A Comparative Analysis of the Determinants of Male, Female, and Total Life Expectancy in United States and China

Abstract

Health is wealth, and ensuring healthy lives and promoting well-being for all are fundamental to achieving Sustainable Development Goal 3 (SDG-3) and fostering sustainable socioeconomic development. Previous studies have primarily examined the economic, social, or institutional determinants of life expectancy (LE) in isolation, with limited attention given to their combined effects within a comprehensive analytical framework for China and United States. To address this gap, the present study investigates the determinants of LE by integrating environmental, economic, social, and institutional factors. Specifically, CO2 emissions (CO2e) represent the environmental dimension; GDP and health expenditure capture the economic dimension; population growth and education reflect the social dimension; the rule of law and government effectiveness represent institutional quality. This study estimates three separate models for male life expectancy (MLE), female life expectancy (FLE), and total life expectancy (TLE) using the ARDL model, with the FMOLS estimator employed to assess the robustness of the long-run results. The findings reveal that CO2e and population growth adversely affect MLE, FLE, and TLE in both China and the United States. In contrast, GDP, health expenditure, education, the rule of law, and government effectiveness significantly improve LE across all three models. From a policy perspective, these countries’ governments should prioritize reducing CO2e through cleaner energy technologies and stricter environmental regulations while continuing to invest in healthcare systems and educational development. Strengthening institutional quality by enhancing the rule of law and improving government effectiveness can further increase the efficiency and accessibility of health services, thereby contributing to longer LE in both China and the United States.

1. Introduction

China and the USA are two of the world’s largest and most influential economies, yet they continue to exhibit notable differences in life expectancy (LE) despite significant global improvements in population health and declining mortality rates. These disparities raise important questions about the factors driving human longevity. In particular, understanding how economic factors (health expenditure and GDP), environmental pressures, population growth, education, and institutional quality, reflected by the rule of law and government effectiveness, interact to influence male, female, and total life expectancy remains an important area of research.

Improving LE is a key objective for countries worldwide, as it reflects the overall health and well-being of their populations. To achieve this goal, governments implement policies aimed at reducing mortality, strengthening healthcare systems, and improving the social, economic, and environmental conditions that contribute to better health outcomes (Girosi & King, 2007; Uddin et al., 2024). LE is one of the most widely used indicators for assessing a country’s overall health status and level of development. LE is the average number of years a newborn is expected to live if current patterns of mortality remain unchanged throughout their lifetime. Beyond measuring longevity, LE reflects the combined influence of healthcare quality, living standards, education, income, nutrition, sanitation, and other socio-economic factors that affect people’s health and survival (Zhang et al., 2025).

Figure 1, shows the trends of life expectancy, total, LE increased in both countries during most of the study period, although the pace and pattern of improvement differed considerably. In the United States, total LE rose from 76.64 years in 2000 to 78.79 years in 2019, reflecting gradual improvements in healthcare, medical technology, and living standards. A sharp decline occurred during 2020–2021, when LE fell to 76.33 years. This reduction was primarily associated with the COVID-19 pandemic, which caused substantial excess mortality. After the pandemic, life expectancy partially recovered, reaching 78.39 years in 2023, although it remained slightly below the pre-pandemic peak. In China, life expectancy followed a more consistent upward trajectory throughout most of the study period. LE increased steadily from 72.29 years in 2000 to 78.20 years in 2022, representing an improvement of nearly 6 years. This sustained increase reflects rapid economic growth, improvements in healthcare services, expanded health insurance coverage, better nutrition, higher educational attainment, and successful public health initiatives. China experienced only a minor decline in 2023, with life expectancy decreasing slightly to 77.95 years.

Figure 1.

Trends of life expectancy at birth, total (years). Source: World development indicator

Figure 2, shows the trend of female LE, female LE increased in both the USA and China over the study period, although the patterns of improvement differed. In the USA, female LE rose from 79.30 years in 2000 to 81.40 years in 2019, reflecting steady improvements in healthcare services, disease prevention, and living standards. A noticeable decline occurred during 2020–2021, when female life expectancy fell to 79.30 years, largely due to the COVID-19 pandemic. It subsequently recovered to 81.10 years in 2023, approaching its pre-pandemic level. In China, female life expectancy showed a more consistent upward trend throughout the study period. It increased from 74.90 years in 2000 to 81.13 years in 2022, representing a gain of more than 6 years. This sustained improvement can be attributed to rapid economic development, expanded healthcare coverage, improved maternal and child health services, higher educational attainment, and better public health policies. A slight decline was observed in 2023, when female life expectancy decreased marginally to 80.93 years. A comparison of the two countries indicates that the United States maintained higher female life expectancy than China throughout most of the study period, although the gap narrowed considerably over time. In 2000, the difference was 4.40 years (79.30 versus 74.90 years), whereas by 2023 it had narrowed to only 0.17 years (81.10 versus 80.93 years). This convergence highlights China’s remarkable progress in improving women’s health and longevity over the past two decades, bringing female life expectancy to a level nearly equal to that of the United States despite beginning from a substantially lower baseline.

Figure 2.

Trends of Life expectancy at birth, female (years). Source: World development indicator

Figure 3 shows male life expectancy, which increased in both the USA and China over the period 2000–2023, although the trends differed across the two countries. In the USA, male life expectancy increased from 74.10 years in 2000 to 76.30 years in 2019, reflecting gradual improvements in healthcare, medical technology, and living conditions. A substantial decline occurred during 2020–2021, when male life expectancy fell to 73.50 years, primarily due to the COVID-19 pandemic. Male life expectancy subsequently recovered to 75.80 years in 2023, remaining slightly below the pre-pandemic level. In China, male life expectancy followed a steady upward trend throughout most of the study period. Male life expectancy increased from 69.85 years in 2000 to 75.49 years in 2021, representing an improvement of more than 5.5 years. This sustained growth reflects China’s rapid economic development, expansion of healthcare services, improvements in nutrition and education, and strengthened public health policies. A slight decline was observed after 2021, with male life expectancy decreasing to 75.20 years in 2023. A comparison of the two countries shows that the USA consistently recorded higher male life expectancy than China, although the difference narrowed considerably over time. In 2000, the gap was 4.25 years (74.10 versus 69.85 years), while by 2023 it had fallen to only 0.60 years (75.80 versus 75.20 years). This convergence indicates that China has made substantial progress in improving male survival and health outcomes, nearly closing the longevity gap with the USA over the past two decades.

