I. INTRODUCTION

In contemporary times, sustainable development is the key to economic growth and development. Rising oil prices and perpetually increasing emissions of greenhouse gases due to burning fossil fuels require urbanization to take into account the minimization of the use of resources and increasing energy efficiency (Chau et al., 2015; Liao et al., 2015). Unconstrained extraction of fossil fuels coupled with increasing rates of population growth have detrimental effects on human life, such as climate change and associated diseases. Energy usage and energy consumption need to be viewed in close conjunction with quality of life to understand their net effects. Various indicators of the quality-of-life index are observed to be highly correlated with energy and electricity consumption per capita (Mazur, 2011). Energy consumption is closely associated with land use patterns, transportation, and population density (Liu et al., 2016). Moreover, access to energy is a pre-condition for employment opportunities, access to health and education services, and consequently poverty reduction (Nadimi & Tokimatsu, 2018). Al-Mulali (2016) reports an improvement of 70% in the quality of life in 198 countries, irrespective of their differences in income. Levels of energy consumption are also observed to play a significant role in the determination of the Human Development Index and the level of sustainable income (Van Tran et al., 2019). Yumashev et al. (2020) find that the volume of energy consumption not only affects the human development index in a particular country, but is also an important factor in determining the level of sustainable development. A recent study by Oluoch et al. (2021) finds that renewable energy consumption is significant and positively correlated with the gross domestic product (GDP) per capita and the education index in the long run, as expected, whereas renewable energy consumption is significant and negatively correlated with CO2 emissions per capita and the life expectancy index in the long run.

Specifically, the Asia-Pacific Economic Cooperation (APEC) region comprises 39% of the world’s population and 54% of the worlds’ GDP. In addition, the APEC region alone was accountable for 58% percent of the global primary energy supply and 57% of energy consumption around the world in 2016. Moreover, this region is also observed to include the four largest energy users in the world, that is, China, Japan, Russia, and the United States (APEC, 2019). With rising levels of industrialization and increasing rates of migration from rural to urban areas, energy consumption is expected to surge in the coming years. The final energy demand of APEC continues to be dominated by fossil fuels, which comprises 64% of final energy demand. However, several factors, such as active government policies for clean energy, price drops in renewable technologies, and cheaply available natural gas, aid in the declining demand for coal in this region. Electric energy access, subjective well-being indicators, local capabilities, and participatory decision-making processes are quintessential in assessing the interaction between electric energy consumption and the quality of life (Castro-Sitiriche & Ndoye, 2013).

This study contributes to the energy literature in two ways. First, we contribute to the literature in terms of correlating governance, transparency, and corruption aspects to the emission intensity of the APEC region. Second, the implications of energy consumption–pollution dynamics have been analyzed by utilizing quantile and threshold regressions, which imply examining regional emissions at different intervals and utilizing governance as a threshold indicator in the region. This paper seeks to contribute to the energy literature differently from the standpoints of the quality of life and governance, since both factors help incentivize sustainable energy consumption and to better control emissions.

The remainder of the paper is structured as follows. Section II discusses the potential data, empirical methods, and findings of the paper. Section III concludes the paper.

II. DATA, METHODS, AND RESULTS

A. Data Collection

This study considers the data on carbon emissions, energy consumption, transparency, the GDP, employment, travel, Internet availability, human development, life expectancy, and corruption for 21 APEC economies from 1995 to 2019. We use the natural logarithm of these measures to avoid spurious regression issues. Data on transparency are obtained from the World Bank’s World Governance Indicators database, while the information on the rest of the variables obtained from the APEC database. Table 1 reports the details of the variables. The data for all the variables in the model are on an annual basis.

Table 1.Definition and Measurement of Variables
 Variables Full Name Definition Unit of measurements CO2 Carbon emissions Emission of pollutants in terms of carbon dioxide. Metric tons EC Energy Consumption Total amount of consumption of different forms of energy. Exo-joules Trans Transparency Transparent governance with predictable enforcement. In index score from 0 to 1 GDP Gross Domestic Product Total final value of goods and services produced in an accounting year. In constant US dollars LFPR Labour force participation Total employment In terms of percentage of total population Int Internet Accessibility of internet services to the people. per 100 people TT Travel Inflows of tourists In numbers HDI Human Development Index Measuring country’s achievement in social and economic dimensions. In index score from 0 to 1 Life Life expectancy It is the life expectancy at bierth in terms of years. In numbers Corr Corruption Perspectives on different forms of corrupt practices. In index score from 0 to 100

Notes: This table shows definitions and measurements of the variables used in our analysis. Author’s own compilation of information.

