## 1nzlea2021002

## Source details

**Canonical URL:** [1nzlea2021002](https://www.imf.org/-/media/files/publications/cr/2021/english/1nzlea2021002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2021/english/1nzlea2021002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2021/english/1nzlea2021002.pdf.json)

---

### The uneven impact of the pandemic on New Zealand’s labor market — key findings
- Pandemic impact was relatively mild by international comparison but disproportionately affected youth, female workers, and some ethnic groups.
- Lockdown and border closure in March 2020; hours worked declined by close to 9 percent in 2020Q2 before recovering in Q3.
- Government wage subsidy scheme covered more than 60 percent of workers, described as the widest coverage among such schemes in the world.
- Aggregate labor market outcomes:
  - Unemployment rate: increased to 4.9 percent in 2020Q4 from 4.1 percent in 2019Q4.
  - Underutilization rate: increased to 11.9 percent in 2020Q4 from 10.1 percent in 2019Q4.
- Sectoral and group heterogeneity:
  - Job losses concentrated in contact-dependent service sectors (e.g., accommodation, food services, recreational services).
  - Youth: largest adjustment — a 2.4 percentage point year‑on‑year increase in the unemployment rate.
  - Female unemployment and underutilization increased more due to job losses in contact-intensive services.
  - Māori and Pacific people more affected than other ethnic groups (Budget Policy Statement 2021 analysis); Asian unemployment increased more sharply than the aggregate but remains below Māori and Pacific unemployment.
  - Workers with lower levels of education appear more severely impacted.
- Wage subsidy effects:
  - International comparisons suggest larger job retention program coverage led to more benign outcomes relative to Okun’s law predictions.
  - New Zealand’s youth unemployment increased by less than predicted by historical Okun’s law estimates, implying wage subsidies and broader policy response were particularly effective for young workers.
  - Male workers tended to benefit more from wage subsidies than female workers.
  - MSD administrative data: proportion of young workers and male workers receiving wage subsidies was higher than other groups.

### Scarring from youth unemployment — empirical magnitudes and implications
- Cohort regression (OECD unemployment rates by five-year age groups for 30 countries, 1960–2019) links cohort unemployment at later ages to unemployment faced during youth (ages 15–24), with age-group dummies and country-year fixed effects; standard errors clustered at the country level.
- Estimated scarring magnitudes (cohort regression results):
  - A cohort that experiences a 1 percentage point higher unemployment rate during its youth suffers about a 0.2 percentage point higher unemployment rate when they are 25 to 29 years of age.
  - The same cohort suffers about a 0.05 percentage point higher unemployment rate when they are 30 to 34 years of age (computed as +0.208 − 0.150).
  - Beyond age 35, the effect of youth unemployment becomes insignificant.
- Coefficients (youth unemployment impact on future unemployment rates):
  - All: 0.208***
  - Male: 0.185***
  - Female: 0.212***
- Age interaction coefficients (reference age group: 25 to 29):
  - Youth unemployment * Age 30 to 34: -0.150*** (All); -0.143*** (Male); -0.142*** (Female)
  - Youth unemployment * Age 35 to 39: -0.212*** (All); -0.197*** (Male); -0.199*** (Female)
  - Youth unemployment * Age 40 to 44: -0.246*** (All); -0.227*** (Male); -0.229*** (Female)
- Employment Protection Legislation (EPL) interaction (column 4):
  - Youth unemployment * EPL: 0.156**
  - Youth unemployment * EPL * Age 30 to 34: -0.120**
  - Youth unemployment * EPL * Age 35 to 39: -0.208***
  - Youth unemployment * EPL * Age 40 to 44: -0.257***
- Conditional implication for New Zealand given relatively mild EPL:
  - Previous recessions increased New Zealand’s youth unemployment rate by about 8-9 percentage points, translating to about a 0.4-0.5 percentage point higher unemployment rate at ages 25 to 29.
  - For a country with average EPL, comparable impact would be about 2.5 percentage points at ages 25 to 29.
- Scarring is larger for women and for countries with less flexible labor markets (regressions for males and females separately show stronger persistence for females).

### Effectiveness of Active Labor Market Policies (ALMPs) — cross-country regression evidence
- Regression setup:
  - Dependent variable: change in unemployment rate.
  - ALMP types included (lagged average of previous two years): training; employment incentives; other policies (direct job creation, start-up incentives, rehabilitation).
  - Panel: OECD countries, sample period 1994-2019. Observations: 545 (All, Female, Young).
- Estimated coefficients (effects on change in unemployment rate; standard errors in parentheses):
  - Training (lagged):
    - All: -0.68 (0.86)
    - Female: -0.65 (1)
    - Young: -3.61* (2.07)
  - Employment Incentives (lagged):
    - All: -2.59** (0.99)
    - Female: -3.02* (1.71)
    - Young: -5.03** (2.1)
  - Other ALMP (lagged):
    - All: 0.47 (1.16)
    - Female: -0.21 (1.32)
    - Young: -0.16 (2.46)
  - GDP growth:
    - All: -0.22*** (0.04)
    - Female: -0.19*** (0.04)
    - Young: -0.42*** (0.11)
  - GDP growth (lagged):
    - All: -0.11*** (0.02)
    - Female: -0.11*** (0.02)
    - Young: -0.22*** (0.03)
- Interpretation:
  - Employment incentives are particularly effective for the overall population and significantly so for young people.
  - Training programs show strong estimated effectiveness for young workers (Young: -3.61*).
  - Other ALMPs (direct job creation, etc.) are not robustly effective across groups in these estimates.

### Design, targeting, and micro evidence on ALMPs
- Empirical and literature findings:
  - Training is often effective for low-skilled workers and for reducing long-term unemployment.
  - Specific skill training tends to have larger effects than general training.
  - Training combined with job search assistance enhances outcomes.
  - Hiring subsidies are more impactful when targeted at small firms, low-wage workers, and jobs with rigid wages, and are more cost-effective if temporary during recessions.
  - Direct job creation is often less effective than training and hiring subsidies (cited literature: Kluve 2010; Card and others 2018; Cahuc 2019; Greenberg and others 2003; Jacobson et al. 2005; Dorsett 2006).

