سال انتشار: 2022

رشته: مدیریت، مهندسی انرژی

گرایش: مدیریت مالی، مدیریت بحران، انرژی های تجدید پذیر

دانلود رایگان این مقاله:

  دانلود مقاله موضوع مصرف انرژی تجدید پذیر

مشاهده سایر مقالات جدید:

مقالات جدید مدیریت 

مقالات جدید مهندسی انرژی 

4. Sample data, sources, and characteristics

The selected sample consists of annual balanced panel data from 37 (19 high income, 11 upper middle-income and 7 lower middle-income) countries and spans the 1987 to 2018 period. The sample data covers two crisis periods – the 1997 and the 2008 crises. The dependent variable proxies pollution captured by carbon dioxide emissions (CO2). Next, the explanatory variables include renewable energy consumption (Renewable Energy), FDI inflows (FDI), exports (Export), import (Import), and GDP (GDP^).14 Sample data definitions, descriptive statistics and sources (list of countries) are presented in Table 1, 2, 3, 4(6). Carbon dioxide (CO2) emissions, measured in million tons, are attributed to the country in which they physically occur. The CO2 emissions data are from the “Our World in Data” database derived from the Global Carbon Project15. Renewable energy consumption, measured in terawatt-hours (TWH), data are from the Our World in Data database16. The inflow of foreign direct investment (FDI) is measured as a percentage of gross domestic product for the year. The FDI data are from United Nations Conference on Trade and Development website17. Exports are exports of goods and services. The Exports data are in current U.S. dollars using natural logarithms. Imports are imports of goods and services. The Imports data are in current U.S. dollars using natural logarithms. Gross domestic product per capita (GDP^) is defined as gross domestic product minus net export. The Exports, Imports, and GDP data are from the World Bank website18. Finally, results are computed for the overall time period19 and for each of three subperiods, periods 1, 2, and 3. Period 1 only includes data spanning the 1987–1996 (inclusive) period. Period 2 (3) spans data for the 1998–2007 (2009–2018) time frame. Next, Table 5 presents standard deviations per unit of output for CO2 emissions and GDP output for the full time period and for each of periods 1–3. These results clearly document that the standard deviation per unit of output of CO2 emissions are larger than corresponding estimates for GDP for all time periods except for the post-1997 crisis period. These results are generally consistent with the findings of Peters et al. (2012). Next, data availability by country and time periods also allows us to conduct a pairwise t-test to determine whether variable means have changed across both crises. The pairwise t-test is a preliminary test to determine if the variable means for CO2 emissions and for renewable energy differ for each category of high income, upper-middle and lowermiddle income countries and across time periods delineated by the crises. If there are no statistically significant differences in each variable mean (CO2 emissions and renewable energy) across time periods and across countries, then there may be no basis to conduct formal tests on the nature of these relationships. If there are significant differences in mean values for CO2 emissions and renewable energy across countries and time periods delineated by the crises, then we can proceed with the formal tests to examine the relationship between the two variables of interest (see Table 5, 6). Table 7 presents these results for differences between pairwise values between periods 2 (post-1997 crisis) and 1 (pre-1997 crisis) for all variables for the overall sample and for each subsample. Similarly, the difference in pairwise values between period 3 (post-2008 crisis) and period 2 (post-1997 crisis but pre-2008 crisis) are presented for all variables and samples. Table 8 presents the paired test results in summary form for ease of interpretation.

5. Empirical results

We first examine whether the sample variables are independent crosssectionally with each other. Table 9 presents the results of the Pesaran"s cross-sectional dependence CD tests and indicate that the null hypothesis of no cross-sectional dependence between sample variables is rejected at the 1% level. To rectify this problem, we conduct the second-generation panel unit root test (Pesaran, 2007). Results documented in Table 10 show that all sample variables are stationary at level. This finding enables us to use the variables to examine the relationships between CO2 emissions and the explanatory variables for the overall sample and for each subsample. As indicated earlier, we adopt the Dynamic Panel Data Model (DPDM) to examine the relationship between the dependent variable and stated explanatory variables. These results are presented in Table 11 for the entire sample and for each subsample, prior to and after each crisis. Table 11 contains the parameter estimates while Table 12 contains the tests for significance of generated estimates. Table 12 shows that all the regression models are significant at the 1% level, with the exception of the results for the high-income sample, post 2008 crisis, using the joint test.

(دقت کنید که این بخش از متن، با استفاده از گوگل ترنسلیت ترجمه شده و توسط مترجمین سایت ای ترجمه، ترجمه نشده است و صرفا جهت آشنایی شما با متن میباشد.)