College Students' Knowledge Of Personal Finance

WORKING PAPER --- PLEASE DO NOT QUOTE WITHOUT PERMISSION

A Survey of College Students’ Knowledge of Personal Finance

Katherine M. Sauer Assistant Professor of Economics University of Southern Indiana 8600 University Blvd. Evansville, IN 47712 Phone: (812) 465-7034 Fax: (812) 465-1044 [email protected]

Gregory P. Valentine Professor, Business and Economic Education University of Southern Indiana 8600 University Blvd. Evansville, IN 47712 Phone: (812) 465-1610 Fax: (812) 465-1044 [email protected]

1

ABSTRACT This paper provides evidence of a lack of financial literacy among college students in southern Indiana. A 44 question survey covering five areas of personal finance (credit, vehicle insurance, banking, life insurance, and housing rental) was given to 234 students at a mid-sized comprehensive university. Results indicate that the average score was 53.5% correct, which is in line with some national studies but lower than the findings from the most recent national assessment of college students’ personal financial literacy. Students scored the lowest on life insurance questions and the highest on housing rental questions; traditionally aged college students scored lower than non-traditional students. Students who are employed scored slightly better than those who did not work and students who paid their own vehicle insurance scored better than those who did not. Regression analysis revealed that several demographic characteristics influenced the students’ test scores. This study suggests that there are opportunities for colleges and universities to implement a personal finance course requirement at the undergraduate level.

ACKNOWLEDGEMENT We are grateful to participants of the 2007 NCEE/NAEE/GATE annual conference in Denver, Colorado on October 3, for helpful comments.

INTRODUCTION Alan Greenspan (2003), Chairman of the Board of Governors of the United States Federal Reserve, stated that in order to achieve an effective and efficient functioning of financial markets, there is a need for more financial education at all levels of society. Recent evidence

2

highlights that the need is growing. In January 2008, the U.S. Federal Reserve reported that by the end of 2007 consumers had amassed $937.5 billion worth of revolving credit at an average interest rate of 14.35%. According to Halperin and Smith (2007), researchers for the Center for Responsible Lending, overdraft fees on bank accounts total $17.5 billion. The mortgage delinquency rate is the highest in two decades, and the number of homes in foreclosure is the highest ever (Mortgage Bankers Association, 2007). The American Bankruptcy Institute reported that in 2007 over 800 thousand consumers declared bankruptcy. CNNMoney (2006) reported that 43% of households are in danger of running short of retirement funds. Given the state of American consumers’ finances, it would seem that Greenspan was correct. Where then are consumers getting their financial education? The Hartford Financial Services Group (2007) recently found that the majority of college students learn the most about personal finance from their parents. In addition, less than half of the students say that their parents make a consistent and conscientious effort to teach them about personal finance. The study further stated that students and parents both agree that college students are not well prepared to deal with the financial challenges that lie ahead. Anecdotal evidence aside, the level of personal financial literacy of young adults has been examined several times by various organizations. The Consumer Federation of America and the American Express Company assessed financial literacy among high school students in 1991 and college students in 1993. They administered a 52 question survey to 428 teenagers at shopping malls in eight metropolitan areas around the country. The students answered 42% of the questions correctly. The survey of college students was given to 1,000 full-time college juniors and seniors at 75 institutions with only 51% of the questions answered correctly. The Jump$tart Coalition nationally surveys high school seniors on personal finance every two years.

3

The 2008 study revealed that the students answered 48.3% of the questions correctly. The survey was given to approximately 6,800 students in 40 states. Also in 2008, Jump$tart Coalition released its first ever survey of college students. The survey was given to 1,030 students nationwide and on average 62% of the questions were answered correctly.

Personal finance literacy has received some attention in the academic literature; however, the focus has been mainly on high school students as opposed to college students. Notable examples of high school studies include Tennyson and Ngyen (2001) and Valentine and Khayum (2005). Tennyson and Ngyen compared financial literacy levels among high school students from states with and without mandates for financial literacy curriculum. They administered the Jump$tart Coalition survey to 1,643 high school students in 31 states. Roughly half of the students in the sample had been exposed to personal finance education. They did not find a significant difference in test scores. Students from states that mandate a personal finance course scored on average 56.9% correct and students from states without a mandate scored 56.5%. Valentine and Khayum used the Consumer Federation of America/American Express survey to study financial literacy among urban and rural high school students. They surveyed 312 high school students in southwestern Indiana. They found no significant difference between the scores of urban and rural students. On average, urban students scored 50% and rural students scored 51%. There are several related strands of literature on college students and personal finance. Financial attitudes and financial behaviors have received attention. Examples include Roberts and Jones’ (2001) work on debt attitudes and Xiao, Noring, and Anderson’s (1995) study of students and credit cards. The impact of financial knowledge has been explored by Borden, Lee, Serido, and Collins (2008); Peng, Bartholomae, Fox, and Cravener (2007); and Fox,

