Examining the asymmetric effects of globalization and tourism on pollution emissions in South Asia

Abstract

The asymmetrical impacts of globalization and tourism on pollution emissions of 5 South Asian countries for the period from 1980 to 2018 are examined through a non-linear autoregressive distributed lag (NARDL) technique, which shows that both short and long-run coefficients are asymmetric. The findings suggest that positive and negative shocks in globalization affect carbon emissions differently in the case of Bangladesh, India, and Pakistan, while similar results are found in the case of Nepal and Sri Lanka in the long run. Furthermore, positive tourism shock, in the long run, ameliorates the environmental quality by reducing carbon emissions in Nepal and Sri Lanka, however, increases the carbon emissions in Bangladesh, India, and Pakistan. While negative tourism shock has an adverse effect on positive shock on carbon emissions in South Asia. The phenomena of globalization and tourism can exert a severe impact in aggravating the pollution emissions that policymakers should forecast and oppose. Based on these findings, some policy suggestions are proposed for South Asian economies.

Introduction

At present, the notable variations have been observed by the scientists in the climate system of the globe. The global surface temperature, according to the report of “State osf Climate”, was 0.38–0.48° C (0.68–0.86° F) above the 1981-2010 average (Blunden and Arndt 2019). Since the pre-industrial period, the global temperature, on average, has risen by 1.1° C as compared with 2011–2015, and the highest average temperatures are recorded during the last 10 years, i.e., 2010–2019 (World Meteorological Organization 2019). Furthermore, the report highlights that, as a prime contributor to global warming, the level of global carbon dioxide has practically quadrupled after the 1960s. Besides, the Intergovernmental Panel on Climate Change (IPCC 2018) underlines that the world climate is enduring significant changes such as the rising lengths of seasons and the level of seas, the declining amount of ice and snow, and fluctuations in global precipitation patterns. This exigency demonstrates pernicious effects on lifestyle, politics, and social, geopolitical, and economic development (Bilgili et al. 2016). Additionally, millions of people have experienced water shortages, hunger, floods, and disease on account of climate change and global warming (Atzori et al. 2018). Currently, the instantaneous CO₂ emissions induced by the inhabitants of this planet during the last 60–70 years, as the majority of the scientists and the researchers have consensus, is the predominant reason for the climate change and global warming (Anderson et al. 2016; Mossler et al. 2017; Koçak et al. 2020).

Nowadays, climate change and global warming are some of the crucial topics that are being widely covered by researchers, scientists, and policy-makers. Also, to explore the determinants of carbon dioxide has become a leading area of interest for the research scholars. The pertinent literature gathers that there are several factors, including but not limited to, financial development, FDI, trade, energy consumption, population, urbanization, and economic growth which exhibit a significant impact on CO₂ emissions (Li and Lin 2015; Cetin et al. 2018; Dong et al. 2018; Nasrollahi et al. 2018; Park et al. 2018).

The tourism sector stimulates economic development by creating jobs, generating income, improving lifestyle and societies, offsetting trade imports through enhancing export function, and transforming the economy as a whole (UNWTO 2017; Sinha et al. 2017; Balsalobre-Lorente et al. 2018). Also, the inflow of international tourists leads to an increase the per capita income and enhancement of the power and transportation sectors through rising socio-economic goods and services (Akadiri et al. 2018; Chon et al. 2013). Therefore, it is theoretically accurate that tourism, along with thriving economic activities, has a close relationship with environmental quality through increasing the quantity of CO₂ emissions via accruing the energy consumption level. The relevant literature reports equivocal findings regarding the link between tourism and the environment. Perhaps, Bach and Gößling (1996) are the first who endeavor to divulge the nexus between tourism and the environment and argue that tourism significantly contributes to environmental degradation by increasing CO₂ emissions. The same findings are also suggested by Goudie and Viles (2013). Besides, the water is used excessively, and natural resources become scanty, and the amount of the waste is increased at natural sites that may lead to the erosion of soil, the surge in the land, water, and air pollution, and eventually may devastate the natural beauty of the globe on account of increasing outflow of tourists (Chan et al. 2018; Latif et al. 2018). Due to the use of electricity, housing facilities, and transports, tourism causes to increase the ratio of CO₂ emissions (Nepal et al. 2019).

In contrast, some researchers advocate that the tourism industry generates favorable effects on the environment by providing indispensable services and promoting innovation and energy efficiency for the country’s development. Hence, tourism is deemed as an instrument of climate protection (Gössling and Hall 2006; Imran et al. 2014; Naradda Gamage et al. 2017; Dogan and Aslan 2017; Paramati et al. 2018; Akadiri et al. 2018). Furthermore, the tourism sector may create a more positive impact on the environmental quality if the authorities of the global economies adopt eco-friendly strategies (Ahmad et al. 2019).