Figure 3.

Trends of Life expectancy at birth, Male (years). Source: World development indicator

There are several economic determinants of LE, such as economic growth (EG), inflation, health expenditure (HE), etc. GDP is broadly recognized as one of the most important determinants of LE. Higher levels of economic development are generally associated with longer lifespans, as EG not only improves living conditions but is reinforced by a healthier and more productive population. Improvements in health promote EG through several key channels. A healthier workforce is more productive, greater investment is made in both physical and human capital, and lower mortality and fertility rates contribute to more sustainable long-term economic development (Chireshe & Ocran, 2020; Li et al., 2025). Public health expenditure refers to government spending allocated to the healthcare sector to improve population health and ensure access to essential health services. Higher public investment in healthcare enhances the availability, affordability, and quality of medical services, enabling individuals to obtain the care they need more effectively (Beylik et al., 2022). Public health expenditure covers a broad range of activities aimed at improving health outcomes, including preventive care, health promotion, medical treatment, rehabilitation services, health administration and regulation, community health programs, and capital investments in healthcare infrastructure. These expenditures support both individual and population health by strengthening healthcare systems and expanding the delivery of essential health services (Oniore et al., 2024; Zhang et al., 2026b).

Environmental conditions, particularly carbon emissions and air pollution, are important determinants of LE. A growing body of evidence shows that prolonged exposure to CO₂ emissions and fine particulate matter (PM2.5) is associated with poorer health outcomes and shorter LE. Deteriorating environmental quality increases the cardiovascular diseases and risk of respiratory ultimately decreasing both the quality and length of human life (Azam et al., 2023; Aziz et al., 2025; Das & Debanth, 2023). PM2.5 refers to fine particulate matter with an aerodynamic diameter of 2.5 micrometers or smaller. Because of their extremely small size, these particles can bypass the body’s natural defense mechanisms and travel deep into the respiratory tract, reaching the alveoli. Some particles may then cross into the bloodstream, allowing them to circulate throughout the body and affect various organs. Long-term exposure to elevated concentrations of PM2.5 has been linked with a wide range of adverse health outcomes, including cardiovascular disorders, respiratory diseases and other systemic complications. Consequently, PM2.5 is widely recognized as one of the most harmful air pollutants and a major risk factor for reduced human health and longevity (Li, X., & Liu, 2021; Zhang et al., 2026a). The limited attention given to health implications is particularly concerning, as growing evidence indicates that air pollution, which is closely associated with CO₂ emissions, is a major risk factor for non-communicable diseases. According to the WHO, air pollution is among the primary environmental threats to human health, donating to millions of premature deaths each year through respiratory illnesses, cardiovascular diseases, stroke, cancer, and other chronic health conditions (Goshua et al., 2022; Shahriyari et al., 2022). A substantial share of these emissions originates from inefficient and fossil fuel-dependent energy systems, especially in rapidly industrializing economies. Although existing approaches are useful for identifying long-term environmental trends, existing approaches often fail to capture the environmental efficiency of energy use, specifically how each additional unit of energy consumption contributes to pollution and associated health risks. This limitation is particularly important in fast-growing economies, where even modest increases in energy demand can generate disproportionately large increases in emissions and adverse health outcomes (Shaari et al., 2026).

Several social factors influence LE, including education, population dynamics, and birth rates. Among these, education is widely regarded as one of the most important determinants of health and longevity. A growing body of literature has shown that higher educational attainment is strongly associated with lower mortality rates, improved health outcomes, reduced disease prevalence, and longer LE. Education enhances individuals’ ability to access and effectively use healthcare services, adopt healthier lifestyles, and make informed decisions about disease prevention and health management (Iyakaremye & Tripathi, 2024). Education contributes to longer LE through both direct and indirect pathways by improving employment opportunities, income, and health awareness (Rogers et al., 1999). Individuals with lower levels of education are generally more likely to experience obesity and other health risks, whereas those with higher educational attainment tend to maintain healthier diets, engage in regular physical activity, and adopt behaviors that promote long-term health and well-being (Brunello et al., 2016). The relationship between population growth and LE has been widely debated across countries and regions. Existing evidence suggests that the impact of population growth on LE is not uniform but depends on each country’s demographic, economic, and social conditions (Ademoh, 2017). Consequently, there is no clear consensus in the literature regarding whether population growth has a positive or negative effect on LE, as its influence varies across different national contexts (Hasnawati et al., 2024; Li et al., 2025).

Institutional quality (IQ) is a key driver of sustainable development and plays an essential role in achieving the SDGs. In particular, it is closely linked to SDG 16, which emphasizes the importance of encouraging peaceful and inclusive societies, ensuring access to justice, and developing effective, accountable, and transparent institutions. Strong institutional frameworks that uphold the rule of law, protect human rights, and minimize corruption create an enabling environment for effective governance, policy implementation, and sustainable economic development. Consequently, countries with higher institutional quality are generally better positioned to achieve long-term social, economic, and development objectives (Uddin et al., 2023). IQ reflects the effectiveness of governance systems in promoting accountability, transparency, regulatory efficiency, and the rule of law, all of which influence a country’s economic and social development (Gebrue, 2024; Yousaf & Riaz, 2024). The institutional framework encompasses both formal institutions, including legal frameworks, public agencies, and government regulations, and informal norms that shape public and organizational behavior. High institutional quality is characterized by effective governance, strong legal enforcement, protection of property rights, and transparent policy implementation. These institutional characteristics play a crucial role in ensuring the efficient allocation of resources and the successful implementation of public policies. In the health sector, institutional quality directly affects the performance of healthcare systems. Weak institutions often result in poor governance, inefficient service delivery, and ineffective use of public health expenditure, particularly in developing economies. Conversely, strong institutions enhance transparency, strengthen accountability, and improve the management of public resources, thereby increasing the efficiency and impact of healthcare investments on population health outcomes (Onisanwa et al., 2026).

