## wpiea2021131-print-pdf

## Source details

**Canonical URL:** [wpiea2021131-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021131-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021131-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021131-print-pdf.pdf.json)

---

### Introduction and context
- The global recession triggered by the COVID-19 pandemic caused a sharp contraction in economic activity and a significant rise in unemployment and underemployment.
- In 2020, 8.8 percent of global working hours were lost relative to the fourth quarter of 2019, equivalent to 255 million full-time jobs—approximately four times greater than during the global financial crisis in 2009 (ILO, 2021).
- The pandemic disproportionately affected low-skilled workers who could not telework and lower-middle-income countries.
- With the expected recovery from 2021, unemployment rates across advanced and emerging market economies are expected to remain elevated (IMF, 2021); full-time job losses will remain above 130 million jobs, relative to the fourth quarter of 2019 (ILO, 2021).
- Persistent weakness in the labor market raises the prospect of hysteresis effects and increased long-term unemployment (Blanchard and Summers, 1986).

### Rationale for public investment in the recovery
- Well-designed recovery packages that include ramped-up spending on public investment can limit long-term labor market scarring.
- Public investment spending multipliers are particularly large during recessions (Auerbach and Gorodnichenko, 2013).
- Historical short-term "job multipliers" referenced:
  - ARRA (2009): about six to eight jobs per US$1 million spent (Wilson, 2012; Garin, 2019; Ramey, 2020).
  - Civilian Conservation Corps (1933–1942): negligible short-term employment impact (Aizer and others, 2020).

### Scope and approach of the paper
- Focus: quantifying the direct employment effect of infrastructure development and maintenance in sectors where social distancing is less relevant during the post-COVID-19 recovery.
- Explicitly analyzed sectors: electricity, roads, schools and hospitals, and water and sanitation in advanced, emerging, and low-income developing economies.
- Rationale: these sectors account for the lion's share of public investment in infrastructure and align with IMF assessments of additional spending required to meet United Nations’ Sustainable Development Goals (SDGs) by 2030 (Gaspar and others, 2019).
- Conceptual link: public infrastructure projects are usually performed through contractors; public investment spending flows to company revenues and subsequently to payroll and employment. Construction firms are assumed to behave similarly in employment efficiency regardless of revenue source, making job elasticity to revenue a first-order estimator of direct job creation at the firm level.
- Reverse causality from employment to revenues considered less likely in construction firms, where revenues are driven by available contracts and competitive bidding.

### Key quantitative findings and estimates
- One percentage point of global GDP in additional spending on public investment can create more than seven million jobs worldwide through its direct employment effects alone.
  - This is about 5.4 percent of the full-time equivalent jobs lost in 2021 relative to the fourth quarter of 2019 (ILO, 2021).
- The total labor impact of one percent of GDP—through direct and indirect macroeconomic effects—is estimated at 20–33 million jobs (IMF, 2020b).
- Job-content estimates applied to a 1 percent of GDP increase in public investment:
  - 4.9 per US$1 million invested for AEs (unweighted average) applied to an increase in investment worth 1 percent of GDP in AEs (ca. US$500 billion in 2020; cf. Appendix 1).
  - 14.8 per US$1 million for EMEs applied to 1 percent of the GDP of EMEs (ca. US$320 billion).
- Sector- and country-income-level direct job creation per US$1 million (midpoints and ranges reported):
  - Advanced Economies (AEs) overall illustrative averages: 6.6 jobs in electricity (low labor-intensity, low mobility scenario reported as 6.6).
  - Emerging Market Economies (EMEs): 10.4–17.2 jobs (range reported for intermediate scenarios across sectors).
  - Low-Income Developing Countries (LIDCs): 16–30.2 jobs (range reported across sectors and scenarios).

### Data and methodology
- Data sources: Compustat and Orbis, matched for the span 1999–2017. Revenues adjusted to constant 2015 U.S. dollars using GDP deflators.
- Sample after filtering:
  - AEs: 43,485 observations for 5,123 firms in 27 advanced economies.
  - EMEs: 4,095 observations for 556 firms in 14 emerging market economies.
  - No firm-level data from LIDCs; LIDC estimates are linear extrapolations from AE and EME estimates.
- Firms retained: at least five annual observations; outliers by revenue and employment dropped.
- Regression specification (equation 1): L_{i,t} = α + β R_{i,t} S_i + X + ε, where L and R are employment and revenue for firm i at time t, S is vector of sector dummies, and X are country (Model 1) or firm (Models 2–4) fixed effects. Regressions clustered at firm level. βs are change in employment per US$1 million of revenue.
- Supply-chain pass-through parameter λ ∈ (0,1) used to scale direct contractor job effects into total sectoral job creation via geometric series: Jobs_s = β_s / (1 − λ). Assumed λ between 35 and 65 percent, midpoint 50 percent.