Globalization is also an essential driver of the CO₂ emissions that has a significant association with the climate change since it affects the economic growth through enhancing international trade, foreign direct investment, transfer of technology, and tourism (Akadiri et al. 2018) that requires an enormous use of energy which results in higher the level of CO₂ emissions (Shahbaz et al. 2018; Figge et al. 2017). Many researchers argue that globalization has a direct relationship with CO₂ emissions by degrading environmental quality. For instance, Figge et al. (2017), Kwabena Twerefou et al. (2017), Shahbaz et al. (2018), and Salahuddin et al. (2019) confirm that the global environment endures deterioration in quality on account of the increasing trend in globalization since globalization induces the global economies to attain economic growth swiftly through enhancing foreign trade and FDI. It leads to increase industrialization in which traditional methods are employed for energy consumption; consequently, the quality of the environment declines (Navarro 1998; Sharma 2011; Lau et al. 2014; Acquaye et al. 2017; Ullah et al. 2020).

However, the findings of many studies disclose that climate quality improves as more globalization occurs among world countries, as Turner and Witt (2001) find that the detrimental effects of CO₂ emissions reductions on account of the higher level of globalization that also boosts the access of tourism. Another research evidence by Cavlovic et al. (2000) infers that globalization can control the emission of carbon dioxide by driving knowledge and innovation that brings economic efficiency; consequently, the globalization moderates the quality of the environment.

Since the world has become a global village, so globalization is playing a pivotal role in improving the economic growth and tourism sector. Furthermore, both indicators, i.e., globalization and tourism, possess the notable effects of environmental quality; thereby, it motivates to study whether both variables predict the environmental quality or not. Although there exists an association between globalization, tourism, and the environment, the ample body of literature focuses on investigating the impact of globalization and tourism on CO₂ emissions separately. Less attention is given to the connection of these factors together so far. Therefore, we choose the South Asian region that has a remarkable contribution to the world economic market and also has the most beautiful and breathtaking places which attract the world’s tourists, as Fig. 1 depicts some most visited places.

Fig. 1

Frequently visited places in South Asian economies

The contribution of our study to the existing literature is fourfold: First, all the previous studies explore the effects of globalization and tourism on CO₂ emissions separately, while the current study considers both drivers in a single model. Second, the majority of the studies employ panel data to examine the dynamic effects that may suffer from the exigency of aggregation bias (Meo et al. 2018). To account for this aggregation bias, our study does the multi-country analysis at the disaggregated (country) level. Third, to the best of our knowledge, this is the first study for the region of South Asian economies that analyzes the dynamic association between globalization, tourism, and CO₂ emissions. Lastly, all the relevant studies deploy the symmetric modeling for seeking the nexus among the interested variables; however, the asymmetric modeling has more power to unveil more detailed and reliable findings (Katrakilidis and Trachanas 2012; Li et al. 2019; Marques et al. 2019). Moreover, in the real world scenario, the asymmetric effects of variables seem more practical, as the behavior of people cannot be predicted with certainty. Therefore, the asymmetric approach of modeling is a more plausible choice in comparison with symmetric one. So, this is the first study that contributes to the literature by inspecting the asymmetric effects of globalization and tourism on CO₂ emissions for the South Asian region economies.

The rest of the study is organized as follows. The “Literature review” section gives a complete literature review on globalization, tourism, and CO₂ emissions. The “Model and methodology” section shows the methodology, variables definition, and data descriptive statistics. The “Results and discussion” section gives the symmetric and asymmetric ARDL results of the short and long run with economic implications. The “Conclusion and policy implication” section concludes the paper with some implications.

Literature review

Since the world has become a global village, the role of globalization in upgrading the tourism sector and GDP growth is pivotal. Furthermore, along with enhancing economic growth, globalization and tourism create considerable effects on environmental quality through increasing the level of energy consumption, via boosting the activities of power and transportation sectors, etc. (Roudi et al. 2019; Akadiri et al. 2018). Therefore, it is crucial to consider the effects of globalization and tourism on CO₂ emissions.

Globalization and environmental pollution

A plethora of literature is available that explores the dynamic association between globalization and CO₂. These studies can be classified into two bunches, panel and time-series data studies. As for the studies which deploy the panel data, Lv and Xu (2018) take the panel of 15 emerging economies from 1970 to 2012 and apply advanced econometric panel approaches. They deduce that globalization ameliorates the deleterious repercussions of carbon dioxide emissions. Likewise, You and Lv (2018) verify the spatial correlation in carbon dioxide among selected 83 economies and the outcome reports that globalization is environmentally friendly since the ratio of CO₂ emissions shrinks on account of the upward trend of globalization in the selected panel. Another study by Lim et al. (2015) for 89 selected countries also confirms that the globalization plays a notable role in enhancing the atmosphere quality.