4

Bartholomae, and Lee (2005) among others. Several studies (e.g. Chen and Volpe (1998) and Markovich and DeVaney (1997)) examine links between financial attitudes and knowledge and explore the effects of student characteristics like gender, ethnicity, and employment. At the college level, financial literacy has been examined by two studies, Danes and Hira (1987) and Chen and Volpe (1998). Danes and Hira questioned 323 students at Iowa State University to determine their level of knowledge of money management, credit cards, and insurance. They found that students had a low level of personal financial knowledge. Chen and Volpe conducted a national study. They surveyed 924 students from 14 college campuses. On their assessment, the average score was 51%. In light of the rather sparse academic literature on college students’ personal financial literacy, the authors sought to contribute the following. First, we benchmarked the level of personal finance literacy among college students in southern Indiana. A survey of personal financial literacy was administered to college students taking an elective personal finance course. Next, we analyzed the scores from each section of the survey (credit, vehicle insurance, banking, life insurance, housing rental) to see where students’ financial knowledge deficiencies specifically lie. Finally, we examined demographic factors to determine if personal financial knowledge differed with student characteristics. METHODOLOGY A survey of personal financial knowledge was given to college students in southern Indiana. From individual students’ scores, the mean percentage correct score was computed to benchmark the level of financial literacy. T-tests were used to analyze any differences in mean scores for various groups of students (e.g. males vs. females). Regression analysis was used to

5

determine whether student characteristics affected their scores. Mean scores, t-tests, and regression analysis was also performed for each of five sections of the survey. The survey consisted of the financial literacy questions developed by the Consumer Federation of America. This instrument was chosen because Chen and Volpe’s 1998 instrument had solely been used in a pilot study while the selected instrument came out of collaborations from Educational Testing Services (ETS), American Express, the Consumer Federation of America, and The Psychological Corporation. The survey instrument covered five areas of personal finance: credit, banking, vehicle insurance, life insurance, and housing rental. In addition to personal finance questions, the survey included demographic questions and questions about students’ exposure to personal finance issues (e.g. credit card usage). College students enrolled in sections of an elective personal finance class at a mid-sized comprehensive public four-year university were included in the survey. Students were given the option of participating or not participating in the survey without any adverse consequences if they chose not to participate; all students participated in the survey. The instrument was administered on the first day of class to those students who were enrolled during the fall semesters of 2006 and 2007 and during the spring semester of 2007. This class is an elective in the Associate Degree in Business program as well as a general elective for all baccalaureate degree programs. The sample was comprised of 234 respondents. Table 1 presents the sample demographics and Table 2 presents the data on experience with personal finance matters. Respondents were predominantly Caucasian, business majors of traditional college age, not married, and without children. A majority of students worked at least one hour per week, had at least one credit card, had some type of bank account, and had a vehicle.

6

The survey contained 44 multiple choice questions related to personal finance. Fourteen questions on credit focused on personal finance terms, where to obtain credit, best indicators of the cost of credit, determining interest, and making payments. The topics covered in the six checking and saving account questions dealt with computing the monthly cost of a checking account, avoiding checking account fees, analyzing the most important factors when shopping for an account, and calculating the various types of interest compounding. The six items on automobile insurance dealt with coverage, costs, and complaint resolution. The twelve questions that dealt with life insurance concerned types of insurance, costs, and problem resolution. The six items about housing rental concerned tenants rights and obligations, landlord rights and obligations, and how to negotiate a lease. A sample question from each section can be found in Table 3. After administering the survey, the mean percent of correct scores for each question, each section, and the entire survey were computed. Standard deviations were also noted. The students’ scores were then grouped by demographic characteristics and t-tests with unequal variances were performed for determining significant difference in scores. Characteristics used for comparison were age, gender, hours worked per week, credit card usage, party responsible for paying vehicle insurance, and college major. These groups were chosen on the basis of previous scholarly works (Chen and Volpe, (1998); Tennyson and Nguyen, (2001); and Valentine and Khayum, (2005)). To determine the effect students’ characteristics and experiences may have had on their survey scores, regression analysis was performed. The students’ scores were regressed on dummy variables for student characteristics and experiences. The regression equation used was

7

Scorei = αi + β1AgeDummy + β2GenderDummy + β3MaritalStatusDummy + β4ChildrenDummy + β5JobDummy + β6EthnicityDummy + β7CreditCardDummy + β8VehicleDummy + β9VehicleInsPayDummy + β10BankingDummy + β11HousingDummy+ β12MajorDummy + errori

