard errors in parentheses; significance: *** p<0.01, ** p<0.05, * p<0.1.

### Annex 4: Wage Phillips Curve estimation — United Kingdom (selected coefficients and statistics)
- Dependent variable: Nominal hourly (percent, year-over-year) wage inflation
- Key coefficient estimates (columns (1)–(4)):
  - t–12 (lagged wage inflation): -0.420*** (0.027); -0.423*** (0.027); -0.427*** (0.026); -0.430*** (0.027)
  - Tightness t–1: 1.967*** (0.555) in Column (1); 1.616** (0.694) in Column (2)
  - (Tightness * low-pay) t–1: 5.744** (2.671) in Column (2)
  - Labor productivity growth t–1: 0.329** (0.118); 0.319** (0.161); 0.334** (0.119); 0.336** (0.161)
  - 4Q moving average of CPI inflation t: 0.002 (0.203); 0.079 (0.250); 0.078 (0.202); 0.092 (0.255)
  - Unemployment rate (%) t–1: -0.245 (0.185) in Column (3); -0.116 (0.116) in Column (4)
  - (Unemployment rate(%) * low-pay) t–1: -0.383** (0.192) in Column (4)
- Constant terms:
  - Column (1): 2.951*** (0.504)
  - Column (2): 2.520*** (0.739)
  - Column (3): 5.098*** (0.996)
  - Column (4): 5.076*** (0.728)
- Sample and fit:
  - Observations: 1,188 in each column
  - R-squared: 0.189 (Col 1); 0.192 (Col 2); 0.188 (Col 3); 0.190 (Col 4)
  - Industry fixed effects: YES
  - Estimation sample: 2001Q4–2020Q1
- Notes:
  - Tightness calculated as V/U ratio.
  - Labor productivity = output per hour worked for the whole economy (ONS).
  - Robust standard errors in parentheses; significance: *** p<0.01, ** p<0.05, * p<0.1.

### Back-of-the-envelope calculations using estimated tightness coefficients
- Procedure (as described for Figure 10, panel 1):
  1. Compute average increase in tightness between 2020Q2 and 2021Q4; multiply by tightness coefficient from Column (1) Table 1 to estimate direct impact on overall wage inflation.
  2. Compute tightness increase between 2020Q2 and 2021Q4 separately for low-pay and other sectors; multiply with respective coefficients from Column (2) Table 1 to estimate direct impact by industry group.
  3. Compute payroll-weighted impact on wage growth using 2021Q4 labor incomes in the two industry groups as weights.

### Annex 4. Box 1: Vacancies in low-wage jobs (Indeed data processing and low-wage classification)
- Data source and period:
  - Indeed job advertisements from January 2019 until November 2021 for Australia, Canada, the United Kingdom, and the United States.
- Vacancy counting and matching:
  - Assume one vacancy per job post (supported by prior studies: Adrjan and Lydon 2019; Adrjan and others 2021; Ens and others 2021).
  - Job title assigned to 4-digit ISCO-08 occupation via UK ONS coding index, after translating titles into English using neural machine translation (noted as particularly relevant for Canada) and stripping non-salient words.
- Matching algorithm steps:
  1. Exact match (word-for-word) with coding index (e.g., 41 percent matched in UK in first step).
  2. For remaining titles with at least three words, approximate match via string similarity (Levenshtein distance), accepting matches with score ≥ 0.8. In the UK this raises classification to 50 percent.
  3. Remaining observations paired via probabilistic algorithms (e.g., Okapi BM25).
- Aggregation and normalization:
  - Daily occupation-level vacancy counts aggregated to monthly series (average weekly vacancies over first four weeks of each month).
  - Each series normalized relative to average monthly value in 2019 (set to 100).
- Low-wage occupation classification:
  - Based on UK four waves of 2019 LFS: compute average hourly wage and employment share for each SOC2010 4-digit, convert to ISCO-08 via concordance with frequency-based splits using employment shares as weights.
  - Rank ISCO-08 codes by average hourly wage; low-wage occupations = those comprising 33 percent of employment starting from bottom.
  - For 4-digit ISCO-08 codes without a SOC2010 pairing, use mean value of other codes in same 3-digit group to assign low-wage label (e.g., if >50 percent of other 4-digit codes in same 3-digit group are low-wage).