### New Zealand — current ALMP spending, pandemic-era support, and recommended policy options
- Historical and pandemic-era ALMP spending:
  - Public spending on ALMPs in New Zealand declined from 0.5 percent of GDP in 2004 to 0.2 percent of GDP in 2017.
  - Since the onset of the pandemic, New Zealand expanded ALMPs including direct job creation, hiring subsidies, and free access to training, alongside a large-scale wage subsidy scheme.
- Pandemic-era income and labor market support (amounts in NZ$):
  - Wage subsidies: NZ$ 14.8 bn
  - Permanent increase in social benefits: NZ$ 2.4 bn
  - COVID-19 leave and self-isolation support: NZ$ 0.1 bn
  - Income Relief Payment: NZ$ 0.6 bn
  - Increase in Winter Energy Payment: NZ$ 0.5 bn
  - Job creation in environmental projects: NZ$ 1.1 bn
  - Targeted Training and Apprenticeship Fund: NZ$ 0.3 bn
  - MSD capacity expansion: NZ$ 0.4 bn
  - Expansion of Flexi-Wage Subsidies: NZ$ 0.3 bn
  - He Poutama Rangatahi – youth employment support: NZ$ 0.1 bn
- Policy recommendations:
  - Increase spending on ALMPs given historically low levels and declining trend pre-pandemic.
  - Expand coverage of training programs beyond narrowly targeted beneficiaries to support broader workers in transition, particularly low-skilled workers and the long-term unemployed.
  - Upgrade training programs to reflect industry needs (Reform of Vocational Education (RoVE) noted as a positive step).
  - Scale up support targeted at displaced low-wage workers if long-term unemployment persists (e.g., expand Flexi-Wage subsidies).
  - Strengthen job search assistance for cost-effective worker reallocation.
  - Review and reinforce quality of interventions; some programs may require improvement (MSD 2019).
  - Emphasize early activation to reduce scarring given that employment likelihood decreases with duration of unemployment.
  - Address inequality and gender gaps:
    - Targeted education support for Māori and Pacific peoples to promote human capital accumulation.
    - Expand childcare support to promote female labor force participation and help close the widened gender gap post-pandemic.

### Conclusions on labor market and policy implications
- Pandemic impact and disparities:
  - The pandemic had a relatively mild yet uneven impact on New Zealand’s labor market, aggravating pre-existing disparities.
  - Before COVID-19, unemployment was significantly higher among youth, women, low-skilled workers, Māori and Pacific peoples.
  - Youth experienced the largest drop in employment during the pandemic.
- Scarring magnitude reiterated:
  - A 1 percentage point increase in youth (ages 15–24) unemployment is associated with a 0.2 percentage point higher unemployment rate between ages 25 and 29.
  - Scarring effects are larger and more persistent for women (female point estimate in cohort regression: 0.212*** vs male 0.185***).
- Policy implication:
  - Continued macroeconomic support together with scale-up and careful targeting of ALMPs (training for young workers, hiring subsidies targeted to low-wage workers, expanded coverage and quality of training, stronger job search assistance) can mitigate distributional effects and reduce scarring while promoting long-term human capital accumulation.

---

### Households — balance sheets, housing, and mortgage vulnerabilities
- Real GDP contracted by 3 percent in 2020.
- Wage subsidy: NZ$14.8 billion or 4.6 percent of GDP.
- Cash rate lowered by 75 bps to 0.25 percent.
- Household disposable income: increased 3 percent y/y in the year ended in September 2020 (from 4.8 percent y/y in December 2019).
- Household deposit growth: close to 10 percent y/y in December 2020 (doubled).
- Household net wealth was 938 percent of household disposable income in June 2020, an increase of 5.4 percent y/y.
  - Housing and land value: 4.2 percent y/y increase.
  - Net financial wealth: 1.3 percent growth.
- House prices (REINZ): surged by 24 percent y/y in March 2021.
- Mortgage servicing and debt:
  - Interest payments on housing loans: 5.6 percent of household disposable income in September 2020 (from 6.2 percent in September 2019).
  - Household debt: 163 percent of household disposable income.
- Affordability heterogeneity:
  - Bottom half of income groups: spent close to 30 percent of household expenditure on housing and household utilities.
  - By dwelling ownership:
    - 27 percent of renters spent more than 40 percent of household disposable income on housing costs.
    - 12 percent of owner-occupier households spent more than 40 percent.
- LVR / DTI and lending patterns:
  - Temporary removal of LVR restrictions increased high LVR lending to investors.
  - Share of first-home buyers increased to 19 percent.
  - Loans with DTI larger than five expanded by 1.8 percentage point of total new loan commitment in first nine months of 2020.
  - Banks provided repayment deferrals to about 8 percent of mortgages outstanding at peak in May 2020; less than one percent remained on deferral by early January 2021.
  - Around NZ$12.8 billion or 4.3 percent of mortgage lending restructured to interest-only loans.
- Vulnerability to interest rate shocks:
  - A one percentage point increase in the mortgage rate is estimated to raise household interest servicing burden to about 7 percent of household disposable income, from 5.6 percent in September 2020.
  - Simulation: a one percentage point increase in mortgage rate would increase total principal and interest repayments from currently 36 percent to 41 percent of household disposable income, exceeding conventional affordability range of 25-40 percent.
- Data limitations:
  - DTI and income levels of investor loans or debt secured on other properties are not published.
- Key reported baseline/implicit numbers from source tables preserved verbatim where presented:
  - “Implicit Interest rate (percent): 3.8 4.8 1.0”
  - “Interest payments (NZ$ bil): 11 14 26 (Percent of household disposable income) 5.6 7.1 1.5”
  - “Housing loan outstanding (NZ$ bil): 290”
  - “Household disposable income (NZ$ bil): 198”
  - Methodological notes and alternative scenario numbers preserved as reported in source.

### Non-Financial Corporations (NFCs), SMEs, and corporate vulnerability
- Pre-pandemic corporate metrics:
  - Size of NFC sector (total assets): declined from peak 350 percent of GDP in 2009 to 306 percent of GDP in 2019.
  - Corporate leverage (debt-to-equity ratio): around 77 percent since 2015.
  - Cash and deposits: about 12 percent of financial assets.
- Firm heterogeneity:
  - Close to one quarter of firms (16 percent of large firms and 30 percent of SMEs) have a liquidity ratio (cash cover of short-term borrowing) of less than one — corresponding to about 12 percent of total corporate debt.
- Pandemic support and outcomes:
  - Wage subsidy program: supported more than 60 percent of jobs through NZ$9.4 billion paid to about 1.5 million unique jobs.
  - Proportion of jobs in accommodation and food services supported by wage subsidy: 93 percent.
  - SMEs revenue grew by 1.4 percent y/y in December 2020 (Xero Small Business Insights), small business jobs grew by 3.2 percent y/y in December 2020.
  - Tourism-dependent Queenstown revenue down 21.4 percent y/y in December 2020.
- Business bankruptcies:
  - Business bankruptcies declined substantially in 2020; expected to rise as support measures are phased out in 2021 and administrative delays are completed.
- EDF and sectoral default signals:
  - Overall EDF in New Zealand rose at the outset of the COVID-19 crisis but less markedly than most advanced economies; EDFs have largely come down since initial rise.
  - Default probability of New Zealand firms in tourism has risen with continued closure of the international border.
- Corporate Vulnerability Index / Debt at Risk findings:
  - “one quarter of total debt in the sample could be at risk based on liquidity vulnerability.”
  - “one half of SMEs debt could be at risk due to higher leverage.”
  - “one tenth of SME debt could be at risk measured by three indicators simultaneously.”
  - Total debt at risk declined substantially according to more than one indicator.
- Liquidity and insolvency simulation results (Orbis-based scenarios):
  - Scenario I (50 percent decline in turnover): “about 5 percent of firms will have liquidity shortfalls after three months.”
  - Scenario II (75 percent decline in turnover for three months): “about 12 percent of all firms, 10 percent of SMEs, and 18 percent of SMEs in services could encounter liquidity shortfalls.”
  - “The share of firms expected to become insolvent is smaller, ranging between 1-7 percent under the two stress scenarios.”
  - Corporate tax rate used in simulation: “28 percent.”
  - Material cost elasticity assumption: “c is assumed at 0.8.”
- SBCS program details preserved verbatim:
  - SBCS “extended until the end of 2023”
  - SBCS “provides small businesses a 5-year loan under eligible criteria which include at least a 30 percent decline in revenue over a 14-day period in the previous 6 months, compared with the same 14-day period a year ago.”
  - “The maximum amount of loan is NZ$10,000 plus NZ$1,800 per full-time-equivalent employee.”
- Example support: “a NZ$400 million Tourism Recovery Package under the Budget 2020.”