where the dummy variables were assigned for each category of student characteristic in Table 1 and Table 2. Separate regressions were run with the overall score and score from each section as the dependent variable. The Shapiro-Francia test confirmed normal data. Because of the wide variety in the variables, the Breusch-Pagan test revealed that heteroskedasticity was a problem. Thus, robust standard errors were used. Multicollinearity was not found to be an issue (no VIF values were greater than 10). FINDINGS The overall mean percent correct score was 53.5%. Mean scores were found to be statistically significantly different for students classified by age, working status, and party responsible for paying vehicle insurance. When breaking the scores down by content section (e.g. banking questions), students scored the highest on the housing rental section and the lowest on the life insurance section. The results from the t-tests for differences in mean scores were varied when classifying students into groups. When regressing student scores on student demographic data, several characteristics were found to have significant impact. Table 4 reports the summary statistics for the survey. The average score was 23.5 correct out of a possible 44 questions, or 53.5% with standard deviation of 0.092. The mean percents correct for the credit, vehicle, and banking sections of the survey were similar with scores of 52.5%, 51.1%, and 53.6% respectively. Students scored the lowest on the life insurance questions (47.9%) and highest on the housing rental section (69.2%). The high and low section scores are perhaps unsurprising. One would not expect many college students to have purchased or dealt with life insurance. As for the higher score on the 8

housing rental questions nearly half of the students in the sample lived in an apartment. Another 30 students are home owners, and it is plausible that they had lived in rental housing before purchasing their home. Sixty-nine students lived in campus housing; one might consider this experience with quasi-rental housing to increase student knowledge on rental issues. In addition, these students likely have friends who live in apartments or are looking to rent an apartment at some point themselves. The scores on the other three sections might be lower than one would expect. With all the types of credit related situations that students encounter (e.g. credit cards, student loans, mobile phone account approval); one would speculate that students would do better on this section. The low score means that students are fundamentally lacking knowledge on many credit related matters. Turning to the banking section, the vast majority of the students had a checking and savings account. Students did well on the questions relating to these accounts but were lacking awareness of bank accounts beyond these two types of accounts. Perhaps most surprising was the low score on the vehicle insurance question. Every student with a vehicle also had insurance. However, vehicle insurance is required in the state of Indiana. One might think it possible that the students purchase insurance without taking time to understand their coverage. To determine if there was a significant difference in the scores of various groups of students, the mean scores were further analyzed. The students were broken down by age, gender, employment, credit card usage, the party responsible for paying for vehicle insurance, and by college major. Students are grouped by gender to check for any differences among the sexes. In the age category, students were categorized by those who were of traditional college age and those who were not. One would expect that older students might have more familiarity with personal finance issues and would thus score higher. Another measure of exposure to personal

9

finance might come through employment. Students earning their own money might have more experience managing their money as well. Credit card use is another natural way to group students by exposure to personal finance issues, as is whether or not the student pays for their own vehicle insurance. Finally, since this survey was given in a College of Business course, students were grouped into business majors and non-business majors. Through their coursework, business students might be expected to have a higher level of knowledge of personal finance issues. Table 5 shows the results of t-tests for data with unequal variances. For gender, major, and number of credit cards, the mean scores were not significantly different. However, as expected, traditionally aged college students scored lower (52.1%) than non-traditional students (58.7%). Also, students who were not currently employed scored lower (51.5%) than students who were working (54.53%). Those who did not pay the cost of their vehicle insurance scored lower (51.6%) than those who did (56.0%). Tables 6 through 10 report the results of t-tests for differences among students, broken down for each section of the survey instrument. Table 6 shows the credit section results. Similar to the overall mean score results, students’ age, working status, and payment for vehicle insurance produced significant differences in mean scores. Again, traditional college students scored lower than non-traditional students (51% vs. 58.3%). Students without jobs scored lower than students with jobs (50% vs. 53.8%). Students who did not pay for their vehicle insurance scored lower than those who did (51.1% vs. 54.4%). For gender, major, and number of credit cards, the mean scores were not significantly different. Table 7 reports the vehicle insurance section. Again, students who personally paid for the cost of their vehicle insurance scored higher than those who did not (55.2% vs. 48.8%) and non-

10

traditional students scored higher than traditional college students (59.4% vs. 48.9%). In addition, females scored higher than males on this section (55.6% vs. 45.4%) and business majors scored higher than non-business majors (53.6% vs. 46.3%). Students who were working did not score significantly different from students who were not. The number of credit cards a student had did not result in different scores. For the banking section, Table 8 reports the findings. In this section, males scored higher than females (60.3% vs. 48.3%). Students who do not pay for their vehicle insurance scored higher than those who do (56.9% vs. 49.8%). There is no particular reason to expect this result. Significant differences in the mean scores were not found for students broken into groups by age, working status, number of credit cards, or college major. Table 9 reports life insurance scores. Non-traditional college students scored higher than their counterparts (55.4% vs. 46%). This result is not surprising. Older students would be more likely to have dealt with or purchased life insurance. Students who paid for their own vehicle insurance scored higher than those who did not (54% vs. 43.3%). There is no immediate reason to expect this result. Business students scored lower than students with other majors (44.8% vs. 53.9%). Gender, working status, and number of credit cards did not produce significant differences in scores. In Table 10, presenting the housing rental section, there was no statistically significant difference among the groups. This is most likely because regardless of other individual characteristics, students had familiarity with housing rental. Table 11 and Table 12 contain the results of the regression analysis. When using dummy variables, the intercept term reflects the reference case. For this analysis, the reference case was Caucasian, male, age 18-22, married, had no children, did not work, not a credit card user, did

