## Executive Summary — wpiea2022086-print-pdf

## Source details

**Canonical URL:** [Executive Summary — wpiea2022086-print-pdf](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022086-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022086-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022086-print-pdf.pdf.json)

---

### Scope and objective
- Presents a statistical framework for estimating the carbon emissions associated with foreign direct investment (FDI) in host economies.
- Produces two sets of estimates:
  - Carbon emissions from capital formation funded by FDI (e.g., construction of new plant and equipment).
  - Direct and indirect carbon emissions from the production of foreign-owned firms (MNEs).

### Data and methodological foundation
- Methodology underpinned by:
  - OECD Inter Country Input Output (ICIO) tables linked to carbon emissions.
  - FDI statistics by industry from the OECD.
  - OECD Analytical Activity of Multinational Enterprises (AMNE) Database.
- Framework also used to develop comparable estimates of carbon emissions in the host economy from operations of domestic owned enterprises (DOEs).

### Key empirical findings — investment impact of FDI (GFCF)
- Coverage and period:
  - Estimates for 23 countries during 2005 to 2015 and 36 industries (ISIC Rev 4); results for 17 countries with more complete annual and industry estimates are presented for comparability.
- Trends:
  - The carbon intensity of gross fixed capital formation (GFCF) financed by FDI has fallen over time.
  - General downward trend in emissions associated with GFCF financed by FDI from 2007 to 2014 despite an increase in FDI flows from 2009.
  - Spike in both emissions and FDI flows in 2015.
  - Interpretation: FDI emissions in GFCF funded by FDI have been falling relative to FDI inflows in later years of the period.
- Industry cumulative shares (average over period):
  - Electricity, gas, and water: average share of 39 percent of total emissions.
  - Construction: 16 percent.
  - Manufacturing of other non-metallic mineral products: 6 percent.
  - Mining and quarrying: 3 percent.
- Industry-level intensity:
  - Electricity, gas, and water had the highest emissions relative to domestic demand; emissions almost three times those of construction.
  - Contributions of electricity, gas, and water were highest in countries relying on fossil fuels.
  - Construction emissions significant in countries with large oil and gas investment projects.
- Country cumulative emissions (top and bottom during period):
  - Highest emissions: Estonia, Netherlands, Hungary, New Zealand, Poland, and Iceland.
  - Lowest emissions: France, Italy, Greece, and Denmark.
- Cross-country and industry patterns:
  - Poland: slight decline in construction and electricity emissions.
  - Czech Republic: much more significant downward trends in construction and electricity.
  - Mexico: upward trend in electricity emissions.
  - United States: no clear trend in electricity emissions.

### Key empirical findings — ongoing operations of MNEs
- Coverage:
  - MNEs operating in 59 countries during 2005 to 2015 in 34 industries.
- Aggregate industry emissions by MNEs (shares):
  - Electricity generation: 15 percent of total MNE emissions.
  - Manufacturing of motor vehicles, trailers, and semitrailers: 13 percent of total MNE emissions.
  - Manufacturing of chemicals and pharmaceuticals: 9 percent of total MNE emissions.
- Shares of emissions within specific industries accounted for by MNEs:
  - Manufacturing of motor vehicles, trailers, and semitrailers: 37 percent.
  - Manufacturing of computer, electronic and optical products: 34 percent.
  - Manufacturing of chemicals and pharmaceutical products: 29 percent.
- Sectoral carbon intensities (tons per 1 million US dollars of output):
  - Highest intensities: electricity; manufacturing of non-metallic mineral products; manufacturing of basic metals; transportation and storage.
- MNEs vs DOEs:
  - Carbon intensities of MNEs were lower than DOEs in almost all industries, except: transport and storage; construction; and some manufacturing subsectors (wood and wood products; machinery and equipment; motor vehicles; textiles).
  - DOEs accounted for the largest share in total emissions and generally had higher carbon intensities.
  - Exceptions exist: in some low carbon intensive countries, MNEs had higher carbon intensities.
- Country-level carbon intensities of MNEs (2005–15 average):
  - Highest MNE carbon intensities: South Africa; China; Saudi Arabia; India; Vietnam.
  - Lowest MNE carbon intensities: Switzerland; Norway; Sweden; Luxembourg; France.
- Country comparisons:
  - DOEs had higher carbon intensities than MNEs in the dozen countries with highest carbon intensities except Indonesia.
  - In low carbon intensity countries, MNEs had higher carbon intensities than DOEs with the exception of Cyprus.
  - Largest differences where DOEs exceeded MNEs: Cyprus and Malta.
  - Largest differences where MNEs exceeded DOEs: Hong Kong SAR, Switzerland, and Iceland.
- Industry and country patterns:
  - In high carbon intensity economies (China, South Africa) electricity, gas, and water had the highest intensity.
  - In low carbon intensity economies (Norway, Switzerland) manufacturing industries had the highest intensities.
  - Differences between MNE and DOE intensity were small in high-intensity economies (except electricity, gas and water) and larger in low-intensity economies.

