## wpiea2021008-print-pdf

## Source details

**Canonical URL:** [wpiea2021008-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021008-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021008-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021008-print-pdf.pdf.json)

---

### I. Introduction
- COVID-19 accelerated the digitization of public services by forcing public agencies to switch to online platforms during lockdowns.
- E-government services can increase efficiency, help businesses and households flourish, and drive economic growth.
- Research question: whether the use of e-government services may help host countries remove obstacles or impediments to attracting FDI inflows.
- Definition:
  - E-government activities provide diverse public services to citizens, investors, and other agencies through automated websites in a time-efficient, convenient, one-source, around the clock platform (Alfred, 2002).
- Hypothesis: improvements in e-government operations can enhance locational attractiveness and subsequently attract more FDI inflows.
- Study scope and data:
  - Sample: 178 host countries.
  - Period: 2003-2018.
- Key empirical result:
  - Digital government in the form of e-government services could stimulate the inflow of FDI.
  - Countries that implement and adopt ICTs are able to attract more FDI inflows than countries with weak ICT infrastructure.
  - Finding is robust under different specifications and different estimation methods.

### II. Literature review and theoretical framework
- Dunning’s “eclectic paradigm”:
  - Firms require ownership ‘O’ advantages, host country must provide location ‘L’ advantages, and firms must possess internalization ‘I’ advantages.
  - ‘L’ advantages determine where firms locate production facilities abroad and depend on economic, social, and political factors (Dunning 1977, 1981, 1988).
- Role of ICTs and e-government:
  - Implementation and adoption of ICTs such as e-government are believed to improve location-specific attractiveness and thus attract more FDI.
- Key citations referenced in the literature:
  - Addison and Heshmati (2003); Azubuike (2006); Portes and Rey (2005); Bekaert (1995); Al-Azzam and Abu-Shanab (2014); Alfred (2002); Valentina (2004); Hirst and Norton (1998) cited in Valentina (2004); Al-Kibsi et al.,2001; Cho and Choi (2004); Vindo (2005); Gani and Sharma (2003).

### II.A The role of e-government — three main mechanisms
- Mechanism 1 — cost, time reductions, and efficiency gains:
  - E-government links government departments into a one-stop site, reducing processing time and costs.
  - Singapore example:
    - 21 different forms pre-1989.
    - 15 to 20 days processing time pre-1989.
    - 23 government departments and agencies involved pre-1989.
    - TradeNet launched in October 1989.
    - One online form after TradeNet.
    - Licenses provided within 15 seconds after TradeNet.
- Mechanism 2 — inclusiveness, accountability, and transparency:
  - E-government increases access to government-collected information, making governments more transparent and an effective tool to fight corruption and reduce red tape.
  - Seoul example:
    - Corruption concerns in 1998.
    - OPEN system introduced in 1999.
    - World Bank citation year: 2001.
- Mechanism 3 — increased access to information and reduction of information asymmetries:
  - E-government enhances investor knowledge of institutional behaviors and operations in host economies.
  - Distance-related information asymmetry is an indirect barrier to investment (Portes and Rey, 2005; Bekaert 1995).
  - Empirical evidence cited:
    - Addison and Heshmati (2003) use phones per 1,000 people as a proxy for IT infrastructure and find IT increases FDI to developing countries.
    - Gani and Sharma (2003) find similar results for high-income countries.

### Main conclusion (summary)
- E-government services are positively correlated with higher FDI inflows into host countries.
- Empirical evidence uses an unbalanced panel of 178 developed and developing countries over the period 2003-2018 and supports the hypothesis that better e-government services stimulate FDI inflows.
- The E-Government Development Index (EGDI):
  - Ranges from 0 (zero readiness) to 1 (full readiness).
  - Composite of: (1) Online Service Index (OSI), (2) Telecommunication Infrastructure Index (TII), and (3) Human Capital Index (HCI).

