JBR-08029; No of Pages Journal of Business Research xxx (2014) xxx–xxx Contents lists available at ScienceDirect Journal of Business Research Credit card behavior, financial styles, and heuristics☆ Hersh Shefrin a,⁎, Christina M Nicols b a b Santa Clara University, Department of Finance, Lucas Hall, 500 El Camino Real, Santa Clara, CA 95053, United States Ketchum Public Relations, 2000 L Street NW, Washington, DC, United States a r t i c l e i n f o Article history: Received September 2013 Received in revised form December 2013 Accepted January 2014 Available online xxxx Keywords: Credit cards Spending Borrowing Financial literacy a b s t r a c t The paper makes four contributions First, the paper provides new data and findings about credit card usage segmentation in respect to spending and borrowing behavior Second, it sets the new findings against the backdrop of the newly emerging literature on financial literacy A great variability occurs in financial literacy across American consumers Third, the paper describes fast and frugal heuristics aimed to help consumers make effective, and in some cases better, budgeting decisions when they use credit cards Fourth, the paper describes the introduction of a new set of online financial tools, offered by a large credit card company, which consumers are now using to make decisions about their spending and borrowing, and links these tools to the heuristics under discussion Fast and frugal heuristics are likely to be especially valuable to consumers with low confidence in their online skills Notably, 25% of credit cardholders report that they have low confidence using online technology to manage their finances, with the corresponding figure being 44% for those most at risk © 2014 Published by Elsevier Inc Introduction By its nature, household decision making is a heuristic enterprise, as most household decision tasks are far too complex to be fully specified, let alone solved through optimization In this paper, the authors discuss how households can use fast and frugal heuristics when using credit cards to engage in spending and borrowing The paper makes four contributions First, it provides new data and findings about credit card usage segmentation in respect to spending and borrowing behavior Second, it sets the new findings against the backdrop of the newly emerging literature on financial literacy Third, it describes fast and frugal heuristics aimed to help consumers make effective, and in some cases better, budgeting decisions when they use credit cards Fourth, it describes the introduction of a new set of online financial tools, offered by a large credit card company, which consumers are now using to make decisions about their spending and borrowing In this regard, it links these tools to the heuristics under discussion As Bendor (2010) points out, Herbert Simon's bounded rationality approach has given rise to several research streams involving heuristics ☆ The authors thank Jay Weiner, Ipsos, for his work in analyzing the data, Gail Hurdis and Paul Hartwick, JPMorgan Chase, for comments, and Edward Schultz, U.S Bank, for sharing his insights with us about financial literacy The authors thank Nathan Berg, our referee, and Shabnam Mousavi, the journal guest editor for this special issue, for their helpful comments on the paper ⁎ Corresponding author E-mail addresses: hshefrin@scu.edu (H Shefrin), Christina.Nicols@ketchum.com (C.M Nicols) Examples of the different streams include the fast and frugal heuristic approach described in Gigerenzer, Todd, and the ABC Research Group (1999), the heuristics and biases approach described in Kahneman, Slovic, and Tversky (1982), and self-control heuristics described in Shefrin and Thaler (1988) Heuristics for household consumer decisions are ubiquitous A large literature in marketing documents how households make use of heuristics that feature, for example, private labels and national brands (Putsis & Dhar, 2001), satisficing in respect to the ordering of cues (Levav, Heitmann, Herrmann, & Iyengar, 2010), and self-control through the choice of which cash denominations to carry (Raghubir & Srivastava, 2009) In respect to self-control, the role of heuristics has also been a subject of study in the literature in economics (Shefrin & Thaler, 1988), and more generally in retirement saving behavior (Benartzi & Thaler, 2007) and borrowing behavior (Karlan & Zinman, 2012) In this regard, the economics literature documents the degree to which many U.S households lack financial literacy (Bernheim, 1995, 1998) A study by Hilgert, Hogarth, and Beverly (2003) finds that most Americans not understand the basic financial concepts of stocks, bonds, and mutual funds Moore (2003) finds that people often fail to understand terms and conditions of consumer loans and mortgages Ameriks, Caplin, and Leahy (2003) report that households with superior financial planning and budgeting skills accumulate wealth at higher rates than other households The economics literature is also developing tools for measuring the degree to which consumers express preferences consistent with a neoclassical preference ordering For example, see Choi, Kariv, Müller, and Silverman (2010) These tools are interesting, but focus more on http://dx.doi.org/10.1016/j.jbusres.2014.02.014 0148-2963/© 2014 Published by Elsevier