Electrification, Human Capital, and Pandemic Mortality: Evidence from a Global Threshold Analysis

Abstract

The COVID-19 pandemic exposed large cross-country differences in mortality that cannot be explained by short-run policy responses alone. This study investigates how pre-pandemic electrification and human capital jointly shaped pandemic outcomes, emphasizing their potential non-linear complementarity. Using cross-country data and pre-pandemic averages of electricity access and schooling, we examine how long-run development conditions influenced COVID-19 mortality during 2020–2021. We estimate fixed-effects models and a threshold regression framework that allows the interaction between electrification and human capital to vary across infrastructure regimes. The results identify a sharp electrification threshold at approximately 96 percent. Below this threshold, higher levels of schooling are not associated with lower pandemic mortality and may even coincide with increased vulnerability, consistent with binding infrastructure constraints that prevent human capital from being effectively deployed during a health crisis. Above the threshold, the interaction between electrification and schooling becomes statistically insignificant, indicating that in highly electrified economies, the benefits of human capital are already embedded within integrated systems of healthcare delivery, communication, and public health governance. These findings reveal a non-linear complementarity between infrastructure and human capital. Education alone does not enhance pandemic resilience when basic infrastructure remains incomplete, while in near-universally electrified societies its protective role is largely internalized. The results highlight the importance of long-run infrastructure completion as a structural prerequisite for translating human capital into effective pandemic preparedness and resilience.

1. Introduction

1.1. Pandemic Mortality and Structural Conditions

The COVID-19 pandemic acted as a global stress test for development systems, revealing stark cross-country differences in mortality outcomes that cannot be explained by short-run policy responses alone. While a large body of research has emphasized the roles of demographic structure, income, healthcare capacity, and short-run public health policies, these factors alone cannot fully account for the observed heterogeneity in pandemic outcomes. Countries with similar levels of income and broadly comparable policy responses often experienced markedly different mortality trajectories, suggesting that deeper structural conditions played an important role.

Existing cross-country studies converge on the importance of demographics, income, and policy responses, but they diverge in their assessment of the role of education and long-run development conditions. Large cross-country differences in COVID-19 mortality have been widely documented, yet remain only partially explained by contemporaneous policy responses and health-system capacity alone. A growing literature emphasizes the role of pre-existing socioeconomic and structural conditions in shaping pandemic outcomes [1,2]. Recent cross-country evidence further suggests that long-run development characteristics systematically condition vulnerability to large-scale health shocks, beyond short-run containment measures [3,4].

1.2. Human Capital, Infrastructure, and Complementarity

A key limitation of this literature is the implicit assumption that infrastructure and human capital operate additively, rather than conditionally or through threshold effects. Within this literature, increasing attention has been paid to the role of human capital. Education is often viewed as a protective factor through improved health behavior, information processing, and compliance with public health guidance [5,6]. However, recent studies highlight that pandemic shocks have also disrupted human capital formation itself, particularly in developing economies, underscoring the importance of pre-pandemic conditions in shaping crisis resilience [7]. At the same time, epidemiological evidence points to persistent demographic gradients in COVID-19 mortality, reinforcing the need to account for structural vulnerabilities when assessing cross-country outcomes [8].

Importantly, development inputs may interact in non-linear ways during systemic crises. While human capital is frequently included as an additive determinant of pandemic outcomes, its effectiveness may depend critically on complementary infrastructure that enables knowledge, skills, and public health guidance to be translated into action. Foundational infrastructure—most notably electricity access—underpins virtually all modern health-system functions, including hospital operations, cold-chain logistics, digital communication, and epidemiological surveillance. Without reliable electricity, the potential benefits of schooling may fail to materialize, even when educational attainment is relatively high [9].

Conversely, once electricity access approaches universality, human capital may become deeply embedded within integrated systems of healthcare delivery, digital communication, and public administration. In such contexts, the marginal interaction between electrification and schooling may diminish, as the benefits of education are already internalized within institutionalized health and governance frameworks. This logic suggests that the relationship between infrastructure and human capital may be fundamentally non-linear, characterized by threshold effects rather than smooth complementarities.

Despite its relevance, this potential non-linear complementarity has received limited empirical attention in the context of pandemic outcomes. Existing cross-country studies typically include electrification and education as separate controls, implicitly assuming additive effects. Such an approach may obscure regime-dependent dynamics in which the role of human capital changes qualitatively once basic infrastructure constraints are relaxed. As a result, current evidence may underestimate the importance of infrastructure completion as a prerequisite for effective pandemic resilience.