The present study investigates the LE determinants in China and the USA by examining the combined effects of economic, environmental, institutional and social factors. Given the growing importance of improving population health and achieving the SDGs, understanding the multidimensional drivers of LE is essential for designing effective public policies. This study makes five significant contributions to the existing literature. First, unlike previous studies that primarily focus on a single dimension, such as economic, environmental, social, or institutional determinants, this study develops a comprehensive framework that simultaneously incorporates environmental, economic, social, and institutional factors. By integrating these dimensions into a unified empirical model, the study provides a more holistic understanding of the complex factors influencing LE. Second, this study conducts a comparative analysis of China and the USA, two of the world’s largest economies with substantially different healthcare systems, socioeconomic structures, environmental conditions, and institutional frameworks. While numerous studies have investigated LE determinants within a single country, comparative evidence between these two major economies remains limited. This cross-country analysis highlights both the common and country-specific determinants of LE, offering valuable insights for policymakers. Third, this study extends the existing literature by examining three separate measures of LE: total LE, male LE, and female LE. Most previous empirical studies have relied solely on aggregate LE, thereby overlooking potential gender-specific differences in the effects of environmental, economic, social, and institutional factors. By estimating separate models for male and female LE, this study provides a more comprehensive understanding of gender-related disparities in health outcomes. Fourth, this study employs the ARDL model to estimate both the long-run and short-run relationships among the variables. To ensure the robustness and reliability of the empirical findings, the FMOLS estimator is also employed as a robustness check. The application of multiple econometric techniques strengthens the credibility of the estimated relationships and confirms the consistency of the results. Finally, this study offers policy recommendations that directly support the achievement of SDG-3, which aims to ensure healthy lives and promote well-being for all at all ages. Based on the empirical findings, the study proposes country-specific policy measures for China and the United States to improve LE through environmental protection, economic development, social progress, and institutional strengthening. These recommendations provide practical guidance for policymakers seeking to enhance public health outcomes while advancing sustainable development objectives.

2. Model and Estimation Technique

2.1. Model and Data

This study examines the determinants of male LE, female LE, and total LE in China and the USA. The empirical model is developed based on the existing literature, particularly the framework proposed by Zhang et al. (2025):

$${MLE}_t={\phi }_0+{{\phi }_{1}CO2e}_t+{{\phi }_{2}GDP}_t+{{\phi }_{3}HE}_t+{{\phi }_{4}PG}_t+{{\phi }_{5}RL}_t+{{\phi }_{6}GE}_t+{{\phi }_{7}EDU}_t+u_1$$ (1)

$${FLE}_t={\phi }_0+{{\phi }_{1}CO2e}_t+{{\phi }_{2}GDP}_t+{{\phi }_{3}HE}_t+{{\phi }_{4}PG}_t+{{\phi }_{5}RL}_t+{{\phi }_{6}GE}_t+{{\phi }_{7}EDU}_t+u_{2\hspace{0.17em}}$$ (2)

$${TLE}_t={\phi }_0+{{\phi }_{1}CO2e}_t+{{\phi }_{2}GDP}_t+{{\phi }_{3}HE}_t+{{\phi }_{4}PG}_t+{{\phi }_{5}RL}_t+{{\phi }_{6}GE}_t+{{\phi }_{7}EDU}_t+u_3$$ (3)

Where is equation (1), (2), and (3) MLE, FLE and TLE signifies the male life expectancy, female life expectancy, and total life expectancy. The CO2e, GDP, HE, PG, RL, GE and EDU represent the carbon emissions, gross domestic product, health expenditure, population growth, rule of law, government effectives and education respectively. The intercept is ${\phi }_0$ , while the ${\phi }_1\hspace{0.17em}to\hspace{0.17em}{\phi }_7$ represent the slop coefficient of the selective independent variables.

This study employs ten variables to examine the determinants of LE. The dependent variables include MLE, FLE, and TLE, all measured in years at birth. The explanatory variables consist of carbon dioxide emissions per capita (CO2e), gross domestic product per capita (GDP; constant 2015 US$), current health expenditure as a percentage of GDP (HE), population growth (PG; annual %), rule of law (RL), government effectiveness (GE), and expected years of schooling (EDU). The institutional quality variables, namely rule of law and government effectiveness, are measured using the Worldwide Governance Indicators (WGI) estimates, while expected years of schooling is measured in years. The data has been obtained from the world bank website. To ensure consistency in the scale of the variables and facilitate the interpretation of the estimated coefficients, FLE, MLE, TLE, CO2e, GDP, and EDU are transformed into their natural logarithmic forms. In contrast, HE and PG are retained in their original forms because they are already expressed as percentages. Likewise, RL and GE are not converted into natural logarithms because these variables contain negative values, making logarithmic transformation inappropriate. Therefore, the empirical analysis is conducted using the logarithmic forms of FLE, MLE, TLE, CO2e, GDP, and EDU, while HE, PG, RL, and GE are included in their original scales.

2.2. Estimation Technique

The first step of the empirical analysis is to examine the stationarity properties of the variables by conducting unit root tests. Establishing whether a time series is stationary is essential before estimating long-run relationships because non-stationary variables may produce misleading or spurious regression results, resulting in unreliable statistical inferences. Testing for unit roots ensures that the selected econometric approach is appropriate and that the estimated relationships are valid. In this study, we used the ADF and PP unit root tests in order to analyze the order of integration of variable components. The ADF test was developed by Dickey and Fuller (1979) and accounts for serial correlation through the use of prior data, while the PP test is a different technique developed by Phillips and Perron (1988), which uses non-parametric techniques. Employing both tests enhances the reliability of the stationarity assessment and provides robust evidence regarding the integration properties of the data. The general specification of the ADF test is presented in Equation (4).

$${\mathrm{\Delta }J}_t=\delta +\beta t+\gamma J_{t-1}+{\displaystyle \sum }_{i=1}^p{\delta }_i{\mathrm{\Delta }J}_{t-i}+e_t$$ (4)

Where in equation (4), According to Gujarati and Porter (2009), the alternative hypothesis (H1): $\varnothing$ <0 implies stationarity, whereas the null hypothesis (H0): $\varnothing$ =0 implies the existence of a unit root (non-stationary series).

Following the unit root analysis, this study examines the existence of a long-run equilibrium relationship among the variables using the Johansen and Juselius, 1990 cointegration test. Testing for cointegration is a crucial step when the variables are non-stationary, as it determines whether they share a stable long-term relationship despite exhibiting short-term fluctuations. If cointegration exists, the variables move together over time and converge toward a common equilibrium path. Ignoring cointegration may lead to misleading regression estimates and incorrect statistical inferences, particularly when analyzing non-stationary time series. The Johansen and Juselius (JJ) cointegration approach is well suited for multivariate time series analysis because it allows the identification of multiple cointegrating vectors within a system of variables. Consequently, it provides a rigorous framework for assessing long-run equilibrium relationships and supports the use of appropriate long-run estimation techniques.

