英语论文网

yt-1 – b1 – b2 xt-1 as the right hand side variable in the error correction model through estimating it with a second least squares regression.

Chapter 5

EMPIRICAL RESULTS

5.1 Introduction

This chapter will provides both the interpretation and the explanations of the results from the test employed to the data defined for the ECOWAS countries.

5.2 Unit Root Tests

The accepted method to test the unit root for stationarity of variables and settle the degree of integration of a time series is to apply the Augmented Dickey Fuller (ADF) and Phillips- Perron (1988) tests. Therefore, these tests for unit root process at the level and at the first differences of all variables have been performed and reported in Table 1.

According to, Table 1, ADF and PP unit root test reveals that GDP, External debt stock and debt Service are all the time series are I(1). However, Cape Varde, Guinea, Liberia, Mali, Senegal as omitted because they have a missing data of more than 20 years.

Table 2: Unit Root Tests for Individual Countries Variables

Source: World Development Indicator (WDI)

Having found out that the time series are I (1), co-integration analysis can be considered to identify the long run relationship between Gross Domestic Product, External Debt Stock and Debt servicing. To perform the standard co-integration test, all economic variables in the model must have the same order of integration (Kulendran, 1996).

5.2 Co Integration Analysis

The rationale of the co integration test is to determine whether a group of non –stationary series is co-integrated or not, as it is discussed in Chapter 4, if any two time series Xt and Y t are both I (1), and if error term, et (the residuals of the time series), which is defined as et =Y- β X t, is I (0), then Xt and Yt are said to be co-integrated.

5.2.1 Johansen Co Integration Test Results

The Johansen co-integrated provides the log likelihood ratio statistics for determining the number (r) of long run relationship between GDP, EDS and DS. If calculated value of the statistics (Johansen trace test statistic) is greater than 95% critical value, the null of r= 0, which indicates no long term relationship, is rejected against the alternative hypothesis.

Now that we have established that GDP, EDS and DS are non stationary at level and integrated to the same order me (1), we can test for the presence of co-integration.

Source-World Bank Development Indicator (WDI)

*, ** and *** denotes statistical significance respectively at 0.01, 0.05, and 0.10 levels of alpha. Optimum Lags are selected by Akaike information Criterion. Figures in parenthesis are t-ratio.

The Johansen co-integration test results are reported in Table 2. Results indicates that there is long run relationship between GDP, external debt stock and servicing in the above 10 ECOWAS countries since the calculated value of the statistics is greater than 5% critical value.

5.2.2 Estimated Long Run Relationship between External debt and Debt servicing on the ECOWAS countries Economic growth

Table 2 reports the co-integration analysis results for the ten countries in the ECOWAS countries. The coefficient estimates for the debt service are negative for Benin and Nige