### Medium-term output, drivers, scenarios, and policy options
- Pre-pandemic drivers:
  - Average potential output growth in 2017-19: 2.9 percent.
  - Contribution of productivity (TFP) to potential output growth by 2019: 0.2 percentage point.
  - Net migration accounted for about 70 percent of working age population growth before the border closure.
- Baseline medium-term impact of the pandemic:
  - Potential output expected to remain about 3¾ percent below pre-COVID trend by 2026.
  - In labor productivity terms, potential level expected to remain about 1.1 percent below the pre-COVID trend in the medium term.
- Contributions to medium-term potential output loss:
  - Reduced labor input (largest channel): net migration decline contributes 1.5 percentage point; increase in NAIRU due to skill mismatches contributes 0.1 percentage point; total reduced labor input contribution: 1.6 percentage points.
  - Productivity slowdown (baseline contribution): 0.8 percentage point.
  - Capital accumulation slowdown: 1.3 percentage points.
- Net migration scenario path (assumption):
  - Net migration remains below pre-COVID trend (approximated by average over 2015-19) by 90 percent in 2021, by 37 percent in 2022, and by 12 percent in 2023, before reaching pre-COVID level in 2024.
- Productivity uncertainty scenarios:
  - Upside scenario:
    - Moderate within-sector productivity gains of about 0.5 percent in the medium term.
    - Potential output loss in 2026: 1½ percent lower than pre-COVID trends.
    - Labor productivity under upside: 1.0 percent higher than the pre-COVID trend.
  - Downside scenario:
    - Within-sector productivity slowdown about twice larger than baseline.
    - Potential output in 2026 under downside: five percent lower than pre-COVID trends.
    - Labor productivity under downside: 1.5 percent lower than the pre-COVID trend.
- Policy guidance:
  - Maintain adequate fiscal and monetary policy stimulus until private demand recovery is firmly entrenched to mitigate output losses and limit scarring.
  - Begin structural reforms now; reforms take significant time to affect medium-term growth and living standards.
  - Productivity-enhancing measures recommended:
    - Product market deregulation and pro-competitive reforms.
    - Improve insolvency framework to facilitate resource reallocation and lift productivity; tailor procedures to SMEs.
    - Scale up government R&D spending and expand R&D tax credit scheme introduced in FY2020.
    - Promote FDI by reducing FDI-related regulation while maintaining control over sensitive areas; attract high quality FDI to generate productivity spillovers.
    - Trade facilitation: advance rulings, automation, information availability, and external border agency cooperation.

### Capital accumulation and investment recommendations
- Infrastructure:
  - Recommendation: scale up infrastructure investment, focusing on telecommunications, electricity, and transportation.
  - Infrastructure spending gap estimated to be between 8 and 23 percent of GDP.
  - New Zealand Infrastructure Commission can play a central role in planning and pipeline management.
- Business investment and financial policies:
  - Encourage FDI and incentivize business investment; scale up government policies that aid business cash flow if downside risks materialize.
  - Government measures noted as appropriate and scalable: business tax loss carry-back scheme; Small Business Cash Flow Scheme; Business Finance Guarantee Scheme.
  - Complementary product market reforms would boost business investment through promoting competition.
- Firm-level determinants of investment (Tobin’s Q model, unbalanced annual panel 1990-2019, 1,443 observations):
  - Estimated coefficients (dependent variable: Investment-to-Lagged Capital; standard errors in parentheses):
    - Cost of Debt (-1): -0.124 (0.171)
    - Debt (-1): -0.157*** (0.046)
    - Liquidity (-1): 0.013* (0.007)
    - Uncertainty (-1): -0.069*** (0.034)
    - Tobin's Q (-1): 0.044*** (0.007)
  - R-squared: 0.058
- Interpretation: Financial variables such as cash flow and uncertainty play an important role in firms’ investment decisions; uncertainty and financial conditions exert significant influence on business investment.

---

*Source: https://www.imf.org/-/media/files/publications/cr/2021/english/1nzlea2021002.pdf*

### References _____________________________________________________________________________ 15

### THE UNEVEN IMPACT OF THE PANDEMIC ON NEW ZEALAND’S LABOR MARKET

### Introduction
- The pandemic’s impact on New Zealand’s labor market has been relatively mild by international comparison but has disproportionately affected the youth, female workers, and some ethnic groups.
- Strong labor market policy support, including training and hiring subsidies, is warranted to facilitate the reallocation of workers affected by the pandemic.
- This paper analyzes the impact of the pandemic on New Zealand’s labor market and discusses policy options across sections B–G.

### Labor Market and Inequality Before the Pandemic
- Aggregate unemployment decreased to 4.1 percent in 2019.
- Youth unemployment rate was 11.3 percent in 2019, similar to the OECD average, and higher than the G7 average of 9.4 percent.
- Female unemployment was higher than male unemployment in most age groups, with a relatively large gap in the age group of 35-44.
- Some ethnic groups faced higher unemployment: historically high unemployment rates among Māori and Pacific people and lower labor force participation rates for these groups.
- Inequality in New Zealand was relatively high compared to other advanced economies (Gini coefficient, latest, 2017-19).
- Okun’s law estimates:
  - The unemployment rate among youth in New Zealand is more sensitive to economic cycles than other age groups, with sensitivity nearly twice as large as the average of all age groups.
  - Low-skilled workers are more prone to lose jobs during downturns.
  - Employment of female workers tended to be less sensitive to macroeconomic conditions in previous business cycles.

