Financial Development, Governance, Environmental Pressure, and Health Expenditure: A Panel Analysis

2.1 Theoretical assessment

Industrialisation has long been hailed as a driver of economic growth, technological advances, and social progress. Nevertheless, the price paid for industrialisation goes well beyond economic wealth: it undermines environmental sustainability and public health (Soni, 2024; Van Tran, Tran, Bui Hoang, & Mai, 2024; Zhang, Zheng, Xia, & Cheng, 2024). The cyclical degradation of ecosystems, increasing pollution, and the subsequent healthcare costs need to be understood within a theoretical framework. Drawing on established theories of the economy and the environment, this study offers a categorical assessment of the relationships among industrial growth, environmental damage, and increasing pressure on health care systems. Being one of the most popular theoretical frameworks that maps the interaction between economic progress and environmental performance, the Environmental Kuznets Curve (EKC) conjecture provides an interesting storyline. In a given framework, nascent and growing economies experience a period of expanding environmental degradation, followed by a decline as mature economies adopt cleaner technologies (Hassan, Yang, Usman, Bilal, and Ullah, 2023). In the early phases of industrialisation, most countries are more focused on GDP growth rates at the expense of protecting ecological integrity, thereby creating significant pollution and ecological loss. With increases in per-capita income, expenditure shifting in favour of less-polluting sectors, and regulatory pressures converging, long-term pollutant emissions decrease. Despite the EKC’s positive outlook, it has limited applicability, particularly in developing economies with weak regulatory systems. The assumption that environmental degradation inevitably shrinks with economic growth does not hold when one considers long-term pollutants such as greenhouse gases and hazardous waste, which only increase as industrial bases expand.

Conversely, the Pollution Haven Hypothesis (PHH) explains that polluter-rich industries often move to less robust jurisdictions or develop there because of the high income and more stringent environmental regulations in the relatively wealthy states (Bradu et al., 2023; Dritsaki and Dritsaki, 2023; Zeeshan, Han, Rehman, Ullah, and Mubashir, 2022). A high degree of unfairness has been created through this dynamic in the distribution of industrial pollution, where the developing world is the victim of ecological devastation and its resultant impacts on the population and their well-being. In most of these locales, morbidities associated with pollution, especially respiratory and cardiovascular diseases, are significantly more common than in urbanised localities, resulting in substantial medical costs. It has been shown empirically that governments that have lax environmental policies serve as an attraction to the most polluting industries and thus contribute to the escalation of the health crisis in the respective countries (Hamid and Wibowo, 2023; Hassan et al., 2023; Tackie, Chen, Ahakwa, and Atingabili, 2022). The PHH thus disputes the wishful thinking that growth can create positive or negative conditions on its own, focusing instead on the necessity of institutional frameworks, multilateral coordination, and coordinated industrial and environmental regulations. One way to further understand the economics of industrial pollution is through the Grossman Health Production Function. This model concludes that health outcomes are a function of multiple interacting inputs, including lifestyle and behaviour, medical care, and environmental context. If industrialisation increases pollution, many individuals and governments will need to invest more in healthcare to prevent pollution-related diseases. The cost of such industrial harm is a negative externality, as producers do not pay the full social costs of the environmental harm they cause. This causes pollution levels above the socially optimal level and additional medical expenditures. These externalities frequently cause governments to step in by subsidizing medical treatment or imposing pollution taxes to internalize them. Nevertheless, without meaningful enforcement mechanisms, industries are allowed to externalize the costs to the environment and, ultimately, our healthcare systems that are overwhelmed with preventable diseases (Dritsaki & Dritsaki, 2023; A. Mujtaba et al., 2022; S. Roy & Khatun, 2022; Y. Shang, Razzaq, Chupradit, An, & Abdul-Samad, 2022).

The discussion of industrialization and environmental degradation should not be divorced from the common talking points on sustainable development worldwide. A multidimensional framework for mapping the impact of Industrial Growth is drawn from the United Nations Sustainable Development Goals (SDGs). SDG 3, which addresses good health and well-being; SDG 9, which promotes sustainable industrialization; and SDG 13, which calls for climate action, are key to considering the impact of pollution on healthcare costs. The dilemma that arises thereupon is how to balance robust economic growth with the damaging environmental safeguards and health needs of the citizenry. Some countries have managed to minimise the financial pressure of medical care due to the process of implementing sustainable industrial relations, in addition to investments in renewable sources of energy, as well as in green technologies.

