1 Introduction

Environmental issues and corresponding climate difficulties have been revealing their effects on a global scale in ever-greater depth. These issues are on track to supplant all others as the most pressing and significant issues of today's century. The task is so great that it has already united the globe around a common denominator of ever-present concern. This denominator comprises numerous activities, including social, demographic, transportation, housing, tourism, industrialization, and international trade. Concerning this, a more comprehensive understanding of the underlying, pivotal components of environmental contamination is now paramount. More specifically, to put solution-oriented policies into practice, demystifying the mechanisms of both reduction of emissions and energy efficiency from a deep perspective in reaching climate neutrality is now the common mission.

Unfortunately, the devastating effects have become even more evident in the last two decades as the globe approaches the 2030 and 2040 targets. Significantly, human-and economically-induced climate-breaking activities speed this markedness up, rendering it the primary focus. Therefore, investigations examining the relationship between human-induced factors such as energy use, urbanization, the growth of population, and energy consumption (ENC) (Adebayo et al., 2022; Anwar et al., 2021; Awan et al., 2022; Ayla & Altıntaş, 2021; Sulaiman & Abdul-Rahim, 2018) and also economic factors consisting of financial development, income, economic growth (ECG), tourism, public expenditures, foreign direct investment, and international trade with environmental pollution (Basoglu & Uzar, 2019; Işik et al., 2016; Kongkuah et al., 2021; Pata et al., 2022; Sharif et al., 2017; Yurtkuran, 2020a, 2020b).

One of the different vital drivers not cited above is an important social institution, the military. Military expenditures (EXM) are a multifaceted component with sociological, political, economic, and environmental outcomes. For this reason, EXM has mostly been the been the subject of the literature due to its different functional interactions, such as with corruption (Ali & Solarin, 2020; Anifowose, 2019; Gupta et al., 2001), income equality (Elveren, 2012; Gül & Torusdag, 2019; Michael & Stelios, 2020; Sharif & Afshan, 2018), employment and unemployment (Canbay & Mercan, 2020; Korkmaz, 2015; Sanso-Navarro & Vera-Cabello, 2015; Zhong et al., 2015), and ECG (Ayla, 2020; Farzanegan, 2014; İpek, 2014; Gokmenoglu et al., 2021; Khalid & Habimana, 2021; Korkmaz & Bilgin, 2017; Manamperi, 2016; Naımoglu & Ozbek, 2022; Phiri, 2019; Ward & Davis, 1992; Zortuk & Karacan, 2019). Most studies have focused on the results of factors other than environmental effects. There is a dearth of evidence on how militarization affects the environment (Chang et al., 2023; Sparrevik & Utstøl, 2020).

As a gauge of international power, expenditures on the military are significant, especially for leading countries. EXM has many functions in terms of triggering environmental damage. First, military operations rely heavily on using fossil fuels as a main source. The harmful tendency of growing consumption of fossil fuels in the military is one of the most effective vitiations of the environment. The period following the Second World War in particular led countries to build a more military-oriented structure. An intensifying global arms race goes hand in hand with militarization, which has turned into a field of aggressive competition between nations. The military movements and products, which have become more technologically complex with the modification of the military industry, have consumed more fossil fuels (Gokmenoglu et al., 2021; Jorgenson et al., 2010). Military equipment such as helicopters, planes, submarines, tanks, warships, vehicles, and others uses sheer amounts of fossil fuels. To keep up processes such as the testing of machines and weapons and the transport of weapons, military equipment, research and development functions, the movement of personnel, land transport, and other military resources, fossil fuels are the most preferred combustible by the sector (Bargaoui & Nouri, 2017; Bildirici, 2017; Clark et al., 2010; Solarin et al., 2018; Ullah et al., 2021). Also, the globe's growing militarization race causes a surge in ENC and, ultimately, CO2 levels (Bildirici, 2017).

However, in the second framework, military production necessitates extracting and processing highly harmful materials, posing the most significant jeopardy to human health and the environment (Gould, 2007). Similarly, military activities that produce different types of toxic waste hurt the environment by destroying natural plants and animals, cutting down trees, and leaving radioactive materials in the biosphere (Singer and Keating, 1999). This process involving the use of chemicals, from the maintenance of military equipment to testing, from training to production, further confirms that the military is one of the largest users of natural resources. Moreover, all military activities, which demand the benefit of an enormous land area, have a multidimensional potential for pollution in the environment's most biologically effective production areas, such as water, air, and soil. Every function of militarization, including extraction, research and development, operation and maintenance, production, distribution, testing, transportation, disposal, setting up, and building, clearly has the prospect of disrupting ecosystems on multiple levels. Each scenario above clearly illustrates the entire ecological destruction process of military expenditures.

EXM is among the most influential variables that cause environmental degradation via human endeavours (Erdoğan et al., 2022; Eregha et al., 2022; Konuk et al., 2023). It is considered one of the most prominent environmentally detrimental human movements (Gould, 2007). The starting point for debating the environmental consequences of EXM depends on the "treadmill of destruction theory" put forth by Hooks and Smith (2005). However, the relationship between EXM and the environment, which has yet to be exposed correctly in terms of its peripheral dimensions, is home to exploring a desired connection in the literature. Although the number of studies investigating the connection is growing, it remains modest (Bildirici, 2017; Konuk et al., 2023; Mughal et al., 2022; Ullah et al., 2021).

So, this gap in the literature is full of opportunities to explore this relationship. This gap is fascinating to investigate, especially for the US, the world leader in the defense field. The US spends heavily on defense. The US's military expenditures are by far higher than those of other nations, and this difference is growing all the time. The country's defense spending is nearing $1 trillion. Worthy of note is that the cycle of the Vietnam War, the Cold War, the Iraq War, and the Afghanistan War accelerates militarization in the US. This global war cycle in the US also supports an upward spending trend for the defense industry. Analogously, with a rise of $71 billion from 2021 to 2022, the US expends more on its military than France, Ukraine, Germany, India, Japan, Russia, Saudi Arabia, China, South Korea, and the United Kingdom, combined (SIPRI, 2023).