### Impact of the Pandemic on New Zealand’s Labor Market
- Lockdown and border closure in March 2020; hours worked declined by close to 9 percent in 2020Q2 before recovering in Q3.
- The government introduced a wage subsidy scheme covering more than 60 percent of workers, the widest coverage among such schemes in the world.
- Aggregate labor market outcomes:
  - Unemployment rate increased to 4.9 percent in 2020Q4 from 4.1 percent in 2019Q4.
  - Underutilization rate increased to 11.9 percent in 2020Q4 from 10.1 percent in 2019Q4.
- Job losses concentrated in contact-dependent service sectors (e.g., accommodation, food services, recreational services); different from previous recessions that affected manufacturing and construction more.
- Uneven impacts by group:
  - Youth employment saw the largest adjustment: a 2.4 percentage point year‑on‑year increase in the unemployment rate.
  - Female unemployment and underutilization increased more due to job losses concentrated in contact-intensive service sectors.
  - Māori and Pacific people have been more affected than other ethnic groups (Budget Policy Statement 2021 analysis); Asian unemployment increased more sharply than the aggregate but remains below Māori and Pacific unemployment.
  - Workers with lower levels of education appear to be more severely impacted.
- The wage subsidy program cushioned the labor market impact:
  - International comparisons suggest countries with larger coverage of job retention programs tended to have more benign labor market outcomes relative to Okun’s law predictions.
  - In New Zealand, the youth unemployment rate increased by less than predicted by historical Okun’s law estimates, implying wage subsidies and broader policy response were particularly effective for young workers.
  - Male workers tended to benefit more from wage subsidies than female workers.
  - MSD administrative data show the proportion of young workers and male workers receiving wage subsidies was higher than other groups.

### Scarring Effects
- Concerns about potential labor-market scarring remain despite a quick recovery.
- Labor market conditions at graduation can have significant medium-term effects on outcomes.
- Cohort analysis using cross-country data (OECD unemployment rates by five-year age groups for 30 countries, 1960–2019) follows the Kawaguchi and Murao (2014) framework:
  - Regression specification links cohort unemployment at later ages to the unemployment faced during youth (ages 15–24), with age-group dummies for 25–29, 30–34, 35–39, and 40–44 and country-year fixed effects; standard errors clustered at the country level.
- Estimated scarring magnitudes:
  - A cohort that experiences a 1 percentage point higher unemployment rate during its youth suffers about a 0.2 percentage point higher unemployment rate when they are 25 to 29 years of age.
  - The same cohort suffers about a 0.05 percentage point higher unemployment rate when they are 30 to 34 years of age (computed as +0.208 − 0.150).
  - Beyond age 35, the effect of youth unemployment becomes insignificant.
- Scarring is larger for women and for countries with less flexible labor markets (regressions for males and females separately show stronger persistence for females).

*1nzlea2021002 - References _____________________________________________________________________________ 15*

### 0.21 percentage point higher for women, slightly more than the estimated 0.19 percentage point for

### 1nzlea2021002 - 0.21 percentage point higher for women, slightly more than the estimated 0.19 percentage point for

### Scarring from Youth Unemployment (Cross-Country Cohort Evidence)
- Youth unemployment coefficient (impact on future unemployment rates):
  - All: 0.208***
  - Male: 0.185***
  - Female: 0.212***
- Age interactions (reference age group: 25 to 29):
  - Youth unemployment * Age 30 to 34: -0.150*** (All); -0.143*** (Male); -0.142*** (Female)
  - Youth unemployment * Age 35 to 39: -0.212*** (All); -0.197*** (Male); -0.199*** (Female)
  - Youth unemployment * Age 40 to 44: -0.246*** (All); -0.227*** (Male); -0.229*** (Female)
- Employment Protection Legislation (EPL) interaction:
  - Youth unemployment * EPL: 0.156** (column 4)
  - Youth unemployment * EPL * Age 30 to 34: -0.120** (column 4)
  - Youth unemployment * EPL * Age 35 to 39: -0.208*** (column 4)
  - Youth unemployment * EPL * Age 40 to 44: -0.257*** (column 4)
- Conditional implication for New Zealand (given relatively mild EPL):
  - Previous recessions increased New Zealand’s youth unemployment rate by about 8-9 percentage points, translating to about a 0.4-0.5 percentage point higher unemployment rate at ages 25 to 29.
  - For a country with average EPL, the comparable impact would be about 2.5 percentage points at ages 25 to 29.

### Effectiveness of Active Labor Market Policies (ALMPs) — Cross-Country Regression Results
- Regression setup:
  - Dependent variable: change in unemployment rate.
  - ALMP types included (lagged average of previous two years): training; employment incentives; other policies (direct job creation, start-up incentives, rehabilitation).
  - Panel: OECD countries, sample period 1994-2019.
- Estimated coefficients (Table 2: effects on change in unemployment rate):
  - Training (lagged):
    - All: -0.68 (standard error 0.86)
    - Female: -0.65 (standard error 1)
    - Young: -3.61* (standard error 2.07)
  - Employment Incentives (lagged):
    - All: -2.59** (standard error 0.99)
    - Female: -3.02* (standard error 1.71)
    - Young: -5.03** (standard error 2.1)
  - Other ALMP (lagged):
    - All: 0.47 (standard error 1.16)
    - Female: -0.21 (standard error 1.32)
    - Young: -0.16 (standard error 2.46)
  - GDP growth:
    - All: -0.22*** (standard error 0.04)
    - Female: -0.19*** (standard error 0.04)
    - Young: -0.42*** (standard error 0.11)
  - GDP growth (lagged):
    - All: -0.11*** (standard error 0.02)
    - Female: -0.11*** (standard error 0.02)
    - Young: -0.22*** (standard error 0.03)
  - Observations: 545 (All, Female, Young)
  - Sample Period: 1994-2019
- Interpretation of results:
  - Employment incentives are particularly effective for the overall population and significantly so for young people in reducing subsequent unemployment.
  - Training programs show particularly strong estimated effectiveness for young workers (Young: -3.61*).
  - Other ALMPs (direct job creation, etc.) are not robustly effective across groups in these estimates.

### Design, Targeting, and Heterogeneity of ALMPs (Literature and Micro Evidence)
- Key empirical and literature findings:
  - Training is often effective for low-skilled workers and for reducing long-term unemployment.
  - Specific skill training tends to have larger effects than general training.
  - Training combined with job search assistance enhances outcomes.
  - Hiring subsidies are more impactful when targeted at small firms, low-wage workers, and jobs with rigid wages, and are more cost-effective if temporary during recessions.
  - Direct job creation is often less effective than training and hiring subsidies (Kluve 2010; Card and others 2018; Cahuc 2019; Greenberg and others 2003; Jacobson et al. 2005; Dorsett 2006).

