Grant Van Der Beken’s Web Page for Econometrics

 

Introduction: 

I am analyzing the correlation between income and the infant mortality rate in the United States.  As income increases the ability to pay for better doctors, hospitals, and prenatal care increases.  However, each American is born under the Constitution, which promises equality for all.  This should equate to the best healthcare options for all patients, and their infants.  All children born under American law should have the same chance at life, regardless of household income. 

            Turns out that income an indirect affect on the infant mortality rate.  However, poverty level does.  States with a high poverty rate tend to have higher rates of infant mortality.  This could mean a few things.  First, government programs to help those in need, are not working when it comes to healthcare.  Second, families below the poverty line are not able to provide for the child as well as families with larger incomes. 

 

Model:

            We would tend to believe that families with higher incomes are better able to provide for infants, thus lowering the infant mortality rate.  As stated, higher incomes lead to a higher standard of living, and access to better healthcare options.  I also looked at the correlations between the infant mortality rate and the percent of the state that had a college degree, and the number of doctors in the state.  Families that have a college degree have a better opportunity at a higher paying job.  We can also assume that the infant mortality rate will go down with parents who are better educated.  I also felt that access to doctors would be a reason that the infant mortality rate would be higher in some states.  States with more doctors raise the state’s average income, and also provide better healthcare to more patients. 

 

Data:

            The data used for this paper came from http://www.census.gov/statab/www/ranks.html.  The dependant variable is the infant mortality rate in percent.  My econometric equation involves the variables: income in dollars, percent of state below the poverty level, percent doctors per state, and percent of state with a college degree.

 

Results:

            As we can see from regression one, annual income with a t stat of 0.454909 is not significantly different from zero, having little effect on the infant mortality rate.  Poverty level has a high t statistic, 2.142687, showing that it has more of an effect on the IMR.  The preferred equation is:

Infant Mortality = B0 * Poverty level + B1 * Degree + B2 * Doctors

           

 

Conclusions: 

Infant mortality depends on the poverty level of the state, the education of the state and the number of doctors in the state.  Annual income seems to have little to do with the IMR based on the regression below.  One could argue that poverty level is associated very closely with income, skewing the results. The point still stands that the higher the state poverty then the higher the infant mortality is going to be. 

 

Dependent Variable: INFANTMORT

Method: Least Squares

Date: 11/05/03   Time: 13:11

Sample: 1 50

Included observations: 50

Variable

Coefficient

Std. Error

t-Statistic

Prob. 

POVERTYLEVEL

0.144224

0.067798

2.127268

0.0388

DOCTORS

0.008495

0.003660

2.320939

0.0248

DEGREE

-0.137935

0.069022

-1.998413

0.0516

C

6.680586

1.658691

4.027626

0.0002

R-squared

0.340984

    Mean dependent var

7.192000

Adjusted R-squared

0.298004

    S.D. dependent var

1.496505

S.E. of regression

1.253849

    Akaike info criterion

3.366932

Sum squared resid

72.31835

    Schwarz criterion

3.519894

Log likelihood

-80.17330

    F-statistic

7.933664

Durbin-Watson stat

2.074726

    Prob(F-statistic)

0.000228