Health Aid and Human Well-being: Exploring the Role of Donor Support in Developing Countries (Evidence from Fifty Developing Countries’ Dynamic Panel Data Analysis)

This study aims to assess the impact of disbursed health aid on key health sector variables in 50 developing countries over 19 years (2002-2020). The variables analyzed include infant mortality rate (IMR), under-5 infant mortality rate (IMRu5), and life expectancy at birth (LifeExp). The study utilizes panel data and employs the Generalized Method of Moments (one-step and two-step GMM) for analysis. The findings reveal that health aid has a significant effect in reducing both IMR and IMRu5. A one percent increase in health aid corresponds to approximately 2.189 and 2.134 fewer infant deaths per 1000 live births and 3.497 and 2.864 fewer under-5 infant deaths per 1000 live births under one-step and two-step GMM, respectively. Additionally, a positive and statistically significant relationship exists between health aid and LifeExp. A one percent increase in health aid is associated with an increase of 0.064 and 0.076 years in LifeExp. The study also examines the impact of health aid on gender-specific health indicators. Health aid reduces both male and female IMR and IMRu5, with a more pronounced impact on male rates. Moreover, health aid has a more significant effect on improving female life expectancy than males. Furthermore, the study compares the effectiveness of multilateral and bilateral health aid. Both types of aid significantly reduce IMR and IMRu5, with bilateral aid being more effective for IMR and multilateral aid for IMRu5. Additionally, multilateral aid has a more substantial impact on enhancing life expectancy in developing countries. The main contribution of this study lies in its comprehensive analysis of the overall impact of health aid and its effects based on gender and donor characteristics. These findings emphasize the importance of Sustainable Development Goal 3 in promoting good health and well-being.


Introduction
Foreign aid plays a crucial role in bridging various gaps, including saving-investment, knowledge, and foreign exchange in the developing world (Harrod, 1939;Chenery, 1966;Bacha, 1990).It serves three fundamental objectives: generating prosperity, sharing prosperity, and sustaining prosperity.Thus, it is instrumental in fostering the concept of "the art of living together" in the 21st century (Lekhak, 2023).Another side, previous studies, such as those conducted by Burnside & Dollar (1998), Wolf (2007), Fielding et al. (2006 and2008), Mishra & Newhouse (2009), Bendavid & Bhattacharya (2014), Arndt et al. (2015), Yogo & Mallaye (2015), Ziesemer (2016) and others, have highlighted a positive relationship between health aid and improved health outcomes in developing countries, suggesting that health aid contributes to enhancing the health sector.For instance: the study by Mishra & Newhouse (2009) concluded that an increase in health aid per capita, ranging from US$1.60 to US$3.20 per year, is associated with a decrease in infant mortality by 1.5 per thousand births.Similarly, the research conducted by Bendavid & Bhattacharya (2014) indicated that a one percent rise in health aid corresponds to a 0.24-month increase in life expectancy and a reduction of 0.14 deaths of children under the age of 5 per 1000 live births.In a similar vein, the study carried out by Arndt et al. (2015) arrived at a similar conclusion, suggesting that foreign aid contributes to heightened life expectancy and reduced infant mortality in recipient countries, a finding that aligns with the discovery of Ziesemer (2016).On a slightly different note, Burnside & Dollar (1998) asserted that aid leads to a reduction in infant mortality, especially when recipients implement sound economic policies.
However, despite these efforts and findings, current statistics present a challenging and concerning picture.For instance, in 2021, sub-Saharan Africa's Infant Mortality Rate (IMR) stood at a mere 50 per 1,000 live births, Under-5 Infant Mortality Rate (IMRu5) at 73 per 1,000 live births, and Life Expectancy at Birth (LifeExp) at 60 years.Similarly, in South Asia, IMR was 31, IMRu5 was 37, and LifeExp was 68 years (World Bank, 2023cc/dd/ee).These figures highlight the persisting health challenges in these regions despite the increased aid in the health sector.
Academic studies have revealed that aid directed toward health can have a mostly positive impact on the health outcomes of recipient nations.However, it is imperative to acknowledge that despite the increased aid allocation, health indicators and the current state of the health sector in developing countries present some confusion regarding the efficacy of aid in this domain.This challenge extends to academics, DPs, and recipients.Consequently, foreign aid targeting the health sector has emerged as a highly debated and significant subject of interest among DPs and scholars due to its impact on health outcomes.This topic remains at the forefront of contemporary discussions, encompassing debates over its effectiveness and controversies.
On the other hand, some previous studies have also found that aid to the health sector has shown a negative and sometimes insignificant relationship with health outcomes.For instance, Herzer & Nagel (2015), Mukherjee &

