Investigating the nexus among sulfur dioxide emission, energy consumption, and economic growth: empirical evidence from Pakistan

Developing countries like Pakistan majorly depend on fossil fuels for achieving higher economic growth but have sloppy environmental rules and regulations in order to attract foreign direct investment (FDI). As a result, energy consumption is considered the primary cause of environmental degradation. Besides CO2 emission, environmental degradation is also associated with emission of sulfur dioxide (SO2). The purpose of this study was to investigate the relationship among SO2 emissions, energy consumption, economic growth, and FDI in Pakistan. By applying the 3SLS method, study has estimated the scale effect, composition effect, and technique effect. The scale effect and technique effect findings indicated that capital stock, FDI, and SO2 emissions all had a significant impact on GDP. When the capital accumulation effects of FDI were considered, the relationship between FDI and stock of capital was found to be positive. According to the technique effect results, FDI, population density, and energy consumption were all significantly related to SO2 emissions. The study came to a conclusion with significant policy implications.


Introduction
Rapid economic growth in developing countries like China, India, and Pakistan has led to increased energy consumption. Developing countries heavily depend on fossil fuels for energy generation and consumption. The massive use of the energy especially from fossil fuels causes severe environmental degradation. Emitters such as CO2, SO2, and particulate matter from energy consumption lead to air pollution. Although other factors also contribute in pollution emission, fossil fuel consumption causes the larger amount of pollution emissions. Yilanci et al. (2020) assert that long-term impact of energy consumption leads to pollution emission. Literature showing the role of energy consumption in CO2 emission is rich and recent (Bakhsh et al. 2017, Guzel and Okumus 2020, Salahuddin et al. 2018, Udemba et al. 2020. Considering SO2 emission, it depends on the sulfur content of the fossil fuels, in addition to the sectoral processes leading to pollutant emission.
Major sources of SO2 emission are energy, transportation, industries, agriculture, forest, and livestock. Figure 1 shows different sources of SO2 emission. Electric utilities including electric power plants and refineries are the major sources of SO2 emission. Point sources excluding electric utilities are vehicles operating on roads, highways, and streets. Residential wood burning, gasoline service stations, dry cleaners, wildfires, and agriculture tilling are among the nonpoint sources of SO2 emission. Non-road sources of SO2 emission include lawn and garden equipment, aircrafts, locomotives, construction equipment, and recreational equipment. The main economic source of SO2 emission is the industrial sector as it consumes a larger quantity of energy sources. This sector highly relies on the consumption of fossil fuels in developing countries including Pakistan. Yang and Shan (2019) argue that industrialization, economic scale, and energy consumption are major contributing factors of SO2 emission in the developing countries. Many studies show a positive relationship between economic growth and SO2 emission (Coggin 2019, Hu et al. 2019, Nadeem et al. 2020, Ramakrishnan et al. 2016, Wu 2019. Ru et al. (2018) and Wang et al. (2016a, b) argue that there exists environmental Kuznets curve (EKC) pattern of relationship between economic growth and SO2 emission. Increasing renewable energy consumption and energy efficiency can reduce pollution emission whereas non-renewable energy consumptions lead to environmental degradation (Destek and Sinha 2020;Hu et al. 2019;Nathaniel et al. 2020;Zafar et al. 2019). Technology, energy mix, sulfur efficiency, industrial structure, and energy efficiency effects are associated with a dwindling SO2 emission (Yang and Shan 2019). Energy prices are also important in mitigating pollutant emissions (Hanif et al. 2019, Nguyen et al. 2020. Stringent energy policy with strong control on energy prices can reduce energy consumption, increase energy efficiency, and decrease pollutant emissions. Pakistan is a developing country and it has diverse customized energy sources, processes, environmental regulation, and economic conditions compared to other developed countries. Therefore, SO2 emission can have distinct emission trajectories. Further, industrial sector is considered the main economic source of SO2 emission as it consumes electricity 26.3% and gas 19.2%. Contribution of industrial sector in gross domestic product is 19.1% (GOP 2021). SO2 emission deteriorates ambient environment quality causing greater threats to human health. Thus, it is imperative to understand the economic growth, energy consumption, and SO2 emission nexus at country level. Nordhaus (2010) and Stern and van Dijk (2017) maintain that it is the useful way to examine the emissions associated with social and economic development.
This study has several departures from the earlier work on FDI, economic growth, and pollution emission. First, we look into SO 2 emission instead of CO 2 because of ambient effects of SO 2 emission. We used SO2 for two main reasons. Firstly, there is direct link between SO2 emission and industrial output. Cherniwachan (2012) asserted that 1% increase in industrial share of output leads to 11.8% increase in per capita SO2 emission. Secondly, the effects of SO2 emission are ambient rather than global. Second departure from the literature is that we employ three-stage least square (3SLS) method which is conceptually stronger and empirically robust, besides being more appropriate to find out trilateral nexus of FDI, environmental pollution, and economic growth. Many studies relating to Pakistan determined relation between FDI and economic growth but very few studies estimated the relation between FDI, economic growth, and environmental pollution (Mahmood and Chaudhary 2012, Wang et al. 2019, Yasmeen et al. 2020). However, we find that these studies do not consider the mechanisms through which foreign investment impacts environmental pollution particularly SO2 emission.
The remaining paper is arranged as follows: The second section provides a brief review of literature. The first subsection of "Material and methods" describes data and sources. Second sub-section provides details on 3SLS method. "Results and discussions" contains findings of 3SLS method including technique effect, scale effect, and composition effect, and findings of the study are compared with the literature as well. "Conclusions and policy implications" gives conclusions and recommendations.