### Regression results (selected coefficients and sample statistics)
- Sample summary statistics (1999–2017, firms with ≥5 observations):
  - AEs (27 countries, 43,485 observations): Mean revenue: 11.0 (millions of 2015 U.S. dollars). Mean employees: 45.1.
  - EMEs (14 countries, 4,095 observations): Mean revenue: 7.7 (millions of 2015 U.S. dollars). Mean employees: 123.4.
- Advanced Economies (Model 2: firm fixed effects, preferred specification) — jobs per US$1 million revenue:
  - Electricity: 2.363 (std. err. 0.206) — ***.
  - Roads: 1.722 (std. err. 0.128) — ***.
  - Schools and hospitals: 0.850 (std. err. 0.125) — ***.
  - Water and sanitation: 1.206 (std. err. 0.235) — ***.
- Emerging Market Economies (Model 2: firm fixed effects) — jobs per US$1 million revenue:
  - Electricity: 7.406 (std. err. 3.175) — **.
  - Roads: 2.287 (std. err. 0.679) — ***.
  - Schools and hospitals: 4.578 (std. err. 1.788) — **.
  - Water and sanitation: 4.965 (std. err. 2.593) — *.
- Robustness: Model 3 excludes 2008–2009; Model 4 uses revenue increases only; results similar to Model 2, indicating job elasticity to revenues not cycle-dependent.

### Job-content multipliers by sector, labor mobility, and labor intensity (selected matrix entries from Table 4)
- Advanced Economies (Jobs per US$1 million):
  - Electricity: High mobility/high intensity 12.2; Medium mobility/medium intensity 6.6; Low mobility/low intensity 3.6.
  - Roads: Medium/medium 6.2; Low/low 2.6.
  - Schools and hospitals: Medium/medium 4.4; Low/low 1.3.
  - Water and sanitation: Medium/medium 5.6; Low/low 1.9.
- Emerging Market Economies (Jobs per US$1 million):
  - Electricity: High/high 23.2; Medium/medium 16.2; Low/low 11.4.
  - Roads: High/high 23.4; Medium/medium 16.4; Low/low 3.5.
  - Schools and hospitals: High/high 22.2; Medium/medium 15.5; Low/low 7.0.
  - Water and sanitation: High/high 35.1; Medium/medium 24.6; Low/low 7.6.
- Low-Income Developing Countries (extrapolated jobs per US$1 million):
  - Electricity: High/high 37.9; Medium/medium 26.6; Low/low 17.5.
  - Roads: High/high 31.4; Medium/medium 22.0; Low/low 7.8.
  - Schools and hospitals: High/high 35.6; Medium/medium 24.9; Low/low 14.3.
  - Water and sanitation: High/high 53.9; Medium/medium 37.8; Low/low 17.8.
- Illustrative benchmarks: for intermediate labor mobility and labor intensity, 3 jobs in schools and hospitals to 6.6 jobs in electricity per US$1 million in AEs; 16 jobs in roads to 30.4 jobs in water and sanitation in LIDCs.

### Green investment and R&D implications
- Green investment job ranges (literature-based):
  - Green electricity: 5–10 jobs per US$1 million.
  - Efficient new buildings (e.g., schools and hospitals): 2.4–12.5 jobs per US$1 million.
  - Green water and sanitation: 5.7–14 jobs per US$1 million.
- Conclusion: job creation multipliers for green investment resemble the high labor-intensity estimates in Table 4.
- R&D public spending (OECD sample) point estimates:
  - Government R&D: 4.837 jobs per US$1 million (standard error 1.281).
  - Higher education R&D: 10.99 jobs per US$1 million (standard error 3.970).
  - Business R&D: 10.55 jobs per US$1 million (standard error 4.609).
  - Non-profit R&D: 4.477 jobs per US$1 million (standard error 2.020).
- Green R&D estimated at 3–8 jobs per US$1 million (IEA, 2020).

### Caveats, limitations, and interpretation
- Dataset illustrative, not statistically representative; most sampled companies are medium-size, unlisted, and have several years of audited financial statements.
- No firm-level data for LIDCs; LIDC estimates are linear extrapolations from AE and EME point estimates.
- Limited data granularity prevents disentangling new investment versus maintenance, part-time versus full-time, skilled versus unskilled, and imported versus local labor.
- Two inefficiencies not captured by reduced-form job estimates:
  - (i) low productivity at public administration (budget-to-contractor losses due to poor project design, red tape, corruption).
  - (ii) low productivity at contractor level (jobs may not translate into public goods).
- β_s coefficients reflect average job content across contracting layers since upstream/downstream firms cannot be disaggregated by industrial code.
- Wage levels, job quality (full vs part-time), and specific green investment types are not unfolded due to data limitations; fixed effects absorb some country- and firm-specific heterogeneity.