11

not have a vehicle, currently resides in a university residence hall, did not have home/rental insurance, and indicated business as a major area of study. For the regression using the overall score as the dependent variable, several student characteristics were significant. From the age dummy variables, being aged 30-39 exerted positive influence on the student’s score. Perhaps this is not unexpected since as people get older they may have had personal experience with more financial instruments. However, this would not explain why being aged 40 and older did not affect scores. By closer inspection, our sample contains only five students from this age group. It is very possible that those five students are not representative of the financial experience of their 40 and older peers in general. Some of the race and marital status variables influenced the test scores; being AfricanAmerican or Hispanic negatively affects scores. This result is very similar to the findings of Murphy (2005). As for marital status and children, being single or divorced positively influenced scores. Having one or two children raised the overall score, but having three children did not affect the score. However, only three students in our sample had three children. Several other variables were found to be significant. Working 1 to 5 hours per week positively impacts scores; curiously, working more than 5 hours did not have an effect on scores. For the credit card variables, having exactly two credit cards negatively affected scores. There does not seem to be an immediate reason for such a phenomenon. Having a checking account did not exert significant influence while having a savings account resulted in a negative impact on scores. Since most students had a checking account, perhaps there was not enough variation in the data for that variable to be significant. For the housing variables, living in a fraternity or sorority house had positive impact on scores. The other living arrangements did not significantly affect the scores. Some of the student major choices had impact on the scores. Health

12

professions, engineering, and liberal arts majors’ exerted positive influence on overall scores. The reference case was also significant and positively impacted overall scores. This means that a married, Caucasian, male, business student, age 18-22, with no children, who was not working, who had no credit cards, no vehicle, who resided in a university residence hall, without home/rental insurance translated into a higher overall score. The results for regressions using section scores as the dependent variables were varied. For the age variables, there was some significance on the scores in the banking section. Being a student in the mid to late twenties produced a negative impact on scores while being aged 30-39 positively impacted the score. In the credit, vehicle insurance, life insurance, and housing sections, age was not a significant variable. Traditionally aged college students were included in the reference case. As for gender, being female had a significant impact on both the vehicle insurance section score and the banking section score. For vehicle insurance, the impact was positive and for banking the impact was negative. Males were included in the reference case. The ethnicity variables were significant in several of the regressions. The AfricanAmerican coefficient was negative and significant in the credit section, banking section, and housing section. Being African-American resulted in lower scores on these sections. The coefficient on the Hispanic variable was only significant in the life insurance section. It produced a negative impact on the life insurance section score. The coefficient on the “other” ethnic variable was not significant in any of the specifications. Caucasian was included in the reference case.

13

In terms of marital status, being divorced was significant in each of the regressions. The coefficient was positive in all regressions. For unmarried students, there was a positive impact on the banking, life insurance, and housing scores. Turning to the number of students’ children, coefficient significance varied by section. For students with one child, the credit, vehicle insurance, and housing scores were positively impacted but the life insurance score was lower. For students with two children, the coefficient was positive and significant solely for the life insurance regression. For students with three children, credit and banking scores were negatively impacted while housing scores were positively impacted. Students with no children were included in the reference case. There was no clear trend for the variables for hours worked. Working 1 to 5 hours per week positively impacted credit and vehicle insurance section scores. Working 6 to 10 hours positively impacted credit section scores. The coefficient on working 11 to 20 hours was positive for the banking section. Working 21 to 30 hours and 31 to 40 hours had positive influence on the vehicle insurance scores. Working over 40 hours a week was not significant. The reference case was working zero hours. The coefficients on credit cards also varied in significance. Students with one or two credit cards had lower scores in the housing section. Having three credit cards resulted in a positive impact on the banking section and housing section scores. The coefficients for four or five credit cards were not significant in any of the regressions. However, having six credit cards produced positive impacts on the credit, banking, and life insurance scores and negative impacts on the housing section scores. The reference case was having no credit cards. For students with vehicles, the coefficients were not significant in any of the regressions. 96.6% of the students in the sample had a vehicle. As for the party responsible for paying for

14

vehicle insurance, students whose parents paid scored lower on the life insurance section. Students whose spouse paid scored higher on the vehicle insurance section. In the sample, only 3 students had spouses who paid the vehicle insurance. Students whose relative paid the vehicle insurance scored lower on the vehicle insurance section, life insurance section, and housing section. Students paying for their own insurance were the reference case. For the banking instruments, the coefficients for students with checking accounts were positive for the banking and housing regressions but negative for the life insurance regression. The coefficient on savings accounts was negative for the credit, vehicle insurance, life insurance, and housing regressions. It was not significant for the banking regression. The reference case was a student without a savings or checking account. Few of the housing coefficients were significant. Having home/rental insurance negatively impacted the vehicle insurance score. Living with parents had positively affected scores in the banking and housing regressions. The same was true for students living alone in an apartment. None of the coefficients for shared apartments or living in a dormitory were significant. For students living in a fraternity or sorority house, the coefficient was positive for the housing regression. Students living in a home that they owned were the reference case. Finally, the students’ choice of major was examined. The education majors had higher scores in the life insurance regression and lower scores in the housing regression. The health professions major coefficients were positive in the credit and vehicle insurance regressions. The scores on the credit, life insurance, and housing sections were higher for engineering majors. The engineering coefficient was negative in the banking regression. For liberal arts majors, the coefficient was positive for the life insurance regression. The life insurance and housing scores were positively impacted by undecided majors. Business majors were the reference case.

