## wpiea2022257-print-pdf

## Source details

**Canonical URL:** [wpiea2022257-print-pdf](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022257-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022257-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022257-print-pdf.pdf.json)

---

### Overview
- Fighting the COVID-19 pandemic required vaccinations; ending it requires equal distribution of vaccinations.
- Scientific consensus: the coronavirus is contagious and mutable, meaning "no one is safe until everyone is safe" and the only way to stop the pandemic is to stamp out the virus simultaneously worldwide.
- The paper examines drivers of COVID-19 vaccination rates across countries during 2021–22 and quantifies the equitable distribution of vaccinations using the principle of equality of opportunity.

### Vaccination distribution statistics (as reported)
- Data source note: Our World in Data; chart notes state "As of August 18, 2022".
- Regional vaccination rates as of mid-August 2022:
  - Sub-Saharan Africa (SSA): 19.5 percent of the region’s population was fully vaccinated.
  - Other emerging markets and developing economies (excluding China): 61 percent fully vaccinated.
  - Advanced economies: 75 percent fully vaccinated.
- China excluded from some series due to delayed reporting beginning August 2021.

### Framing of equity and responsibility
- Key question: who is more responsible for unequal vaccination rates—vaccine suppliers or recipient countries?
- Equitable distribution framed via equality of opportunity: vaccination rates should not depend on circumstances beyond a country’s control.
- Operational assumption: delivery of vaccine doses is treated as the primary factor beyond a country’s control; administering rates capture factors within a country’s control (including hesitancy and distribution capacity).

### Vaccine logistics (process and constraints)
- End-to-end supply chain stages:
  - Recipient countries find suppliers and sign purchase agreements (COVAX can act as intermediary).
  - Suppliers produce vaccines and schedule deliveries; production disruptions and export restrictions can cause delays.
  - Countries prepare distribution schemes, health infrastructure, train health workers, and generate demand.
  - Cold-chain requirements and hard-to-reach populations complicate in-country distribution.
- Examples of disruptions:
  - Export controls on inputs (e.g., syringes and vials) and industrial accidents hampered manufacturing.
  - In April 2021, India halted exports of AstraZeneca vaccines produced by the Serum Institute of India, affecting supplies to COVAX; the ban was rescinded in October 2021.

### Conceptual decomposition of vaccination outcomes
- Two key milestones: delivery of vaccine doses (d_i) and vaccine administration (a_i).
- Accounting identity:
  - v_i = d_i ∙ a_i
  - In logarithms: log v_i = log d_i + log a_i.
- Empirical forms:
  - log v_i = f(log d_i, X_i)
  - log a_i = g(X_i)
  - X_i is a vector of economic, health, and demographic factors.
- Linear regression specification for equitable distribution:
  - log v_i = β X_i^R + γ X_i^NR + ε_i
    - X_i^R: domestic factors within country’s control (in logs), including COVID-19 cases and deaths, GDP per capita, infrastructure quality, urbanization, government spending on health, demographic structure, and human development index.
    - X_i^NR: factors beyond country control; here contains only delivery of contracted or agreed vaccine doses (in logs).
    - ε_i captures unobservables such as vaccine hesitancy.
  - Equitable distribution constructed as:
    - log v_i^* = β X_i^R + C, where C ensures ∑ v_i = ∑ v_i^*.

### Inequality measurement and decomposition
- Inequality index chosen: Gini index as measure of overall inequality.
- Decomposition approach: Shapley value decomposition following Shorrocks (2013) to attribute shares of overall inequality to contributing factors.
- Interpretation: the share of vaccine delivery in overall vaccination inequality is the ratio of its contribution to the overall Gini index; since delivery is treated as the only factor beyond country control, its contribution represents inequitable distribution of global vaccinations.
- Annex robustness: Shapley value decomposition results are robust to using other inequality indices, such as a MN-measure (Magdalou and Nock 2011), which attaches higher weights to shortfalls from the equitable distribution and encompasses mean log deviation as a special case.

