Economy and Pollution by CO2 Emissions - Kuznets Environmental Curve in Asia Pacic Countries

These authors contributed equally to this work Abstract The economies of the Asia Pacific region (APEC) are among the most dynamic, with the highest levels of commercial activity and environmental degradation in the world. One way to study trends in the relationship between the economy and pollution is the Environmental Kuznets Curve (EKC). This perspective assumed that in the long-term, higher levels of wealth would be achieve lower pollution levels. The aim was analyze whether the EKC behavior occurs for the APEC economies. The method used was an econometric model for carbon dioxide (CO2) emissions and GDP. The results showed stationary unit root in first differences, this suggest the presence of EKC behavior was explained with a high probability level that the GDP coefficient would reach an inflection point to become negative and significant.


Introduction
The Asia Pacific region is better understood, if one considers that the economies of the 21 countries that make up the Asia Pacific Cooperation Forum (APEC) are found in it which cover 47% of international trade; represents 39% of the world's population and contributes around 60% of GDP, (APEC, 2019). This explains why the economies of this region have registered sustained economic growth for more than 20 years (World Bank 2019).
Economic growth has come at the expense primary goods consumption, natural resources and the deterioration of the environment. The region is one of those that generate the largest ecological footprint of the planet, consumption of resources over the bio-capacity to regenerate them, in Wakernagel and Ress, (1998) concept. The ecological footprint of the countries located in the Pacific Basin predominantly exceeds their bio-capacity, that is, they have ecological debt, (Footprintnetwork 2021;Wackernagel & Rees 1998).
This situation has led to concern about environmental deterioration, considering several sections on ecological, environmental and energy aspects as challenges that humanity will face in the coming years, in the Millennium Development Goals (Motala, et al. 2015). In such a way that one of the great planetary challenges; sustainable development; Recognized at the Rio Summit (Cumbre de Rio) since 1992, almost five decades after its completion, it still maintains as one of the pending subjects of humanity, the reduction of environmental deterioration.
In this way, a boost in environmental protection policies is to be expected, even in commercial and economical aspects. However, the economic integration policies of APEC have focused in aspects such as: cooperation, tariff reduction, education, science and technology, new communication technologies development, among other items. The integration has been in economic, commercial and investment terms, remaining a perspective of domination of large economies such as China in Asia and North America (Hernández 2013).
Then, environmental policy is still a secondary issue in APEC integration agenda, however the last years of twenty century, were included declarations against environmental deterioration, as an example some APEC countries have suggested lists and classifications for environmental goods in the non-agricultural market access negotiation group, identifying more than 109 products to be considered in the trade liberalization agreement, emphasizing in environmental goods according to their end use (Martínez, 2006).
The main concern of APEC's economic integration was economic growth from a neoclassical paradigm, which relies on technological innovation to reduce long-term negative environmental impacts and maintain growth rates (Guilló & Magalhaes 2018). But what has been environmental pollution consequences? Or How have long-term economic growth and environmental impact behaved in APEC countries? Considering GDP and CO2 emissions (per inhabitant) as an approximation to these variables and taking as a reference the Environmental Kuznets Curve (EKC).
There is an implicit idea in the EKC model that economic development is a process that progresses in phases to consolidate itself. An initial stage development and environmental degradation are incipient. Second phase of development is accompanied by high rates of environmental degradation. In a third stage, as economic processes incorporate technological improvements, materials substitution and greater social demands, they will promote changes in the laws and environmental management, with which the environmental impact would tend to decrease with increases in incomes and development.
The aim of present is to verify the trend in the behavior and interaction between the variables GDP and carbon dioxide (CO2), per inhabitant in APEC countries. To achieve the aim, the article is organized into five sections. The first describes the relationship between economy, development and environmental quality, discussing the theoretical positions, relevant background work for the study in the countries of the Pacific Basin. The third part describes the econometric techniques, for the fourth part to describe the results, discuss them and finally the conclusions are presented.