Brief literature review
Sustainable development and environmental protection are c e n t r a l t o t h e w o r l d w i d e d e v e l o p m e n t a g e n d a .
Environmental degradation elevates concerns about global climate change (Salahuddin et al. 2018;Kasman and Duman 2015). The role of foreign direct investment (FDI) inflows in developing countries is of paramount importance in increasing economic growth. There is rich literature examining relationship between economic growth and FDI (see Anwar and Nguyen 2010;Bakhsh et al. 2017;Nadeem et al. 2020;Tsang and Yip 2007 (2018) find huge contribution of FDI to economic growth in the host countries. Increasing economic growth happens at the cost of environmental pollution. Developing countries relax environmental rules and regulations to attract FDI inflows. Most importantly, FDI inflows are made in energy and industrial sectors using natural resources namely oil, gas, and coal. Hassaballa (2014) shows bidirectional causality between FDI and energy consumption in the developing nations. Resultantly, FDI has positive impact on SO2 emission in the host country (Zhu et al. 2017). Thus FDI is related to pollution emissions thereby adversely affecting environment Shahbaz et al. 2015;Tang and Tan 2015). Nadeem et al. (2020) find a negative long-term relationship between FDI and SO2 emission. This happens because pollutant emission rises with the progression of the economy, reaches a maximum point, and then starts declining with further economic growth ). In addition to adverse impacts, FDI inflows are also considered to have an important contribution in promoting better environmental awareness and environmental friendly technology in the developing countries (Zeng and Eastin 2012).
The relationship between FDI and pollution emission through an increase in income or energy consumption in the existing literature is still debatable because of mixed findings with regard to directionality (Omri and Kahouli 2014;Saboori et al. 2012;Shahbaz et al. 2015). There are some studies showing unidirectional relationship between FDI and pollution emission (Jaunky 2011, Nasir andRehman 2011), others find bidirectional relationship (Halicioglu 2009;Soytas and Sari 2009), and there are some studies showing no relationship (Richmond and Kaufmann, 2006). Apergis and Payne (2009) and Zhang (2011) show mixed results of relationship between carbon emissions and income level of countries.
A huge empirical research is available on investigating the relationship between energy use and per capita income mainly through causality analysis at aggregation and disaggregation levels (Bakhsh et al. 2017;Bowden and Payne 2010;Ghali and El Sacka 2004;Luqman et al. 2019;Mehrara 2007;Yuan et al., 2008).  used dynamic simultaneous equation models to explore the causation link between pollution emissions, economic development, and FDI and found bidirectional correlation. Cleaner technologies, spillovers, productivity increases, and transformational abilities can all be brought by capital accumulation (Lee 2013).
Inconclusive relationship between FDI, pollutant emission, and economic growth can be because of regional differences in environmental regulation, given the different sources, processes, and economic factors. Zhao et al. (2017) asserted that environmental regulation and land use are significant factors on the spatial pattern of SO2 emission. Thus, the effects of FDI on pollutant emission can be heterogeneous across countries.

