Association of government effectiveness, logistics performance, IT systems and income with COVID-19 mortality

Abstract

The COVID-19 pandemic has unprecedentedly shaken the public health system worldwide. It has been one of the greatest humanitarian crises faced by all countries, regardless of their economic prosperity. However, some countries have been able to minimize the deaths caused by the coronavirus even in the face of a large number of cases, while others have failed to control the death rate even in a comparatively small number of cases. This study explores possible causes of this disparity using cross-sectional data from 126 countries associated with demography, governance, income level, the extent of ICT maturity and the geographical divide. The results of this study suggest that while government effectiveness is negatively associated with the COVID-19 death rate, the logistics performance of governments is positively linked to the COVID-19 mortality rate. The ICT maturity proxied through online service delivery did not confirm its association with the COVID-19 mortality rate. This study informs that poverty and the location of countries do not necessarily influence COVID-19 deaths. Hence, it behoves governments to focus on improving government effectiveness and putting in place more effective and efficient mobility systems, healthcare supply chains and digital administration to address the global health crisis posed by the COVID-19 pandemic and mitigate its harsh effects, including mortality.

1. Introduction

Coronavirus disease (COVID-19), an infectious disease caused by the SARS-CoV-2 virus, still plagues the world. Affected by it to varying degrees, countries have struggled to prevent and control it. While some affected individuals have become seriously ill, requiring medical attention [1], most infected people have developed mild to moderate illness and recovered without hospitalization.

According to the data from WHO, over 636,440,663 people have been reported as infected worldwide as of November 25, 2022 and 6,606,624 people have died. The coronavirus death rate has been 1.03%. Three years into the pandemic since the last month of 2019, death rates of the disease have varied across the countries and regions: Yemen has recorded 18.75%, Peru, 5.17%; Mexico, 4.64%; China, 0.314%; USA, 1.1%; Singapore, 0.079% and Bhutan, 0.034%.

Once the virus was identified and the virus-caused health hazards were experienced, countries have been paying a lot of medical attention. Also, medical scientists and social scientists alike have undertaken studies from diverse perspectives both at pharmacological and non-pharmacological levels. Studies conducted at patients-level on the impact of COVID-19 by varied dimensions - such as, sex differences [[2], [3], [4], [5]] underlying diseases [[6], [7], [8], [9]] and physical fitness [[10], [11], [12]] - have come up with varied findings. Studies by these dimensions have contributed to further analysis and understanding of the risk factors and fatalities associated with COVID-19.

However, evidence obtained from pharmacological and patient-level studies alone does not suffice when it comes to the adoption of policies to reduce COVID-19 risks and fatalities. Scholars have, therefore, undertaken beyond-patient level studies to understand COVID-19 fatalities and policy effectiveness based on such dimensions as mobility and lockdown [13,14] unitary and federated political structure [15,16], government types [17,18] availability of hospital beds [19,20], testing of virus [19,20], government effectiveness [[21], [22], [23], [24], [25], [26]], government policies for non-pharmaceutical interventions [27] and social vulnerability and differences [19,28,29].

There have been very few studies that interpret findings by country variations. Fewer still, if any, are cross-sectional studies that analyze the associations between COVID-19 deaths and exploring the state of the use of ICT in the fight against the pandemic. This study aims to fill this gap by taking into account beyond-patient factors - such as demography (age cohorts), governance (government effectiveness and logistic performance proxied through Logistics Performance Index), income (extreme poverty), income level of countries (low-income countries, upper-middle-income countries and high-income countries) and geographic divide (Europe and Central Asia, Middle East and North Africa, East Asia and Pacific, Sub-Saharan Africa, Latin America and Caribbean Countries and South Asia) - in the analysis of COVID-19 deaths. Another concurrent aim is to generate lessons and insights that can be applied to other health emergencies of the proportion of the pandemic.

This paper explores if there are any causal relationships between COVID-19 and the variables and to what extent they explain the correlations.

1.1. Government effectiveness and addressing COVID-19 mortality rate

For the purpose of this study, government effectiveness has been defined, following Kaufmann, Kraay and Mastruzzi [30], as the quality of public and civil services, the delivery of services independent of political pressures, the quality of policy formulation and implementation and the credibility of the government's commitment to such policies. This study, however, uses effectiveness as a generic term, as government effectiveness in terms of responses to public concerns.

Governments across the globe have perceived COVID-19 as an unprecedented public crisis and therefore have deployed agencies at various levels and capacities to tackle it [31]. They have rescued their citizens from abroad with priorities and have paid due attention to more effective management of health resources such as quarantines, isolation centres, ventilators, dedicated COVID-19 hospitals, and policies restricting the export of medicines, which are helpful for COVID-19 control.

