## IMF Working Papers — What Matters for Job Findings and Separation in the Long Run

## Source details

**Canonical URL:** [IMF Working Papers — What Matters for Job Findings and Separation in the Long Run](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022172-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022172-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022172-print-pdf.pdf.json)

---

### Overview
- Aggregate labor market outcomes depend on flows into and out of unemployment and job-to-job transitions.
- Non-employment rate is defined as "the share of working age population that is either unemployed or not participating in the labor force."
- Data: anonymized New Zealand Household Labour Force Survey (HLFS) microdata (quarterly, 2006Q1 to 2021Q1), rotating panel design, roughly 15,000 households (~30,000 individuals over age 15).
- Samples:
  - Broad sample: all individuals aged 15-64 including those not in the labor force (measures non-employment).
  - Narrow sample: excludes NILF (standard unemployment definition).

### Methodology
- Micro-level regressions:
  - Transition dummies from quarter t to t+1: job-finding f_{i,t} and job-exit s_{i,t}.
  - Panel probit specifications: Pr(Y_{i,t}=1 | X_{i,t}, Z_{i,t}, G_{t}) = Φ(α + β X_{i,t} + γ Z_{i,t} + σ G_{t} + ε_{i,t}) and alternative with time fixed effects Pr(Y_{i,t}=1 | X_{i,t}, Z_{i,t}, G_{t}) = Φ(α + β X_{i,t} + γ Z_{i,t} + δ_{t} + ε_{i,t}).
  - Aggregate control G_{t}: annual (four-quarter average) GDP growth per capita; robustness uses Reserve Bank of New Zealand output gap.
  - Core categorical X: Education (less than high-school; high-school; post-school but less than bachelors; bachelors or more), Ethnicity (European; Māori; Pacific People; Others), Age (15–24; 25–39; 40–54; 55–64), Gender (male; female).
  - Controls Z: family/household type, regional council dummies, seasonality, unemployment duration (for finding), part-time indicator and industry (for exit).
- Counterfactual steady-state decomposition:
  - Steady-state approximation: u_{i} ≈ s_{i} / (s_{i} + f_{i}).
  - Construct counterfactual u by replacing group-specific conditional f or s with another group’s conditional f or s (marginal effects from regressions).
  - Example formulas: u_{f} − u_{f}^{c,finding} ≈ s_{f}/(s_{f}+f_{f}) − s_{f}/(s_{f}+f_{m}).

### Key empirical findings — job finding and separation (micro regressions)
- Job finding:
  - Older workers, non-European ethnic groups (Māori and Pacific people), and less educated workers tend to have lower job finding rates after controls.
  - Unemployment duration: unemployed >1 year have finding rate about 17 percentage points lower than those unemployed <1 year.
  - Never-employed/new entrants have finding rate about 14 percentage points lower than those unemployed <1 year.
  - NILF individuals (broad sample) have lower finding rates than unemployed actively looking.
  - Single parents tend to have lower finding rates.
- Job separation:
  - Young workers, Māori and Pacific people, and less-educated individuals tend to have higher separation rates.
  - Part-time jobs: people with part-time jobs are "two and a half times" more likely to transition to unemployment or out of the labor force compared to full-time jobs.
  - Single parents tend to have higher separation rates.
  - Occupations: unskilled laborers and community and personal services have higher exit rates vs managers and professionals.
  - Industry: government and health jobs have lower exit rates; rental and real estate, and administrative support services have higher exit rates.
- Gender:
  - Broad sample: women have a somewhat lower job finding rate — "about 1 percentage point lower" — and higher probability of exiting employment to unemployment or out of the labor force.
  - Narrow sample: job-finding disadvantage for women is insignificant; women have slightly lower job exit rate than men.
  - Interpretation: women are less likely than men to move from outside the labor force to employment and are more likely to leave employment to NILF.

### Heterogeneous cyclical responses
- Specification with interaction terms: Pr(Y_{i,t}=1 | X_{i,t}, Z_{i,t}, G_{t}) = Φ(α + β X_{i,t−1} + γ Z_{i,t−1} + σ G_{t} + λ X_{i,t} * G_{t} + ε_{i,t}).
- Age and cyclicality:
  - Job finding rates for young workers are more cyclical; 55-64 least cyclical for finding.
  - Prime-age and, for some specs, older workers (55-64) tend to have more cyclical exit rates (smaller differences than for finding).
- Ethnicity:
  - Māori and Pacific people tend to have more procyclical job finding rates compared to European and other groups.
  - No significant differences in cyclicality of job exit rates by ethnicity.
- Qualification:
  - Workers with only high school completion tend to have more cyclical job exit rates.
  - No large differences in job finding rate cyclicality by qualification.
- Gender:
  - No significant gender differences in cyclicality of unemployment after controls.