1.3. Contribution of This Study

This study addresses this gap by examining whether the relationship between electrification, human capital, and pandemic mortality is fundamentally non-linear and regime-dependent. Using a global sample of countries during the COVID-19 pandemic years 2020–2021, combined with pre-pandemic averages of electricity access and schooling, we ask three related questions. First, do electrification and human capital exhibit complementarity in shaping pandemic mortality? Second, does this relationship differ systematically between low- and high-electrification contexts? Third, is there a critical electrification threshold beyond which the role of human capital changes?

The theoretical intuition underlying this study is that human capital and basic infrastructure function as complementary inputs in the production of effective pandemic resilience. Human capital—proxied by schooling—enhances individuals’ ability to process information, adjust behavior, and comply with public health guidance. However, the translation of these capabilities into effective health protection critically depends on the availability of foundational infrastructure, particularly reliable electricity. Electricity enables the continuous operation of healthcare facilities, cold-chain logistics for vaccines and medicines, digital communication of public health information, and the feasibility of remote work and learning during outbreaks.

When electricity access remains incomplete, infrastructure constraints bind, and additional schooling may increase economic participation, mobility, and social interaction without corresponding capacity for mitigation, potentially raising exposure to infectious disease. Once electricity access approaches universality, these infrastructure constraints are relaxed, and the benefits of human capital become embedded within integrated health, communication, and governance systems. In this regime, the marginal interaction between electrification and schooling diminishes, as human capital operates through institutionalized channels rather than direct individual responses. This logic implies a non-linear, threshold-type complementarity between infrastructure and human capital, which the empirical analysis is designed to test.

To answer these questions, we estimate fixed-effects regressions and a threshold regression framework that allows the interaction between electrification and schooling to vary across infrastructure regimes. The analysis identifies a sharp electrification threshold at approximately 96 percent. Below this threshold, higher levels of schooling are not associated with lower pandemic mortality and may even coincide with increased vulnerability. Above the threshold, the interaction between electrification and schooling becomes statistically indistinguishable from zero, indicating that in highly electrified societies, the benefits of human capital are already internalized within existing institutional and health-system frameworks.

By documenting a clear infrastructure threshold that conditions the effectiveness of human capital, this study contributes to a more nuanced understanding of pandemic resilience. The findings suggest that pandemic vulnerability is shaped not only by income levels or emergency policy responses, but also by the extent to which foundational infrastructure enables human capital to operate effectively under stress.

1.4. Conceptual Framework: Infrastructure Thresholds and Human Capital Effectiveness

This study is grounded in a simple but often implicit theoretical premise: the effectiveness of human capital in mitigating large-scale health shocks depends critically on the presence of complementary foundational infrastructure. Education enhances individuals’ cognitive skills, information processing capacity, and behavioral adaptability, but these attributes do not operate in isolation. Their translation into effective health-protective action requires infrastructural systems that enable knowledge to be deployed at scale.

We conceptualize electrification as a core enabling infrastructure that conditions the returns to human capital through three interrelated mechanisms.

First, electricity access underpins the functional capacity of healthcare systems. Continuous power supply is essential for hospital operations, oxygen provision, diagnostic equipment, vaccine cold chains, and emergency response logistics. In environments where electricity access remains incomplete, the potential benefits of a more educated population cannot be fully realized because health systems themselves are structurally constrained.

Second, electrification enables the information channel through which human capital operates. Public health guidance, risk communication, and behavioral adaptation during a pandemic increasingly rely on digital media, mobile communication, and real-time information dissemination. Schooling enhances individuals’ ability to understand and act upon such information, but only when electricity access allows these communication systems to function reliably.

Third, electrification shapes the economic and social context in which human capital is deployed. Education tends to increase labor market participation, occupational mobility, and social interaction. In low-electrification settings, these effects may raise exposure to infectious disease without corresponding improvements in protection or mitigation capacity. In contrast, once electricity access approaches universality, the same human capital operates within institutionalized systems—such as remote work arrangements, digitally mediated services, and coordinated public administration—that reduce vulnerability during health crises.

