The Regression Model of a Medium-Term Strategy for Raising the Gross Domestic Product - Assignment Example

The intention of this paper is to provide an analysis of all the possibilities available before the forthcoming meeting of the government committee charged with formulating a medium-term strategy for raising the country’s per capita Gross Domestic Product (GDP) over the next 15 years. The present meeting is considered to be of crucial significance since new priorities are to be decided upon with the current economic crisis changing the whole situation demanding a reappraisal of the same. At its previous meeting the committee had decided to devote the government allocated US$ 2 billion fund to policy measures that will increase the country’s enrolment rate in secondary education. Nevertheless this decision would now be seriously required to be revisited since this recession has inflicted serious damage especially upon the financial and banking sector. With banks making loud claims for an emergency intervention on the part of the government the committee needs to prioritise sectors on the basis of the likely chances to promote and raise the country’s per capita Gross Domestic Product (GDP). This paper provides with an econometric model in order to facilitate the committee to look at all the possibilities and arrive at a decision in this regard. The Regression Model For constructing the model, six independent variables are considered viz, normalized value of per capita GDP in 1990, percentage of secondary school enrollment rate, government share of real GDP per capita in 1990, variable related to the openness of the market of the country, rate of inflation and credit. The rate of growth of per capita GDP at constant prices during the period from 1990 to 2005 is taken as the dependent variable. For normalizing the values of per capita GDP, natural logarithms of the values are taken for estimation. Likewise, the values of secondary school enrollment and inflationary rate have been normalized by converting into logarithmic values. The model for the estimation is: dlypci = 1 + 2lypc90i + 3lsecedi + 4govgdpi + 5openi + 6infli + 7crediti + ui where ypc05 = per capita GDP (constant prices, chain series) in 2005 ypc90 = per capita GDP (constant prices: chain series) in 1990 dlypc = ln(ypc05) – ln(ypc90) lypc90 = ln(ypc90) seced = percentage of secondary school age population enrolled at secondary school in 1990 lseced = ln(seced) govgdp = government share of real GDP per capita in 1990 open = openness = ratio of (exports+imports) to GDP in 1990 cpi90 = consumer price index value in 1990 cpi85 = consumer price index value in 1985 infl = five-year inflation rate = ln(cpi90) – ln(cpi85) credit = ratio of private credit by deposit money banks and other financial institutions to GDP in 1990 The empirical investigation through this model intends to examine whether the growth of per capita Gross Domestic Product of the countries during the period from 1990 to 2005 is influenced by these six important dependent variables. The per capita GDP in 1990 at constant prices is taken as one of the dependent variables in order to check whether the initial achievement in development has any significant impact on the further development process. In the model, it is also examined whether the possession of income by the government (or government revenue) has any positive and crucial role in enhancing the income level of the whole country. For estimating this, government share of real GDP per capita in 1990 has been taken as a variable in the model. In order to understand, whether inflationary trend has stimulated the growth rate of per capita GDP, the difference in the values of consumer price indices in 1990 and 1985, has been estimated. For making the crucial policy decision of whether boosting secondary school enrolment or bailing out the banks, secondary school enrolment ratio and ratio of private credit by deposit money banks and other financial institutions have