After confirming the presence of cointegration, this study employs the ARDL model to estimate both the short-run dynamics and long-run associations among the variables. The ARDL approach is particularly advantageous because it can be applied when the variables are integrated of mixed orders, provided that none is integrated beyond the first order, that is, the variables are either I (0) or I (1) (Pesaran et al., 2001). In addition, the ARDL model effectively captures the dynamic adjustment process from short-run deviations toward long-run equilibrium, making it a flexible and reliable approach for time series analysis. The general specification of the ARDL model is presented below for the three dependent variables examined in this study:

$${\mathrm{\Delta }MLE}_t={\beta }_0+{\delta }_1{MLE}_{t-1}+{\delta }_2{CO2e}_{t-1}+{\delta }_3{GDP}_{t-1}+{\delta }_4{HE}_{t-1}+{\delta }_5{PG}_{t-1}+{\delta }_6{RL}_{t-1}+{\delta }_7{GE}_{t-1}+{\delta }_8{EDU}_{t-1}+{\displaystyle \sum }_{i=1}^n{\vartheta }_1{\mathrm{\Delta }MLE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_2{\mathrm{\Delta }CO2e}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_3{\mathrm{\Delta }GDP}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_4{\mathrm{\Delta }HE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_5{\mathrm{\Delta }PG}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_6{\mathrm{\Delta }RL}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_7{\mathrm{\Delta }GE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_8{\mathrm{\Delta }EDU}_{t-1}+e_t$$ (5)

$${\mathrm{\Delta }FLE}_t={\beta }_0+{\delta }_1{FLE}_{t-1}+{\delta }_2{CO2e}_{t-1}+{\delta }_3{GDP}_{t-1}+{\delta }_4{HE}_{t-1}+{\delta }_5{PG}_{t-1}+{\delta }_6{RL}_{t-1}+{\delta }_7{GE}_{t-1}+{\delta }_8{EDU}_{t-1}+{\displaystyle \sum }_{i=1}^n{\vartheta }_1{\mathrm{\Delta }FLE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_2{\mathrm{\Delta }CO2e}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_3{\mathrm{\Delta }GDP}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_4{\mathrm{\Delta }HE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_5{\mathrm{\Delta }PG}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_6{\mathrm{\Delta }RL}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_7{\mathrm{\Delta }GE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_8{\mathrm{\Delta }EDU}_{t-1}+e_t$$ (6)

$${\mathrm{\Delta }TLE}_t={\beta }_0+{\delta }_1{TLE}_{t-1}+{\delta }_2{CO2e}_{t-1}+{\delta }_3{GDP}_{t-1}+{\delta }_4{HE}_{t-1}+{\delta }_5{PG}_{t-1}+{\delta }_6{RL}_{t-1}+{\delta }_7{GE}_{t-1}+{\delta }_8{EDU}_{t-1}+{\displaystyle \sum }_{i=1}^n{\vartheta }_1{\mathrm{\Delta }TLE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_2{\mathrm{\Delta }CO2e}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_3{\mathrm{\Delta }GDP}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_4{\mathrm{\Delta }HE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_5{\mathrm{\Delta }PG}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_6{\mathrm{\Delta }RL}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_7{\mathrm{\Delta }GE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_8{\mathrm{\Delta }EDU}_{t-1}+e_t$$ (7)

The ARDL bounds testing method is used to determine whether the variables have a long-term connection. The null hypothesis assumes that no long-run equilibrium association exists among the variables ( $H_0:{\delta }_1={\delta }_2={\delta }_3={\delta }_4={\delta }_5={\delta }_6={\delta }_7={\delta }_8=0$ ).

On the other hand, cointegration is implied by the alternative hypothesis. The critical value boundaries and the computed F-statistic are used to make the choice. The null hypothesis, which states that there is no long-term link, cannot be rejected if the calculated F-statistic is less than the lower critical bound. In contrast, the null hypothesis is rejected and cointegration between the variables is confirmed if the F-statistic is greater than the upper critical bound. However, the test produces an equivocal result when the F-statistic falls between the lower and higher critical boundaries, and more data would be needed to establish if a long-term association exists. The Error Correction Model (ECM), which includes both the immediate impacts of changes in the explanatory variables and the rate at which deviations from the long-run equilibrium are corrected, is used to assess the short-run dynamics once cointegration has been established. After short-term shocks, the coefficient of the ECM should be statistically significant and negative, suggesting convergence toward the long-run equilibrium. The general specification of the ECM is presented as follows:

$${\mathrm{\Delta }MLE}_t={\displaystyle \sum }_{i=1}^n{\vartheta }_1{\mathrm{\Delta }MLE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_2{\mathrm{\Delta }CO2e}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_3{\mathrm{\Delta }GDP}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_4{\mathrm{\Delta }HE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_5{\mathrm{\Delta }PG}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_6{\mathrm{\Delta }RL}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_7{\mathrm{\Delta }GE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_8{\mathrm{\Delta }EDU}_{t-1}+{\omega }_1\text{EC}{M}_{t-1}+u_t$$ (8)

$${\mathrm{\Delta }FLE}_t={\displaystyle \sum }_{i=1}^n{\vartheta }_1{\mathrm{\Delta }FLE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_2{\mathrm{\Delta }CO2e}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_3{\mathrm{\Delta }GDP}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_4{\mathrm{\Delta }HE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_5{\mathrm{\Delta }PG}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_6{\mathrm{\Delta }RL}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_7{\mathrm{\Delta }GE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_8{\mathrm{\Delta }EDU}_{t-1}+{\omega }_1\text{EC}{M}_{t-1}+u_t$$ (9)

$${\mathrm{\Delta }TLE}_t={\displaystyle \sum }_{i=1}^n{\vartheta }_1{\mathrm{\Delta }TLE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_2{\mathrm{\Delta }CO2e}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_3{\mathrm{\Delta }GDP}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_4{\mathrm{\Delta }HE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_5{\mathrm{\Delta }PG}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_6{\mathrm{\Delta }RL}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_7{\mathrm{\Delta }GE}_{t-1}+{\displaystyle \sum }_{i=0}^n{\vartheta }_8{\mathrm{\Delta }EDU}_{t-1}+{\omega }_1\text{EC}{M}_{t-1}+u_t$$ (10)

A negative coefficient of the $EC{M}_{t-1}$ , indicates the existence of a stable long-run equilibrium association among the variables. It also reflects the speed at which short-run deviations from equilibrium are corrected over time. Specifically, a significant negative coefficient implies that any temporary disequilibrium between the dependent and explanatory variables gradually adjusts and converges toward the long-run equilibrium following a short-term shock (Iheanacho, 2016).

