Econometrics Term Paper - Life Expectancy

...

However, the author was not done with his research and discusses another model he used, which consist of ten variables. His model included urban population, urban population growth, fertility rate, mortality rate, CO2 GDP, cuberrootGNI, HIV, school enrollment, access to telephones, and electricity produced by nuclear plants. This was a very useful model because the author was able to conclude that a ten percent increase in urban population growth is associated with a three year increase in life expectancy. In addition the author was able to conclude that CO2 GDP and the fertility rate also had a huge impact on life expectancy. The coefficient on CO2 GDP was -1.3, which means that a one-unit increase in CO2 GDP is associated with 1.3 decrease in life expectancy. The coefficient on fertility rate was -0.8, which means a one unit increase in fertility rate is associated with 0.8 year decrease in life expectancy. These results clearly empathize how valuable this information can be in order to increase life expectancy in many countries.

The Data:

I have found information on many of Bryant University’s library website databases, which includes researching on Ebsco, Econlit, and Jstor to further my knowledge of life expectancy. I have also used a very helpful “life expectancy analysis” paper written by D. Husek & V. Fucik’s, in order gain more insight into the expected future lifetime of an individual. In addition, Audrey Baer’s “Predicting Life Expectancy” analysis has really been a useful tool for me because it offers several tables of data. Her data examines the role that multiple independent variables have on life expectancy and she was able draw several conclusions based on these variables. I decided to conduct research on 189 countries because I wanted to analyze as many countries as I could in order to receive more accurate information. The research taken from these 189 countries were taken from 1990 to 2004 simply because I found a lot of sufficient data associated with life expectancy during this time period. Another useful resource I used was Karnjana Sanglimsuwan’s article called, “The Relationship between Health and Environment: Econometric Analysis”. This reference article really helped build a better understand of how life expectancy depends on improved water sources, infant mortality, population density, health status, and many more variables. However, it was clear after reading this article that I must include improved water sources as well as infant mortality rate in my research paper. This article also provided a couple of data tables that further my knowledge of these variables

Empirical Model:

Life Expectancy = B0 + B1 improved water sources + B2 physicians + B3 HIV + B4 health expenditure

My dependent variable in this model is life expectancy and it is measure as life expectancy at birth for both genders. This variable simply explains the number of years a newborn infant would live if previous patterns of mortality, at the time of birth, were to hold constant through the person’s life.

My first explanatory variable in my model is improved water sources and is measured by the percent of the population that has access to an improved water source. This variable is very significant to my research paper because it serves as my focus variable. However, it is important to realize that improved water sources include piped water, public tabs or standpipes, tube wells, protected drug wells, rainwater collection, and more.

My next explanatory variable is Physicians and it is measured as the number of physicians per 1,000 people. This variable is very important because the amount of physicians in each country can significantly contribute to the life expectancy of people living within each country.

My next explanatory variable in my model is HIV and it is measured as the percent of the population between the ages of 15-49 who are infected with HIV. This variable is also significantly important because this disease is very popular in developing countries and can have huge implications on a person’s health and life expectancy

My final explanatory variable in my model is health expenditure per capita measured in US dollars and it is the sum of both the public and private health expenditures as a ratio of total population. However, this data does not include provision of water and sanitation but does include emergency aid, health services, and nutrition activities. In addition, the data was logged in order to reduce heteroskadicity and due to the high dollar amounts. This is very significant for my regression because it excludes provision of water and sanitation because my dependent variable is improved water sources.

Empirical Results:

The next step in my research was to run the regression of life expectancy against my dependent and explanatory variables; improved water sources, physicians, HIV, and health expenditure. In the first table there is a basic summary of the number of observations, the minimum and maximum, standard deviation, as well as the mean of all the variables included in my model. However, some variables had a lot more observations than other variables did and I noticed that life expectancy as well as improved water sources had about 2,600 more observations than physicians. In addition, I also found it interesting that the maximum number of physicians were so low compared to the other variables.

Variable

Obs

Mean

Std. Dev

Min

Max

Life Expectancy

4506

66.71851

10.10128

26.76378

83.33195

Improved Water

4599

82.58056

18.74789

13.2

100

Physicians

1907

1.88446

1.404881

.007

7.739

...