B. Empirical Framework

The study uses the following empirical model:

\begin{align} {CO2}_{{it}} &= \alpha_{1}{EC}_{{it}} + \alpha_{2}{Trans}_{{it}} + \alpha_{3}{GDP}_{{it}} \\ &\quad + \alpha_{4}{LFPR}_{{it}} + \alpha_{5}{Int}_{{it}} + \alpha_{6}{TT}_{{it}} \\ &\quad + \alpha_{7}{HDI}_{{it}} + \alpha_{8}{Life}_{{it}} + \alpha_{9}{Corr}_{{it}} \\ &\quad + \varepsilon_{{it}} \end{align} \tag{1}

where $$CO2$$ is the main outcome variable and $${EC},$$ $${Trans},$$ $${GDP},$$ and $${LFPR}$$ are the major explanatory variables. The rest of the variables include control variables, such as internet availability (Int), travel human development (TT), life expectancy (Life), and corruption (Corr), and I and t denote the 21 APEC economies and the years 1995 to 2019, respectively. The term $$\varepsilon_{{it}}$$ is the stochastic disturbance. The major objective of this study is to estimate $$\alpha_{1},$$ $$\alpha_{2},$$ $$\alpha_{3},$$ and $$\alpha_{4}$$ in model (1). We expect energy consumption and the GDP to contribute positively to energy consumption, and transparency to impact emissions negatively. Other control variables, such as labor force participation, human development, and corruption, are expected to impact carbon emissions positively over the years.

We estimate equation (1) by utilizing ordinary least squares estimates followed by quantile regression (Koenker & Bassett, 1978). Quantile estimates enable us to examine the effects of carbon emissions in the region at different intervals. Furthermore, following Hansen (1999), we estimate the threshold regression model by utilizing transparency as the threshold variable.

C. Results

The regression results are displayed in Table 2. The coefficients of energy consumption are statistically significant and positive. There is further evidence of a positive association between corruption and emissions in the region, while improvements in transparency in governance notably help decrease pollution. Further robust evidence is obtained from the positive relation between human development, labor force participation, and carbon emissions. Overall, the trend reflects unsustainable growth in the region coupled with institutional failures.

Table 2.Regression Estimates
 CO2 I II III IV EC 0.156* (0.011) 0.291* (0.047) 0.166* (0.011) 0.242* (0.045) Transparency -0.116* (0.023) 0.079* (0.028) -0.122* (0.024) 0.096* (0.026) GDP -0.057* (0.020) 0.005 (0.004) -0.066* (0.022) 0.001 (0.001) LFPR 0.423* (0.175) -0.282* (0.117) 0.498* (0.178) -0.009 (0.008) Internet -0.010 (0.009) 0.016* (0.007) 0.102* (0.025) 0.055* (0.009) Travel -0.088* (0.019) -0.008 (0.007) -0.315* (0.058) 0.037** (0.020) HDI 3.416* (0.133) 0.869* (0.177) 3.164* (0.136) 1.273* (0.174) Life expectancy -3.720* (0.445) -2.518* (0.510) -3.634* (0.538) -0.959** (0.521) Corruption 0.108* (0.015) 0.009 (0.007) 0.101* (0.035) 0.076* (0.016) Constant 4.508* (0.705) 5.473* (0.915) 4.398* (0.755) 1.607*** (0.697) Time Effect No No Yes Yes Country Effect No Yes No Yes R2 0.825 0.901 0.846 0.903 F test 271.24* 609.86* 81.72* 576.78*

Notes: This table shows the regression estimates. (*), (**) and (***) indicate the levels of significance at 1%, 5% and 10% respectively.

Next, we focus on the variance of emission intensity as a measure of emissions, where growth and institutional factors can have different levels of appeal on the empirical front. We find that the coefficients of energy consumption are positive and significant at the 1% level (see Table 3). As mentioned above, we also note that corruption and declining institutional quality result in higher emissions.

