Is ignorance bliss? Consumer accuracy in judgments about credit ratings.
There is considerable evidence to suggest that in general, consumers are unaware or misinformed about basic financial and economic principles--information they need in order to make important decisions such as buying a home and planning for retirement (Chen and Volpe 1998; Hogarth and Hilgert 2002; Lee and Hogarth 1999; Mandell 2006). This issue has gained more attention recently as a result of concerns about borrowers in the subprime market obtaining mortgages that they cannot afford. There is little known about the specific nature of this lack of awareness and misinformation.
Credit scores are designed to measure credit risk at a particular point in time and are based on models that use information in consumer credit reports maintained at the credit reporting agencies to predict future payment behavior (Fair Isaacs 16). Lenders use credit scores to help them make lending decisions, although each lender may differ in terms of the level of risk it finds acceptable, and this may even vary for different types of loans. Higher scores indicate lower credit risks or higher credit quality, although no single cutoff score is used by all lenders.
Previous research on biases in judgment and decision making has shown that individuals tend to display overconfidence about their knowledge and ability (Kahneman and Tversky 1996; Lichtenstein and Fischhoff 1977). Due to overconfidence, people often believe that they know more than they actually do, and this can have negative consequences. For example, previous research has shown that overconfidence affects decision making because it causes individuals to overweight their own judgments relative to other decision inputs. The purpose of this study was to examine overconfidence in consumers' self-assessments of their credit rating. Our research questions ask: How likely are consumers to overestimate their credit ratings? How do consumers who overestimate their credit ratings differ from those who have a more-accurate estimate of their credit quality? Is there a relationship between overestimating one's credit rating and financial behaviors such as budgeting, saving, or investing?
Financial literacy has grown in importance to researchers and policy makers in recent years. Much of the policy debate has centered on educating consumers and encouraging legislators to include financial literacy in public school curriculums. Research in this area has focused on measuring the extent to which consumers lack financial knowledge and the resulting consequences (see Braunstein and Welch 2002, for a review). For example, in a nationwide survey of twelfth graders, Mandell (2006) found an average score of 52.4% correct on a test of personal finance basics. In another example, Lee and Hogarth (1999) found that approximately 40% of mortgage borrowers did not understand the interest rates associated with their loans.
Several studies have also linked consumer financial knowledge with responsible financial behavior. For example, Chang and Hanna (1992) found that increased levels of financial information resulted in more-efficient decisions. Hogarth and Hilgert (2002) and Hilgert, Hogarth, and Beverly (2003) found that consumers who are financially knowledgeable are more likely to behave in financially responsible ways. Similarly, Perry and Morris (2005) found that consumers with higher levels of financial knowledge were more likely to budget, save, and plan for the future. The present study examines the relationship between the consumer knowledge of financial and credit principles and the tendency to overestimate one's credit rating.
Previous research has examined self-assessed credit ratings (Ards, Ha, and Myers 2006; Ards and Myers 2001; Betsey 2006; Courchane, Galley, and Zorn 2008). According to Ards and Myers (2001) and Betsey (2006), African American consumers disproportionately underestimate their credit ratings.
Ards, Ha, and Myers (2006) link the tendency to misperceive credit ratings with borrowers' prior experience with loan denials. Courchane, Gailey, and Zorn (2008) examine the relationship between self-assessed credit ratings, actual credit ratings, and loan terms. While these authors found that consumers tend to overestimate their credit rating, this had no significant impact on the interest rates they ultimately paid for mortgage loans. This study builds on these prior studies by examining the relationship between credit rating misperceptions, overall financial knowledge, and financial behaviors.
This study draws on previous research on overconfidence in decision making. Kahneman and Tversky (1996) conceptualize overconfidence as a judgmental or perceptual bias that may result in errors. This bias leads individuals to overestimate their own knowledge or ability relative to objective assessments of knowledge or ability. Overconfidence is particularly prevalent when tasks are difficult or nearly impossible (Fischhoff, Slovic, and Lichtenstein 1977; Kahneman and Tversky 1996), and overconfidence tends to be less common when tasks are easier. Lichtenstein and Fischhoff (1977) found that the most-knowledgeable subjects were actually underconfident. These findings persist whether difficulty is defined in terms of previous subjects' performance or in terms of independent assessments of correctness (Lichtenstein and Fischhoff 1977; Lichtenstein, Fischhoff, and Phillips 1982).