Data and Variables Selection
This study employs dynamic panel data analysis to assess the impact of health aid on various health outcomes.Specifically, it investigates the influence of health aid (including entire, bilateral, and multilateral health aid) on infant mortality rate (including overall, female-specific, and male-specific rates), under-5 mortality rate (including overall, female-specific, and male-specific rates), and life expectancy at birth (including overall, female-specific, and male-specific rates).The analysis utilizes a panel dataset spanning 19 years, from 2002 to 2020, and focuses on fifty low-income and lower-middle-income countries across different regions, namely sub-Saharan Africa (comprising 24 countries), Asia (comprising 16 countries), Latin America and the Caribbean (comprising 4 countries), and the Middle East and North Africa (comprising 6 countries).The selection of these countries is based on three criteria: membership in the Development Assistance Committee (DAC) aid recipient countries, classification as low-income and lower-middle-income countries according to the World Bank, and the availability of relevant data.
Total health aid per capita (Gross Disbursements, Constant Prices US$, 2021, HealthAidP), total multilateral health aid per capita (Gross Disbursements, Constant Prices US$, 2021, HealthAidMulP), total bilateral health aid per capita (Gross Disbursements, Constant Prices US$, 2021, HealthAidBilP) are taken as the main explanatory variables.Aid per capita is utilized as a measure, considering that larger countries require more significant resources than smaller countries to improve health coverage.And this metric indicates the average amount of health aid allocated to each individual in the recipient country for improving their healthcare system.The previous studies by Ziesemer (2016), Mishra & Newhouse (2009), Williamson (2008), and Masud & Yontcheva (2005) have taken per capita health aid in their research.It is important to note that the aid commitment does not guarantee total disbursement.Therefore, this study specifically focuses on analyzing disbursed health aid to avoid potential biases in the findings.The aid variables are transformed into logarithmic terms to normalize the data.Previous research conducted by Yogo & Mallaye (2015), Mukherjee & Kizhakethalckal (2013), Burguet & Soto (2012), Chauvet et al. (2008), Gyimah-Brempong & Asiedu (2008), and Masud & Yontcheva (2005) have also considered disbursed aid in their respective studies.
The study used lagged Infant Mortality Rate, overall/female/male (per 1,000 live births, (IMR/IMRF/IMRM)), lagged Under-5 Infant Mortality Rate, overall/female/male (per 1,000 live births, (IMRu5/IMRFu5/IMRMu5)), and lagged Life Expectancy at birth, total years/female/male (LifeExp/LifeExpF/LifeExpM) as control variables.Similarly, the study has taken Hospital Beds (per 1,000 people) and Physicians (per 1,000 people) as structural characteristics of the health system control variables.The earlier studies by Mukherjee & Kizhakethalackal (2013), Mishra & Newhouse (2009), and Williamson (2008) have taken these variables as a control variable.The literacy rate of adult females (% of females ages 15 and above, LitRFe) and the Primary School Completion Rate (PCR, in %) are considered variables to control for the education level of females and the general awareness of people in society regarding health concerns, respectively.The earlier studies by Mishra & Newhouse (2009) and Mukherjee & Kizhakethalackal (2013) have taken these variables as control variables, respectively.
This study considers population (lnPop) as a control variable for Life expectancy at birth (LifeExp) because higher population levels can pose challenges related to healthcare resources and infrastructure, malnutrition, food security, pressure on natural resources, and limited economic opportunities.These factors collectively significantly impact health outcomes and, consequently, life expectancy.The previous studies by Herzer (2019), Arndt et al. (2015), Wilson (2011), Mishra &Newhouse (2009), andBoone (1996) have taken population as a control variable.
Previous studies conducted by Arndt et al. (2015), Han & Koenig-Archibugi (2015), Bendavid & Bhattacharya (2014), Sweeney et al. (2014), Mukherjee & Kizhakethalackal (2013), Wilson (2011), Williamson (2008), Masud & Yontcheva (2005), and Boone (1996) have extensively examined the economic and governance factors, incorporating them as control variables to address the economic and governance dimensions of selected countries.This study also considers good governance as a highly influential control variable for aid effectiveness, so it decided to use two of the World Bank's Worldwide Governance Indicators: Government Effectiveness (GE) and Control of Corruption (CC), to capture governance issues adequately (WGI, 2023).Additionally, to account for economic factors, the study has included GDP per capita (lnGDPCap, constant 2015 US$) as a logarithmic term for data normalization, along with Government's Current Health Expenditure (Hexp%GDP).
The data for this study were obtained from multiple sources.The World Development Indicators (WDI) of the World Bank (World Bank, WDI, 2022), the Worldwide Governance Indicators from the World Bank, and the Creditor Reporting System (CRS) of the Organization for Economic Co-operation and Development/Development Assistance Committee (OECD/DAC) were utilized.Appendix A provides a summary of the variables used in the study, along with their respective data sources and periods.

Methodology and Estimation Strategy
The issue of endogeneity poses a significant challenge when examining aid effectiveness, and this concern has been acknowledged by eminent scholars such as Burnside & Dollar (2000), Hansen &Trap (2001), andCollier &Dollar (2002).Similarly, previous studies on health aid effectiveness, including Ziesemer (2016), Yogo & Mallaye (2015), Han & Koenig-Archibugi (2015), Sweeney et al. (2014), Afridi & Ventelou (2013), Wilson (2011), Mishra &Newhouse (2009), andGyimah-Brempong &Asiedu (2008), have employed the GMM to address this issue.This study also adopts GMM as a dynamic panel estimator to address endogeneity, omitted variable bias, unobserved panel heterogeneity, and data measurement errors (Roodman, 2009).GMM is particularly suitable for "small T, large N" panel datasets (Roodman, 2009).Within the GMM framework, the study employs the system GMM approach proposed by Arellano & Bover (1995) and Blundell & Bond (1998), which corrects endogeneity by introducing more instruments, transforming the instruments to make them uncorrelated (exogenous) with fixed effects.To ensure the robustness of the findings, the study utilizes the one-step system GMM and two-step system GMM estimators.In evaluating the validity of the GMM results, the study conducts three diagnostic tests, including the Hansen (1982) J test and Sargan (1958) test for over-identifying restrictions, testing for autocorrelation/serial correlation of the error term (with a focus on AR (2)), and ensuring that the number of instruments is less than or equal to the number of groups (i.e., Z ≤ N).
The basic estimation equation is as follows: In the equation, Y it represents the dependent variable at a time t , Y it-1 represents the lagged value of the dependent variable at a time t-1 , X' it represents a set of exogenous variables at a time t , Z' it represents another set of control variables at a time t , d t represents year dummy effects, and ε it represents the error term.The coefficients β 0 , β 1 , β 2 , and β 3 represent the parameters to be estimated in the model.
The detailed final model for each variable is given below based on the above estimation equation.1) For Infant Mortality Rate (Entire/Female-wise/Male-wise/ Multilateral and Bilateral-wise) 2) For under-5 Infant Mortality Rate (Entire/Female-wise/Male-wise/ Multilateral and Bilateral-wise)

