Working Paper Series: Special Edition of 2016 to 2018 Interns

4.3 Methodology 4.3.1 Panel Least Squares

The results of the Durbin-Watson test for autocorrelation indicates that there is an absence of serial correlation among the variables. Four different tests were used to check the presence of unit root: i. Levin, Lin and Chu; ii. Im, Pesaran and Shin W-Sta;, iii. ADF-Fisher Chi-square; and iv. PP- Fisher Chi-square. The Levin, Lin and Chu test assumes a common unit root process while the other tests assume individual unit root process. Table 3 in the appendix shows a summary of the unit root tests results. The findings indicate that all variables except consumption activity were statistically significant in levels, indicating the absence of a unit root in these variables and rendering them stationary. Government consumption returned a p-value of more than 5 per cent for all of the unit root tests, presenting evidence of a common unit root and an individual root. In order to make the variable stationary, the variable was first differenced. Additionally, the White cross-section method was applied to control for heteroscedasticity. The Hausman test is used to determine whether the Fixed Effects model or the Random Effects Model was the more suitable. The null and alternative hypotheses for the Hausman Test are as follows: 0 : ℎ 1 : ℎ In models 1 and 2, the Hausman test returned a p-value of 0.3702 and 0.2657. Given the insignificance of these results, the null hypothesis was not rejected and the random effects model was deemed to be appropriate. These results can be seen in table 4 of the appendix. 4.3.2 Debt Sustainability Analysis The Debt Sustainability Analysis Framework is a forward-looking analysis of debt that is used to assess the primary balance necessary to maintain debt at a certain level. Also, it calculates the primary balance needed to reduce debt to a target level within a specified number of years. The Debt Sustainability Analysis was used for all eight ECCU countries in order to forecast the likely debt path of member countries after the occurrence of an intense natural disaster. The natural disaster coefficient from model 2 was used as a shock within the DSA to evaluate whether the

162