### Counterfactual decomposition results — contributions to broad-sample non-employment differences (units: percentage points)
- Gender:
  - Female minus Male: Difference in data 10.5; Difference due to finding rate 1.2; Difference due to separation rate 3.8; Difference from both 4.8; Unexplained difference 5.7.
  - Note: "5.7 percentage points of the difference in non-employment is not explained by the marginal effect of gender on finding or separation rates."
  - Part-time employment-to-labor force ratio for female workers is, on average, 20 percentage points higher than that for male workers for the sample period.
- Age (vs 40–54 reference):
  - Youth (15–24) minus 40 to 54: Difference in data 31.0; Difference due to finding rate -1.0; Difference due to separation rate 15.9; Difference from both 15.3; Unexplained difference 15.7.
  - 25 to 39 minus 40 to 54: Difference in data 4.2; Difference due to finding rate -0.8; Difference due to separation rate 5.0; Difference from both 4.4; Unexplained difference -0.2.
  - 55 to 64 minus 40 to 54: Difference in data 9.2; Difference due to finding rate 5.9; Difference due to separation rate 4.0; Difference from both 9.1; Unexplained difference 0.1.
- Ethnicity (European / Māori / Pacific relative to Others):
  - European minus Others: Difference in data -4.4; Difference due to finding rate 0.4; Difference due to separation rate -4.3; Difference from both -3.8; Unexplained difference -0.6.
  - Māori minus Others: Difference in data 10.8; Difference due to finding rate 3.8; Difference due to separation rate 3.6; Difference from both 7.1; Unexplained difference 3.8.
  - Pacific people minus Others: Difference in data 14.5; Difference due to finding rate 4.3; Difference due to separation rate 3.9; Difference from both 7.8; Unexplained difference 6.7.
  - Youth population shares: Māori youth account for 30.2 percent of working age population; Pacific people youth account for 31.7 percent; European people youth account for 19.9 percent.
- Education (relative to lowest qualification group):
  - Bachelors minus low: Difference in data -27.6; Difference due to finding rate -8.8; Difference due to separation rate -3.6; Difference from both -13.6; Unexplained difference -14.0.
  - Post school minus low: Difference in data -24.1; Difference due to finding rate -6.1; Difference due to separation rate -3.6; Difference from both -10.4; Unexplained difference -13.7.
  - Secondary school minus low: Difference in data -10.8; Difference due to finding rate -5.1; Difference due to separation rate -3.0; Difference from both -8.4; Unexplained difference -2.4.

### Aggregate context and calibration
- Over 2006–2021, aggregate unemployment rate has averaged around 5 percent, with peaks after the global financial crisis and during COVID.
- Using average aggregate transition rates in steady-state formula yields about 3.6 percent unemployment (sample-average unemployment ~5 percent), reflecting sample composition, two recessions, and finding-rate shifts.

### Policy implications and recommendations
- For groups where higher separation rates dominate (youth; female workers):
  - Policies supporting longer employment duration—e.g., training programs—are important.
  - Training can increase duration of employment in addition to reducing unemployment duration.
- For groups where lower finding rates dominate (older workers; Māori; Pacific people):
  - Policies supporting job finding—e.g., job search assistance and targeted hiring credits—are important.
- For groups with more cyclical behavior (youth; Māori; Pacific):
  - Ensure sufficient coverage of income insurance.
  - Intensify targeted labor market policies for these groups during downturns.

_Implied source: 5. Conclusion, wpiea2022172-print-pdf_

### 1. Introduction ........................................................................................................

### 1. Introduction

### Overview
- Aggregate labor market outcomes depend on flows into and out of unemployment and job-to-job transitions.
- More flexible labor markets exhibit greater turnover: higher job-finding rates (probability of an individual moving from being unemployed to employed) and higher job-separation rates (probability of an individual moving from being employed to unemployed).
- The paper uses anonymized New Zealand Household Labour Force Survey (HLFS) microdata to study how worker characteristics affect job finding and separation rates and how these differences translate into unemployment and non-employment rates.
- Non-employment rate is defined as "the share of working age population that is either unemployed or not participating in the labor force."