Result
The study employed multiple models to assess the robustness of its findings regarding IMR, IMRu5, and LifeExp at birth.Specifically, six models were developed and analyzed for IMR, while four and five models were utilized for IMRu5 and LifeExp, respectively.Among these models, model six served as the primary model for the IMR.In contrast, model four and five were designated as the primary model for IMRu5 and LifeExp at birth, respectively.The study employed the Akaike Information Criterion (AIC) to determine the appropriate lag structure of the models.
Regarding the investigation of the effects of health aid, the study considered a two-year lag at all levels.Health aid encompasses various factors, including health policy and administrative management, medical education/training, medical research, medical services, basic health care, basic health infrastructure, basic nutrition, infectious disease control, population policies/programs, reproductive health, and more (OECD/CRS, 2022).Consequently, it requires a significant amount of time for the impact of health aid on health outcomes to become evident.
Similarly, the lagged corruption of control (L.CC) variable was included in the analysis due to its substantial influence on the current year's results, as the corruption scenario of the previous year has a significant impact.The study also incorporated lagged government effectiveness (L.GE) since government policies and strategies typically require at least one year to yield tangible outcomes.Furthermore, lagged current health expenditure as a percentage of GDP (L.Hexp%GDP) was utilized to examine its effect on country-level health outcomes.This variable requires a minimum of one year to manifest its impact.Finally, lagged GDP per capita (L.lnGDPCap) was included in the analysis to examine its effect on a country level.It typically takes at least one year to observe the manifestation of this impact.The descriptive statistics of variables are presented in Appendix B. The results of both estimations, as presented in Table 1, consistently demonstrate a negative and significant effect of health aid (lnHealthAidP) on IMR across all models.This finding suggests that health aid contributes to a reduction in IMR.Specifically, a one percent increase in health aid is associated with approximately 2.189 and 2.134 decreases in infant deaths per 1000 live births in the final model, as indicated by the one-step and two-step GMM methods, respectively.These associations are statistically significant at the one and five percent levels.These findings align with and support previous literature, including the works of Herzer (2019), Kotsadam et al. (2018), Wilson (2011), Gomanee et al. (2005aGomanee et al. ( , 2005b)), Gupta et al. (2002), Arndt et al. (2015), Mishra & Newhouse (2009), Wolf (2007), Bhaumik (2005), Burnside & Dollar (1998).However, Boone (1996) also concluded that only specific political regimes aid contributes to lower infant mortality rates.
The lagged Infant Mortality Rate (L.IMR) is statistically significant in both estimation methods.Furthermore, the Female Literacy Rate (LitRFe) exhibits a negative association with IMR in both methods, with statistical significance observed in the one-step GMM.This implies that higher levels of female literacy reduce IMR in developing countries, corroborating existing literature such as the works of Mishra & Newhouse (2009).Additionally, according to the findings of Masud & Yontcheva (2005), a 1 percent decrease in female illiteracy is associated with a 0.52 percent decrease in infant mortality.
The availability of Hospital beds (HosBeds) and the number of Physicians (Physicians) show negative associations with IMR across all models in both estimation methods.However, statistical significance is observed only in model three for both variables in both methods.These results indicate that health structural characteristics are crucial in significantly reducing IMR.These findings align with and support existing literature, such as the works of Mukherjee & Kizhakethalackal (2013) and Mishra & Newhouse (2009).
The lagged Health Expenditure as a percentage of GDP (L.Hexp%GDP) demonstrates a negative association with IMR in both estimation methods.It is statistically significant at the five and ten percent levels in the one-step and two-step GMM methods, respectively.These findings highlight the importance of government health expenditure in significantly reducing IMR.This aligns with and reinforces existing literature, such as the works of Wolf (2007), World Bank & IMF (2005), Gomanee et al. (2005b), Rajkumar &Swaroop (2002), andGupta et al. (1999).
Similarly, lagged Government Effectiveness (L.GE) is negatively associated with IMR, although statistical significance is observed only in the two-step GMM at the ten percent level.This underscores the crucial role of sound government policies, strategies, and effective civil services in reducing IMR.These findings align with and support existing literature, such as the works of Mishra & Newhouse (2009), because they concluded that health aid demonstrates higher effectiveness in reducing infant mortality rates in countries with higher institutional quality.
Likewise, the positive association of lagged Control of Corruption (L.CC) with IMR suggests that a nation's malgovernance scenario creates an unfavorable environment for health outcomes, finding alignment with existing literature such as the work of Wolf (2007).However, the association is statistically significant in model five of one-step GMM.
The AR (2) and Hansen Statistic results indicate no second-order serial correction and no issues with over-identifying restrictions, respectively.Additionally, the number of instruments used is lower than the number of groups.Based on this comprehensive analysis, both estimations consistently indicate that health aid significantly reduces the IMR in developing countries.*** p<0.01, ** p<0.05, * p<0.1 (Source: Author's own computation using system GMM).

Comparison Result of the Effect of Health Aid on Female and Male Infant Mortality Rate
Across both estimation methods (Table 2-3), the impact of health aid (lnHealthAidP) on the female and male Infant Mortality Rate (IMRF, IMRM) is consistently negative and statistically significant in both categories.This indicates that health aid significantly reduces IMR among females and males in developing countries.Interestingly, the analysis reveals that health aid is more effective in reducing male infant mortality rates.In the male category, health aid demonstrates statistical significance in all models for both estimation methods.However, in the female category, it is statistically significant only in the one-step GMM.
Under the one-step GMM method, a one percent increase in health aid is linked to a significant reduction of 1.793 per 1000 live births in the male IMR in the final model.Simultaneously, it decreases by 2.103 per 1000 live births in the female category.In contrast, the two-step GMM method shows a decrease of 2.204 per 1000 live births in the male IMR for every one percent increase in health aid in the final model.For the female category, the reduction is 1.362 per 1000 live births, but it lacks statistical significance.
Moreover, the lagged female and male Infant Mortality Rates (L.IMRF and L.IMRM) exhibit statistical significance in both categories.The Female Literacy Rate (LitRFe) shows a negative association with female and male IMR in both estimation methods, indicating that higher levels of female literacy contribute to reducing infant mortality rates.The lagged health expenditure as a percentage of GDP (L.Hexp%GDP) exhibits a negative relationship with female and male IMR in both estimation methods.It achieves statistical significance at the five and ten percent levels in the two-step GMM for the female and male categories, respectively.These findings suggest that health aid, along with factors such as hospital beds, physicians, female literacy, and health expenditure as a percentage of GDP, is crucial in reducing both female and male infant mortality rates; however, males benefit more than females.