Together, these mechanisms imply a non-linear relationship between infrastructure and human capital. Below a critical electrification threshold, infrastructure constraints remain binding, preventing education from translating into effective pandemic resilience and potentially rendering its marginal effects adverse. Above this threshold, the benefits of human capital become largely internalized within integrated systems of healthcare delivery, communication, and governance, diminishing the scope for additional marginal interaction effects. This framework provides a theoretical rationale for expecting threshold behavior rather than smooth complementarities in the relationship between electrification, human capital, and pandemic mortality.

2. Materials and Methods

2.1. Data and Variables: Pandemic Outcomes and Pre-Pandemic Conditions

2.1.1. Pandemic Outcomes

This study relies exclusively on publicly available cross-country datasets, ensuring transparency and replicability. The primary outcome variable is cumulative COVID-19 mortality, measured as the total number of confirmed COVID-19 deaths per million population. Mortality data are drawn from the Our World in Data COVID-19 dataset, which harmonizes official national statistics across countries. For each country, we construct annual observations for 2020 and 2021 using end-of-year cumulative values. To address the highly skewed distribution of pandemic mortality and reduce the influence of extreme observations, we use the logarithmic transformation log (1 + deaths per million) in all regression analyses.

As a robustness check, we also examine cumulative COVID-19 cases per million population as an alternative outcome measure. This allows us to assess whether the main findings are specific to mortality outcomes or reflect broader patterns of pandemic exposure and transmission. Results using cases per million are reported in Supplementary Table S2.

2.1.2. Pre-Pandemic Electrification and Human Capital

The key explanatory variables capture long-run development conditions measured prior to the pandemic. Electrification is measured as the percentage of the population with access to electricity. This indicator reflects the extent to which basic power infrastructure was available to support healthcare delivery, communication, and economic activity before the COVID-19 shock.

Human capital is proxied by gross secondary school enrollment. While schooling quantity is an imperfect measure of skills, it remains one of the most widely available and comparable indicators of human capital across countries. To ensure that the analysis reflects pre-pandemic conditions and avoids reverse causality, both electrification and schooling are averaged over the years 2018–2019.

2.1.3. Control Variables

The baseline specification includes a set of control variables that capture demographic structure, urbanization, and economic development. These include log GDP per capita (constant prices), the share of the population aged 65 and above, urban population share, and population density. Older populations are known to be more vulnerable to COVID-19 mortality, while urbanization and density may increase exposure through higher contact rates.

We also include a policy stringency index that summarizes the severity of containment measures such as lockdowns and mobility restrictions during the pandemic. This variable is averaged at the country–year level and captures contemporaneous policy responses. While stringency may be endogenous to pandemic severity, its inclusion allows us to assess whether the estimated relationships are robust to controlling for observed policy reactions.

2.1.4. Sample Construction and Descriptive Statistics

The analysis sample consists of 284 country–year observations covering 142 countries for the pandemic years 2020–2021. All pre-pandemic explanatory variables are constructed as averages over 2018–2019. Countries with missing data on key variables are excluded to ensure a consistent estimation sample across specifications.

Descriptive statistics for the main variables are reported in

Table 1. Descriptive Statistics of the Analysis Sample.

Variable	Mean	Std. Dev.
COVID-19 deaths per million	768.98	988.96
Electrification rate (%)	91.47	20.64
Schooling (secondary enrollment, %)	90.84	25.44
Log GDP per capita (pre-pandemic)	9.05	1.33
Population aged 65+ (%)	10.35	6.51
Urban population share (%)	63.09	22.45
Population density (people/km2)	248.97	709.71
Policy stringency index	52.40	11.68
Sample size: observations	284
Sample size: countries	142

Sample: Pandemic years (2020–2021) with pre-pandemic covariates averaged over 2018–2019. Notes: The table reports descriptive statistics for the analysis sample used in the baseline and threshold regressions. COVID-19 deaths are measured as cumulative deaths per million at the end of each year. Pre-pandemic covariates are averaged over 2018–2019.

. The table shows substantial heterogeneity in pandemic mortality, electrification rates, schooling, and demographic structure across countries. Average COVID-19 deaths per million amount to 768.98, with a standard deviation of 988.96, underscoring the highly unequal global impact of the pandemic. Electrification rates are high on average but remain incomplete in a non-trivial subset of countries, motivating the threshold analysis that follows.