2.2 Empirical assessment

A. Carbon dioxide emissions and health expenditure

The former group speaks about the positive relationship. The relationship between carbon dioxide emissions and health expenditure is analyzed across total, government, and household health expenditure. For example, Raihan et al. (2022) show that increased carbon emissions intensity is associated with higher healthcare costs and worsening environmental conditions. Furthermore, Pichler, Jaccard, and Weisz (2019) believe that chronic diseases can also be precipitated by carbon emissions, requiring long-term medical therapy, continuous medical intervention, and increased medical costs. As the publications by Wang, Dong, and Dong (2021) and Zhao, Jiang, Dong, and Dong (2021) show, suboptimal air quality due to carbon emissions can lead to the emergence of various diseases, and people also need health assistance, which increases the cost of health care. Vyas, Mehta, and Sharma (2023) found that a limited budget allocated to healthcare may affect long-run capital spending on health, leading to a significant increase in health expenditures. The strong, positive relationship between CO2 and healthcare spending posits that CO2 increases healthcare spending, as supported by Kutlu and Örün (2023).

The report outlines the health risks to the population from increased CO₂ emissions and environmental deterioration. As demonstrated by Dritsaki and Dritsaki (2023), G7 CO₂ emissions are positively correlated with health expenditure, as CO₂ emissions add to the costs of health care and to the negative impacts of environmental pollution on human health and the economy. According to Chaabouni and Saidi (2017), the correlation between health expenditure and CO₂ emissions is significant and positive, indicating the negative impact of pollution on human health and the costs imposed on society. Interestingly, the findings suggest that a 1 percent increase in CO₂ emissions is linked with a high increase in healthcare bills by 2.5 percent. Apergis, Gupta, Lau, and Mukherjee (2018) assess the influence of CO₂ emissions on state healthcare expenditure between 1966 and 2009 and find a positive relationship, but only in states that spend more on healthcare. The emissions of CO₂ and expenditure on health are influenced by complex regional factors such as climate and energy requirements. The possible advantage of CO₂ reduction is healthcare savings. The report outlines the need for a coordinated effort and political will to cut CO₂ emissions cost-effectively and to address the complex relationship between environmental issues and health care expenditures. The quantitative impact of CO₂ emissions in China on healthcare expenditure (HCE) is that, though not as much as income, the impact of CO₂ emissions on the extension of HCE is positive, particularly at higher quantiles. As demonstrated by Zeeshan et al. (2022), family health expenditure increases with rising CO₂ emissions, highlighting the health hazards of environmental pollution and its economic costs. Another statistically significant, non-zero positive correlation between CO₂ emissions, environmental pollution, and household spending in Chinese health is also found in their study. On the other hand, there is no confirmation of an adverse nexus between carbon dioxide emissions and health spending.

B. Industrialisation and health expenditure

Having two lines of evidence on the relationship between health expenditure and industry is taken over. A positive correlation is observed in the former group, indicating that health spending increases with industrial development. For example, Raghupathi and Raghupathi (2020), Jakovljevic et al. (2017), and Hassan et al. (2023) associate health-care costs with industrialisation and economic growth, driven by adverse environmental effects. On the other hand, the fragmentation of roles and worsening working conditions are forms of industrialisation of health that can also raise costs. Abbas Khan et al. (2019) found that, as trade volume increases, CO₂ emissions rise, thereby increasing health expenditures. Y. The study by Shang et al. (2022) also revealed that the health burden worsens due to carbon emissions from urbanization and industrialization, leading to increased health costs. In the article, Hassan et al. (2023) demonstrate that health-care expenditure increased in the 10 leading countries from 1995-2018, driven by industrialization. Their findings reveal a positive correlation between industry and health costs, and that clear policies are required to address the health-economic implications of industrialization. Nadeem, Ali, Khan, and Guo (2020) note that the effect is complex. Industrial clusters that lead to pollution can increase or reduce residents’ health expenditures. Industrialization influences the cost of health care through rising incomes and the negative impacts of pollution. Kraipornsak (2017) determined that increased use of green energy in logistics and business processes can reduce spending on general health and increase labour productivity, as well as environmental health. The authors suggested that the creation and implementation of green technologies have a strong positive effect on the environment and personal health (Dong, Xue, Xiao, and Liu, 2021; Shahzad et al., 2020). Tackie et al. (2022) applied a PMG-ARDL model to demonstrate that industrialization is associated with lower panel expenditure on health and recommended that industrialized countries invest more in health. Their statistics confirm that industrialization and health expenditure in West African economies are positively linked.