Health service infrastructures and human resources meant for normal times may not suffice when it comes to addressing the pandemic-induced health crises. In the UK, for example, a temporary hospital with 4000 beds was established in nine days when the COVID-19 cases increased rapidly [32]. As the virus peaked, Spain tripled its ICU capacity [33] and Spain and France turned to medical students and retired doctors for support [34]. WHO [1] has encouraged governments to introduce new health infrastructures and mechanisms, such as the operation of isolation centres and quarantines. Governments have even imposed mobility restrictions [35,36] on their people to tackle the crisis besides taking actions by perceived urgency and the corresponding national capacities. Since COVID-19-caused mortality depends on various health conditions, it is important to effectively limit its spread, provide well-managed and regulated isolation centres for the infected and ensure due medical care for those with symptoms [35].

Governments and their agencies have, accordingly, mobilized available resources, both public and private [27,37]. They have shortened procedural burdens to streamline resources and develop new technologies (such as testing KITs). Governments have collaborated with private sectors to develop ICT-based tracing and tracking tools, promote innovative medical care and, more importantly, develop and distribute vaccines [27,[37], [38], [39], [40]]. They have taken measures to ensure the compliance of stay-at-home orders and effective contact tracing, opened up channels for state-society interactions and encouraged community volunteering to fight COVID-19 [41,42]. These efforts are argued to strengthen community solidarity, which plays a key role in implementing policy decisions of governments [43].

Studies suggest that there is a close relationship between government effectiveness and COVID-19 death rates. Greater government effectiveness is associated with lower COVID-19 mortality rates [44,45]. Governments that forged cross-sector collaborations have been more effective in large-scale crisis management, such as the COVID-19 pandemic [23]. Policy decisions taken following consultations across administrative levels, policy areas and sectors have become more instrumental and effective in addressing COVID-19 [46]. In addition, adaptive governance has proved to be a valuable policy option vis-à-vis COVID-19 [47], as adaptive public officials, health scientists and institutions are trusted by people [[48], [49], [50]]. Governance adaptiveness, prompt actions, and regulatory facilitation have become instrumental in breaking the chain of the spread of the virus, controlling the ongoing humanitarian crisis, minimizing the loss of life and preventing further damage to economies.

Similarly, non-effective policies have to do with Covid-19 mortality. Policies that fail to combine the containment interventions turn out to be ineffective so do not explain the success of COVID-19 control [51]. Also, the policies applied untimely to contain the disease are non-effective [52]. Given the contexts, it is hypothesized:

H1

Government effectiveness is negatively associated with the COVID-19 mortality rate.

1.2. Association of COVID caused death with economic inequality and poverty

Some studies, such as by Elgar et al. [53] and Wildman [54], have found a positive association between income inequality and COVID-19 mortality rate meaning that greater economic inequality results in more death rates. Demenech et al. [55] and Boettke & Powell [56] also conclude that a better economic status of countries enables them to better mitigate the effect of COVID-19, including deaths. It is primarily because sound economies have better infrastructure, resources and crisis management skills required to deal with the pandemic. COVID-19 severity and resultant mortality directly depend on individuals’ health conditions or comorbidities. Certain diseases such as diabetes [[57], [58], [59]], high blood pressure [60,61] and heart disease [62,63] increase the risk of COVID-19 mortality. The burden of disease, in general, differs according to the countries’ income levels, with low-income countries being more vulnerable than middle- and high-income countries [52]. This burden applies to the case of COVID-19 mortality as well.

The level of the economy has also had a bearing on a country’s capacity to effectively test, track or trace and then report a true figure of COVID-19 patients and deaths therefore rich countries should support poor countries in this regard [64]. Despite WHO’s urge to report COVID-19 cases and deaths transparently and realistically, some low-income countries are found to have been unable to do so because of the lack of resources, particularly the testing infrastructure [27]. When a patient cannot be tested for a disease, the cause of death cannot be ascertained, as low testing and high positivity rate significantly increase COVID-19 mortality [65].

Hence, it is also hypothesized:

H2

Low income economy countries have higher COVID-19 death rates

The general perception that people from economically sound countries have better immunity is a misconception [66]. They are also vulnerable to COVID-19 and the resultant mortality as they are less health-conscious than people from developed countries. COVID-19 has a greater toll in more developed nations [54,67] and economic variables are not significantly associated with a decrease or increase in deaths by COVID-19 [68].