Data and source
Different sources of SO2 emission in Pakistan include energy, industrial processes, agriculture, land use change and forestry, and wastes. SO2 emission has increased from 775 thousand tons in 1994 to 1041 thousand tons in 2015 recording an increase of 34.3%. In order to explore the empirical relation among FDI, energy consumption, SO 2 emission, and economic growth, annual data was used for the period of 1975 to 2015. We were interested to consider the relationship between SO2 emission, energy consumption, and economic growth at different sector levels. However, the data on SO2 emission is not available on sector levels from 2011 to onward. Therefore, we planned the study at the country level. Table 1 provides information on definition of variables and sources of data. Different sources were used to collect data on the real gross domestic product (US$ million), gross fixed capital formation (US million dollar), capital intensity, investment (US$ million), industrialization (% of GDP), FDI (US$ million), and population density (persons living in one square kilometer). Data on energy (kt as energy use equivalent to oil) was taken from the World Development Indicators (WDI). Data on SO 2 emissions (giga tons) was taken from two sources. WDI was used to collect data from 1975-2010, whereas the remaining data from 2011 to 2015 was obtained from GOP (2018a, b). Standard method was used to generate missing values of SO2 emission for some years. For smooth and easy interpretation of the results, log form of data was used in the estimation. Copeland and Taylor (2003) suggest that emission can have three types of effects, namely scale effect, technique effect, and composition effect. Scale effect refers gross domestic product, whereas ratio of industrial output to gross domestic product is known as composition effect. Technique effect is defined as the ratio of pollution to industrial value added. When we examine the scale effect, technique effect, and composition effect, there is a problem of simultaneity. Because of the existence of correlation between unobserved disturbances across a simultaneous equations system as it is the case in the present study, 3SLS technique was employed to estimate environmental effects of FDI. Figure 2 shows steps used in 3SLS empirical method. 3SLS technique has advantages over single equation estimates and 2SLS (see Bao et al. 2011). Decomposition of pollution emission into three effects can be written as:

Empirical methods
Equation 1 is the identical equation showing pollution indicators resulting from FDI (Bao et al. 2011). Sulfur dioxide (SO2) is taken as pollution indicator. SO2 is dependent variable in Eq. 1, whereas explanatory variables include scale effect, composition effect, and technique effect.
A system of six equations is used to simultaneously estimate three effects of FDI on emissions and growth in Pakistan. The impact of FDI on the economic scale is measured by using Eq. 2 (Bao et al. 2011).
Here gdp shows gross domestic product of Pakistan, FDI is shown by fdi, stock of capital is represented by k, and polu is SO2 emission in Pakistan. a 0 to a 5 are parameters to be estimated. ε is white noise error term. We expect a negative sign for coefficient of pollution emission variable because of its damaging effect. FDI is directly related to economic growth as indicated by Cole et al. (2011), whereas physical capital in the country also increases indirectly with an increase in FDI (Zhang et al. 2004). Equation 3 shows estimation procedure of the indirect effect of FDI on financial advancement (Bakhsh et al. 2017).
One and two periods of lagged GDP values are used to measure the impact of economic growth and environmental Here, technique effect is shown by tech, fdi shows FDI, pd represents population density, eng is energy consumed in Pakistan, time is time period as a proxy for technology development in the country. β 0 to β 5 are parameters to be estimated. The increase in per capita income is expected to increase economic activity, which in turn leads to an increase in demand for a healthier environment. Further, an increase per capita income also implies allocation of resources for protecting the environment and abating pollution as indicated by Grossman and Krueger (1995) and Pao and Tsai (2010). Equation 5 is used to estimate this type of the effect while considering composition effect.
comp shows composition effect and it is measured as industrial output to GDP, k/l is capital to labor ratio, δ 0 to δ 4 are parameters to be estimated. Higher physical capital stock increases industrial output, which leads to rising level of SO2 emission. The effect of GDP on dependent variable can be of two types. First, it indicates higher level of industrialization and second, it represents the increased demand for cleaner environment when per capita GDP rises due to industrialization.
Before taking decision about FDI inflows in any developing host country, the developed countries consider time and country-specific factors. Equation 6 is used to examine the factors affecting FDI inflows in Pakistan.
System of equations has endogeneity issue; thus, we have taken one period lagged value of FDI in the model to avoid endogeneity. We include domestic investment, industrialization, and pollution emission level to see the effect of decision to invest in Pakistan.
The above system of equations shows mechanisms by which FDI may have influence on pollution emissions in the host nation. The direct effects of FDI on pollution emission can be measured by the coefficients, namely d 1 , β 1 + δ 1 , and δ 1 . FDI affects indirectly pollution emissions through its effect on GDP when greenhouse gas emissions and industrial composition vary with economic development. The indirect effect can be jointly measured by a1 and β 5 (δ 4 ). In addition, the indirect impact of FDI on toxic greenhouse gas emissions can be measured by the adverse effect on economic development, which has effects on both the scale effect and the composition effect. Furthermore, feedback implication of emissions (SO2 in the present study) on economic growth is also considered.
We also estimated the composite effect of foreign direct investment on the emission of SO2 in Pakistan using methods used by Bao et al. (2011). The total effect of FDI on SO2 emission can be estimated using Eq. 7: Scale effect is estimated using Eq. 8. All identities given in equation are defined earlier.
The technique effect is determined using Eq. 9: The composition effect of FDI on (SO2) emissions is found by using Eq. 10: a 4 , β 1 , and δ 1 are the direct parameters of the scale, technique, and composition effects of the FDI respectively. Capital accumulation emphasizes the importance of FDI inflows in the country. It is given by the coefficient of capital accumulations. Thus, capital accumulation accelerates economic activity through scale effect. The formula d 1 (a 1 + δ 2 ) measured indirect effect of capital accumulation of FDI in the country. The technical and compositional effects are used to show the level of economic growth and emissions of pollution. Scale effect is used to show the feedback impact of pollution emissions on economic growth. The scale effect can have impact on technique effect and composition effect (β 2 + δ 3 ), and we can see the total feedback effect by the measure β 2 (1 + δ 1 + δ 2 ). All notations used in the above equations are defined in Appendix Table 8.