### Annex 5. Box 2: Receding immigration and rising vacancies in Canada and the United Kingdom
- Data sources and mapping:
  - Uses Indeed vacancies and Labour Force Surveys for Canada and the United Kingdom.
- Canada mapping details:
  - Canadian LFS: occupation at 2-digit National Occupational Classification (NOC).
  - Indeed vacancies coded at 4-digit ISCO-08 converted to 2-digit NOC using Statistics Canada conversion table.
  - Conversion method: many-to-many mapping at 4-digit level; conversion is “modal” (select 2-digit NOC with most pairings). In bi-modal matches, duplicate 4-digit ISCO-08 code and assign each instance weight 0.5 to rescale vacancies.
- United Kingdom mapping details:
  - UK LFS: occupation at 4-digit SOC 2010, but foreign-worker share very volatile at that level; aggregate occupations to 3-digit level for stability.
  - Indeed vacancies already coded at 4-digit SOC 2010 and aggregated to 3-digit level.
- Definitions and measurement for Figure 2.2:
  - Low-skill occupations defined based on share with at least an undergraduate degree in 2019: occupations comprising up to approximately 50 percent of total employment from the bottom of ranking designated as low-skill.
  - Change in share of foreign workers uses yearly average for 2019 and average over first three quarters of 2021 (to reduce volatility).
  - Growth rate of vacancies measured as 2021Q3 relative to 2019 average.

### Annex 6. Box 3: Great Resignation — quit rates and tightness measures
- United States:
  - Quit rates from JOLTS.
  - Labor market tightness calculated using JOLTS vacancy levels and unemployment from US CPS.
  - In Figure 3.2, linear fitted lines between quit rates and labor market tightness use 2019 (pre-COVID) data.
- United Kingdom:
  - Quit rates computed from 2-quarter longitudinal UK LFS:
    - Identify “job leavers”: employed in Q1 and non-employed or employed in Q2 with tenure < three months.
    - Within job leavers, “quits” = separations reported as “voluntary quit.”
    - Quit rate measured as ratio of quitters to total employment in Q1.

*International Monetary Fund — sdnea2022001 (Annex materials as presented in the source PDF).*