### Key findings (summary)
- Job finding:
  - Older workers, non-European ethnic groups (Māori and Pacific people), and less educated workers tend to have lower job finding rates after controlling for the business cycle and other factors.
  - The length of unemployment spells affects job finding probabilities.
- Job separation:
  - Young workers, Māori and Pacific people, and less-educated individuals tend to have higher separation rates.
- Gender differences:
  - Differences in separation and finding rates for women compared to men depend on the sample used.
  - In a broad sample including people not in the labor force, labor market dynamics for women are worse than for men.
  - In a narrow sample of employed and unemployed people only, women are not worse off, indicating women are less likely than men to move from outside the labor force to employment and more likely to leave the labor force from employment.
- Cyclical heterogeneity:
  - Youth workers tend to have more cyclical job finding rates.
  - Older workers tend to have more cyclical separation rates.
  - Māori and Pacific people tend to have more cyclical job finding rates.
  - Workers with only high school completion tend to have more cyclical job exit rates compared to those with higher degrees.
- Accounting for unemployment/non-employment differences:
  - Higher unemployment or non-employment rates for youth and workers aged 25-39 can be attributed to higher separation rates rather than lower finding rates.
  - For workers aged 54-65, higher unemployment and non-employment rates compared to other prime-aged workers are mainly attributable to lower job finding rates.
  - Ethnicity: both separation and finding rates contribute to dispersion in unemployment/non-employment rates.
  - Education: finding rates play a more important role in explaining differences across education levels.

### Relation to literature and contribution
- Builds on literature on heterogeneity in labor market outcomes; controls simultaneously for multiple individual characteristics using granular microdata.
- Uses a decomposition approach based on a steady state relation between the unemployment rate and finding and separation rates to quantify the relative importance of job finding and exit rates in explaining variation in unemployment rates across groups.
- Extends benchmark regression models to explore cyclical heterogeneity across groups.

### Paper structure (as stated)
- Section II: describes heterogeneity across workers in New Zealand and overview of HLFS microdata.
- Section III: analyzes drivers of heterogeneity in job finding and separation rates and assesses contributions to unemployment rate differences.
- Section IV: explores cyclical behavior of job finding and separation rates.
- Section V: concludes.

### Context and aggregate indicators
- Over the sample period 2006-2021, the aggregate unemployment rate has averaged around 5 percent, with peaks after the global financial crisis and during COVID.
- Aggregate finding and separation rates have fluctuated with the business cycle.
- Some worker groups—female, youth, and workers with non-European ethnicities—have persistently higher unemployment rates than the aggregate.

### Data: Household Labour Force Survey (HLFS) micro-level data
- HLFS characteristics:
  - Quarterly survey collecting employment status and personal-level information from 15,000 households (approximately 30,000 individuals over age 15) living in private dwellings in New Zealand.
  - Main data source for official quarterly employment and unemployment estimates.
  - Contains sectors, regions, unemployment spells, and rich individual characteristics suitable for granular analysis.
- Rotating panel design:
  - Repeatedly surveys respondents in the same dwellings over a set of consecutive quarters.
  - One-eighth of the rotation panel is rotated out each quarter and replaced by a new panel of dwellings.
  - There is always an overlap of seven-eighths of the panels from one quarter to another—roughly 13,000 households and 26,000 individuals.
  - The longitudinal feature allows analysis of job finding and job separation at the individual level from one quarter to the next.
- Sample attrition and adjustment:
  - Actual number remaining over two consecutive quarters is smaller due to moving or nonresponse; over 2013—2016, the response rate was around 85 percent.
  - Young (less than 30 years of age), non-European, and sole-person households have higher likelihood of missing the survey, implying potential selection bias.
  - To compensate for higher drop-out rates, sampling weights of sub-populations associated with age, ethnicity and household type are scaled up.
  - Scaling-up approach (iterative):
    - Calculate quarterly population estimates by age-ethnicity-household type using HLFS sampling weights and take averages over two consecutive quarters.
    - Repeat calculations for those remaining in the survey over two adjacent quarters.
    - The ratio between the above population estimates is the scaling factor used to adjust the sampling weights.

*IMF Working Papers — What Matters for Job Findings and Separation in the Long Run*