### Trends in selected industries (MNE carbon intensities, 2005–15)
- Manufacturing of basic metals:
  - Highest intensities in 2015: South Africa, India, Russian Federation, China, Saudi Arabia.
  - General downward trend except for India and South Africa.
- Electricity:
  - MNEs in South Africa, Saudi Arabia, India, Estonia, and China showed a general downward trend.
- Manufacture of other non-metallic mineral products:
  - Countries showing general downward trends: Vietnam, Thailand, Philippines, India, China; exception: Vietnam.
- Transport and storage:
  - Highest MNE intensities: Malta, Singapore, Saudi Arabia, China, Thailand.
  - Decline in middle of period followed by an uptick later in period.

### Exports: emissions embodied in MNEs output
- Industry-level observation:
  - Industries with high carbon intensities among MNEs (transport; manufacture of basic metals; chemicals and pharmaceuticals; non-metallic mineral products) have export intensities between 30 and 60 percent.
  - Several low carbon intensity industries (accommodation; manufacture of textiles; electrical equipment; machinery; computer and electronic products) also have high export shares.
- Country-level observation:
  - Countries with fairly low carbon intensities can have large shares of MNE output exported.
  - Examples where MNEs have relatively low emission intensities but more than half of output is exported: Ireland; Hungary; Iceland; Luxembourg; Cyprus; Slovenia; Switzerland; Costa Rica.
  - Notable exceptions with both high carbon intensity and high export shares: China; Vietnam; Thailand.
- Comparative export shares:
  - Export-related emissions shares are higher for MNEs compared to DOEs in all countries.
  - In countries with lower carbon intensities, the gap between MNEs and DOEs export shares tends to be higher (exceptions include Luxembourg and Lithuania).
  - In higher carbon intensity economies, gaps are smaller (exceptions include China, India, South Africa, Vietnam, Thailand).

### Policy implications and recommendations
- Home country policies:
  - Policies that incentivize domestic direct investors to meet high emissions standards in host economies could be an important tool in reducing global emissions, given the high carbon emission intensities of MNEs in high export intensity industries.
  - Such policies could induce firms to use lower carbon production functions and technology at home and abroad, and to demand lower carbon infrastructure and transportation in host economies.
  - Encouraging firms to reduce emissions along their supply chains could lead them to demand that their suppliers reduce their carbon emissions.
- Host economy policies:
  - Remove barriers to investment in environmental goods and services industries as well as in low carbon technologies to promote positive spillovers and knowledge and technology transfer to the domestic economy.
  - Include an analysis of the impact on carbon emissions as part of FDI attraction strategies.
- Corporate disclosure:
  - Developing a standard for companies to disclose their carbon emissions will provide valuable information to better understand the role of all enterprises, both MNEs and DOEs, in carbon emissions.
- Data improvements recommended:
  - Develop statistics that identify the FDI flows that are used to expand capacity in the host economy.
  - Develop statistics that identify carbon emissions by MNEs.
  - Expand country coverage to enable a more comprehensive analysis of the impact of offshoring of MNEs on global carbon emissions.