### Welfare, political economy, and state-capacity considerations
- AEs generally have the human and physical capital and can borrow at record-low interest rates to scale up public investment; EMEs and LIDCs face higher borrowing costs, elevated public debt, and capacity constraints.
- SOEs often undertake public investment and have on average higher job intensity than private firms. In OECD countries, SOEs represent on average 4.7 percent of the labor force versus 15.8 percent by the general government.
- Job intensity for SOEs is found to be 30 percent higher than for their private counterparts (Baum and others, 2019).
- Political economy and institutional capacity constraints in EMEs and LIDCs complicate scaling up public investment despite potentially higher employment multipliers.
- Public-private partnerships are not a panacea when fiscal space is limited; private-sector investment decisions are driven by profitability and enforcement of property rights.

### Policy implications and suggested analytical tools
- Use reported estimates (e.g., Table 4) to assess employment impact and trade-offs of alternative uses of fiscal space:
  - Trade-offs between short-term current spending and long-term capital spending.
  - Trade-offs between social and physical capital sectors.
  - Enhancing employment impact via higher labor mobility (e.g., training, flexible arrangements).
  - Enhancing employment impact via higher labor content (e.g., infrastructure maintenance, green investment).
- Design green recovery policies carefully to avoid backlash and account for distributional effects and required skills.
- Consider correcting for budget-to-contractor losses using Public Investment Management Assessment (PIMA) data to adjust upward bias in job estimates.
- Future research extensions noted: quantification of employment impact by country; employment impact in digital infrastructure; employment impact by green versus brown investment as industrial classifications evolve.

*wpiea2021131-print-pdf - References .........................................................................................................*

### References .............................................................................................................

### wpiea2021131-print-pdf - References .............................................................................................................

### Introduction and context
- The global recession triggered by the COVID-19 pandemic caused a sharp contraction in economic activity and a significant rise in unemployment and underemployment.
- In 2020, 8.8 percent of global working hours were lost relative to the fourth quarter of 2019, equivalent to 255 million full-time jobs—approximately four times greater than during the global financial crisis in 2009 (ILO, 2021).
- The pandemic disproportionately affected low-skilled workers who could not telework and lower-middle-income countries.
- With the expected recovery from 2021, unemployment rates across advanced and emerging market economies are expected to remain elevated (IMF, 2021); full-time job losses will remain above 130 million jobs, relative to the fourth quarter of 2019 (ILO, 2021).
- Persistent weakness in the labor market raises the prospect of hysteresis effects and increased long-term unemployment (Blanchard and Summers, 1986).

### Rationale for public investment in the recovery
- Well-designed recovery packages that include ramped-up spending on public investment can limit long-term labor market scarring.
- Public investment spending multipliers are particularly large during recessions (Auerbach and Gorodnichenko, 2013).
- Historical estimates of short-term "job multipliers" from fiscal stimulus vary:
  - The American Recovery and Reinvestment Act (ARRA) of 2009 is estimated to have yielded about six to eight jobs per US$1 million spent in the short term (Wilson, 2012; Garin, 2019; and Ramey, 2020).
  - The Civilian Conservation Corps (1933–1942) had negligible short-term employment impact (Aizer and others, 2020).

### Scope and approach of the paper
- Focus: quantifying the direct employment effect of infrastructure development and maintenance in key sectors for the post-COVID-19 recovery phase, where social distancing is less relevant.
- Explicitly analyzed sectors: electricity, roads, schools and hospitals, and water and sanitation in advanced, emerging, and low-income developing economies.
- Rationale for sector choice: these sectors account for the lion's share of public investment in infrastructure and align with IMF assessments of additional spending required to meet United Nations’ Sustainable Development Goals (SDGs) by 2030 (Gaspar and others, 2019).
- Public infrastructure projects are usually performed through contractors (state-owned or private), so public investment spending flows to company revenues and subsequently to payroll and employment.
- Construction companies’ revenues come from public and private contracts; construction firms are argued to behave similarly in employment efficiency regardless of revenue source, making job elasticity to revenue a first-order estimator of direct job creation impact of public investment at the firm level.
- Reverse causality from employment to revenues is considered less likely in construction firms, where revenues are driven by available contracts and competitive bidding.