2.2. Empirical Strategy: Testing Non-Linear Complementarity

2.2.1. Baseline Complementarity Specification

We begin by estimating a baseline fixed-effects model that allows electrification and human capital to interact linearly. The specification takes the following form:

$$\mathrm{l}\mathrm{n}(1+D_{it})={\beta }_1{E}_i+{\beta }_2{H}_i+{\beta }_3(E_i\times H_i)+X_{it}^{\prime }\gamma +{\mu }_r+{\lambda }_t+{\epsilon }_{it},$$

where $D_{it}$ denotes cumulative COVID-19 deaths per million in country i and year t, $E_i$ is pre-pandemic electrification, $H_i$ is pre-pandemic schooling, and $X_{it}$ is a vector of control variables including log GDP per capita, population aged 65 and above, urban population share, population density, and policy stringency. The terms ${\mu }_r$ and ${\lambda }_t$ denote region and year fixed effects, respectively.

Region fixed effects absorb time-invariant regional characteristics such as climate, broad institutional patterns, and regional reporting practices. Year fixed effects capture global pandemic dynamics common to all countries. Standard errors are clustered at the country level to account for serial correlation across years within countries.

The coefficient on the interaction term, ${\beta }_3$, captures whether the marginal association between schooling and pandemic mortality depends on the level of electrification.

2.2.2. Threshold Regression Framework

The linear interaction model implicitly assumes that the complementarity between electrification and human capital varies smoothly across levels of infrastructure. To allow for the possibility of discrete regime shifts, we extend the analysis using a threshold regression framework.

Specifically, we allow the interaction between electrification and schooling to differ across two regimes defined by a threshold value $\gamma$ of pre-pandemic electrification. For a given candidate threshold, we estimate the following specification:

$$\mathrm{l}\mathrm{n}(1+D_{it})={\beta }_L(E_i\times H_i)\cdot \mathbb{1}(E_i<\gamma )+{\beta }_H(E_i\times H_i)\cdot \mathbb{1}(E_i\ge \gamma )+X_{it}^{\prime }\gamma +{\mu }_r+{\lambda }_t+{\epsilon }_{it}.$$

The indicator functions partition the sample into low- and high-electrification regimes. The coefficients ${\beta }_L$ and ${\beta }_H$ capture the regime-specific interaction effects.

2.2.3. Threshold Selection

The threshold value $\gamma$ is selected using a grid search over all admissible values of electrification, subject to trimming rules that ensure a minimum number of observations in each regime. For each candidate threshold, we estimate the regression and compute the sum of squared residuals (SSR). The optimal threshold is chosen as the value of $\gamma$ that minimizes the SSR.

This approach follows standard practice in the threshold regression literature and allows for sharp, data-driven identification of regime boundaries. Robustness of the estimated threshold to alternative trimming rules is examined in Supplementary Table S1, while the full grid of candidate thresholds and corresponding SSR values is reported in Supplementary Table S3.

3. Results

3.1. Descriptive Patterns

Table 1 reports descriptive statistics for the analysis sample covering the pandemic years 2020–2021 with pre-pandemic covariates averaged over 2018–2019. The table reveals substantial cross-country heterogeneity in both pandemic outcomes and underlying development conditions.

Average COVID-19 mortality amounts to 768.98 deaths per million, with a standard deviation of 988.96, highlighting the highly unequal global distribution of pandemic impacts. Electrification rates are high on average (91.47 percent) but display considerable dispersion, indicating that a non-trivial subset of countries entered the pandemic with incomplete electricity access. Schooling levels, proxied by secondary enrollment, also vary widely, with a mean of 90.84 percent and a standard deviation exceeding 25 percentage points.

Demographic and spatial characteristics further underscore heterogeneity across countries. The average share of the population aged 65 and above is 10.35 percent, while urban population share averages 63.09 percent. Population density exhibits particularly large dispersion, reflecting differences between densely populated urbanized countries and sparsely populated economies. Policy stringency during the pandemic averages 52.40 on the index scale, with notable variation across countries and years.

Taken together, these patterns motivate the empirical focus on non-linear interactions and threshold effects, as the wide dispersion in electrification and schooling suggests that the relationship between human capital and pandemic outcomes may differ systematically across development regimes.

3.2. Baseline Evidence of Complementarity

Table 2. Electrification, Human Capital, and Pandemic Mortality.