15

CONCLUSIONS and RECOMMENDATIONS This paper provides evidence of a lack of financial literacy amount college students in southern Indiana. The average score on a personal finance literacy survey was 53.5%. This is in line with some national studies but lower than the findings from the most recent national assessment of college students’ personal finance literacy. This paper further analyzed students’ financial literacy by examining students’ scores for questions in specific areas of personal finance: credit, vehicle insurance, banking, life insurance, and housing. Scores on the credit, vehicle insurance and banking sections were all in the low 50% range. Students did slightly worse on the life insurance section (48%) and much better on the housing section (69%). Regression analysis was used to see if students’ demographic characteristics or familiarity with financial instruments impacted their scores on the financial literacy survey. Results were varied. While students in southern Indiana demonstrated financial literacy consistent with the knowledge of students nationwide, the lack of personal finance knowledge is far from laudable. This study suggests that there are opportunities to design and add a personal finance requirement to the university curriculum.

REFERENCES American Bankruptcy Institute. (2007). Annual business and non-business filings by year (1980-2007). Retrieved January 11, 2008, from http://www.abiworld.org/AM/AMTemplate.cfm?Section=Home&TEMPLATE=/CM/Co ntentDisplay.cfm&CONTENTID=51718 Borden, L., Lee, S. , Serido, J. , & Collins, D. (2008). Changing college students’ financial knowledge, attitudes, and behavior through seminar participation. Journal of Family and Economic Issues, 29(1), 23-40. Chen, H. & Volpe, R. (1998). An analysis of personal financial literacy among college students. Financial Services Review, 7(2), 107-129. 16

CNNMoney.com (2006, June 6). Study: 43% won’t have enough in retirement. Retrieved November 1, 2007, from http://money.cnn.com/2006/06/06/retirement/risk_index/index.htm Consumer Federation of America. (1991). High school student consumer knowledge: A nationwide test. Washington, DC. Consumer Federation of America. (1993). College student consumer knowledge: A nationwide test. Washington, DC. Danes, S., & Hira, T. (1987). Money management knowledge of college students. The Journal of Student Financial Aid, 17(1), 4–16. Fox, J., Bartholomae, S., & Lee, J. (2005). Building the case for financial education. The Journal of Consumer Affairs, 39(1), 195–214. Greenspan, A. (2003). The importance of financial and economic literacy. Social Education. 67(2), 70-72. Halperin, E. & Smith, P. (2007). Out of balance: Consumers pay $17.5 billion per year. Center for Responsible Lending. Retrieved November 1, 2007, from http://www.responsiblelending.org/issues/overdraft/reports/page.jsp?itemID=33341925 Hartford Financial Services Group, The. (2007, April 21). New survey by the Hartford reveals financial literacy communication gap among college students and parents. Retrieved January 11, 2008, from http://ir.thehartford.com/releasedetail.cfm?releaseid=237682 Jump$tart Coalition for Personal Financial Literacy. (2008). 2008 Survey of personal financial literacy among high school students. Retrieved May 30, 2008, from http://www.jumpstart.org/fileindex.cfm Jump$tart Coalition for Personal Financial Literacy. (2008). 2008 Survey of personal financial literacy among college students. Retrieved May 30, 2008, from http://www.jumpstart.org/fileindex.cfm Markovich, C., & DeVaney, S. (1997). College seniors’ personal finance knowledge and practices. Journal of Family and Consumer Sciences, 89(2), 61–65. Murphy, A.J. (2005). Money, money, money: An exploratory study on the financial literacy of black college students. College Student Journal, 39(3), 478–488. Mortgage Bankers Association. (2007, December 6). Delinquencies and foreclosures increase in latest MBA national delinquency survey. Retrieved January 11, 2008, from http://mortgagebankers.org/NewsandMedia/PressCenter/58798.htm Peng, T., Bartholomae, S., Fox, J., & Cravener, G. (2007). The impact of personal finance 17

education delivered in high school and college courses. Journal of Family and Economic Issues, 28(2), 265–284. Roberts, J., & Jones, E. (2001). Money attitudes, credit card use, and compulsive buying among American college students. The Journal of Consumer Affairs, 35(2), 213–240. Tennyson, S. & Nguyen, C. (2001). State curriculum mandates and student knowledge of personal finance. Journal of Consumer Affairs, 35(2), 241-262. U.S. Federal Reserve. (2008, January 10). Consumer credit. Federal Reserve Statistical Release. Retrieved January 11, 2008, from http://federalreserve.gov/releases/g19/20080110/ Valentine, G. & Khayum, M. (2005). Financial literacy skills of students in urban and rural high schools. Delta Pi Epsilon Journal, 67(1), 1-10. Xiao, J., Noring, F., & Anderson, J. (1995). College students’ attitudes towards credit cards. Journal of Consumer Studies and Home Economics, 19(3), 155–174.