T. Shang, Samour, Abbas, Ali, and Tursoy (2024) noted that lower health expenditure can be achieved through pollution-driven industrial clusters, provided technological innovation, increased employment, expanded medical services, and environmental infrastructure are put in place to improve the situation. Nevertheless, there are still locations where expenditures are higher due to rising pollution or higher incomes. Ampon-Wireko et al. (2022) found a unidirectional causal relationship between public expenditure on health and industrialisation, meaning that the higher the level of industrialisation, the greater the health spending. This implies that health care is a subject in which more resources are invested in more developed nations. Another body of research links the concept of green industrialisation to health expenditure. The primary objective of green industrialisation is to minimise environmental damage, while also improving population health and reducing healthcare costs. Renewable energy and cleaner technologies reduce respiratory and cardiovascular disease hospitalisations by reducing air pollution (Sarfraz, Ivascu, and Cioca, 2022). Water is available responsibly and without pollution, which preserves clean water supplies, reducing waterborne illnesses and expenses (Ferreira, Graziele, Marques, and Goncalves, 2021). In industrial processes, safer chemicals reduce occupational illnesses, thereby minimising healthcare costs. Environmentally-friendly transportation reduces air pollution and traffic crashes, enhances respiratory health, and decreases expenditures (Tchapchet Tchouto et al., 2024). Safety procedures at the workplace reduce accidents and injuries, thereby reducing absenteeism and health care costs (Hadi and Nayeri, 2023). Green industrialisation reduces the impact of polluters and toxins on the environment, thereby preventing disease and associated health costs (Bradu et al., 2023). Sustainable agriculture and healthier food can support population health and reduce health spending (Ashraf et al., 2021).

C. FDI and health expenditure

Proving the correlation between Foreign Direct Investment and total, government, and household health expenditure. The article by Ehsani, Dashtban Farouji, Khoshnoodi, and Dashtban Farouji (2023) has conducted an empirical model study with the help of the nonlinear autoregressive distributed lag (NARDL) model that revealed positive and significant results, indicating the positive impact of Foreign Direct Investment (FDI) on health expenditure and, consequently, population health and life expectancy in the long term. The statistics indicate that FDI enhances economic growth, leading to high spending on health and health promotion. According to research by Alziyani and Murad (2021), African health spending increases with FDI, good governance, and larger city populations. The results indicate that FDI increases health spending in regions, thereby enhancing health development. Similarly, Giammanco and Gitto (2019) suggest that FDI rises with EU public health expenditure. The connection between money and commitment to healthcare infrastructure is highlighted in the project (Barkat, Alsamara, Al Kwifi, & Jarallah, 2024). According to Immurana (2021), Foreign Direct Investment (FDI) has a positive impact on the African health sector, enhancing a country’s health capacity in the short and long term and contributing to life expectancy and mortality. It means that higher levels of FDI can enhance health outcomes, and policies to promote these investments should be emphasised, along with additional steps to maximise welfare benefits, including health benefits. Unver and Erdogan (2015) note that health spending in these areas can be negatively affected by foreign direct investment (FDI). A study has shown that foreign direct investment (FDI) enhances the long-term health of the Bangladeshi people. The inquiry claims that FDI enhances health outcomes and that healthcare and hygiene policies can maximise its impact on the country’s health sector.

D. Clean energy and health expenditure

The first category shows a positive correlation, indicating a relationship between renewable energy and total health expenditure (both government and household). The results of the study conducted by Zhu (2023) demonstrate that when renewable energy sources are used in cooking, including electricity, natural gas, liquid gas, methane, or solar energy, the health outcome of rural people also improves since discomfort decreases, and physical activity increases. The use of clean energy did not influence self-reported health, bronchitis, asthma, or medical costs, yet it has a positive effect and may lead to cost savings in rural regions. According to this study by Zhongwei and Liu (2022), the use of clean energy, including renewable energy, raises the Chinese life expectancy. This positive outcome demonstrates that promoting renewable energy production and consumption can enhance health. Therefore, clean energy projects can improve the general health of the population and reduce healthcare costs. Li, Ozturk, Majeed, Hafeez, and Ullah (2022) emphasise that renewable energy can improve health outcomes, including reducing chronic disease rates. Thus, the fact that clean energy is more health-promising refers to potential cost savings in health-related costs. It demonstrates that clean-energy logistics enhances the sustainability and profitability. Green energy can benefit both health and the economy by promoting sustainability. Greener energy can reduce health care costs, as environmental performance adversely affects health expenditure. ICT and renewable energy in Pakistan cut on health expenditure. The research by Shahzad et al. (2020) indicated that health-related costs are reduced due to clean and renewable energy and breakthroughs in ICT. According to the findings, politicians are advised to invest in renewable energy and ICT to improve the environment and reduce healthcare costs—the researchers of A. Khan, Chenggang, Hussain, and Kui (2021) found that countries that used renewable energy-based Belt and Road Initiative (B&RI) projects had better environmental quality and lower CO2 emissions. Cleaner energy could help reduce healthcare costs by improving the environment. Clean energy protects the environment and reduces healthcare costs, enhancing sustainability and health. In the study, Ullah, Rehman, Khan, Shah, and Khan (2020) note that Renewable energy (RE) saves money in Pakistan by reducing health costs. The statistics indicate that the utilisation of renewable energy decreases health expenditure. This implies that the inclusion of renewable energy sources in the energy mix will lower CO₂ emissions and health costs.