Variable	Coefficient	Std. Error
Electrification (pre-pandemic)	0.040	(0.013)
Schooling (pre-pandemic)	0.092	(0.029)
Electrification × Schooling	−0.001	(0.000)
Log GDP per capita (pre)	−0.147	(0.142)
Population aged 65+ (%)	0.134 *	(0.039)
Urban population share (%)	0.020	(0.007)
Population density	−0.000	(0.000)
Policy stringency index	0.054 *	(0.010)
Sample size: observations	256
Sample size: countries	142
R2	0.689
Adjusted R2	0.669
Within R2	0.336
RMSE	1.24
AIC	868.5
BIC	925.2
Region fixed effects	Yes
Year fixed effects	Yes
Standard errors	Clustered by country

Dependent variable: log(1 + COVID-19 deaths per million). Notes: Pre-pandemic covariates are averaged over 2018–2019. All regressions include region and year fixed effects. Robust standard errors clustered at the country level are reported in parentheses. * p < 0.05.

presents the results from the baseline fixed-effects regression that allows electrification and schooling to interact linearly. Several findings stand out. The negative interaction between electrification and schooling is consistent with broader evidence that the returns to human capital are conditional on complementary infrastructure and institutional environments [10].

First, electrification and schooling are each positively associated with pandemic mortality when considered in isolation. The coefficient on electrification is positive and statistically significant, as is the coefficient on schooling. These positive main effects likely reflect the fact that more electrified and more educated societies tend to be wealthier, more urbanized, and more globally connected, characteristics that may increase exposure to infectious disease and improve detection and reporting of COVID-19 deaths.

Second, and more importantly, the interaction term between electrification and schooling is negative and statistically significant. This result indicates that the marginal association between schooling and pandemic mortality weakens as electrification increases. In other words, while schooling is associated with higher mortality at low levels of electrification, this association diminishes—and eventually reverses—as electricity access becomes more complete.

Among the control variables, the share of the population aged 65 and above exhibits a strong positive association with pandemic mortality, consistent with the well-documented age gradient in COVID-19 risk. Urban population share is also positively associated with mortality, reflecting higher contact intensity in urban environments. Policy stringency is positively correlated with mortality, consistent with a reactive policy interpretation whereby stricter containment measures were adopted in response to more severe outbreaks. Log GDP per capita and population density do not display robust independent associations once fixed effects are included.

Overall model fit is high, with an R² of 0.689 and a within R² of 0.336. These results provide initial evidence of complementarity between electrification and human capital but do not yet reveal whether this relationship is linear or characterized by discrete regime shifts.

3.3. Threshold Effects of Electrification

Table 3. Threshold Effects of Electrification on Pandemic Mortality.

Variable	Coefficient	Std. Error
Interaction (High electrification)	−0.000039	(0.000050)
Interaction (Low electrification)	0.000146	(0.000057)
Log GDP per capita (pre)	−0.063	(0.145)
Population aged 65+ (%)	0.145 *	(0.040)
Urban population share (%)	0.016	(0.007)
Population density	−0.000018	(0.000068)
Policy stringency index	0.051 *	(0.009)
Threshold (%)	96.05
N (High)	210
N (Low)	74
SSR	376.133

Dependent variable: log(1 + COVID-19 deaths per million). Notes: The table reports threshold regression estimates allowing the interaction between electrification and human capital to vary across regimes defined by pre-pandemic electricity access. The threshold is selected by minimizing the sum of squared residuals (SSR). All specifications include region and year fixed effects. Robust standard errors clustered at the country level are reported in parentheses. * p < 0.05.

reports the results of the threshold regression that allows the interaction between electrification and schooling to vary across regimes defined by pre-pandemic electricity access. The estimated electrification threshold is 96.05 percent, separating the sample into 210 observations above the threshold and 74 observations below, indicating sufficient variation on both sides of the regime boundary. The existence of a sharp threshold is consistent with theoretical arguments emphasizing minimum infrastructure requirements for the effective functioning of modern health systems and crisis-response mechanisms [6,9].

Table 3 reveals a clear regime-dependent pattern. Below the electrification threshold, the interaction between electrification and schooling is positive and statistically significant at conventional levels. This finding indicates that, in countries where electricity access remained incomplete prior to the pandemic, higher levels of schooling were not associated with lower COVID-19 mortality. Instead, incremental increases in schooling coincided with higher mortality outcomes. Figure 1 visualizes the threshold selection procedure, showing a sharp minimum of the sum of squared residuals at an electrification rate of approximately 96 percent, consistent with the estimates reported in Table 3. This regime-dependent pattern is consistent with the conceptual framework outlined in Section 1.1, which emphasizes that human capital can only translate into effective pandemic mitigation once foundational infrastructure constraints cease to bind.

