英语论文网

nt variable in both models. The explanatory variables are external debt service payment and external debt stock.

4.4 MODELS SPECIFICATION

The models to investigate the effect of external debt and external debt servicing on the economic growth of ECOWAS with the dependent variable as gross domestic product while the explanatory variables are external debt service payment. So that:

MODEL

gdp = f (eds, ds)

Where gdp - Gross Domestic Product

eds - External debt stock

ds - Debt service payment

4.5 METHODOLOGY

The method employed for construction, estimation and analysis will be econometric models. The model attempt to investigate the relationships between the external debt and servicing and economic growth of each ECOWAS countries using an annual time series data for each ECOWAS (1960 to 2008), The following tests will be conducted in the estimation of individual country data of the ECOWAS countries: ADF and PP Unit root test, Johansen – Juselius Co-integration test and Error correction Model.

4.6 UNIT ROOT TESTS

Prior to econometric modeling estimates and in to order to empirically test for co-integration between economic growth, external debt stock and servicing variables there is a need to examine, as an initial step, if series are integrated and have a unit root by using Augmented Dickey – Fuller(ADF) (1979) and Phillips- Perron tests (PP). Phillips and Perron (1988) propose a non parametric method of controlling for higher –order serial correlation in a series. Thus, the test regression for the Phillips- Perron (PP) test is the AR (1).

The PP test makes an adjustment to the t-statistic of the coefficient from the AR(1) regression to account for the serial correlation in εt .The Augmented Dickey-Fuller (ADF) (1979) and Phillips –Perron (PP) unit root test is used in examining the stationary nature of the data series and test the null hypothesis that a time series is I(1). It consists of running a regression of the first difference of the series against the series lagged once, lagged difference terms, and optionally, a constant and a time trend. This can be expressed as:

Δyt = β1yt-1 + β2Δyt-1 + β3Δyt-2 + β4 + β5t (1)

The test for a unit root is directed on the coefficient of yt-1 in the regression. If the coefficient is considerably different from zero then the hypothesis that y contains a unit root is rejected. Rejection of the null hypothesis implies stationarity. If the calculated ADF statistic is greater than McKinnon's critical value then the null hypothesis is not rejected and it is concluded that the considered variable is non-stationary, i.e. has at least one unit root. Then, the testing is re-applied after transforming the series into first differenced form. If the null hypothesis of non stationarity can be discarded, it can be concluded that the time series is integrated of order one, I (1).

4.7 CO-INTEGRATION TESTS

After the order of integration is determined in the data, co-integration between the series should be tested to identify any long run relationship. “Co-integration means that despite being individually nonstationary a linear combination of two or more time series can be stationary” (Guajarati, p.730). Co-integration is the statistical proposition of the existence of a long term r