Single country-level studies have found that poverty has a positive association with COVID-19 death rates [69,70]. Davies [71] suggests that a lower socioeconomic status of a person has a positive association with the underlying medical conditions and comorbidities that trigger deaths. Similarly, the study by Strang et al. [72] of the socio-economically marginalized communities in Stockholm finds that socio-economically weaker communities are at higher risks of COVID-19 mortality than other groups despite them being equal in all other indicators. Poorest population groups in Mexico have also been found to have a lower rate of survival from COVID-19 compared to their affluent neighbours [73]. COVID-19 death rates of individuals living in extreme poverty in the United States were higher than that of those not classified as poor [69,74]. The poor and marginalized have been hit hardest by COVID-19 because of them having no access to the public healthcare system [75,76]. Therefore, it is hypothesized:

H3

Poverty is positively associated with the COVID-19 mortality rate.

1.3. Regional differentiation in COVID-19 mortality

Regional diversity is one of the important measures of understanding cultural similarity at the citizen level. Global regions are believed to share some similarities in the feature of public infrastructure. They also feature similarly in terms of government efficiency in addressing public concerns and in the delivery of public services. Despite such assumptions, studies have found that there are striking differences between the continents and within the continent of Africa itself when it comes to COVID-19 deaths and effects [77]. Such significant variations in COVID-19 deaths among global regions result because of their demographic structures [78] that create spatial vulnerabilities to which COVID-19 patients succumb [79]. Some factors such as quality of air, global interconnectedness, trends in urbanization and health expenditures as well as the policies implemented to mitigate the pandemic influence regionally uneven mortality rate [52]. It is hypothesized:

H4

There is a regional differentiation in the COVID-19 death rates

1.4. Logistics infrastructures and COVID-19

Governments across the globe have adopted mobility restrictions as one of the preventive measures for addressing the COVID-19 pandemic. Mobility restrictions clamped at the community level to break the chain of the pandemic and borrow time to prepare for other measures to fight the coronavirus have varying results. In the early days, mobility restrictions, and lockdown, in particular, were significant in controlling COVID-19 cases [36]. Nevertheless, countries could not prolong such restrictions as they resulted in negative effects on economies [80]. Some countries have, therefore, adopted non-lockdown measures having equal or more degrees of containment effectiveness [28]. The countries that lifted the lockdown measures or gave economic priority in the early days enabled high mobility than the countries with restrictions increased the chance of vulnerable populations getting infected by coronavirus – helping both increasing COVID-19 cases and mortality. Those that adopted lockdowns were required to scale up other measures following relaxation or lifting the lockdowns [81]. Those who failed to do so saw a post-lockdown resurgence of cases [82].

Mobility restrictions and the measures of lockdown imposed by governments for the prevention of COVID-19 resulted in panic buying and increased home consumption, which ultimately impacted transport volume and freight capacity dynamics [83] besides disrupting logistics networks [84]. Governments however have to put extra efforts to manage logistics. Compelled to compromise on multiple fronts, such as financial, public support, international collaboration and cooperation, governments have to work more responsibly to ensure efficient logistics arrangements with more attention to healthcare services. Efficient healthcare logistics supply helps minimize sensitive coronavirus cases and also COVID-19 related deaths. Countries that have been able to adopt smarter strategies and policies in terms of essential supplies have been more efficient and effective to moderate the effect of COVID-19 [36,85]. We, therefore, hypothesize:

H5

Efficient and effective logistics performance is positively associated with the COVID-19 death rate.

1.5. Use of technologies to reduce COVID-19 mortality

Countries that have applied digital technologies in pandemic management and response, especially in terms of planning, surveillance, testing, contact tracing, quarantine and health care, have also been successful in containing COVID-19 [38,86,87]. Leveraging digital technologies has particularly been crucial for protection when the spread rate of COVID-19 has become very high [88]. The UN acknowledges that technologies have made quarantine more livable for millions of people, besides helping the countries to fight the COVID-19. The application of integrated IT system contributed to the investigation of epidemiological issues besides saving resources by automating the overall tracking processes [27].

However, some countries did not effectively utilize the ICTs primarily due to political reasons. Due to lack of political will, central, provincial as well as local governments of some countries faced political and economic sacrifices exposing the country’s insufficient infrastructure including those relating with ICT. The United States is an example in this regard [89]. In some cases, the protocol controlled applications of policies turned out to be inefficient as to addressing COVID-19 pandemic. Technological interventions that require protocols have only marginally contributed to tackling it as the application of technologies in protocol-controlled situations takes time [90].

Studies show that the ‘Internet of Things’ is found to be helpful for COVID-19 patients to identify symptoms [91]. Geographic Information Systems (GIS) and big data technologies have proved to be supportive in many other respects [92]. Likewise, Artificial Intelligence is found to be important to detect the cluster of cases and predict where this virus will concentrate and affect in future [93]. In the same vein, unmanned aerial vehicles (UAVs) have been helpful to variously facilitate a coping strategy during the COVDI-19 outbreak and subsequent lockdowns [94]. Agreeing with Zeng et al. [95] that efficient and intensive use of modern technologies should be at the centre of the COVID containment strategy of governments, it is hypothesized:

H6

The richness of the use of information technology is negatively associated with the COVID-19 mortality rate.