Following the estimation of the ARDL model, a series of diagnostic tests are conducted to evaluate the validity and stability of the estimated model. These tests assess important assumptions, including serial correlation, heteroskedasticity, model specification, normality of the residuals, and parameter stability. Satisfying these diagnostic criteria enhances the reliability and robustness of the estimated short-run and long-run coefficients. To further verify the consistency of the empirical findings, this study employs the Fully Modified Ordinary Least Squares (FMOLS) estimator as a robustness check. FMOLS is widely used to estimate long-run relationships among cointegrated variables and provides reliable estimates in the presence of non-stationary time series data. Developed by Phillips and Hansen (1990), the FMOLS estimator applies non-parametric corrections to address potential endogeneity and serial correlation in the error term, thereby producing unbiased, consistent, and efficient long-run parameter estimates. Comparing the FMOLS results with the ARDL estimates enables the study to assess the robustness and stability of the long-run relationships and strengthens the reliability of the empirical findings. Figure 4 shows the flow chart of Estimation technique.

Figure 4.

Estimation technique

3. Results and Discussion

3.1. Results

Table 1, shows the descriptive statistics for China and USA. For China, the mean value of MLE, FLE, TLE, CO2e, GDP, HE, PG, RL, GE and EDU are 4.2931, 4.3634, 4.3276, 1.8278, 8.6806, 4.6319, 0.4949, -0.40000, 0.2101 and 2.5634 respectively. The JB statistics confirms that all variables are normally distributed except PG. For USA, the mean value of MLE, FLE, TLE, CO2e, GDP, HE, PG, RL, GE and EDU are 4.3233, 4.3887, 4.3558, 2.8374, 10.9138, 15.6982, 0.784, 1.5257, 1.5026 and 2.7882 respectively. The JB statistics confirms that all variables are normally distributed except PG. Moreover, the USA demonstrates higher Mean values for GDP, HE, RL, GE, EDU, CO2e, and life expectancy measures, with lower Std. Dev. in most cases. China shows greater variability in GDP and CO2e. PG is the only variable in both countries that violates normality according to the JB test (P-value < 0.10), while all other variables satisfy the normality assumption.

Table 1.

Descriptive Statistic

China
Variable	Mean	Median	Maximum	Minimum	Std. Dev.	Skewness	Kurtosis	JB	P-value
Male life expectancy (MLE)	4.2931	4.297	4.3239	4.2464	0.0247	-0.3957	1.9003	1.8359	0.3993
Female life expectancy (FLE)	4.3634	4.3671	4.3961	4.3161	0.0257	-0.3356	1.8251	1.831	0.4003
Total life expectancy (TLE)	4.3276	4.3313	4.3593	4.2807	0.0251	-0.3615	1.857	1.8291	0.4007
CO2 Emissions (CO2e)	1.8278	2.0139	2.2225	1.0609	0.3665	-0.9285	2.5292	3.6698	0.1596
Gross domestic product (GDP)	8.6806	8.7789	9.4322	7.7131	0.5487	-0.3381	1.8162	1.8585	0.3948
Health expenditure (HE)	4.6319	4.472	5.9439	3.6174	0.6186	0.5644	2.5859	1.4455	0.4854
Population growth (PG)	0.4949	0.5657	0.788	-0.1038	0.2272	-1.3957	4.1032	9.0088	0.0111
Rule of law (RL)	-0.4000	-0.4679	0.0103	-0.6999	0.2039	0.6889	2.337	2.3381	0.3107
Government effectiveness (GE)	0.2101	0.1295	0.8079	-0.1588	0.2986	0.3802	1.907	1.7729	0.4121
Education (EDU)	2.5634	2.5818	2.7395	2.293	0.1389	-0.457	2.0909	1.6618	0.4356

Table 2 reports the results of the unit root tests, including the ADF and PP tests. For China, the ADF and PP results indicate that TLE, HE, PG, RL, GDP, GE, and EDU are non-stationary at level. However, after taking the first difference, all variables become stationary in both tests, confirming that they are integrated of order one, I (1). For the United States, the ADF and PP tests show that TLE, CO2e, GDP, HE, PG, RL, and GE are stationary at level. Nevertheless, after first differencing, all variables are found to be stationary, indicating that the series are either I (0) or I (1).

Table 2.

Unit Root Tests

China
Variable	ADF		PP
	I(0)	I(1)	I(0)	I(1)
Male life expectancy (MLE)	-3.1690*	-3.1491*	-3.8807*	-2.9184**
Female life expectancy (FLE)	-1.9920***	-4.3065*	-1.9330***	-3.5154*
Total life expectancy (TLE)	-0.7410	-2.8108*	-0.0171	-2.985**
CO2 Emissions (CO2e)	-4.3891*	-6.8054*	-5.8238*	-4.163*
Gross domestic product (GDP)	-1.4485	-5.7460*	-1.8353	-6.632*
Health expenditure (HE)	0.5790	-5.1201*	0.9125	-4.608*
Population growth (PG)	0.2141	-3.1524*	0.9281	-1.902***
Rule of law (RL)	0.0543	-4.2997*	0.1858	-4.016*
Government effectiveness (GE)	-0.7654	-5.5141*	-0.2942	-5.124*
Education (EDU)	-0.8715	-6.5060*	-1.3436	-3.701*

Note. *, ** & *** represent the 1%, 5% & 10% significance level.

Table 3 shows the ARDL bounds test, For China, the F-statistics for Model 1, Model 2, and Model 3 exceed the upper bound critical value of 3.91 at the 1% significance level. This indicates the rejection of the null hypothesis of no long-run relationship and confirms the existence of cointegration among the variables in all three models. Similarly, for the United States, the F-statistics for Model 1, Model 2, and Model 3 are also greater than the 3.91 critical value at the 1% level. Therefore, the null hypothesis of no cointegration is rejected, confirming a stable long-run relationship among the variables in each model.