### Empirical findings and regression evidence
- Sample and period reiterated:
  - 178 developed and developing countries.
  - Period: 2003-2018.
- EGDI observations:
  - EGDI examples: Central African Republic 0.07; U.S. 0.88.
  - Regional EGDI averages (2020): Europe (0.82), Asia (0.64), Americas (0.63), Oceania (0.51), Africa (0.39).
  - Top EGDI countries (2020) include: Denmark, Republic of Korea, Estonia, Finland, Australia, Sweden, United Kingdom, New Zealand, U.S., Netherlands, Singapore, Iceland, Norway, Japan.
- Regression results:
  - The coefficient of the e-government variable is positive and significant at the 1.0 percent level in the Fixed Effects estimations for the full sample.
  - Results robust to inclusion of institutional quality, corruption, and political risk controls, and to sample restrictions (developing countries only; excluding small states).
  - Alternative estimation methods (Pooled OLS and Random Effects) yield consistent significance of the E-Gov index on FDI inflows.
- Control variable findings:
  - log (GDP per capita) is positive and highly significant at the 1.0 percent level in all models.
  - Real GDP growth and population growth are positive.
  - Degree of openness to international trade is positive but insignificant in the Fixed Effects models.
  - Quality of institutions is positive and statistically significant in all models.
  - Corruption and political risk are insignificant in Fixed Effects specifications; some specifications show unexpected signs for inflation and political risk.

### Policy implications and recommendations
- Development of digital government services should be a top priority for countries aiming to attract more FDI.
- Prioritizing e-government can:
  - Lower costs of doing business.
  - Raise public service efficiency and effectiveness.
  - Enhance a country’s attractiveness to foreign investors.

### APPENDIX I — The E-Government Development Index
- EGDI definition and aggregation:
  - EGDI = 1/3 (OSI_normalized + TII_normalized + HCI_normalized).
  - EGDI measures readiness and capacity of national institutions to use ICTs to deliver public services.
- Component indices:
  - Online Service Index (OSI): composite normalized score from an Online Service Questionnaire.
  - Telecommunication Infrastructure Index (TII): arithmetic average of five indicators:
    - estimated Internet users per 100 inhabitants
    - number of main fixed telephone lines per 100 inhabitants
    - number of mobile subscribers per 100 inhabitants
    - number of wireless broadband subscriptions per 100 inhabitants
    - number of fixed broadband subscriptions per 100 inhabitants
  - Human Capital Index (HCI): four components:
    - adult literacy rate
    - combined primary, secondary and tertiary gross enrolment ratio
    - expected years of schooling
    - average years of schooling
- Normalization:
  - Z-score standardization applied: Z = (x − η) / σ, where x is the raw score, η is the mean of the sample, σ is the standard deviation of the sample.
  - After Z-score standardization, each component index is normalized to fall between 0 and 1.
  - Overall EGDI is the arithmetic average of the three normalized component indices.
- Source for EGDI methodology: United Nations (2020): “E-Government Survey 2020: Digital Government in the Decade of Action for Sustainable Development”.

### APPENDIX II — Definitions of Variables and Descriptive Statistics
- Variable definitions and sources:
  - FDI inflows: The World Bank’s World development Indicators, 2020
  - RGDPG: Real GDP growth — The World Bank’s World development Indicators, 2020
  - Inflation: Annual inflation rate — The World Bank’s World development Indicators, 2020
  - Population: General government gross debt as % of GDP — The World Bank’s World development Indicators, 2020
  - Openness: Exports and Imports as a % of GDP — The World Bank’s World development Indicators
  - E-Gov Index: E-Government Development Index — The UN’s e-Government Knowledgebase
  - Political Risk: Political Terror Scale-US State Department, 1 (very low), and 5 (very high)
  - Quality of Institutions: Index measuring law and order ranging from 0 (very low) to 6 (very high) — International Country Risk Guide
  - Corruption level: Corruption level 0 (very low) to 6 (very high) — International Country Risk Guide
- Sample: 178 host countries: 2003-2018
- Selected descriptive statistics (Appendix Table 1):
  - Log (FDI/PoP): Obs 1,478; Mean 4.75; Std. Dev. 2.06; Min -5.2; Max 11.3
  - log (GDP per capita): Obs 1,478; Mean 8.44; Std. Dev. 1.46; Min 5.3; Max 11.6
  - Real GDP growth: Obs 1,478; Mean 4.23; Std. Dev. 5.31; Min -33.1; Max 123.1
  - Open: Obs 1,478; Mean 88.04; Std. Dev. 49.19; Min 0.2; Max 437.3
  - Inflation: Obs 1,478; Mean 5.94; Std. Dev. 8.74; Min -31.5; Max 157.0
  - Population growth: Obs 1,478; Mean 1.49; Std. Dev. 1.55; Min -9.1; Max 16.5
  - E-government index: Obs 1,478; Mean 0.45; Std. Dev. 0.21; Min 0.0; Max 0.9
  - Online services: Obs 1,478; Mean 0.39; Std. Dev. 0.26; Min 0.0; Max 1.0
  - Institutional quality: Obs 1,102; Mean 3.65; Std. Dev. 1.28; Min 0.5; Max 6.0
  - Corruption index: Obs 1,102; Mean 3.41; Std. Dev. 1.14; Min 0.0; Max 6.0
  - Political Risk: Obs 1,476; Mean 2.44; Std. Dev. 1.10; Min 1.0; Max 5.0
- Selected correlations (Appendix Table 2):
  - log (FDI/Pop) with E-Government index: 0.69
  - log (FDI/Pop) with Online services: 0.53
  - log (FDI/Pop) with Log (GDP per capita): 0.79
  - E-Government index with Online services: 0.91
  - E-Government index with Log (GDP per capita): 0.85
  - Institutional quality with Corruption index: -0.67