However, Shahbaz et al. (2018) investigate the nexus between CO₂ and globalization for 25 advanced economies and argue that environmental pollution accrues as the countries become more globalized. Similarly, Salahuddin et al. (2019) use the data of 44 SSA economies and confirm that globalization is one of those factors which escalate the level of CO₂ emissions. Also, Kwabena Twerefou et al. (2017) apply the GMM technique and infer that globalization shows a significant impact on environmental degradation. Le et al. (2016) inspects the CO₂-globalization nexus for selected 98 economies and employ trade openness as a proxy for globalization. The findings suggest that the environmental quality worsens as the economies adopt the trade-openness policy. Furthermore, Li et al. (2015) take the panel of 134 economies and conclude that globalization declines the climate quality while deploying the air visibility for environmental quality measurement.

Besides, many researchers suggest the globalization is a mixed blessing for the global environment due to having mixed effects on CO₂ emissions. The findings of Figge et al. (2017), employing the mega set of 171 economies, unveil that social globalization has a negative association with environmental deterioration while its ratio increases due to economic globalization. Another study by Doytch and Uctum (2016) deploys FDI as a proxy for globalization and indicates that the sectorial FDI exhibits positive as well as the negative impact on CO₂ emissions. Additionally, Dreher (2006) use different environmental proxies and explore that the harmful effects of water pollution and sulfur dioxide reduce on account of globalization; however, it has no significant impact on round wood production and CO₂. Many other researchers (See; Frankel and Romer 1999; Antweiler et al. 2001; Baek et al. 2009; Managi and Kumar 2009; Boulatoff and Jenkins 2010; Shahbaz et al. 2012; Shahbaz and Leit 2013) take the sample of different economies and also affirm that globalization exhibits mixed impact on CO₂. Intriguingly, the results of Martens and Raza (2010) reveal that globalization possesses an insignificant association with the environment when they utilize the different environmental sustainability indexes for 150 economies.

As far as the time-series studies are concerned, Shahbaz et al. (2018) disclose that globalization, energy consumption, and economic growth are the cause to degrade the environment by increasing the CO₂ emissions in Japan. Likewise, Shahbaz et al. (2015) check the effects of globalization for India and support that globalization enlarges the pernicious impact of carbon dioxide in both periods. Besides, some researchers report the mixed impact of globalization on the global environment. Shahbaz et al. (2016) take the data of 19 African economies and deduce that globalization demonstrates the mixed effects on CO₂. Applying the ARDL approach for the period of 1971–2014, Ahmed et al. (2019) also find that globalization has no association with the ecological footprint, while it aggravates the environment of the Malaysian economy by surging carbon footprint. However, Xu et al. (2018) assess the impact of globalization on Saudi Arabia, and the findings indicate that CO₂ emissions have no significant link with globalization. Since the findings of all relevant studies are mixed, therefore, it is transparent that the effects of globalization on the global environment are inconclusive.

Tourism and environmental pollution

Again, a large number of studies are available that analyze the effects of the tourism sector on CO₂ emissions. Also, these studies can be categorized into two groups, panel data, and time-series studies. Whereas the panel data studies are concerned, Koçak et al. 2020) explore the CO₂-tourism nexus for top 10 visited economies and gather that the tourism industry produces favorable effects on the quality of the environment. Paramati et al. (2018), Dogan and Aslan (2017), and Lee and Brahmasrene (2013) take different data samples for EU economies to divulge the dynamic association between CO₂ and tourism. All these studies reach a consensus that tourism developments have a significant decreasing effect on carbon dioxide emissions. Also, Akadiri et al. (2018) report the same findings for 16 small developing economies, deploying panel Granger causality testing technique.

On the other hand, Paramati et al. (2017a, b), Zaman et al. (2017), and León et al. (2014) argue that the upward trend of tourism in selected developed and developing economies is environmentally unfriendly. Likewise, the results by Shakouri et al. (2017) indicate that Southeast Asian economies endure climate degradation on account of the increasing ratio of foreign tourists. Moreover, Dogan and Aslan (2017) and Zaman et al. (2017) also suggest that the improvement in the tourism sector has an increasing association with CO₂ emissions in OECD and transition economies, respectively. Interestingly, Paramati et al. (2017a, b) employ the sample of Western and Eastern European economies from 1995 to 2013 and conclude that tourism development has increasing and decreasing effects on CO₂ emissions in Eastern Europe and Western Europe, respectively.