2. Methods

For this study, cross-sectional data from 126 countries was used (see Appendix for the list of the countries). The variables used for the empirical validity of the hypotheses stated above were built from several sources. The effect of the pandemic is not limited to any demographic or national boundary. People of all ages and nations have suffered the same force and intensity. Even the government effectiveness – measured against the Government Effectiveness (GE) score defined by Kaufmann et al. [30], with 2.5 indicating the most effective government and −2.5 the least effective government – has been unable to check the force of the pandemic. In our sample, for example, Yemen has the least effective government (with a GE score of −2.279421568) and Switzerland has the most effective government (with a 1.95212 GE score). However, Switzerland has had higher COVID-19 mortality than Yemen until January 1, 2021 – the date that covers the sample of this study.

Mortality rate is the dependent variable of this study. ‘Mortality rate’ is the measure of the number of deaths caused by COVID-19 per 1,000 population sample study.

In order to test the hypothesis, we used the following model

$$\left(COVID-19MortalityRate\right)={\beta }_0+\beta X+\gamma Z+\epsilon$$

Where ${\beta }_0$ is the constant term, $X$ is the vector of independent variables (government effectiveness, logistic performance, share of the population aged 65 or above, online service maturity, number of hospital beds per 1000 population and extreme poverty) and $Z$ is the vector of control variables (geographic regions and income level of countries). And $\epsilon$ is the error term. The COVID-19 mortality rate, the dependent variable, refers to the deaths per 1,000 cases recorded on January 1, 2021, before the start of the official vaccination against COVID-19 b y any country. By the date, most countries had experienced the first wave of the pandemic and some countries were about to start to administer vaccinations [96]. This study has analyzed the mortality rate differences using the mortality rate data from OurWorldInData (https://ourworldindata.org).

The Logistics Performance Index was used to measure the quality of transportation infrastructure using the data retrieved from https://lpi.worldbank.org. Regional and income dummy variables were taken from the World Bank’s regional and income categories (as of September 1, 2021) from https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending groups. Similarly, the data on the number of hospital beds per 1,000 population was taken from OurWorldInData, whereas the percentage of the population above 65 years of age is used from the World Bank Indicators found at https://data.worldbank.org/https://data.worldbank.org/indicator/SP.POP.65UP.TO.ZS).

Extreme poverty is proxied through the World Development Indicators’ percentage of the population living under USD 1.9 a day (see https://ourworldindata.org/extreme-poverty).

The use of technology is proxied through the Online Service Indicator taken from the e-Government survey (e-Gov survey 2018) available at https://publicadministration.un.org/egovkb/en-us/Data-Center. The variable measures e-commerce development, the level of e-commerce practice of government agencies as well as how government agencies, businesses and citizens interact electronically. The proxy can also be used to measure a particular country’s online activities.

In the regional analysis, the North American region was not included as the data was available from only two countries (Canada and the USA) whose use would present a distorted regional picture because of the limited sample. Similarly, the countries having 0 death rates (as of January 1, 2021) were excluded.

3. Results

To find out beyond-patient factors associated with the death rate of COVID-19 patients across countries, the study explored whether there was a causal relationship between the COVID-19 death rate and government effectiveness. Whether extreme poverty (both in terms of ‘income level’ and ‘individual head count’) and the geographic regions explained the high mortality rate of COVID-19 patients was also investigated. Then the association of transportation facilities (proxied through Logistic Performance Index) with the reduction in COVID-19 death rate was tested. Finally, whether the state of online service maturity would affect the COVID-19 death rate was explored.

3.1. Descriptive statistics

Table 1 summarizes maximum and minimum values and standard deviation across the variables based on 126 observed countries. The range between the maximum (2.794743 per 1,000 population) and the minimum (0.000163 per 1,000 population) rate of COVID-19 mortality varied widely across countries whereas the mean rate was 0.3457. The mean Logistics Performance Index was 2.86469. Similarly, the mean value of online Service Maturity in the observed countries was 0.61469 whereas the mean value of the population earning less than 1.9 USD a day was 12.99415. See Table 1 for more data on the variables.

Table 1.

Descriptive data on predictor variables used in the study.

