Term Project

The Effects Geographic Location and Age
Have on
A Credit Score

Nicholas Rogers
EC315
10 March 2013

TABLE OF CONTENTS

PURPOSE STATEMENT AND MODEL………………………………………………………..3

DEFINITION OF VARIABLES…………………………………………………………………3

DATA DESCRIPTION…………………………………………………………………………..5

PRESENTATION AND INTERPRETATION OF RESULTS………………………………….7

WORKS CITED…………………………………………………………………………………10
DATA SOURCES……………………………………………………………………………….10
REGRESSION ANALYSIS AND CORRELATION MATRIX……………………attachment 1

PURPOSE STATEMENT AND MODEL

With the current state of the economy and job market in the United States, many individuals are shocked to hear that credit scores are becoming even more important now than before not only when it comes to making purchases or financial decisions, but also in other aspects of their everyday life. According to Forbes.com staffer, Heather Struck, in A Bad Credit Score Affects A Lot More Than Credit, ones FICO score can affect their ability to get a job, their insurance rates, and even how their personal relationships work out (2011). Some people have differing opinions on what can hurt or help a credit score. To keep from going into each of those possible outcomes, I am just going to pay attention to geographic location and age of the person seeking credit. The purpose of this project is to see what, if any, effect where a person lives and his age will have on his credit score. The dependent variable, CREDIT SCORE, is determined by the independent variables of geographic location (geo_loc) and age. The primary independent variable in this relationship is age or length of credit history because it has the most weight in the calculation between the two variables. According to a graph from MyFico.com, this is responsible for 15% of how a FICO score is obtained (MyFico.com, 2013). A person’s age is important to factor into a credit score. In general, the older a person is the higher credit score he will have. This study will show what relationship, if any, where a person lives and how old he is has on his credit score. The model is: Credit_score = consumer_age + geo_loc DEFINITION OF VARIABLES The dependent variable Credit_score is the purchasing power and the risk to creditors a consumer possesses. An individual’s credit score can be determined by the use of several factors. Whether the person meets all the criteria or not, he will still have a credit score. A lower score will result in not paying attention or simply not applying for credit. On the opposite end, if attention is paid to all the important factors and actions are swift, a higher score will be the result. The primary independent variable in this relationship is age or length of credit history because it has the most weight in the calculation between the two variables. Longer credit age and payment history will allow for increased credit scores. “Older borrowers have an advantage in maintaining a higher credit score not only because of the ability to have a more extensive payment history and older credit age, but also because they have had more time to clear negative debt or marks from their credit report if they had a less than spotless payment history” (freescore, 2012). This is not to say that younger people cannot have a high credit score, rather it is not necessarily that common. For this study, the median age of members of each state will be used to see how much age plays a part in one’s credit score. I expect to see that this independent variable will have a positive coefficient. The relationship between age and the independent variable is direct. All other things equal and if the individual takes care of what he is supposed to, the older he gets the higher his credit score will become until it reaches the highest echelon. The next independent variable is geographical location or a person’s region of residence. Just a bit of common sense shows that different areas of the United States will have different credit scores. Things such as cost of living in a certain area, taxes paid to the state, and the nature of the respective state’s economy can all have an effect on the average credit score for someone from that state or area. The southern states tend to have lower costs of living and their economic state seems to be affected more with shifts in the economy. According to Experian, the national average credit score is 687, and southern states typically report lower average credit scores (Governing.com, 2011). Since it is already known that the south has lower credit scores than the north, on average, this study will break the states into two different regions, the East and the West. Breaking these groups into the east and west, I expect to see a negative coefficient. I think there is an inverse relationship between geographic location and credit score.

DATA DESCRIPTION This study will first consider the correlation between age and credit scores. Data is reported based upon information gathered from the Census bureau. The table below outlines median age by states.
|State |
|AL |37.4 |
|AK |32.8 |
|AZ |35 |
|AR |37 |
|CA |34.8 |
|CO |35.7 |
|CT |39.5 |
|DE |38.4 |
|FL |40 |
|GA |34.7 |
|HI |37.5 |
|ID |34.1 |
|IL |36.2 |
|IN |36.8 |
|IA |38 |
|KS |35.9 |
|KY |37.7 |
|LA |35.4 |
|ME |42.2 |
|MD |37.7 |
|MA |39 |
|MI |38.5 |
|MN |37.3 |
|MS |35 |
|MO |37.6 |
|MT |39 |
|NE |35.8 |
|NV |35.5 |
|NH |40.4 |
|NJ |38.8 |
|NM |35.6 |
|NY |38.1 |
|NC |36.9 |
|ND |36.3 |
|OH |38.5 |
|OK |35.8 |
|OR |38.1 |
|PA |39.9 |
|RI |39.2 |
|SC |37.6 |
|SD |36.9 |
|TN |37.7 |
|TX |33 |
|UT |28.8 |
|VT |41.2 |
|VA |36.9 |
|WA |37 |
|WV |40.5 |
|WI |38.2 |
|WY |35.9 |