Considering the studies that use the time-series data, Nepal et al. (2019) scrutinize the nexus between CO₂ and tourism industry employing the data for 1975–2014. The findings unveil that the climate gets worse due to CO₂ emissions as more tourists visit Nepal. Taking the monthly data of the USA for 1996–2015, Raza et al. (2017) also research the effects of tourism on CO₂ by using the Wavelet transforms method. They confirm that tourist arrivals are the significant cause of increasing carbon dioxide. Similarly, For Greece, Işik et al. (2017) deploy the ARDL technique and conclude that financial development, trade, and tourism affect the environment adversely.

Likewise, Sharif et al. (2017), employing the data for the period of 1972–2013, report that CO₂ emissions surge on account of the upward trend of tourist arrivals in Pakistan. Also, De Vita et al. (2015) inspect the dynamic association between CO₂ and tourism industry in Turkey. The results reveal that the environmental quality degrades as CO₂ emissions rise due to tourist arrivals. Besides, Katircioglu (2014) apply the ARDL approach to inquire about the relationship between CO₂ and the tourism sector. The outcome discloses that the ratio of carbon dioxide accrues as more foreigners visit Cyprus. Another study by Solarin (2014) also uses the ARDL technique and confirms that the Malaysian tourism sector significantly deteriorates the quality of climate.

On the contrary, Naradda Gamage et al. (2017) examine the impact of tourism on carbon dioxide, taking the sample for the period of 1974–2013. Applying the DOLS method, they argue that the tourism industry plays a notable role in enhancing the environment quality by decreasing CO₂. Besides, some researchers find mixed findings such as Azam et al. (2018) use the sample of three economies and deduce in Malaysia, CO₂ emissions demonstrate a direct association with tourism. However, the environment of Singapore and Thailand gets better as more people visit these economies. Similarly, Sghaier et al. (2019) infer that the detrimental effects of CO₂ emissions decline and increase in Egypt and Tunisia, respectively, due to tourist arrivals, while the study suggests no significant nexus between CO₂ and tourism in Morocco.

The aforementioned literature regarding tourism-CO₂ nexus reveals the inconclusive results. Furthermore, most of the studies employ panel data that may be suspected of aggregation bias (Saeed Meo et al. 2018; Nosheen et al. 2019); thus, the country-level analysis may bring more productive results. Also, less attention is given to the South Asian economies, and the ample body of the pertinent literature deploys symmetric modeling; however, asymmetric modeling has more power to explore more detailed and reliable findings (Bildirici and Turkmen 2015; Saeed Meo et al. 2018). Therefore, considering the potential asymmetric effects of globalization and tourism in the South Asian economies, the current study contributes to the existing literature by exploring the asymmetric impacts of tourism and globalization on CO₂ emissions for South Asian region countries at the disaggregated level (Table 1).

Table 1 Summary of literature

↓, ↑, and × indicate the negative, positive, and no impact of the focused variable on CO₂, respectively

Model and methodology

In this study, we first want to see the impact of globalization and international tourism on CO₂ emission by using a linear approach. To this end, following the literature, we have constructed model 1 as given below:

$$ {\mathrm{CO}}_{2,t}={\beta}_0+{\beta}_1{\mathrm{Glob}}_{\mathrm{t}}+{\beta}_2{\mathrm{Tour}}_t+{\beta}_3{\mathrm{GDP}}_t+{\varepsilon}_t $$

(1)

According to Eq. 1, annual CO₂ emission in Pakistan depends upon the globalization (Glob_t) and international tourism (Tour_t), GDP per capita (GDP_t), and error term (ε_t). However, Eq. 1 is a long-run model and it only gives us long-run estimates irrespective of methodology. To get short-run estimates, we redesign Eq. 1 into a format of error correction as specified in Eq. 2.

$$ {\Delta \mathrm{CO}}_{2,t}={\beta}_0+{\sum}_{k=1}^p{\theta}_k{\Delta \mathrm{CO}}_{2,t-k}+{\sum}_{k=0}^P{\pi}_k{\Delta \mathrm{Glob}}_{t-k}+{\sum}_{k=0}^P{\updelta}_{\mathrm{k}}{\Delta \mathrm{Tour}}_{t-k}+{\sum}_{k=0}^P{\lambda}_k{\mathrm{GDP}}_{t-k}+{\alpha}_1{\mathrm{CO}}_{2,t-1}+{\alpha}_2{\mathrm{Glob}}_{t-1}+{\alpha}_3{\mathrm{Tour}}_{t-1}+{\alpha}_4{\mathrm{GDP}}_{t-1}+{\varepsilon}_t $$

(2)

Equation 2 is commonly known as the ARDL model presented by Pesaran et al. (2001). The coefficients attached to ∆ signs give us short-run results. On the other hand, the coefficients α₂–α₄, normalized on α₁, give us long-run findings. This method is quite beneficial in the sense that it gives both short-run and long-run estimates in a single equation. Moreover, we do not need to worry about whether our concerned variables are of I(0), I(1), or mixture of both, as this method can take care of the integrating properties of variables.