Inc Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx consistency of choice than on the quality of the decisions The main focus of the present paper is on steps consumers can take to improve the quality of their financial decisions In this regard, see the normative discussion in Agnew (2010) Lusardi (2010) summarizes findings from the National Financial Capability Study (NFCS), which was commissioned by the FINRA Investor Executive Foundation, and in which she played a lead role The NFCS comprises three linked studies, one focused on national traits, one focused on comparisons state-by-state, and one focused on military personnel Lusardi (2010) describes the results from the national survey, which pertains to (1) making ends meet; (2) planning ahead; (3) managing financial products; and (4) financial knowledge and decision making In terms of making ends meet, the first study finds that approximately half of the Americans surveyed report difficulty in keeping up with monthly expenses Roughly a quarter report overdrawing their checking accounts, and of these, about three quarters admit to being challenged in making ends meet As for planning ahead, 51% of Americans fail to accumulate precautionary savings Only 42% have attempted to assess their retirement savings needs Only 41% of those with financially dependent children have set aside funds for college education Credit cards are among the financial products addressed within the NFCS Interestingly, 68% of those surveyed reported that they possess credit cards Of these, 51% indicated that in some months they carried a balance and paid interest, 29% indicated that in some months they paid the minimum amount due, and 23% indicated that in some months they incurred a fee for late payment In contrast, 54% report that they always pay their credit card balances in full Notably, 28% of those possessing credit cards appear to be challenged in making ends meet, which contrasts with 49% for the full population When it comes to financial knowledge, Americans delude themselves Although 37% rate their overall knowledge of finance at the high end, with the corresponding percentage for mathematics being 46%, the survey found that people overrate their abilities on both dimensions For those who rated themselves at the highest end in both finance and mathematics, only half could correctly perform two calculations pertaining to interest rates and inflation respectively The financial service landscape is changing in notable ways To begin with, studies such as the NFCS are documenting the spectrum of financial literacy across households In addition, the credit card act passed in 2009, with the acronym CARD (the Credit Card Accountability, Responsibility, and Disclosure Act), introduced a series of major changes, most of which became effective in February 2010 Even before the passage of CARD, financial institutions, such as banks, credit card companies, and mutual funds had begun plans to offer financial tools online This is especially important in the case of banks and credit card companies, as these tools can access household transaction data In this regard, Mui (2010) discusses innovations at credit card firms Chase, Citi, Discover, and American Express These innovations involve options for lowering interest payments Mui notes that innovations at Chase have attracted the most attention because they feature important budgeting tools to help credit cardholders manage their balances over time As part of a consulting team, the authors worked with Chase to develop a system to help credit card users engage in some self-diagnosis, in order to ascertain how best to use the budgeting tools offered by Chase In this paper, the authors describe some of the key findings of the research underlying this effort, and use the data to analyze important heuristic features In light of the NFCS findings, the central questions addressed in the paper ask whether households can make use of fast and frugal heuristics (Gigerenzer and Gaissmaier, 2011; Gigerenzer et al., 1999; Todd and Gigerenzer, 2000), to achieve greater self-awareness about their credit card financial styles, and in consequence to employ financial tools to make better decisions about their spending and borrowing The remainder of the paper is organized as follows: The first section describes a fast and frugal heuristic for identifying credit card financial styles The second section provides an overview of different types of financial styles in practice The third section describes the data and associated key characteristics The fourth section explains how the authors applied cluster analysis to identify financial styles The fifth section contrasts the results from cluster analysis with those from the application of fast and frugal heuristics The sixth section describes the Chase Blueprint program, for which the authors developed their style analysis Notably, this section discusses research findings about consumer reaction to the relationship between style analysis and the program A final section provides conclusions Developing fast and frugal heuristics Consider the objective of helping credit cardholders make better financial decisions As a first step, suppose one seeks to categorize credit cardholders based upon repayment characteristics, general attitude to budgeting, and reliance on budgeting heuristics