Multiple Regression Analysis: Empirical Project - Essay Example

 Multiple Regression Analysis: Empirical Project Abstract Several factors have impacts on the GDP and they include; consumptions, investments, disposable income, exports among others. This study was meant to determine the impact of an increase consumption, investment, disposable income and exports (all in Billion Dollars) on the GDP growth. To make this a reality, economic data (secondary) was used and aimed to estimate the economic model shown below; GDP = β0 + β1C + β2I + β3PDI + β4NT In the above economic model, GDP was considered as the dependent variables (DV) while consumptions, investment, disposable income and net exports were considered as the independent variables (IV).Further, simple regression models of the nature GDP = β0 + β1ya (where yas are the IV) were estimated. From the results, investments and disposable income have a statistically significant impact on GDP while consumption and net exports are not significant at 5% level of significance. Considering each IV alone Vs GDP, then it is significant at 5% level of significance. Table of Contents Abstract 1 Table of Contents 2 List of Tables 2 List of Figures 2 Introduction 3 Literature Review 3 Data 4 Model 4 Results 4 Multiple Regression-Full Model 5 Simple Regressions 7 Conclusions 13 References 14 Appendix 15 List of Tables Table 1: Heteroskedasticity Test of the Full Model 6 Table 2: Full Model Coefficients with Robust Errors 6 Table 3: Endogenity Test Results 7 Table 4: Correlations 7 Table 5: Step-wise Regression Results 8 Table 6: Model with Consumption and GDP 8 Table 7: Model with Investments and GDP 10 Table 8: Model with Investments and GDP 11 Table 9: Model with Net Exports and GDP 13 List of Figures Figure 1: Consumption Vs GDP Line of Best Fit 9 Figure 2: GDP Vs Investments Line of Best Fit 11 Figure 3: GDP Vs Disposable Income Line of Best Fit 12 Figure 4: GDP Vs Net Exports Line of Best Fit 13 Introduction The GDP growth of a country is determined by several factors which include consumption expenditures, investments (domestic and external), disposable income, net exports (resulting from the difference between exports and imports) among many others. This study was meant to determine the impact of an increase consumption expenditures, investments, disposable income and exports on the GDP growth. To make this a reality, economic data (secondary) was used and aimed to estimate the economic model where GDP was considered as the dependent variables (DV) while consumptions, investment, disposable income and net exports were considered as the independent variables (IV). Literature Review GDP is the cumulative total amount of goods and services which a country produces within a given year. GDP growth changes year-in-year out depending on the economic conditions (amount of investments, consumption levels, disposable incomes and net exports). Other factors determine a country’s GDP growth and include stability of politics, environmental factors (global warming) and regional stability among others. According to Fischer (1993), a country’s GDP is also driven by the amount of spending the government does. Mankiw et al. (1992) supports Fischer’s argument introducing the aspect of inflation in GDP growth arguing that inflation is the phenomena of rising trends of prices in a country. As the prices go up, citizens tend to spend a lot of their savings than before. This leads lead to fewer or no savings which translates to less investment capital within a country and thus low or no GDP growth. Data Secondary data was used for this research study. This is so because it is easy to get such data as compared to survey and also since the kind of data cannot be gotten through a survey as it is historical data. However, survey data has more weight and makes tangible inferences as compared to secondary data utilized in this case. The data used here can be downloaded from: http://www.oswego.edu/~kane/econometrics/data.htm and is made up of the following variables; Year (Time variable); GDP (DV); Consumption (IV); Investments (IV); Disposable Income (IV); and Net Exports (IV). Model The study aimed at estimating the following economic model; GDP = β0 + β1C + β2I + β3PDI + β4NT. In this economic model, GDP was considered as the dependent variables (DV) while consumptions, investment, disposable income and net exports were considered as the independent variables (IV).Further, simple regression models of the nature GDP = β0 + β1ya (where yas are the IV) were estimated. These models are valid and reliable in the determination of how much the GDP changes with a unit increase in each IV holding the others (IVs) constant. Results Given that this is time series data, this first test carried out was to determine the presence of serial autocorrelation using the Durbin-Watson d-statistic (5, 74) = .3566969 (Gore & Altman 1992). From this result, it is evident that serial autocorrelation is not a problem. Further, heteroskedasticity test was carried out to determine whether the error terms of the full model have a constant variance (homoskedastic). Table 1: Heteroskedasticity Test of the Full Model From table 1 above, it is evident that the error terms of the full model do not have a constant variance; are not homoskedastic [Chi-square (14) = 46.03, p Read More