Next, in this study, our foremost objective is to find the asymmetric impact of Glob_t and Tour_t on CO₂ in Pakistan. For this purpose, Shin et al. (2014) suggested the breaking down of relevant variables into its positive and negative parts. Hence, following the same procedure, in this study, we have broken down our related variables into positive (Glob⁺, Tour⁺) and negative (Glob⁻, Tour⁻) sections.

$$ {{\mathrm{Glob}}^{+}}_t=\kern0.5em {\sum}_{n=1}^t\Delta {{\mathrm{Glob}}^{+}}_t=\kern0.5em {\sum}_{n=1}^t\max\ \left(\Delta {{\mathrm{Glob}}^{+}}_t,0\right) $$

(3a)

$$ {{\mathrm{Glob}}^{-}}_t=\kern0.5em {\sum}_{n=1}^t\Delta {{\mathrm{Glob}}^{-}}_t=\kern0.5em {\sum}_{n=1}^t\min\ \left(\Delta {{\mathrm{Glob}}^{-}}_t,0\right) $$

(4a)

$$ {{\mathrm{Tour}}^{+}}_t=\kern0.5em {\sum}_{n=1}^t\Delta {{\mathrm{Tour}}^{+}}_t=\kern0.5em {\sum}_{n=1}^t\max\ \left(\Delta {{\mathrm{Tour}}^{+}}_t,0\right) $$

(3b)

$$ {{\mathrm{Tour}}^{-}}_t=\kern0.5em {\sum}_{n=1}^t\Delta {{\mathrm{Tour}}^{-}}_t=\kern0.5em {\sum}_{n=1}^t\min\ \left(\Delta {{\mathrm{Tour}}^{-}}_t,0\right) $$

(4b)

After breaking down the variables, we will then put these positive and negative shocks into Eq. 2 in place of our standard variables, and this equation takes the form of non-linear ARDL model as depicted below in Eq. 5.

$$ {\Delta \mathrm{CO}}_{2,t}={\beta}_0+{\sum}_{k=1}^P{\theta}_k{\Delta \mathrm{CO}}_{2,t-k}+{\sum}_{k=0}^P{\psi}_k\Delta {{\mathrm{Glob}}^{+}}_{t-k}+{\sum}_{k=0}^P{\pi}_k\Delta {{\mathrm{Glob}}^{-}}_{t-k}+\sum \limits_{k=0}^P{\delta}_k\Delta {{\mathrm{Tour}}^{+}}_{t-k}+\sum \limits_{k=0}^P{\eta}_k\Delta {{\mathrm{Tour}}^{-}}_{t-k}+\sum \limits_{k=0}^P{\lambda}_k{\mathrm{GDP}}_{t-k}+{\alpha}_1{\mathrm{CO}}_{2,t-1}+{\alpha}_2{{\mathrm{Glob}}^{+}}_{t-1}+{\alpha}_3{{\mathrm{Glob}}^{-}}_{t-1}+{\alpha}_4{{\mathrm{Tour}}^{+}}_{t-1}+{\alpha}_5{{\mathrm{Tour}}^{-}}_{t-1}+{\alpha}_6{\mathrm{GDP}}_{t-1}+{\varepsilon}_t $$

(5)

Equation 5 is commonly known as non-linear ARDL model of Shin et al. (2014). It works in the same way as the linear ARDL model. The NARDL has several advantages over the traditional cointegration techniques, including the following: first, it can produce efficient findings even for the small sample size. Second, it does not require the stationary test for modeled variables. Last, it estimates effective coefficients even if the proposed variables are integrated at I(0), I(1), or I(0) and I(1). Once we get the estimates from Eq. 5, we need to perform an additional test, i.e., the Wald test to check the validity of asymmetry in the effects of variables, in the short and long run. In the short run, the Wald test confirms the presence of combined asymmetry in the effects of short-run estimates if Σѱ_k ≠ Σπ_k and δ_k ≠ η_k. Similarly, the long-run asymmetries are approved by the Wald test if $\raisebox{1ex}{${\alpha_2}^{+}$}\!\left/ \!\raisebox{-1ex}{${\alpha}_1$}\right.$≠ $\raisebox{1ex}{${\alpha_3}^{-}$}\!\left/ \!\raisebox{-1ex}{${\alpha}_1$}\right.,\raisebox{1ex}{${\alpha_4}^{+}$}\!\left/ \!\raisebox{-1ex}{${\alpha}_1$}\right.$≠ $\raisebox{1ex}{${\alpha_5}^{-}$}\!\left/ \!\raisebox{-1ex}{${\alpha}_1\ $}\right.$.