It is possible that the overconfidence effect results from an individual's inability to estimate the degree of difficulty associated with a task. In the context of a consumer's credit rating, overconfidence might be due to consumer's lack of knowledge about how difficult it is to obtain and maintain good credit. Several researchers have attempted to uncover its basic psychological causes and ways to improve the calibration of judgments made by decision makers. For example, Ferrell and McGoey (1980) suggest that individuals have feelings about outcomes that must be translated into judgments that are often numerical or otherwise highly specific and are often unable to do this without feedback. Similarly, Griffin and Tversky (1992) suggest that judgments are often based on beliefs about individual events rather than base rates or probabilities based on outside information. This implies in our case that consumers may have feelings or beliefs about their credit quality and what it should be but are unable to map these feelings into the same variables used by credit rating agencies. Thus, these beliefs result in overconfidence, which in turn leads them to overestimate their creditworthiness, or credit quality.
Previous research in the overconfidence literature has also examined the effects of overconfidence judgments. For example, Biais et al. (2005) found that overconfidence leads to the "winners curse" and earned fewer profits in financial market trading. Camerer and Lovallo (1999) found similar evidence that inflated self-assessments have a negative effect on performance. They found that overestimating one's chances of success on a new venture led to excessive market entry and financial losses. In the present study, we examine the relationship between overestimating one's credit ratings and financial behaviors, such as budgeting, saving, and investing.
In previous studies, consumer financial knowledge has been found to be associated positively with their income and education level (Hogarth and Hilgert 2002; Kinsey and McAlister 1981; Mandell 2006; Perry and Morris 2005). Previous research has also examined the relationship between race and ethnicity, financial literacy, and financial outcomes (Ards and Myers 2001; Betsey 2006; Courchane, Gailey, and Zorn 2008; Perry and Morris 2005). For example, according to Perry and Morris (2005), the effects of knowledge on financial behavior can vary by race or ethnicity. The present study tests whether these demographic characteristics are related to a consumer's tendency to overestimate his/her credit rating.
Overconfidence has been described as a judgment bias because individuals are more likely to overestimate than to underestimate their knowledge or performance abilities, particularly with difficult tasks (Schraw and Roedel 1994). Thus, we expect that due to the ambiguous and perhaps complicated nature of credit ratings, consumers will be poorly calibrated and exhibit an overconfidence bias. We propose that consumer decision makers are biased toward overestimation when making judgments about credit rating. In addition, we propose that consumers who underestimate their credit ratings differ from other consumers in terms of knowledge, knowledge acquisition, and demographic characteristics. We also propose that overestimating one's credit rating may negatively affect responsible financial decision making.
H1: Consumers will be more likely to overestimate their credit quality than to underestimate their credit quality.
In general, there is little consistent evidence in the calibration literature to support individual differences in overconfidence (Lichtenstein and Fischhoff 1977; Lichtenstein, Fischhoff, and Phillips 1982). For example, there have been some findings that suggest that experts can be just as overconfident, and sometimes more overconfident than novices (Lichtenstein, Fischhoff, and Phillips 1982). However, there is evidence in the literature on judgment biases that calibration improves with increased knowledge, and increased knowledge in general is associated with less overconfidence and, in the extreme, underconfidence (Lichtenstein and Fischhoff 1977). In this study, we test whether consumers with more knowledge, expertise or experience in financial matters are less likely to overestimate their credit ratings.
H2: There is negative relationship between overestimating one's credit rating and financial knowledge.
This model also proposes that this overestimation is a function of an individual's prior learning experiences. Perry and Ards (2002) found that learning from high school, college, or financial training courses resulted in higher levels of financial knowledge. These authors also found positive effects of learning from experience on financial knowledge. However, learning through word-of-mouth sources had no effect on financial knowledge.
H3: There is a negative relationship between overestimating one's credit rating and formal financial education.
We also examine the effects of overestimating one's credit rating on behavioral outcomes. Previous research has found that overconfidence has a negative effect on decision quality (Biais et al. 2005; Camerer and Lovallo 1999). Courchane, Gailey, and Zorn (2008) find some evidence that inaccurate self-assessments of credit quality lead to increased probabilities of being denied credit, experiencing a "bad" financial event, or having a higher annual percentage rate on a mortgage. We propose that individuals who have mistaken beliefs about their credit ratings are less likely to budget, save, or invest--behaviors that may exacerbate credit problems.