Comparison Result of the Effect of Multilateral and Bilateral Health Aid on Infant Mortality Rate
In both estimation methods (Tables 4-5), the impact of multilateral and bilateral health aid (HealthAidMulP, HealthAidBilP) on the IMR consistently shows a negative and statistically significant relationship.This indicates that multilateral and bilateral health aid significantly reduce the IMR in developing countries.However, the analysis reveals that bilateral health aid is more effective, as the coefficients associated with bilateral aid are higher than those for multilateral aid in both estimation methods.
Under the one-step GMM method, a one percent increase in bilateral health aid is associated with a substantial reduction of 3.738 per 1000 live births in the IMR in the final model.In contrast, a one percent increase in multilateral health aid reduces 2.531 per 1000 live births.Similarly, under the two-step GMM method, a one percent increase in bilateral health aid is associated with a 2.319 per 1000 live births reduction in the IMR in the final model.The deduction is slightly lower at 2.072 per 1000 live births for the multilateral category.These findings highlight the more substantial impact of bilateral health aid on reducing the IMR.This contradicts the study of Masud & Yontcheva (2005) because their research concluded that NGO aid significantly reduces IMR.In contrast, bilateral aid does not have any significant impact on IMR.Similarly, the result of this study contrasts with the finding of Mukherjee & Kizhakethalackal (2013) because their research concluded that multilateral health aid has no statistically significant impact on IMR.One of the main reasons behind this differing result is a methodological problem.Mukherjee & Kizhakethalackal (2013) used semiparametric regression with one period lag of health aid instead of the current period health aid to subdue the endogeneity effect, which might fail to address the endogeneity issue successfully.
The lagged Infant Mortality Rate (L.IMR) exhibits statistical significance in multilateral and bilateral types.Furthermore, the lagged health expenditure as a percentage of GDP (L.Hexp%GDP) demonstrates a negative association with the IMR in multilateral and bilateral categories.It achieves statistical significance at the ten percent level in the two-step GMM for both types.
Overall, the analyses demonstrate that the efforts of development partners, both multilateral and bilateral, in reducing the IMR are heading in the right direction and significantly contribute to the reduction of IMR in developing countries.These findings slightly differ from existing literature, such as the works of Masud & Yontcheva (2005) and Mukherjee & Kizhakethalackal (2013).The results of both estimations, as presented in Table 6, consistently indicate a significant and negative impact of health aid (lnHealthAidP) on the IMRu5 across all models.These findings suggest that health aid plays a crucial role in reducing IMRu5.Based on the one-step and two-step GMM methods, a one percent increase in health aid corresponds to a decrease of 3.497 and 2.864 per 1000 live births in IMRu5 in the final model, respectively.These associations are statistically significant at the ten percent level.These results support previous literature, including the studies conducted by Herzer (2019), Han & Koenig-Archibugi (2015), Bendavid & Bhattacharya (2014), Wilson (2011), Fielding et al. (2008), Wolf (2007), Bhaumik (2005), and Gupta et al. (2002).
The lagged under-5 infant mortality rate (L.IMRu5) demonstrates statistical significance in both methods.The Female Literacy Rate (LitRFe) also shows a negative relationship with IMRu5 in both methods, with statistical significance observed in the two-step GMM.This implies that higher levels of female literacy contribute to a reduction in IMRu5 in developing countries.
In all models and estimation methods, the lagged Health Expenditure as a percentage of GDP (L.Hexp%GDP) shows a negative association with IMRu5.It achieves statistical significance at the one and five percent levels in the one-step and two-step GMM methods, respectively.These findings emphasize the importance of government health expenditure in significantly reducing IMRu5, consistent with existing literature such as Wolf (2007), World Bank & IMF (2005), Rajkumar & Swaroop (2002), and Gupta et al. (1999).
The AR (2) and Hansen Statistic results show no second-order serial correction and no concerns with over-identifying restrictions, respectively.Additionally, the number of instruments used is lower than the number of groups.Based on this comprehensive analysis, both estimations consistently indicate that health aid significantly reduces IMRu5 in developing countries.