been taken as variables in the model. Model Estimation The model estimated with data from 50 countries is as follows: dlypci =0.757 + -0.199 lypc90i + 0.25 lsecedi + 0.01 govgdpi + 0.0110 openi +0.114 infli + 0.212 crediti { Number of obs = 50, F( 6, 43) = 1.77, Prob > F = 0.1291, R-squared = 0.1977, Adj R-squared = 0.0857, Root MSE = .27658} The estimated values of ratio of secondary school enrolment and ratio of private credit by deposit money banks and other financial institutions are 0.25 and 0.21 respectively which implies that these two variables influence more on the per capita GDP than other four variables under our consideration. The coefficient value of the variable, ratio secondary school enrolment seems to be higher than the other five coefficients in the model. That is, there exists 25 percent influence of secondary school enrollment ratio on the per capita GDP of the country while the influence of credit is 21 percent. The coefficient value of per capita GDP in 1990 seems to be negative which points out that the average value of per capita GDP during the15 years and the value of per capita GDP in 1990 has an inverse relationship. This negative relationship between these variables may be because higher the value of per capita GDP in 1990, lesser will be the difference between the values of two years. The coefficient values of the variables, openness and government share in the per capita GDP seem to be very low. The overall strength of the regression model, which is shown by R-Squared and Adjusted R-Squared Tests does not seem very powerful as the value of R-Square is not high. The detailed regression results are shown in the following table. Tests of Hypothesis: (i) H0:2=3=4=5=6=7=0 against H1:j0 for at least one j(2...7), using a significance level of 0.05. (ii) H0:2=0 against H0:20 using a significance level of 0.05. (iii) H0:3=0 against H0:3>0 using a significance level of 0.05. (iv) H0:7=0 against H0:7>0 using a significance level of 0.1. Table 1. Regression Results Source | SS df MS -------------+------------------------------ Model | .810476713 6 .135079452 Residual | 3.28926078 43 .076494437 -------------+------------------------------ Total | 4.0997375 49 .083668112 dlypc Coef. Std. Err. T P>t [95% Conf. Interval] lypc90 -.1997314 .0748916 -2.67 0.011 -.3507648 -.048698 lseced .250282 .0966267 2.59 0.013 .0554157 .4451483 govgdp .0100052 .0062559 1.60 0.117 -.002611 .0226215 Open90 .0001565 .0015954 0.10 0.922 -.0030609 .003374 Infl .1142594 .1060633 1.08 0.287 -.0996376 .3281564 credit .2124701 .139727 1.52 0.136 -.0693162 .4942563 _cons .7566435 .404901 1.87 0.068 -.0599171 1.573204 While looking at P values and t values of the coefficients, only three variables (i.e, the per capita GDP in 1990, secondary school enrollment, and government share in the per capita GDP in 1990) seem to be significant at 0.05 level of significance. Thus, the empirical evidence shows that the co-efficient of the variable, ratio of secondary school education seems to significant and the rejects the null hypothesis at 0.05 confidence level. The coefficient of credit variable does not seem to be significant at 0.05 confidence level and hence it is not possible to reject the null hypothesis. Hence strong recommendation is made for spending the sum of US$ 2 billion more effectively on boosting secondary school education than bailing out the banking sector. Test for Homoscedasticity White’s Test White test is used to test whether the regression estimation satisfies the assumption of homoscedasticity or there is the presence of Heteroscedasticity (Hamilton 1994; Wooldridge 2002). Whites test for Ho: homoskedasticity Test statistic: TR^2 = 15.90219, with p-value = P(Chi-square(2) > 3.256389) = 0.9548 Whites general test statistic : 15.90219 Chi-sq(27) P-value = .9548 The result shows that the residuals do not have the Heteroskedasticity as the probability value accept the null hypothesis that there is no heteroskedasticity (Table). NormalityTest