### New Zealand: Current Policy Stance, Spending, and Recommended Policy Options
- Historical and pandemic-era spending on ALMPs:
  - Public spending on ALMPs in New Zealand declined from 0.5 percent of GDP in 2004 to 0.2 percent of GDP in 2017.
  - Since the onset of the pandemic, New Zealand expanded ALMPs including direct job creation, hiring subsidies, and free access to training, alongside a large-scale wage subsidy scheme.
- Pandemic-era income and labor market support measures (Table 3: amounts in NZ$):
  - Wage subsidies (NZ$ 14.8 bn)
  - Permanent increase in social benefits (NZ$ 2.4 bn)
  - COVID-19 leave and self-isolation support (NZ$ 0.1 bn)
  - Income Relief Payment (NZ$ 0.6 bn)
  - Increase in Winter Energy Payment (NZ$ 0.5 bn)
  - Job creation in environmental projects (NZ$ 1.1 bn)
  - Targeted Training and Apprenticeship Fund (NZ$ 0.3 bn)
  - MSD capacity expansion (NZ$ 0.4 bn)
  - Expansion of Flexi-Wage Subsidies (NZ$ 0.3 bn)
  - He Poutama Rangatahi – youth employment support (NZ$ 0.1 bn)
- Policy recommendations to address emerging labor market issues and inequalities:
  - Increase spending on ALMPs given historically low levels and declining trend pre-pandemic.
  - Expand coverage of training programs beyond narrowly targeted beneficiaries to support broader workers in transition, particularly low-skilled workers and the long-term unemployed.
  - Upgrade training programs to reflect industry needs (Reform of Vocational Education (RoVE) is a positive step).
  - Scale up support targeted at displaced low-wage workers if long-term unemployment persists (e.g., expand Flexi-Wage subsidies).
  - Strengthen job search assistance for cost-effective worker reallocation.
  - Review and reinforce quality of interventions; some programs may require improvement (MSD 2019).
  - Emphasize early activation to reduce scarring given that employment likelihood decreases with duration of unemployment.
  - Address inequality and gender gaps:
    - Targeted education support for Māori and Pacific peoples to promote human capital accumulation.
    - Expand childcare support to promote female labor force participation and help close the widened gender gap post-pandemic.

### Conclusions (Key Quantitative and Qualitative Takeaways)
- Pandemic impact and disparities:
  - The pandemic had a relatively mild yet uneven impact on New Zealand’s labor market, aggravating pre-existing disparities.
  - Before COVID-19, unemployment was significantly higher among youth, women, low-skilled workers, Māori and Pacific peoples.
  - Youth experienced the largest drop in employment during the pandemic.
- Scarring magnitude (cross-country cohort estimate reiterated):
  - A 1 percentage point increase in youth (ages 15–24) unemployment is associated with a 0.2 percentage point higher unemployment rate between ages 25 and 29.
  - Scarring effects are larger and more persistent for women (female point estimate in cohort regression: 0.212*** vs male 0.185***).
- Policy implication:
  - Continued macroeconomic support together with scale-up and careful targeting of ALMPs (training for young workers, hiring subsidies targeted to low-wage workers, expanded coverage and quality of training, stronger job search assistance) can mitigate distributional effects and reduce scarring while promoting long-term human capital accumulation.

*Source: https://www.imf.org/-/media/files/publications/cr/2021/english/1nzlea2021002.pdf*

### Annex II. Additional Results on Impact of Youth Unemployment

### Annex II. Additional Results on Impact of Youth Unemployment on Future Unemployment Rates

### Table II.1 — Regression results: Impact of youth employment rate on future employment rates
- Sample: 30 countries, 1960 to 2019. Data on employment rate at the country-cohort-age level from OECD. EPL = Employment Protection Legislation Index (OECD).
- Estimation outcomes (columns labeled (1) All, (2) Male, (3) Female, (4) EPL Interaction):
  - Youth employment rate:
    - (1) 0.170** (standard error reported as (0.0705))
    - (2) 0.169*** (standard error (0.0513))
    - (3) 0.195* (standard error (0.111))
    - (4) 0.0927 (standard error (0.0820))
  - Reference age group: 25 to 29.
  - Youth employment * Age dummy for 30 to 34:
    - (1) -0.0920** (0.0398)
    - (2) -0.142*** (0.0491)
    - (3) -0.0614 (0.0444)
    - (4) -0.0556 (0.0350)
  - Youth employment * Age dummy for 35 to 39:
    - (1) -0.133** (0.0519)
    - (2) -0.191*** (0.0664)
    - (3) -0.0872 (0.0565)
    - (4) -0.0868* (0.0465)
  - Youth employment * Age dummy for 40 to 44:
    - (1) -0.144** (0.0554)
    - (2) -0.220*** (0.0768)
    - (3) -0.0598 (0.0633)
    - (4) -0.106* (0.0540)
  - Youth employment * EPL:
    - (4) 0.131 (0.105)
  - Youth employment * EPL * Age dummy for 30 to 34:
    - (4) -0.0923* (0.0511)
  - Youth employment * EPL * Age dummy for 35 to 39:
    - (4) -0.155** (0.0598)
  - Youth employment * EPL * Age dummy for 40 to 44:
    - (4) -0.185** (0.0683)
  - Constant:
    - (1) 66.57*** (3.138)
    - (2) 73.60*** (2.656)
    - (3) 58.47*** (4.674)
    - (4) 70.81*** (3.768)
  - Observations: 14,406 (all columns)
  - R-squared:
    - (1) 0.940
    - (2) 0.930
    - (3) 0.932
    - (4) 0.948

### A. Introduction — Context and scope
- Key observations:
  - New Zealand’s households and corporates have so far weathered the COVID-19 shock relatively well due to:
    - Relative strength of balance sheets at the outset.
    - Strong cash flow support from expansionary fiscal and monetary policies.
    - Loan deferrals from sound and well-capitalized banks.
  - Inherent vulnerabilities remain, notably:
    - Corporate vulnerabilities in hospitality and transportation due to international border closure.
    - High household debt relative to advanced economies and banks’ concentrated exposure to residential mortgages (60 percent of total loans).
  - The full effect will be clearer after temporary support measures end; stress tests indicate household exposure to higher interest rates and corporate vulnerabilities from renewed revenue losses.
- Paper structure:
  - Sections B and C examine (i) COVID-19 impact on households by income and indebtedness; and (ii) vulnerability of NFC sector and SMEs vs larger firms. Section D discusses implications and policy support options.