### 3. Labor Market Dispersion: Heterogeneity

### 3. Labor Market Dispersion: Heterogeneity

### A. Methodology: Micro-level regressions
- Data:
  - Panel dimension of the HLFS microdata; surveyed individuals/households can stay up to eight consecutive quarters.
  - Baseline "broad" sample: all individuals aged 15-64 including those not in the labor force (NILF) — effectively measuring non-employment and transitions into/out of employment (including transitions into/out of NILF). Robustness check: "narrow" sample excludes NILF (standard unemployment definition).
- Transition dummies constructed from quarter t to t+1:
  - Job-finding dummy f_{i,t} = 1 if unemployed in t and employed in t+1; = 0 otherwise.
  - Job-exit dummy s_{i,t} = 1 if employed in t and unemployed in t+1; = 0 otherwise.
- Regression specification:
  - Panel probit: Pr(Y_{i,t}=1 | X_{i,t}, Z_{i,t}, G_{t}) = Φ(α + β X_{i,t} + γ Z_{i,t} + σ G_{t} + ε_{i,t}) (Equation 1).
  - Robustness: alternative using time fixed effects: Pr(Y_{i,t}=1 | X_{i,t}, Z_{i,t}, G_{t}) = Φ(α + β X_{i,t} + γ Z_{i,t} + δ_{t} + ε_{i,t}) (Equation 2).
  - Aggregate control G_{t}: annual (four-quarter average) GDP growth per capita; robustness check uses Reserve Bank of New Zealand output gap.
- Core categorical X variables (included in all regressions):
  - Education: four groups — less than high-school, high-school, post-school but less than bachelors, bachelors or more.
  - Ethnicity: four groups — European, Māori, Pacific People, Others (multiple-response ethnicity dummies used).
  - Age: four groups — 15 to 24; 25 to 39; 40 to 54; 55 to 64.
  - Gender: male, female.
- Controls Z_{i,t}:
  - Family/household type: individual, couple, couple with children, single parent.
  - Regional council dummies; quarter-fixed effects for seasonality (excluded when time fixed effects used).
  - Finding-rate regressions additionally include unemployment duration (unemployed <1 year; unemployed >1 year; never employed) and, in broad sample, a NILF dummy.
  - Exit-rate regressions additionally include part-time indicator and industry of employment, plus family type and region.

### B. Results of micro-level regressions (summary)
- General: Significant roles for gender, age, ethnicity, and education in both job-finding and job-exit rates. Broad and narrow samples yield broadly similar results, with differences driven by exits to NILF/retirement.
- Gender:
  - Job-finding: women have a somewhat lower job finding rate than men — "about 1 percentage point lower" for the broad sample (insignificant for narrow sample).
  - Job-exit: in the broad sample, women have a higher probability of exiting employment to unemployment or out of the labor force; in the narrow sample, women have a slightly lower job exit rate than men.
  - Interpretation: women are more likely to exit into NILF; when unemployed they are not at a disadvantage finding work relative to men, but are less likely to find employment from out of the labor force.
- Age:
  - Compared to youth (15–24, reference), other age groups have significantly lower exit rates in both samples.
  - Narrow sample: 55–64 cohort has the lowest exit rate; broad sample: 40–54 cohort has the lowest exit rate — reflecting retirement exits to NILF among older workers.
  - Finding rates: differences between youth and prime-age workers are small and often insignificant; older workers have lower job finding rates in both samples.
- Ethnicity:
  - Māori and Pacific people: higher job exit rates and lower job finding rates (both broad and narrow samples), after controlling for other factors.
  - Europeans: significantly lower job separation rates in all specifications and higher job finding rates in some specifications.
- Education:
  - Higher education → lower job exit rates and higher finding rates. Finding rate increases monotonically with education (both samples).
  - Broad sample: people with bachelor’s degrees have almost twice as high a finding rate and 20 percent lower separation rate compared to individuals who have only completed high school.
- Other determinants — finding rates:
  - Unemployment duration: unemployed >1 year have a finding rate about 17 percentage points lower than those unemployed <1 year.
  - Never-employed/new entrants have a finding rate about 14 percentage points lower than those unemployed <1 year.
  - In broad sample, NILF individuals have lower finding rates than unemployed actively looking.
  - Single parents tend to have lower finding rates (narrow and broad) after controls.
- Other determinants — exit rates:
  - Part-time jobs: people with part-time jobs are "two and a half times" more likely to transition to unemployment or out of the labor force compared to full-time jobs.
  - Single parents tend to have higher separation rates (narrow and broad).
  - Occupation: unskilled laborers and community and personal services have higher exit rates compared to managers and professionals.
  - Industry: government and health jobs have lower exit rates; rental and real estate, and administrative support services have higher exit rates.

### C. Methodology: Counterfactual analysis for steady state unemployment rates
- Steady-state approximation (Shimer): u_{i} ≈ s_{i} / (s_{i} + f_{i}) (Equation 3), where:
  - u_{i} = group i’s steady state non-employment/unemployment rate;
  - s_{i} = job separation rate for group i;
  - f_{i} = job finding rate for group i.
- Approach:
  - Use group-specific conditional finding and separation rates from regressions (marginal effects controlling for covariates) to construct counterfactual steady state non-employment/unemployment rates.
  - Counterfactuals change either the finding rate or the separation rate of a group to that of another group (e.g., women given men’s conditional finding rate).
- Counterfactual formulas illustrated:
  - Impact of changing women’s finding rate to men’s: u_{f} − u_{f}^{c,finding} ≈ s_{f}/(s_{f}+f_{f}) − s_{f}/(s_{f}+f_{m}) (Equation 4).
  - Impact of changing women’s separation rate to men’s: u_{f} − u_{f}^{c,separation} ≈ s_{f}/(s_{f}+f_{f}) − s_{m}/(s_{m}+f_{m}) with separation counterfactual form given (Equation 5).
- Notes:
  - f_{m} denotes the conditional finding rate for men at the average level of other covariates (not the raw male finding rate).
  - Using average aggregate job finding/separation rates over the sample with Equation 3 yields a steady state unemployment rate of about 3.6 percent (sample average unemployment ~5 percent).