Variables	Max	Min	Mean	Std Dev	Observation
COVID-19 Mortality Per 1,000	2.794743	0.000163	0.34057	0.44927	126
Overall Logistic Performance Index	4.201444	2.046158	2.86469	0.55047	126
Government Effectiveness	1.95212	−2.27942	−0.04664	1.005838	126
Online Service Maturity	1	0.0764	0.61469	0.261884	126
Extreme Poverty (% of population earning less than 1.9 USD a day)	79.52603	0.001893	12.99415	20.93876	126
Percentage of Population Aged 65 years and Above	28.39727	1.9857	10.0049	6.998773	126
Number of Hospital Bed Per 1000 Population	129.8	1	27.836	23.9999	126
Europe and Central Asia	–	–	–	–	42
The Middle East and North Africa	–	–	–	–	12
East Asia and the Pacific	–	–	–	–	12
Sub-Saharan Africa	–	–	–	–	36
Latin America and Caribbean Countries	–	–	–	–	19
South Asia	–	–	–	–	5
Low Income Countries	–	–	–	–	21
Upper Middle-Income Countries	–	–	–	–	35
Lower Middle-Income Countries	–	–	–	–	35
High-Income Countries	–	–	–	–	35

The correlation between the variables of interest leads to biased results and weakens the viability of the hypotheses and the results. To avoid such a possible situation, a number of statistical tests were performed. As Table 2 suggests, some variables have some degree of correlation, while others have not.

Table 2.

Correlation between the predictor variables.

Variables	COVID-19 Mortality per 1,000	Overall Logistic Performance Index	Government Effectiveness	Online Service Maturity	Extreme Poverty (% of the population earning less than 1.9USD a day)	Percentage of Population aged 65 years and above	Number of Hospital beds per 1,000 population
COVID-19 Mortality Per 1,000	1
Overall Logistics Performance Index	0.439255	1
Government Effectiveness	0.381553	0.845962	1
Online Service Maturity	0.46039	0.761767	0.819413	1
Extreme Poverty (% of population earning less than 1.9 USD a day)	−0.40687	−0.49356	−0.63599	−0.63255	1
Percentage of Population Aged 65 Years and Above	0.539882	0.767738	0.795989	0.71654	−0.59487	1
Number of Hospital Bed Per 1,000 Population	0.299732	0.534818	0.556371	0.529248	−0.4647	0.718881	1

The variance inflation factor (VIF) for Model 1 shows that the model is suspected of multicollinearity (VIF 5.65 for Model 1 and VIF 4.22 for Model 2 see Table 3). The use of data received from different sources may be the reason for the multicollinearity issue. Another possible reason could be the nature of the data because the indicator - government effectiveness - is directly related to other independent variables used for this study. The multicollinearity issue can be removed by applying Cook’s Distance approach, as it removes influential outliers existing in the dataset. However, the persistent outlier in the dataset is a point of interest that reflects the impact of independent predictor variables on each other. Theoretically, a VIF value of 10 or more is considered a serious issue in terms of multicollinearity. However, with the VIF being less than 6, there is no serious multicollinearity among the independent variables engaged in the study. Therefore, the use of Ordinary Least Square (OLS) methods can serve to test the hypotheses of this study.

Table 3.

Result of variance inflation factor test result.

Variables	Model 1		Model 2
	Tolerance	VIF	Tolerance	VIF
Overall Logistic Performance Index	0.2388197	4.187259	0.2753535	3.631695
Government Effectiveness	0.1769824	5.65028
Government Effectiveness (Squared)			0.662589	1.509231
Extreme Poverty (% of population earning less than 1.9USD a day)	0.5154722	1.939969	0.4695136	2.129864
Number of Hospital Bed Per 1,000 Population	0.4779902	2.092093	0.4756002	2.102606
Percentage of Population Aged 65 Years and above	0.2329569	4.292639	0.2364505	4.229214
Online Service Maturity	0.2847318	3.512077	0.2943173	3.397694

All the regressors are assumed to be independent of each other. It is the basic assumption of the OLS model and the residual is normally distributed across observations. The Durbin Watson test (DWT) was performed to confirm that the basic model is free from autocorrelation. In the test, the autocorrelation value is found to be −0.1325491, the D-W test value to be 2.262595 and a p-value to be 0.118 for Model 1. For Model 2, the autocorrelation value is −0.06239553, the D-W test value is 2.122434 and the p-value is 0.492. This suggests that the residuals are free from the autocorrelation issue and, as such, reject the null hypothesis. This reality confirms that the data are independent. Moreover, the p-value from the Breusch-Pagan test (BP 4.6749, p-value 0.5861 for Model 1 and BP 4.0604 and p-value 0.6685 for Model 2) also confirms that the estimation has a consistent variance of error terms, which ensures the absence of heteroscedasticity. These tests further confirm that the OLS regression approach serves well and the result is not biased for a testing hypothesis based on the predictor variables used in this study.