H4: There is a negative relationship between overestimating one's credit rating and budgeting on a regular basis.
H5: There is a negative relationship between overestimating one's credit rating and saving/investing on a regular basis.
The following analysis is based on data from the Freddie Mac Consumer Credit Survey collected in 2000. This survey was designed to collect data on attitudes, behaviors, knowledge, and experiences with credit and financial management. The data set includes detailed information on individual and household characteristics of consumers randomly drawn from mail panels from two independent research firms. Approximately twenty-three thousand surveys resulted in an initial response rate of 52%. Addresses from respondents' returned surveys were matched with FICO credit files, and then, each respondent was assigned to a FICO score category. These categories correspond to those that are used for risk-based pricing in the lending industry (Fair Isaacs Corporation 2003; Straka 2000). The following analysis is based on 9,471 observations that could be matched to nonmissing credit score data.
The main items used in the analysis are listed in Appendix 1. Self-assessed credit rating, compared to FICO credit score category to create the dependent variable, was measured on a 5-point scale ranging from "very bad" to "very good."
Independent variables include measures of general financial knowledge, source of financial knowledge, financial management behaviors (i.e., budgeting and saving/investing), and demographic characteristics. Financial knowledge was measured using the number correct on a 15-item financial knowledge quiz. This quiz has been used in previous studies on financial literacy (e.g., Ards and Myers 2001; Betsey 2006; Perry and Morris 2005) and measures knowledge of basic economic and financial concepts as well as consumer credit principles.
Source of financial knowledge, that is, financial learning variables are self-reported items that measure, on a 5-point Likert scale ranging from "nothing" to "a great deal," how much the respondent learned about financial matters from various sources. These sources include: formal training, that is, high school or college courses; training courses outside school; word of mouth, that is, friends, peers, or spouses; parents; or "hard knocks," that is, difficult financial experiences. These were converted to dummy indicators for respondents who reported learning "some," "a fair amount," or "a lot" from these sources.
Financial behavior is captured by two variables. These are also self-reported Likert-type items that ask on a 1-5 scale how often respondents follow a budget and save or invest out of each paycheck. The responses ranged from "never" to "always." The actual interval scale values for these variables were included in the analysis.
Income is included as a series of dummy variables--one for annual incomes under $35K, another for incomes between $36K and $55K, and the omitted category that includes incomes over $55K. Level of education is also a dummy variable coded as 1 for those who attended at least some college and 0 otherwise. Race and ethnicity, gender, and homeownership are captured by dummy variables. Age is included in the model as an interval scale value.
Demographic characteristics of the Consumer Credit Survey sample are presented in Table 1, including the distribution of respondents by gender, income, level of education, housing tenure, and race and ethnicity. Approximately 56 percent of respondents were female, and the annual median family income was between $35,000 and $44,000. Those in the highest-income category, that is, income greater than $100,000, are underrepresented since the sample was originally designed to oversample individuals with lower and moderate incomes. Twenty-seven percent of respondents were college educated, and 56 percent of respondents were homeowners.
Table 2 reports correlations and descriptive statistics for the major independent variables. Self-assessed credit ratings are significantly and positively correlated with income and saving/investing behavior. General financial knowledge, income, and level of education are also significantly correlated. These relationships are tested formally in a logistic regression analysis.
The dependent variable in this model is overestimation of one's credit rating. In our model, overestimation occurs when the respondent's self-assessed credit rating is higher than the individuals' actual credit rating. Respondents who meet this overestimation criterion are indicated by a binary variable and include those who rated their current credit record as "good" or "very good" but had FICO scores below 620. The group of respondents who underestimate their credit rating includes those who rate themselves as having "very bad" or "bad" credit but have FICO scores over 620. "Accurate" consumers are those who neither underestimate nor overestimate their credit ratings. This approach to measuring overconfidence judgments has been applied in previous studies. For example, Oskamp (1962) computed an expected percentage of correct decisions in order to compare the level of accuracy with the level of confidence. Similarly, Schraw and Roedel (1994) measure the overconfidence bias as the difference between average performance and average estimated performance for a set of questions.