Results and discussions
Foreign direct investment is associated with toxic greenhouse gas emissions and economic development in the host country. To examine these effects, we employed 3SLS. A system of six equations is considered in the present study; however, Eq. 1 is identity equation, so it is not possible to estimate it directly. The remaining five equations are measured to determine the impact of FDI on growth and emissions of SO2. Table 2 shows estimates of scale effect. We find that the coefficient of fdi is negative and statistically significant. Pollution variable, i.e. SO2, is an important factor in the growth of GDP and therefore is the key concern of the present study. Copeland and Taylor (2003) point out that pollution is an important factor of production in the country, besides labor and capital. An increase in SO2 emission increases economic growth significantly. Pollution emission in developing countries including Pakistan is positively related to economic growth. Ample evidence of empirical work shows that economic growth and greenhouse gas emissions are positively associated (Ali et al. 2019;Alvarado et al. 2018;Ma et al. 2016;Nadeem et al. 2020). Lax environmental rules and regulations in Pakistan induce firms to emit pollutants in the environment. Capital (k) is also statistically significant, implying that it has positive relation with economic growth. Table 3 shows the estimates of capital accumulation effect of FDI in Pakistan. We observe a positive association between FDI and capital accumulation which is statistically significant. Bakhsh et al. (2017) also found similar relationship in their study. Economic growth during the previous year also affects physical capital accumulation in the present year. To consider this effect, we have taken lagged value of gross domestic product in the capital accumulation effect (see Eq. 3). We find that previous year economic growth is significant and negatively related with capital accumulation in Pakistan. To see the influence of FDI on economic growth, it is estimated by using Eq. 3; we find 0.1287 total scale effect of foreign direct investment while considering SO2 emission. This scale effect is larger than estimate of Bakhsh et al. (2017) for CO2 emission.
Results of technique effect for SO2 emission and FDI are given in Table 4. A positive and significant coefficient of FDI implies that FDI inflows in the country increase marginal pollution damage. Foreign investors do not use environmental friendly technology in the developing countries because they are more interested in low cost production benefits (Lan et al. 2012). Thus FDI causes a significant environmental degradation in the host countries due to lax environmental monitoring and regulations. Countries with slow rates of increasing consumption activities are found having moderate economic growth rates and emissions (Díaz-Vázquez and Cancelo 2009;Luqman et al. 2019). Wang et al. (2016a, b) argue that environmental degradation is related with positive association between high energy consumption and emissions. Population density has a strong and substantial effect on SO2 emissions. This significant positive relationship  2152.9300*** ***and ** are levels of significance at 1% and 5%, respectively shows that a rise in population density also increases marginal environmental damage. Increased environmental damage associated with increasing population density warrants for pollution control measures to protect environment. Many studies also find this type of relationship (Doytch and Uctum 2016;Zhang and Zhou 2016). Gross domestic product (gdp) is negatively related with marginal pollution damage. Yavuz (2014) finds a long-term reverse effect of economic growth on pollution emissions and suggests that the economic activity itself would contribute to a decrease in emissions from pollution. Energy consumption is positively related with marginal environmental damage in the present study. Alvarado et al. (2018) and Cherni and Jouini (2017) explain a rise in fossil fuel consumption as being related to economic growth. Ample evidence shows an increase in energy consumption leading to pollution emission (Baek 2016, Behera and Dash 2017, Tang and Tan 2015, Zhu et al. 2016. The use of fossil fuels is the primary source of pollution emissions, namely SO2 and CO2. Since energy consumption is highly linked with economic growth, the developing countries can reduce its detrimental impacts through energy mix. Pakistan's dependence on thermal sources like coal, gas, and LNG has declined over the last few years, whereas the share of renewable energy is increasing over the