Comparision Result of the Effect of Health Aid on Under-5 Female and Male Infant Mortality Rate
In both estimation methods (Tables 7-8), the impact of health aid (lnHealthAidP) on the under-5 Infant Mortality Rate for females (IMRFu5) and males (IMRMu5) consistently shows a negative and statistically significant relationship in both categories.These results indicate that health aid significantly reduces under-5 infant mortality rates among females and males in developing countries.Interestingly, the analysis reveals that health aid is particularly more effective in reducing male under-5 infant mortality rates.In the male category, health aid demonstrates statistical significance in almost all models for both estimation methods.However, in the female category, it is not statistically significant in the final model of the one-step GMM estimation.
In the one-step GMM method, a one percent increase in health aid is associated with a significant reduction of 3.800 per 1000 live births in the male under-5 Infant Mortality Rate (IMRMu5) in the final model.Simultaneously, it reduces 3.148 per 1000 live births in the female category, although this result does not reach statistical significance.On the other hand, the two-step GMM method reveals a decrease of 3.235 per 1000 live births in the male under-5 Infant Mortality Rate (IMRMu5) for every one percent increase in health aid in the final model.In the female category, the reduction is 2.555 per 1000 live births.
Furthermore, the lagged female and male under-5 Infant Mortality Rates (L.IMRFu5 and L.IMRMu5) demonstrate statistical significance in both categories.Additionally, the Female Literacy Rate (LitRFe) shows a statistically significant negative association with both female and male IMRu5 in both estimation methods, indicating that higher levels of female literacy contribute to reducing under-5 infant mortality rates.Moreover, the lagged health expenditure as a percentage of GDP (L.Hexp%GDP) reveals a negative relationship between female and male IMRu5 in both estimation methods.It achieves statistical significance at the five percent level in the two-step GMM for the female and male categories.These findings suggest that health aid, along with factors such as female literacy and health expenditure as a percentage of GDP, is crucial in reducing female and male under-5 infant mortality rates.9-10) reveal that multilateral and bilateral health aid (HealthAidMulP, HealthAidBilP) has a negative and statistically significant effect on the Under-5 Infant Mortality Rate (IMRu5).This implies that both multilateral and bilateral health aid plays a significant role in reducing IMRu5 in developing countries.However, when using the two-step GMM method, the effect of bilateral health aid on IMRu5 is not statistically significant.This suggests that multilateral health aid is more effective, consistently showing statistically significant results in all models and estimation methods.
Using the one-step GMM method, a one percent increase in multilateral health aid (HealthAidMulP) is associated with a 0.445 per 1000 live births reduction in the IMRu5 in the final model.Similarly, bilateral health aid (HealthAidBilP) is related to a decrease of 0.477 per 1000 live births.Under the two-step GMM method, a one percent increase in multilateral health aid (HealthAidMulP) is associated with a 0.220 per 1000 live births reduction in the IMRu5 in the final model.In the bilateral category, it is reduced by 0.145 per 1000 live births, but this reduction is not statistically significant.The lagged Infant Mortality Rates (L.IMRu5) also show statistical significance in multilateral and bilateral categories.
Similarly, in both categories, the lagged health expenditure as a percentage of GDP (L.Hexp%GDP) is negatively correlated with IMRu5 and statistically significant at the one percent level in both one-step and two-step GMM.These analyses demonstrate that the efforts of development partners (multilateral and bilateral) in reducing IMRu5 are moving in the right direction and significantly contribute to its reduction.The results obtained from both estimation methods (Table 11) provide compelling evidence regarding the impact of second-period lagged health aid (L2.lnHealthAidP) on Life Expectancy at Birth (LifeExp).Across all models and estimation techniques, a consistently positive and statistically significant relationship is observed, indicating that health aid substantially increases LifeExp.Specifically, a one percent increase in health aid (HealthAidP) is associated with a noteworthy increase of 0.064 and 0.076 years in the LifeExp, as observed in the final model using the one-step and two-step GMM methods, respectively.These findings reinforce and align with existing literature, such as the seminal works of Herzer (2019), Ziesemer (2016), Arndt et al. (2015), Herzer & Nagel (2015), Bendavid & Bhattacharya (2014), Wilson (2011), Mishra & Newhouse (2009), and Williamson (2008).
Furthermore, the lagged Life Expectancy at Birth (L.LifeExp) exhibits statistical significance in both estimation methods.The population (lnPop) demonstrates a positive association with LifeExp in both estimation techniques.These findings suggest that maintaining an optimal population level enhances LifeExp in developing countries.Once again, these findings are consistent with and further support existing literature, such as the influential studies conducted by Mishra & Newhouse (2009).Both estimation methods show a positive and statistically significant relationship between the lagged GDP per capita (L.lnGDPCap) and LifeExp.This empirical evidence suggests that a country's economic situation is crucial in improving health outcomes.The findings align with the consistent results reported by several reputable studies, including Ziesemer (2016).
The variables of Hospital beds (HosBeds) exhibit a consistent positive association with Life Expectancy (LifeExp) in all models across both estimation methods.However, it is essential to note that the result for hospital beds is not statistically significant in the two-step GMM method.In both estimation methods, the lagged health expenditure as a percentage of GDP (L.Hexp%GDP) positively correlates with LifeExp and is statistically significant at the ten percent level.This indicates that increased government expenditure on healthcare significantly contributes to improving LifeExp.This finding further aligns with and reinforces the existing literature, including the study conducted by Ziesemer (2016).
The Primary Completion Rate (PCR) demonstrates a positive association with Life Expectancy (LifeExp) in both estimation methods, but it is statistically significant only in the two-step GMM method.This finding highlights the significance of literacy and awareness in fostering a healthier societal environment.This aligns with the conclusions drawn by previous studies conducted by Verschoor (2002), which emphasized that public expenditures on social services (education, health, and sanitation) have a positive effect on health care for the poor, further reinforcing the existing body of evidence.
The AR (2) and Hansen Statistic result indicates no second-order serial correction and no problem with over-identifying restrictions, respectively.And the number of instruments is less than the number of groups.Based on this comprehensive analysis, both estimation methods consistently reveal that health aid is crucial in increasing Life Expectancy (LifeExp) in developing countries.Note.Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 (Source: Author's own computation using system GMM).