Theoretical aspects of EKC
The relationship between the economy and the environment has been emphasized since the 1970s. After the Club of Rome declared that the greatest challenges for humanity in the last quarter of a century would be: 1) degradation of natural resources in quality and quantity; 2) increased demand for energy, supplies and food; and 3) an increase in income inequality, (Meadows, et al. 1972 y Mesarovic andPestel, 1975 A decisive factor in the analysis of the EKC is the income elasticity in relation to CO2 emissions, which implies that reaching the determined level of income would produce changes in the preferences of consumers who would be willing to spend a greater proportion of their income. income in environmental quality, as if it were a luxury good (Baldwin, 1995;Selden & Song, 1994).
Some less optimistic working papers highlighted that empirical repercussions could approach to theoretical aspects to EKC, without significant repercussions on economic policy, Stern (2004), Selden & Song (1994) y Wagner (2008. It is argued that the behavior of the EKC, despite finding the expected ratio per inhabitant, indicates that the global emission of pollutants will continue to increase (Selden & Song, 1994).
There is skepticism of both theoretical and empirical works, for the proposition of a simple and predictable relationship between pollution and per capita income. However, there may be an inverted U shape in relation to urban environmental concentrations of some pollutants and income, although this should be more rigorously tested in time series or panel data methods (Stern 2004). On the other hand, other authors comment that the studies referring to EKC are wide and varied, at the same time that they have had applications in severals economies of the world, despite being such a widespread analysis with econometric tools, but they might not be adequate, since they can lead to contradictory conclusions (Wagner, 2008). Alternatively, some studies propose the use of non-parametric techniques such as polynomial splines, which are presented as an alternative method when decomposing the time series and examine the EKC using confidence bands and simulations (Wang, 2011).
A study for 182 countries, applying panel data and cross section, found EKC hypothesis is acceptable in both approaches, although inflection points in the cross sections are not entirely reliable. It is also highlighting the importance of adequately addressing problems such as heterogeneity, structural changes and spatial interaction and the issue of causality was not reviewed (Atwi, Barberán, Mur & Angulo, 2018).
Particularly for some countries in APEC, studies can be seen that demonstrate the EKC with different methodological approaches, which are shown in Table 1. Although it is a highly studied hypothesis for different countries and regions of the world, the techniques and variables used give an important diversity to the analyzes. The studies presented in Table 1 are just some studies for the region in the last decade, it can be seen that the use of macroeconomic variables such as the emission of Greenhouse Gases (GHG), (mainly CO2), the GDP, GDP^2 and Renewable Energy Consumption. It is emphasized that the work of Al-Mulai, Saboori and Ozturk (2015) CAK is not demonstrated, whereas it is only partially demonstrated for 16 Asian countries (Wen-Cheng, 2017), because that non-fossil energy consumption was decisive, an aspect related to energy supply policies.
According to the above, there are different arguments for and against the validation of the EKC hypothesis, so it is important to continue generating empirical evidence that contributes to the generation of knowledge that provides elements on the relationship between economic activity and environmental quality.
From a reflective perspective EKC approaches contain arguments that serve as theoretical support for the hypothesis, but that are not largely operative in the economic, social and institutional circumstances in developing countries. In particular, inequality in income distribution and the fragility of the institutional framework in environmental matters weaken the main foundations that support this hypothesis, making it less important to wait and grow to achieve improvements in environmental quality (Zilio, 2012).
Although conclusive results are not presented, however, most studies show similar trends.
While it is worth noting that the variants and the assumptions of the models have differences, which influence the results obtained. The EKC continues to be a reference for interpreting relationship and behavior between environmental quality, economic growth and development styles. With analysis, EKC could be show phenomena such as technology advanced, structural composition of GDP, environmental regulations and social preferences. But is necessary to know the context and other issues about the countries to study.

The theoretical model of the Environmental Kuznets Curve
It is possible to define a parametric form of the EKC hypothesis defined as the relationship between CO2 pollutant emissions and income per inhabitant, it can be described in a model as follows:  (Balsalobre, et al. 2016).