4. Discussion

The result estimation using OLS regression revealed a negative association between government effectiveness and COVID-19 death rate, with the coefficient of government effectiveness in the base models (coeff. −0.237 in Model 1 and -0.088 in Model 2). The coefficients belonging to the government effectiveness are significant in all estimated models (see Table 4). The hypothesis (H1) is therefore accepted. This result leads to the conclusion that effective governments are indispensable to reduce COVID-19 mortality rates. This conclusion is in line with the findings of other researchers, including Goldfinch et al. [48], Lazarus, Binagwaho et al. [49], Liang et al. [44] and Serikbayeva et al. [45].

Table 4.

Regression results.

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Constant	−0.686 (0.329) *	−0.124 (0.241)	−0.993 (0.349) **	−0.518 (0.259) *	−0.704 (0.361).	−0.274 (0.346)
Overall Logistic Performance Index	0.166 (0.121)	0.085 (0.114)	0.253 (0.114) *	0.241 (0.11) *	0.224 (0.117).	0.131 (0.113)
Government Effectiveness	−0.237(0.077) **		−0.198(0.075) **		−0.251(0.081) **
Government Effectiveness (Squared)		−0.088(0.034) *		−0.096(0.032) **		−0.067(0.036) .
Extreme Poverty (% of population earning less than 1.9 USD a day)	−0.004(0.002) .	0 (0.002)	−0.003 (0.003)	0.002 (0.003)	−0.001 (0.003)	0.002 (0.003)
Number of Hospital Bed Per 1000 Population	−0.004(0.002) *	−0.004(0.002) *	−0.001 (0.002)	−0.001 (0.002)	−0.004(0.002) *	−0.004(0.002) *
Percentage of Population Aged 65 years and above	0.043(0.01) ***	0.041(0.01) ***	0.022(0.011) *	0.018(0.01).	0.037(0.011) ***	0.039(0.011) ***
Online Service Maturity	0.452(0.233).	0.029 (0.232)	0.297 (0.233)	−0.159 (0.222)	0.321 (0.228)	−0.043 (0.229)
Europe and Central Asia			0.389(0.183) *	0.346(0.181) .
The Middle East and North Africa			0.231 (0.184)	0.189 (0.182)
East Asia and the Pacific			−0.087 (0.191)	−0.17 (0.188)
Sub-Saharan Africa			0.192 (0.191)	−0.015 (0.186)
Latin America and Caribbean Countries			0.502(0.172) **	0.451(0.17) **
Low-Income Countries					−0.244 (0.219)	−0.054 (0.212)
Upper Middle Income Countries					0.146 (0.134)	0.224(0.133) .
Lower Middle Income Countries					−0.206 (0.173)	−0.065 (0.167)
Adjusted R Squared	0.3447	0.3295	0.488	0.4958	0.4005	0.3707
Number of Observations Included in the study	126	126	126	126	126	126

Note: Signif. Codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ [All bold faces are significant at least 10% level of significance].

Government effectiveness is the state of good governance, a state in which the public administration delivers on its duties and promises and wins over people’s trust. In the context of COVID-19, it specifically means improvement in public health communication and sanitation literacy, robust surveillance and reporting, preparing people for possible waves of the pandemic, strengthening the overall health system, linking social equity in health sectors and ensuring that confinement and de-confinement strategies are comprehensive [49]. The current pandemic demands effective and competent leadership [97] not only to rise to the challenge at hand but also to imagine and prepare for similar challenges in future.

An effective government does not act alone. It creates an enabling environment for multiple actors to engage and mobilize. In large-scale crises such as the COVID-19 pandemic, governments should be able to build cross-sectoral collaboration for effective delivery [23,46].

The analysis of country-level income shows that there is no significant link between poverty and COVID-19 mortality, which is inconsistent with the findings of some previous studies [69,[71], [72], [73]]. Poverty proxied through the percentage of the population earning less than USD 1.9 a day is statistically significant in Model 1 but not significant in other models. The signs of its coefficients are also inconsistent. This means poverty, whether measured through ‘income level’ or individual ‘headcount,’ has no significant role in increasing the COVID-19 mortality rate, suggesting that both hypotheses (H2 and H3) are rejected.

COVID-19 infection per se is not related to socio-economic status. However, an individual’s chances and choices of getting exposed to possible transmitters of the disease may depend on and be linked to economic deprivation. This study, however, did not test how the degree of poverty would interact with death rates over time. It instead suggests that a comprehensive analysis should be carried out of pharmacological (patient-level) data to understand the impact of poverty on the COVID-19 mortality rate, which ultimately would help to understand the relation between poverty-linked living style to COVID-19 severity. Further studies are also necessary to see the interconnections and linkages of healthy habits, balanced diet, obesity, health consciousness and COVID-19 mortality, as socioeconomic status is also linked with the quality of life.