years. The share of nuclear and hydel sources is also increasing in e n e r g y m i x ( G O P 2 0 2 1 ) . A l t e r n a t i v e E n e r g y Development Board (AEDB) in Pakistan has the responsibility to explore and develop renewable energy sources in the country. Several initiatives are taken to provide the induce environment for the sustainable generation and growth of alternative renewable energy in the country. Salient features of Alternative Renewable Policy 2019 include achieving a target of 20% on-grid capacity from alternative renewable energy by 2025 and 30% capacity by 2030, building solar and wind power projects through competitive bidding process, envisaging development of large scale alternative renewable projects throughout the country with active participation of the provinces, and forming the basis of on-grid capacity procurements through indicative generation capacity expansion plan. FDI raises pollution, demonstrating that the technological impact is incapable of regulating or reducing industrial pollution emissions, like SO2 as foreign firms use obsolete technologies in the manufacturing process due to weak environmental regulations. Although environmental laws for preventing or reducing pollution emissions exist, these laws are not implemented in true letter and spirit in order to attract FDI in the country. All this would have a detrimental longterm impact on economic growth (Jayanthakumaran et al. 2012).
Estimates of composition effects are reported in Table 5. Results show that a rise in FDI and a contribution of industrial production in the economy has positive and statistically significant impacts. The ratio of capital to labor is an important compositional impact indicator. Taking into account the k/l coefficient, we find that it increases the share of industrial production in the economy. It happens primarily because the manufacturing sector produces more toxic contamination as a result of capital-intensive production (Antweiler et al. 2001). Table 6 shows the estimates of determinants of FDI. Coefficient of investment implies that more investment by public and private sectors attracts more investment, possibly due to indication of conducive environment for production and cost advantages. Self-accumulation effects of FDI are estimated by taking lagged value of FDI. Statistically significant and positive coefficient shows that economic conditions of the country during the previous years attracted more FDI. Low cost of production and slack environmental standard operating procedures are the possible reasons. Like other developing countries, special economic zones (SEZs) are planned in Pakistan as an industrial policy to attract local and foreign direct investment in order to build, diversify, and upgrade  Chi 2 404.8800*** ***, **, and * are levels of significance at 1%, 5%, and 10%, respectively industries (GOP 2021). The SEZs are considered more efficient to attract FDI compared to other policy initiatives. However, SEZs are assumed to bring new technology in the industrial production with little impact on environmental degradation. Table 7 reports various effects of FDI on SO2 emission and these estimates are based on Bao et al. (2011) formulae discussed in the previous section. We find a negative estimate of total effect of FDI on SO2 emission. In the present study, the scale effect of FDI is positive. It indicated that an increase in FDI is causing a rise in gross domestic product. Our finding is consistent with that of Bao et al. (2011), who also find negative relationships between environmental pollution and FDI. The country with lax environmental rules, regulations, and management always exhibits negative technique effect even it is economically progressing (Mahmood and Chaudhary 2012). However, indirect composition effect is positive, implying that FDI inclusively increases the host country's share of industrial sector in GDP.
The present study has a few limitations. It considered data at country level to examine economic growth, energy consumption, and SO2 emission. It would have been great if SO2 emission at different sectors and or provincial levels were considered. Since the data on SO2 emission at sectoral and provincial levels is not available, future research should consider this limitation when the data becomes available. This study is limited to only Pakistan. Expanding the study at South Asian level can yield interesting findings. Taking only SO2 emission is another limitation of the study. In the future, all greenhouse gas emissions should be considered while examining the economic growth, SO2 emission, and energy consumption nexus.