Comparison Result of the Effect of Health Aid on Female and Male Life Expectancy at Birth
In both estimation methods (Table 12-13), the analysis reveals a positive impact of second-period lagged health aid (L2.lnHealthAidP) on both female and male Life Expectancy at Birth (LifeExpF, LifeExpM) in both categories.However, this impact is statistically significant in both estimation methods only for the female category.These findings indicate that health aid has a more substantial and effective influence on the life expectancy of females.One percent increase in health aid (HealthAidP) is associated with a noteworthy increase of 0.083 and 0.089 years in the LifeExpF, as observed in the final model of both one-step and two-step GMM methods.This highlights the positive impact of health aid on enhancing female life expectancy.Furthermore, the lagged female and male Life Expectancy (L.LifeExp) demonstrate statistical significance in both categories.
The population (lnPop) positively associates both female and male Life Expectancy (LifeExp) in both estimation techniques.However, this association is statistically significant only for the female category.Similarly, both estimation methods reveal a positive and statistically significant relationship between the lagged GDP per capita (L.lnGDPCap) and LifeExp for both the female and male categories.However, in model three of the two-step GMM method, this relationship is statistically significant only for the male category.These findings indicate that GDP per capita significantly influences life expectancy.Furthermore, the Hospital beds (HosBeds) variable consistently positively associates with LifeExp in all models and estimation methods for female and male categories.However, this association is statistically significant only for the female category in the one-step GMM method.
In both estimation methods, the analysis reveals a positive correlation between the lagged health expenditure as a percentage of GDP (L.Hexp%GDP) and Life Expectancy (LifeExp) in both the female and male categories.This indicates that higher health expenditure as a percentage of GDP is associated with increased life expectancy for both genders.The statistical significance of this relationship varies across models and estimation methods.For the female category, the lagged health expenditure is statistically significant in all models of both estimation methods.This reinforces the importance of allocating a higher proportion of GDP to healthcare to improve female life expectancy.On the other hand, for the male category, the statistical significance is observed only in model four of the one-step GMM method.
The Primary Completion Rate (PCR) exhibits a positive association with Life Expectancy (LifeExp) in both estimation methods and both the female and male categories.In the female category, the PCR is statistically significant in both the one-step and two-step GMM methods.This finding underscores the importance of female literacy and education in promoting overall female well-being.The results indicate that a one percent increase in the PCR is associated with a 0.005 and 0.0057 years increase in female LifeExp in the one-step and two-step GMM methods, respectively, at the ten percent level.In the male category, the statistical significance of the PCR is observed only in the two-step GMM method.Specifically, the results indicate that a one percent increase in the PCR is associated with a 0.004 years increase in male LifeExp under the two-step GMM method, significant at the five percent level.
One promising outcome of this analysis is the positive and significant association between health aid and the improvement of female healthcare in the developing world.This finding highlights the crucial role that health aid plays in positively impacting female well-being and health outcomes.Additionally, it is worth noting that although the analysis focuses on female healthcare, the results also demonstrate positive associations between health aid and healthcare outcomes for males.This implies that health aid interventions have broader positive effects on healthcare, benefiting both genders in the developing world.In both estimation methods (Table 14-15), the analysis reveals a positive and statistically significant impact of second-period lagged multilateral and bilateral health aid (L2.lnHealthAidMulP,L2.lnHealthAidBilP) on Life Expectancy at Birth (LifeExp) in developing countries.The estimation result showed that bilateral health aid has almost statistical significance at a ten percent level in both estimation methods.In contrast, multilateral health aid is statistically significant only in the two-step GMM method.In the two-step GMM method, a one percent increase in multilateral health aid (HealthAidMulP) is associated with a substantial rise in 0.099 years in LifeExp, as observed in the final model.Similarly, bilateral health aid (HealthAidBilP) is associated with a significant increase of 0.062 years.The higher coefficient of multilateral health aid compared to bilateral health aid emphasizes the greater effectiveness of multilateral aid in enhancing LifeExp in developing countries.
The lagged Life Expectancy at Birth (L.LifeExp) exhibits statistical significance in both estimation methods.The population (lnPop) demonstrates a positive association with Life Expectancy (LifeExp) in estimation techniques under multilateral and bilateral health aid.However, this association is statistically significant only for the multilateral category.Similarly, in both estimation methods, the analysis reveals a positive correlation between the lagged health expenditure as a percentage of GDP (L.Hexp%GDP) and LifeExp in both the multilateral and bilateral categories.However, this relationship is statistically significant in all models of both estimation methods only for the multilateral category.
The Primary Completion Rate (PCR) demonstrates a positive association with Life Expectancy (LifeExp) in both estimation methods within the multilateral and bilateral categories.The analysis reveals that the PCR is statistically significant in both estimation methods under the bilateral category.In contrast, it is only statistically significant in the one-step GMM method within the multilateral category.
Despite the mixed results, the analysis underscores the positive relationship between multilateral and bilateral health aid and LifeExp in developing countries.The findings indicate that the efforts of development partners, both multilateral and bilateral, are moving in the right direction and significantly contribute to improving overall well-being and increasing life expectancy.*** p<0.01, ** p<0.05, * p<0.1 (Source: Author's own computation using system GMM).
The positive and significant relationship of health aid across analyzed health outcomes suggests that the health aid architecture policy, including health policy and administrative management, medical education/training, medical research, medical services, basic health care, basic health infrastructure, basic nutrition, infectious disease control, population policies/programs, reproductive health, and more (OECD/CRS, 2022), is heading in the right direction.
Furthermore, previous and ongoing approaches taken by development partners, both multilateral and bilateral, to enhance health outcomes have been effective.Bhatia et al., 2019).
Due to these efforts, developing regions have made notable achievements in their health systems.For instance, between 1990 and 2021, sub-Saharan Africa and South Asia experienced substantial decreases in IMR from 107 to 50 and 92 to 31 per 1,000 live births from 1990 to 2021, respectively, similarly IMRu5 from 179 to 73 and 130 to 37 per 1,000 live births.Correspondingly, in the same period, both regions witnessed significant progress in LifeExp from 50 to 60 and 59 to 68 total years, respectively (World Bank, 2023a,b,c).
In contrast, the health aid effect for LifeExp, the impact of health aid on IMR and IMRu5 shows a statistically significant relationship at the initial stage.This distinction can be attributed to the different aid mechanisms employed.From the development partners' (DPs) perspective, for instance, a notable portion of health aid, which focuses on IMR and IMRu5, targets particular interventions such as vaccination campaigns, maternal healthcare, and nutrition programs.Such focused efforts can lead to quicker improvements in IMR and IMRu5 compared to the more complex and multifaceted factors influencing LifeExp.For instance, between 1980 and 2021, sub-Saharan Africa and South Asia experienced substantial increases in measles immunization (% of children ages 12-23 months) from 6% to 68% and 1% to 87% from 1980 to 2021, respectively.Similarly, DPT immunization (% of children ages 12-23 months) increased from 5% to 71% and 6% to 85% (World Bank, 2023d,e).And according to the World Health Organization (WHO), immunization currently prevents 3.5-5 million deaths every year from diseases such as diphtheria, tetanus, pertussis, influenza, and measles (WHO, 2023).Similarly, focusing on the importance of immunization, UNICEF emphasized that "vaccines are the world's safest method to protect children from life-threatening diseases" (UNICEF/Immunization, 2023).