We find that about 63 percent of respondents made accurate self-assessments of their credit rating. Figure 1 shows the distribution of self-assessed credit ratings for categories of actual FICO credit scores. Consumers with high FICO scores, that is, in categories over 620, are more likely than consumers with low FICO scores to report that they have good or very good credit ratings. These respondents would be considered "accurate" in their self-assessed credit ratings. At the same time, consumers with low FICO scores are more likely than consumers with high FICO scores to report that their credit rating is bad or very bad. However, the remaining 36.9 percent of respondents either overestimated or underestimated their credit quality. Approximately 32.3 percent of consumers overestimate their credit rating, while only 4.6 percent of consumers underestimate their credit rating. These results are consistent with previous research on the overconfidence bias (e.g., Lichtenstein, Fischhoff, and Phillips 1982) and support H1.
A binary logistic regression model, appropriate for categorical dependent variables, was estimated where a dummy variable for overestimating credit quality was regressed on the following: the respondent's score on a financial knowledge quiz, financial learning source, the tendency to keep a budget, and the tendency to save or invest regularly (Allison 1999, p. 6). Independent variables also include the respondent's annual income level, the level of education, gender, race and ethnicity, and homeownership as background variables. Results of this regression are presented in Table 3.
According to the -2 log L and Score criteria for assessing model fit, the overall model including all predictors is significant at the p < .001 level. The pseudo [R.sup.2,] used to measure strength of association in logistic regression, is .1352. It is important to note that although this is a low value, the pseudo [r.sup.2] is not a measure of variance and is not directly analogous to [r.sup.2] used in ordinary least squares regressions. The variables with the strongest effects on [r.sup.2] included: saving/investing, the college dummy variable, the homeowner dummy variable, and the African American dummy variable. The [r.sup.2] for a model including only these variables was 1094. These low [r.sup.2] values suggest that there may be additional important omitted variables.
With respect to H2, those who overestimated their credit rating scored significantly lower on the financial knowledge quiz. In other words, individuals with lower levels of knowledge about financial principles are more likely to fall into the overestimation category. Thus, H2 is supported.
Findings for H3 were mixed. The effects of formal financial training, word-of-mouth information, and learning from parents on overestimation were not significant. However, individuals who reported learning from difficult financial experiences were significantly more likely to overestimate their credit ratings. One possible explanation is that consumers do not necessarily learn from difficult financial experiences, that is, "hard knocks," and that information gained from this source may be misleading.
Both H4 and H5 were supported. Individuals who overestimated their credit quality were also significantly less likely to follow a budget or save] invest out of every paycheck when differences in demographic characteristics and financial knowledge were held constant.
All the demographic variables were statistically significant predictors of overestimating credit rating. Significant and positive coefficients on the income independent variables suggest that individuals with lower incomes are more likely to overestimate their credit ratings. At the same time, college-educated individuals were significantly less likely to overestimate their credit rating. In terms of race and ethnicity, African American and Hispanic/Latino respondents were more likely to overestimate their credit ratings, while Asian American respondents were less likely to do so. In addition, the homeownership, age, and male coefficients were all negative, suggesting that homeowners, older individuals, and males are less likely to overestimate their credit ratings.
Table 3 also reports odds ratios for each independent variable. An odds ratio is the increase or decrease in odds of overestimating one's credit rating that can be attributed to each independent variable. The further the odds ratio is away from 1.0, the greater the impact the variable has on the odds of overestimating one's credit rating. For example, the odds ratio for the homeownership dummy variable is 0.651, which means that the odds of estimating one's credit rating decreases by a factor of 0.651 if one is a homeowner. The odds ratio for having an income under $35,000 a year is 1.683, which suggests that having a low income will increase the odds of overestimation by 68%. These odds ratios can be used to compare the relative impact of the independent variables on overestimating credit ratings. Several of the demographic variables, including income, education, being a member of a minority group, and being a homeowner, had large effects on the dependent variable. However, the effects of financial knowledge and budgeting were relatively small compared to the effects of difficult financial experiences and regular saving behavior.
These findings support previous studies in judgment and decision making that demonstrate that individuals are more likely to be overconfident about their knowledge or abilities than underconfident, especially with difficult or ambiguous tasks. These results indicate that 32 percent of respondents who would be considered high risk by most lenders mistakenly believe that they have average or above-average credit.