### Data sources and construction
- Vaccine delivery: monthly deliveries by vaccine type, November 2020 to March 2022, from Airfinity.
- Vaccination rates, COVID-19 cases and deaths: daily data from Our World in Data; monthly cumulative vaccination rate uses the highest vaccination rate reported in each month for each country.
- Vaccination uptake: average vaccination coverage from Airfinity (sourced from WHO and UNICEF) for polio, diphtheria tetanus toxoid and pertussis, hepatitis B, measles, and Hib disease.
- Median travel time: Nelson et al (2019) pixel-level travel time to nearest city (population between 0.5 and 1 million); country median of pixel values used as proxy for difficulty reaching remote populations.
- Economic, health, demographic indicators: GDP per capita, infrastructure quality, urban share, health spending, share of 15+ population aged 15-64, share aged 65+ from World Bank WDI; Human development index from UNDP.
- Summary from Table 1 (as of 2022Q1): Vaccination rates and cumulative delivery (per 100 people) are lowest in sub-Saharan Africa (SSA) and highest in advanced economies. SSA has the lowest cumulative cases and deaths (per million), GDP per capita, health expenditure per capita, infrastructure quality score, human development index, and urban population share. SSA share of 15+ population is 61 percent, compared to 84 percent in advanced economies and 75 percent in other EMDEs.

### Findings — Vaccine delivery and vaccination rates
- Worldwide, about 76 percent of delivered vaccine doses have been administered.
- Cumulative delivery and vaccination rates are highly correlated; most country observations are close to the 45-degree line in delivery vs vaccination plots (Figure 3, as of 2022Q1).
- Delivery rates in SSA are markedly lower than in the rest of the world.
- Regression evidence (log vaccination rate on log cumulative delivery, with SSA dummy and interaction) across dates:
  - Log cumulative delivery coefficients: 1.16*** (2021Q2), 1.18*** (2021Q3), 1.13*** (2021Q4), 1.16*** (2022Q1).
  - SSA dummy and interaction with cumulative delivery are statistically insignificant across quarters; interaction is negative in 2021Q2 and 2021Q3 and positive in 2021Q4 and 2022Q1 (all statistically insignificant).
- Interpretation: Despite a slow start, SSA countries administered delivered vaccines at a similar—if not faster—speed compared to the rest of the world since the second half of 2021.

### Findings — Role of vaccine type and source
- mRNA-based vaccines tend to give a strong boost to vaccination rates; other vaccine types either reduce vaccination rates or have no statistically significant impact (Annex II).
- Vaccine source matters:
  - In SSA countries, only vaccine deliveries from COVAX have positive and statistically significant impact on vaccination rates.
  - In the rest of the world, direct purchases from manufacturers and, to a lesser extent, bilateral donations are associated with higher vaccination rates (Annex II).
- Implication: Low delivery rates in SSA reflect a combination of limited funds and insufficient donations.

### Factors affecting vaccination rates (multivariate evidence)
- Sequential regressions (as of 2022Q1) show individual factors positively associated with vaccination rates: higher COVID-19 cases, higher COVID-19 deaths, higher GDP per capita, better infrastructure quality, higher health spending, older population, and higher human development index.
- Example elasticity interpretation: Column 1 implies when cumulative delivery increases by 10 percent, vaccination rates increase by 12.8 percent.
- When all factors are included together (single regression, column 9), only cumulative delivery and quality of infrastructure remain statistically significant determinants of vaccination rates.
- The last-row calculation: if each individual factor increases from the 25th percentile to the 75th percentile, vaccination rates would increase by about 100 percent.
- Policy implication: Vaccine logistics—international delivery and domestic distribution infrastructure—are key to getting people vaccinated.

### Administering rates (administered doses / delivered doses)
- Historical vaccination uptake correlates with higher average administering rates, but there is substantial heterogeneity; some countries with low historical uptake achieved high COVID-19 administering rates.
- No correlation between marginal administering rates (Δa_i / Δd_i over 2022Q1) and vaccination rates as of 2021Q4.
- Regression evidence (administering rate as dependent variable, as of 2022Q1):
  - Columns 1-8: higher historical vaccination uptake, higher COVID-19 cases and deaths, higher GDP per capita, better infrastructure quality, higher health spending per capita, older population, and higher human development index are associated with higher administering rates.
  - If each individual factor increases from 25th percentile to 75th percentile, administering rate increases by about 20 percent.
  - In single regression controlling for all factors, coefficients become imprecisely estimated.
- Robustness: Results qualitatively unchanged when administering rates measured at 2021Q3 and 2021Q4.