*Source: IMF working paper (wpiea2021008-print-pdf).*

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

### References

### I. Introduction
- COVID-19 accelerated the digitization of public services globally by forcing public agencies to switch to online platforms during lockdowns.
- E-government services can increase efficiency, help businesses and households flourish, and drive economic growth.
- Research question: whether the use of e-government services may help host countries remove obstacles or impediments to attracting FDI inflows.
- Definition and scope:
  - E-government activities provide diverse public services to citizens, investors, and other agencies through automated websites in a time-efficient, convenient, one-source, around the clock platform (Alfred, 2002).
  - Hypothesis: improvements in e-government operations can enhance locational attractiveness and subsequently attract more FDI inflows.
- Study scope and data:
  - Empirical assessment uses data for 178 host countries, both advanced and developing, over the period 2003-2018.
- Key empirical result:
  - The empirical results support the hypothesis that digital government in the form of e-government services could stimulate the inflow of FDI.
  - Specifically, countries that implement and adopt ICTs are able to attract more FDI inflows than countries with weak ICT infrastructure.
  - Finding is robust under different specifications and different estimation methods.
- Paper organization:
  - Section 2: literature review and relationship between e-government and FDI inflows.
  - Section 3: model and econometric estimation method.
  - Section 4: empirical results.
  - Final section: conclusions and policy implications.

### II. Literature review — theoretical framework
- Dunning’s “eclectic paradigm” (1977, 1981, and 1988):
  - Firms require ownership ‘O’ advantages that are transferable to foreign markets.
  - Host country must provide location ‘L’ advantages that allow firms to reap benefits of ‘O’ advantages.
  - Firms must possess internalization ‘I’ advantages to reduce transactions costs.
  - ‘L’ advantages determine where firms locate production facilities abroad and depend on economic, social, and political factors.
- Role of ICTs and e-government:
  - Implementation and adoption of ICTs such as e-government are believed to improve location-specific attractiveness and thus attract more FDI.

### II.A The role of e-government — three main mechanisms
- Mechanism 1 — cost, time reductions, and efficiency gains:
  - E-government links government departments into a one-stop site, reducing processing time and costs compared with traditional operations.
  - Example: prior to 1989, obtaining an import/export license in Singapore required filling out 21 different forms and waiting 15 to 20 days for 23 government departments and agencies to process the request; after TradeNet launched in October 1989, only one online form was required and licenses were provided within 15 seconds (Al-Kibsi et al.,2001).
- Mechanism 2 — inclusiveness, accountability, and transparency:
  - E-government increases access to government-collected information, making governments more transparent and an effective tool to fight corruption and reduce red tape.
  - Example: Seoul’s Online Procedures Enhancement (OPEN) system introduced in 1999 as an anti-corruption web portal following corruption concerns raised in 1998; cited sources include Cho and Choi, 2004 and the World Bank, 2001.
- Mechanism 3 — increased access to information and reduction of information asymmetries:
  - E-government enhances investor knowledge of institutional behaviors and operations in host economies.
  - Information asymmetry associated with greater distance is an indirect barrier to investment; distance impedes portfolio diversification and international asset flows (Portes and Rey, 2005; Bekaert 1995).
  - Empirical literature: Addison and Heshmati (2003) use phones per 1,000 people as a proxy for IT infrastructure and find IT increases FDI to developing countries; Gani and Sharma (2003) find similar results for high-income countries.