The concern of ensuring the country’s security outweighs the concern of protecting the environment, making it very sophisticated to integrate an environmental management system into a wide range of defense activities (Chang et al., 2023; Ramos & De Melo, 2005). However, compared to others, the US has already taken steps to leave this argument behind. As the world leader in the development of the US military defense industry, it is steadfast in incorporating environmentally friendly technologies into all stages of the overall military industry process. While it was emphasized that the US was the world’s chief producer of biological, chemical, and nuclear weapons in the 2000s (Gould, 2007), nowadays, the country supports the use and production of more environmentally friendly military equipment. The US's defense cares about the environment, consequential that it can be said that there is a trade-off between spending more on the military and shielding the environment for the US.

EXM affecting ecology positively or negatively is a curious topic in the building consensus debates. Being the top spender in the military is the fundamental motivation for preferring the US for the present paper. Therefore, the country is an excellent sample for examining the relationship between EXM and the environment. The limited number of studies in the literature dealing with this sample also strongly reinforces this assertion (Ben Youssef, 2023; Bildirici, 2017; Solarin et al., 2018). In this vein, using data from 1970 to 2018 for the US, this paper aspires to explore the relationship between EXM and the environment.

Considering all of these perspectives, this study intends to make a significant contribution with the following motivations for the growing literature: (a) This study explores the literature and studies on US military expenditure, focusing on the country's critical role in the global field. (b) Contrary to studies that used CO2 (Ahmed et al., 2020a, 2020b; Bildirici, 2017; Erdogan et al., 2022; Jorgenson & Clark, 2016), the study prefers to use ecological footprints (E) (Chang et al., 2023; Eregha et al., 2022), a broader environmental measure to catch the multifaceted effects generated by militarization on environmental degradation in the best way. (c) Thirdly, the study allows, to a large extent, both war and non-war periods to be considered together, considering the extensive time period of 1970–2018. (d) Of course, analyzing an extensive time period can bring along many structural changes. So, advanced econometric methods like Maki-cointegration test (MCT) and CS unit root test (URT-CS), both of which allow multiple structural breaks, are utilized in analysis for more rational results. (e) Lastly, in the analysis, unlike the traditional causality analysis tests, the Time-varying causality (TVC) test, which argues that causality will not be constant over time, pioneered the determination of causality relationships between variables. Therefore, the current paper focuses on exploring the long-run relationship between the E, EXM, exports, imports, urbanization, and agricultural area for the US over the period 1970–2018 by using one of the next-generation cointegration tests, MCT.

The leftover section of the research is scheduled as follows: these four main headings, respectively: literature, methodology, and data; results section; and lastly, the conclusion part.

2 Literature review

During the production process, both energy and natural resources are utilized, resulting in the release of atmospheric pollutants that bring about climate change (Ozdemir, 2021). The devastating effects of climate change on the environment have become more evident with increasing droughts, floods, severe hurricanes, and melting glaciers. Therefore, in recent years, researchers (Khan et al., 2018b; Kim et al., 2019) have focused on the variables that affect environmental performance in developed as well as developing nations. The impacts of economic activity on the environment have been a well-known topic for research since the 1960s, and many researchers have looked into them, particularly since the 1990s.

Despite the growing body regarding research on environmental economics, the relationship between EXM and environmental damage initially got less focus. On the other hand, the ecological literature has recently started researching EXM as a factor influencing environmental quality. Hooks and Smith (2005) demonstrate how operations and EXM cause environmental change employing the Treadmill of Destruction Theory. This theory highlights that military activities result in increased ENC and harmful waste, which harm the environment. Among the researchers who examine the relationship between military actions and the environment in the context of the Treadmill of Destruction theory are Jorgenson and Clark (2009), Jorgenson et al. (2010), Clark et al. (2010), and Givens (2014). According to these studies, EXM harms environmental quality. In a related study, Bildirici (2017) investigates the causal and long-run relationship among CO2, EXM, ECG, and ENC in the US. The research findings indicate a positive correlation between EXM and CO2. Results from the MWALD technique point to evidence of a BC relationship between real income and EXM, real income and ENC, real income and CO2, and ENC and EXM, as well as a relationship between CO2 and EXM. Bildirici (2018) reports in another study that military expenditures cause environmental degradation by increasing greenhouse gas emissions in G7 countries. She also figured out a bidirectional causality bidirectional causality (BC) between EXM and GHG emissions. Bradford and Stoner (2017) report a fragile and varying relationship between EXM and CO2 in 62 countries. They notice that the economic development levels of countries mitigate the impact of EXM on CO2, with more developed economies' military spending having significantly greater net effects on emissions.

Solarin et al. (2018) examine the impact of EXM as well as ECG, population, ENC, renewable and non-renewable ENC, urbanization, trade openness, and financial development on CO2 in the US by using the STIRPAT model. They suggest the varying effects of EXM on environmental contamination. Based on this result, they conclude that EXM can cause pollution due to the enormous amount of fossil fuels used in the country’s military sector. On the other hand, according to the results of research and development in the military sector, EXM can reduce emissions. Further, the results reveal that ECG, ENC, non-renewable ENC, population growth, and urbanization also raise CO2 in the long term, whereas financial development, renewable ENC, and trade openness lower emissions. Gokmenoglu et al. (2021) examine the relationship between EXM, financial development, ENC, ECG, and environmental depletion in Turkey. Environmental degradation is a representation of E and CO2. The estimation results posit that EXM, energy use, and ECG significantly hasten environmental depletion, whereas financial development improves the environment. The results of causality test reveal a BC between EXM and ECG as well as a unidirectional causality (UC) from EXM to CO2, and E. Ahmed et al. (2020a) investigate Pakistan's ECG, EXM, and E. The research outcome implies the adverse impact of EXM on the environment by raising E.

The results of Hacker and Hatemi-J’s (2012) bootstrap causality test show UC from EXM to ECG, but UC from EXM to the E. According to Qayyum et al. (2021), EXM significantly raises E in the long and short run for four South Asian countries, including Pakistan. Additionally, the Granger causality test results show a one-way causal relationship between militarization, armed conflict, and E. They suggest that in South Asia, the conflict between the two dominant nations in the region, Pakistan and India, has resulted in a steady rise in EXM, which will exacerbate the region's already unfavorable climatic conditions and cause environmental degradation. However, Ullah et al. (2021) indicate that expenditures on the military boost the health of the environment by diminishing CO2 in Pakistan and India. Isiksal (2021) also reports that in the top 10 nations having the greatest EXM, EXM contributes to environmental depletion, whereas renewable energy increases environmental quality. Isiksal (2021) argues that armed forces continue to burn fossil fuels even in times of peace, which causes CO2 to build up in the atmosphere and produce harmful pollutants.