How might one develop a fast and frugal style heuristic for the categorization task? What might be a reasonable set of cues, alternatives, stopping rules, and decision rules in respect to credit card behavior? To fix ideas, consider a specific example, involving cues for repayment characteristics, general attitude to budgeting, and reliance on budgeting heuristics Cues According to the NFCS findings, the most important characteristic of repayment behavior is the frequency with which a holder pays only the minimum due, pays the balance in full, pays something in between, or is occasionally in default, thereby incurring late fees One cue for repayment behavior is the answer to the following question: When it comes to paying your credit card(s), you tend to pay the entire balance on the card each month, or just the minimum due? Possible answers might be restricted to 1) minimum due; 2) entire balance; or 3) something in between An indicator of attitude toward household budgeting is the degree to which a holder believes that it is important to be in control of his or her finances One cue for importance of control is the answer to the following question: When it comes to managing finances, how important is it to you that managing your finances is completely within your control, even if it involves considerable effort on your part? Possible answers might be 1) unimportant; 2) important; 3) something in between As for budgeting heuristics, Shefrin and Thaler (1988) propose that households base their spending on how they mentally categorize their wealth into particular mental accounts Examples of categories are current income, liquid assets, home equity, and future income In the Shefrin–Thaler framework, households establish a pecking order for funding, with current income at the top of the pecking order and future income at the bottom Specifically, households spend first from current income, and only after this account is fully depleted, reluctantly “break into” the liquid asset account The idea of reluctance is measured as a mental setup cost, so that the household only breaks into the account if the benefit from doing so is large enough to compensate for the setup cost of doing so In a similar vein, Raghubir and Srivastava (2009) provide evidence that people can control impulse purchases by only carrying large denominations of cash, effectively placing large denominations into mental accounts that they are reluctant to invade Raghubir and Srivastava call this tendency the “denomination effect” Notably, credit cards rather than cash, checks, or debit cards are the most preferred mode of payment by households When it comes to selfcontrol heuristics, credit card usage declined from 87% in 2007 to 56% in 2009, with a corresponding rise in the use of debit cards This is consistent with the use of a heuristic to increase self-control in respect to taking on debt At the same time, some households achieve self-control by Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx using multiple credit cards to help themselves budget, and dedicate particular cards for specific purchases Through expenditure limits on these cards, or by not carrying some cards all the time, households can institute mental accounting principles to control their spending One cue for mental accounting-based credit card spending is the answer to the following question: Do you tend to use a single credit card for all of your expenses, or you use different cards for different reasons, such as one card for everyday expenses and another for emergencies only? Possible answers might be (1) single credit card or (2) multiple credit cards Alternatives Consider the following four style alternatives for a fast and frugal categorization heuristic, in which a non-compensatory classification tree is used to assign credit cardholders to one of four styles Low control minimum payers High control minimum payers Full balance paying multiple cardholders Full balance paying single cardholders Below is an intuitive description of these four categories Low control minimum payers are likely to be most at risk when it comes to financial behavior that is lacking in discipline These credit cardholders report that financial control is not important to them, and usually or always make the minimum payment on their monthly credit card balances A lack of financial literacy for credit cardholders in this category makes them highly vulnerable to falling into debt for extended periods of time, and paying high fees and interest High control minimum payers also make the minimum payment on their monthly credit card balances, usually or always However, this group reports that being in control of their finances is important Credit cardholders falling into this group are likely to be more purposeful and disciplined in using credit card debt to make ends meet Full balance paying multiple cardholders routinely pay most if not all of their monthly balances, and in addition use multiple credit cards Credit cardholders in this category are likely to be more sophisticated than the first two groups, and to rely on mental accounting heuristics to achieve financial discipline Full balance paying single card users almost always pay their balance in full, and plausibly not need the complexity of heuristics based on multiple credit cards Credit cardholders in