### D. Results of counterfactual analysis (summary)
- Overall patterns:
  - Higher separation rates of young workers play a disproportionate role in explaining heterogeneity across age groups.
  - Differences in finding rates are more important in explaining differences by education level.
  - Both finding and separation rate differences matter for differences across ethnicities.
- Gender counterfactuals (broad sample, parameters from Column 1 of Tables 2 and 3):
  - Observed difference: non-employment rate is 10 percentage points higher for women compared to men (Column 1, Table 4).
  - Lower female finding rate explains 1.2 percentage points of this difference (Column 2).
  - Higher female separation rate explains 3.8 percentage points of this difference.
  - (Implication: a portion of the 10 percentage point gap is explained by these conditional finding/separation differences; remaining gap reflects other factors.)
- Aggregate calibration note:
  - The steady-state calculation using sample-average aggregate transition rates implies about 3.6 percent unemployment, below the sample average ~5 percent, possibly reflecting the short sample, two recessions, and significant shifts in the finding rate.

*IMF Working Paper chapter: 3. Labor Market Dispersion: Heterogeneity*

### 5.7 percentage points of the difference in non-employment is not explained by the marginal effect of gender on

### wpiea2022172-print-pdf - 5.7 percentage points of the difference in non-employment is not explained by the marginal effect of gender on

### Gender: drivers of non-employment and unemployment gaps
- "5.7 percentage points of the difference in non-employment is not explained by the marginal effect of gender on finding or separation rates."
- More than half of the gender difference in non-employment is explained by other factors included in the regressions in Tables 2 and 3 (for example, higher part-time work and less stable occupations among women).
- The part-time employment-to-labor force ratio for female workers is, on average, 20 percentage points higher than that for male workers for the sample period.
- Narrow-sample counterfactual (Table 5): difference in unemployment rates between females and males is much smaller because the narrow sample does not capture the larger share of female workers not in the labor force; job finding and separation rates for females are not very different from males in the narrow sample, so those rates do not explain the unemployment gap; other controlled factors (e.g., higher share of part-time jobs for female workers) likely explain the remaining difference.

### Age: importance of exit and finding rates across cohorts
- Counterfactuals (Table 4) compare cohorts to 40 to 54 year olds (lowest non-employment in data).
- Youth (15 to 24 year olds): non-employment rate is more than 30 percentage points higher than 40 to 54 year olds, with higher separation rate explaining more than half of the difference.
- 25 to 39 year olds: higher separation rate explains the entire difference in non-employment rate compared to 40 to 54 year olds.
- Older cohort (55 to 64): both lower finding and higher separation play an important role in higher non-employment.
- Table 5 (narrow sample, unemployment rates): for workers aged 15-24 and 25-39, job separation rates drive differences relative to workers aged 40-54, although these differences are much smaller than non-employment differences in the broad sample; for workers aged 55-64, differences relative to 40-54 are much smaller for unemployment rates.

### Ethnicity: contribution of finding and separation rates and residual factors
- Māori non-employment rate is 10.8 percentage points higher than for non-Māori.
- Of that 10.8 percentage points: 3.8 percentage points can be explained by lower finding rates for Māori; 3.6 percentage points can be explained by higher separation rates for Māori.
- A significant portion of the Māori–non-Māori non-employment gap remains unexplained by ethnicity-specific finding and separation rates, reflecting other characteristics controlled in regressions (e.g., higher part-time work share, age distribution, qualification levels).
- Youth population shares: Māori youth account for 30.2 percent of working age population; Pacific people youth account for 31.7 percent; European people youth account for 19.9 percent.
- Similar pattern for Pacific people: higher non-employment partially explained by finding and separation rates, with other factors contributing.
- For other non-European ethnicities: finding rates are lower but there is no large difference in exit rates; "Others" have non-employment about 4 percentage points higher, most of which is explained by lower finding rate.
- Table 5 (narrow sample, unemployment rates): higher unemployment rates for Māori and Pacific people are explained partially by job finding and separation rates, although other factors also contribute.