### Key statistics and exact figures cited
- 178 host countries (sample size).
- Period: 2003-2018.
- Dunning years: 1977, 1981, 1988.
- Singapore example:
  - 21 different forms pre-1989.
  - 15 to 20 days processing time pre-1989.
  - 23 government departments and agencies involved pre-1989.
  - TradeNet launched in October 1989.
  - One online form after TradeNet.
  - Licenses provided within 15 seconds after TradeNet.
- Seoul example:
  - Corruption concerns in 1998.
  - OPEN system introduced in 1999.
  - World Bank citation year: 2001.
- Empirical citations:
  - Addison and Heshmati (2003).
  - Azubuike (2006).
  - Portes and Rey (2005).
  - Bekaert (1995).
  - Al-Azzam and Abu-Shanab (2014).
  - Alfred (2002).
  - Valentina (2004).
  - Hirst and Norton (1998), cited in Valentina, 2004.
  - Al-Kibsi et al.,2001.
  - Cho and Choi, 2004.
  - Vindo, 2005.
  - Gani and Sharma (2003).

_Italic: wpiea2021008-print-pdf - References_

### conclusion. IT achievement and diffusion, such as mobile phones and internet hosts, are

### wpiea2021008-print-pdf - conclusion. IT achievement and diffusion, such as mobile phones and internet hosts, are

### Main conclusion
- E-government services are positively correlated with higher FDI inflows into host countries.
- The empirical evidence, using an unbalanced panel of 178 developed and developing countries over the period 2003-2018, supports the hypothesis that better e-government services stimulate FDI inflows.
- The E-Government Development Index (EGDI) ranges from 0 (zero readiness) to 1 (full readiness) and is a composite of: (1) the Online Service Index, (2) the Telecommunication Infrastructure Index (TII), and (3) the Human Capital Index (HCI).

### Empirical findings
- Sample and period:
  - 178 developed and developing countries
  - Period: 2003-2018
- Key EGDI observations:
  - EGDI examples: Central African Republic 0.07; U.S. 0.88.
  - Regional EGDI averages (2020): Europe (0.82), Asia (0.64), Americas (0.63), Oceania (0.51), Africa (0.39).
  - Top EGDI countries (2020): Denmark, Republic of Korea, Estonia, Finland, Australia, Sweden, United Kingdom, New Zealand, U.S., Netherlands, Singapore, Iceland, Norway, Japan.
- Regression evidence:
  - The coefficient of the e-government variable is positive and significant at the 1.0 percent level in the Fixed Effects estimations for the full sample, indicating that accessibility of government information online stimulates FDI inflows.
  - Results are robust to inclusion of institutional quality, corruption, and political risk controls, and to sample restrictions (developing countries only; excluding small states).
  - Alternative estimation methods (Pooled OLS and Random Effects) yield consistent significance of the E-Gov index on FDI inflows.
- Control variables:
  - log (GDP per capita) is positive and highly significant at the 1.0 percent level in all models.
  - Real GDP growth and population growth are positive, implying forward-looking foreign investors.
  - Degree of openness to international trade is positive but insignificant in the Fixed Effects models.
  - Quality of institutions is positive and statistically significant in all models.
  - Corruption and political risk are found to be insignificant in the Fixed Effects specifications; results for inflation and political risk show some coefficients with unexpected signs in some specifications.

### Mechanisms and interpretation
- E-government reduces processing time and bureaucratic inefficiencies by linking government departments and improving information flow across agencies.
- E-government increases inclusiveness, accountability, and transparency, which helps reduce red tape and opportunities for corruption.
- Lower administrative costs and improved potential returns on investment make countries with stronger e-government more attractive to FDI.

### Policy implications and recommendations
- The development of digital government services should be a top priority for countries aiming to attract more FDI.
- Prioritizing e-government can yield multiple benefits: lowering costs of doing business, raising public service efficiency and effectiveness, and enhancing a country’s attractiveness to foreign investors.