3.1 Conceptual framework and mode specification of the study

The conceptual framework of this study (see Figure 1) is grounded in the health production perspective, which treats health as an outcome shaped by both medical inputs and non-medical conditions. In this view, health expenditure is not determined only by the structure of the health system. It also responds to broader economic, environmental, and institutional forces. Prior work based on the health production function shows that health outcomes depend on medical resources, as well as socio-economic, financial, and physical conditions, while environmental variables and socio-economic status act as non-medical influences on health demand. This provides a suitable theoretical base for linking financial development and environmental pressure to health expenditure.

Figure 1. Conceptual framework of the study.

Within this framework, financial development affects health expenditure through two main channels. The first is a resource channel. A deeper financial system improves access to credit, capital mobilisation, insurance coverage, and public financing capacity. These changes can expand households’ ability to pay for healthcare and governments’ and private providers’ ability to finance medical infrastructure and services. In this sense, financial development can raise health expenditure by expanding access to formal health financing and increasing the supply of health-related investment. The second is a structural channel. Financial development often stimulates industrial activity, trade expansion, and investment flows. These processes can improve income and production, but they may also intensify pollution and ecological stress when growth depends on carbon-intensive production. Studies of the finance-environment nexus show that financial development may initially worsen environmental quality, although the relationship may later improve as economies adopt cleaner technologies and stronger regulation.

Environmental pressure is introduced as the main transmission mechanism linking economic expansion to rising healthcare costs. The pollution-health literature shows that exposure to air pollution and related environmental risks increases disease incidence, hospitalisation, and treatment demand. Extensions of the Grossman-type health demand model explicitly include pollution as a determinant of health needs, showing that environmental degradation reduces health capital and increases the demand for medical care. This implies that carbon emissions, industrial pollution, and other environmental stressors raise health expenditure by increasing both preventive and curative healthcare needs.

The framework also assigns an important role to governance and the energy transition. Governance conditions how effectively financial resources are allocated, how strictly environmental regulations are enforced, and how efficiently health expenditure is converted into better health outcomes. Renewable energy, in turn, reduces environmental pressure by lowering reliance on fossil-fuel-intensive production and consumption. Thus, the study expects industrialisation, trade openness, foreign direct investment, and carbon emissions to increase health expenditure, while renewable energy is expected to reduce it. Financial development may raise health expenditure directly through financing capacity and indirectly through environmental change. This integrated framework strengthens the theoretical basis of the study by showing that health expenditure is the outcome of a connected system in which finance, environment, production structure, and institutions jointly shape health system costs over time.

The theoretical framework of the study investigates the relationships among health expenditure (HE), industrialisation (IND) and environmental degradation (CO 2), as well as foreign direct investment (FDI) and financial development (FD), clean energy (CE) and trade openness (TO). Using the significant economic and environmental theories, we can see the picture of these relations. The environmental Kuznets Curve (EKC) hypothesis is frequently used to evaluate the relationship that exists between industrialisation and health spending. The implications of this hypothesis are as follows: when an economy begins to grow and becomes industrialised, the rate of environmental degradation increases; however, this decreases once an economy has reached a certain level of development. A lack of environmental awareness, increased regulation and the adoption of greener technologies have caused this. There is also a possibility that industrialisation will increase CO2 emissions, negatively impacting health and, therefore, spending on health (Dam, Durmaz, Bekun, and Tiwari, 2024; Y. Shang et al., 2022; J. Wang et al., 2021; Zeeshan et al., 2022). One of the significant aspects of the global economy is FDI. It can enhance growth and industrialisation. There are, however, environmental risks when multinational companies take advantage of lax environmental rules in host nations, a scenario described by the Pollution Haven Hypothesis (PHH). According to this assumption, FDI can increase pollution levels, which in turn can raise healthcare costs through ecological destruction (Chireshe and Ocran, 2020; A. Khan et al., 2021; T. Shang et al., 2024). The role of financial development is two-fold. In the long term, it is possible to reduce health expenditure by investing sufficiently in clean technologies and medical infrastructure.

