**5. Conclusions**

With the implementation of the Paris Agreement signed by the HICs and NICs, the results show different effects on each country. Some countries have efficiently utilized their renewable energy resources, but some are not doing so, while others have maintained their levels and it seems they are not moved by the Agreement terms. The use of the undesirable model has successfully calculated each country's position in terms of the utilization of their resources while considering greatly a very bad factor, which is the carbon emission.

Knowing the economic capabilities of the highly industrialized nations, the results show that Russia is the least efficient among them. A mean efficiency score of 0.1801 verifies the country's lack of attention to its renewable energy program. In fact, in 2015, renewable energy only comprised 20% of the total installed power capacity, which mostly came from bioenergy and hydropower. The country must focus on other sources such as geothermal, wind, and solar photovoltaic (PV) [42]. Canada, with a mean efficiency score of 0.4221 manages to place second to the last among the HICs. A report from 2015 reveals that Vancouver alone sourced its 69% of energy from fossil fuels and 31% from renewable sources [43]. The situation in Japan is a little different, as due to the nuclear disaster from a major earthquake years ago, the country aims their energy to be completely 100% supplied by renewable energy, especially in their local regions, by 2020 [44]. Additionally, a report states that there is a decrease in the use of fossil fuels in 2018, 78% from 81% share. However, a dependency of the country on nuclear energy still exists and is evident from the increase of its generation from 2.8% to 4.7% of the same year [45]. This can be one of the reasons for Japan's efficiency score drop from 2016 to 2018 and its acquisition of an average of 0.6671 throughout the studied period. The same goal is seen in Italy's energy program. With the introduction of the 20-20-20 EU goals, the Italian governmen<sup>t</sup> aims to decrease the emission of greenhouse gases by 20%, improve energy savings to up to 20%, and attain a 20% generation of renewable energy by 2020 [46]. The result of this study shows that Italy has quite improved its renewable energy utilization with increased and maintained efficiency scores from 2015 to 2018. While reaching a perfect efficiency of 1.0 in 2018, the average score becomes 0.7108 due to lower scores acquired from previous years. These scores imply that the German governmen<sup>t</sup> has effectively improved its renewable energy utilization throughout the study period. Gaining an efficiency score of 1.0 throughout the study period, France, the United Kingdom, and United States, have been consistently utilized their renewable energies even before or after the Paris agreement. Their energy and sustainability programs can be a benchmark for other HICs that aim to improve their use of own renewable resources.

In contrast with the efficiency scores of the HICs, none from the list of NICs is able to ge<sup>t</sup> a consistent perfect efficiency. In spite of this, Indonesia manages to acquire perfect efficiency during the period of 2015 to 2018 rising from low-e fficiency scores. Getting the highest mean e fficiency score of 0.7645 among the NICs, this can be due to the fact that the country is considered to have the biggest resources of geothermal energy in the world [47]. South Africa demonstrates an interesting e fficiency score as it can be seen that during the period of 2013 to 2014, the country was able to gain a perfect efficiency. This performance is attributable to their noticeable increase in renewable energy production in previous years [48]. However, this performance by South Africa was not maintained since the country's e fficiency dropped tremendously in succeeding years resulting in an average e fficiency score of only 0.5185. Mexico, Turkey, Brazil, Malaysia, and Thailand are performing consistently at a low-level efficiency, with a minimum of 0.2278 and a maximum of 0.2734 mean scores. These results can be caused by the following factors: Mexico's dependency on fossil fuels wherein 85% of their total power was produced in 2012 [49]; Turkey's power sector industry as the biggest contributor to CO2 emissions and a high reliance to coal-powered electricity generation accounting to 37.2% of their total electricity [50]; the old structure of Brazil's energy sector which limits their capacities in handling renewable energy demands, and other political factors that hinder the country's development to successfully execute their national energy programs [51]; and, Malaysia and Thailand's high proportional use of crude oil, coal, and natural gas, giving less importance to renewable energies [52,53].

The HIC group is expected to keep increasing the e fficiency level by 7.76% from the past period to the projected period. This can be due to the expected development to the future renewable energy developments by Japan and Italy. However, the NIC group is expected a bit decline in the average efficiency with negative 1.23% due to the e fficiency drop performance by Turkey and Mexico along with the low e fficiency performances by other countries.

Future implications from the result of combined data gathered using grey prediction and the undesirable DEA model shows that the HICs and NICs will continuously follow the trend of the efficiency scores for all these countries with Germany joining the group of perfectly e fficient ones together with Indonesia, France, United Kingdom, and the United States. All other HICs and NICs will perform otherwise, especially if they will not concede significant importance to renewable energy. Most of the developments in the use of the renewable energy begin with the improvement in the political and legal structure surrounding its use and regulation, providing incentives or tax exemptions to private power producers to shift in using the renewable energy. Countries can also build strong partnerships with non-governmen<sup>t</sup> and international organizations that are focusing on providing assistance to countries that want to achieve sustainable and renewable energy production. Strengthening the support to the research and development sector to determine the suitable sites to deploy renewable energy sources such as solar, wind and wave energy. Long-term strategic plans for renewable energy development that will also consider the economic, social, and environmental impacts to the country.

This study contributes as a method to mathematically evaluate the energy utilization e fficiency of HICs and NICs based on available public data. The DEA undesirable model treatment of the CO2 emission factor as a less important factor made this study di fferent from the others. Hence, this model is the most appropriate method to evaluate the energy sector that uses data with undesired factors. The results of this research may guide each country for improvement of its production and consumption towards sustainable renewable energy development. This can also help policymakers, governmen<sup>t</sup> agencies and the energy sectors to address the problem in the existence of bad outputs such as CO2 emissions. For future studies involving quantitative measurements, the authors recommend combining DEA with a qualitative evaluation approach such as the one described in Wang et al. [54], which uses the fuzzy analytical hierarchy process for analysis to improve studies of this kind.

**Author Contributions:** Conceptualization, C.-N.W. and H.T.; data curation, H.T. and D.H.D.; formal analysis, H.T. and D.H.D.; funding acquisition, C.-N.W.; investigation, H.T. and D.H.D.; methodology, C.-N.W., H.T., and D.H.D.; project administration, C.-N.W.; writing—original draft, H.T. and D.H.D.; writing—review and editing, C.-N.W. and H.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was partly supported by National Kaohsiung University of Science and Technology, and project number 108-2622-E-992-017-CC3 from the Ministry of Sciences and Technology in Taiwan.

**Acknowledgments:** The authors appreciate the support from Taiwan National Kaohsiung University of Science and Technology, Philippines Technological University of the Philippines Taguig, and Taiwan Ministry of Sciences and Technology.

**Conflicts of Interest:** The authors declare no conflict of interest.