In Nigeria, Eregha et al., (2022) demonstrate that EXM boosts E, accelerating the decline of the environment. The causality findings reveal a UC from EXM to E but a BC between EXM and ECG. They claim that Nigeria is an ecologically challenged country. Therefore, with rising EXM and E, the country's biocapacity continues to decline. Furthermore, Kwakwa (2022) indicates that Ghana's increasing EXM is contributing to higher CO2. In another very recent study, Ahmed et al. (2022) report that EXM has contributed positively to CO2, indicating that EXM has deteriorated the environment in OECD nations. Mughal et al. (2022) reveal evidence supporting the positive impact of EXM on E in the NEXT-11 economies, implying that EXM increases environmental deterioration.

On the other hand, Konuk et al. (2023) report for G7 countries that rising EXM reduces CO2. They argue that, particularly in the US, greater funding for research and development efforts and incentives for using environmentally friendly military equipment have helped mitigate the detrimental effects of EXM on the environment. Chang et al. (2023) indicate that increased EXM worsened environmental harm by raising E in 15 RCEP economies. Pata et al. (2023) examine EXM, ECG, CO2, and renewable energy use in 15 NATO countries. The empirical results in their study are consistent with the results of most other studies, which indicate that EXM contributes to growing CO2.

Muhammad et al. (2023) suggest that EXM is increasing environmental emissions in the short term. However, Ben Youssef (2023) investigates the relationship among renewable ENC, EXM, net energy imports, arms exports, gross domestic product, and CO2 in the US. The estimation results reveal a long-term positive impact of EXM on renewable ENC but a long-term negative impact on net energy imports and CO2. He claims that the military sector in the US is helping promote the use of renewable energy and combat the effects of climate change.

Agriculture is a crucial factor in the process of degrading the environment due to rising energy usage, land use, and CO2 (Dogan, 2016). Agriculture produces around 20% of global CO2 (FAO, 2020), significantly contributing to environmental degradation. Although many scholars have extensively researched the relationship between environmental performance and various indicators (e.g., ECG, urbanization, ENC, and industrialization, among others), recent research has examined the role of agriculture on environmental contamination, which was one of the factors overlooked in previous literature. Dogan (2016) examines the impact of agriculture on Turkey's CO2 within the context of the EKC hypothesis and finds the presence of EKC. The empirical results also reveal a significant mitigating effect of agriculture on CO2. Similarly, Rafiq et al. (2016) reveal that agriculture boosts environmental quality by lowering CO2 in 53 countries.

Sarkodie and Owusu (2016) suggest that Ghana's agriculture is a significant contributor to increased CO2. In another single-country study on Tunisia, agriculture contributes to higher CO2, according to Ben Jebli and Ben Youssef (2017a). Conversely, the authors (2017b) demonstrate that in five North African countries, agriculture enhances environmental quality. Liu et al. (2017a) suggest that agriculture in Malaysia, Indonesia, Thailand, and the Philippines effectively decreases CO2. Liu et al. (2017b) reveal that agriculture pollutes the environment in the BRICS countries. Doğan (2018) demonstrates that agriculture increases environmental pollution levels. According to Gökmenolu and Taşpnar (2018) and Ullah et al. (2018), agriculture in Pakistan has a BC with CO2, and it lowers the welfare of the environment. Waheed et al. (2018) also report that agriculture contributes to the worsening of the environment in Pakistan.

In conflict with the results of these studies, Khan et al. (2018a) note that agriculture reduces greenhouse gas emissions in Pakistan. According to Agboola and Bekun (2019), agriculture in Nigeria has a marginally beneficial impact on CO2. Ben Jebli and Ben Youssef (2019) claim that there exists a UC relationship between agriculture and CO2 in Brazil and that agriculture minimizes damage to the environment.

Conversely, Gokmenoglu et al. (2019) find that agriculture generates degradation of the environment and a UC from agriculture to CO2. Balsalobre-Lorente et al. (2019) assert that agriculture produces environmental damage, and that there exists UC from agriculture to CO2 for BRICS countries. Olanipekun et al. (2019), Qiao et al. (2019), and Aydogan and Vardar (2020) indicate that agriculture contributes to environmental degradation in 11 Central and West African regions, 19 G20 countries, and 7 emerging countries, while Sarkodie et al. (2019) determine that agriculture lowers CO2, which lessens environmental devastation in 14 African countries. Prastiyo and Hardyastuti (2020) suggest that Indonesian agriculture significantly reduces CO2, with a BC between agriculture and CO2. However, Udemba (2020) claims that agriculture harms the environment by increasing E in India. Furthermore, empirical results suggest UC from agriculture to E.

According to Pata (2021), agriculture has a positive but statistically insignificant contribution to environmental contamination in China and Brazil. Additionally, the findings show that there are BC between agriculture and CO2 in Russia as well as between agriculture and E in India and China. Salari et al.'s 2021 study revealed that agriculture at the 25th and 50th quantile levels significantly improved E in 21 emerging countries. As a result, they argue that agriculture generates environmental damage. Usman et al. (2021) investigate how agriculture in BRICS-T strengthens the potential for the region's E to increase, showing an additional 1% in agricultural boosts the region's E level by 0.2201%.

Furthermore, the causality analysis reveals the persistence of the feedback effect between agriculture and E. According to Muoneke et al. (2022), agriculture, despite its nonlinear structure, benefits the environment both in the short and long terms, but it also contributes to environmental depletion through economic practices like livestock raising and growing crops, as well as integrated activities like using fossil fuels for agribusiness-related transportation, making fertilizer, polluting the water supply, using pesticides, destroying forests, and making biofuels.