Table 3.Quantile Regression
 CO2 I II III EC 0.223* (0.019) 0.162* (0.014) 0.116* (0.014) Transparency -0.102* (0.040) -0.066* (0.029) -0.143* (0.029) GDP -0.032 (0.031) -0.018 (0.016) 0.005 (0.005) LFPR 0.670* (0.300) 0.055 (0.047) 0.685* (0.216) Internet -0.037 (0.029) -0.019 (0.017) -0.012 (0.011) Travel -0.064** (0.034) -0.070* (0.024) -0.086* (0.024) HDI 2.959* (0.228) 3.133* (0.164) 3.985* (0.173) Life expectancy -1.398** (0.762) -3.507* (0.550) -3.945* (0.549) Corruption 0.125* (0.025) 0.102* (0.018) 0.050* (0.018) Constant -0.086 (0.083) 4.942* (0.872) 3.838* (0.370) Quantiles 25th 50th 75th Pseudo R2 0.638 0.612 0.593

Notes: This table shows the quantile regression estimates. (*), (**) and (***) indicate the levels of significance at 1%, 5% and 10% respectively.

There is little evidence of a positive association between human development and employment and pollution across quantiles. This necessarily suggests that rising human development in terms of income growth leads to more lavish expenditures and negates green growth. Other empirical evidence on growth and internet availability do not show any substantial significant relation with the rate of emissions.

Table 4.Threshold Regression
 CO2 I II III IV EC 0.276* (0.037) 0.271* (0.032) 0.243* (0.022) 0.281* (0.047) Transparency -0.121* (0.024) -0.111* (0.020) -0.101* (0.013) 0.015 (0.014) GDP 0.007 (0.006) 0.007 (0.006) 0.005 (0.004) 0.008 (0.007) LFPR 0.224* (0.086) 0.214* (0.074) 0.188* (0.074) 0.243* (0.111) Internet 0.012* (0.006) 0.011* (0.004) 0.011* (0.003) 0.016* (0.007) Travel -0.002 (0.002) -0.002 (0.002) -0.001 (0.001) -0.008 (0.006) HDI 0.321* (0.140) 0.301* (0.122) 0.281* (0.102) 0.864* (0.175) Life expectancy -1.100* (0.407) -0.993* (0.367) -0.976* (0.359) -2.556* (0.504) Corrupt -0.005 (0.004) -0.005 (0.004) -0.005 (0.004) 0.008 (0.006) Constant 2.498* (0.729) 2.122* (0.663) 1.896* (0.557) 5.121* (0.899) Threshold variable Transparency Transparency Transparency Transparency Trimming 0.01 0.01 0.01 0.05 0.05 0.05 0.10 0.10 0.10 0.25 0.25 0.25 R2 0.505 0.493 0.481 0.428 F test 89.24* 89.10* 86.70* 40.40*

Notes: This table shows the threshold regression estimates. (*), (**) and (***) indicate the levels of significance at 1%, 5% and 10% respectively.

Considering transparency as the main threshold variable, we propose a single-threshold model with emissions as the outcome variable. Our analysis suggests that energy consumption in the existing phase is positively influencing emissions. The GDP exhibits no significant impact on emissions due to differences in environmental standards and regulations (see Table 4). Improved human development and employment exert positive impacts on carbon emissions, as evident from existing and previous models. Surprisingly, we note two areas of divergence from our earlier findings. First, considering transparency as a threshold variable, the impact of corruption on emissions is almost nonexistent. Second, we find two different results in terms of a positive and a negative significant association between human development, life expectancy, and emissions. This is solely because the APEC region has a large young to middle-aged working population spending more on luxuries.

III. CONCLUSION

In this study, we test the asymmetric impacts of growth, energy consumption, human development, and institutional factors upon the carbon emissions for the APEC region. In terms of mechanisms, the results suggest that regional economies need to achieve sustainability, especially in terms of human development and employment. Our results surprisingly reveal that a rise in human development and employment results in higher emissions. In addition, improvements in governance and facilities such as tourism are essential to attain sustainability in the long run. Our overall findings show that reforms at the institutional level and other developmental fronts need to ensure sustainability to control pollution, in addition to reforms in the energy sector.