### B. What Has Been the Impact on Household Balance Sheets?
- Fiscal and monetary support and near-term outcomes:
  - Real GDP contracted by 3 percent in 2020.
  - Wage subsidy: NZ$14.8 billion or 4.6 percent of GDP.
  - Cash rate lowered by 75 bps to 0.25 percent.
  - Household disposable income increase: 3 percent y/y in the year ended in September 2020 (from 4.8 percent y/y in December 2019).
  - Household deposit growth: close to 10 percent y/y in December 2020 (doubled).
  - Mortgage and SME six-month principal and interest repayment deferrals provided.
- Net worth and housing wealth:
  - Household net wealth was 938 percent of household disposable income in June 2020, an increase of 5.4 percent y/y.
    - Housing and land value: 4.2 percent y/y increase.
    - Net financial wealth: 1.3 percent growth.
  - House prices (REINZ): surged by 24 percent y/y in March 2021.
  - New Zealand’s increase in real house prices is close to the 90th percentile among advanced economies.
- Mortgage servicing and debt:
  - Interest payments on housing loans: 5.6 percent of household disposable income in September 2020 (from 6.2 percent in September 2019).
  - Household debt: 163 percent of household disposable income.
  - Mortgage repayment deferrals and restructuring supported ability to service mortgages.
- Heterogeneity and affordability:
  - Bottom half of income groups spent close to 30 percent of household expenditure on housing and household utilities.
  - By dwelling ownership:
    - 27 percent of renters spent more than 40 percent of household disposable income on housing costs.
    - 12 percent of owner-occupier households spent more than 40 percent.
- LVR, DTI, and lending patterns:
  - Temporary removal of LVR restrictions increased high LVR lending to investors.
  - Share of first-home buyers increased to 19 percent.
  - Loans with DTI larger than five expanded by 1.8 percentage point of total new loan commitment in first nine months of 2020.
  - High DTI (higher than five) mortgage lending by owner-occupiers gained share in 2020.
  - LVR restrictions were reinstated at pre-existing levels from March 2021 and further tightened for investor loans effective May 2021.
- Ongoing and potential vulnerabilities (monitoring warranted):
  - Repayment deferrals and restructuring:
    - Banks provided repayment deferrals to about 8 percent of mortgages outstanding at peak in May 2020.
    - Less than one percent of mortgage lending remained on deferral by early January 2021.
    - Around NZ$12.8 billion or 4.3 percent of mortgage lending restructured to interest-only loans.
  - Vulnerability to interest rate shocks:
    - A one percentage point increase in the mortgage rate is estimated to raise household interest servicing burden to about 7 percent of household disposable income, from 5.6 percent in September 2020.
    - Most mortgages are in 2 to 3-year fixed rates, so benchmark rate changes impact borrowers faster.
    - Simulation (principal and interest loan using average household disposable income and median national house prices): a one percentage point increase in mortgage rate would increase total principal and interest repayments from currently 36 percent to 41 percent of household disposable income, exceeding conventional affordability range of 25-40 percent.
  - Baseline/implicit interest rate table (reported values):
    - Implict Interest rate (percent): 3.8 4.8 1.0 (table headings indicated Sep-20(% change) and other columns; these values are presented in source text)
    - Interest payments (NZ$ bil): 11 14 26 (Percent of household disposable income) 5.6 7.1 1.5
    - Housing loan outstanding (NZ$ bil): 290
    - Household disposable income (NZ$ bil): 198
  - Alternative mortgage rate impacts (reported values):
    - Mortgage rate (percent) 1/: 3.3-1+1+2 (table formatting in source indicates scenarios: 3 6 3 2 4 1 4 6 — source text presents numbers 36 32 41 46; these are quoted from source in context as “36324146” under the Alternative table)
    - Note: source provides methodological notes: 1/ Average 1-year residential mortgage rate (end-of-period). 2/ Based on a principal and interest loan of 30-year maturity with a down payment of 20 percent of the house price, using national median house prices and estimated annual household disposable income in end-2020.
  - Vulnerabilities from rising house prices:
    - Higher house prices have partially offset benefits from lower mortgage rates for new home buyers; share of principal repayments in total mortgage repayments increased from 2019 to 2020.
  - Vulnerability of lower-income borrowers:
    - Median annual household income: about NZ$88,000 in June-2020.
    - Owner-occupiers with gross income at NZ$90,000 or less accounted for about 20 percent of owner-occupier loans.
    - About 23 percent of loans with DTI higher than five are concentrated in this group.
    - Two-thirds of owner-occupier loans are concentrated in the bottom five income groups.
  - Data limitations:
    - DTI and income levels of investor loans or debt secured on other properties are not published.

### C. How Vulnerable Are Non-Financial Corporations (NFCs)?
- Pre-pandemic corporate balance-sheet position:
  - Size of NFC sector (total assets): declined from peak 350 percent of GDP in 2009 to 306 percent of GDP in 2019.
  - Corporate leverage (debt-to-equity ratio): around 77 percent since 2015.
  - Cash and deposits: about 12 percent of financial assets.
  - Interpretation: relatively low leverage, profitable, and moderate liquidity compared internationally.
- Industry and firm-size heterogeneity:
  - Contact-intensive industries not generally more leveraged, with exceptions: retail trade and accommodation industries.
  - Small firms more profitable than larger firms overall; accommodation and food services rank low on profitability.
  - Firm-level data: close to one quarter of firms (16 percent of large firms and 30 percent of SMEs) have a liquidity ratio (cash cover of short-term borrowing) of less than one — corresponding to about 12 percent of total corporate debt.
  - Note on SME classification: paper uses threshold at 75th percentile of revenue for classification of SMEs due to sparse availability of firms’ employee number in Orbis data.
- Pandemic impact and policy support:
  - NFCs have so far weathered the pandemic downturn well due to large and prompt support measures.
  - Wage subsidy program:
    - Supported more than 60 percent of jobs through NZ$9.4 billion paid to about 1.5 million unique jobs.
    - Proportion of jobs in accommodation and food services supported by wage subsidy: 93 percent.
  - Other support: loan and mortgage deferral programs; Government’s COVID-19 Small Business Cashflow Loan (SBCS) and Business Finance Guarantee Schemes (BFGS).
  - Quick adaptability of some small businesses to tech-driven solutions aided resilience.
- SME performance and recovery indicators:
  - Xero Small Business Insights (XSBI) indicators:
    - SMEs revenue grew by 1.4 percent y/y in December 2020 (compared to Australia small firms at -2.7 percent).
    - Small business jobs grew by 3.2 percent y/y in December 2020.
    - All industries recorded job numbers higher than pre-crisis levels in March 2020, except hospitality:
      - Hospitality jobs remained below pre-crisis levels, down by 4.9 percent since March 2020.
    - Tourism-dependent Queenstown revenue down 21.4 percent y/y in December 2020.

*International Monetary Fund — Annex II. Additional Results on Impact of Youth Unemployment on Future Unemployment Rates (source content provided)*

### 14.      Business bankruptcies have declined substantially in 2020 amidst the pandemic

### 14.      Business bankruptcies have declined substantially in 2020 amidst the pandemic

### Business bankruptcies and drivers
- Business bankruptcies declined substantially in 2020 during the pandemic recession.
- The decline is observed in other advanced economy peers such as Australia.
- Primary drivers:
  - Government COVID-19 response measures, including banks’ loan deferral scheme and the government’s Business Debt Hibernation scheme which allowed businesses to place their existing debts on hold.
- Outlook:
  - Corporate bankruptcies are expected to rise as support measures are phased out in 2021 and administrative processes that delayed insolvency are completed.

### Default probabilities and sectoral impacts (EDF analysis)
- Expected Default Frequency (EDF) by Moody’s measures the probability a firm with publicly traded equity will default over a specific period (typically one year) when market value of assets falls below liabilities payable.
- Key observations:
  - Overall EDF in New Zealand rose at the outset of the COVID-19 crisis, though less markedly than most other advanced economies.
  - EDFs have largely come down since the initial rise, but some industries remain disproportionately affected.
  - The default probability of New Zealand firms in tourism has risen with the continued closure of the international border.
  - EDFs in selected other sectors remain above pre-crisis levels.