*Source: IMF working paper (conclusion section of wpiea2021008-print-pdf).*

### APPENDIX I

### APPENDIX I — The E-Government Development Index

### EGDI definition and aggregation
- The E-government development index (EGDI) measures the readiness and capacity of national institutions to use ICTs to deliver public services.
- EGDI is a survey composite index based on a weighted average of three normalized indices:
  - EGDI = 1/3 (OSI_normalized + TII_normalized + HCI_normalized)

### Component indices and their composition
- Online Service Index (OSI)
  - The OSI is a composite normalized score derived on the basis on an Online Service Questionnaire.
- Telecommunication Infrastructure Index (TII)
  - The TII is an arithmetic average composite of five indicators:
    - estimated Internet users per 100 inhabitants
    - number of main fixed telephone lines per 100 inhabitants
    - number of mobile subscribers per 100 inhabitants
    - number of wireless broadband subscriptions per 100 inhabitants
    - number of fixed broadband subscriptions per 100 inhabitants
- Human Capital Index (HCI)
  - The HCI consists of four components:
    - adult literacy rate
    - the combined primary, secondary and tertiary gross enrolment ratio
    - expected years of schooling
    - average years of schooling

### Normalization and Z-score standardization
- Prior to normalization, a Z-score standardization procedure is implemented for each index to ensure equal influence of the three component indexes.
- Standard Z-score calculation for each component indicator:
  - Z = (x − η) / σ
    - x is a raw score to be standardized
    - η is the mean of the sample
    - σ is the standard deviation of the sample
- After Z-score standardization, each component index is normalized to fall between 0 and 1.
- The overall EGDI is derived by taking the arithmetic average of the three normalized component indices.

*Source: United Nations (2020): “E-Government Survey 2020: Digital Government in the Decade of Action for Sustainable Development”, New York, United States.*

---

### APPENDIX II — Definitions of Variables and Descriptive Statistics

### Variable definitions and sources
- FDI inflows: Foreign direct investment inflows — The World Bank’s World development Indicators, 2020
- RGDPG: Real GDP growth — The World Bank’s World development Indicators, 2020
- Inflation: Annual inflation rate — The World Bank’s World development Indicators, 2020
- Population: General government gross debt as % of GDP — The World Bank’s World development Indicators, 2020
- Openness: Exports and Imports as a % of GDP — The World Bank’s World development Indicators
- E-Gov Index: E-Government Development Index — The UN’s e-Government Knowledgebase
- Political Risk: Political Terror Scale-US State Department, 1 (very low), and 5 (very high) — http://politicalterrorscale.org/
- Quality of Institutions: Index measuring law and order ranging from 0 (very low) to 6 (very high) — International Country Risk Guide, the Political Risk Groups
- Corruption level: Corruption level 0 (very low) to 6 (very high) — International Country Risk Guide, the Political Risk Groups

### Sample
- Sample: 178 host countries: 2003-2018

### Appendix Table 1: Descriptive Statistics (selected variables)
- Observations and summary statistics
  - Log (FDI/PoP): Obs 1,478; Mean 4.75; Std. Dev. 2.06; Min -5.2; Max 11.3
  - log (GDP per capita): Obs 1,478; Mean 8.44; Std. Dev. 1.46; Min 5.3; Max 11.6
  - Real GDP growth: Obs 1,478; Mean 4.23; Std. Dev. 5.31; Min -33.1; Max 123.1
  - Open: Obs 1,478; Mean 88.04; Std. Dev. 49.19; Min 0.2; Max 437.3
  - Inflation: Obs 1,478; Mean 5.94; Std. Dev. 8.74; Min -31.5; Max 157.0
  - Population growth: Obs 1,478; Mean 1.49; Std. Dev. 1.55; Min -9.1; Max 16.5
  - E-government index: Obs 1,478; Mean 0.45; Std. Dev. 0.21; Min 0.0; Max 0.9
  - Online services: Obs 1,478; Mean 0.39; Std. Dev. 0.26; Min 0.0; Max 1.0
  - Institutional quality: Obs 1,102; Mean 3.65; Std. Dev. 1.28; Min 0.5; Max 6.0
  - Corruption index: Obs 1,102; Mean 3.41; Std. Dev. 1.14; Min 0.0; Max 6.0
  - Political Risk: Obs 1,476; Mean 2.44; Std. Dev. 1.10; Min 1.0; Max 5.0
- Source: Author's calculations