### Methodological and data limitations; avenues for future work
- Main limitations:
  - Absence of FDI data that distinctly finances GFCF; available FDI may finance M&As, transit capital via SPEs, acquisition of financial assets.
  - To address transit capital, estimates excluded countries with large offshore financial centers: Luxembourg, Netherlands, and Ireland.
  - Absence of separate direct carbon emissions data for MNEs and DOEs; direct emission intensities of MNEs assumed equal to DOEs in same industry.
  - Geographical bias for GFCF of FDI indicator to OECD countries.
  - Analysis limited to period 2005–2015 due to data availability.
  - Input–output framework limitations: no dynamic interactions, fixed input structures, no factor supply constraints, no budget constraints, assumption of no lags in deployment of FDI funds.
  - Direct emissions based on IEA Tier 1 method; some countries may have Tier 2 or Tier 3 estimates that were not used.
- Suggested future work:
  - FDI estimates that distinguish between the use of FDI resources for acquisition of assets versus for greenfield investment and capacity extension.
  - Expanded information on the role of MNEs in carbon emissions such as actual estimates of direct carbon emissions of MNEs by activity/sector.
  - Use of models to capture dynamic interactions.
  - Sensitivity analysis on the differences in direct emissions between MNEs and DOEs to inform how to structure climate related incentive schemes.
  - Explore whether there are spillovers in the form of reduced carbon intensity at DOEs from the operations of MNEs in the host economy.
  - Better understand the relationship between FDI’s role in production and export diversification and the country’s carbon emissions.

### Illustrative numerical examples from the annex (method demonstration)
- Emissions (by final use) totals:
  - Agriculture: 259.7339782 (HFCE) and 51.94679565 (exports) — subtotal 312
  - Manufacturing: 262.8778718 (HFCE) and 262.8778718 (exports) — subtotal 526
  - Services: 194.050786 (HFCE) and 48.51269649 (exports) — subtotal 243
  - Total embodied emissions (HFCE + exports): 1080
- Input-output / multiplier example:
  - Leontief inverse (multipliers) row totals:
    - Agriculture total: 2.42
    - Manufacturing total: 2.73
    - Service total: 2.59
  - Direct emissions coefficients: 0.30 (Agriculture), 0.2 (Manufacturing), 0.1 (Services)
  - Direct and indirect emissions coefficients (emissions multipliers) totals:
    - Total 0.519 (Agriculture), 0.526 (Manufacturing), 0.485 (Services)
- Simple output allocation table (selected entries):
  - Agriculture contributed to Agriculture: multiplier 1.30, direct emissions 0.3, partial direct and indirect 0.389
  - Manufacturing contributed to Manufacturing: multiplier 1.59, direct emissions 0.2, partial direct and indirect 0.317
  - Services contributed to Services: multiplier 1.58, direct emissions 0.1, partial direct and indirect 0.158

*Source: Executive Summary of wpiea2022086-print-pdf*

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

### wpiea2022086-print-pdf - Executive Summary

### Scope and objective
- Presents a statistical framework for estimating the carbon emissions associated with foreign direct investment (FDI) in host economies.
- Produces two sets of estimates:
  - Carbon emissions from capital formation funded by FDI (e.g., construction of new plant and equipment).
  - Direct and indirect carbon emissions from the production of foreign-owned firms (MNEs).

### Data and methodological foundation
- Methodology underpinned by:
  - OECD Inter Country Input Output (ICIO) tables linked to carbon emissions.
  - FDI statistics by industry from the OECD.
  - OECD Analytical Activity of Multinational Enterprises (AMNE) Database.
- Framework also used to develop comparable estimates of carbon emissions in the host economy from operations of domestic owned enterprises (DOEs).

### Key empirical findings
- Trend in FDI-financed capital formation:
  - The carbon intensity of gross fixed capital formation (GFCF) financed by FDI has fallen over time.
  - Reductions driven in most countries by declines in the carbon intensity of the electricity, gas, and water industry.
- Relative magnitude of emissions:
  - Carbon emissions from the ongoing operations of foreign-controlled firms (MNEs) are larger than those associated with their capital formation.
- Industry-level high emitters among MNEs:
  - Manufacturing; transport and storage; and electricity, gas, and water had the highest overall emissions and emission intensities among MNEs.
- Comparison MNEs vs DOEs:
  - DOEs accounted for the largest share in total emissions and generally had higher carbon intensities.
  - Exceptions exist: in some low carbon intensive countries, MNEs had higher carbon intensities.
- Relationship between emission and export intensity:
  - For MNEs, high emissions intensities were accompanied by high export intensities in mining; transport and storage; and manufacturing industries.

### Policy implications and recommendations
- Home country policies that incentivize domestic direct investors to meet high emissions standards in host economies could be an important tool in reducing global emissions, given:
  - The high carbon emission intensities of MNEs in high export intensity industries.
- Data improvements to enhance estimate quality:
  - Develop statistics that identify the FDI flows that are used to expand capacity in the host economy.
  - Develop statistics that identify carbon emissions by MNEs.
- Expand country coverage to enable a more comprehensive analysis of the impact of offshoring of MNEs on global carbon emissions.