Nevertheless, rapid economic growth in the absence of environmental regulations may lead to greater environmental degradation, thereby increasing health spending. The use of clean energy is traditionally associated with reduced environmental damage and improved health outcomes for the population. With the adoption of renewable energy sources, CO2 emissions will decrease significantly, reducing the adverse effects of pollution on health and potentially lowering healthcare costs. Trade openness may also significantly influence health spending by affecting economic growth and environmental quality. Greater trade can potentially drive economic growth, though at the cost of some problems (including environmental degradation at the outset and greater health spending). In the long run, openness to trade may facilitate the exchange of cleaner technologies and other environmental best practices, thereby reducing health spending.

All in all, the theoretical development of the study shows that different effects interact, indicating that industrialisation, FDI, financial development, clean energy, and trade openness influence health expenditures. These aspects affect health spending through negative environmental degradation and economic development. These relationships can only be understood deeply when policymakers aim to achieve a balance among economic development, human health, and environmental sustainability.

The objectives of the study are to evaluate the Impact of industrialisation (IND), environmental degradation (CO2), foreign direct investment (FDI), financial development (FD), clean energy (CE), and trade openness (TO) on health expenditure (HE) in Lower Middle-income countries and higher-middle-income nations for the period 1995-2020. The general form of empirical relations is presented below;

After the natural log transformation of Equation (1), it can be rewritten in regression form in the following manner, see Equation (2):

Where the coefficients of ${\beta }_1\dots \dots ..{\beta }_7$ explain the changes in HE due to variations in target variables, ${\alpha }_0$ explains the constant term and ${\epsilon }_{\mathrm{i},\mathrm{t}}$ for white noise in the equation.

3.2 Measures of variables

Carbon dioxide (CO2) emissions are the first explanatory variable, reflecting the release of CO2 into the atmosphere. This mainly occurs due to human activities such as industrial production, transportation, and energy production. Carbon dioxide emissions are typically measured in metric tons per capita or in total carbon dioxide emitted over a given period. Other sources are official environmental agencies within a country, international bodies such as the World Bank, and research institutions. The increase in carbon dioxide emissions is one of the causes of environmental pollution, which has adverse consequences for human health, including respiratory diseases, cardiovascular diseases, and other health complications. Research shows that the costs of treating pollution-related diseases, including hospitalisation, medication, and medical care, may increase health spending. Therefore, the cost of treating pollution-induced illnesses can strain healthcare budgets, leading to increased government spending on healthcare systems and facilities. Nevertheless, other research has shown that applying controls to curb pollution and encouraging clean energy programmes can reduce CO2 emissions and, consequently, medical expenditures for diseases caused by pollution.

Furthermore, improvements in environmental health can be achieved through increased research and development funding to develop pollution-mitigation technologies, which eventually translate into lower healthcare costs (Eckelman et al., 2020). Government initiatives to address pollution-related health challenges can also create employment opportunities in the healthcare sector, leading to economic growth and, indirectly, reducing healthcare spending by boosting tax collection. However, our findings show that carbon dioxide emissions and health expenditure are positively related. This increase in CO2 emissions is expected to raise healthcare costs due to the adverse health effects of pollution, with β1 = 1 = 2HE/2CO2.

The second independent variable is industrialisation, which refers to the process of economic development that entails the development of an economy’s manufacturing and industrial sectors (Hauge & Chang, 2019). One of the possible methods to evaluate the state of the process of industrialization is through various indicators, including the ratio of industry to the Gross Domestic Product (GDP), Study (Tulchinsky & Varavikova, 2014; Dembe, 2001) went further and claimed that industrialization might increase pollution rates, occupational health risk, and exposure to harmful substances which subsequently translates into greater healthcare costs. Accidents, injuries, and occupational diseases may also result from the development of industrial operations and require medical care and recovery. Quite the contrary, some studies have shown that the dividends of industry growth may lead to the development of new technologies in healthcare provision, such as healthcare equipment and medicines, as well as more efficient treatment strategies. It can eventually improve healthcare outcomes and reduce the costs of long-term care. Further, industrialisation may help enhance living standards, sanitation, and access to healthcare, thereby achieving better health outcomes and possibly lowering the cost of developing widespread healthcare spending to treat preventable illnesses. However, in our investigation, industrialisation will have a positive impact occasionally on health expenditure owing to possible growth in pollution diseases as well as work health risks.