The study will also use average credit score data from different states from 2013. Data is reported based upon information gathered from a large database of credit scores. This particular information came from creditreport.com. This information will be used to create a correlation analysis for the study to determine if location affects an individual’s credit score. The tables below outline average credit scores by state.
[pic]

PRESENTATION AND INTERPRETATION OF RESULTS After completing the regression analysis the regression equation is as follows:
Avg_credit_score = 546.88 – 7.29(geo loc) + 3.78(age). Now that we have the regression equation, we can determine just how the variables relate to each other using a coefficient of determination, the adjusted R2. The adjusted R2 is used to measure the strength of the relationship between the set of independent variables and the credit score and also for the number of variables in the regression equation. This particular coefficient of determination shows that the independent variables listed account for 19.5% of the variance of credit scores. This seems to be a relatively low percentage, but is not surprising since there are so many variables that can affect one’s credit score. The variables used were just two of the many variables that could have been used. The next thing we could do to see how the independent variable correlates to the dependent variable is to run a F test. The 5-step hypothesis test will be used. 1) State the hypothesis: H0: U1 = U2 H1: the mean scores are not equal 2) Select the level of significance: 0.05 3) Determine the test statistic: the test statistic will follow the F distribution 4) Formulate the decision rule: df in the numerator = 2 Df in the denominator = 47 The critical value for 2 df in the numerator and 47 df (infinite df) in the denominator at a 0.05 significance level is 3.00. The rejection rule is to reject the null if the computed value of F exceeds 3.00. 5) Select the sample, perform the calculations, and make a decision. Using the ANOVA table, the computed value of F is 6.93 which is greater than the critical value of 3.00. The null hypothesis would be rejected. Using the p-value approach, the decision can be made on whether or not to reject the null hypothesis. From our critical value of 3.00, using the z table we get a value of .4987. From that, we subtract .4987 from .5000 (.5000-.4987) and get a p-value of 0.0013. Using this information, we would reject the null hypothesis because the p-value is less than the significance level of 0.05. Now that we know that the means are not equal, we can see that the independent variables listed do not bear the same weight when dealing with how a credit score is calculated. After completing the F-test, we know that the independent variables do not account for the same level of significance when it comes to obtaining a credit score. Now, we can see if what happened with the regression analysis is what was expected to happen when the study was started. Looking at the regression output the coefficients came out just as expected. Geographic location has an inverse relationship with credit scores. So, living in one part of the country will either mean that a credit score will be reduced or improved. Because the p-value of geographic location is rather large, this shows that while the relationship in inverse, it does not necessarily have a strong relationship to credit score. On the other hand, the age variable has a direct (positive) relationship with how credit scores are determined. Already stated, was that the older the person is, generally the better his credit score will be. This variable has a small p-value, which means that it is significant in determining the credit score of an individual. Again as stated previously, the age of a person is the most significant variable of the ones listed. Any time a multiple regression analysis is conducted; multicollinearity has to be taken into account. So, just what is multicollinearity? This is when independent variables are correlated. The general rule to see if this is a problem that exists, the correlation must be between -.70 and .70. Looking at a correlation matrix, the largest correlation is 0.57 and the smallest is 0.09. Since the numbers do not fall outside of the range from the general rule, it can be concluded that there is no problem with multicollinearity between the two variables of age and geographic location when it comes to determining a credit score.

WORKS CITED
Average Credit Scores by State. (2011). Retrieved February 17, 2013, from governning.com: http://www.governing.com/gov-data/economy-finance/average-credit-score-by-state.html

How Your Age Affects Your Credit Score. (2005). Retrieved February 17, 2013, from BCSAlliance.com:http://www.bcsalliance.com/credit_averagescores.html

National Average Credit Score. (2013). Retrieved February 17, 2013, from Money-zine.com: http://www.money-zine.com/Financial-Planning/Debt-Consolidation/National-Average- Credit- Score/

The Relationship between Credit Scores and Age. (2012). Retrieved February 17, 2013, from FreeScore.com: http://www.freescore.com/credit-score-and-age.aspx

What’s in my FICO Score. (2013). Retrieved January 27, 2013, from MyFico: http://www.myfico.com/crediteducation/whatsinyourscore.aspx

Struck, H. (2011, July 20). A Bad Credit Score Affects A Lot More Than Credit. Retrieved January 27, 2013, from Forbes: http://www.forbes.com/sites/heatherstruck/2011/07/20/credit-score-fico-can-hurt-you/

DATA
Average Credit Scores. (2013). Retrieved March 09, 2013, from CreditReport.com: http://www.creditreport.com/creditscores/creditratings/average-credit-scores.aspx

Median age, by state. (2010, June 10). Retrieved March 09, 2013, from USAToday: http://usatoday30.usatoday.com/news/nation/census/median-age-by-state.htm