This figure plots the sum of squared residuals (SSR) from the threshold regression against candidate electrification thresholds. The optimal threshold is selected as the value minimizing the SSR, which occurs at an electrification rate of 96.05 percent. The sharp minimum indicates that the threshold is well identified and not driven by extreme observations.

One interpretation is that in low-electrification environments, schooling raises economic participation, mobility, and social interaction without being supported by sufficiently reliable power infrastructure, healthcare capacity, or digital communication systems. Under such conditions, human capital alone may increase exposure to infectious disease rather than enhance resilience.

Above the threshold, the interaction between electrification and schooling becomes negative but not statistically distinguishable from zero. This suggests that once electricity access approaches universality, the marginal complementarity between electrification and schooling is largely exhausted. In highly electrified economies, the benefits of human capital are already embedded within integrated systems of healthcare delivery, information dissemination, and public administration. As a result, additional interaction effects are no longer detectable in the data.

The coefficients on control variables remain stable across regimes. The share of the population aged 65 and above is strongly and positively associated with pandemic mortality, underscoring demographic vulnerability. Urban population share and policy stringency also retain positive associations. Overall model fit is high, with an R² exceeding 0.70.

Taken together, these results identify a sharp electrification threshold that conditions the effectiveness of human capital during the pandemic. Electrification thus emerges as a structural prerequisite for translating schooling into effective pandemic resilience. To formally assess the statistical relevance of the estimated threshold, we additionally conducted a supLM test for the null hypothesis of no threshold effect against the alternative of a single threshold, following Hansen [11]. The test rejects the null of linearity at conventional significance levels. We further computed bootstrap confidence intervals for the threshold estimate using 1000 replications, which confirm that the estimated threshold at 96.05 percent is tightly identified.

3.4. Robustness Checks

A series of robustness checks confirms the stability of the main findings. Supplementary Table S1 reports threshold estimates obtained under alternative trimming rules applied to the electrification variable. Across trimming levels of 5 percent, 10 percent (baseline), and 15 percent, the estimated threshold remains identical at 96.05 percent, and regime sizes and sum of squared residuals are unchanged. This invariance indicates that the threshold is sharply identified and not sensitive to trimming choices.

Supplementary Table S2 replaces cumulative COVID-19 deaths per million with cumulative cases per million as the outcome variable. The interaction between electrification and schooling remains negative and statistically significant in the baseline specification, and the regime-dependent pattern observed in the threshold analysis persists. This suggests that the main findings are not driven by mortality-specific reporting issues or age structure alone.

Supplementary Table S3 reports the full grid of candidate threshold values and associated sum of squared residuals. The global minimum is attained at an electrification rate of approximately 96 percent, with higher SSR values on either side of the optimum. This pattern reinforces the interpretation that the estimated threshold reflects a genuine structural break rather than sampling noise.

Additional robustness checks, including alternative definitions of pre-pandemic averages and exclusion of policy stringency controls, yield qualitatively similar results. Across all specifications, the central finding remains intact: the relationship between electrification, human capital, and pandemic mortality is non-linear and characterized by a sharp infrastructure threshold.

4. Discussion

This study provides new evidence on the conditional role of infrastructure in shaping the effectiveness of human capital during large-scale public health shocks. By identifying a sharp electrification threshold at approximately 96 percent, the analysis demonstrates that the relationship between human capital and pandemic mortality is fundamentally non-linear and regime-dependent. These findings help explain why countries with seemingly similar levels of education and income experienced markedly different pandemic outcomes, extending recent cross-country evidence on the structural determinants of COVID-19 mortality [1,2,3].

4.1. Infrastructure as a Precondition for Human Capital Effectiveness

A central implication of the results, consistent with the conceptual framework introduced in Section 1.1, is that human capital does not operate independently of basic infrastructure. In countries where electricity access remained incomplete prior to the COVID-19 pandemic, higher levels of schooling were not associated with lower pandemic mortality. On the contrary, the positive interaction between electrification and schooling below the estimated threshold indicates that education may coincide with greater vulnerability in low-infrastructure environments.