The third explanatory variable is foreign direct investment (FDI). It occurs when a foreign entity invests its capital into the domestic economy as a way of setting up business or purchasing assets. FDI is measured by aggregating FDI inflows or as a percentage of a nation’s GDP. FDI has many sources of data, among them national investment promotion agencies, central banks and international bodies such as UNCTAD and OECD. According to some studies, FDI can enhance economic growth, advance infrastructure, and generate jobs (Zekarias, 2016). This, in its turn, may increase accessibility to healthcare services and funding.

The fourth independent variable is clean energy, incorporating renewable sources such as solar, wind, and hydroelectric power, providing options beyond traditional energy sources that depend on fossil fuels. Consumption of Clean energy may be measured either as a share of overall energy consumption or in units such as terawatt-hours. Information on the use of clean energy can be found through various sources, such as national energy agencies, international organisations such as IRENA, and energy research institutions. Previous evidence indicates that clean energy has the potential to reduce air pollution, respiratory diseases, and the medical costs associated with pollution-related illnesses. The investment in clean energy technologies will also generate employment, boost economic development, and improve the population’s health outcomes, thereby reducing health spending. Other sources, however, assert that a short-term investment in clean energy infrastructure and technology may incur these costs at the outset, temporarily increasing healthcare expenditure. We have analysed that clean energy will save on health expenditure by minimising pollution-related diseases and health care costs.

The dependent variable is health expenditure, defined as the monetary amount spent on the referral and support of healthcare services (Ke, Saksena, and Holly, 2011). It is measured by three major elements: total health expenditure, government health expenditure, and household health expenditure. Health spending is typically defined as a percentage of GDP or, in absolute terms, as the total amount spent on healthcare within a given economy. Government spending on health is the money the government allocates to fund healthcare facilities and services (Babatunde, 2018). Household spending on health is the expenditures made by people and families to cover health services, including those not covered by insurance.

3.3 Estimation strategies

To ensure econometrically valid and robust inference, this study adopts a comprehensive multi-stage estimation strategy explicitly designed for heterogeneous, cross-sectionally dependent panel data. The empirical framework sequentially examines slope heterogeneity, cross-sectional dependence, unit roots, and cointegration, and then proceeds to dynamic long-run estimation using the Dynamic Common Correlated Effects (DCCE) approach.

To assess whether slope coefficients differ across cross-sectional units, the study applies the slope heterogeneity (SH) test proposed by Bersvendsen and Ditzen (2021). This test evaluates whether individual-specific slope estimates significantly deviate from the pooled estimator.

Consider the baseline panel regression:

The null hypothesis $H_0:{\beta }_i=\beta \forall i$ implies slope homogeneity, whereas rejection indicates heterogeneous slope behaviour across cross-sections.

1. Cross-Sectional Dependence Test

Given the strong likelihood of interdependence among economic units, cross-sectional dependence (CD) is tested using the statistic developed by Juodis and Reese (2022), which generalizes Pesaran’s CD test.

Let ${\widehat{u}}_{\mathit{it}}$ denote residuals obtained from the panel regression. Pairwise correlation coefficients are computed as:

The CD test statistic is given by:

Under the null hypothesis $H_0:{\widehat{\rho }}_{\mathit{ij}}=0$ , the statistic converges to a standard normal distribution. Rejection implies the presence of cross-sectional dependence, necessitating estimators robust to standard shocks.

To determine the order of integration, the study employs the panel unit root test of Herwartz and Siedenburg (2008), which accommodates heteroskedasticity, cross-sectional dependence, and structural instability.

The augmented regression is specified as:

Rejection of $H_0$ indicates that the series is stationary.

Long-run equilibrium relationships are examined using the Westerlund and Edgerton (2008) cointegration test, which is robust to cross-sectional dependence and structural breaks.

The test is based on the Durbin–Hausman principle applied to the estimated error-correction term:

The group-mean test statistic is defined as:

After establishing cointegration, long-run coefficients are estimated using the Dynamic Common Correlated Effects (DCCE) estimator proposed by Chudik and Pesaran (2015).