Empirical model
Starting from equation [1], the effect of the income level on CO2 has been estimated using an autoregressive panel data cointegration model, with distributed lags. The expression takes the following form: If the variables are cointegrated they tend to move together maintaining a long-term equilibrium, that is, a stationary relationship between variables, the cointegration between variables implies, according to the Granger representation theorem, (Granger 1998) that the system admits a representation in the form of error correction mechanism (Engle & Granger, 1987). For this reason, instead of estimating equation [3], its parametric transformation has been used: A model is established with an error correction mechanism, all the variables are expressed in differences and there is a term that collects the adjustment of the deviations of the dependent variable with respect to its long-term equilibrium value. The existence of an error correction mechanism representation is not subject to the most common problems of a regression, (given the cointegration) since all the variables that enter the equation are stationary.
The test of the cointegration between emissions, income and income squared can be If the parameter associated with the error correction mechanism, using the residuals of the cointegration equation, is statistically significant "t" it can be stated, according to the Granger representation theorem, that the series are cointegrated (Granger 1998). The evidence in favor of the EKC requires that the coefficient β1 is positive and the coefficient β2 is negative, since in this way a quadratic relationship in the shape an inverted U is obtained.

EKC applied in APEC Contries
The behavior of the variables GDP and CO2, per inhabitant, maintains a behavior with variable growth rates throughout the more than five decades shown in Figure 1. It can be highlighted that in terms of graphic representation, CO2 emissions are more stable than GDP, that is; less variation in growth rates. In the same sense, GDP increases to a greater extent, which suggests the CAK curve, however a reduction in CO2 emissions has not yet been noticed.

1. Econometrical analysis CO2 and PIB (per capita) (
In the analysis of time series variables, it is important to know the order of integration and verify if there is cointegration between the variables, to avoid obtaining spurious results. The econometric literature suggests that unit root tests on panel data have greater power than time series unit root tests. According to Baltagi (1995), when combining the time series with the cross-sectional data, there is a greater number of observations, more degrees of freedom, more variability, less collinearity and greater efficiency. However, it is important to first apply some cross-section dependency test before unit root tests with panel data (Pesaran, 2007).   -5.449*** ∆PIB C -13.710*** -5.867*** ∆PIB cuadrado C -13.326*** -5.354** Note 1: The tests used were: Augmented Dickey Fuller Cross Section (CADF) and Im, Pesaran & Shin (2003) and Cross Section (CIPS) Pesaran (2007).
Note 2: *** denotes the rejection of the null hypothesis at the 1% level of significance, **5% level of significance. C denotes constant and CT denotes constant and trend. Table 3 shows the results of the new unit root tests with panel data that are consistent in the presence of dependency on the cross sections. The evidence indicates that the variables present unit roots in levels, but are stationary when taking the first difference at a significance level of 1%. To test the presence of an equilibrium or long-term relationship between the integrated variables of the same order, in this research two cointegration tests with panel data are used: the Pedroni (2004) and Fisher tests proposed by Maddala and Wu (1999). Note: *** and ** denote rejection of the null hypothesis of no cointegration at 1%, and 5% significance, respectively.
In the results of the Pedroni test shown in Table 4, the null hypothesis of no cointegration is rejected for most of the statistics with a significance level of 1%. This is confirmed by the results of the Kao and Fisher tests (Tables 5 and 6). In the case of Fisher's test, it is indicated that at least two cointegration relationships exist between the analyzed variables, that is, there is a long-term equilibrium relationship between the three variables; CO2, GDP and GDP2 (Table 4).
According to Table 7, with the FMOLS and DOLS estimators all the variables are statistically significant. The long-term elasticity of per capita GDP is positive, which implies that as economic activity increases, carbon dioxide emissions also increase.
Furthermore, the hypothesis of the environmental Kuznets curve is supported by the results for these countries, since the coefficient of GDP is positive and that of GDP squared is negative, both statistically significant at 1%. Note: ***, **, and * denote statistical significance at the 1, 5, and 10% level, respectively.
According to Tables 7 and 8, the results show that the causal relationship goes from economic activity to emissions in the long term at a significance level of 1%. These results are also confirmed in the short term, there is a unidirectional causal relationship that goes from economic activity to polluting emissions. The above can be interpreted as that economic activity contains important information that helps to better predict the behavior of emissions.
V. Discussion of results.
The econometric tests implemented to check if the CAK shows a quadratic trend for the variables GDP and CO2. In other words, the hypothesis of the relationship between economic growth and environmental quality, measured from carbon dioxide emissions, (both variables relativized per inhabitant) per unit of product is confirmed. Therefore, it can be affirmed that, in the long term, pollutant emissions per unit of production maintain a tendency to grow to a lesser extent than the growth of the economy. This situation would suggest that productivity is moving towards a trend of less negative environmental impact, with regard to CO2 pollution. The implications this has would be a function of the degree of technological development and the prevailing economic structure in APEC.
Although they have reduced emissions per unit produced, it must be taken into account that pollution as a whole is greater, since the wealth generated or GDP is also greater, The difference in the increases between the two variables suggests that the Asia Pacific countries are improving aspects such as regulatory frameworks and technologies, although there is still a heterogeneous rate. CO2 emissions per unit of GDP / inhab, are increasingly efficient, which can be attributed to technological innovation, as well as increasingly rigid regulations, it is also appreciated that this relationship is relatively slow, so that It would be necessary for innovations to be more frequent, of greater scope for their diffusion and above all of greater impact. In the analysis of Figure 4, it shows that the trend in a comparison between countries for the year 2013 is growing and positive, that is; the higher the GDP, the higher the polluting emissions.
The productive structure of the economies analyzed has a close and determining relationship in terms of the behavior of the relationship between the variables analyzed.
That may be the situation that Hong Kong has very low emissions and high-income levels, given that its productive structure is predominantly tertiary, focused on activities mainly of an administrative and financial nature.