Data

Data employed for this analysis cover the period 1980–2018 for five selected South Asian countries, i.e., Bangladesh, India, Nepal, Pakistan, and Sri Lanka. This group of South Asian countries was selected for one reason because these regions have well-experienced rapid globalization and tourism industry with significant emissions of carbon. The data were obtained from three sources: the World Bank and KOF. The detailed variable descriptions are described in Table 2. The data has one dependent (CO₂), two independent variables (Glob and Tour), and one control variable (GDP). The detailed descriptive statistics are given in Table 3. The statistics indicate that India is generating the highest carbon emissions with the highest value of 1.101 metric tons per capita, while Sri Lanka is generating the lowest mean value of 0.215 metric tons per capita every year, which shows that India is at the top for carbon emitters in South Asian countries. Moreover, 55.91 is the highest, and 40.55 is the lowest mean value of the globalization of Sri Lanka and Nepal, respectively. There is a significant fluctuation observed in international tourism receipts, which is from 111.471 (lowest value) in Bangladesh and 11,759.04 (highest value) in India. The mean values of GDP per capita vary from 470.911 (Nepal) to 1984.517 (Sri Lanka).

Table 2 Variables description and source

Table 3 Descriptive statistics

Results and discussion

For traditional unit root tests, ADF and PP tests were executed to decide the stationarity property of the series in Table 4. ADF and PP tests have the null hypothesis of non-stationary and the alternative hypothesis of stationary, while all variables are accepted the null at the level. This implies that all the variables are stationary at the first difference I(1). Thus, the ARDL methodology is assumed in this study.

Table 4 Unit root tests

Regarding cointegration, the outcomes specified in Table 5 show that the F-statistics is lying between the lower bound and upper bound values in the symmetric ARDL model in Bangladesh, India, Nepal, Pakistan, and Sri Lanka. This implies that there is no cointegration existing in the linear ARDL model while the null hypothesis of no cointegration has been rejected for asymmetric ARDL in all South Asian countries, suggesting the cointegration exists only in non-linear ARDL model.

Table 5 Asymmetric cointegration

The asymmetric ARDL findings of South Asian countries are reported in Table 6. In results, short-run globalization results show that positive change has a significant positive impact on pollution in Nepal and Pakistan, while negative impact in India. It indicates that 1% increase in globalization would increase environmental pollution to 0.316% and 0.711% in Nepal and Pakistan; however, environmental pollution 0.810% would fall in India in the short run. While negative shock has adverse positive results in India and Sri Lanka, negative shocks have a significant negative impact on carbon emissions in Pakistan. Similarly, the results also indicate that 1% fall in globalization would increase environmental pollution to 0.091% and 0.443% in India and Sri Lanka, while environmental pollution would fall by 0.020% in Pakistan in the short run. The past values of positive and negative shocks of globalization on carbon emissions have significance in sign and magnitude. In how many South Asian countries’ short-term impacts convert into significant and meaningful long-run effects? Therefore, the long-run results of panel B provide the answer and as we observed that positive shocks of globalization maintained the results for India and Nepal. While positive shocks of globalization effects are lost in Bangladesh, Pakistan, and Sri Lanka, the negative shock of globalization effects is also maintained in Bangladesh, India, and Sri Lanka.

Table 6 Short and long-run estimates

The long-run coefficients of Glob⁺ are − 0.417, − 0.600, and − 0.087 in Bangladesh, India, and Pakistan, respectively. These results show that positive shocks have an adverse influence on carbon emissions. For instance, 1% increases in Glob⁺ shrunk carbon emission by 0.417% in Bangladesh, 0.600% in India, and 0.087% in Pakistan. A similar result is found in OECD and developed countries (Tamazian et al. 2009; Balsalobre-Lorente et al. 2019) noted that positive shock in globalization is more environment friendly, possibly because globalization is considered as a tool for improving technical progress. It implies that globalization enables developing economies to achieve energy-saving advanced technology from developed nations with low amount of CO₂ discharges.