Shapiro-Wilk W test is used to check whether the model satisfies the assumption of linearity. The results are given in the following table. Shapiro-Wilk W test Variable Obs W V z Prob>z dlypc 50 0.91706 3.900 2.903 0.00185 lypc90 50 0.93390 3.109 2.419 0.00779 lseced 50 0.91626 3.938 2.923 0.00173 govgdp 50 0.93038 3.274 2.530 0.00571 infl 50 0.97639 1.110 0.223 0.41180 infl 50 0.97639 1.110 0.223 0.41180 credit 50 0.83159 7.920 4.413 0.00001 The results shows that the normality assumption of the OLS method exists in this regression estimation also. Conclusion The empirical evidence based on the observations from 50 countries throws light on the need of investing the sum of US $ 2 billion for enhancing secondary school enrollment than for boosting up of banking sector. References Hamilton, J. D. (1994). Time Series Analysis. Princeton, NJ: Princeton University Press. Gujarati, Damodar (2004). Basic Econometrics, Fourth Edition, New York, The McGraw-Hill Companies. Wooldridge, J.(2002). Introductory Econometrics: A Modern Approach, USA, Thomsun Business Information. Appendix – The data used for the estimation of Regression Model dlypc lypc90 lseced govgdp open 90 infl credit 0.168679 8.578219 4.110874 10.85 73.97 0.135922 0.401753 0.391244 10.05234 4.406719 13.46 28.85 0.278506 0.946479 0.292834 7.387808 2.944439 8.18 17.81 -0.11736 0.212692 0.256822 10.10883 4.634729 14.84 124.59 0.489905 0.348254 0.141694 8.963319 3.637586 21.34 13.39 0.377779 0.240326 0.332023 6.830971 1.94591 38.37 59.15 0.242289 0.175308 -0.04945 7.904781 3.332205 10.67 30.56 0.324603 0.275423 0.303555 10.14778 4.615121 15.21 49.94 0.175224 0.766266 0.674793 9.064156 4.290459 16.17 47.41 0.056461 0.468582 1.212102 7.564835 3.89182 20.27 23.82 -0.26788 0.86382 -0.22166 7.969244 3.091042 12.74 63.45 0.372058 0.402152 0.164475 8.493511 4.007333 21.28 41.55 -0.60756 0.121859 0.374862 8.187316 4.330733 7.41 62.33 -0.11368 0.278849 0.113376 6.756874 2.639057 18.38 27.95 -0.18109 0.231029 0.195971 10.07144 4.59512 16.86 32.71 0.466514 0.915517 0.18329 10.11047 4.584967 12.02 40.66 0.486977 0.93393 0.195406 7.137676 3.583519 18.12 58.88 -0.39903 0.046946 0.402867 9.742274 4.532599 14.13 36.71 0.488817 0.345739 0.348799 9.345009 4.369448 27.65 36.52 0.233354 0.448724 0.519587 7.601697 3.78419 28.29 17.05 -0.13171 0.255768 0.417537 8.076177 3.78419 18.32 46.59 -0.23778 0.367796 0.5122 8.646666 4.007333 13.88 75.76 1.136515 0.28709 0.182053 10.05055 4.418841 13.32 42.58 0.573789 0.480762 0.121066 10.18054 4.574711 10.71 16.86 0.383969 1.915641 -0.0215 7.631063 3.178054 8.41 43.02 -0.19427 0.298436 0.616012 9.384983 4.49981 10.16 32.56 0.290338 0.904025 -0.21659 6.976759 2.890372 12.09 57.23 -0.15877 0.147346 0.231642 6.841306 2.079442 6.72 55.7 0.143391 0.12597 0.671753 9.03824 4.025352 13.87 139.83 -0.14057 0.670275 0.353629 6.780513 1.94591 19.82 45.73 0.444871 0.123941 0.124496 8.411804 3.555348 10.7 44.93 0.293485 0.133424 0.260194 7.281909 3.496508 16.32 31.53 -0.08693 0.117449 0.281977 10.11126 4.787492 17.61 78.34 0.455658 1.398607 0.301187 7.200022 3.218876 7.02 56.44 -0.93321 0.118722 0.298385 7.79397 3.135494 18.53 32.09 -0.09135 0.237437 0.354024 8.300141 4.204693 12.71 24.54 0.696544 0.041801 0.182378 8.127319 4.290459 13.53 74.32 -0.08518 0.195622 0.565592 8.881093 4.394449 20.19 27.72 -0.42264 0.016446 -0.08262 10.02202 3.78419 17.74 79.73 -0.23838 0.637771 0.194014 8.976521 4.304065 22.27 38.4 0.444144 0.836308 0.422161 9.858065 4.644391 11.87 27.62 0.645984 0.751116 0.525316 8.055535 4.304065 23.42 54.8 -0.03674 0.177204 0.71807 6.862538 3.178054 6.41 29.33 1.13796 0.059481 0.356955 7.504722 3.951244 23.84 71.3 -0.07881 0.070054 0.472006 8.595204 3.401197 11.93 90.5 0.07221 0.72401 0.284566 8.587898 3.850148 15.27 24.63 0.411392 0.13066 0.455629 6.606785 2.564949 32.61 27.08 -0.4526 0.024429 0.33108 9.987024 4.442651 16.48 36.97 0.404541 1.130044 0.078276 9.224906 3.555348 21.96 46.47 -0.45851 0.231344 -0.79122 8.462777 3.912023 13.32 56.76 -0.03619 0.197089 Read More