Conclusions
The formulation of the initial hypothesis on a long-term trend for the economies located in the Asia Pacific region, assumes that a relationship could be found between the variables GDP per inhabitant and CO2 per inhabitant. The results suggest that the series can cointegrate, so that both have information that allows knowing and predicting their behavior. Although, the assumption that both variables adjust to a U-shaped relationship in the long term, the heterogeneity of the economies analyzed must be taken into account, as well as their productive structures. With the available information, it was possible to corroborate that the analysis of panel data and cross-section for the period analyzed indicates that valid and unbiased information for the EKC was produced.
The importance of the Asia Pacific region lies, among other things, in the growing trade flow in the region driven by the economies of China and the United States. Therefore, it is relevant to identify the terms and conditions adopted by international trade in this macroregion of the planet. In the event that the commercial relationship integrates various aspects such as: environment, flow of people, technology transfer, financing for development in strategic areas of countries with greater backwardness. The information on the way in which development is configured will be transcendent and for this, this article showed the relationship between two key elements for the future commercial relationship: income and polluting emissions. One of the greatest challenges for the region will be precisely to reduce the disparity between the economies in order to continue fostering the commercial expansion of the APEC region.
One of the elements that remain pending and probably for the next few years will be of vital importance, it will undoubtedly be the energy policy of the countries. In the case of the Asia Pacific region, which at the moment is heterogeneous the use of free sources of fossil fuels, the trend is towards decarbonization and the promotion of renewable and alternative energies (wind, solar, hydroelectric, nuclear and tidal wave, etc.) which is likely to drive a reduction in pollution.
Finally, the growth of GDP per inhabitant has advanced in terms of efficiency, emitting less CO2 per unit of product, however, in absolute values this is reflected in air pollution of capital magnitudes, this is multiplying emissions by three in a few more than 50 years.
However, China and the United States participate with 70% of the emissions of the APEC region, although adding Russia and Japan the figure grows to 83% of a sample of APEC region, as well as the wealth generated. In APEC there is a high concentration in few economies that generate the highest proportion of both variables (GDP and CO2). It is because of this situation that these countries should be the ones promoting the use of cleaner technologies and stricter environmental regulations, which will have implications for areas such as trade and regional and world production.

Conflicts of Interest
The authors declare no conflict of interest.