this category might carry a credit card balance only if they use their card for particularly large purchases, especially those that are unexpected This specific heuristic relies on three cues Later in the paper, the authors discuss other cues Decision rule This particular categorization task involves four options and three cues, a situation that lends itself to an elimination heuristic such as QuickEst, or some variation Consider the following possibility Suppose a credit cardholder answers the first question by saying that he or she makes only the minimum payment each month This answer eliminates categories and 4, as these categories feature credit cardholders routinely paying most, if not all, of their monthly balances The second question, relating to the importance of control, is then used to decide whether to assign the holder to either category or category If the holder answers the second question with the answer “important,” then the heuristic assigns him (or her, henceforth him) to category 2, “high control minimum payers.” Otherwise, the heuristic assigns him to category 1, “low control minimum payers.” On the other hand, if the credit cardholder answers the first question by saying that he pays more than the minimum payment, then the heuristic uses the answer to the third question to assign him to either category or category If the credit cardholder reports that he uses multiple credit cards, then the heuristic assigns him to category 3, “full balance paying multiple credit cards.” If the credit cardholder reports that he uses a single credit card, then the heuristic assigns him to “full balance paying single credit card.” This non-compensatory classification tree for typing consumers into one of four styles is similar to the Breiman-classification procedure for possible heart attack patients which Todd and Gigerenzer (2000) use at the outset of their article about how simple heuristics can make people smart General advice heuristic The style outcome produced by the classification tree is intended as a cue for improving household budgeting, pointing the credit cardholder to the identification of a planning tool that is likely to prove most useful Below is an example of advice, depending on the category into which credit cardholders fall Low control minimum payers: use a planning tool to avoid paying interest on everyday purchases, such as groceries and fuel High control minimum payers: use a planning tool to manage the timeframe over which credit card balances will be repaid Full balance paying multiple cardholders: simplify budgeting by using a planning tool to track spending across consumption categories, along with a more efficient use of the number of cards Full balance paying single cardholders: use a planning tool to isolate large credit card purchases, especially those associated with unplanned expenditures The prescriptions that are imbedded in the above heuristic are generic in nature Later in the paper, the authors discuss prescriptions that are more specific Cues and styles There is a rich psychology literature on personality traits: See John, Naumann, and Soto (2008) This literature has inspired the development of instruments designed to identify financial personalities In this section, the authors describe some of these instruments, with the intent to provide some background for the analysis in this paper and to highlight some of the major cues used in these instruments The financial personality quizzes available to credit cardholders are quite varied Examples include VISA UK's “Better Money Skills,” FNBO Direct, First Bank of Omaha's “What's Your Savings Style?” Jordan Goodman's “Master Your Money Type,” and Louisiana State University's “What Color Is Your Money?” A typical test is “Better Money Skills” offered by VISA UK For illustrative purposes, the authors discuss some of its key features Its stated purpose is to allow credit cardholders to test themselves in order to identify their financial personalities They so by taking an online test in which they choose answers to eleven online questions that closely reflect their attitudes and experiences with money Three of the test questions are: When I get my credit card statement, I normally repay: A The minimum — and sometimes that comes from another credit card B As much of the outstanding amount as possible C The entire balance My monthly budget is: A An ideal rather than a reality — I never have enough money to go round B Usually OK but I sometimes overspend C Very occasionally a struggle but I generally stick to it Savings are: A What savings? I spend every penny that comes in B There to pay for important things, such as a deposit for a home of our own or a new kitchen C Part of my monthly budget plan Notice that all three questions offer three choices of answers, labeled A, B, and C Indeed, answers to all eleven questions in “Better Money Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx Skills” are structured in this manner, and the preponderance of A, B, or C responses is used to assign the test taker to one of three financial personality profiles, designated A, B, or C The personality profiles are as follows: A Money is for enjoying, as far as you're concerned You may consider luxuries such as new computers, cars or designer clothes as modern essentials, which can make it hard to cut