18

Table 1 Student Characteristics: Demographics (N = 234) n

percentage

18 - 22 23 - 29 30 - 39 40 and up

186 34 9 5

79.5% 14.5% 3.8% 2.1%

male female Ethnicity Caucasian African Am. Hispanic Other Marital Status not married married divorced Number of Children none one two three Hours Worked 0 hours 1-5 hours 6-10 hours 11-20 hours 21-30 hours 31- 40 hours over 40 hours Major business education health professions engineering liberal arts undecided

104 130

44.4% 55.6%

219 9 4 2

93.6% 3.8% 1.7% 0.9%

168 35 31

71.8% 15.0% 13.2%

216 4 11 3

92.3% 1.7% 4.7% 1.3%

82 7 25 42 42 24 12

35.0% 3.0% 10.7% 17.9% 17.9% 10.3% 5.1%

154 11 14 7 14 34

65.8% 4.7% 6.0% 3.0% 6.0% 14.5%

Age

Gender

19

Table 2 Student Characteristics: Financial Experience (N = 234) n Number of Credit Cards None 92 One 70 Two 45 Three 10 Four 6 Five 5 Six 6 Vehicle Yes 226 no 8 Vehicle Insurance insured 226 uninsured 0 no vehicle 8 Who Pays for Vehicle Insurance student 100 parent 116 spouse 3 relative 7 no vehicle 8 Checking Account Yes 221 No 13 Savings Account Yes 174 No 60 Housing Situation home owner 30 with parents 49 single apartment 43 shared apartment 38 Greek system house 5 dormitory 69 Home or Rental Insurance No 183 Yes 51

percentage 39.3% 29.9% 19.2% 4.3% 2.6% 2.1% 2.6% 96.6% 3.4% 96.6% 0.0% 3.4% 42.7% 49.6% 1.3% 3.0% 3.4% 94.4% 5.6% 74.4% 25.6% 12.8% 20.9% 18.4% 16.2% 2.1% 29.5% 78.2% 21.8%

20

Table 3 Sample Questions Survey Section

Question

Credit

The most important factors that lenders use when deciding to approve a loan are a) marital status and number of children. b) education and occupation. c) age and gender. d) bill-paying record and income.

Vehicle Insurance

If a car is stolen, what type of insurance coverage pays for its replacement? a) uninsured motorist b) liability c) comprehensive d) collision

Banking

Which deposit account usually pays the most interest? a) certificate of deposit b) passbook savings account c) NOW account d) money market account

Life Insurance

A policy loan is not available on which type of life insurance? a) term life b) whole life c) universal life d) variable life

Housing Rental

If a tenant if forced to break a 12-month lease after 9 months because of a family emergency, the tenant is legally obligated to pay the landlord a) no additional rent b) 1 month additional rent c) 2 months additional rent d) 3 months additional rent

21

Table 4 Summary Statistics

mean score mean percent correct

total

credit

vehicle insurance

bank accounts

life insurance

housing rental

23.5 53.5%

7.3 52.5%

3.1 51.1%

3.2 53.6%

5.7 47.9%

4.1 69.2%

median score median percent correct

24

7

3

3

6

5

54.5%

50.0%

50.0%

50.0%

50.0%

83.3%

standard deviation

0.092

0.108

0.216

0.201

0.157

0.221

35 14 44

11 3 14

6 0 6

6 1 6

10 1 12

6 1 6

maximum score minimum score possible score

22

Table 5 Analysis of Overall Mean Scores

n

average score

18-22 23 and older

186 48

52.11% 58.71%

male female Working not working working Credit Cards none one or more Vehicle Insurance other pays pay own Major business other ** significant at 5%

104 130

53.21% 53.67%

Age

Gender

82 152

51.50% 54.53%

92 142

52.89% 53.84%

126 100 154 80

difference -0.066**

t statistic -3.831

p value 0.000

-0.005

0.938

0.350

-0.030**

-2.505

0.013

-0.010

-0.763

0.446

0.043**

-3.493

0.001

-0.016

-1.118

0.266

51.75% 56.03% 52.94% 54.49%

23

Table 6 Analysis of Mean Scores, Credit Section

n

average score

18-22 23 and older

186 48

50.96% 58.33%

male female Working not working working Credit Cards none one or more Vehicle Insurance other pays pay own Major business other ** significant at 5%