### Corporate debt at risk (Box 1 — Corporate Vulnerability Index / Debt at Risk)
- Framework:
  - Corporate vulnerability index (CVI), based on Debt at Risk (DaR), estimates total debt outstanding associated with firms deemed financially vulnerable across four leading-indicator aspects of balance-sheet vulnerability.
  - Extends DaR to multiple indicators to capture intensity of debt at risk (share of debt vulnerable by one or more indicators simultaneously).
- Selected thresholds for corporate financial vulnerability:
  - Debt service capacity: DaR1: Interest coverage ratio (ICR<1)
  - Leverage: DaR2: Debt to Asset (> 90th percentile within the same industry in 2010-19)
  - DaR3: Net debt (total debt minus cash and cash equivalent) to EBIT (> 90th percentile within the same industry in 2010-19)
  - Rollover/Liquidity: DaR4: Quick ratio, current assets minus inventories over current liabilities (<10th percentile within the same industry in 2010-19)
  - Profitability: DaR5: Return on assets (net income/total assets) (<10th percentile within the same industry in 2010-19)
- Findings:
  - About one quarter of total debt in the sample could be at risk based on liquidity vulnerability.
  - Over one half of SMEs’ debt could be at risk due to higher leverage.
  - About one tenth of SME debt could be at risk measured by three indicators simultaneously.
  - Total debt at risk declined substantially according to more than one indicator.
  - In contrast, about one-tenth of SME debt considered at risk by three indicators.

### Risk of liquidity shortages and equity shortfalls (Box 2 — simulation)
- Simulation setup:
  - Based on firm-level Orbis data and OECD (2020) methodology.
  - Scenario I: 50 percent decline in turnover.
  - Scenario II: 75 percent decline in turnover.
  - Horizons: one- to three-months.
  - Assumptions: elasticity of material costs to sales c = 0.8; corporate tax rate 28 percent; operating revenue, wages, interest payments treated as one-twelfth of annual data; depreciation & amortization used as proxy for debt principal repayment.
  - Omissions: liquidity shortfall simulation omits about 8 percent of firms with negative cash and cash equivalent; equity shortfall simulation excludes about 9-10 percent of firms with zero or negative equity.
- Simulation results:
  - Under a 50 percent decline in turnover, about 5 percent of firms will have liquidity shortfalls after three months.
  - Under a 75 percent fall in turnover for three months:
    - about 12 percent of all firms,
    - 10 percent of SMEs, and
    - 18 percent of SMEs in services
    could encounter liquidity shortfalls.
  - The share of firms expected to become insolvent ranges between 1-7 percent under the two stress scenarios.
  - Services firms are far more prone to liquidity shortfalls and a higher share face equity shortfalls under a severe drop in turnover.
  - Estimates are likely lower bounds due to sample bias toward healthier firms, use of pre-COVID-19 balance sheet data, and likely less stringent balance-sheet insolvency criteria.

### Policy implications and recommendations
- Timing and targeting of support:
  - Phasing out policy support needs to be well-timed, with refocusing on viable loans and solvent firms.
  - Blanket emergency cashflow relief was appropriate at the pandemic’s depth; next phase warrants a more targeted approach to firms that are viable.
  - Support measures such as the SBCS are designed to be temporary and are capped at low levels to reduce risks of supporting unviable firms.
- Recommended instruments and focus:
  - Greater reliance on bank lending channel, supplemented with targeted credit guarantee schemes, to ensure credit reaches viable projects that generate adequate investment return.
  - Apply a higher expected rate of return for supporting firms in severely affected sectors such as tourism to foster restructuring and resource reallocation.
  - SME financing and targeted approaches: improve design and access to SME financing, guaranteed credit, and small-scale cashflow grants to limit scarring and preserve viable SMEs.
- Macroprudential and housing-related policy:
  - Actively use macroprudential policy to safeguard financial stability, including tools to potentially limit households’ DTI exposure and secure adequate debt serviceability buffers at the RBNZ’s disposal.
  - Address structural vulnerabilities from banks’ concentration in housing loans.
  - Implement comprehensive policy measures to support housing affordability, including ensuring a swift and adequately large supply response.
- Data and monitoring:
  - Improve timely availability of data and consider additional data collection to fill gaps.
  - On housing loans: make available DTI and income profile of investors and size/distribution of interest-only loans.
  - On NFCs: increase frequency and timeliness of the Enterprise Survey; incorporate monthly indicators of SME performance from commercial sources.

### Key statistics and figures (preserved verbatim)
- “one quarter of total debt in the sample could be at risk based on liquidity vulnerability.”
- “one half of SMEs debt could be at risk due to higher leverage.”
- “one tenth of SME debt could be at risk measured by three indicators simultaneously.”
- Simulation outcomes:
  - “about 5 percent of firms will have liquidity shortfalls after three months.” (50 percent turnover decline)
  - “about 12 percent of all firms, 10 percent of SMEs, and 18 percent of SMEs in services could encounter liquidity shortfalls.” (75 percent turnover decline)
  - “The share of firms expected to become insolvent is smaller, ranging between 1-7 percent under the two stress scenarios.”
- SBCS details:
  - “extended until the end of 2023”
  - “provides small businesses a 5-year loan under eligible criteria which include at least a 30 percent decline in revenue over a 14-day period in the previous 6 months, compared with the same 14-day period a year ago.”
  - “The maximum amount of loan is NZ$10,000 plus NZ$1,800 per full-time-equivalent employee.”
- Example support measure:
  - “a NZ$400 million Tourism Recovery Package under the Budget 2020” allocated to protect strategic tourism businesses, drive domestic tourism through regional events and lift digital capability in the tourism industry.
- Corporate tax rate used in simulation: “28 percent.”
- Material cost elasticity assumption: “c is assumed at 0.8.”

*Source: 1nzlea2021002 - 14.      Business bankruptcies have declined substantially in 2020 amidst the pandemic (IMF staff, content provided in source PDF).*

### 2.      This paper analyzes the medium-term impact of the pandemic on New Zealand’s

### 1nzlea2021002 - 2.      This paper analyzes the medium-term impact of the pandemic on New Zealand’s

### B. Drivers of Growth in New Zealand Before the Pandemic
- Average potential output growth in 2017-19: 2.9 percent.
- Contribution of productivity (TFP) to potential output growth by 2019: 0.2 percentage point.
- Growth drivers and constraints:
  - Strong working age population growth driven by a surge in net migration.
  - Capital accumulation was weak relative to population growth, producing slower capital deepening and sectoral capital shallowing that weighed on labor productivity growth.
- Net migration contribution: net migration accounted for about 70 percent of working age population growth before the border closure.

### C. Lessons from Previous Recessions in Advanced Economies
- Historical average medium-term output loss after recessions: 4½ percent below pre-recession trend (after 5 years).
- For large recessions (top quartile; peak-to-trough GDP decline above 4.25 percent):
  - Output remains 11 percent below pre-recession trend.
  - Total factor productivity (TFP) almost 6 percent below trend.
  - Capital stock about 7 percent below trend.
- New Zealand experience:
  - Output per capita remained nearly 20 percent below the pre-GFC trend by end-2019.
  - Slower capital accumulation, shorter hours worked, and weaker productivity growth contributed to the medium-term output loss.