According to Balogh and Jámbor (2017), foreign trade (EI), is associated with increased pollution through transboundary pollution transfer and the migration of the manufacturing sector. The present situation demonstrates that pollution happens during the production of items for EI and their usage in different countries (Karış & Kaya, 2021). Many empirical studies have replied the relationship of EI with environmental depletion. Based on a country's features, overseas trade may positively or negatively affect environmental pollution. While some studies find that EI harms environmental quality (Chebbi et al., 2011; Farhani & Ozturk, 2015; Halicioglu, 2009; Nathaniel, 2021; Nathaniel et al., 2021; Omri et al., 2015; Sabir & Gorus, 2019; Shahbaz et al., 2017; Weber et al., 2008; Yan & Yang, 2010). Others (Al Mamun et al., 2014; Dogan & Seker, 2016; Essandoh et al., 2020; Kim et al., 2019; Usman et al., 2021; Zhang et al., 2017) have noted that EI improves environmental quality. Using modern panel data estimators such as CUP-FM and CUP-BC, Ahmed et al. (2020b) discover that in G7 countries, urbanization boosts the E, imports increase environmental degradation, and exports decrease environmental degradation. Nathaniel et al. (2020) also reveal that urbanization significantly exacerbates environmental degradation. Conversely, EI improves environmental quality.

As the adverse effects of pollution emissions and E induced by environmental degradation on wildlife, human health, and economic sustainability grow, it is critical to identify the contributors shaping environmental depletion gauges. EXM, agriculture, and EI are three major contributors to environmental degradation. Most of the research focuses on CO2 as the most often used gauge for measuring environmental depletion, with less emphasis on E. CO2 does not indicate water, soil, or other environmental problems, but E offers more detailed information regarding environmental challenges (Pata, 2021). The E comprises six components: carbon footprints, developed land, grazing land, farmland, forest land, and fishing grounds. Because the E considers multiple resource stocks, its research will be more potent in modelling policy measures for the sustainability of the economy, considering environmental quality, and controlling environmental degeneration (Udemba, 2020). Therefore, this examination prefers the ride of the E as a wide gauge of environmental degradation.

According to Brown University (2019), vehicles used in war zones emit hundreds of thousands of metric tonnes of nitrogen oxides, carbon monoxide, hydrocarbons, and sulphur dioxide in addition to CO2. Atmospheric pollution from military weapons and vehicles adversely affects public health among civilians and US soldiers in war zones. Heavy military vehicles generate more dust than usual, exposing soldiers to toxins breathed in from this dust, resulting in respiratory system disorders. Depleted uranium in oil and ammunition in military vehicles contaminates water supplies in war zones. Animal and bird populations have been adversely affected by the degradation of forests and the consumption of natural resources.

Research on the drivers of environmental degradation has recently advanced as a result of recent advances in econometrics, with many studies adopting new variables and focusing on different countries. Despite extensive research on environmental factors across various regions, countries, and groups, scholars remain divided on the subject. The lack of consensus motivates additional research to solve the puzzle. Therefore, this type of research is critical. This study aims to investigate the relationship between EXM, EI, agriculture, and E in the US for the period 1970–2018. Ultimately, the research will help advance and fill the gap in the body of literature on this subject. The present research provides five unique contributions to the current literature: First and foremost, while the present study is one of the first to empirically examine the relationship between EXM, EI, agriculture, and E, particularly for the US sample, it also aims to promote the current literature. Second, the time period of this study for the US is more extended. Thirdly, MCT, which enables five breaks from the next generation tests, was employed in the study on the relationship between EXM, E, and EI because the time period is broad. The reliability of the analysis's findings rises as a result of the adoption of this method, which accounts for breaks. Fourth, TVC approach is used instead of traditional causality tests to analyze relationships between series, examining different causality relationships in different periods. This method's strengths include focusing on the intertemporal changes of the causal relationship between variables and providing data regarding the stability of these relationships. Finally, advanced econometric techniques such as the URT-CS and MCT were used to achieve robust estimates (Fig. 1).

Fig. 1
figure 1

The comprehensive overview of the critical findings of the literature

Source: Created by the authors 

Full size image

3 Econometric method and methodology

The relationship between EXM, EI, and E variables is investigated for the US economy in the study's analytical section. Since EI consists of the amount of exports and imports, the analyzes are evaluated within the scope of two models: the first model is based on exports, and the second model is based on import variables. In addition, it would be more useful to examine the environmental impact of exports and imports independently using two different models.

Time series approach including structural breaks is employed for annual data from 1970 to 2018 in the paper. The study's main hypothesis is that there is a long-run relationship among EXM, EI, and E with structural breaks. Such a case, the research methodology is ascertained once the variable set and model of the variables used within the framework of the hypothesis are supplied. After the theoretical framework, the outcomes of the tests are delivered.

3.1 Data set and model

  • Based on the common data constraints among the model's series, annual data of the US for the 1970–2018 periods were taken as the basis for the analysis. The main motives for choosing the US can be summarized as follows:

  • Being the country in the world that allocates the highest share of the budget to EXM within the scope of security.

  • Ranking first in the world based on EI volume.

  • When considered in terms of national income, it can be expressed as a country where environmental impacts are more intense due to its production and consumption dimensions and its size in the world economy.

In short, it is crucial to choose the US as the sample country because it is one of the biggest players in international trade and because the policies it implements will be crucial in addressing the environmental issues brought about by global production.

The hypothesis is tested through two models. The variables are determined in line with the existing studies. E, the most inclusive indicator variable, by including many environmental factors, constitutes the dependent variable of both models. As independent variables, exports of goods and services (EGS) (% of GDP) and imports of goods and services (IGS) (% of GDP) which are the two magnitudes of EI, are included in the model. Additionally, as a security variable, expenditure on military for US (EXM) (% of GDP) and the ratio of agricultural land to total land area (AL), which is shown as an environmental variable and known to affect the E, are enclosed in the model. Finally, scrutinized model incorporates the annual growth of urban population (UPG) as a control variable, which is known to influence the E. The source for E is the Global Footprint Network, and for the others (EGS, IGS, EXM, AL, UPG) is World Development Indicators.

The study investigates the impact of EI and EXM on ecosystem deterioration because of recent natural destruction, to reveal policy implications for the US economy, which has the largest power in the world economy. For this purpose, the models established to determine the effect of EXM and EI on E are shown in Eqs. 1 and 2.

$${E}_{t}={\beta }_{0}+{\beta }_{1}{EGS}_{t}+{\beta }_{2}{EXM}_{t}+{\beta }_{3}{AL}_{t}+{\beta }_{4}{UPG}_{t}+{\varepsilon }_{t}$$
(1)
$${E}_{t}={\beta }_{0}+{\beta }_{1}{IGS}_{t}+{\beta }_{2}{EXM}_{t}+{\beta }_{3}{AL}_{t }+{\beta }_{4}{UPG}_{t}+{\varepsilon }_{t}$$
(2)

In the models, ɛ gives the error term.