104 130

52.47% 52.47%

Age

Gender

82 152

50.00% 53.80%

92 142

51.09% 53.37%

126 100 154 80

difference -0.074**

t statistic -3.633

p value 0.001

0.000

0.000

1.000

-0.038**

-2.547

0.012

-0.023

-1.581

0.115

-0.034**

-2.327

0.021

-0.015

-0.940

0.349

51.08% 54.43% 51.94% 53.48%

24

Table 7 Analysis of Mean Scores, Vehicle Insurance Section

n

average score

18-22 23 and older

186 48

48.92% 59.38%

male female

104 130

45.35% 55.64%

Age

Gender

Working not working working Credit Cards none one or more Vehicle Insurance other pays pay own Major business other ** significant at 5%

82 152

47.56% 52.96%

92 142

49.46% 52.11%

126 100 154 80

difference -0.105**

t statistic -3.359

p value 0.001

-0.103**

-3.802

0.000

-0.054

-1.970

0.050

-0.027

-0.951

0.343

-0.064**

-2.189

0.030

0.073**

2.352

0.020

48.81% 55.17% 53.57% 46.25%

25

Table 8 Analysis of Mean Scores, Banking Section

N

average score

18-22 23 and older

186 48

54.30% 51.04%

male female

104 130

60.26% 48.33%

Age

Gender

Working not working working Credit Cards none one or more Vehicle Insurance other pays pay own Major business other ** significant at 5%

82 152

51.83% 54.61%

92 142

50.72% 55.52%

126 100

56.88% 49.83%

154 80

difference 0.033

t statistic 0.862

p value 0.392

0.119**

4.811

0.000

-0.028

-1.042

0.299

-0.048

-1.856

0.065

0.071**

2.595

0.010

0.001

0.048

0.962

53.68% 53.54%

26

Table 9 Analysis of Mean Scores, Life Insurance Section

n

average score

18-22 23 and older

186 48

45.97% 55.38%

male female

104 130

47.52% 48.21%

Age

Gender

Working not working working Credit Cards none one or more Vehicle Insurance other pays pay own Major business other ** significant at 5%

82 152

46.54% 48.63%

92 142

49.18% 47.07%

126 100

43.32% 54.00%

154 80

difference -0.094**

t statistic -3.474

p value 0.001

-0.007

-0.330

0.742

-0.021

-0.939

0.349

0.021

1.018

0.310

-0.107**

-5.561

0.000

-0.090**

-4.390

0.000

44.81% 53.85%

27

Table 10 Analysis of Mean Scores, Housing Rental Section

n

average score

18-22 23 and older

186

68.10%

48

73.26%

male female

104 130

67.15% 70.77%

82 152

68.50% 69.52%

92 142

70.11% 68.54%

Age

Gender

Working not working working Credit Cards none one or more Vehicle Insurance other pays pay own Major business other ** significant at 5%

126 100

67.99% 70.83%

154 80

70.13% 67.29%

difference -0.052

t statistic -1.508

p value 0.136

-0.036

-1.243

0.215

-0.010

-0.356

0.722

0.016

0.525

0.600

-0.028

-0.972

0.332

0.028

0.938

0.350

28

Table 11 Regression Results: Demographic Independent Variables Independent Variables age 23 - 29

Dependent Variable Overall Score Credit Vehicle Ins. Banking 0.0073 0.0281 0.0378 -0.1419** (0.46) (1.08) (1.13) (-4.33) age 30 - 39 0.0776** 0.0193 0.0408 0.1644** (3.31) (0.50) (0.76) (2.54) age 40 and up -0.0108 0.0157 -0.0527 0.0735 (-0.28) (0.51) (-0.93) (1.01) female -0.0115 -0.0084 0.0675** -0.0831** (-1.05) (-0.55) (2.16) (-2.76) African American -0.0975** -0.0895** 0.0118 -0.1624** (-2.92) (-2.82) (0.15) (-2.26) Hispanic -0.1421** -0.0342 -0.1368 -0.0284 (-3.93) (-0.61) (-1.73) (-0.24) other -0.0051 -0.0933 -0.0683 0.1099 (-0.12) (-1.01) (-0.33) (1.41) not married 0.1073** 0.0329 0.0792 0.1101** (5.49) (1.27) (1.72) (2.51) divorced 0.1589** 0.0727** 0.2305** 0.1433** (6.46) (2.16) (4.27) (2.42) one child 0.1186** 0.0848** 0.4414** 0.1398 (5.93) (2.63) (7.49) (1.86) two children 0.0754** 0.0551 0.0236 0.0626 (2.35) (1.1) (0.34) (0.91) three children -0.0883 -0.1779** 0.1789 -0.6583** (-1.91) (-3.24) (1.59) (-4.78) education 0.0183 0.0323 -0.0190 -0.0486 (0.60) (0.59) (-0.29) (-0.60) health professions 0.0791** 0.1167** 0.1051** -0.0127 (4.09) (4.32) (2.68) (-0.30) engineering 0.0911** 0.1731** -0.2925** 0.1065 (4.04) (5.56) (-4.39) (1.53) liberal arts 0.0746** 0.0344 0.0887 0.0514 (3.18) (1.29) (1.13) (0.97) undecided 0.0280 -0.0349 0.0365 -0.0209 (1.75) (-1.62) (0.83) (-0.56) Note: Robust standard errors were used. t-statistics in parenthesis. ** significant at 0.05 or better