These individuals who overestimate their credit ratings have lower incomes, have less formal education, are less likely to own their homes, and are more likely to be African American or Hispanic and female. Taken together, one explanation for these results could be that minority consumers in general have less experience with financial markets, and this in turn affects their propensity to overestimate their credit rating. We also find that formal financial training and learning from others do not affect the propensity to overestimate, although bad financial experiences do. Those who overestimate their credit rating are less knowledgeable in general about basic principles about the economy and financial markets.
In addition, consumers who overestimate their credit ratings are less likely to budget and save/invest. This is consistent with previous research on the effects of overconfidence. For example, Landier and Thesmar (2003) show that firms started by optimistic entrepreneurs (who have a higher tendency to overestimate their firm's chances of success relative to others in the same business category) tend to grow less, die sooner, and be less profitable. In sum, it appears that overestimating credit ratings is partly a function of a lack of financial sophistication. Although it is difficult to draw conclusions about cause and effect relationships using these data, it is conceivable that people who have overestimated their credit rating take less care in managing their finances. It is also possible and consistent with previous research that the reason these individuals have credit blemishes in the first place is a result of not keeping a budget or saving regularly.
There are important limitations that should be taken into account when interpreting these results. The data used for these analyses were collected before the Fair and Accurate Credit Transactions act, a Federal law, was passed in 2003. Since 2005, when Fair and Accurate Credit Transactions fully went into effect, the three major credit reporting agencies are required to provide consumers with a free copy of their own credit report every twelve months. In addition, Fair and Accurate Credit Transactions requires that mortgage lenders provide consumers with a disclosure notice that includes their credit scores, range of scores, credit bureaus, scoring models, and factors affecting their scores (Beales 2004). As a result of this legislation, consumers are now able to access information on how factors such as the timeliness of credit card payments in the recent past affect their credit scores. Thus, the proportion of consumers who inaccurately assess their credit ratings is likely to be smaller than reported in this study.
In addition, because the sample is limited to members of a commercial mail panel, respondent characteristics may not reflect the characteristics of nonrespondents.
This study provides empirical evidence of negative consequences of overconfidence, that is, overestimation in consumer judgments about credit quality. People who do not know their credit rating are more likely to overestimate than to underestimate their credit quality. This tendency toward overestimation may lead consumers to be less cautious in their financial decision making. In addition, this overestimation bias is more likely to affect groups of particular concern to policy makers, including less financially sophisticated, lower-income, African American, and Hispanic consumers.
Given these findings, it is critical for policy makers and consumer advocates to encourage consumers take advantage of the Fair and Accurate Credit Transactions act requirements by obtaining a copy of their credit report and to keep abreast of changes in their credit rating. This will undoubtedly improve the accuracy of consumer's self-assessments of their credit quality. Several studies in the judgment and decision-making literature have found that feedback tends to decrease overconfidence (Arkes et al. 1987; Mahajan 1992). Policy makers and consumer organizations involved in financial literacy education could use public service campaign strategies and other social marketing approaches to promote this behavior.
Another implication of these findings is that financial knowledge is multifaceted. According to these results, general knowledge of financial and economic principles is related to the accuracy of judgments about one's own credit ratings, and both these are linked to financial behaviors such as saving and investing.
Further research should investigate what kinds of financial knowledge have the most impact on financial behavior and outcomes. In addition, future research should examine to what extent consumers understand credit ratings and how knowledge of credit ratings affects financial decisions.
APPENDIX 1 Selected Consumer Credit Survey Variables Subjective Credit Rating How would you rate your current credit record? Very bad  Bad  About average  Good  Very good  Financial Knowledge (Correct answers are marked with an 'x') What will be the impact on the interest rate people pay on a loan if they ...? Rate would No Rate would be higher impact be lower a. always eventually pay off their debts, but are sometimes late on monthly bills [x]   b. get someone else to co-sign the loan with them   [x] c. have never borrowed money before [x]   d. offer the lender some collateral for the loan   [x] What will be the impact on a person's credit rating if they... ? Rate would No Rate would be higher impact be lower a. charge lots of money on several credit cards, and make the minimum payments each month [x]   b. have a good payment record and apply for many new credit cards [x]   c. skip a student loan payment [x]   d. never borrow money or use a credit card for anything [x]   e. miss a couple of loan payments but make them up, plus interest, the next month [x]   f. have a legal judgment over a disputed bill [x]   Do you think that the following statements are generally correct, or not correct? Not Correct correct a. People will be financially better off if the cost of living increases by more than income  [x] b. It is financially worthwhile to borrow money for an investment if the interest rate on the loan is less than the expected return [x]  c. The cheapest way to use a credit card is to pay off the bill in full each month [x]  d. Over the long run, people can expect to earn more money by investing in stocks than by putting month into U.S. savings bonds [x]  e. Investing $1,000 a year for 10 years will earn the same amount of money as investing $2,000 a year for 5 years if the interest rate is the same for both investments  [x] Financial Behaviors How often do you do the following? Never Seldom Sometimes Often Always a. Follow a budget b. Save or invest money out of each paycheck Source: Financial Learning How much have you learned about managing money and using credit from...? Very A Fair Nothing little Some amount A lot a. high school and/or college courses b. training courses and/or seminars taught outside school c. difficult financial experiences / "school of hard knocks" d. parents e. spouse/domestic partner f. friends and peers g. work h. TV or radio
Allison, Paul D. 1999. Logistic Regression Using the SAS System: Theory and Application. Cary, NC: SAS.