### Education: finding rates are relatively more important
- Comparison to group that has not graduated high school (lowest qualification).
- Group with bachelor’s degree: non-employment rate is 27.6 percentage points lower than lowest qualification group.
- Of that 27.6 percentage points: higher job finding rate explains 8.8 percentage points; lower separation rate explains 3.6 percentage points.
- About half of the 27.6 percentage points difference is not explained by education level; other factors (share of part-time employment, age distribution, ethnicity, industry) also play roles.
- For post-secondary degree and secondary school education groups: large differences with lowest qualification are explained by higher job finding rates and lower separation rates, with other factors contributing.
- Table 5 (narrow sample, unemployment rates): job finding rates explain a larger part of the differences across qualification groups, except for the group with secondary school education.

### Heterogeneous response to business cycles
- Equation (6) specification: Pr(Y_{i,t}=1 | X_{i,t}, Z_{i,t}, G_{t}) = Φ(α + β X_{i,t−1} + γ Z_{i,t−1} + σ G_{t} + λ X_{i,t} * G_{t} + ε_{i,t}), where X_{i,t} * G_{t} is vector of interaction terms of individual-characteristic dummies and the business cycle indicator (annual per capita GDP growth rate or output gap); λ captures heterogeneous responses of worker groups to business cycles; β gauges time-invariant heterogeneity after controlling for cyclical heterogeneity.
- Tables 6 and 7 (parameters associated with business cycle interacted with individual characteristics, λ) show some heterogeneity in job finding and separation rates across groups.
- Age and cyclicality:
  - Job finding rates for young workers are more cyclical.
  - Job finding rates for older workers (55-64) are least cyclical.
  - On job separation, prime-age workers and, for some specifications, older workers aged 55-64 tend to have more cyclical exit rates (difference is smaller than for finding rates).
  - This pattern aligns with Gielen and Ours (2006): cyclical workforce adjustment mainly via worker entry for young workers and via separations for old workers.
- Ethnicity and cyclicality:
  - Māori and Pacific people tend to have more procyclical job finding rates compared to European people and other ethnic groups (broad sample, after controls).
  - No significant differences in cyclicality of job exit rates by ethnicity.
- Qualification and cyclicality:
  - Workers with only high school completion tend to have more cyclical job exit rates.
  - No large differences in job finding rate cyclicality by qualification.
- Gender and cyclicality:
  - No significant gender differences in cyclicality of unemployment after controlling for other factors (contrasts with some prior studies).
- Robustness:
  - Appendix Tables 1 and 2 report parameters for time-invariant individual characteristics; results in Section III are robust to addition of interaction terms of business cycle indicators and individual characteristics.

*IMF WORKING PAPERS What Matters for Job Findings and Separation in the Long Run — INTERNATIONAL MONETARY FUND*

### 5. Conclusion

### 5. Conclusion

### Methodology and scope
- Analysis uses novel anonymized individual-level Household Labour Force Survey data for New Zealand covering 2006Q1 to 2021Q1.
- Includes a rich set of control variables and analyzes individual characteristics within region, industry, and job type (full-time or part-time).
- Decomposes unemployment rates into gross flows and examines contributions of job finding rates and separation rates by calculating marginal impacts of individual characteristics on job finding and separation rates.
- Investigates heterogeneity of individual responses to cyclical fluctuations.
- Analysis conducted for two samples:
  - Narrow sample: transitions into and out of unemployment for people in the labor force.
  - Broad sample: includes transitions into and out of the labor force.

### Key empirical findings
- Individual characteristics play a significant role in explaining job finding and separation probabilities.
- Age groups:
  - Higher separation rate of young workers explains the larger part of the unemployment rate gap between youth and prime age workers.
  - Lower job finding probabilities of older workers explain part of the gap between workers aged 55-64 and younger prime age workers.
  - For the narrow sample, the oldest cohort (55 to 64 years) has the lowest exit rate.
  - For the broad sample, prime age workers (40-54 years) have the lowest exit rates; this difference reflects retirement transitions from employment to out of the labor force rather than into unemployment.
- Gender:
  - Higher separation rate of female workers plays a large role in explaining gender gaps.
  - Finding rates for women are similar to men in the narrow sample, but significantly lower for the broad sample, indicating women are less likely than men to transition from out of the labor force to employment.
- Ethnicity:
  - Higher separation rates for Māori and Pacific workers play a large role in their elevated non-employment rates.
  - Differences in finding rates between non-European and European workers also contribute to some of the ethnicity gap.
- Education:
  - Higher qualifications (Bachelor and above; Post secondary) are associated with higher job-finding probabilities and lower job-exit probabilities in regressions (see Tables 2 and 3).
- Cyclical heterogeneity:
  - Youth workers, Māori and Pacific people tend to have more cyclical job finding rates after controlling for other factors.
  - Older workers aged 55-64 and workers with high school education tend to have more cyclical job exit rates.
- Robustness:
  - Results are robust to alternative specifications that control for heterogeneous responses to the business cycle.