Conclusions and policy implications
Economic growth and FDI are significantly related in the host country. FDI increases environmental pollution such as SO2 emissions. Results supported the pollution haven hypothesis, and nonexistence of a U-shaped relationship between FDI and SO 2 emissions. It points out that SO 2 emissions will continue to increase with increase in FDI inflows even after reaching threshold level in Pakistan.
Therefore, results indeed strongly demand the environmental friendly energy policy.
Results of technique effect showed that SO2 emission and capital had positive effect on gross domestic product, whereas FDI was negatively related to economic growth. Technique effect method showed that FDI, population density, and energy consumption are positively related with SO2 emission and supported the existence of pollution haven hypothesis. Furthermore, the results showed that technical effect is much larger than growth effect, indicating that FDI damages the environment of the country than speeding up the economic growth. In composition effect, FDI is positively related with industrial output, showing larger share of production of capital intensive goods. FDI in the previous years and investment by public and private sector in the host country were significant factors of FDI inflows. However, we found the negative impact of SO2 emission on FDI inflows.
In spite of environmental degradation linked with FDI inflows, the developing countries provide various incentives including lax environmental rules to foreign investors for speeding economic growth, building capital stock, and generating opportunities to the skilled and unskilled labor force. Findings of the study reveal that the recipient of FDI should emphasize the impacts of FDI on environmental quality because degrading environment costs the economy in the long run. The need is also to develop environmental friendly technology required during production processes. This necessitates the role of research and development activities. Recent estimates showing annual loss of PakRs450 billion owing to environmental degradation are important for policymakers and others to recognize the importance of environmental concerns in Pakistan. Pakistan has started taking strategic measures to control environmental degradation. Similarly private sector especially industrial sector has to play a crucial role in overcoming SO2 emission. Although regulatory bodies exist to monitor pollution emissions, there is still a need to strengthen these institutions for effective implementation of environmental rules and regulations.
Economic growth is function of four factors of production such as land, labor, capital, and organization. Copeland and Taylor (2003) introduced pollution as additional factor of production. Results of this study concluded that pollution is one of significant factors of production and helps to speed up the economic growth. From a policy perspective, this research has significant implications. First, the findings show that energy use has a long-term dramatic effect on SO2. As a result, in light of the Paris Agreement, policymakers should change their respective energy policies and shift toward renewable energy resources. Second, the findings revealed that environmental degradation surpasses economic growth, necessitating the imposition of FDI limitations and the restructuring of environmental regulatory laws. Finally, all domestic producers should be sensitized to less pollute the environment through an environmental capacity-building campaign.  Fdi fdi stands for foreign direct investment. An individual or a company residing in one country makes investment in another country is known as FDI. It can be in the form of establishing business operations or buying assets in the target country.
Inv inv stands for investment. It is the investment made by public and private sources in a country. This is used to attract FDI in the country.
Gdp gdp shows gross domestic product. It is the value of all final goods and services produced in a country during the specific period of time like during one year.
Pol polu indicates SO2 emission. This greenhouse gas is a colorless and it is emitted from burning of oil and coal and industrial processes containing sulfur.

Eng
Energy is the power we derive from using physical or chemical resources to be used for heating, cooling and lighting purposes or for running machines. k k shows gross fixed capital formation. It shows availability of building, machinery, equipment, etc. to be used in the production process. It is among factors of production.
k/l It represents capital intensity. It shows the amount of capital in relation to other factors of production like labor, land, etc.

Ind
It is used for industrialization in the country. This shows the transformation of economy from primary goods production to manufacturing goods. We can say transformation of economy from agricultural goods production to manufactured goods.

Pd
It represents population density. This shows number of persons living per square km in the country.

Tech
It shows technique effect. It is the ratio of pollution to industrial value added in the country.

Scale
It is scale effect. It is used to see the impact of FDI on gross domestic production.
Comp comp is used to represent composition effect. It is estimated by taking ratio of industrial output to gross domestic product.