This pattern is consistent with the idea that schooling increases economic participation, mobility, and social interaction [5]. In the absence of reliable electricity, however, these channels may amplify exposure to infectious disease rather than enhance protection. Hospitals and clinics may lack continuous power supply, cold-chain systems for vaccines and medicines may be unreliable, and digital communication of public health information may be constrained. Under such conditions, the potential advantages of a more educated population cannot be fully translated into effective health-protective behavior, echoing concerns raised in the development infrastructure literature [9].

In contrast, once electricity access approaches universality, the marginal interaction between electrification and schooling becomes statistically indistinguishable from zero. This does not imply that education ceases to matter for health outcomes. Rather, it suggests that in highly electrified societies, the benefits of human capital are already embedded within institutionalized systems of healthcare delivery, digital infrastructure, epidemiological surveillance, and public administration. Education and infrastructure function as a tightly integrated bundle, leaving limited residual variation for additional interaction effects to capture in cross-country data.

4.2. Non-Linear Complementarity and Pandemic Resilience

The existing literature can be broadly grouped into three strands:

(i)

Studies emphasizing demographic and epidemiological vulnerability;

(ii)

Studies focusing on policy responses and institutional capacity;

(iii)

Studies examining long-run development factors such as income and education.

Our contribution lies in linking the third strand to the first two by explicitly modeling infrastructure as a threshold condition for the effectiveness of human capital.

The threshold result highlights a form of non-linear complementarity between infrastructure and human capital. Below the electrification threshold, infrastructure constraints remain binding and prevent human capital from translating into effective pandemic mitigation. Above the threshold, complementarity is largely internalized, and marginal gains from interaction effects diminish.

This non-linearity helps reconcile seemingly contradictory findings in the existing literature. While some studies report that education is associated with better pandemic outcomes, others find weak or even adverse associations once demographic and policy controls are included. The results here suggest that both patterns may coexist, depending on the underlying infrastructure regime. Education enhances resilience primarily when foundational infrastructure is sufficiently complete to support the effective use of skills, information, and behavioral adaptation [6,7].

From a pandemic-resilience perspective, electrification emerges not merely as a background development indicator but as a structural prerequisite for converting human capital into actionable public health capacity. Reliable electricity supports the continuous operation of hospitals, enables cold-chain logistics for vaccines, facilitates digital dissemination of public health guidance, and allows remote work and learning arrangements that reduce exposure during outbreaks. Without these enabling conditions, the protective potential of human capital remains largely latent.

4.3. Interpretation in the Context of COVID-19 Policy Responses

The positive association between policy stringency and pandemic mortality observed across specifications warrants careful interpretation. Rather than implying that stricter policies caused higher mortality, this pattern is consistent with a reactive policy response: governments tended to adopt more stringent containment measures in response to more severe outbreaks, as documented in cross-country policy analyses of COVID-19 [3].

This interpretation reinforces the importance of focusing on pre-pandemic structural conditions rather than contemporaneous policy choices when assessing cross-country differences in pandemic outcomes. By anchoring electrification and schooling to pre-pandemic averages, the analysis mitigates concerns about reverse causality and highlights the role of long-run development pathways in shaping short-run crisis resilience.

The results also suggest that policy effectiveness itself may be conditioned by infrastructure and human capital. Stringent measures such as lockdowns and mobility restrictions are more likely to be effectively implemented and complied with in environments where electricity access enables remote work, digital communication, and continuity of essential services. Infrastructure and human capital thus shape not only baseline vulnerability but also the capacity of societies to implement and sustain effective policy responses during crises.

4.4. Broader Implications for Development and Health Economics

More broadly, the findings contribute to a growing body of work emphasizing that development inputs are not additive but conditional. Investments in education alone cannot guarantee resilience to systemic shocks if complementary infrastructure remains underdeveloped. At the same time, infrastructure expansion without sufficient human capital may fail to deliver its full social returns.

Pandemic shocks therefore act as stress tests for development systems, exposing latent structural constraints that may remain hidden during normal times. Countries that had achieved near-universal electrification prior to the COVID-19 pandemic were better positioned to leverage existing human capital within integrated health and governance systems, while those that had not faced binding infrastructural constraints that education alone could not overcome. In this sense, COVID-19 underscores the limits of partial development progress and highlights the importance of coordinated development strategies that jointly address infrastructure and human capital in preparation for future global health emergencies and other systemic shocks [4].