*Source: Executive Summary of wpiea2022086-print-pdf*

### Introduction

### Introduction

### Effects of FDI on host economies and environment
- FDI has been associated with rising wages, higher productivity, productivity and knowledge spillovers to domestic firms, export diversification and introduction of new industries, and increasing growth (literature cited).
- FDI can affect carbon emissions through multiple channels:
  - by increasing the scale of economic activity;
  - by contributing to demand for addressing climate change;
  - by diffusing low-carbon knowledge and technology across borders.
- Three views on FDI and environmental outcomes are noted:
  - Environmental Kuznets curve argument: as FDI increases incomes, demand for environmental quality rises and environmental damage will eventually fall.
  - Pollution havens hypothesis: FDI is associated with higher carbon emissions especially in low-income countries because low pollution standards attract pollution-intensive FDI; proponents recommend mandatory environmental conduct requirements.
  - Pollution halos argument: FDI is cleaner than domestic investment because it deploys cleaner technologies, better management, and higher standards, facilitating transfer of low-carbon technology.

### Research focus and key questions
- The paper develops a framework to estimate contributions of FDI and foreign-owned firms to carbon emissions using industry-level information on production, trade, investment, carbon emissions, and distinctions between MNEs and DOEs.
- The framework produces estimates of:
  - direct emissions from investment-financed capital formation;
  - emissions from operations of foreign-owned firms;
  - whether production and embodied emissions meet domestic demand or are exported.
- The three explicit key questions the framework aims to answer for a given country:
  i. What is the effect on carbon emissions of direct investment that finances investments in new productive capacity, such as new plant and equipment?
  ii. What is the contribution to emissions from the operations (i.e., economic activity) of foreign owned enterprises in host economies?
  iii. Does the production of foreign-owned firms, as well as the emissions embodied in that production, meet domestic demand or is it exported to meet foreign demand?

### Methodology and data — overview
- Framework components:
  - A. Estimating the investment effect of FDI on carbon emissions (carbon emissions in supply to GFCF of FDI).
    - Uses input–output analysis central equation to determine carbon emissions embodied in supplies to gross fixed capital formation (GFCF).
    - Direct emissions: IEA production-based estimates of carbon emissions from fuel combustion using Tier 1 methods based on the 2006 IPCC Guidelines.
    - Indirect emissions: embodied emissions in inputs (e.g., cement used in construction).
    - Steps summarized:
      - Step 1: Obtain total emissions emitted during production for each industry for each country.
      - Step 2: Estimate direct emission coefficient Cdirect for each industry as total emissions divided by output.
      - Step 3: Estimate total carbon emission coefficients (direct + indirect) via:
        Ctotal = (I−A)−1 · Cdirect  (Equation (1))
        where Ctotal is (n×1) vector, Cdirect is (n×1) vector, A is input coefficient matrix, I is identity matrix, and (I−A)−1 is the Leontief inverse.
      - Step 4: Estimate total carbon emissions associated with GFCF:
        Carbon emissions of GFCF = Ctotal × final use in GFCF  (Equation (2))
      - Step 5: Apportion GFCF emissions to FDI:
        Carbon emissions of GFCF of FDI = Ctotal × final use in GFCF × (FDI/GFCF)  (Equation (3))
    - Industry-level estimates of carbon emissions in supply to GFCF of FDI are divided by respective industry-level final demand for domestic products for comparability.
  - B. Estimating effect of ongoing operations of foreign-owned enterprises on carbon emissions.
    - Uses OECD Analytical AMNE Database to track production activity of MNEs and DOEs by industry and country.
    - Computes total emission coefficients for combined MNEs & DOEs using:
      CMNEs&DOEs total = (I − AMNEs&DOEs)−1 · CMNEs&DOEs Direct  (Equation (4))
    - Ctotal MNEs&DOEs is split into Ctotal MNEs and Ctotal DOEs to estimate:
      Carbon emissions in MNEs output for final demand = CMNEs total × MNEs Output for FD  (Equation (5))
      Carbon emissions in DOEs output for final demand = CDOEs total × DOEs Output for FD  (Equation (6))
  - C. Estimating emissions of MNEs and DOEs embodied in exports.
    - Emissions associated with exports:
      Carbon emissions of Exports of MNEs = CMNEs total × Exports of MNEs  (Equation (7))
      Carbon emissions of Exports of DOEs = CDOEs total × Exports of DOEs  (Equation (8))
- Data sources and coverage (Table 1 summarized):
  - Carbon emissions: IEA production-based emissions, period 2005–15.
  - Output: OECD National Accounts Database, period 2005–15.
  - Input coefficients: OECD Input Output Database, period 2005–15.
  - GFCF: OECD National Accounts Database, period 2005–15.
  - Inward FDI of non-SPEs: OECD FDI financial flows database, period 2005–15.
  - Final demand: OECD Input Output Database, period 2005–15.
  - MNEs and DOEs final demand and exports: OECD Analytical AMNE database, period 2005–15.
  - MNEs and DOEs input coefficients: OECD Intercountry Input Output Tables from the Analytical AMNE database, period 2005–15.
- Use of Analytical AMNE Database:
  - Provides transaction matrices for domestic-owned and foreign-owned firms in 59 countries plus rest of world, covering 34 sectors for 2005–2016.
  - Database structure splits intermediate consumption and final demand by ownership (domestic vs foreign) and separates value-added and gross output by ownership.