### Excess doses of vaccines
- Potential measure of unused vaccines: cumulative delivered doses minus cumulative administered doses, d_i (1 − a_i), but this may overstate excess due to reporting lags and timing mismatches.
- United States example: excess doses rising steadily since April 2021, reaching almost 200 per 100 people in April 2022.
- Across countries: advanced economies (e.g., United States, United Kingdom) have accumulated far more excess doses than developing countries.
- Global implication: reallocating unused doses from countries with excess to those with demand can substantially increase recipient countries' vaccination rates at the same administering rate.

### Equitable distribution and actual vaccination rates
- Equitable distribution constructed from domestic factors implies:
  - log v_i^* = β X_i^R + C with ∑ v_i = ∑ v_i^*.
- Figure 6 (as of 2022Q1) comparison:
  - Equitable distribution of vaccinations is close to—but not identical with—equal distribution.
  - Most countries under the equitable distribution scenario are at 100 doses per 100 people, reflecting global vaccine supply capacity in early 2022.
  - Most SSA countries would have higher vaccination rates under the equitable distribution.

### Evolution of vaccination inequality and decomposition
- Global vaccination inequality (Gini index across countries) has declined over time but remained as high as 0.4 at the end of 2022Q1.
- Shapley value decomposition (Figure 8) of contributors to global vaccination inequality (based on column 10 of Table 3):
  - Vaccine delivery is the dominant factor, accounting for between 70 and 80 percent of global vaccination inequality during 2021Q2–2022Q1.
  - Other health, economic, and demographic factors (COVID-19 cases, GDP per capita, urban share, share of 15+ population) play a minor role.
  - Residual (capturing unobserved factors such as vaccine hesitancy) increases slightly over 2021, peaking at only 13 percent of global vaccination inequality.
- Conclusion: Vaccine delivery (access to doses) is far more important for global vaccination inequality than vaccine hesitancy or other domestic factors.
- Under the MN-measure, the inequitable component driven by vaccine delivery accounted for about 90 percent of the overall global vaccination inequality in the period shown.

### Vaccine types, sources, and regional implications (additional regression evidence)
- Higher share of mRNA vaccines is associated with higher vaccination rates (positive association strengthens when the United States is excluded).
- Higher shares of AstraZeneca and Chinese vaccines are associated with lower vaccination rates in some specifications; however, none of the vaccine-type coefficients are statistically significant in the full-sample specifications reported.
- For SSA, the interaction term of the SSA dummy and share of mRNA vaccines is statistically significant, indicating that mRNA vaccines might boost vaccination rates in SSA.
- Regression evidence on delivery sources (log vaccination rate, doses per 100 people, as of 2022Q1):
  - Manufacturers delivery: coefficient 0.33***.
  - Bilateral donations delivery: coefficient 0.20**.
  - Interaction SSA × Log COVAX delivery: positive and statistically significant (e.g., 0.82** and 1.30*** in reported specifications).
  - SSA dummy is negative and statistically significant in several specifications (e.g., SSA -3.59***, -4.22***, -4.30***), reflecting lower vaccination rates net of delivery and other controls.

### Robustness checks
- Results robust to population-weighted regressions: Log cumulative delivery coefficients reported as 1.28***, 1.37***, 1.30*** in weighted specifications.
- Adjusting for vaccine types and using share of people with at least one dose as dependent variable (adjusted deliveries dividing multi-dose vaccines by 2 except J&J): Log cumulative delivery coefficients reported as 1.21***, 1.36***, 1.33*** in adjusted specifications.
- Infrastructure quality score remains statistically significant in specifications that include multiple covariates.

### Conclusion and policy implications
- Uneven global vaccination rates: low- and middle-income countries have much lower vaccination rates than advanced countries.
- Delivery of vaccine doses accounts for about 80 percent of global vaccination inequality.
- Vaccine hesitancy accounts for at most about 13 percent of global vaccination inequality.
- Delivery of vaccine doses and quality of infrastructure appear to matter the most for vaccination rates, after controlling for factors within countries’ control.
- SSA countries received vaccine deliveries later than other regions, but their administering rates are comparable to those of other countries when deliveries are available.
- Policy recommendations:
  - Strengthen global collaboration and systems for faster and better pandemic response to reach global vaccination targets.
  - Stronger multilateral institutions with high-level political support to coordinate global response to health shocks.
  - Rapid and adequate financing windows for global public goods activated at the onset of pandemics.
  - Diversified regional manufacturing capacity with arrangements to share technology and know-how.
  - Improved regulatory frameworks to allow speedy delivery of existing and novel tools worldwide.
  - Countries should invest in strengthening health systems to prevent, detect and respond to future threats.
  - The international community should allocate additional resources to strengthen health systems and vaccine logistics—only collective action can improve outcomes for future pandemics.