While long-run coefficients of Glob⁺ in Nepal and Sri Lanka are 0.718 and 0.714, it signifies that a positive shock of globalization has a positive effect on carbon emissions. The results suggest that 1% increase in globalization escalated carbon emission by 0.718 % in Nepal and 0.714% in Sri Lanka. These results are consistent with the findings of Koengkan et al. (2019) for Latin America and the Caribbean economies, who noted that positive shocks of globalization have a negative influence on environmental quality. Moreover, Sabir and Gorus (2019) have similar results found in Bangladesh, India, Nepal, Pakistan, and Sri Lanka, which noted that globalization slows down environmental quality globally as well in South Asia. The positive shock of globalization has a positive influence on carbon emissions. It suggests that globalization leads to higher pressure on the environment. The examined results indicated that globalization leads to the formation of pollution-intensive industries in developing countries due to weaker environmental law enforcement (Copeland and Taylor 2004). The examined results also suggested that globalization depletes and destroys the natural resources that cause environmental quality.

The negative shock of Glob⁻ has a positive influence on carbon emissions in Bangladesh, India, Nepal, and Sri Lanka, except Pakistan. The negative shocks estimated of Bangladesh, India, Nepal, and Sri Lanka are 0.200, 0.208, 0.487, and 0.005, respectively. Therefore, 1% decrease in globalization raised carbon emission by about 0.200%, 0.208%, 0.487%, and 0.005% in Bangladesh, India, Nepal, and Sri Lanka, respectively. However, this result is reserved by the globalization theory and is also reliable with Koengkan et al. (2019) for Latin America and the Caribbean economies.

In the short run, positive shock (ΔTour_t⁺) has mixed results on carbon emission in South Asian countries, indicating that 1% increase in tourism would lead to an increase in pollution of 0.291% in Nepal and 0.189% in Sri Lanka while falling environmental pollution 0.470% in Bangladesh and 0.437 in Pakistan. The tourism negative shock (ΔTour_t⁻) also shows that environmental pollution increases by 0.013% in Bangladesh while the adverse result is found in Nepal and fall in 0.005% in Nepal. From all South Asian non-linear model in panel B, we see that positive globalization effects are maintained, while the impact of negative globalization shock does not exist in Nepal.

Therefore, long-run outcomes show that increased tourism has a significant negative effect on carbon emission in Bangladesh, India, and Pakistan. Therefore, 1% increases in tourism reduced CO₂ by about 0.609%, 0.549%, and 0.174% in Bangladesh, India, and Pakistan. Our results on tourism gain support from Lee and Brahmasrene (2013), Zhang et al. (2016), and Paramati et al. (2017a, b), who noted that tourism has a positive impact on the environment quality. The possible reason for this finding is that tourism is a sub-sector of the service sector, and it is cleaner than the agricultural and industrial sectors because the tourism is comparatively less energy consuming than other sectors (Begum et al. 2015; Bilgili et al. 2017). The findings advise that carbon emissions start to decrease as an economy moves from the agricultural and industrial sectors to the service sector. As the economy then improvements from an industrial to services, carbon emissions begin to decline. Adverse results are found in Nepal and Sri Lanka. The possible reason of negative shock is that tourism increases the arrivals and departures of tourists and drives more public transport services (Dogan and Aslan 2017; Sharif et al. 2017). It implies that transportation is the leading factor in CO₂ emissions. One of another justification is tourists demand a variety of infrastructure services such as restaurants, accommodation, hotels, roads, airports, ports, railways, and telecommunications. It concludes that infrastructure contributes significantly to increase carbon emissions (Lee and Brahmasrene 2013).

Furthermore, long-run estimates show that the effect of negative changes in tourism on CO₂ is significantly positive in Bangladesh, India, and Pakistan, while significantly negative in Nepal. The results suggest that 1% fall in tourism raised CO₂ by about 0.009%, 0.077%, and 0.901% in Bangladesh, India, and Pakistan, respectively, but reduced CO₂ in Nepal by about 0.091%. The outcomes suggest that globalization enables the tourism industry that has direct, indirect, as well as complementary negative effects on carbon emissions. The result implies that tourist demands clean, green natural beauty, and that raises the environmental quality. A possible reason for Nepal’s results is that the tourist activities produce various kinds of waste, including water waste, littering, damage to the natural sites, and depletion of natural resources, therefore resulting to negative shock causing risk in environmental quality, especially in the LDCs (Michailidou et al. 2016). Similarly, long-run results further show that GDP significantly contributed to an increase in carbon emissions in Bangladesh, India, Nepal, and Pakistan, while GDP is reducing carbon emissions in Sri Lanka. Almost all GDPs have similar results on carbon emissions in the short run, except Bangladesh.