### Methodological and data limitations
- For carbon emissions from GFCF of FDI:
  - Absence of FDI data that distinctly finances GFCF; available FDI may finance M&As, transit capital via SPEs, acquisition of financial assets.
  - To address transit capital, estimates excluded countries with large offshore financial centers: Luxembourg, Netherlands, and Ireland.
  - Comparative estimates based on operating entities (excluding SPEs) showed similar trends but were not fully available across sample.
  - Improvements depend on statistical updates enabling decomposition of FDI by use in host country.
- For operations of MNEs:
  - Absence of separate direct carbon emissions data for MNEs and DOEs; direct emission intensities of MNEs were assumed equal to DOEs in same industry.
  - Total-emissions variation between MNEs and DOEs driven mainly by industry distribution and sourcing patterns reflected in input coefficients.
  - Better firm-level or location-specific emissions data would allow relaxing the equal-intensity assumption.
  - Initiatives to improve corporate emissions reporting could ease this limitation.
- Other limitations:
  - Geographical bias for GFCF of FDI indicator to OECD countries.
  - Analysis limited to period 2005–2015 due to data availability.
  - Input–output framework limitations: no dynamic interactions, fixed input structures, no factor supply constraints, no budget constraints, assumption of no lags in deployment of FDI funds.
  - Direct emissions based on IEA Tier 1 method; some countries may have Tier 2 or Tier 3 estimates that were not used.

### Key results — investment impact of FDI (GFCF)
- Coverage: estimates for 23 countries during 2005 to 2015 and 36 industries (ISIC Rev 4); results for 17 countries with more complete annual and industry estimates are presented for comparability.
- Trends:
  - General downward trend in emissions associated with GFCF financed by FDI from 2007 to 2014 despite an increase in FDI flows from 2009.
  - Spike in both emissions and FDI flows in 2015.
  - Interpretation: FDI emissions in GFCF funded by FDI have been falling relative to FDI inflows in later years of the period.
- Industry cumulative shares (average over period):
  - Electricity, gas, and water industry: average share of 39 percent of total emissions.
  - Construction: 16 percent.
  - Manufacturing of other non-metallic mineral products: 6 percent.
  - Mining and quarrying: 3 percent.
- Industry-level intensity:
  - Electricity, gas, and water had the highest emissions relative to domestic demand; emissions almost three times those of construction.
  - Contributions of electricity, gas, and water were highest in countries relying on fossil fuels.
  - Construction emissions significant in countries with large oil and gas investment projects.
- Country cumulative emissions (top and bottom during period):
  - Highest emissions: Estonia, Netherlands, Hungary, New Zealand, Poland, and Iceland.
  - Lowest emissions: France, Italy, Greece, and Denmark.
- Cross-country and industry patterns:
  - Poland: slight decline in construction and electricity emissions.
  - Czech Republic: much more significant downward trends in construction and electricity.
  - Mexico: upward trend in electricity emissions.
  - United States: no clear trend in electricity emissions.