### Key quantitative findings and estimates
- One percentage point of global GDP in additional spending on public investment can create more than seven million jobs worldwide through its direct employment effects alone.
  - This is about 5.4 percent of the full-time equivalent jobs lost in 2021 relative to the fourth quarter of 2019 (ILO, 2021).
- The total labor impact of one percent of GDP—through direct and indirect macroeconomic effects—is estimated at 20–33 million jobs (IMF, 2020b).
- Historical/empirical short-term job estimates referenced:
  - ARRA (2009): about six to eight jobs per US$1 million spent.
- Additional noted (incomplete) estimate in source text: US$1 million of public spending on infrastructure can create 3–

*Italic source attribution: wpiea2021131-print-pdf - References .........................................................................................................*

### 6.6 jobs in advanced economies, 10.4–17.2 jobs in emerging market economies, and 16–30.2 jobs

### wpiea2021131-print-pdf - 6.6 jobs in advanced economies, 10.4–17.2 jobs in emerging market economies, and 16–30.2 jobs

### Key findings
- Estimated direct job creation per US$1 million of public investment (midpoint and ranges reported elsewhere in the paper):  
  - 6.6 jobs in advanced economies (AEs) for electricity (low labor-intensity, low mobility scenario reported as 6.6 in Table 4).  
  - 10.4–17.2 jobs in emerging market economies (EMEs) (range reported for intermediate scenarios across sectors in Table 4).  
  - 16–30.2 jobs in low-income developing countries (LIDCs) (range reported across sectors and scenarios in Table 4).  
- Green investments have higher employment impacts; job creation multipliers for green investment resemble the high labor-intensity estimates in Table 4.
- Employment impact is inversely correlated with country income level: construction of hospitals and schools is less labor-intensive in AEs than in EMEs and LIDCs.
- R&D public spending generates jobs in R&D with point estimates (OECD sample):  
  - Government R&D: 4.837 jobs per US$1 million (standard error 1.281).  
  - Higher education R&D: 10.99 jobs per US$1 million (standard error 3.970).  
  - Business R&D: 10.55 jobs per US$1 million (standard error 4.609).  
  - Non-profit R&D: 4.477 jobs per US$1 million (standard error 2.020).

### Data and methodology
- Data sources: Compustat and Orbis, matched for the span 1999–2017. Revenues adjusted to constant 2015 U.S. dollars using GDP deflators.
- Sample after filtering:  
  - AEs: 43,485 observations for 5,123 firms in 27 advanced economies.  
  - EMEs: 4,095 observations for 556 firms in 14 emerging market economies.  
  - No firm-level data from LIDCs; LIDC estimates are linear extrapolations from AE and EME estimates.
- Firms retained: at least five annual observations; outliers by revenue and employment dropped.
- Sector coverage: electricity, roads, schools and hospitals (construction of institutional buildings), water and sanitation.
- Regression specification (equation 1): L_{i,t} = α + β R_{i,t} S_i + X + ε, where L and R are employment and revenue for firm i at time t, S is vector of sector dummies, and X are country (Model 1) or firm (Models 2–4) fixed effects. Regressions clustered at firm level. βs are change in employment per US$1 million of revenue.
- Supply-chain pass-through parameter λ ∈ (0,1) used to scale direct contractor job effects into total sectoral job creation via geometric series: Jobs_s = β_s / (1 − λ). Assumed λ between 35 and 65 percent, midpoint 50 percent.

### Sample summary statistics (1999–2017, firms with ≥5 observations)
- Advanced Economies (27 countries, 43,485 observations):  
  - Mean revenue: 11.0 (millions of 2015 U.S. dollars).  
  - Mean employees: 45.1.
- Emerging Market Economies (14 countries, 4,095 observations):  
  - Mean revenue: 7.7 (millions of 2015 U.S. dollars).  
  - Mean employees: 123.4.

### Regression results (selected coefficients from Table 3)
- Advanced Economies (Model 2: firm fixed effects, preferred specification): jobs per US$1 million revenue:  
  - Electricity: 2.363 (std. err. 0.206) — statistically significant at 1% (***).  
  - Roads: 1.722 (std. err. 0.128) — ***.  
  - Schools and hospitals: 0.850 (std. err. 0.125) — ***.  
  - Water and sanitation: 1.206 (std. err. 0.235) — ***.
- Emerging Market Economies (Model 2: firm fixed effects): jobs per US$1 million revenue:  
  - Electricity: 7.406 (std. err. 3.175) — **.  
  - Roads: 2.287 (std. err. 0.679) — ***.  
  - Schools and hospitals: 4.578 (std. err. 1.788) — **.  
  - Water and sanitation: 4.965 (std. err. 2.593) — *.
- Robustness: Model 3 excludes 2008–2009 (Global Financial Crisis); Model 4 uses revenue increases only; results similar to Model 2, indicating job elasticity to revenues not cycle-dependent.