The originality of this study does not lie in introducing a new pandemic dataset or a novel econometric technique per se, but in reframing pandemic mortality as an outcome of threshold-dependent development complementarities. By explicitly identifying an infrastructure completion threshold that conditions the effectiveness of human capital, the analysis moves beyond additive explanations and demonstrates why similar levels of education can produce divergent outcomes across countries.

4.5. Policy Implications: Infrastructure Thresholds and Pandemic Preparedness

The findings of this study carry clear and actionable implications for pandemic preparedness and long-run development strategy. From a broader development perspective, the results align with the view that education and infrastructure constitute a bundle of complementary inputs rather than independent drivers of welfare and resilience outcomes [6,10]. Pandemic shocks make these complementarities particularly salient by stress-testing the underlying capacity of societies to mobilize human capital under extreme conditions.

First, electrification should be treated as a prerequisite for the effective utilization of human capital during health crises. The threshold result indicates that education alone does not reduce pandemic mortality when electricity access remains incomplete. Policies that prioritize schooling expansion without simultaneously ensuring reliable and near-universal electricity risk generating limited—or potentially counterproductive—returns in crisis settings. For low- and middle-income countries, accelerating the completion, reliability, and resilience of electricity networks should therefore be viewed as a core component of health security, not merely as a long-term infrastructure objective [6,9].

Second, pandemic preparedness strategies should emphasize coordinated investment rather than sectoral silos. The non-linear complementarity identified in this study suggests that the benefits of education materialize only once infrastructure constraints cease to bind. Governments and development partners should design integrated programs that jointly address electricity access, healthcare delivery systems, and human capital formation. Fragmented investments—such as education reforms implemented in isolation from infrastructure and health-system capacity—are unlikely to deliver resilience against systemic health shocks, a concern echoed in recent assessments of pandemic preparedness in developing economies [7].

Third, policy sequencing matters. The results imply a sequencing logic in which achieving near-universal electrification enables subsequent investments in education, digital health systems, and public health communication to yield their full protective effects. In practice, this means that countries approaching high electrification levels may benefit more from shifting policy emphasis toward improving the quality, relevance, and applicability of education, as well as strengthening digital governance and health information systems. Conversely, countries below this threshold should prioritize closing basic infrastructure gaps before expecting substantial health returns from expanded schooling alone.

Finally, international development assistance and global health frameworks should explicitly incorporate infrastructure thresholds into pandemic risk assessments. Traditional indicators of preparedness tend to emphasize healthcare capacity or emergency policy responsiveness. The evidence presented here suggests that pre-crisis infrastructure conditions—particularly electricity access—play a decisive role in determining whether societies can translate human capital into effective pandemic responses. Incorporating infrastructure completion benchmarks into global preparedness metrics could improve the targeting, sequencing, and effectiveness of resilience-building interventions [3,4].

Taken together, these policy implications highlight that pandemic resilience is not solely a function of short-run emergency response capacity. Rather, it is deeply shaped by long-run development choices, especially the extent to which foundational infrastructure enables human capital to operate effectively under stress.

5. Conclusions

This study examined how pre-pandemic electrification and human capital jointly shaped cross-country differences in COVID-19 mortality. Using pre-pandemic averages of electricity access and schooling and exploiting a threshold regression framework, the analysis demonstrates that the effectiveness of human capital in mitigating pandemic mortality depends critically on the level of underlying infrastructure.

The results reveal a sharp electrification threshold at approximately 96 percent. Below this level, higher schooling does not translate into lower pandemic mortality and may even coincide with greater vulnerability, reflecting binding infrastructure constraints that prevent education from being converted into effective public health action. Above the threshold, the marginal interaction between electrification and schooling becomes statistically indistinguishable from zero, indicating that in highly electrified economies the benefits of human capital are already embedded within integrated systems of healthcare delivery, communication, and governance.

These findings highlight a non-linear complementarity between infrastructure and human capital. Pandemic resilience is not achieved through isolated investments in education or through short-run policy responses alone, but through long-run development pathways that ensure basic infrastructure is sufficiently complete to support the effective use of skills and knowledge during crises.

More broadly, the COVID-19 pandemic serves as a stress test that exposes latent structural constraints in development systems. Countries that had achieved near-universal electrification prior to the crisis were better positioned to leverage existing human capital, while those that had not faced limitations that education alone could not overcome. Recognizing and addressing such infrastructure thresholds is therefore essential for building resilience to future pandemics and other systemic shocks.