### Key results — ongoing operations of MNEs
- Coverage: MNEs operating in 59 countries during 2005 to 2015 in 34 industries.
- Aggregate industry emissions by MNEs (shares):
  - Electricity generation: 15 percent of total MNE emissions.
  - Manufacturing of motor vehicles, trailers, and semitrailers: 13 percent of total MNE emissions.
  - Manufacturing of chemicals and pharmaceuticals: 9 percent of total MNE emissions.
- Shares of emissions within specific industries accounted for by MNEs:
  - Manufacturing of motor vehicles, trailers, and semitrailers: 37 percent.
  - Manufacturing of computer, electronic and optical products: 34 percent.
  - Manufacturing of chemicals and pharmaceutical products: 29 percent.
- Sectoral carbon intensities (tons per 1 million US dollars of output):
  - Highest intensities: electricity; manufacturing of non-metallic mineral products; manufacturing of basic metals; transportation and storage.
  - Carbon intensities of MNEs were lower than DOEs in almost all industries, except: transport and storage; construction; and some manufacturing subsectors (wood and wood products; machinery and equipment; motor vehicles; textiles).
- Country-level carbon intensities of MNEs (2005–15 average):
  - Highest MNE carbon intensities: South Africa; China; Saudi Arabia; India; Vietnam.
  - Lowest MNE carbon intensities: Switzerland; Norway; Sweden; Luxembourg; France.
- Country comparisons of MNEs vs DOEs:
  - DOEs had higher carbon intensities than MNEs in the dozen countries with highest carbon intensities except Indonesia.
  - In low carbon intensity countries, MNEs had higher carbon intensities than DOEs with the exception of Cyprus.
  - Largest differences where DOEs exceeded MNEs: Cyprus and Malta.
  - Largest differences where MNEs exceeded DOEs: Hong Kong SAR, Switzerland, and Iceland.
- Industry and country patterns:
  - In high carbon intensity economies (China, South Africa) electricity, gas, and water had the highest intensity.
  - In low carbon intensity economies (Norway, Switzerland) manufacturing industries had the highest intensities.
  - Differences between MNE and DOE intensity were small in high-intensity economies (except electricity, gas and water) and larger in low-intensity economies.

### Trends in selected industries (MNE carbon intensities, 2005–15)
- Manufacturing of basic metals:
  - Highest intensities in 2015: South Africa, India, Russian Federation, China, Saudi Arabia.
  - General downward trend except for India and South Africa.
- Electricity:
  - MNEs in South Africa, Saudi Arabia, India, Estonia, and China showed a general downward trend.
- Manufacture of other non-metallic mineral products:
  - Countries showing general downward trends: Vietnam, Thailand, Philippines, India, China; exception: Vietnam.
- Transport and storage:
  - Highest MNE intensities: Malta, Singapore, Saudi Arabia, China, Thailand.
  - Decline in middle of period followed by an uptick later in period.

### Exports: emissions embodied in MNEs output
- Industry-level observation:
  - Industries with high carbon intensities among MNEs (transport; manufacture of basic metals; chemicals and pharmaceuticals; non-metallic mineral products) have export intensities between 30 and 60 percent.
  - Several low carbon intensity industries (accommodation; manufacture of textiles; electrical equipment; machinery; computer and electronic products) also have high export shares, making their combined effect on domestic emissions significant.
- Country-level observation:
  - Countries with fairly low carbon intensities can have large shares of MNE output exported.
  - Examples where MNEs have relatively low emission intensities but more than half of output is exported: Ireland; Hungary; Iceland; Luxembourg; Cyprus; Slovenia; Switzerland; Costa Rica.
  - Notable exceptions with both high carbon intensity and high export shares: China; Vietnam; Thailand.
- Comparative export shares:
  - Export-related emissions shares are higher for MNEs compared to DOEs in all countries.
  - In countries with lower carbon intensities, the gap between MNEs and DOEs export shares tends to be higher (exceptions include Luxembourg and Lithuania).
  - In higher carbon intensity economies, gaps are smaller (exceptions include China, India, South Africa, Vietnam, Thailand).