*Source: wpiea2022257-print-pdf*

### 1.    Introduction

### 1.    Introduction

### Overview
- Fighting the COVID-19 pandemic required vaccinations; ending it requires equal distribution of vaccinations.
- Scientific consensus: the coronavirus is contagious and mutable, meaning "no one is safe until everyone is safe" and the only way to stop the pandemic is to stamp out the virus simultaneously worldwide.
- The paper examines drivers of COVID-19 vaccination rates across countries during 2021–22 and quantifies the equitable distribution of vaccinations using the principle of equality of opportunity.

### Vaccination distribution statistics (as reported)
- Data source note: Our World in Data; chart notes state "As of August 18, 2022".
- Regional vaccination rates as of mid-August 2022:
  - Sub-Saharan Africa (SSA): 19.5 percent of the region’s population was fully vaccinated.
  - Other emerging markets and developing economies (excluding China): 61 percent fully vaccinated.
  - Advanced economies: 75 percent fully vaccinated.
- China excluded from some series due to delayed reporting beginning August 2021.

### Framing of equity and responsibility
- Key question: who is more responsible for unequal vaccination rates—vaccine suppliers or recipient countries?
- The paper frames equitable distribution via equality of opportunity: vaccination rates should not depend on circumstances beyond a country’s control.
- Delivery of vaccine doses is treated as the primary factor beyond a country’s control; administering rates capture factors within a country’s control (including hesitancy and distribution capacity).

### Vaccine logistics (process and constraints)
- End-to-end supply chain stages highlighted:
  - Recipient countries find suppliers and sign purchase agreements (COVAX can act as intermediary).
  - Suppliers produce vaccines and schedule deliveries; production disruptions and export restrictions can cause delays.
  - Countries prepare distribution schemes, health infrastructure, train health workers, and generate demand.
  - Cold-chain requirements and hard-to-reach populations complicate in-country distribution.
- Examples of disruptions:
  - Export controls on inputs (e.g., syringes and vials) and industrial accidents hampered manufacturing.
  - In April 2021, India halted exports of AstraZeneca vaccines produced by the Serum Institute of India, affecting supplies to COVAX; the ban was rescinded in October 2021.

### Conceptual decomposition of vaccination outcomes
- Two key milestones: delivery of vaccine doses (d_i) and vaccine administration (a_i).
- Accounting identity used:
  - v_i = d_i ∙ a_i
  - In logarithms: log v_i = log d_i + log a_i. (Equation (1) in source)
- Interpretation:
  - d_i is largely beyond a recipient country’s control (timing and quantity of deliveries).
  - a_i reflects domestic efforts, infrastructure, and vaccine hesitancy.
- Two empirical forms used:
  - log v_i = f(log d_i, X_i) (Equation (2))
  - log a_i = g(X_i) (Equation (3))
  - X_i is a vector of economic, health, and demographic factors.

### Equitable access definition and empirical implementation
- Principle: distribution is equitable if vaccination rates do not depend on circumstances beyond a country’s control.
- Operational assumption for tractability: vaccine delivery is the only factor beyond recipient countries' control; administering rates and other factors are within control (noting caveats that some domestic factors may be hard to change quickly and securing deals can affect deliveries).
- Linear regression specification used to compute equitable distribution:
  - log v_i = β X_i^R + γ X_i^NR + ε_i (Equation (4))
    - X_i^R: vector of domestic factors within country’s control (in logs), including COVID-19 cases and deaths, GDP per capita, infrastructure quality, urbanization, government spending on health, demographic structure, and human development index.
    - X_i^NR: vector of factors beyond country’s immediate control; in this paper X_i^NR contains only delivery of contracted or agreed vaccine doses (in logs).
    - ε_i captures unobservables such as vaccine hesitancy.
  - Equitable distribution constructed as:
    - log v_i^* = β X_i^R + C, where C ensures ∑ v_i = ∑ v_i^* (Equation (5)).
    - v^* is a redistribution of vaccination rates based on domestic factors, holding total administered doses fixed.