3.2 Method

The methodological flow shown in Fig. 2 is heeded to investigate the relationship between EXM, EI, and E variables in the US economy.

Fig. 2
figure 2

Sequency of methodology

Full size image

3.2.1 Descriptive statistics

While examining the relationship between E and EXM, EGS, IGS, UPG, AL, the normal distribution of the series included in the model is interpreted according to the statistics of kurtosis, skewness, and the Jarque-Berra test.

According to Table 1, since all statistics are less than 3, the series are kurtotic. In terms of skewness values, E, AL, and IGS variables are negatively skewed (right) since they are less than zero, while EGS, EXM, and UPG variables are positively skewed (left).

Table 1 Descriptive statistics for the variables
Full size table

3.2.2 Unit root test (URT)

The URT involved in the analysis are the Augmented Dickey–Fuller (ADF) unit root test (URT-ADF), the Kwiatkowski, Phillips, Schmidt, and Shin (KPSS) unit root test (URT-KPSS), and the Carrion-i Silvestre unit root test (URT-CS), which allows capturing five breaks.

Table 2, which presents the URT-ADF results, shows that all variables are unit rooted at their level values. All variables must be stationary at the same level in order to conduct a cointegration analysis of the series. The EXM becomes stationary at the 5% significance level, and the others become stationary at the 1% significance level at the first difference I(1). In the URT-KPSS, it is concluded that E and EXM are stationary at the 5% significance level and the UPG variable is stationary at the 1% significance level, while the others are stationary at the first difference I(1). When the variables are taken the difference, it is understood that all variables become stationary at the 1% significance level. It is observed that the stationarity of the variables (E, EXM, and UPG), which are stationary at their level values, strengthens when they are taken a difference. The essential level of I(1) is ensured for the cointegration test.

Table 2 Analysis of URT-ADF and URT-KPSS
Full size table

Based on the outcomes from the URT-CS presented in Table 3, it is evident that all variables exhibit unit roots at their level values. In order to make the series stationary, the method of taking a difference was not applied in here as in classical unit root tests (URT-ADF, URT-KPSS). In these tests, the series appear stationary at level, but their significance is heightened when differenced. Since the URT-CS consider structural breaks, the difference of the series was not taken, considering that the reason for the non-stationarity in the series was not a structural break. This finding aligns with the prerequisites for the cointegration test, as all variables are classified as I(1) based on the classical URT results, which identified unit roots at their level values.

Table 3 Results of the URT with multiple structural breaks
Full size table

3.2.3 Maki cointegration test (MCT)

The long-run relationship between EXM, EI and E variables is investigated with the MCT. This method is useful for the long-run period of the study as it allows up to five breaks like the new-generation tests. The strengths of this test and the main reasons for its selection can be stated as follows (Maki, 2012: 2013).

  • It is a new-generation test that takes into account structural break dates in economies with significant social, economic, and political fluctuations.

  • The MCT, allowed 5-structural breaks, analyzes the realized breaks endogenously.

  • The test's algorithm works differently than those of other cointegration tests. In the analyses, a possible break point is taken in each period, and the t statistic is calculated first, and the points where the t value is the minimum are ataken as the break focus.

An examination of past studies divulges that cointegration relationships between EXM and different macroeconomic variables have been investigated. On this point, the relationship between ECG and EXM has been investigated by Kollias et al. (2007), Narayan and Singh (2007), and Augier et al. (2017). Hooker and Knetter (1994), Malizard (2014), and Azam et al. (2016) investigate the relationship between EXM and unemployment, while Dunne and Smith (2020) investigate the relationship between EXM and investment spending.

The present research contributes significant additions at both the macro and micro levels to the current body of literature. The main focus of the study was to look at how macro variables, such as EI, EXM, and urbanization, and micro variables, such as agriculture and the environment, affect the changes in global competitiveness. The significance of production costs holds considerable weight in the context of global economies. The implementation of environmental measures significantly affects EI by incurring substantial costs. In order to achieve macroeconomic goals, the budgetary weight of EXM is crucial. As a result, this research is anticipated to create a notable grant to the literature because it examines the relationship between EXM, the environment, and EI variables, as well as global competition with macro and micro variables, using the MCT, which allows for five breaks (Table 4).

Table 4 MCT test results (model 1)
Full size table

According MCT, allowed maximum 5 breaks, results of the model, including the EGS variable, a long-run relationship is found at the 1% level in Model 0 and at the 10% level in Model 2. Therefore, there exists a long-run relationship among the independent variables of EXM, UPG, AL, EGS, and E with structural breaks (Table 4).

Break dates indicate the significant impact of major events or structural changes brought about by crises on variables. The reliability of analysis results that take breaks into account increases. In the results of Model 1, two break dates are determined as 1979 and 2011. When the break dates are taken into account, tensions with Russia led to an augmentation in EXM in the US in 1979 as the effects of the Cold War continued. From another point of view, the rise in oil prices in 1979 also led to economic difficulties. 2011 was a year in which the effects of the 2008–2009 global economic crisis continued and economic stability was affected (Table 5).

Table 5 MCT test results (model 2)
Full size table

According MCT, allowed maximum 5 breaks, results of the model, including the IGS, a long-run relationship is detected in Model 0, Model 2, and Model 3. Therefore, there exists a long-run relationship among the independent variables of EXM, UPG, AL, IGS, and E with structural breaks. The Bd are the late 1970s, early 1980s, early and late 1990s, and 2007. Evaluations of the breaks observed in the variables at these dates reveal that the rise in oil prices and the energy crisis in the US in the late 1970s resulted in the stagflation process. The 1980s were a period of increased EXM due to the Cold War. With the collapse of the Soviet Union, in the 1990s, the US became a world superpower. In addition, this period was a period of advances in information and communication technologies, the reshaping of global foreign trade agreements, and the achievement of ECG and stability. In 2007, a bubble in the real estate sector, widespread financial mismanagement, and the outbreak of the 2008 global financial crisis can be stated as the reasons for the aforementioned breakdowns (Table 5).

3.2.4 Long run analysis: estimation of cointegration coefficients

When the cointegration test affirms a long-run relationship between the variables, long-run and short-run relationships can be estimated using various methods. The FMOLS method, which Phillips and Hansen (1990) introduced to the literature, yields reliable and consistent coefficient estimates for small samples. After solving the endogeneity problem in a time series, it estimates the long-run effect of independent variables on the dependent variable. It is also asymptotically efficient and unbiased (Yurdakul, 2018: 61).