### Quantitative decomposition highlights (broad sample non-employment, units in percentage points)
- Gender:
  - Female minus Male: Difference in data 10.5; Difference due to finding rate 1.2; Difference due to separation rate 3.8; Difference from both 4.8; Unexplained difference 5.7.
- Age:
  - Youth (15-24) minus 40 to 54: Difference in data 31.0; Difference due to finding rate -1.0; Difference due to separation rate 15.9; Difference from both 15.3; Unexplained difference 15.7.
  - 25 to 39 minus 40 to 54: Difference in data 4.2; Difference due to finding rate -0.8; Difference due to separation rate 5.0; Difference from both 4.4; Unexplained difference -0.2.
  - 55 to 64 minus 40 to 54: Difference in data 9.2; Difference due to finding rate 5.9; Difference due to separation rate 4.0; Difference from both 9.1; Unexplained difference 0.1.
- Ethnicity:
  - European minus Others: Difference in data -4.4; Difference due to finding rate 0.4; Difference due to separation rate -4.3; Difference from both -3.8; Unexplained difference -0.6.
  - Māori minus Others: Difference in data 10.8; Difference due to finding rate 3.8; Difference due to separation rate 3.6; Difference from both 7.1; Unexplained difference 3.8.
  - Pacific people minus Others: Difference in data 14.5; Difference due to finding rate 4.3; Difference due to separation rate 3.9; Difference from both 7.8; Unexplained difference 6.7.
- Education:
  - Bachelors minus low: Difference in data -27.6; Difference due to finding rate -8.8; Difference due to separation rate -3.6; Difference from both -13.6; Unexplained difference -14.0.
  - Post school minus low: Difference in data -24.1; Difference due to finding rate -6.1; Difference due to separation rate -3.6; Difference from both -10.4; Unexplained difference -13.7.
  - Secondary school minus low: Difference in data -10.8; Difference due to finding rate -5.1; Difference due to separation rate -3.0; Difference from both -8.4; Unexplained difference -2.4.

### Policy implications and recommendations
- Because job separation rates play a larger role for youth workers and females:
  - Policies that support longer employment duration, such as training programs, would be important.
  - Training can increase duration of employment in addition to reducing unemployment duration.
- Because job finding rates play a larger role for older workers, Māori and Pacific people:
  - Policies that support job finding, such as job search assistance and targeted hiring credits, remain important.
- Because some worker groups (young, Māori, Pacific) display more cyclical behavior:
  - Sufficient coverage of income insurance for these groups would be essential to promote economic stability.
  - Intensified labor market policies targeted to these groups during downturns would be warranted.

_Implied source: 5. Conclusion, wpiea2022172-print-pdf_

### References

### References

### Core themes covered
- Labor market dynamics and unemployment: Clark, K. B., Summers, L. H., Holt, C. C., Hall, R. E., Baily, M. N., & Clark, K. B. (1979). Labor market dynamics and unemployment: a reconsideration. Brookings Papers on Economic Activity, 1979(1), 13-72.
- Unemployment dynamics and cyclical behavior: Elsby, M., Hobijn, B., and Sahin, A. (2013). Unemployment Dynamics in the OECD. The Review of Economics and Statistics, MIT Press 95 (2): 530-548; Shimer, R. (2005). The cyclical behavior of equilibrium unemployment and vacancies. American Economic Review, 95(1), 25-49; Shimer, R. (2012). Reassessing the ins and outs of unemployment. Review of Economic Dynamics, 15(2), 127-148.
- Job finding and separation cyclicality: Fujita, S., & Ramey, G. (2009). The cyclicality of separation and job finding rates. International Economic Review, 50(2), 415-430; Hairault, J. O., Le Barbanchon, T., & Sopraseuth, T. (2015). The Cyclicality of the Separation and Job Finding Rates in France. European Economic Review, 76, 60-84.
- Matching, search, and macro labor theory: Yashiv, E. (2007). Labor search and matching in macroeconomics. European Economic Review, 51(8), 1859-1895; Barnichon, R., & Figura, A. (2015). Labor market heterogeneity and the aggregate matching function. American Economic Journal: Macroeconomics, 7(4), 2.
- Heterogeneity and demographics: Ahn, H. J., & Hamilton, J. D. (2020). Heterogeneity and unemployment dynamics. Journal of Business & Economic Statistics, 38(3), 554-569; Gielen, A. C., & van Ours, J. C. (2006). Age-specific cyclical effects in job reallocation and labor mobility. Labour Economics, 13(4), 493-504; Zanin, L. (2014). On Okun’s law in OECD countries: An analysis by age cohorts. Economics Letters, 125(2), 243-248.
- Inequality, gender, and wage gaps: Blau, F. D., & Kahn, L. M. (2017). The Gender Wage Gap: Extent, Trends, and Explanations. Journal of Economic Literature, 55(3), 789-865; Sin, I., Stillman, S., & Fabling, R. (2020). “What Drives the Gender Wage Gap? Examining the Roles of Sorting, Productivity Differences, Bargaining and Discrimination. “Review of Economics and Statistics, 1-44.
- Policy interventions and labor programs: Cahuc, P., Carcillo, S., & Le Barbanchon, T. (2019). The effectiveness of hiring credits. The Review of Economic Studies, 86(2), 593-626; Fitzenberger, B., Orlanski, O., Osikominu, A., & Paul, M. (2013). Déjà Vu? Short-term training in Germany 1980–1992 and 2000–2003. Empirical Economics, 44(1), 289-328; Gorry, A. (2013). Minimum Wages and Youth Unemployment. European Economic Review, 64, 57-75.
- Education, training, and transitions: Gritz, R. M. (1993). The impact of training on the frequency and duration of employment. Journal of Econometrics, 57(1-3), 21-51; Farber, H. S. (1999). Alternative and Part-time Employment Arrangements as a Response to Job Loss. Journal of Labor Economics, 17(S4), 142-169.