Table 3.

ARDL Bounds Test

China
​	Test statistic	Value	Signif.	I(0)	I(1)
Model 1	F-statistic	7.926*	10%	1.7	2.83
Model 2	F-statistic	5.156*	5%	1.97	3.18
Model 3	F-statistic	10.853*	1%	2.54	3.91

Note. *, ** & *** represent the 1%, 5% & 10% significance level.

Table 4, shows the ARDL results for China and USA. In model 1, where the dependent variable is MLE, I model 2 the dependent variable is FLE, and model 3 the dependent variable is TLE. For both china and USA, the ARDL long run and short run results shows that CO2e and PG have negative effect on FLE, MLE and TLE. While the GDP, HE, RL, GE and EDU have positive effect on FLE, MLE and TLE.

Table 4.

ARDL Results

China
	Model 1			Model 2			Model 3
Long run
Variable	Coefficient	Std. Error	Prob.	Coefficient	Std. Error	Prob.	Coefficient	Std. Error	Prob.
CO2 Emissions (CO2e)	-0.862	0.140	0.000	-0.228	0.044	0.000	-0.795	0.348	0.059
Gross domestic product (GDP)	0.99	0.277	0.000	0.295	0.120	0.063	0.933	0.471	0.085
Health expenditure (HE)	0.022	0.267	0.938	0.346	0.163	0.084	0.144	0.057	0.040
Population growth (PG)	-0.766	0.265	0.038	-0.035	0.274	0.44	-0.142	0.044	0.000
Rule of law (RL)	0.717	0.245	0.029	0.368	0.155	0.061	0.634	0.142	0.000
Government effectiveness (GE)	0.077	0.013	0.000	0.252	0.103	0.050	0.865	0.194	0.000
Education (EDU)	0.963	0.471	0.050	0.203	0.020	0.000	0.518	0.211	0.040

For China, the coefficient of CO2e is -0.862, -0.228, and -0.795, indicating that a 1% increase in CO2e reduces male life expectancy (MLE), female life expectancy (FLE), and total life expectancy (TLE) by 0.862%, 0.228%, and 0.795%, respectively. The coefficient of GDP is 0.99, 0.295, and 0.933, suggesting that a 1% increase in GDP increases MLE, FLE, and TLE by 0.99%, 0.295%, and 0.933%, respectively. The coefficient of health expenditure (HE) is 0.022, 0.346, and 0.144, implying that a 1% increase in HE increases MLE, FLE, and TLE by 0.022%, 0.346%, and 0.144%. The coefficient of population growth (PG) is -0.766, -0.035, and -0.142, indicating that a 1% increase in PG reduces MLE, FLE, and TLE by 0.766%, 0.035%, and 0.142%. The coefficient of rule of law (RL) is 0.717, 0.368, and 0.634, showing that a 1-unit improvement in RL increases MLE, FLE, and TLE by 0.717%, 0.368%, and 0.634%, respectively. The coefficient of government effectiveness (GE) is 0.077, 0.252, and 0.865, indicating that a 1-unit increase in GE increases MLE, FLE, and TLE by 0.077%, 0.252%, and 0.865%. Similarly, the coefficient of education (EDU) is 0.963, 0.203, and 0.518, suggesting that a 1% increase in higher education increases MLE, FLE, and TLE by 0.963%, 0.203%, and 0.518%, respectively.

For the USA, the coefficient of CO2e is -0.270, -0.634, and -0.362, indicating that a 1% increase in CO2e reduces MLE, FLE, and TLE by 0.270%, 0.634%, and 0.362%, respectively. The coefficient of GDP is 0.520, 0.429, and 0.711, suggesting that a 1% increase in GDP increases MLE, FLE, and TLE by 0.520%, 0.429%, and 0.711%, respectively. The coefficient of health expenditure (HE) is 0.027, 0.197, and 0.248, implying that a 1% increase in HE increases MLE, FLE, and TLE by 0.027%, 0.197%, and 0.248%, respectively. The coefficient of population growth (PG) is -0.039, -0.067, and -0.035, indicating that a 1% increase in PG reduces MLE, FLE, and TLE by 0.039%, 0.067%, and 0.035%, respectively. The coefficient of rule of law (RL) is 0.400, 0.089, and 0.554, suggesting that a 1-unit improvement in RL increases MLE, FLE, and TLE by 0.400%, 0.089%, and 0.554%, respectively. The coefficient of government effectiveness (GE) is 0.323, 0.611, and 0.390, indicating that a 1-unit increase in GE increases MLE, FLE, and TLE by 0.323%, 0.611%, and 0.390%. Finally, the coefficient of education (EDU) is 0.711, 0.299, and 0.463, showing that a 1% increase in education increases MLE, FLE, and TLE by 0.711%, 0.299%, and 0.463%, respectively.

The results show that CO2e negatively affect life expectancy in both China and the USA, indicating that environmental degradation harms public health in both economies. The negative effect appears stronger in China, particularly for male and total life expectancy, suggesting that pollution exposure may be relatively higher or more harmful. Economic growth (GDP) improves life expectancy in both countries, but the impact is stronger in China, indicating that economic development still plays a larger role in improving health outcomes in emerging economies. Health expenditure positively influences life expectancy in both countries, although the effect is more consistent and statistically stronger in the USA, reflecting the advanced healthcare system. Population growth generally reduces life expectancy in both countries, but the negative effect is more pronounced in China, likely due to pressure on healthcare and public services. Institutional quality indicators, including rule of law and government effectiveness, show positive effects in both countries, highlighting the importance of strong institutions for improving health outcomes. Finally, education variable significantly enhances life expectancy in both countries, confirming that better education improves health awareness, healthcare access, and overall quality of life.

For China, the ECM coefficient is -0.025, -0.024, and -0.004 for the MLE, FLE, and TLE models, respectively. This indicates that 2.5%, 2.4%, and 0.4% of the short-run disequilibrium are corrected toward the long-run equilibrium in each period, showing a relatively slow adjustment speed toward equilibrium. For the USA, the ECM coefficient is -0.094, -0.029, and -0.062 for the MLE, FLE, and TLE models, respectively. This suggests that 9.4%, 2.9%, and 6.2% of the short-run deviations are corrected toward the long-run equilibrium in each period, indicating a faster adjustment process compared to China. The negative and significant ECM coefficients confirm the existence of a stable long-run relationship among the variables, while the results suggest that the USA adjusts more quickly toward long-run equilibrium than China.