Ards, Sheila, Inyuck Ha, and Samuel Myers. 2006. The Effect of Bad Credit on Loan Denial Rates. APPAM Fall Research Conference Proceedings. Madison, WI, November 2.
Ards, Sheila and Samuel Myers. 2001. The Color of Money: Bad Credit, Wealth, and Race. American Behavioral Scientist, 42 (2): 223-239.
Arkes, Hal R., Caryn Christensen, Cheryl Lai, and Catherine Blumer. 1987. Two Methods of Reducing Overconfidence. Organizational Behavior and Human Performance, 39 (1): 133-144.
Beales, J. Howard III 2004. Prepared Statement of the Federal Trade Commission on Identity Theft and Social Security Numbers before the Subcommittee on Social Security of the Committee on Ways and Means. Washington, DC: U.S. House of Representatives, June 15. http://www.ftc.gov/os/testimony/040615idtheftssntest.pdf.
Betsey, Charles. 2006. Believed Truth: How Perceptions May Hinder Equal Access to Credit. APPAM Fall Research Conference Proceedings. Madison, WI, November 2.
Biais, Bruno, Denis Hilton, Karine Mazurier, and Sebastien Pouget. 2005. Judgmental Overconfidence, Self-Monitoring, and Trading Performance in an Experimental Financial Market. The Review of Economic Studies, 72 (251): 287-312.
Braunstein, Sanndra and Carolyn Welch. 2002. Financial Literacy: An Overview of Practice, Research, and Policy. Federal Reserve Bulletin, 88 (11): 445457.
Camerer, Colin and Dan Lovallo. 1999. Overconfidence and Excess Entry: An Experimental Approach. American Economic Review, 89 (1): 306-318.
Chang, Regina Y. and Sherman Hanna. 1992. Consumer Credit Search Behavior. Journal of Consumer Studies and Home Economics, 16 (1): 207-227.
Chen, Haiyang and Ronald P. Volpe. 1998. An Analysis of Personal Financial Literacy among College Students. Financial Services Review, 17 (2): 107-129.
Courchane, Marsha, Adam Galley, and Peter Zorn. 2008. Consumer Credit Literacy: What Price Perception? Journal of Economics and Business, 60 (1-2): 125-138.
Fair Isaacs Corporation. 2003. Understanding Your Credit Score. www.myFICO.com.
Ferrell, William R. and P. J. McGoey. 1980. A Model of Calibration for Subjective Probabilities. Organizational Behavior and Human Performance, 26 (1): 32-53.
Fischhoff, Baruch, Paul Slovic, and Sarah Lichtenstein. (1977). Knowing with Certainty: The Appropriateness of Extreme Confidence. Journal of Experimental Psychology: Human Perception and Performance, 3 (4): 552-564.
Griffin, Dale and Amos Tversky. 1992. The Weighing of Evidence and the Determinants of Confidence. Cognitive Psychology, 24 (3): 411-435.
Hilgert, Marianne A., Jeanne M. Hogarth, and Sondra Beverly. 2003. Household Financial Management: The Connection between Knowledge and Behavior. Federal Reserve Bulletin, 89 (July): 309-322. www. federalreserve.gov/pubs/bulletin/2003/07031ead.pdf.