### Job-content multipliers by sector, labor mobility, and labor intensity (selected entries from Table 4)
- Advanced Economies (Jobs per US$1 million):
  - Electricity: High mobility/high intensity 12.2; Medium mobility/medium intensity 6.6; Low mobility/low intensity 3.6. (Table lists specific matrix entries including 6.6 as low intensity/low mobility entry.)
  - Roads: Medium/medium 6.2; Low/low 2.6.
  - Schools and hospitals: Medium/medium 4.4; Low/low 1.3.
  - Water and sanitation: Medium/medium 5.6; Low/low 1.9.
- Emerging Market Economies (Jobs per US$1 million):
  - Electricity: High/high 23.2; Medium/medium 16.2; Low/low 11.4.
  - Roads: High/high 23.4; Medium/medium 16.4; Low/low 3.5.
  - Schools and hospitals: High/high 22.2; Medium/medium 15.5; Low/low 7.0.
  - Water and sanitation: High/high 35.1; Medium/medium 24.6; Low/low 7.6.
- Low-Income Developing Countries (extrapolated jobs per US$1 million):
  - Electricity: High/high 37.9; Medium/medium 26.6; Low/low 17.5.
  - Roads: High/high 31.4; Medium/medium 22.0; Low/low 7.8.
  - Schools and hospitals: High/high 35.6; Medium/medium 24.9; Low/low 14.3.
  - Water and sanitation: High/high 53.9; Medium/medium 37.8; Low/low 17.8.
- The paper highlights illustrative benchmarks: for intermediate labor mobility and labor intensity, 3 jobs in schools and hospitals to 6.6 jobs in electricity per US$1 million in AEs; 16 jobs in roads to 30.4 jobs in water and sanitation in LIDCs.

### Green investment and R&D implications
- Literature-based estimates for green investments (reported ranges):  
  - Green electricity: 5–10 jobs per US$1 million.  
  - Efficient new buildings (e.g., schools and hospitals): 2.4–12.5 jobs per US$1 million.  
  - Green water and sanitation: 5.7–14 jobs per US$1 million.
- Conclusion: job creation multipliers for green investment resemble high labor-intensity estimates in Table 4.
- R&D: public R&D generates higher-skilled jobs; green R&D estimated at 3–8 jobs per US$1 million (IEA, 2020); R&D job-content estimates broadly consistent with investment-job estimates reported.

### Caveats, limitations, and interpretation
- Dataset should be treated as illustrative rather than statistically representative: most sampled companies are medium-size, unlisted, and have several years of audited financial statements.
- No firm-level data for LIDCs; LIDC estimates are linear extrapolations from AE and EME point estimates.
- Limited data granularity prevents disentangling: new investment versus maintenance, part-time versus full-time, skilled versus unskilled, and imported versus local labor.
- Two inefficiencies not captured by reduced-form job estimates: (i) low productivity at public administration (budget-to-contractor losses due to poor project design, red tape, corruption), and (ii) low productivity at contractor level (jobs may not translate into public goods).
- The β_s coefficients reflect average job content across contracting layers since upstream/downstream firms cannot be disaggregated by industrial code.
- Wage levels, job quality (full vs part-time), and specific green investment types are not unfolded due to data limitations; fixed effects absorb some country- and firm-specific heterogeneity.

### Welfare and political economy considerations
- AEs generally have the human and physical capital and can borrow at record-low interest rates to scale up public investment; EMEs and LIDCs face higher borrowing costs, elevated public debt, and capacity constraints.
- SOEs often undertake public investment and have on average higher job intensity than private firms. In OECD countries, SOEs represent on average 4.7 percent of the labor force versus 15.8 percent by the general government.
- Political economy and institutional capacity constraints in EMEs and LIDCs complicate scaling up public investment despite potentially higher employment multipliers.

*Source: Author’s own estimations based on Compustat and Orbis (paper organized as Sections II–V; sample period 1999–2017; regressions clustered at firm/country levels as specified in source tables).*

### 4.8 percent, Hungary 4.2 percent, France 3.5 percent, Finland 3.5 percent, the Czech Republic

### wpiea2021131-print-pdf - 4.8 percent, Hungary 4.2 percent, France 3.5 percent, Finland 3.5 percent, the Czech Republic

### Main findings on public investment and employment
- Public investment can support employment in addition to creating infrastructure; the emphasis on job creation is particularly relevant given the dramatic impact on labor markets.
- An increase in public investment equivalent to 1 percent of GDP could directly create more than seven million jobs in AEs and EMEs through direct employment effects alone (i.e., about 5.4 percent of the full-time equivalent jobs lost in 2021 relative to fourth quarter of 2019; ILO 2021, January).
- Job content estimates applied:
  - 4.9 per US$1 million invested for AEs (unweighted average) applied to an increase in investment worth 1 percent of GDP in AEs (ca. US$500 billion in 2020; cf. Appendix 1).
  - 14.8 per US$1 million for emerging markets applied to 1 percent of the GDP of EMEs (ca. US$320 billion).
- The impact could be higher for green investment and for investments in R&D with a higher labor intensity.
- Job intensity for SOEs is found to be 30 percent higher than for their private counterparts (Baum and others, 2019); possible explanations include larger size of SOEs and an implicit employment remit. This is especially important in EMEs and LIDCs, where SOEs account for more than half of all infrastructure project commitments (IMF, 2020a).