### D. Impact of the Pandemic on New Zealand’s Medium-Term Output
- Overall assessment:
  - The pandemic is expected to have a sizable and lasting impact on medium-term potential output, with all main drivers affected: labor inputs, capital accumulation, and productivity growth.
  - Potential output is expected to remain about 3¾ percent below pre-COVID trend by 2026.
  - In labor productivity terms, potential level is expected to remain about 1.1 percent below the pre-COVID trend in the medium term.
- Labor input:
  - Reduced net migration is the largest channel of loss.
  - Net migration decline is expected to contribute 1.5 percentage point to potential output losses in the medium term.
  - Increase in NAIRU due to skill mismatches contributes 0.1 percentage point to potential output losses.
  - Total reduced labor input contribution: 1.6 percentage points.
  - Scenario assumption for net migration path: net migration remains below pre-COVID trend (approximated by average over 2015-19) by 90 percent in 2021, by 37 percent in 2022, and by 12 percent in 2023, before reaching pre-COVID level in 2024.
- Productivity:
  - Aggregate productivity expected to slow, with within-sector productivity slowdown partially offset by reallocation to more productive sectors.
  - Baseline contribution of productivity slowdown to potential output loss: 0.8 percentage point.
- Capital accumulation:
  - Expected slowdown in capital accumulation contributes 1.3 percentage points to potential output loss.
  - Trend capital accumulation is expected to decline due to reduced labor supply, weaker productivity growth, and heightened uncertainty affecting investment.
- Scenarios illustrating uncertainty around productivity:
  - Upside scenario:
    - Firms adopt new technologies and ICT investment; shift to a more productive, digitalized economy.
    - Moderate within-sector productivity gains of about 0.5 percent in the medium term.
    - Potential output loss in 2026: 1½ percent lower than pre-COVID trends.
    - Labor productivity under upside: 1.0 percent higher than the pre-COVID trend.
  - Downside scenario:
    - Weaker productivity spillovers and lost innovation due to muted migration and subdued innovative investment.
    - Slowdown in productivity within sectors about twice larger than the baseline.
    - Potential output in 2026 under downside: five percent lower than pre-COVID trends.
    - Labor productivity under downside: 1.5 percent lower than the pre-COVID trend.

### E. Policy Options to Address the Medium-Term Fallout of the Pandemic
- Overarching guidance:
  - Maintain adequate fiscal and monetary policy stimulus until private demand recovery is firmly entrenched to mitigate output losses and limit scarring.
  - Begin structural reforms now; reforms take significant time to affect medium-term growth and living standards.
- Productivity-enhancing structural reforms and measures:
  - Product market regulations:
    - Scope for improvements in some areas and upstream network sectors by streamlining administrative burdens and simplifying regulations.
    - Where state ownership is inefficient, consider reducing public ownership or improving state-owned enterprises.
    - Pro-competitive reforms to encourage firm innovation and investment.
  - Insolvency framework:
    - Improve insolvency regime to facilitate resource reallocation and lift productivity.
    - International indicators point to room for improvement; insolvency resolution takes longer and can be more costly compared to some advanced economies.
    - Procedures tailored to SMEs could expedite restructuring.
  - Innovative investment:
    - R&D spending has been historically low relative to peers (both government and business).
    - Scale up government spending on R&D and expand innovative capacities of Crown Research Institutes.
    - Expand and strengthen the R&D tax credit scheme introduced in FY2020.
    - Global collaboration to help technology diffusion.
  - Foreign direct investment (FDI) and global connectivity:
    - FDI in New Zealand has been relatively low as a percent of GDP.
    - Ongoing reform of the Overseas Investment Act should focus on promoting FDI by reducing FDI-related regulation while maintaining control over sensitive areas.
    - Aim to attract high quality FDI to generate productivity spillovers.
  - Trade facilitation:
    - Further improve areas such as advance ruling; automation (e.g., electronic signature); information availability (e.g., customs hotlines); and external border agency cooperation.
    - Reduce time and costs associated with the logistical process of exporting and importing goods.

*Prepared by IMF staff based on the chapter content.*

### 17.      Capital accumulation. To promote capital accumulation, New Zealand should scale up

### 17. Capital accumulation.

### Infrastructure gap and investment
- Recommendation: scale up infrastructure investment, focusing on closing gaps in telecommunications, electricity, and transportation.
- Assessment:
  - The planned increase in infrastructure investment is a welcome development.
  - The New Zealand Infrastructure Commission can play a central role in planning and pipeline management.
  - The infrastructure spending gap is estimated to be between 8 and 23 percent of GDP.
- Regional impact: Infrastructure investment would also reduce regional dispersion in New Zealand.

### Business investment and financial policies
- Recommendation: encourage FDI and incentivize business investment; scale up government policies that aid business cash flow if downside risks materialize.
- Government measures noted as appropriate and scalable:
  - business tax loss carry-back scheme
  - Small Business Cash Flow Scheme
  - Business Finance Guarantee Scheme
- Complementary reforms: Product market reforms would boost business investment through promoting competition.

### Labor market and human capital
- Recommendation: facilitate sectoral reallocation and support workers severely affected by the pandemic, including youth, to minimize scarring and promote long-term human capital accumulation.
- Policy direction:
  - Transition from wage subsidies to active labor market policies such as training and targeted hiring subsidies.
  - The scale and scope of active labor market policies can be further expanded, as spending in this area has been relatively low in New Zealand in international comparison.

### Firm-level determinants of investment (empirical findings)
- Model: Tobin’s Q model augmented with firm-level financial variables and uncertainty; unbalanced annual panel from 1990-2019 with total observations of 1,443 (financial sector firms excluded).
- Key estimated coefficients (dependent variable: Investment-to-Lagged Capital):
  - Cost of Debt (-1): -0.124 (0.171)
  - Debt (-1): -0.157*** (0.046)
  - Liquidity (-1): 0.013* (0.007)
  - Uncertainty (-1): -0.069*** (0.034)
  - Tobin's Q (-1): 0.044*** (0.007)
- Model statistics:
  - R-squared: 0.058
  - Sample Period: 1990-2019
  - Number of Observation: 1,443
- Interpretation: Financial variables such as cash flow and uncertainty play an important role in firms’ investment decisions (Annex I). At the aggregate level, uncertainty and financial conditions exert significant influence on business investment; at micro level results are mixed across studies.

### Effect of product market reforms on investment
- Method: panel local projection model focused on advanced economies; sample of 18 advanced economies from 1990 to 2016 (see Annex II).
- Finding: Product market reforms increase business investment (responses of business investment to product market reforms reported in Section E).

*Source: 1nzlea2021002 - 17. Capital accumulation. To promote capital accumulation, New Zealand should scale up*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2021/english/1nzlea2021002.pdf_