Hogarth, Jeanne M. and Marianne A. Hilgert. 2002. Financial Knowledge, Experience and Learning Preferences: Preliminary Results from a New Survey on Financial Literacy. Consumer Interest Annual, 48 (1). http://www.consumerinterests.org/files/public/FinancialLiteracy-02.pdf.
Kahneman, Daniel and Amos Tversky. 1996. On the Reality of Cognitive Illusions. Psychological Review, 103 (3): 582-591.
Kinsey, Jean and Ray McAlister. 1981. Consumer Knowledge of the Costs of Open-End Credit. Journal of Consumer Affairs, 15 (2): 249-270.
Landier, A. and Thesmar, D. 2003. Financial Contracting with Optimistic Entrepreneurs: Theory and Evidence. Unpublished Paper, University of Chicago. http://www.chicagogsb.edu/capideas/may04/optimism.html.
Lee, Jinkook and Jeanne M. Hogarth. 1999. The Price of Money: Consumers Understanding of APRs and Contract Interest Rates. Journal of Public Policy and Marketing, 18 (Spring): 66-76.
Lichtenstein, Sarah and Baruch Fischhoff. 1977. Do Those Who Know More Also Know More about How Much They Know? Organizational Behavior and Human Performance, 20 (1): 159-183.
Lichtenstein, Sarah, Baruch Fischhoff, and Lawrence D. Phillips. 1982. Calibration of Probabilities: The State of the Art to 1980 in Judgment under Uncertainty: Heuristics and Biases, edited by Daniel
Kahneman, Paul Slovic, and Amos Tversky (306-334). Cambridge, UK: Cambridge University Press.
Mahajan, Jayashree. 1992. The Overconfidence Effect in Marketing Management Predictions. Journal of Marketing Research, 29 (3): 329-342.
Mandell, Lewis. 2006. Financial Literacy: Are We Improving ? Results of the 2006 National Jump$tart Survey. Washington, DC: Jumpstart Coalition.
Oskamp, Stuart. 1962. The Relationship of Clinical Experience and Training Methods to Several Criteria of Clinical Prediction. Psychological Monographs, 76 (28): 20-31.
Perry, Vanessa Gall and Sheila Ards. 2002. CreditSmart@ Research: How Consumers Learn about Credit and Implications for Credit Education. Social Marketing Quarterly Journal, 8 (3): 3-16.
Perry, Vanessa G. and Marlene D. Morris. 2005. Who Is in Control? The Role of Self-Perception, Knowledge and Income in Explaining Consumer Financial Behavior. Journal of Consumer Affairs, 39 (2): 299-313.
Schraw, Gregory and Teresa DeBacker Roedel. 1994. Test Difficulty and Judgment Bias. Memory and Cognition, 22 (1): 63-59.
Straka, John. 2000. A Shift in the Mortgage Landscape: The 1990s Move to Automated Credit Evaluations. Journal of Housing Research, 11 (2): 207-232.
Vanessa Gail Perry is an assistant professor in the Department of Marketing, George Washington University School of Business, Washington, DC (email@example.com).
The views expressed in this paper are those of the author and do not necessarily reflect those of Freddie Mac, its management, shareholders, or Board of Directors.
TABLE 1 Demographic Characteristics of the Consumer Credit Survey Sample Gender (%) Male 43.7 Female 56.3 Annual family income (%) Under $15,000 9.3 $15,000 to $24,999 15.6 $25,000-$34,999 19.2 $35,000-$44,999 16.4 $45,000-$54,999 15.9 $55,000-$64,999 10.6 $65,000-$74,999 6.8 $75,000-$100,000 4.7 Over $100,000 1.5 100.0 Level of education (%) Some schooling 4.0 Associate's degree 24.2 High school diploma or equivalent 34.5 Some college 10.4 Graduated college 19.0 Graduate or professional degree 7.9 100.0 Tenure (%) Own 56.5 Rent 43.5 Race/ethnicity (%) African American 12.5 Asian 3.6 Hispanic/Latino 2.2 White 71.2 100.0 TABLE 2 Correlations and Descriptive Statistics * Financial Self-Assessed Knowledge Credit Rating Score Income