*Economic Growth and Environmental Pollution; Testing the EKC Hypothesis in Brazil DOI: http://dx.doi.org/10.5772/intechopen.104388*

may proceed with cointegration analysis (1). In **Table 2**, the lag order selection criteria are used to decide which criterion best fits the study—the icteric (\*) number with the lowest value is chosen as the criterion for selecting the delays. One SC reported 24.55327\* under lag, according to the table. With icteric (\*) in lag two, four criteria were identified: LR-36.39728, FPE-1.55e-05\*, AIC-2.560914\*, and HC - 2.507760\*. Among the criteria, the AIC with the number 2.560914 is the lowest number with icteric (\*) under lag two. As a result, the Akaike Information Criterion is chosen as the lag order for future estimations.

**Table 3** provides Vector Auto-regression estimates-VAR estimates. Given that C (1) is the coefficient of the first lag of CO2 and it is a log-log formation, the interpretation will be on elasticity form. The coefficient signs tell us the direction of the impact; negative signs indicate a decrease, and positive signs indicate an increase.


*I- denote intercept, I&T- represent intercept and trend (\*) Significant at the 10%(\*\*)Significant at the 5%(\*\*\*) Significant at the 1%(No)Not significant. Lag length based on SIC. Probability based on MacKinnon (1996) one sided pvalue.*

#### **Table 1.**

*Unit root test.*


*\* Indicates lag order selected by the criterion, LR: sequential modified LR test statistics (each test at 5% level), FPE: Final prediction error, AIC: Akaike information criterion, SC: Schwarz information criterion, HQ: Hannan-Quinn information criterion.*

#### **Table 2.**

*Lag order selection criteria.*


#### **Table 3.**

*Vector auto-regression estimates.*

Therefore, looking at CO2, we can see that CO2 strongly influences itself at LGGDPSQ (1). The past realization of CO2 is associated with a 100% increase in CO2 emissions on average certris paribus. LGGDPSQ(1) recorded t-statistic value of 1.96983\* and its corresponding coefficient of 0.004306. Therefore, a percent increase in GDP per capita is associated with a 0.431 percent increase in carbon dioxide emissions on average. The findings from **Table 4**- the VAR estimates indicated that economic growth increase environmental pollution cetris paribusThe Wald test is shown in **Table 4**. **Table 4** shows that the P-value is significant at the 1% level; as a result, we reject the null hypothesis; (5) = C (6), since the Wald test shows that the coefficient of log GDP data to the first and second lags of GPD have a statistical impact on the log of CO2. We may conclude that GDP GDPSQ has a combined significant effect on CO2 based on the results of the Walt test. It can also be observed in **Table 5** that no root lies outside the unit circle. This indicates that VAR meets the stability requirement.

The Pairwise Dumitrescu Hurlin panel causality (PDHPC) and Pairwise Granger causality tests are estimated in **Table 6**. Economic growth and carbon dioxide emissions are bidirectional, according to the PDHPC. The Zbar-stat and its probability values (2.21122, 0.0270; 3.41367, 0.0006) respectively verified this. This finding