back Many of you will have a great time, though — until your overspending catches up with you You need to remember that credit isn't free cash and that all your bills ultimately have to be paid You might want to avoid temptation, by saving a small but regular amount to pay off your debts — and learning to say no B Money is for sharing with others and for living a comfortable life You're quite well organized financially but are likely to overspend when you're stressed or unhappy — and you can excuse your extravagance if it's designed to make other people happy You might want to start saving little and often so you can afford to indulge those you love, drawing up a budget that prioritizes essentials over treats — and learn to say no more often C Money is for security, which is fine when you have enough of it to feel safe When you're short, though, you can feel panic-stricken You're more than capable of sticking to a sensible budget, making repayments on time and saving for the future You might want to start paying off debts before you add to savings, asking for advice and help if you're in trouble — and learn to manage your level of debt sensibly These three personality profiles are informative, and tied to general advice However, they are also vague in so far as serving as cues for specific actions associated with spending and borrowing “Better money skills” also offers a test taker the opportunity to assess his or her degree of their financial literacy After receiving a profile, the test taker is then given a profile and additional information to help educate him or her about finances For present purposes, the key issue concerns the nature of the other nine questions, in so far as cues are concerned These questions relate to issues such as: being in control of household finances, what the test taker would with a sizeable windfall, how the test taker would handle an unexpected bill, how carefully the test taker chooses new debt, and reasons for discomfort when managing spending and borrowing The other tests mentioned above offer similar types of questions, with similarly structured financial profiles as outputs There are also tests that focus on other financial dimensions, and therefore provide other types of questions/cues MoneySense offers two such quizzes The first is the MoneySense Self-Analysis Quiz which addresses issues related both to saving and investing, but not to spending per se The questions in this test pertain to having rainy day accounts, adequate life insurance, diversified investments, a current will, etc The quiz consists of ten questions, with answers on a 0-to-5 point scale The total number of points is used to assign test takers to one of categories: Poor, Fair, Good, and Excellent The second is what MoneySense calls its Personality Quiz, and it focuses exclusively on investment This quiz was developed by Statman and Wood (2004) The quiz consists of sixteen questions, with four choices for each question, which then map into the following four personality profiles: artisans (good instincts will prevail), idealists (money just isn't the top priority), guardians (discipline is the key to security), and rationals (cool reason conquers all) Data: attitudes, behavior, efficacy, and credit cards The data for this study come from two surveys respectively conducted in May and June of 2009 Both surveys focus on eliciting characteristics and behavior patterns of credit cardholders that are associated with specificity of financial goals, spontaneity of purchasing behavior, tendency to procrastinate, perceived control, importance of control, confidence in managing household finances, confidence in using online technology, monthly payment behavior, use of cards for large expenses, number of credit cards employed, and general decision style The May 2009 survey, conducted by Opinion Research Corporation, provides information about consumers' general propensities along with associated demographics The sample for this survey consists of 1047 U.S adults, comprising 501 men and 546 women 18 years of age and older Opinion Research Corporation conducts the online omnibus study twice a week among a demographically representative U.S sample of 1000 adults, aged 18 years and above The June 2009 survey was conducted among a sample of 4026 U.S adults, comprising 2087 men and 1939 women 18 years of age and older Ipsos, a market research firm, conducted the online survey among a demographically representative U.S sample of adults on June 11–15, 2009 With a sample of 4026, one can say with 95% certainty that the overall results are within ± 1.5% of what they would have been had the entire population of adults in the country been surveyed The margin of error for specific demographic segments will be lower Also, Ipsos collects comparable information on demographics compared to Opinion Research Corporation, but they display the information in their banners differently – for example, Ipsos offers fewer breaks in age groups – 18–34, 35–54, and 55 and over instead of the more narrow age segments done by ORC For this reason, the results reported pertain to the May survey, unless otherwise indicated The present section describes univariate analysis involving the demographics based on the May survey, organized into four subsections In most respects, the results from the May survey and