DOLS, another method developed by Stock and Watson, estimates coefficients by taking into account the priors and lags of the first differences of the variables (Al-Azzam & Hawdon, 1999: 7). Estimators are determined using Monte Carlo simulations. DOLS estimators are more efficient for models with a small number of observations and a heterogeneous structure (Pata & Tütüncü, 2017: 43).

The "Canonical Cointegration Regression (CCR)" method developed by Park (1992) eliminates the quadratic deviations obtained from the OLS estimator. It is also based on transformations of the variables in the cointegration regression. This estimator, which is closely related to the FMOLS method and allows for asymptotic chi-square testing, uses the first differences of variables instead of their level values. In this method, similar to the FMOLS estimator, error terms and long-run covariance matrices are obtained. In contrast to FMOLS, CCR also demands a consistent estimator of the simultaneous covariance matrix (Yurtkuran, 2020a, 2020b: 1537).

In the long-run analysis, the break dates encountered in the cointegration test have been taken into account by creating dummy variables. Dummy variables have been created by assigning a value of zero to the years up to the date of the break and one to the other years. Tables 6 and 7 present the results.

Table 6 Long run cointegration coefficients (model 1)
Full size table
Table 7 Long run cointegration coefficients (model 2)
Full size table

The results in Table 6 are analyzed:

  • According to the FMOLS and the CCR estimators, the relationship between EXM of US and E is inverse. This result demonstrates that the US defence sector has contributed to the improvement of environmental quality by pursuing sustainable environmental goals and investing in environmentally friendly military practices.

  • According to the three estimators, there is an inverse relationship between AL and E. Based on these results, it is determined that practices such as sustainable agricultural practices, organic farming, and efficient use of water resources contribute to the improvement of environmental quality in the US.

  • According to the three estimators, there is an inverse relationship between EGS and E. The fact that exported products are not consumed within the country is expected to reduce E since it does not increase environmental waste and natural resource use.

  • There is no relationship between UPG and E, according to the three estimators.

  • The statistically significant results of one of the dummy variables indicate that there have been significant changes in EXM, EI, and E in 2011, as determined by the cointegration method.

  • The results in Table 7 are analyzed;

  • There is an inverse relationship between EXM of US and E, according to the FMOLS and CCR estimators. The increase in the defence budget has supported environmentally friendly technologies in the military field and contributed to the quality of the environment through the use of green technology without leading to the consumption of natural resources.

  • According to three estimators, there is no relationship between AL and E.

  • There is a positive relationship between IGS and E, according to the FMOLS and DOLS estimators. Consumption of imported products within the country is expected to increase E as it increases environmental waste and natural resource utilization.

  • According to the DOLS estimator, there is a positive relationship between urbanization and E. The increase in the urban population is likely to create environmental problems as it leads to an increase in the need for food and housing.

The statistically significant dummy variables indicate that there have been significant changes in EXM, EI, and E in 1984, 1998, and 2007, as determined by the cointegration method.

The results of the long-run coefficients for Model 1 and Model 2 are presented in Figs. 3 and 4.

Fig. 3
figure 3

Long run coefficients for model 1

Full size image
Fig. 4
figure 4

Long run coefficients for model 2

Full size image

3.2.5 Time-varying causality (TVC) test results

The causality test with time-varying parameters introduced by Hacker and Hatemi-J (2006) segments the sample and assesses each segment separately. This approach accounts for the possibility that structural shifts might alter the parameters, thereby affecting the direction of causality over time. Recognizing that the interrelations among variables may evolve throughout the study period, this method deviates from conventional techniques. In a subsequent step, this time-varying bootstrap causality test, which builds on the Toda–Yamamoto (1995) methodology, is employed and its results are analyzed.

The dynamics between variables, as highlighted by Tang (2008) and Arslantürk et al. (2011) are subject to temporal shifts. Notably, political and economic incidents have a significant impact on the causal dynamics. This necessitates an analysis of the variable relationships using a TVC approach, allowing for the exploration of different causality relations across various time frames. In the context of Hacker and Hatemi-J (2006) causality analysis using the Rolling Window technique, the term 'windows' denotes specific sub-periods.

In socio-economic analysis, the causal relationship between variables is usually examined with standard tests. Causality methods in these tests are used to analyze the overall period. In the context of global economies, it is important to acknowledge that the causal relationship between variables can occasionally lose its validity within certain time periods. For the reasons mentioned above, TVC analysis is employed in this study. The primary benefit of TVC tests lies in their ability to track alterations in the causal interactions among variables over time. Additionally, they offer insights into the consistency of these causal links.

In the study analyzing the causality relationship among EXM, EI, and E variables using TVC methods, the blue lines in Fig. 5 represent the periodically calculated test statistic value of the hypotheses, while the red solid line describes the critical value of the test. A causal relationship exists between the variables in the ranges, with the blue line above the red line. The test statistic calculated according to the results of the TVC test is determined by the Hatemi-J information criterion and obtained with 10,000 iterations using bootstrap.

Fig. 5
figure 5

TVC test results

Full size image

A causal relationship exists between E and EXM between the dates indicated in Table 8. The reason for the causality relationship is that increased EXM due to military activities and equipment use during the Cold War period caused environmental impacts, and a causality relationship from EXM to E emerged in the late 1980s and early 1990s. In addition, considering that international activities for environmental awareness have increased awareness of environmental issues with events such as the publication of the Brundtland report in 1987 and the Rio Earth Summit in 1992, the causality relations in question may be the reason why these causal relationships have turned into a mutual relationship after these dates and especially in 2003. The bidirectional relationship reveals that there is a cyclical interaction between EXM and E, with EXM affecting E and E affecting EXM.

Table 8 Direction and dates of causality
Full size table

A causal relationship exists between E and AL between the dates indicated in Table 8. These causal relationships have been observed since 2000s. The reason for the causal relationship is that climate change caused by environmental impacts and the decrease in water resources have had some effects on agricultural activities. Therefore, R&D efforts to improve environmental impacts in agriculture have paved the way for sustainable agriculture. Organic farming practices may have contributed to the determination of the environment-agriculture interaction as of the 2000s.

