This section describes the econometrics methods that we use to access the relationship between FDI and economic growth

Subject:EconomicsPrice: Bought3

Share With

This section describes the econometrics methods that we use to access the relationship between FDI and economic growth. We use simple ordinary least square (OLS) regressions and the empirical analysis is conducted by using annual data on FDI and economy growth in Malaysia over the 1970-2005 periods. We use annual data from IMF International Financial Statistics tables, published by International Monetary Fund to find out the relationship between FDI and economic growth in Malaysia case. 5.1 OLS framework Growth = a + BFDI, (1) Where the dependent variable, Growth, cquals to real GDP growth or real GNP growth, and FDI is gross private capital inflows to a country. We use both GDP and GNP for dependent variables in order to test the robustness of the findings From the equation above, the positive sign of coefficient for FDI represent that there is positive relationship between FDI and economy growth. If there is an increase in FDI infow, there will led and enhance the economic growth in Malaysia. In contrast , if the FDI is negative correlation to economic growth, it will not help in GDP growth in a country The hypothesis is stated as below Hypothesis 1: Ho: B = 0 HB 0 The null hypothesis B = 0 (there are no relationship between foreign direct investment (FDI) and real gross domestic production (RGDP) ) or real Gross National income RGNI) against its alternative B+ 0, if less than lower bound critical value (0.05), then we do not reject the null hypothesis. Conversely, if the t-statistic value greater than 5 percent critical value, then we reject the null hypothesis and conclude that there are significant relationship between independent variable and dependent variable. 5.2 Diagnostic Testing On the other hand, we also apply the diagnostic testing to test the series whether the series are free from autocorrelation. heteroscedasticity and normality problem. Hypothesis 2: Ho: There are autocorrelation between members of series of observations ordered in time Hy: There are not autocorrelation between members of series of observations ordered in time Hypothesis 3: Ho: There are constant variances for the residual term H : There are no constant variance for the residual term. The null hypothesis from hypothesis two and three are do not existing autocorrelation and heteroscedasticity against its alternative do existing autoregression and heteroscedasticity. If the computed p-value is greater than 0.05 significant levels, then we do not reject the null hypothesis and conclude that there does not existing autocorrelation and heteroscedasticity. Conversely, if the computed p-value is less than 0.05 significant levels, the we reject the null hypothesis and conclude that there are existing autocorrelation and heteroseedasticity problem 5.3 Unit Root Test The first step of constructing a time series data is to determine the non-stationary property of each variables, we must test each of the series in the levels (log or real GDP or GNP and log of FDI) and in the first difference (growth and FDI First, the ADF test with and without a time trend. The latter allows for higher autocorrelation in residuals. That is, consider an equation of the form: AX-B+X+ P.AX. + rate). However, as pointed out earlier, the ADF tests are unable to discriminate well between non-stationary and stationary series with a high degree of autoregression In consequences, the Phillips-Perron (PP) test (Phillips and Perron, 1988) is applied. The PP test has an advantage over the ADF test as it gives robust estimates when the series has serial correlation and time-dependent heteroscedasticity. For the PP test, estimate the equation as below: ? AX,-a +Axx1 + 0 (---)+ 4x + (iii) 2 In both equations (ii) and (iii), A is the first difference operator and C., and ca are covariance stationary random error terms. The lag length n is determined by Akaike's Information Criteria (AIC) (Akaike, 1973 to ensure serially uncorrelated residuals (for PP test) is decided according to Newley-West's (Newley and West, 1987) suggestions. Hypothesis 4: He: Series contains a unit root H: Series is stationary In ADF and Phillips Perron tests, the null hypothesis of non-stationarity is tested the t-statistic with critical value calculated by MacKinnon (1991). The outcome suggests that reject null hypothesis which can conclude the series is stationary. Both ADF and PP test are applied following Engle and Granger (1987) and Granger (1986) and subsequently supplemented by the PP test following West (1988) and Culver and Papell (1997). Besides that, Kwiatkowski, Phillips, Schmidt, and Shin (1992) introduce such a test, and do it by choosing a component representation in which the time series under study is written as the sum of a deterministic trend, a random walk, and a stationary error. The null hypothesis of trend stationary corresponds to the hypothesis that the variance of the random walk equals zero. As one could expect their results are frequently supportive of the trend stationarity hypothesis contrary to those traditional unit root tests. Hypothesis 5: He: Series is stationary H: Series contains a unit root In KPSS test, the null hypothesis of stationarity is tested. The outcome suggests that do not reject hypothesis, which can conclude the series is stationary. Besides, testing for stationarity is so important in time series data is to avoid spurious regression problem and violate of assumption of the Classical Regression Model.