### Limitations and factors likely to bias estimates
- The employment estimates may underestimate actual job creation because:
  - Firms with less than five observations are excluded, missing employment increases in cyclical companies that form during fiscal expansion and disappear during consolidation and which likely have higher elasticity between revenue and employment.
  - Effects in LIDCs are linearly extrapolated from AEs and EMEs; this relationship may be convex, so the impact on employment may increase exponentially the lower the country’s income per capita, since infrastructure development is more labor-intensive and the labor force is less specialized in LIDCs (Tanzi, 2019).
  - Indirect labor impact and spillovers (including Keynesian multiplier effects into other sectors) are not included. The IMF (2020b) estimates that a 1 percent of GDP increase in public investment in AEs and EMEs has the potential to create, directly and indirectly, between 20 and 33 million jobs.
  - Estimates are pooled and do not distinguish between new projects and maintenance, or between skilled and unskilled labor; ceteris paribus, maintenance projects and projects with a higher unskilled labor component would create more jobs than estimated here.
- The job creation estimates may be overstated due to the degree of waste from budget allocations to actual contracted public investment. The Public Investment Management Assessment (PIMA) database can serve as a tentative first approach to correct for this upward bias.

### Public-private partnerships and state capacity
- Public-private initiatives are not a panacea when fiscal space is limited:
  - Public-private partnerships are not “free money”; private-sector investment decisions are driven by profitability and enforcement of property rights.
  - Fiscal capacity tends to be correlated with state capacity. Public and private sectors are complementary in infrastructure when there is sufficient state capacity to provide effective regulation and safeguard property rights.
  - If there is sufficient fiscal space and state capacity, the private sector can contribute superior technology and efficient management in particular cases; if there is no fiscal space, private-sector capital is unlikely to solve institutional and financing deficiencies.

### Green investment considerations
- Policies on greening the recovery should be carefully designed to avoid backlash.
- Clean-energy infrastructure has been found to be labor intensive in the short term (Garrett-Peltier, 2017), although not all green investments create jobs quickly (Popp and others, 2020).
- Some green investments are not job rich in the long term and require specific skills; for example, windmills are capital intensive and produced in only a few countries.
- While green investment offers clear global welfare gains (Hicks-Kaldor efficiency), distributional effects and Pareto efficiency for low-income countries are debatable.

### Policy implications and tools for policymakers
- The estimates presented (e.g., in Table 4 of the source) can help policymakers assess employment impact and trade-offs of alternative uses of fiscal space:
  - Trade-offs between short-term current spending and long-term capital spending.
  - Trade-offs between social and physical capital sectors.
  - The enhancing employment impact of higher labor mobility (e.g., through training and flexible arrangements).
  - The enhancing employment impact of higher labor content (e.g., in infrastructure maintenance and green investment).
- Policies that provide short-term sectoral preference may have longer-term implications (e.g., prioritizing roads may crowd out electricity investment needed for digital infrastructure).
- Future research extensions noted:
  - Quantification of employment impact by country.
  - Employment impact in digital infrastructure.
  - Employment impact by green versus brown investment as industrial classifications evolve.

### Appendix 1 — sample coverage summary (selected exact figures)
- Sample includes firm-level data from 38 advanced economies and 63 emerging market economies, totaling 101 economies representing 95 percent of global GDP.
- Aggregate sample totals:
  - Employment: 2,204.4 (millions)
  - GDP: 83,218.7 (US$ billions)
- Selected country entries (Employment in millions; GDP in US$ billions):
  - United States: 156.9; 21,439.5
  - China: 775.3; 1,4140.2
  - India: 32.0; 2,935.6
  - Japan: 67.4; 5,154.5
  - Germany: 42.0; 3,863.3
  - Brazil: 92.1; 1,847.0
  - United Kingdom: 32.8; 2,743.6
  - France: 25.6; 2,707.1
  - Italy: 23.3; 1,988.6
  - Russia: 72.7; 1,637.9
  - Hungary: 4.5; 170.4
  - Czech Republic: 5.3; 247.0
  - Finland: 2.6; 269.7
  - Slovak Republic: 2.4; 106.6
  - Mexico: 54.4; 1,274.2
  - Indonesia: 127.1; 1,111.7