The ARDL diagnostic analysis indicated that, for both China and the USA, the Jarque–Bera, LM test, ARCH test, and Ramsey RESET F-statistics were statistically insignificant. These results suggested that the ARDL life expectancy models satisfied the key diagnostic conditions, including normal distribution of residuals, absence of autocorrelation, and no heteroscedasticity. Overall, the findings confirmed that the estimated models were stable and well specified. The detailed diagnostic results, including the Jarque–Bera test, CUSUM, and CUSUM of squares, are presented in Appendix, see Figures A1–A6. Figure 5 shows the summary of the Empirical Finding.

Figure 5.

Empirical finding

Table 5 presents the robustness analysis using the FMOLS estimator. The findings indicate that CO2e and PG have negative effect on MLE, FLE, and TLE, while GDP, HE, RL, GE and EDU have positive effect on MLE, FLE, and TLE. These results are consistent with the findings obtained from the ARDL model.

Table 5.

Robustness Analysis (FMOLS)

China
Variable	Model 1			Model 2			Model 3
	Coefficient	Std. Error	Prob.	Coefficient	Std. Error	Prob.	Coefficient	Std. Error	Prob.
CO2 Emissions (CO2e)	-0.951	0.169	0.000	-0.642	0.036	0.000	-0.639	0.036	0.000
Gross domestic product (GDP)	1.689	0.175	0.000	0.348	0.119	0.009	0.349	0.121	0.010
Health expenditure (HE)	0.024	0.012	0.029	0.028	0.012	0.068	0.030	0.017	0.096
Population growth (PG)	-0.152	2.525	0.658	-0.027	0.032	0.416	-0.028	0.033	0.403
Rule of law (RL)	0.986	0.343	0.016	0.385	0.043	0.000	0.382	0.044	0.000
Government effectiveness (GE)	0.485	0.038	0.000	0.653	0.048	0.000	0.306	0.057	0.000
Education (EDU)	0.236	0.024	0.000	0.978	0.361	0.015	0.961	0.368	0.018

3.2. Discussions

The empirical findings reveal notable differences in the determinants of LE between China and the USA, reflecting their distinct stages of economic development, healthcare systems, environmental conditions, and institutional capacities. Although several variables influence life expectancy in both countries, the magnitude and significance of these effects vary considerably.

Environmental degradation, measured by CO2e, has a stronger adverse effect on life expectancy in China than in the USA. The larger coefficients for China suggest that environmental pollution poses a greater public health challenge, likely due to rapid industrialization, urban expansion, and higher exposure to environmental hazards. In contrast, the comparatively smaller impact in the USA may reflect stricter environmental regulations, cleaner production technologies, and more effective pollution control policies. The results align with the findings of Zhang et al. (2025) examined the impact of Air pollution on LE, and they found that Air pollution has negative effect on LE in China. The United States’ air quality and LE were studied by Hill et al. (2019). The study found that states with higher PM2.5 levels typically show lower average LE across a variety of LE outcomes and additive models.

GDP contributes positively to life expectancy in both countries; however, its influence is more pronounced in China. This finding implies that continued economic expansion in China generates substantial improvements in living standards, healthcare access, and public infrastructure, all of which translate into better health outcomes. In the USA, where income levels and healthcare infrastructure are already well established, the additional health gains from economic growth appear relatively smaller, suggesting diminishing marginal returns of income on longevity. Chetty et al. (2016) examined the association between income and LE in the USA from 2001–2014. The findings show that higher income is associated with longer LE across the income distribution. Considerable geographic disparities exist among low-income individuals, with LE differing by approximately 4.5 years between areas with the highest and lowest longevity. Between 2001 and 2014, changes in LE also varied substantially across regions, ranging from gains of more than four years to declines of over two years. Geographic differences in LE among the lowest income quartile were strongly associated with smoking behavior but showed no significant relationship with healthcare access, environmental conditions, income inequality, or labor market characteristics. Our finding is consistent with Adebayo et al. (2024), who examined the effects of health expenditure and income on LE in the United States. The study found that higher income, educational attainment, food production, and HE positively influence LE.

The contribution of health expenditure also differs across the two countries. In the United States, healthcare spending consistently improves male, female, and total life expectancy, highlighting the important role of healthcare investment in enhancing population health. In contrast, the long-run effect of health expenditure in China is less consistent, indicating that increases in healthcare spending alone may not be sufficient to maximize health outcomes without improvements in the efficiency and accessibility of healthcare services. Our finding is similar with the finding of, Liu and Zhong (2022), who examined the HE, renewable energy and LE nexus in china, and they found that HE has positive effect on LE in china. Li et al. (2025) studied the social and Economic determinants of LE in China, and they found that HE has positive effect on LE in China. Population growth exhibits a more pronounced negative influence in China than in the United States. This finding suggests that demographic pressures place greater strain on China’s healthcare system, public resources, and environmental quality. Although population growth also reduces life expectancy in the United States, its overall impact is comparatively weaker, reflecting greater institutional capacity to accommodate demographic changes.

Institutional quality positively influences life expectancy in both countries, but the relative importance of its components differs. Government effectiveness appears to have a stronger influence in the United States, indicating that efficient public administration and effective policy implementation play a critical role in improving health outcomes. Conversely, the rule of law exerts a relatively stronger effect in China, suggesting that institutional reforms aimed at strengthening governance, regulatory enforcement, and legal accountability can substantially improve population health. The finding is same with the finding of Shah et al. (2021), who examined the effects of environmental quality and public environmental expenditure on LE in China using quarterly data from 1999Q1 to 2017Q4. The study found that improvements in environmental quality and increased public spending on environmental protection significantly enhanced life expectancy. Hadipour et al. (2023) investigated the role of IQ in health outcomes using panel data from 158 countries. The study reported that stronger institutional quality significantly reduced infant mortality and increased life expectancy. Zhang et al. (2026a) examined the impact of environmental degradation, information and communication technology (ICT), health expenditure, and institutional quality on life expectancy in China. The study found that CO2 emissions significantly reduced life expectancy, whereas ICT, health expenditure, and institutional quality had positive and significant effects.