From the recipient's perspective, the possible reason is that over the last three decades, the increase in female literacy rates and primary completion rates in developing countries has contributed to a healthier and more aware society, thereby making health aid more effective in reducing IMR and IMRu5.For instance, between 1990 and 2020, adult female literacy rates (% of females ages 15 and above) in sub-Saharan Africa and South Asia increased significantly from 41% to 61% and 32% to 65%, respectively.Similarly, the primary completion rate (% of the relevant age group) rose from 54% to 71% in sub-Saharan Africa and from 64% to 92% in South Asia (World Bank, 2023f,g).
Regarding LifeExp, the relationship between LifeExp and second-period lagged health aid reveals a significant positive association.Several factors contribute to this finding.Firstly, life expectancy represents a long-term phenomenon, reflecting the cumulative effects of diverse factors over extended periods.Consequently, the impact of health aid on life expectancy may not be immediately evident but instead becomes more apparent in the long run.
Furthermore, the architecture and policies governing health aid programs often encompass a range of diverse strategies.The practical implementation of these strategies, leading to noticeable improvements in healthcare infrastructure, services, and health outcomes, may require considerable time.As a result, there is a lagged effect where the benefits of health aid interventions materialize gradually, contributing to improved life expectancy over time.Health aid programs emphasize long-term interventions, such as establishing sustainable healthcare systems, professional training, and large-scale public health campaigns.These initiatives may necessitate adequate time to reach full operational capacity and showcase their influence on life expectancy outcomes.
Additionally, life expectancy is influenced by a complex interplay of socioeconomic, environmental, and healthcare factors.The intricate nature of these influences may further contribute to the observed lagged effect in the relationship between health aid and life expectancy.
In comparing the impact of health aid on females and males, IMR and IMRu5 indicated that males benefited more than females in both outcomes.This phenomenon is attributed to socio-economic factors.In the developing world, several complex factors contribute to this disparity, with discrimination and gender bias being significant among them.In certain cultures and communities, male children may receive preferential treatment in terms of healthcare resources and opportunities for survival.In contrast, female children may not receive the same level of attention and care as their male counterparts.Earlier studies by Chaudhuri (2012Chaudhuri ( & 2015) ) emphasized the phenomena of 'son preference' and the 'gender gap,' while Fuse & Crenshaw (2006) highlighted it as a 'social structure and female infanticide' phenomenon.Similarly, some previous studies found that gender bias is the leading cause behind it, evident in India (Kishor, 1993;Murthi et al., 1995;Griffiths et al., 2000) and Bangladesh (D 'Souza & Chen, 1980;Muhuri & Preston, 1991;Rahman et al., 1992;Rahman & DaVanzo, 1993).
In contrast, when comparing the impact of health aid on female and male life expectancy, the findings indicate that health aid has a more substantial and compelling influence on the life expectancy of females.Several factors contribute to the lower significance of health aid in males.Lifestyle and behavioral factors are among the leading causes.Men tend to engage in riskier behaviors, such as smoking and excessive alcohol consumption, linked to health issues like heart disease, liver disease, and certain types of cancer.For example, Brønnum & Davidsen's study (2022) found that heavy smoking (15+ cigarettes daily) reduced a man's life expectancy by 8.2 to 11.7 years in high and low-education groups.Occupational hazards may also play a role, as men often work in physically demanding and hazardous occupations, such as mining, construction, and transportation, which increases the risks of accidents, injuries, and exposure to harmful substances.
Similarly, cardiovascular diseases, including heart disease and stroke, are the leading causes of death globally, and men tend to develop these conditions at an earlier age than women.Seifarth et al. (2012) highlighted that sociological and biological perspectives contribute to females living longer, on average, than males.The trend of long-run data further supports these observations.Data from 1960 to 2021 show that the life expectancy of females in high-income countries increased from 71 to 83 years, in middle-income countries from 47 to 73 years, and in low-income countries from 43 to 65 years.In contrast, male life expectancy in high-income countries increased from 66 to 77 years, in middle-income countries from 44 to 68 years, and in low-income countries from 40 to 60 years.These data consistently demonstrate that, across all income categories, females have outlived males over the decades (World Bank/LifeExpM/F, 2023).
Regarding the education side control variables, female literacy has shown a significant and robust relationship in reducing IMR and IMRu5.There are several reasons behind this positive impact.Firstly, it increases access to health information, increasing awareness of prenatal and postnatal care, vaccinations, and proper hygiene practices.Secondly, female literacy enables women to acquire knowledge about family planning and birth spacing, contributing to improved maternal and child health outcomes.Additionally, it fosters awareness about nutrition and childcare practices, promoting healthier growth and development in children.Furthermore, female literacy is crucial in women's empowerment and increased decision-making power within households.It can help reduce child marriage rates and, most importantly, break the cycle of poverty, leading to improved health outcomes for the entire community.These factors collectively contribute to the direct and indirect decrease in IMR and IMRu5.
Similarly, the primary completion rate (PCR) also plays a positive role in reducing IMR and IMRu5.However, when considering its role in increasing life expectancy in developing countries, PCR equips individuals with knowledge, skills, and empowerment.This leads to improved healthcare utilization, better health practices, and increased access to economic opportunities, ultimately contributing to better health outcomes and longer life expectancy for the population.Previous studies by Diallo et al. (2023) Based on the study's findings, it can be concluded that a statistically significant relationship exists between health aid and the different health outcomes (IMR, IMRu5, and LifeExp).The study further indicates that health aid effectively enhances health outcomes in developing countries.However, the finding shows that females are less behind than males talking benefit from health aid due to social-culture phenomenon, especially in IMR and IMRu5.The significant result of education variables (LitRFe and PCR) highlighted the education and social awareness to enhancing the health sector.Furthermore, the robust relationship between health structural characteristics (HosBeds and Physicians) and health outcomes (IMR, IMRu5, and LifeExp) underscores the importance of investing in health-related infrastructures.Similarly, the significant relationship of economic variables (Hexp%GDP and GDPCap) with health outcomes (IMR, IMRu5, and LifeExp) highlights the importance of economic factors in enhancing a nation's overall health structure.And the results of governance indicators underscore the significant role of governance in fostering health outcomes.These findings have important implications for policy measures to improve health outcomes in the developing world and ensure aid effectiveness.
From the recipient's point of view, implementing a robust monitoring and evaluation system for health aid programs is crucial to ensure the effective utilization of resources.Additionally, advocating against gender biases is essential to give equal priority to infants (both female and male) in society.Regarding female literacy, the recipient's government should prioritize improving access to education and offer gender-sensitive incentives and scholarships.Addressing issues of early marriage and child labor, creating safe learning environments, providing adult literacy programs, and utilizing media campaigns can further promote female education.To enhance healthcare access and quality, policymakers must prioritize investments in health infrastructure.Recognizing the importance of economic factors in overall health structure, promoting economic development and growth is vital to increase access to healthcare services and resources.For improving governance, recipient governments should focus on transparency and accountability, maintaining the rule of law, streamlining bureaucracy, promoting citizen engagement, and fostering collaboration.These strategies will lead to better service delivery to citizens.From the development partners' perspective, targeted health aid, enhancing health infrastructures, and fostering collaborative partnerships are critical for improving health outcomes in developing countries.By focusing on these key policy measures, health aid can significantly impact healthcare access, quality, and overall health system development in recipient countries.