*Source: WEO and IMF.*

### REFERENCES

### REFERENCES

### Green recovery and climate-related fiscal policy
- Allan, Jennifer, Charles Donovan, Paul Ekins, Ajay Gambhir, Cameron Hepburn, David Reay, Nick Robins, Emily Shuckburgh, and Dimitri Zenghelis, 2020, “A Net-Zero Emissions Economic Recovery from COVID-19,” Smith School Working Paper 20-01 (Oxford Smith School of Enterprise and the Environment).
- Barbier, Edward, 2010, “Green Stimulus, Green Recovery and Global Imbalances,” World Economics Journal, Vol. 11, No. 2, pp.149–77.
- Garrett-Peltier, Heidi, 2017, “Green versus brown: Comparing the employment impacts of energy efficiency, renewable energy, and fossil fuels using an input-output model,” Economic Modelling, Vol. 61, pp. 439–47.
- Houser, Trevor, Shashank Mohan, and Robert Heilmayr, 2009, “A Green Global Recovery? Assessing US Economic Stimulus and the Prospects for International Coordination,” Policy Brief 09-3, (Washington: Peterson Institute for International Economics).
- International Energy Agency (IEA), 2020, “Sustainable Recovery,” World Energy Outlook Special Report (Paris, France).
- Jacobs, Michael, 2012, “Green Growth: Economic Theory and Political Discourse,” Working Paper No. 92 (London: Grantham Research Institute on Climate Change and the Environment at The London School of Economics and Political Science).
- Muro, Mark, Adie Tomer, Ranjitha Shivaram, and Joseph W. Kane, 2019, Advancing Inclusion through Clean Energy Jobs. Metropolitan Policy Program (Washington: Brookings Institution).
- Popp, David, Francesco Vona, Giovanni Marin, and Ziqiao Chen, 2020, “The Employment Impact of Green Fiscal Push: Evidence from the American Recovery Act,” NBER Working Paper No. 27321 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Coalition of Finance Ministers for Climate Action, The, 2020, “Better Recovery, Better World: Resetting Climate Action in the Aftermath of the COVID-19 Pandemic,” (Washington: The World Bank Group).

### Fiscal policy, public investment, and multipliers
- Auerbach, Alan J., and Yuriy Gorodnichenko, 2013, “Output Spillovers from Fiscal Policy,” American Economic Review, Vol. 103, No. 3, pp.141–46.
- Gaspar, Vitor, David Amaglobeli, Mercedes Garcia-Escribano, Delphine Prady, and Mauricio Soto, 2019, “Fiscal Policy and Development: Human, Social, and Physical Investments for the SDGs,” Staff Discussion Notes No. 19/03 (Washington: International Monetary Fund).
- International Monetary Fund, 2015, “Making Public Investment More Efficient,” IMF Policy Paper (Washington: International Monetary Fund).
- International Monetary Fund, 2018, “Public Investment Management Assessment—Review and Update,” IMF Policy Paper (Washington: International Monetary Fund).
- International Monetary Fund, 2019, “How to Mitigate Climate Change,” October 2019 Fiscal Monitor (Washington: International Monetary Fund).
- International Monetary Fund, 2020a, “Policies to Support People During the COVID-19 Pandemic,” April 2020 Fiscal Monitor (Washington: International Monetary Fund).
- International Monetary Fund, 2020b, “Policies for the Recovery,” October 2020 Fiscal Monitor (Washington: International Monetary Fund).
- International Monetary Fund, 2021, “A Fair Shot,” April 2021 Fiscal Monitor (Washington: International Monetary Fund).
- Ramey, Valerie A., 2020, “The Macroeconomic Consequences of Infrastructure Investment,” NBER Working Paper No. 27625 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Wilson, Daniel J., 2012, “Fiscal Spending Jobs Multipliers: Evidence from the 2009 American Recovery and Reinvestment Act,” American Economic Journal: Economic Policy, Vol. 4, No. 3, pp. 251–82.
- Tanzi, Vito, 2019, “The Limits of Stabilization Policies,” Acta Oeconomica, Vol. 69, pp. 141–51.
- Tanzi, Vito, and Hamid Davoodi, 1998, “Corruption, Public Investment, and Growth,” in The Welfare State, Public Investment, and Growth, ed. by Hirofumi Shibata and Toshihiro Ihori (Tokyo: Springer).

### Employment, labor markets, and targeted programs
- Aizer, Anna, Shari Eli, Adriana Lleras-Muney, and Keyoung Lee, 2020, “Do Youth Employment Programs Work? Evidence from the New Deal,” NBER Working Paper No. 27103 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Emrath, Paul, 2015, “Subcontracting: Three-Fourths of Construction Cost in the Typical Home,” National Association of Home Builders (NAHB) Special Studies Series (Washington: National Association of Home Builders).
- Garin, Andrew, 2019, “Putting America to Work, Where? Evidence on the Effectiveness of Infrastructure Construction as a Locally Targeted Employment Policy,” Journal of Urban Economics, Vol. 111, pp.108–31.
- International Labour Organisation, 2021, “ILO Monitor: COVID-19 and the World of Work. 7th Edition,” Briefing note, (Geneva: International Labour Organization).
- Papanikolaou, Dimitris, and Lawrence D.W. Schmidt, 2020, “Working Remotely and the Supply-side Impact of Covid-19,” NBER Working Paper No. 27330 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Mummolo, Jonathan, and Erik Peterson, 2018, “Improving the Interpretation of Fixed Effects Regression Results,” Political Science Research and Methods, Vol. 6, No. 4, pp. 829–35.
- Muro, Mark, Adie Tomer, Ranjitha Shivaram, and Joseph W. Kane, 2019, Advancing Inclusion through Clean Energy Jobs. Metropolitan Policy Program (Washington: Brookings Institution).

