Determinants of GDP
My topic is that how dependent variable, GDP, is affected by independent variables which include eight quantitative variables and one categorical variable. The quantitative variables are exports of goods and services, foreign direct investment, unemployment rate, value of agriculture, merchandise exports, value of service, value of industry and total labour force. Also, my categorical independent variable is the type of countries (i.e. these countries are classified into developing counties and developed countries). In my excel file, I collected data about GDP and these independent variables from World Data Bank[1]. Then I downloaded and organized specific data of 68 countries within year 2000, and I regarded these 68 countries as observations in my paper. The definitions of variables of my main interest are introduced in the following paragraph.
Firstly, I will briefly introduce the major variables in my paper. GDP (gross domestic product) is my dependent variable. I name it GDP in my excel file, which is generally defined by World Data Bank, “[GDP] is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products.” (World Data Bank, n.d.). Also, my category variable is “Country_Type” which is in column G of datacopy sheet, and the classifications of countries are from Wikipedia[2]. One of criteria Wikipedia uses to classify country is total income of a country: a country which has gross national income per capita below US$4,036 is treated as a developing country; a developed country has gross national income per capita more than US$4,036. The amounts of income may influence individuals', organizations' or countries' expense or demend, indirectly changing production of various sectors. That is the main reason why I use type of country as a categorical variable.
Next, I transform the category variable into dummy variables: when a country is a developing country, I code it “1” in column L; if not, it is “0”. I use “if function (=IF($K2=$L$1,1,0))” and drag it down to code the column L. In column M (i.e. “DevelopedCountry”), if the country is a developed country, I code it “1”; if not, this is “0”. Exports of goods and services, merchandise exports, total labor force, foreign direct investment and unemployment rate which is called “Export”, “Merchandise_Exports”, “LaborForce”, “Foreign_Investment” and “UnemploymentRate” respectively in the excel file are defined by common sense. According to World Data Bank, value of agriculture is the net output of a sector which contains forestry, hunting, and fishing, as well as cultivation of crops and livestock production; this is called as “Agriculture_Value” in my excel file(World Data Bank, n.d.) . “Service_Value” and “Industry_Value” are collected in my excel; however, after running regression model, I found that the two variables do not significantly affect GDP. Thus, I will not talk about them in the word paper in detailed.
In the full regression model, I regress GDP on all independent variables: “Exports”, “Foreign_Investment”, “UnemploymentRate”, “agriculture value”, “merchandise exports”, “Service_Value”, “Industry_Value”, “LaborForce” and “DevelopingCountry” for a sample of 68 countries. The “DevelopedCountry” is my base case and is omitted from the full regression. The output of full regression model is following:
[pic 1]
In the output of full regression, all statistics are assumed within 95% confidence level and alpha= 0.05. R square is 0.95228143 which means that about 95.23% of the variability in GDP is predicted from the nine independent variables. Thus, this is a good model which has a regression line almost fits the data to support my following analysis.
To find the relationships between GDP and the nine independent variables, I start with F test. My null hypothesis is that GDP is not linearly related to any of the explanatory variables (exports of goods and services, foreign direct investment, unemployment rate, value of agriculture, merchandise exports, value of service, value of industry and total labour force) in the population. In addition, my alternative hypothesis is that GDP is linearly related to at least one of the explanatory variables in the population. In the output, we can easily know: significance F = 5.97E-35= 0.0000000000000597 < 0.05. Therefore, I should reject the null hypothesis. The conclusion is that GDP is linearly related to at least one of the explanatory variables.