Funding
None.

Informed Consent
Obtained.

Provenance and Peer Review
Not commissioned; externally double-blind peer reviewed.Infant mortality rate, male is the number of male infants dying before reaching one year of age per 1,000 male live births in a given year.

WDI, 2002-2020
Mortality rate, under-5 (per 1,000 live births) The under-five mortality rate is the probability per 1,000 that a newborn baby will die before age five.

WDI, 2002-2020
Mortality rate, under-5, female (per 1,000 live births) Under-five mortality rate, female is the probability per 1,000 that a newborn female baby will die before reaching age five.

WDI, 2002-2020
Mortality rate, under-5, male (per 1,000 live births) Under-five mortality rate, male is the probability per 1,000 that a newborn male baby will die before reaching age five.

WDI, 2002-2020
Life expectancy at birth, total (years)/ female(years)/ male (years) Life expectancy at birth is a measure that quantifies the anticipated number of years an infant would live if the prevailing mortality patterns at the time of their birth remained constant over the course of their entire life.Source: Author's own computation.

Figure 1 .
Figure 1.Disbursed Health Aid in Developing Countries Source: Prepared by Author.
3.1.1Health Aid and Infant Mortality Rate 3.1.1.1The Effect of Health Aid on Infant Mortality Rate per capita represents the gross amount of health aid disbursed to the health sector.Per capita is determined by dividing the total gross disbursement of health sector foreign aid provided by Official Donors to individual countries by the total population of the recipient country.The calculation uses constant 2021 US dollars to account for inflation and ensure comparability over time.aid per capita represents the gross amount of multilateral health aid disbursed to the health sector.Per capita is determined by dividing the total gross disbursement of health sector foreign aid provided by multilateral to individual countries by the total population of the recipient country.The calculation uses constant 2021 US dollars to account for inflation and ensure comparability over time.Prices US$, 2021) Total bilateral health aid per capita represents the gross amount of bilateral health aid disbursed to the health sector.Per capita is determined by dividing the total gross disbursement of health sector foreign aid provided by bilateral to individual countries by the total population of the recipient country.The calculation uses constant 2021 US dollars to account for inflation and ensure comparability over time.inpatient beds that are accessible in a wide range of healthcare facilities, such as public and private hospitals, both general and specialized, as well as rehabilitation centersliteracy rate refers to the proportion of females aged 15 and above who can read and write.students successfully completing the last year of (or graduating from) primary school in a given year is divided by the number of children of official graduation age in the population.represents the division of gross domestic product by the population count at a given point in time.This metric provides insight into the magnitude of an economy.WDI, 2002-2020Control of CorruptionControl of corruption captures perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption.GE) captures: the quality of public service, the quality of civil service and how far it is independent of political pressures, the process and quality of policy formulation and implementation, the government's credibility and commitment to such policies.Total population, which counts all residents regardless of legal status or citizenship.The values shown are midyear estimates.

Table 1 .
The Effect of Health Aid on Infant Mortality Rate

Table 3 .
The Effect of Health Aid on Male Infant Mortality Rate Dependent Variable: -Male Infant Mortality Rate (IMRM)

Table 4 .
The Effect of Multilateral Health Aid on Infant Mortality Rate Dependent Variable: -Infant Mortality Rate (IMR)

Table 5
. The Effect of Bilateral Health Aid on Infant Mortality Rate Dependent Variable: -Infant Mortality Rate (IMR)

Table 6 .
The Effect of Health Aid on Under-5 Infant Mortality Rate Dependent Variable: -Under-5 Infant Mortality Rate (IMRu5)

Table 8 .
The Effect of Health Aid on Male under-5 Infant Mortality Rate Dependent Variable: -Male under-5 Infant Mortality Rate (IMRMu5)

Table 9 .
The Effect of Multilateral Health Aid on under-5 Infant Mortality Rate

0057* (0.003) (0.0039)
Comparison Result of the Effect of Multilateral and Bilateral Health Aid on Life Expectancy at Birth

Table 14 .
The Effect of Multilateral Health Aid on Life Expectancy at Birth Dependent Variable: -Life Expectancy at Birth (LifeExp)

Policy Measures, Limitations, and Suggestions for Future Research
Makuta & O'Hare (2015)2007)l (2013) Erhijakpor (2007)have emphasized the significant contribution of female literacy in reducing IMR and IMRu5 in developing countries.The significant and robust relationship between health structural characteristics (HosBeds and Physicians) and health outcomes (IMR, IMRu5, and LifeExp) underscores the importance of investing in health-related infrastructures.Investing in such infrastructures helps enhance healthcare accessibility, improve disease management, and promote preventive care, ultimately contributing to a healthier and more resilient society.These findings follow and bolster the existing literature, exemplified by the research ofMukherjee & Kizhakethalackal (2013),Mishra & Newhouse (2009), andAnyanwu & Erhijakpor (2007).Similarly, the significant and robust relationship of economic variables (Hexp%GDP and GDPCap) with health outcomes (IMR, IMRu5, and LifeExp) highlights the importance of economic factors in enhancing a nation's overall health structure.More specifically, a higher percentage of GDP allocated to government health expenditure can improve healthcare accessibility, preventive care, maternal and child health services, early detection and treatment, research advancements, poverty reduction, and crisis preparedness.All of these factors combined can positively impact reducing IMR and IMRu5 and increasing life expectancy in a population.On the other hand, GDP per capita plays a critical role in shaping the overall socioeconomic conditions, affecting healthcare, nutrition, and public health measures.For example, it helps improve healthcare infrastructure, access to quality healthcare services, better nutrition, food security, investment in education and awareness, and the reduction of poverty and inequality.Addressing these aspects in the health sector can reduce IMR and IMRu5 and enhance life expectancy in a country.The findings of previous studies byRahman et al. (2018),Makuta & O'Hare (2015),Linden & Ray (2017),Farag et al. (2013),Bokhari et al. (2007), andAnyanwu & Erhijakpor (2007)have all arrived at the same conclusion.Linden & Ray (2017) discovered a statistically significant relationship between log GDP per capita and life expectancy over the last four decadesacross 148 countries using OLS and quantile regressions, whileBokhari et al. (2007)emphasized the undeniable contribution of economic growth and government spending on health, particularly for IMRu5 and maternal mortality;Rahman et al. (2018)found a significant correlation between public health expenditure leading to reduced IMR and increased GDP per capita, resulting in higher life expectancy, as didFarag et al. (2013), who demonstrated the noteworthy impact of government health spending in reducing both IMR and IMRu5, whileAnyanwu & Erhijakpor (2007)further supported this notion, revealing the significant role of health expenditure in decreasing IMR and IMRu5 across 47 African countries; andMakuta & O'Hare (2015)underscored the robust influence of public health expenditure in reducing IMRu5 and promoting increased life expectancy.Finally, the findings regarding governance control variables (L.CC and L.GE), which indicate a positive effect of L.CC and a negative relationship of L.GE with health outcomes (IMR, IMRu5, and LifeExp) underscore the significant role of governance in fostering health outcomes in developing countries.Corruption within healthcare systems can have far-reaching consequences for health outcomes, exacerbating health disparities, reducing healthcare service quality and accessibility, and undermining public health efforts to improve community well-being.On the other hand, government effectiveness is crucial in promoting better health outcomes through