As regards COVID-19 having geographic or regional correlation, coefficients indicate that some regions – for example, ‘Europe and Central Asia’, and ‘Latin America and Caribbean’ – have significantly positive effects while others have insignificant associations, thereby validating the related hypothesis (H4) of the study. Similar findings were observed in earlier studies [98]. Global regions vary by climatic conditions, such as temperature, humidity, rainfall and the presence or absence of sunlight. Studies offer varied - and at times conflicting - findings relating to correlations of these variables with COVID-19 transmission, infection and subsequent deaths. While the study of Chen et al. [99] and Shi et al. [100] concludes that COVID-19 transmission has a negative correlation with air temperature and humidity conditions, the findings by Gupta et al. [101] confirm there is a positive association between the two variables. Our study however does not test how the factors such as altitude, temperature, humidity, rainfall and other climatic situations of the global regions would interact with COVID-19 mortality rates.

The logistics performance index of the country is negatively associated with the COVID-19 death rate, thereby validating the hypothesis (H5) related to this variable. The study findings also concur with that of Anser et al. [102] that an effective and trustworthy healthcare supply chain is of paramount importance to reducing the COVID-19 death toll and case fatality ratio. What this indicates at the very least is the need for an uninterrupted supply of medical instruments and healthcare services and supplies before and during the pandemic.

Findings of numerous studies suggest that effective use of ICT can enable governments to address COVID-19 related issues, such as contact tracing, follow-up, logistic planning and management and others that require quick decisions [38,87,91,103]. As part of e-governance, mature online services enhance government control, involve citizens, contribute to high transparency, minimize corruption and increase convenience [104] by influencing what Ata-Agboni & Olufemi [105] call as the three areas of e-governance: e-administration (for improving government processes), e-services (for connecting citizens with their governments) and e-society (for building interactions with and within civil society).

However, our result does not confirm that there was an association between ICT and mortality rate prior to the administration of COVID-19 vaccination. In our study, the base model (i.e. Model 1) is positively associated with mature online services. However, its association with other models is insignificant. This leads us to conclude that the null hypothesis associated with online services (H6) is invalid. A study by Wei et al. [65] shows that OECD countries, which have relatively better ICT infrastructure and services than that of the non-OECD countries, had higher mortality rates until April 2020 than many non-OECD countries [71]. This reality suggests that the prevalence of better ICTs and related services alone does not suffice to effectively tackle COVID-19 when the countries lack proper knowledge and experience to deal with the pandemic.

The number of hospital beds is also negatively associated with the COVID-19 death rate. Countries having higher numbers of hospital beds per 1,000 population have fewer COVID-19 deaths per case. Our study - in line with previous studies by Sen-Crowe et al. [106] and Xie et al. [107] - confirms that COVID-19 mortality rates are affected by multiple factors, hospital resources and bed capacity being one of them.

This study finds that there is a clear correlation between age and COVID-19 deaths. A country with a higher share of aged populations has been found to have a higher rate of COVID-19 death, as have done previous studies [3,108,109]. That COVID-19 generates a higher mortality burden on old-age people is also reconfirmed.

These confirmations, contrasts and insights add following four lessons to the literature on COVID-19. These lessons can be applied to other COVID-19-like health emergencies as well.

First, protecting people from COVID-19-like pandemics requires immediate action to break the chain of transmission, more so in countries where the health system is not robust [110]. A region why the developed countries saw the disproportionately high rate of mortality in the early days of the pandemic is the delay in imposing lockdowns and other ‘social distancing’ measures to protect the economy from an adverse impact. In times of pandemics, it is not the economy but human lives that should count.

Second, fighting pandemics effectively requires arming the global health community with the knowledge needed to predict the possible outbreak of a pandemic of the force and consequence of COVID-19. Doing so calls for investment in preventive research, which seems to have suffered because of diverging interests of the designers of the global health architecture of the day [111]. The adage that prevention is better than cure is nowhere more relevant than in the fight against pandemics like COVID-19.

Third, the privatized healthcare system – that prizes those who can pay and denies services and insurance to the poor and those with pre-existing conditions [112] – has become obsolete. While this state of health affairs affects all the people around the world, the effect becomes harsher in developed countries where privatization is dominant.

Four, public health authorities should integrate ICT into public health governance and management. Doing so helps them enhance health surveillance systems, map impending risks and take informed decisions about preventive and curative measures that should be in place to tackle the risk. In the case of COVID-19, countries utilized their available ICT applications and technologies during the early stage, however, there were not any proven approaches to tackle with the COVID-19.