### References

### References

### Cited works
- Adrjan, P., and R. Lydon. 2019. “Clicks and Jobs: Measuring Labour Market Tightness Using Online Data.” Economic Letter No. 6/EL/19, Central Bank of Ireland, Dublin.
- Adrjan, P., G. Ciminelli, A. Judes, M. Koelle, C. Schwellnus, and T. Sinclair. 2021. “Will It Stay or Will It Go? Analysing Developments in Telework during COVID-19 Using Online Job Postings Data.” OECD Productivity Working Papers 30, Organisation for Economic Co-operation and Development, Paris.
- Alon, T., S. Coskun, M. Doepke, D. Koll, and M. Tertilt. 2021. “From Mancession to Shecession: Women’s Employment in Regular and Pandemic Recessions.” NBER Macroeconomics Annual 36.
- Ando, S., and others. Forthcoming. “Labor Market Fallout of the COVID-19 Crisis and Associated Policy Options.” IMF Departmental Paper, International Monetary Fund, Washington, DC.
- Armantier, O. Goldman, L. Koşar G., and van der Klaauw, W. 2021 “An Update on How Households Are Using Stimulus Checks.” Liberty Street Economics, Federal Reserve Bank of New York, New York.
- Attinasi, M. G., A. Bobasu, and A. Manu. 2021. “The Implications of Savings Accumulated during the Pandemic for the Global Economic Outlook.” ECB Economic Bulletin 5.
- Barrero, J. M., N. Bloom, and S. J. Davis. 2021. “Why Working from Home Will Stick.” NBER Working Paper No. w28731, National Bureau of Economic Research, Cambridge, MA.
- Bilbiie, F. G. Eggertsson, G. Primiceri, and A.Tambalotti. 2021 “’Excess Savings’ Are Not Excessive.” Liberty Street Economics, Federal Reserve Bank of New York, New York.
- Bluedorn, J., F., Caselli, N. J. Hansen, I. Shibata, and M. M. Tavares. 2021. “Gender and Employment in the COVID-19 Recession: Cross-Country Evidence on ‘She-cessions’.” Covid Economics: Vetted and Real-Time Papers 76 (April): 87–109.
- Borup, D., and E. Montes Schütte. 2020. “In Search of a Job: Forecasting Employment Growth Using Google Trends.” Journal of Business & Economic Statistics (40)1: 1–38.
- Byrne, D., and Z. Zekaite. 2018: “Missing Wage Growth in the Euro Area: Is the Wage Phillips Curve Non-Linear?” Central Bank of Ireland Economic Letter, Dublin, Press Release, November 6.
- Coombs, K., A. Dube, C. Jahnke, R. Kluender, S. Naidu, and M. Stepner. 2021. “Early Withdrawal of Pandemic Unemployment Insurance: Effects on Earnings, Employment and Consumption.” Harvard Business School Working Paper, No. 22-046, Boston, MA.
- Dingel, J. I., and B. Neiman. 2020. “How Many Jobs Can Be Done at Home?” Journal of Public Economics 189 (C).
- Ens, E., L. Savoie-Chabot, K. See, and S. L. Wee, 2021. “Assessing Labour Market Slack for Monetary Policy.” Bank of Canada Staff Discussion Paper No. 2021-15, Ottawa.
- Fabrizio S., D. Gomes, and M. M. Tavares. 2021. “COVID-19 She-cession: The Employment Penalty of Taking Care of Young Children.” Covid Economics 72: 136–66.
- Galí, J. 2011. “The Return of the Wage Phillips Curve.” Journal of the European Economic Association 9 (3): 436–61.
- International Monetary Fund (IMF). 2021. “Recessions and Recoveries in Labor Markets: Patterns, Policies, and Responding to the COVID-19 Shock.” World Economic Outlook, Washington, DC, April.
- Kaplan, G., B. Moll, and G. L. Violante. 2020. “The Great Lockdown and the Big Stimulus: Tracing the Pandemic Possibility Frontier for the US.” NBER Working Paper No. w27794, National Bureau of Economic Research, Cambridge, MA.
- Nickel C., E. Bobeica, G. Koester, E. Lis, and M. Porqueddu. “Understanding Low Wage Growth in the Euro Area and European Countries.” ECB Occasional Paper Series No. 232, European Central Bank, Frankfurt.
- Pizzinelli, C., and I. Shibata. 2022. “Has COVID-19 Induced Labor Market Mismatch? Evidence from the US and the UK.” IMF Working Paper No. 2022/005, International Monetary Fund, Washington, DC.
- Şahin, A., J. Song, G. Topa, and G. L. Violante. 2014. “Mismatch Unemployment.” American Economic Review 104: 3529–64.
- US Federal Reserve System. 2021. “The Beige Book: Summary of Commentary on Current Economic Conditions by Federal Reserve District – May 2021.” Washington, DC.
- Woloszko, N. 2020, "Tracking Activity in Real Time with Google Trends." OECD Economics Department Working Papers No. 1634, OECD Publishing, Paris.

*Labor Market Tightness in Advanced Economies Staff Discussion Note No. SDN/2022/001*

---


_Source: https://www.imf.org/-/media/files/publications/sdn/2022/english/sdnea2022001.pdf_