### Inequality measurement and decomposition
- Inequality index chosen: Gini index as measure of overall inequality.
- Decomposition approach: Shapley value decomposition following Shorrocks (2013) to attribute shares of overall inequality to contributing factors.
- Interpretation: the share of vaccine delivery in overall vaccination inequality is the ratio of its contribution to the overall Gini index; since delivery is treated as the only factor beyond country control, its contribution represents inequitable distribution of global vaccinations.

### Key findings and policy implications (summary statements from the introduction)
- The paper finds that access to vaccines, as measured by delivery of vaccine doses, was the dominant reason behind vaccine inequality, accounting for more than 70 percent of the vaccine inequality.
- Other socio-economic factors—per capita income, urbanization, spending on health, demographic structure, and human development index—play a relatively minor role.
- Policy implications highlighted:
  - Ensuring timely production and delivery of vaccine doses plays a crucial role in reducing unequitable distribution of vaccinations.
  - Improving domestic distribution networks and promoting proper vaccination campaigns are also key to driving up vaccination rates.
- Related empirical findings from cited literature:
  - Agarwal and Reed (2022) find that 60-75 percent of the delay in vaccine deliveries to low- and middle-income countries is attributable to their signing purchase agreements later than high-income countries.

### Paper roadmap
- Section 2: reviews related literature and provides overview of vaccine logistics.
- Section 3: discusses methodological approach and data.
- Section 4: presents results.
- Section 5: concludes.

*Source: wpiea2022257-print-pdf - 1.    Introduction*

### Annex I provides further details on the Shapley value decomposition. We show that the results of

### wpiea2022257-print-pdf - Annex I provides further details on the Shapley value decomposition. We show that the results of

### Annex: Shapley value decomposition robustness
- Results of the Shapley value decomposition are robust to using other inequality indices, such as a MN-measure (Magdalou and Nock 2011).
- The MN-measure attaches higher weights to shortfalls from the equitable distribution (Hufe et al 2021) and encompasses mean log deviation as a special case when the equitable distribution is the equal distribution.
- Note: Most inequality indices require non-negative inputs; a log transformation is sometimes needed.

### Data sources and construction
- Vaccine delivery: monthly deliveries by vaccine type, November 2020 to March 2022, from Airfinity.
- Vaccination rates, COVID-19 cases and deaths: daily data from Our World in Data; monthly cumulative vaccination rate uses the highest vaccination rate reported in each month for each country.
- Vaccination uptake: average vaccination coverage from Airfinity (sourced from WHO and UNICEF) for polio, diphtheria tetanus toxoid and pertussis, hepatitis B, measles, and Hib disease.
- Median travel time: Nelson et al (2019) pixel-level travel time to nearest city (population between 0.5 and 1 million); country median of pixel values used as proxy for difficulty reaching remote populations.
- Economic, health, demographic indicators: GDP per capita, infrastructure quality, urban share, health spending, share of 15+ population aged 15-64, share aged 65+ from World Bank WDI; Human development index from UNDP.
- Summary from Table 1 (as of 2022Q1): Vaccination rates and cumulative delivery (per 100 people) are lowest in sub-Saharan Africa (SSA) and highest in advanced economies. SSA has the lowest cumulative cases and deaths (per million), GDP per capita, health expenditure per capita, infrastructure quality score, human development index, and urban population share. SSA share of 15+ population is 61 percent, compared to 84 percent in advanced economies and 75 percent in other EMDEs.

### Findings — Vaccine delivery and vaccination rates
- Worldwide, about 76 percent of delivered vaccine doses have been administered.
- Cumulative delivery and vaccination rates are highly correlated; most country observations are close to the 45-degree line in delivery vs vaccination plots (Figure 3, as of 2022Q1).
- Delivery rates in SSA are markedly lower than in the rest of the world.
- Regression evidence (log vaccination rate on log cumulative delivery, with SSA dummy and interaction) across dates:
  - Log cumulative delivery coefficients: 1.16*** (2021Q2), 1.18*** (2021Q3), 1.13*** (2021Q4), 1.16*** (2022Q1).
  - SSA dummy and interaction with cumulative delivery are statistically insignificant across quarters; interaction is negative in 2021Q2 and 2021Q3 and positive in 2021Q4 and 2022Q1 (all statistically insignificant).
- Interpretation: Despite a slow start, SSA countries administered delivered vaccines at a similar—if not faster—speed compared to the rest of the world since the second half of 2021.