### Key empirical sources for New Zealand and regional focus
- New Zealand-specific labor market analyses and data: Ball, C., Groshenny, N., Karagedikli, Ö., Ozbilgin, M., & Robinson, F. (2020). “Low Wage Growth and Job-to-job Transitions: Evidence from Administrative Data in New Zealand.”; Hyslop, D., Rice, A., & Skilling, H. (2019). Understanding labour market developments in New Zealand, 1986-2017 (No. DP2019/02). Reserve Bank of New Zealand; Jacob, P., & Robinson, F. (2019). "Suite as! Augmenting the Reserve Bank’s output gap indicator suite," Reserve Bank of New Zealand Analytical Notes series AN2019/08, Reserve Bank of New Zealand.
- Inequality and outcomes for Māori and Pacific graduates: Theodore, R., Taumoepeau, M., Kokaua, J., Tustin, K., Gollop, M., Taylor, N., ... & Poulton, R. (2018). Equity in New Zealand university graduate outcomes: Māori and Pacific graduates. Higher Education Research & Development, 37(1), 206-221.
- Policy evaluation and minimum wage impacts in New Zealand: Hyslop, D., & Stillman, S. (2007). Youth minimum wage reform and the labour market in New Zealand. Labour Economics, 14(2), 201-230.

### Theoretical and methodological foundations cited
- Equilibrium wage stickiness and employment fluctuations: Hall, Robert E. (2005). Employment Fluctuations With Equilibrium Wage Stickiness. American Economic Review 95(1): 50-65.
- Life-cycle flows and worker transitions: Choi, S., Janiak, A., & Villena‐Roldán, B. (2015). Unemployment, Participation and Worker Flows Over the Life‐Cycle. The Economic Journal, 125(589), 1705-1733.
- Family, immigration, and labor supply: Baker, M., & Benjamin, D. (1997). The role of the family in immigrants' labor-market activity: an evaluation of alternative explanations. American Economic Review, 705-727; Alesina, A., Algan, Y., Cahuc, P., & Giuliano, P. (2015). Family values and the regulation of labor. Journal of the European Economic Association, 13(4), 599-630; Altonji, J. G., & Blank, R. M. (1999). Race and gender in the labor market. Handbook of labor economics, 3, 3143-3259.

### Select other referenced studies
- Anand, R., Kothari, S., Kumar, N., & Papi, L. (2016). South Africa: Labor Market Dynamics and Inequality1. IMF Working Papers, 2016(137).
- Hoynes, H., Miller, D. L., & Schaller, J. (2012). Who suffers during recessions?. Journal of Economic Perspectives, 26(3), 27-48.
- Gritz, R. M. (1993). The impact of training on the frequency and duration of employment. Journal of Econometrics, 57(1-3), 21-51.
- Yashiv, E. (2007). Labor search and matching in macroeconomics. European Economic Review, 51(8), 1859-1895.

*What Matters for Job Finding and Separation in the Long Run? Evidence from Labor Market Dynamics in New Zealand — Working Paper No. WP/2022/###*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022172-print-pdf.pdf_