*Source: IMF Working Paper (Introduction and Methodology/Results sections from the supplied PDF).*

### Conclusions and Policy Implications

### Conclusions and Policy Implications

### Key empirical findings on FDI and emissions
- The paper presents an experimental approach for estimating the effect of FDI on carbon emissions in host economies through their investment, production, and export related activities.
- The approach:
  - makes use of already available data;
  - is intuitive, and easy to follow and replicate for many countries as data becomes available.
- Estimates were comparable within and between industries and countries.
- Emissions from the investment activity of FDI:
  - have generally fallen between 2005 and 2015;
  - are concentrated in electricity, construction, and manufacturing industries.
- Emissions from ongoing operations of FDI companies (MNEs) were greater than emissions from investment activity funded by FDI.
- For emissions from ongoing operations of MNEs:
  - high carbon emissions were mainly in manufacturing; transport; and the electricity, gas, and water industries;
  - the same industries had the highest carbon intensities.
- The construction industry had a large effect on emissions for FDI-funded GFCF, but was not a major contributor to emissions from the ongoing operations of MNEs.
- Comparison of MNEs and DOEs:
  - MNEs were generally found to have lower contributions to overall emissions and carbon intensities in most sectors compared to DOEs;
  - within individual industries in some countries, MNEs had higher carbon intensities compared to DOEs, especially in countries with low overall carbon intensities.
- Trade and exports:
  - MNEs in industries with high carbon intensities tended to have high export intensities, suggesting an important role of foreign demand.
  - MNEs in countries with the highest export intensities generally had lower carbon intensities with some key exceptions, including China, Vietnam, and Thailand.
- The work highlights the important role of firms operating in one economy but owned by investors in another economy in global carbon emissions.

### Policy implications and recommendations
- For home countries:
  - Policies that incentivize their domestic direct investors to meet high environmental and emissions standards at both home and foreign operations could be important to reducing emissions globally and not just domestically.
  - Such policies could induce firms to use lower carbon production functions and technology at home and abroad, and to demand lower carbon infrastructure and transportation in host economies.
  - Encouraging firms to reduce emissions along their supply chains could lead them to demand that their suppliers reduce their carbon emissions.
- For host economies:
  - Remove barriers to investment in environmental goods and services industries as well as in low carbon technologies to promote positive spillovers and knowledge and technology transfer to the domestic economy.
  - Include an analysis of the impact on carbon emissions as part of FDI attraction strategies.
- Corporate disclosure:
  - Developing a standard for companies to disclose their carbon emissions will provide valuable information to better understand the role of all enterprises, both MNEs and DOEs, in carbon emissions.

### Methodological limitations and avenues for future work
- Noted methodological and data limitations to the framework; future work could address these limitations.
- Possible areas for future work:
  - FDI estimates that distinguish between the use of FDI resources for acquisition of assets versus for greenfield investment and capacity extension.
  - Expanded information on the role of MNEs in carbon emissions such as actual estimates of direct carbon emissions of MNEs by activity/sector.
  - Use of models to capture dynamic interactions.
  - Sensitivity analysis on the differences in direct emissions between MNEs and DOEs to inform how to structure climate related incentive schemes.
- Additional analytical routes using resulting estimates:
  - Explore whether there are spillovers in the form of reduced carbon intensity at DOEs from the operations of MNEs in the host economy.
  - Better understand the relationship between FDI’s role in production and export diversification and the country’s carbon emissions.

### Illustrative numerical examples from the annex (method demonstration)
- Emissions (by final use) totals:
  - Agriculture: 259.7339782 (HFCE) and 51.94679565 (exports) — subtotal 312
  - Manufacturing: 262.8778718 (HFCE) and 262.8778718 (exports) — subtotal 526
  - Services: 194.050786 (HFCE) and 48.51269649 (exports) — subtotal 243
  - Total embodied emissions (HFCE + exports): 1080
- Input-output / multiplier example:
  - Leontief inverse (multipliers) row totals:
    - Agriculture total: 2.42
    - Manufacturing total: 2.73
    - Service total: 2.59
  - Direct emissions coefficients: 0.30 (Agriculture), 0.2 (Manufacturing), 0.1 (Services)
  - Direct and indirect emissions coefficients (emissions multipliers) totals:
    - Total 0.519 (Agriculture), 0.526 (Manufacturing), 0.485 (Services)
- Simple output allocation table (selected entries):
  - Agriculture contributed to Agriculture: multiplier 1.30, direct emissions 0.3, partial direct and indirect 0.389
  - Manufacturing contributed to Manufacturing: multiplier 1.59, direct emissions 0.2, partial direct and indirect 0.317
  - Services contributed to Services: multiplier 1.58, direct emissions 0.1, partial direct and indirect 0.158

*Measuring Carbon Emissions of Foreign Direct Investment in Host Economies — Working Paper No. WP/2022/086*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022086-print-pdf.pdf_