### Findings — Role of vaccine type and source
- mRNA-based vaccines tend to give a strong boost to vaccination rates; other vaccine types either reduce vaccination rates or have no statistically significant impact (Annex II).
- Vaccine source matters:
  - In SSA countries, only vaccine deliveries from COVAX have positive and statistically significant impact on vaccination rates.
  - In the rest of the world, direct purchases from manufacturers and, to a lesser extent, bilateral donations are associated with higher vaccination rates (Annex II).
- Implication: Low delivery rates in SSA reflect a combination of limited funds and insufficient donations.

### Factors affecting vaccination rates (multivariate evidence)
- Sequential regressions (as of 2022Q1) show individual factors positively associated with vaccination rates: higher COVID-19 cases, higher COVID-19 deaths, higher GDP per capita, better infrastructure quality, higher health spending, older population, and higher human development index.
- Example elasticity interpretation: Column 1 implies when cumulative delivery increases by 10 percent, vaccination rates increase by 12.8 percent.
- When all factors are included together (single regression, column 9), only cumulative delivery and quality of infrastructure remain statistically significant determinants of vaccination rates.
- The last-row calculation: if each individual factor increases from the 25th percentile to the 75th percentile, vaccination rates would increase by about 100 percent.
- Policy implication: Vaccine logistics—international delivery and domestic distribution infrastructure—are key to getting people vaccinated.

### Administering rates (administered doses / delivered doses)
- Historical vaccination uptake correlates with higher average administering rates, but there is substantial heterogeneity; some countries with low historical uptake achieved high COVID-19 administering rates.
- No correlation between marginal administering rates (Δa_i / Δd_i over 2022Q1) and vaccination rates as of 2021Q4.
- Regression evidence (administering rate as dependent variable, as of 2022Q1):
  - Columns 1-8: higher historical vaccination uptake, higher COVID-19 cases and deaths, higher GDP per capita, better infrastructure quality, higher health spending per capita, older population, and higher human development index are associated with higher administering rates.
  - If each individual factor increases from 25th percentile to 75th percentile, administering rate increases by about 20 percent.
  - In single regression controlling for all factors, coefficients become imprecisely estimated; contrast with Table 3 results where vaccine logistics are significant determinants of vaccination rates.
- Robustness: Results qualitatively unchanged when administering rates measured at 2021Q3 and 2021Q4.

### Excess doses of vaccines
- Potential measure of unused vaccines: cumulative delivered doses minus cumulative administered doses, d_i (1 − a_i), but this may overstate excess due to reporting lags and timing mismatches.
- United States example: excess doses rising steadily since April 2021, reaching almost 200 per 100 people in April 2022 (Panel (a), Figure 5).
- Across countries (Panel (b), Figure 5): advanced economies (e.g., United States, United Kingdom) have accumulated far more excess doses than developing countries.
- Global implication: reallocating unused doses from countries with excess to those with demand can substantially increase recipient countries' vaccination rates at the same administering rate.

### Equitable distribution and actual vaccination rates
- Figure 6 compares actual vaccination rates (as of 2022Q1) with estimated equitable distribution:
  - Equitable distribution of vaccinations is close to—but not identical with—equal distribution.
  - Most countries under the equitable distribution scenario are at 100 doses per 100 people, reflecting global vaccine supply capacity in early 2022.
  - Most SSA countries would have higher vaccination rates under the equitable distribution.

### Evolution of vaccination inequality and decomposition
- Global vaccination inequality (Gini index across countries) has declined over time but remained as high as 0.4 at the end of 2022Q1 (Figure 7).
- Shapley value decomposition (Figure 8) of contributors to global vaccination inequality (based on column 10 of Table 3):
  - Vaccine delivery is the dominant factor, accounting for between 70 and 80 percent of global vaccination inequality during 2021Q2–2022Q1.
  - Other health, economic, and demographic factors (COVID-19 cases, GDP per capita, urban share, share of 15+ population) play a minor role.
  - Residual (capturing unobserved factors such as vaccine hesitancy) increases slightly over 2021, peaking at only 13 percent of global vaccination inequality.
- Conclusion: Vaccine delivery (access to doses) is far more important for global vaccination inequality than vaccine hesitancy or other domestic factors.