Life ins. 0.0170 (0.48) 0.0814 (1.79) -0.0397 (-0.84) 0.0005 (0.03) -0.0689 (-1.80) -0.4419** (-7.67) 0.0452 (0.40) 0.1553** (4.41) 0.1883** (4.08) -0.1023** (-2.05) 0.1974** (3.23) -0.0897 (-1.25) 0.1282** (3.01) 0.0617 (1.92) 0.1216** (2.76) 0.1426** (2.98) 0.0837** (2.78)

Housing 0.0580 (1.70) 0.1563 (1.79) -0.0569 (-0.33) -0.0502 (-1.46) -0.2178** (3.00) 0.0869 (0.87) 0.0487 (0.36) 0.2100** (3.79) 0.2449** (3.59) 0.2950** (3.44) -0.0563 (-0.43) 0.4263** (3.00) -0.1299** (-2.58) 0.0918 (1.57) 0.2069** (2.79) 0.0415 (0.63) 0.1043** (2.30)

29

Table 12 Regression Results: Financial Experience Independent Variables Independent Variables one credit card two cards three cards four cards five cards six cards has a vehicle parent pays vehicle ins spouse pays vehicle ins relative pays vehicle ins has checking account has savings account has home/rental ins live with parents single apartment shared apartment Greek System house dormitory constant

Dependent Variable Vehicle Ins. Banking 0.0367 -0.0299 (1.06) (-0.85) -0.0612 -0.0699 (-1.77) (-1.66) -0.0538 0.2042** (-0.91) (4.40) -0.1431 0.0953 (-1.15) (1.74) 0.1211 0.0961 (1.41) (0.97) -0.1858 0.2998** (-1.96) (2.93) -0.0616 0.0294 (-0.76) (0.34)

Overall Score -0.0178 (-1.35) -0.0509** (-3.19) 0.0178 (0.76) 0.0069 (0.23) 0.0087 (0.34) 0.0026 (0.07) 0.0039 (0.13)

Credit 0.0007 (0.04) -0.0034 (-0.15) -0.0150 (-0.37) 0.0149 (0.31) -0.0026 (-0.06) 0.1259** (2.70) 0.0288 (0.60)

Life ins. -0.0291 (-1.02) -0.0349 (-1.38) -0.0715 (-1.37) -0.0118 (-0.17) -0.0750 (-1.59) 0.1287** (2.01) -0.0874 (-1.63)

Housing -0.0806** (-2.05) -0.1649** (-3.44) 0.1581** (3.15) 0.0871 (0.85) 0.0032 (0.05) -0.6463** (-5.99) 0.1683 (1.93)

-0.0196 (-1.56)

0.0053 (0.29)

-0.0176 (-0.55)

0.0336 (0.92)

-0.0498** (-2.13)

-0.0727 (-1.74)

-0.0075 (-0.25)

0.0171 (0.38)

0.2417** (3.35)

-0.0282 (-0.33)

-0.0462 (-0.79)

-0.2161 (-2.28)

-0.1014 (-1.98) -0.120 (-0.65) -0.0681** (-5.84) -0.0223 (-1.57 0.0171 (0.77) (0.0235 (0.95) 0.0313 (1.40) 0.0956** (2.99) 0.0227 (0.98) 0.4757** (12.09)

0.0175 (0.45) -0.0352 (-1.19) -0.0409** (-2.55) 0.0079 (0.36) -0.0501 (-1.67) -0.0295 (-0.95) -0.0122 (-0.42) 0.0659 (1.00) 0.0456 (1.26) 0.5004** (7.96)

-0.2472** (-2.24) 0.0663 (0.92) -0.1865** (-6.40) -0.1064** (-2.31) 0.0155 (0.25) 0.0097 (0.15) 0.0239 (0.42) 0.1359 (1.70) 0.0320 (0.49) 0.4671** (3.52)

0.0156 (0.24) 0.1552** (3.53) -0.0045 (-0.05) -0.0069 (-0.19) 0.1369** (2.59) 0.1192** (2.17) 0.0894 (1.67) -0.0428 (-0.39) -0.0221 (-0.40) 0.2385** (2.44)

-0.1163** (-2.35) -0.1674** (-2.83) -0.0576** (-2.61) -0.0201 (-0.62) -0.0432 (-0.97) -0.0280 (-0.65) 0.0167 (0.36) 0.1178 (1.77) -0.0205 (-0.50) 0.6503** (7.82)

-0.3207** (-3.81) 0.1072** (2.35) -0.1013** (-3.54) -0.0288 (-0.84) 0.1778** (2.43) 0.1685** (2.18) 0.1162 (1.66) 0.2211** (2.39) 0.09165 (1.23) 0.3147** (2.93)

234 0.5008

234 0.6028

234 0.4953

n 234 234 234 Rsquared 0.6543 0.5123 0.5341 Note: Robust standard errors were used. t-statistics in parenthesis. ** significant at 0.05 or better

30