Self-assessed credit rating 1.00 Financial knowledge score .18# 1.00 Income .32# .32# 1.00 Level of education .21# .27# .29# Budget .16# .04# -.02 Save/invest .35# .18# .31# Formal training .06# .02 .00 Word of mouth .07# .04# .11# Hard knocks -.13# .13# .02 Parents .07# .02 .05# Mean 3.32 7.99 4.00 SD 1.25 2.92 2.02 Minimum 1.00 0.00 1.00 Maximum 5.00 15.00 9.00 Level of Save/ Education Budget Invest Formal Self-assessed credit rating Financial knowledge score Income Level of education 1.00 Budget .05# 1.00 Save/invest .23# .23# 1.00 Formal training .06# .08# .08# 1.00 Word of mouth .02 .03# .08# .13# Hard knocks .02 .00 -.02 .10# Parents .05# .06# .07# .15# Mean 3.42 3.26 3.32 0.55 SD 1.35 1.29 1.46 0.50 Minimum 1.00 1.00 1.00 0.00 Maximum 6.00 5.00 5.00 1.00 Word of Hard Mouth Knocks Parents Self-assessed credit rating Financial knowledge score Income Level of education Budget Save/invest Formal training Word of mouth 1.00 Hard knocks .10# 1.00 Parents .24# .06# 1.00 Mean 0.76 0.75 0.70 SD 0.43 0.43 0.46 Minimum 0.00 0.00 0.00 Maximum 1.00 1.00 1.00 * Note: Bold coefficients are significant at p < .01 is indicated with #. TABLE 3 Logistic Regression Results. Dependent Variable: Overestimate Credit Rating (Overestimated Credit Rating: Self-Assessed Credit Rating = Good or Excellent and FICO <620), n = 9,471 Wald Parameter Estimate SE [chi square] Intercept 0.9725 0.2174 20.0183 Homeowner -0.4298 0.0522 67.6952 Age -0.0275 0.00464 35.0398 Male -0.2301 0.0549 17.5859 African American 0.8536 0.0604 199.9759 Hispanic 0.3769 0.0655 33.1309 Asian -0.3444 0.0984 12.2427 Income under $35K 0.5206 0.0883 34.7385 Income under $65K 0.3163 0.0867 13.3076 College -0.5301 0.0549 93.3421 Financial quiz score -0.0562 0.00898 39.1464 Source Formal financial training 0.0148 0.05 0.0878 Word of mouth 0.00904 0.0587 0.0237 "Hard knocks" 0.3205 0.0591 29.3774 Parents -0.043 0.0548 0.6158 Keep a budget? -0.0687 0.0194 12.5907 Save or invest regularly? -0.1899 0.0179 112.8297 Probability Parameter > [chi square] Odds Ratio Intercept <0.0001 Homeowner <0.0001 0.651 Age <0.0001 0.973 Male <0.0001 0.794 African American <0.0001 2.348 Hispanic <0.0001 1.458 Asian .0005 0.709 Income under $35K <0.0001 1.683 Income under $65K .0003 1.372 College <0.0001 0.589 Financial quiz score <0.0001 0.945 Source Formal financial training .767 1.015 Word of mouth .8776 1.009 "Hard knocks" <0.0001 1.378 Parents .4326 0.958 Keep a budget? .0004 0.934 Save or invest regularly? <0.0001 0.827 Model fit statistics: [R.sup.2] = .1332; maximum resealed [R.sup.2] = .1862; likelihood ratio [chi square] = 1,322.9939, 16 dfs, and p < .0001; score [chi square] = 1,232.8282, 16 dfs, and p < .0001: Wald [chi square] = 1,075.2847, 16 dfs, and P < .0001. FIGURE 1 Subjective Credit Ratings Compared to Objective Credit Ratings Very Bad (< 579) 59.5% 9.3% Bad ('580-619) 39.7% 19.9% Average ('620-679) 17.9% 45.6% Good ('680-749) 3.7% 79.2% Very Good (> = 750) 1.0% 92.5% Note: Table made from bar graph.
|Printer friendly Cite/link Email Feedback|
|Author:||Perry, Vanessa Gail|
|Publication:||Journal of Consumer Affairs|
|Date:||Jun 22, 2008|
|Previous Article:||Toward a general model of consumer empowerment and welfare in financial markets with an application to mortgage servicers.|
|Next Article:||Saving for retirement: the effects of fund assortment size and investor knowledge on asset allocation strategies.|
|Sorted: Meet The Vulture..; EAGLE BOSS PREYED ON 'CREDIT RISKS'.|
|RECORD VIEW : Tell it straight.|
|By the numbers. (Briefing).|
|Jill joins smoke alarm campaign.|