The error correction term (ECT) is negatively significant with coefficients equal to − 0.430, − 0.331, − 0.570, − 0.491, and − 0.711 for the Bangladesh, India, Nepal, Pakistan, and Sri Lanka. These values entail that the speed of adjustment towards the long-run equilibrium is about 43% over each year for Bangladesh, 33% for India, 57% for Nepal, 49% for Pakistan, and 71% for Sri Lanka. However, diagnostic numbers endorse that the asymmetric models are usable and reliable. For instance, the Jacque-Bera (JB), LM, Breusch-Godfrey (BG), RESET test statistics indicate normality of the residuals, free of autocorrelation and heteroscedasticity problems, and correct functional form in NARDL model. Also, CUSUM and CUSUM square tests endorse the stability of NARDL-estimated models. In the Wald test, globalization and tourism variables hold short and long-run asymmetric assumptions in South Asian countries except for two measures. The dynamic multiplier graphs, based on the aforementioned NARDL model, are illustrated in Fig. 2. The asymmetry curves show the non-linear mixture of the dynamic multipliers due to positive and negative shocks of globalization and tourism.

Fig. 2

Dynamic multiplier graphs of South Asian countries

Conclusion and policy implication

The complexities of the asymmetrical effects of globalization and tourism on environment quality were studied for five South Asian countries, from 1980 to 2018 by using the non-linear ARDL approach. The results indicate that globalization components have an asymmetric impact on carbon emissions. The positive shock of globalization exerts a positive impact on CO₂ emissions in Nepal, Sri Lanka, and Pakistan while the adverse effect is found in Bangladesh, India, and Pakistan in the long run. The phenomena of the negative shock of globalization could exert a positive effect on carbon emissions in Bangladesh, India, and Sri Lanka, while an insignificant adverse result is found in Pakistan. Nevertheless, while the asymmetric effects were statistically significant in South Asia, they were different in magnitude and direction. Besides, a 1% increase in globalization decreased carbon emission by 0.417% in Bangladesh, 0.600% in India, and 0.087% in Pakistan while it increased 0.718% in Nepal and 0.714% in Sri Lanka.

These results of asymmetry suggest that positive and negative fluctuations in tourism affect carbon emissions differently. Indeed, positive fluctuations in tourism are positive signs on pollution in the case of Nepal and Sri Lanka while negatively significant in Bangladesh, India, and Pakistan in the long run with anticipated economic signs. Besides, a 1% increase in tourism can be decreased 0.609% in Bangladesh, 0.211% in India, and 0.174% in Pakistan while it increased carbon emissions by 0.504% in Nepal and 0.643% in Sri Lanka. Negative0 fluctuations in tourism are statistically significant, with a positive sign in the long term in Bangladesh, India, and Pakistan while adverse on carbon emissions in Nepal. The short-run asymmetric results of globalization and tourism also have significance.

Based on the short run, our finding recommends that governments integrate the rule and regulation of environmental pollution with globalization policies in Nepal and Pakistan. Similarly, Nepal and Sri Lanka governments make tourism polices in short run that reduce the pollution that may be possible if governments ban private transports and start only green transport in tourist places. Therefore, the green transport approach is a solution for environmental authorities. A tight environmental policy could be helpful to decline environmental hazards in Nepal, Pakistan, and Sri Lanka in the short run.

Based on the long run, our finding suggested that the government should encourage foreign investors to invest in green and clean energy projects in Nepal and Sri Lanka to protect environmental quality through innovation and international collaborations. Policymakers also consider “globalization” as a key instrument in their environmental policy context to improve the quality of environmental health in Bangladesh, Nepal, and Sri Lanka in the long run. Countries like Bangladesh, Nepal, and Sri Lanka should take incentives from developed countries in the transportation and construction sector. Nepal and Sri Lanka governments should redesign the transportation and infrastructure policy in the long term. Also, transport and construction rules which are well defined in tourism regions may be adopted to upgrade the environmental quality. Developed and South Asian countries have significant differences in terms of social, economic, institutional, infrastructure, human capital, technological, and environmental awareness (Chan et al. 2018). Therefore, the tourism sector should plan and develop low-carbon tourism products in South Asia and other developing regions based on developed countries. Furthermore, clean energy is the best option to reduce pollution in tourism regions. India and Pakistan could improve the tourism sector by lowering the environmental pollution. The South Asian region could attract clean and green technology from the developed countries. This study left a potential conclusion from different quantiles that is an avenue for future research. One direction of future research work is environmental pollution asymmetries, which can be investigated with macroeconomic indicators in South Asia.