The cross-sectionally augmented distributed lag (CS-ARDL) specification is:

The long-run coefficient vector is recovered as:

2. Endogeneity Correction via DCCE-IV

To mitigate endogeneity arising from lagged dependent variables, the study employs an instrumental-variable-adjusted DCCE estimator following Ditzen (2018). Historical lags of regressors are used as instruments:

This approach yields consistent and efficient estimates even in the presence of feedback effects and weak exogeneity.

4.1 Pre – estimation assessment

This section investigates cross-sectional dependency and the slope of the heterogeneity test, following Juodis and Reese (2022) and Bersvendsen and Ditzen (2021). Table 1 displays the results with two-panel outputs for the CD test and the SH test, respectively. According to the test statistics, all variables are cross-sectionally dependent and exhibit heterogeneity in their properties.

The study performed a panel unit root test and a cointegration test with a structural break, following Helmut Herwartz, Maxand, Raters, and Walle (2018) and Westerlund and Edgerton (2008). The output of the panel unit root and cointegration test is reported in Table 2. From the results, it is apparent that all variables become stationary after the first difference. Furthermore, the panel cointegration test established a long-run association in the empirical relations.

4.2 Empirical model estimation with DCE and DCE-IV

The coefficient, see Table 4, for industrialization was positive and statistically significant at the 1% level, suggesting that industrialization may increase health care costs. Precisely, a 10% increase in industrial output will result in a 1.826% increase in HE cost with DCE and a 1.729% increase with DCE-IV estimation. Our findings are consistent with the existing literature (Rastegar, 2004; Shen, Wang, & Shen, 2021; Zhou et al., 2020). Effects of industrial processes on health may be explained by their impact on the environment and health, such as air pollution. Most of the time, pollution is associated with high healthcare costs, as it requires greater spending. Policymakers should thus consider the impacts of industrialisation on healthcare budget allocation. In doing so, they will be able to minimise adverse effects on the population’s health and ensure that the healthcare industry is ready to address increased demand for services. FDI also has a positive impact on healthcare costs, with the effect statistically significant at the 1% level. An 10% increase in FDI results in a nearly 1.509% increase in healthcare spending when employing the DCE technique and a 1.735% increase using the DCE-IV associated technique. This could be because FDI drives economic growth and increases the demand for healthcare services, and this also brings new medical technologies and practices that initially increase the cost high. The results are consistent with those of Alziyani & Murad (2021), Ehsani et al. (2023), Van Tran et al. (2024), and Zekarias (2016). The two variables, renewable energy consumption and healthcare costs, are statistically significant at the 1 percent level. There is a 10% increase in renewable energy consumption, resulting in a 1.723% reduction in healthcare spending in the DCE model and a 1.003% reduction in the DCE-IV model. According to studies, increased renewable energy will reduce healthcare costs, likely by decreasing pollution and its associated health issues.

Healthcare expenditures show positive correlations with the CO2 emission coefficients in both models, and these correlations are significant at the 1% level. This means that spending in the healthcare sector will increase as the environment continues to be degraded. Namely, a change of 10 0.795 0.5 stayed costs by a significant margin of 6.691, a 10 0.5 rate of increase in CO 2 emissions, increases healthcare expenses by 0.795 0.5 ratio and 6.691 0.5 ratio respectively, with respect to the DCE and DCE-IV techniques. These conclusions highlight that the effects of pollution on health services and costs are significant, and these arguments effectively support the implementation of policies to reduce emissions (Dritsaki and Dritsaki, 2023; Hamid and Wibowo, 2023; Ullah et al., 2020).

Financial development positively affects healthcare costs, with the effect statistically significant at the 1% level. A 10% increase in financial development leads to a 0.995% increase in healthcare expenditures with the DCE method and a 1.598% increase with the DCE-IV method. The study by Chireshe & Ocran (2020) and Xu & Tan (2020) contended that financial development improves access to healthcare services, leading to higher utilisation and, consequently, higher costs. Trade openness is positively associated with healthcare costs, and the association is statistically significant at the 1% level. A 10% increase in trade openness results in a 1.575% increase in healthcare expenditures with the DCE method and a 1.683% increase with the DCE-IV method. Trade openness can drive economic growth and increase healthcare demand, but it may also introduce health risks through the exchange of goods and services that require greater healthcare spending.