There is a causal relationship between E and UPG between the dates indicated in Table 8. The causality relationship has been in evidence since the end of the 1980s. The 1980s and 1990s in the US were a period of rapid urbanization. Rapid urbanization has created processes involving environmental problems. Therefore, environmental policies aimed at regulating the relationship between urbanization and the environment may have shaped the emergence of the causal relationship between these variables.

A causal relationship exists between E and EGS between the dates indicated in Table 8. The causality relationship was found in the 1990s and 2000s. The fact that environmental awareness increased gradually in the years in question has paved the way for the formation of the effects of environmental awareness in production processes. Since the 1990s, environmental regulations have become stronger in the US. Therefore, the regulations of exporting enterprises to reduce environmental impacts in their production processes may have led to the formation of a causal relationship. The fact that enterprises have turned towards production in accordance with environmental standards, from energy efficiency to environmentally friendly production, has revealed the emergence of the relationship between the environment and EI.

A causal relationship is observed between E and IGS between the dates indicated in Table 8. The reason for the causal relationship is that the increase in environmental awareness practices in the 1990s and 2000s encouraged practices to ensure that goods and services transferred to the country from abroad comply with environmental standards. Thus, the relationship between IGS and the environment emerged during this period.

4 Discussion

With the Industrial Revolution, efforts to enhance mass production resulted in disregarding environmental effects. On the other hand, with the liberalization movements in world trade, the development of trade, and the increase in production, the impacts of EI on the environment have started to be discussed and researched. Furthermore, practices concerning a country's security are among the most crucial concerns that have remained relevant from the past to the present. Defense is a critical concern for all countries around the world. This is because a country's security practices necessitate planning to meet the needs of the age in order to preserve people's welfare, maintain peace and stability, and have a role in global developments. As a result, the development of weapon systems, military operations, and defense logistical activities all have environmental consequences. These activities use a lot of energy and influence the environment.

This paper investigates the relationship between EXM, EI, and E for the US economy. We focus on the US, as it leads the world in EXM, EGS, and IGS. We examined the relationships among EXM, EI, and E through time series analysis techniques that allow structural breaks for 1970–2018. Therefore, determining the environmental impacts of defense and foreign trade-oriented production in the US economy motivates this study.

The study investigates the long-term relationship between variables through two models. E is preferred as a dependent variable that represents broad environmental impacts. Models 1 and 2 include exports and imports as independent variables, respectively. The other independent variables in both models are EXM and AL, while the UPG is used as a control variable. The stationarity analysis of the variables was performed in light of the URT-ADF, URT-KPSS, and the URT-CS with multiple structural breaks. According to the findings, the variables were stationary at the I(1) level. The cointegration test by Maki (2012) investigated the long-term relationship between EXM, EI and E. Findings from both models suggest a cointegration relationship between the variables with structural breaks, which indicates a long-run equilibrium relationship. This finding is supported by the work of Ahmed et al. (2020a), Dogan et al. (2019), Gokmenoglu et al. (2020), Udemba (2020), and Cutcu et al. (2023).

The FMOLS, DOLS, and CCR methods were employed to estimate the long-run coefficient. The results obtained from Model 1 and Model 2 posit that the FMOLS and CCR estimators indicate that the EXM of US harm the E at the 1% significance level, revealing the contribution of EXM in the US to environmental quality. This finding is consistent with the work of Solarin et al. (2018), Ullah et al. (2021), Ben Youssef (2023), and Konuk et al. (2023). However, it found that the results differ from the findings of Bildirici (2018), Gokmenoglu (2019), Ahmed et al. (2020a), Qayyum et al. (2021), Eregha et al., (2022), and Chang et al. (2023). Different results were yielded in the literature due to the diversification of the selected country groups, the method employed, or the period handled. Because it is well known that the priorities and practices that different countries throughout the world place on environmental quality differ in the defense industry, notably in the US, it is well known that the country prefers a higher technological level than other nations on the globe and that it utilizes its capacity to produce environmentally friendly goods for the military. The energy demand in the defense field is relatively high. As a result, it can be stated that R&D activities for the US defense sector positively impact environmental quality by using environmentally friendly military equipment and mediating the development of intensive environment-friendly technologies that will meet the high energy demand in the defense field. The Department of Defence (DOD) is dedicated to increasing energy efficiency and lowering environmental impact by encouraging green technologies and environmentally friendly solutions (NDST, 2022). This means the USA Army has focused on adopting the first-ever advanced technology in climate strategy. Primarily, the plan is to reduce the military's emissions by 50% by 2030. Tactical Vehicle Electrification Kits (TVEK), which reduce average fuel consumption by approximately 25% while providing an extra advantage in electricity output, are just one of the most important examples of this. Again, in these 2030- and 2050-focused strategies, the military aims to incorporate advanced hybrid-electric technologies into future military systems and ultimately achieve production of an increasing number of all-electric vehicles, like fully-electric tactical vehicles, by 2050. These and similar advanced technology and low-carbon defence industry strategies are likely to help address challenges such as military-related air pollution, water depletion, and infectious diseases that threaten human health besides environmental degradation (Department of the Army, 2022).

Similarly, according to FMOLS, DOLS, and CCR estimator results, an inverse relationship exists between AL and E in Model 1. This finding is supported by the work of Dogan (2016), Rafiq et al. (2016), Liu et al., (2017a, 2017b), Khan et al. (2018a), Ben Jebli and Ben Youssef (2019), and Mahmood et al. (2019). However, Ullah et al. (2018), Udemba (2020), and Usman et al. (2021) obtained different results. This finding could be attributed to problems faced in taking environmental quality in agricultural areas into account nationally or to a lack of attention placed on such planning. Therefore, differences in country, period, and method can effectively differentiate the results achieved. Thus, practices such as water resource efficiency, using environmentally friendly technologies in energy demand, organic farming, and so on help improve environmental quality.

The three estimators indicate that the relationship between EGS and E is inverse. The research by Ahmed et al. (2020b), Dogan et al. (2019), Muhammad et al. (2020), and Cutcu et al. (2023) supports this conclusion. This finding differs from the result of Liu et al. (2018). This difference may be due to a different country, period, or methodology. The advanced technological structure of US export products enhances environmental quality in the production process. According to FMOLS and DOLS estimators, model 2 results show a positive relationship between IGS and E. This finding is supported by the work of Dogan et al. (2019) Salman et al. (2019), and Ahmed et al. (2020b). This result is different from the study by Cutcu et al. (2023). These findings may be due to country differences, time periods, and methods. Imports are carried out over long distances, adversely affecting environmental quality through the use of fossil fuels in transportation. In addition, importing products from countries with weak environmental standards harms environmental quality.