### State capacity, governance, and public enterprises
- Baum, Anja, Clay Hackney, Paulo Medas, and Mouhamadou Sy, 2019, “Governance and State-Owned Enterprises: How Costly is Corruption?,” IMF Working Paper No. 19/253 (Washington: International Monetary Fund).
- Besley, Timothy, and Torsten Persson, 2009, “The Origins of State Capacity: Property Rights, Taxation, and Politics,” American Economic Review, Vol. 99, No. 4, pp.1218–44.
- Besley, Timothy, and Torsten Persson, 2010, “State Capacity, Conflict, and Development,” Econometrica, Vol. 78, No. 1, pp.1–34.
- Organization for Economic Cooperation and Development, 2017, The Size and Sectoral Distribution of State-Owned Enterprises (Paris: OECD Publishing).
- Organization for Economic Cooperation and Development, 2013, “Employment in General Government and Public Corporations,” in Government at a Glance 2013 (Paris: OECD Publishing).
- Moszoro, Mariano, Gonzalo Araya, Fernanda Ruiz-Nuñez, and Jordan Schwartz, 2015, “What Drives Private Participation in Infrastructure Developing Countries?,” in Public Private Partnerships for Infrastructure and Business Development, ed. by Stefano Caselli, Guido Corbetta, and Veronica Vecchi (New York: Palgrave Macmillan).
- Moszoro, Mariano, 2018, “Public–Private Monopoly,” The B.E Journal of Economic Analysis & Policy, Vol. 18, No. 2, pp.1–15.
- Sargent Jr., John F., 2020, “Federal Research and Development (R&D) Funding: FY2020,” CRS Report R45715 (Washington: Congressional Research Service).

### Methodology, measurement, and empirical techniques
- Kézdi, Gabor, 2004, “Robust Standard Error Estimation in Fixed-Effects Panel Models,” Hungarian Statistical Review, Special Number 9, pp. 96–116.
- Wooldridge, Jeffrey M., 2003, “Cluster-Sample Methods in Applied Econometrics,” American Economic Review, Vol. 93, No. 2, pp.133–38.
- Mummolo, Jonathan, and Erik Peterson, 2018, “Improving the Interpretation of Fixed Effects Regression Results,” Political Science Research and Methods, Vol. 6, No. 4, pp. 829–35.
- Popp, David, Francesco Vona, Giovanni Marin, and Ziqiao Chen, 2020, “The Employment Impact of Green Fiscal Push: Evidence from the American Recovery Act,” NBER Working Paper No. 27321 (Cambridge, Massachusetts: National Bureau of Economic Research).

### COVID-19 policy, social spending, and recovery strategies
- International Monetary Fund, 2020a, “Policies to Support People During the COVID-19 Pandemic,” April 2020 Fiscal Monitor (Washington: International Monetary Fund).
- International Monetary Fund, 2020b, “Policies for the Recovery,” October 2020 Fiscal Monitor (Washington: International Monetary Fund).
- International Monetary Fund, 2021, “A Fair Shot,” April 2021 Fiscal Monitor (Washington: International Monetary Fund).
- Coalition of Finance Ministers for Climate Action, The, 2020, “Better Recovery, Better World: Resetting Climate Action in the Aftermath of the COVID-19 Pandemic,” (Washington: The World Bank Group).
- Yackovlev, Irene, Zuzana Murgasova, Fei Liu, Gohar Minasyan, and Ke Wang, 2020, “How to Operationalize IMF Engagement on Social Spending during and in the aftermath of the COVID-19 Crisis,” IMF How-To Note No. 20/02 (Washington: International Monetary Fund).
- International Labour Organisation, 2021, “ILO Monitor: COVID-19 and the World of Work. 7th Edition,” Briefing note, (Geneva: International Labour Organization).
- Papanikolaou, Dimitris, and Lawrence D.W. Schmidt, 2020, “Working Remotely and the Supply-side Impact of Covid-19,” NBER Working Paper No. 27330 (Cambridge, Massachusetts: National Bureau of Economic Research).

*References list as provided in the source PDF.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021131-print-pdf.pdf_