4.3 Coefficients estimation with NARDL

Among the high-income group, the long-run asymmetric coefficients (see Table 5, Panel A) indicate that health expenditure is positively associated with CO2 emissions (both positive and negative), with coefficients of 0.1212 and 0.1325, respectively, suggesting that higher CO2 emissions are associated with greater health expenditure. Coefficients of -0.1238 and -0.0999, respectively, for positive and negative changes in renewable energy consumption (REC), imply a negative correlation between changes in REC and health care spending. Health spending is positively affected by changes in foreign direct investment (FDI) (0.1269), though negative FDI has a minor positive effect (0.0971). There is a complicated link between industrial development (IND) and trade openness (TO). IND changes, whether good or negative, increase health spending, with the negative impact being more pronounced (0.1333).

Nevertheless, trade openness shows a negative coefficient for positive changes (-0.0991) and a substantial drop for negative changes (-0.1158), suggesting that health spending is reduced in the long term by more open trade policies. However, the trend is different for the low-income group. Health spending is also substantially increased by positive CO2 emissions (0.1006), but at a slower pace than the high-income group; on the other hand, harmful CO2 emissions have a somewhat more significant effect (0.1165). Health spending is substantially increased (0.1159) by positive REC changes but only slightly (0.0983) by negative REC changes. Foreign direct investment (FDI) has a favorable influence on both positive and negative changes in the low-income group. However, the adverse effect is much higher (0.1737) than in the high-income group. Similarly, health spending is favorably affected by industrial growth, with a higher positive effect (0.0999) than a negative one (0.0849). Not only does trade openness have a positive effect on health expenditure (0.0722) in the low-income group, but it also has an adverse effect (0.1053). This shows that low-income groups are more affected by trade dynamics.

Panel B shows that extra insights are revealed in the near term by the asymmetric consequences. Positive CO2 emissions significantly raise health spending in high-income nations (0.0394), whereas harmful CO2 emissions significantly lower it (-0.0474). Positive CO2 emissions have little effect on low-income nations (0.003), but harmful emissions considerably raise health care costs (0.0441). Both positive and negative changes in renewable energy consumption continue to have favourable consequences; however, low-income groups are hit more by negative changes (0.0261) than high-income groups. Foreign direct investment (FDI) has a favourable effect regardless of the direction of the flow and the recipient’s income level. However, low-income recipients are particularly walloped by negative FDI changes (0.0321). Health spending is positively affected by industrial growth across all income levels, with low-income nations being more sensitive to its adverse effects (0.0513). In high-income nations, trade openness raises health spending by 0.047, whereas in low-income countries it has a negligible, negative effect of -0.0103. In conclusion, nations with lower incomes react more strongly (0.0523) to financial progress, although this effect is statistically significant across both categories. From the adjustment speed to equilibrium (cointEq (-1)), when health spending deviates from its long-run path, low-income countries restore long-term equilibrium more quickly (-0.3042) than high-income countries (-0.4161).

We further test the duality of causality by conducting directional causality tests in Model 1 ( Table 3) and find diverse relationships among these variables. The relationship between health expenditure (HE) and carbon dioxide emissions (CO2) is also bidirectional, as shown by a critical W-Stat and Zbar-Stat at both HE <==> CO2. That is to say, the variations in health expenditure could anticipate CO2 emissions and vice versa. Also, there is bidirectional causality between HE and IND; CO2 and TO indicate each other’s predictive influence. It is also noted that throughout the countries under test. At the same time, there exists a unidirectional causal configuration from CO2 to HE via REC and FDI, and neither direction can predict any of those variables.

Model 2 showed causal relationships between the variables as well. Some of them were significantly higher or lower compared to Model 1. Obviously, the two-way causality between HE and CO2 is essential for retaining and suggesting an intimate connection between health expenditure and carbon dioxide emissions (for instance, []). Furthermore, precipitation is bidirectionally causal with HE and IND, and CO2 is bidirectionally causal with FD, allowing them to predict each other. REC to HE unidirectional causality is established, indicating that health expenditure can be predicted by renewable energy consumption moving from the past into the future, while not evolving vice versa. A more complex interplay between these variables is suggested by the current model, which shows that CO2 Granger-causes REC and that REC Granger-causes CO2. Our results also provide evidence for the sparsely linked causal network among health expenditure, CO2 emissions, renewable energy consumption, and some economic dimensions.

4.4 Robustness assessment

The following section deals with the assessment of the robustness of the estimation through different techniques, such as FGLS, PCSE, and FMOLS, as well as endogeneity assessment with IV estimation. The output reported in Table 6 showed a similar vine of association, as observed in the earlier estimation. Thus, it assesses the robustness and internal consistency of the empirical model.

The endogeneity problem in the study has also been addressed using IV techniques, as shown in Table 7. An endogeneity problem is absent.