5. Conclusions

Our results have suggested that a higher COVID-19 mortality rate is associated with lower government effectiveness hence, improving government effectiveness may contribute to reducing COVID-19 related mortality. Effective governments led by capable leaders willing to forge links with all stakeholders can strengthen the overall health system in general and the one needed to deal with the ongoing COVID-19 pandemic in particular. And hence, capable leaders can be catalytic to reducing COVID-19 mortality. Moreover, ensuring the engagement of multi-stakeholders and receiving their contribution as and when required empowers governments to fight and control health hazards of newer variants of COVID-19 such as Omicron and other strains that possibly could emerge in future.

Since better logistics performance of a country has to do with lesser COVID-19 mortality rate, governments should, therefore, work towards ensuring a more effective and efficient healthcare supply chain during the pandemic. Political authorities and bureaucracies skilled at online and digital administration can activate the mechanisms of good governance through which to manage crises such as the one posed by COVID-19 and mitigate its harsh effects, including mortality. Here comes the role of ICT both in the governance and management of the health care system. However, except in a few countries, ICT has not been effectively applied and leveraged.

Poverty and geographic locations, per se, do not play significant roles to determine the rate of COVID-19 deaths. However, lifestyles and health habits of a particular region may influence the impact of COVID-19, an area up for further research by social and medical scientists.

The global fight against COVID-19 and its future variants can be effective if governments pay more attention to good governance, better logistics performance and online service deliveries simultaneously. To this end, governments should timely revisit and strictly implement related policies in coordination with and support from all stakeholders at national and sub-national functional lines.

The policies that govern public mobility and health infrastructure require prompt review and introspection. In particular, the mass transit systems should be reimagined to make them compatible with COVID-19-like pandemics. Likewise, the cure-focused health infrastructure of the day should be overhauled in order to restore prevention to its position of primacy. Or else, the world will continue to remain vulnerable to the pandemics whose causes and origins remain unknown and contested.

5.1. Limitations and way forward

Given the limitation of the scope of the study, it has not properly explored two questions that are crucial to understanding the force and consequences of COVID-19. First, what prevented the countries with mature ICT infrastructure from using it to tackle the spread of COVID-19, especially during the first wave, which saw a huge loss? Critical interrogation of this question is important because the effective use of ICT is believed to improve public decision-making and its impact. Second, why is improved logistics, such as road infrastructure, positively associated with COVID-19 mortality? An informed understanding of underlying issues related to these questions can be helpful not only in the fight against other waves of COVID-19 but also in possible future pandemics.

APC waived - waiver code: Heliyon - Contractual payment discounts (subject to not being used for the year already).

Author contribution statement

Dipendra Prasad Pant; Mukunda Raj Kattel: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper.

Bikram Acharya: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Data availability statement

Data referenced in article.

Declaration of interest’s statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.

Appendix.

Countries included in the study.

Albania	The central African Republic	Egypt	Haiti	Lesotho	Morocco	Rwanda	Thailand
Algeria	Chad	El Salvador	Honduras	Liberia	Myanmar	Senegal	Togo
Angola	Chile	Estonia	Hungary	Lithuania	Nepal	Serbia	Trinidad and Tobago
Argentina	China	Fiji	Iceland	Luxembourg	Netherlands	Sierra Leone	Tunisia
Armenia	Colombia	Finland	Indonesia	Madagascar	Niger	Slovakia	Turkey
Australia	Comoros	France	Iran	Malawi	Nigeria	Slovenia	Uganda
Austria	Congo	Gabon	Iraq	Malaysia	Norway	Somalia	Ukraine
Bangladesh	Costa Rica	The Gambia	Ireland	Maldives	Pakistan	South Africa	United Kingdom
Belgium	Cote d'Ivoire	Georgia	Israel	Mali	Panama	South Korea	Uruguay
Benin	Croatia	Germany	Italy	Malta	Paraguay	Spain	Uzbekistan
Bosnia and Herzegovina	Cyprus	Ghana	Jamaica	Mauritania	Peru	Sri Lanka	Vietnam
Brazil	Democratic Republic of Congo	Greece	Japan	Mauritius	Philippines	Sudan	Yemen
Bulgaria	Denmark	Guatemala	Jordan	Mexico	Poland	Sweden	Zambia
Burkina Faso	Djibouti	Guinea	Kazakhstan	Moldova	Portugal	Switzerland	Zimbabwe
Burundi	Dominican Republic	Guinea Bissau	Kenya	Mongolia	Romania	Syria
Cameroon	Ecuador	Guyana	Latvia	Montenegro	Russian Federation	Tajikistan