*Sources: Airfinity, Our World in Data, Nelson et al. (2019), World Bank World Development Indicators, UNDP, and IMF staff calculations.*

### 5.    Conclusion and policy Implications

### 5.    Conclusion and policy Implications

### Uneven global vaccination rates and drivers
- COVID-19 vaccination rates vary greatly across countries, with low- and middle-income countries having much lower vaccination rates than advanced countries.
- Delivery of vaccine doses accounts for about 80 percent of global vaccination inequality.
- Vaccine hesitancy accounts for at most about 13 percent of global vaccination inequality.
- Delivery of vaccine doses and quality of infrastructure appear to matter the most for vaccination rates, after controlling for factors within countries’ control (health care system, transportation network, demographic structure, and development status).
- Sub-Saharan African countries received vaccine deliveries later than other regions, but their rates of administering vaccines are comparable to those of other countries when deliveries are available.

### Quantitative trends in inequality (summary from alternative measures)
- Global vaccination inequality declined over the course of 2021 and into 2022 (both Gini index and MN-measure).
- Under the MN-measure, the inequitable component driven by vaccine delivery accounted for about 90 percent of the overall global vaccination inequality in the period shown.
- Both measures depict the same landscape: inequality is declining over time but remains high, with vaccine delivery as the dominant driver.

### Vaccine types, sources, and regional implications
- Regression evidence (as of 2022Q1) on vaccine type:
  - Higher share of mRNA vaccines is associated with higher vaccination rates (positive association strengthens when the United States is excluded).
  - Higher shares of AstraZeneca and Chinese vaccines are associated with lower vaccination rates in some specifications; however, none of the vaccine-type coefficients are statistically significant in the full-sample specifications reported.
  - For Sub-Saharan Africa (SSA), the interaction term of the SSA dummy and share of mRNA vaccines is statistically significant, indicating that mRNA vaccines might boost vaccination rates in SSA.
- Source of vaccine deliveries matters:
  - SSA countries predominantly received vaccine doses from COVAX.
  - Advanced economies relied primarily on direct purchases from manufacturers.
  - Non-SSA emerging market and developing economies sourced most doses directly from manufacturers.
- Regression evidence on delivery sources (log vaccination rate, doses per 100 people, as of 2022Q1):
  - Manufacturers delivery: coefficient 0.33*** (t-statistics reported in table).
  - Bilateral donations delivery: coefficient 0.20** (t-statistics reported in table).
  - COVAX delivery: coefficients vary across specifications; interaction SSA × Log COVAX delivery is positive and statistically significant (e.g., 0.82** and 1.30*** in reported specifications), suggesting COVAX deliveries have been particularly important for SSA vaccination rates.
  - SSA dummy is negative and statistically significant in several specifications (e.g., SSA -3.59***, -4.22***, -4.30***), reflecting lower vaccination rates net of delivery and other controls.

### Robustness checks
- Results are robust to:
  - Population-weighted regressions: coefficients on vaccine delivery remain statistically significant and quantitatively similar (e.g., Log cumulative delivery coefficients reported as 1.28***, 1.37***, 1.30*** in weighted specifications).
  - Adjusting for vaccine types and using share of people with at least one dose as dependent variable (adjusted deliveries dividing multi-dose vaccines by 2 except J&J): coefficients on adjusted cumulative delivery remain statistically significant and similar (e.g., Log cumulative delivery coefficients reported as 1.21***, 1.36***, 1.33*** in adjusted specifications).
- Infrastructure quality score remains statistically significant in specifications that include multiple covariates, reinforcing the role of logistics.

### Policy implications and recommendations
- Strengthen global collaboration and systems for faster and better pandemic response to reach global vaccination targets.
- Key areas for improvement highlighted (drawn from Agarwal et al (2022) as cited):
  - Stronger multilateral institutions with high-level political support to coordinate global response to health shocks.
  - Rapid and adequate financing windows for global public goods activated at the onset of pandemics.
  - Diversified regional manufacturing capacity with arrangements to share technology and know-how.
  - Improved regulatory frameworks to allow speedy delivery of existing and novel tools worldwide.
- Countries should invest in strengthening health systems to prevent, detect and respond to future threats.
- The international community should allocate additional resources to strengthen health systems and vaccine logistics—only collective action can improve outcomes for future pandemics.

*IMF Working Paper — Equitable access to vaccines: Myth or reality? — Section 5, Conclusion and policy Implications*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022257-print-pdf.pdf_