Education enhances life expectancy in both countries, but its impact is considerably larger in China. This result indicates that improvements in educational attainment continue to generate substantial health benefits through increased health awareness, healthier lifestyles, and greater access to economic opportunities. In contrast, the smaller educational effect observed in the United States suggests that the country has already achieved relatively high levels of educational attainment, leaving less room for additional gains in life expectancy. Luy et al. (2019)examined the contribution of increasing education levels to improvements in life expectancy in Italy, Denmark, and the United States over the period 1990–2010. Using a decomposition analysis, the study found that rising educational attainment significantly contributed to gains in life expectancy, in addition to reductions in mortality. The authors concluded that education plays a vital role in improving population health and should be considered an important indirect health policy. Sart et al. (2024) investigated the effects of education and income inequalities, ICT, CO2 emissions, and real GDP per capita on life expectancy in the new European Union member states during 2010–2022. Employing a fixed-effects regression model, the study found that education inequality, income inequality, and CO2 emissions reduced life expectancy, whereas ICT development and economic growth significantly enhanced longevity. The authors emphasized that reducing educational and income disparities is essential for improving population health and increasing life expectancy.

4. Conclusions and Policy Recommendations

This study explored the determinants of life expectancy (male, female and Total) in China and USA by incorporating a comprehensive set of institutional, economic, environmental and social determinant of LE in China and USA. This study utilized the ARDL and FMOLS estimators from 2000 to 2023. The ARDL long and Short run estimates showed that CO2e and population growth have negative effect on MLE, FLE and TLE in China and USA, while the health expenditure, GDP, education, rule of law and government effectiveness have positive effect on MLE, FLE and TLE in China and USA.

4.1. Policy Recommendations for China

First, China should accelerate the transition from coal-intensive energy production toward renewable and low-carbon energy sources. The stronger negative effect of environmental degradation on life expectancy supports reducing emissions from coal-based power generation and energy-intensive industries. Second, industrial pollution should be controlled through stricter emission standards, continuous pollution monitoring, and stronger enforcement in highly industrialized regions. Reducing pollution exposure can improve air quality and population health. Third, healthcare investment should be expanded, particularly in rural and less-developed regions. The positive contribution of health expenditure to life expectancy supports greater investment in primary healthcare facilities, medical equipment, healthcare professionals, and essential medical services. Fourth, preventive healthcare should receive greater attention through disease screening, vaccination, health monitoring, and community-based healthcare programs. Greater emphasis on prevention can reduce the long-term burden of chronic diseases and improve population health. Fifth, China should strengthen policies addressing population ageing by expanding elderly healthcare, long-term care facilities, and community-based support services. These measures can help address increasing healthcare needs associated with demographic ageing. Sixth, investment in education should continue with greater emphasis on health literacy. The positive role of education in life expectancy supports programs that improve awareness of nutrition, disease prevention, environmental risks, and healthy lifestyles. Seventh, economic development should be accompanied by broader improvements in living conditions and public services. The positive relationship between economic conditions and life expectancy suggests directing economic gains toward healthcare, sanitation, housing, and social protection for lower-income and rural populations. Eighth, environmental and health regulations should be integrated into regional development planning. Highly industrialized areas should apply stronger environmental assessments to projects with substantial pollution and energy requirements. Ninth, institutional capacity should be strengthened through better coordination between central and local governments. Clear responsibilities, transparent monitoring, and effective enforcement can improve the implementation of environmental, healthcare, education, and population policies. Tenth, regional disparities in health outcomes should be incorporated into policy design. Public resources should prioritize areas with lower healthcare access, higher pollution exposure, and weaker socioeconomic conditions to support more balanced improvements in life expectancy.

4.2. Policy Recommendations for the USA

First, the USA should strengthen efforts to reduce transportation and industrial emissions through effective enforcement of environmental standards and greater adoption of low-emission technologies. The negative effect of environmental degradation on life expectancy supports continued action against major pollution sources. Second, pollution-control policies should focus on communities facing relatively high environmental health risks. Stronger environmental monitoring and regulatory enforcement can reduce exposure to industrial emissions and other major pollution sources. Third, healthcare resources should be allocated more efficiently toward preventive and primary healthcare. The positive effect of health expenditure on life expectancy supports greater emphasis on disease prevention, early diagnosis, and accessible primary care. Fourth, affordable healthcare access should be expanded among lower-income and underserved populations. Reducing financial and geographic barriers to medical services can help address persistent differences in health outcomes across communities. Fifth, socioeconomic disparities should be considered in health policy design. The positive association between income and life expectancy supports policies that improve household economic security, employment opportunities, and basic living conditions. Sixth, healthcare services for an ageing population should be strengthened through long-term care, geriatric services, home-based care, and community healthcare programs. These measures can address increasing healthcare demand associated with demographic ageing. Seventh, education and health literacy should be strengthened through schools and community programs. Greater awareness of preventive healthcare, nutrition, physical activity, disease prevention, and environmental health risks can support longer life expectancy. Eighth, greater emphasis should be placed on medical innovation and efficient healthcare delivery. Advances in diagnosis, treatment, and healthcare management can improve health outcomes when combined with accessible and affordable medical services. Ninth, coordination between federal, state, and local authorities should be improved in health and environmental governance. Better institutional coordination can facilitate timely responses to regional health risks and strengthen policy implementation. Tenth, evidence-based monitoring should be incorporated into long-term health policy. The relatively faster adjustment toward long-run equilibrium observed for the USA supports regular monitoring of environmental, economic, and healthcare conditions to facilitate timely policy responses.

4.3. Limitation of the Study

This study has several limitations that provide directions for future research. First, the analysis is limited to China and the USA, which restricts the generalizability of the findings to other countries and regions. Future research could extend the analysis to a broader group of economies to examine whether the identified relationships remain consistent across different economic, institutional, and environmental settings. Second, the study covers the period from 2000 to 2024 based on data availability. Future studies could use longer time periods as additional data become available to capture longer-term changes and provide more comprehensive evidence on the determinants of life expectancy.

ORCID iD

Tian Liu https://orcid.org/0000-0001-8698-3279

Ethical Considerations

This study was conducted using publicly available secondary data obtained from the World Bank’s World Development Indicators (WDI) database Therefore, ethical approval and informed consent were not required in accordance with institutional and international ethical guidelines.

Consent to Participate

This study used publicly available secondary data and did not involve human participants, patients, or identifiable personal information.

Data Availability Statement

Data will made available on request.*