### Appendix Table 2: Correlation matrix (selected pairwise correlations)
- log (FDI/Pop) correlations
  - with E-Government index: 0.69
  - with Online services: 0.53
  - with Log (GDP per capita): 0.79
  - with Real GDP growth: -0.10
  - with Population growth: -0.16
  - with Open: 0.50
  - with Inflation: -0.23
  - with Institutional quality: 0.54
  - with Corruption index: -0.58
- E-Government index correlations
  - with Online services: 0.91
  - with Log (GDP per capita): 0.85
  - with Real GDP growth: -0.20
  - with Population growth: -0.39
  - with Open: 0.26
  - with Inflation: -0.25
  - with Institutional quality: 0.58
  - with Corruption index: -0.67
- Online services correlations
  - with Log (GDP per capita): 0.66
  - with Real GDP growth: -0.16
  - with Population growth: -0.31
  - with Open: 0.15
  - with Inflation: -0.24
  - with Institutional quality: 0.45
  - with Corruption index: -0.58
- Log (GDP per capita) correlations
  - with Real GDP growth: -0.16
  - with Population growth: -0.24
  - with Open: 0.31
  - with Inflation: -0.26
  - with Institutional quality: 0.64
  - with Corruption index: -0.68
- Real GDP growth correlations
  - with Population growth: 0.14
  - with Open: 0.00
  - with Inflation: -0.06
  - with Institutional quality: -0.05
  - with Corruption index: 0.18
- Population growth correlations
  - with Open: -0.04
  - with Inflation: 0.07
  - with Institutional quality: -0.13
  - with Corruption index: 0.19
- Open correlations
  - with Inflation: -0.09
  - with Institutional quality: 0.27
  - with Corruption index: -0.25
- Inflation correlations
  - with Institutional quality: -0.23
  - with Corruption index: 0.28
- Institutional quality correlations
  - with Corruption index: -0.67

### Appendix Table 3: Country sample
- Country list includes (partial extract): Afghanistan, Chile, Guyana, Mauritania, Rwanda, United Kingdom, Albania, China, Haiti, Mauritius, St. Kitts and Nevis, Tanzania, Algeria, Colombia, Honduras, Mexico, St. Lucia, United States, Angola, Comoros, Hungary, Mongolia, Sao Tome and Principe, Burkina Faso, Antigua and Barbuda, Congo, Rep., Iceland, Moldova, Samoa, Uruguay, Azerbaijan, Congo, Dem. Rep., India, Morocco, Saudi Arabia, Uzbekistan, Argentina, Costa Rica, Indonesia, Mozambique, Senegal, Venezuela, RB, Australia, Croatia, Iran, Islamic Rep., Oman, Serbia, Yemen, Rep., Austria, Cyprus, Iraq, Namibia, Seychelles, Zambia, Bahamas, The, Czech Republic, Ireland, Nauru, Sierra Leone, Bahrain, Benin, Italy, Nepal, Singapore, Bangladesh, Denmark, Cote d'Ivoire, Netherlands, Slovak Republic, Armenia, Dominica, Jamaica, Vanuatu, Vietnam, Barbados, Dominican Republic, Japan, New Zealand, Slovenia, Belgium, Ecuador, Kazakhstan, Nicaragua, South Africa, Bhutan, Egypt, Arab Rep., Jordan, Niger, Zimbabwe, Bolivia, El Salvador, Kenya, Nigeria, Spain, Bosnia and Herzegovina, Equatorial Guinea, Korea, Rep., Norway, Sudan, Botswana, Ethiopia, Kuwait, Micronesia, Fed. Sts., Suriname, Brazil, Eritrea, Kyrgyz Republic, Marshall Islands, Eswatini, Belize, Estonia, Lao PDR, Palau, Sweden, Solomon Islands, Fiji, Lebanon, Pakistan, Switzerland, Brunei Darussalam, Finland, Lesotho, Panama, Tajikistan, Bulgaria, France, Latvia, Papua New Guinea, Thailand, Myanmar, Gabon, Liberia, Paraguay, Togo, Burundi, Georgia, Libya, Peru, Tonga, Belarus, Gambia, The, Lithuania, Philippines, Trinidad and Tobago, Cambodia, Germany, Luxembourg, Poland, United Arab Emirates, Cameroon, Ghana, Madagascar, Portugal, Tunisia, Canada, Kiribati, Malawi, Guinea-Bissau, Turkey, Cabo Verde, Greece, Malaysia, Timor-Leste, Turkmenistan, Central African Republic, Grenada, Maldives, Qatar, Tuvalu, Sri Lanka, Guatemala, Mali, Romania, Uganda, Chad, Guinea, Malta, Russian Federation, Ukraine.

*Source: Author's calculations*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021008-print-pdf.pdf_