Welfare implication of determinants affecting aggregate consumption expenditures in Rwanda

3.3.3.1 Interpretation of the long run Model

From the above table, results given by Eviews8 highlight the follows:

LNGCE= 0.371652+0.867339LNGDP-0.032394INT-0.040527INF-0.025767LNEXCHR

? From the above output of long run equation, first, the gross domestic product (GDP) is positively related to gross consumption expenditure (GCE) as expected in theory. This means that, when GDP increases by one more units, GCE increases by 86% by considering other variables constant. (Ceteris paribus)

57

? Second, from the above table, the interest rate is negatively related to GCE which means that: increases by one more units in interest rate, decreases gross consumption expenditure (GCE) by 3.2%, ceteris paribus.

? Third, from our findings, the inflation rate is negatively related to GCE which means that: an increase by one more units in inflation rate decreases the GCE by 4.0% ceteris paribus.

? Fourth, from our findings, the exchange rate is negatively related to GCE which means that: an increase by one more unit in exchange rate (depreciation) decreases the GCE by 2.5% ceteris paribus.

The probabilities of coefficients show that all explanatory variables except that of the constant, are statistically significant because they are less than 0.05 or less than 5%.

The coefficient of Determination R-squared (R²) equals 99.8% indicates that the explanatory variables contribute significantly in explaining gross consumption expenditure. This is the indicator of a well fitted model. The Prob (F-statistic) of 0.0000 allow us to reject the null hypothesis which states that all explanatory variables do not collectively explain the explained variable and therefore all explanatory variables collectively influence the gross consumption expenditure.

3.3.4. Vector Auto-regression Estimates (Short run relationship) The system equation for the short run is:

Equation: LNGCE = C(1)*LNGCE(-1) + C(2)*LNGCE(-2) + C(3)*LNGDP(-1)

+ C(4)*LNGDP(-2) + C(5)*INT(-1) + C(6)*INT(-2) + C(7)*INF(-1) + C(8)

*INF(-2) + C(9)*LNEXCH(-1) + C(10)*LNEXCH(-2) +

C(11) Observations: 19
R-squared	0.994725	Mean dependent var	7.254886
Adjusted R-squared	0.988131	S.D. dependent var	0.738783
S.E. of regression	0.080487	Sum squared resid	0.051826
Durbin-Watson stat	1.816104

Source: World Bank indicators1995-2015 and author's computation

Table 7: Short run relationship effect of changes in GDP, INT, INF, and EXCHR on Gross Consumption Expenditure

58

3.3.4.1. Interpretation of the short run equation and the coefficients of probabilities

From the above findings, the probability except C (1) and (C3) shows that individually, explanatory variables are not statistically significant to influence the dependent variable because they are greater than 5% level of significance except the GDP which has a probability of 4.1%. From the above results obtained from E-views8 software, not all conditions of a good model are observed:

- The coefficients are different from zero which is good for the model

- The coefficient (C1) of residuals is positive.

The determination R-squared (R²⁾ is greater than 99.4% which also good for the model and from the output coefficients are not statistically significant to influence the dependent variable except the GDP. This is because their probabilities are greater than 5% level of significance except that of GDP which is 4.1% and this means that these independent variables are not significant to influence gross consumption expenditure in the short run.