Multiple Regressions Empirical Project - Essay Example

Multiple Regressions Empirical Project May Table of Contents May Table of Contents 2 List of Tables 2 List of Figures 2 Executive Summary3 Introduction 4 Literature Review 4 Economic theory 4 Data 5 Hypothesis 5 The Model 5 Empirical Results 5 Autocorrelation Test 5 Heteroskedasticity Test 6 Multiple Regression Analysis 7 Endogenity Test of the Full Model 8 Simple Linear Regressions 8 Conclusion 11 References 12 List of Tables Table 1: Heteroskedasticity Test 6 Table 2: Full Model Coefficients 7 Table 3: Model with Consumption as the Explanatory Variable 8 Table 4: Model with Foreign Direct Investment as the Explanatory Variable 9 Table 5: Model with Net Exports as the Explanatory Variable 10 List of Figures Figure 1: Autocorrelation Test Plot 6 Figure 2: Histogram of Residuals 7 Figure 3: Consumption as Explanatory of GDP Plot 9 Figure 4: FDI as Explanatory of GDP Plot 10 Figure 5: NE as Explanatory of GDP Plot 11 Executive Summary This study was meant to determine the impact of responsible consumption, foreign direct investments and net exports. To make this a success, secondary data downloadable from http://www.oswego.edu/~kane/econometrics was used. From the results it was found out that a billion dollar increase in consumption holding direct foreign investment and net exports constant leads to 1.41 billion dollar increase in GDP. Further, a billion dollar increase in direct foreign investment holding consumption and net exports constant leads to 0.18 billion dollar decrease in GDP. A billion dollar increase in net exports holding consumption and direct foreign investments constant leads to 0.47 billion dollar increase in GDP. Considering consumption alone, it was found out that a billion dollar increase in consumption leads to 1.43 billion dollars increase in GDP and that consumption levels explains 99.84% of the total variations in GDP [r2 (60) = 0.9984]. Further, taking foreign direct investments alone, it was found that a billion dollar increase in foreign direct investment leads to 5.33 billion dollars increase in GDP. This model was found to be significant at 5% level of significance and that FDI explains 96.39% of the total variations in GDP Lastly, a billion dollar increase in net exports led to 17.47 billion dollars decrease in GDP and the model with NE alone was found to be significant at 5% level of significance and that NE explains 54.41% of the total variations in GDP. Introduction This study aimed at determining the impact of responsible consumption, foreign direct investments and net exports and employed the use of secondary data to proof the objectives. Different writers have argued that consumptions and investments are the key variables on which the GDP depends most. However, other variables like irresponsible consumptions, political un-rests, environmental degradation, and lack of government priorities translate to irresponsible spending are some other factors which should be taken care of for GDP to grow. Literature Review GDP is the cumulative amount of goods/services which a country produces within a given year (Hall and Mishkin 1982; Hill 1992). When GDP changes, then a country is said to have experienced economic growth (if positive change) and economic melt-down (if negative growth-previous year’s performance is better than current year’s). Factors like level of consumption, direct foreign investments and net exports are some of the factors which contribute to positive GDP growth, hence economic growth (Haron 2005). High direct foreign investments and high net exports translate to more GDP for a country. On the other hand, more irresponsible consumptions, political un-rests, environmental degradation, lack of government priorities translate to irresponsible spending. These are some of the factors which impact negatively on the economic growth of a country. The well being of a country’s citizenry determines the level of development in that country. High GDP growth translate to more jobs Economic theory In this case, the economic model is depicted by the fact that responsible consumption and an increase in direct foreign investment and net exports do not affect GDP growth while responsible consumption and an increase in direct foreign investment and net exports affects GDP growth. Data To ascertain the above arguments, secondary data; downloadable from http://www.oswego.edu/~kane/econometrics was used. In this data set, China’s GDP (from 1941 to 2002 in Billion Dollars) was considered as a dependent variable (DV) while consumption, direct foreign investment (in Billion Dollars) and net exports (in Billion Dollars) were considered as explanatory variables. Hypothesis The Empirical research paper attempted to proof the following hypothetical statements; H0: Responsible consumption and an increase in direct foreign investment and net exports do not affect GDP growth H1: Responsible consumption and an increase in direct foreign investment and net exports affects GDP growth The Model The empirical research project attempted to predict the following models; GDP = β0 + β1C+ β2DFI + β3NE (Where C = Consumption, DFI = Direct Foreign Investment and NE = Net Exports). In this multiple regression equation, GDP was the dependent variable while consumption, Direct Foreign Investment and Net Exports were considered as the within subject factors. Simple regression models of the form; GDP = β0 + β1 Bi (Where the Bi are the explanatory variables). Empirical Results Autocorrelation Test The first test performed was the autocorrelation test carried to determine the existence of serial correlation of the GDP over the years. Figure 1: Autocorrelation Test Plot From figure 1, it is evident that the error terms are un-correlated and thus the GDP values have no serial correlation. Heteroskedasticity Test A heteroskedasticity test was carried out White test as shown below. According to Gore and Altman (1992); Freed et al. (1991), if the error terms are not homoskedastic, then the model coefficients are not reliable and thus heteroskedasticity should be tested for and eliminated. Table 1: Heteroskedasticity Test From the above results, it is evident that the model errors are not homoskedastic [Chi-square (9) = 32.40, p = 0.0002]. These results can also be confirmed by figure 2 results. Figure 2: Histogram of Residuals This can be corrected using robust error in the model (see table below). Multiple Regression Analysis Table 2: Full Model Coefficients From table 2, the model was found to be significant at 5% level of significance [F (3, 58) = 9210.77, p Read More