Model 2 results reveal a positive relationship between UPG and E, according to the DOLS estimator. The research of Wang and Dong (2019), Ahmed et al., (2020a, 2020b), Nathaniel et al. (2020), and Chen et al. (2022) supports this finding. This finding differs from those of Lv and Xu (2019) and Nathaniel et al. (2020). This may be due to the differences in the countries, time periods, and procedures used in these investigations. Factors such as increased energy demand with urbanization, unplanned urbanization, destruction of natural areas, and pollution reduce environmental quality.

The outcomes of Hacker and Hatemi-J's (2006) time-varying causality test reveal a bidirectional causal relationship for all variables at the pertinent dates. The research of Ahmed et al., (2020a, 2020b), Wang and Dong (2019), Pata (2021), Usman and Makhdum (2021)), Bildirici (2017), and Wang and Dong (2019) supports these findings. On the other hand, this result differs from the findings from research by Ahmed et al., (2020a, 2020b), Gokmenoglu et al. (2019), Ben Jebli and Ben Youssef (2019), Quyyum et al. (2021), and Eregha et al., (2022). When the causality results are evaluated, causality relations are found in the 1980s and 1990s and increasingly in the 2000s in the 1970–2018. Based on these findings, the fact that the 1980s and 1990s were years when awareness of the environment began to rise and environmental efforts gained attention in the US may have exposed this association. In the 2000s, the fact that environmental issues gained a global dimension and gained increasing importance indicates that countries like the US, which has a strong economy that shapes the world, increased their efforts in this field. Especially in the US, which ranks high in military expenditures, factors such as the problem of diminishing water resources caused by climate change, increasing natural disasters, and loss of biodiversity are considered security threats, resulting in the US contributing to the increase in policies and projects on environmental issues.

Determining the steps to be taken for environmental quality and knowing the effect of these steps on the result are among the most critical issues of recent times. Although everyone recognizes the necessity of enhancing environmental quality, the limited number of studies on the practices and the determination of their effects constituted the motivation for this study. This study aims to contribute to this gap in the literature. For this purpose, the study investigates the impact of security, one of the most critical issues for all countries worldwide, and sustainable environmental quality on each other. Investigating these relations in the US, the economy followed worldwide within the scope of the environment, security, and foreign trade, increasing the importance of the emerging results. The break tests used to establish the relationships enable extensive results to be obtained, highlighting the additional significance of the research. In addition, the results obtained will play a role in revealing the results in defense, which is an important sector worldwide, and in contributing to the existing literature on the subject. The aim is to provide new knowledge to policymakers and academics and inspire other countries to adopt new practices in the areas of environment, safety, and foreign trade.

5 Conclusions

The need for all nations to collaborate for a sustainable future requires research into the security, foreign trade, and environmental challenges that all countries face today. Researching the issue in the US, which ranks first in the world in the fields of security, foreign trade, and environment, is critical for giving strategy and result-oriented information to all stakeholders in evaluating the findings acquired.

Based on the findings on the relationship among military expenditures, foreign trade, and E variables for the US, the improving effect of military expenditures, exports, and agriculture on environmental quality suggests that policymakers should share their environmental practices with other countries and be pioneers. Practices aimed at making the planet we live on more livable and leaving a sustainable world for future generations can achieve their goals with the practices of other countries in this field. In this context, the necessity of supporting the indispensable activities of all countries in defense with environmentally friendly processes comes to the fore. Defense is vital to a country's national security. Efforts to improve environmental quality are an essential issue on which the whole planet must act together for the sustainability of the current period and future generations, both at the global and national scales. In this regard, the deterioration of environmental quality causes countries to face security threats due to the disasters it causes, from climatic change to natural disasters. Therefore, in today's world, the relationship between factors related to security issues and the environment reveals a process that needs to be considered. In this context, policymakers should quickly introduce environmentally friendly defense policies by incorporating environmental threats into national security policies. In addition, the relationship between the environment and security requires steps to be taken towards arrangements that include organizations for cooperation in the international arena. Thus, sharing reports on the environmental impacts of security practices at the country level with stakeholders worldwide can contribute to progress towards this goal.

The evidence indicates a concerning trend of a deteriorating relationship between imports and environmental quality, underscoring the urgent need for global collaboration and an innovative approach to improvement efforts. When countries are unable to implement environmentally friendly production methods and engage in import activities, they effectively transfer their environmental pollutants to other nations. This highlights the necessity for collective action and a broader perspective in addressing these issues on a global scale. Therefore, it is critical to harmonize environmental norms for import practices on a global scale. In this context, renewable energy sources should be used for the energy sources employed in production. Efficient production policies should be developed for the use of raw materials and in every step of production by producers, thus preventing waste and the consumption of resources. The environmental damage should be minimized through advanced technology by manufacturing companies. Investing in biological capacity is appropriate to reduce firms' dependence on external resources for foreign trade. Countries and companies must adopt regulations while considering the EF of the countries where they conduct most of their business. Companies should organize training to raise environmental awareness among employees and plan activities to increase ecological awareness.

In addition, the deteriorating impact of urbanization on environmental quality in the US shows that policymakers should prioritize the steps to be taken in this field. The steps towards environmentally friendly urbanization practices should be implemented as soon as possible. In order to protect the environment and ensure that it is less affected or, if possible, not adversely affected by the military weapons and ammunition used, priority should be given to R&D activities, especially for the development of vehicles with armor that does not contain depleted uranium and armor-piercing ammunition (MHS, 2023).

On the other hand, a comprehensive examination of the effects of US practices in military expenditures, exports, and agriculture that contribute to environmental quality and guide other countries can accelerate ecological activities.

However, in the US, civil society organizations should take on the task of supporting and collaborating with environmental sustainability efforts around the world due to advances in environmental quality. Reporting and informing the world about environmental progress can accelerate achieving desired results. Thus, new research in this field can lead the way in revealing both the current situation and the results of new generation time series and panel data analyses for different countries or country groups. As a result of the inclusive results acquired, new countries and improved approaches will contribute to expanding the literature in this sector, and helpful policy recommendations can be made.