the June survey are mutually consistent Rather than aggregate the two datasets, the authors report any differences in results as ranges, with the first entry referring to the May survey and the second to the June survey Attitudes to budgets and planning Self-control is a determining factor in consumers' planning skills Self-control was assessed by posing the following question: When you have to an unpleasant task, you tend to it right away or are you more likely to put it off? About half (49–45%) answered that they would the task right away, 30–37% responded that they would put it off, and the other responses were neutral Seniors are the most likely to the task right away, and younger consumers are more likely than others to put off the task In this regard, 60% of those over age 65 indicate that they would the task right away, in contrast to 42% between the ages of 18 and 24 (To simplify exposition, the authors provide the range, but omit values for intermediate values for age, education, etc.) For those putting off the task, the corresponding responses were 23% for those over age 65 and 38% for those between the ages of 18 and 24 When asked whether they tend to set specific financial goals or more general financial goals, 40–50% of consumers report that they set general goals, whereas 31–24% report setting specific goals Middle-income and affluent consumers are more likely than lower, lower-middle and upper-middle income consumers to set specific financial goals Interestingly, younger households and college graduates are more likely to set specific financial goals than others Caucasian consumers are more likely than other groups to set general financial goals When asked how important it is that managing their finances be completely within their control, even if it involves considerable effort on their part, 81–74% respond that it is important The older is a consumer, the more likely he or she is to say that it is very important to have complete control Hispanic consumers are more likely than others to say that it is very important to have complete control In this regard, 89% of those aged 65 and over so respond, whereas 75% of those between the ages of 18 and 24 respond in this way As for ethnicity, 87% of Hispanic consumers respond in this fashion, compared to 82% of Caucasians and 73% of African Americans In the June survey, those with higher incomes, college graduates, and women are more prone to respond by saying that control is very important Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx When asked about whether they rely on instinct and intuition, facts and logic, or concern for others in making financial decisions, the respective percentages from the June data are 18, 72, and 10 General budgeting behavior When asked whether they tend to stay within a set monthly budget or to have no set budget, 55% say that they stay within a set budget, 24% say that they have no set budget, and the remaining responses are neutral In terms of demographics, 67% of Hispanic consumers report staying within a set budget, compared to 54% of Caucasian consumers and 56% of African Americans When asked about being more of a deliberate spender, usually planning their purchases, or a spontaneous spender, usually spending if they happen to see something they like, 51–49% respond that they are deliberate spenders Those responding that they are spontaneous spenders comprise 24–30% of consumers In this regard, seniors, who are typically on fixed incomes, are the most deliberate spenders Similarly, college graduates are more likely than others to be deliberate spenders Younger consumers are more likely than others to be spontaneous spenders, as are Hispanic consumers In this regard, 62% of those 65 and older are deliberate spenders, in contrast to 47% of those between the ages of 18 and 24 In respect to level of education, 57% of college graduates report being deliberate spenders, compared to 47% for high school incompletes, 43% for of high school graduates, and 52% for college incompletes In respect to ethnicity, 35% of Hispanic consumers report being spontaneous spenders, compared to 24% of Caucasian consumers and 14% of African American consumers Perceptions about budgeting efficacy When asked how confident they are that they manage their money well, 66–64% respond that they have high confidence and 15–19% indicate that they have low confidence Seniors and affluent consumers are the most likely to indicate that they have high confidence (The June survey also identifies college graduates as judging themselves to be in control of their finances.) In this regard, 82% of those aged 65 and over have high confidence, compared to 60% of those between the ages of 18 and 24 When asked how confident are consumers in using online technology to manage their finances, 56–52% express high confidence, whereas 19–25% express low confidence Not surprisingly, younger, higher income, college graduates are more confident in using online technology In this regard, 69% of those between the ages of 18 and 24 express high confidence, whereas only 46% of those aged 65 and over The results for income and education are similar When asked whether they are surprised by the balance on their monthly credit card statements, or the balance is as expected, 70% respond that the balance is as expected, and 8% respond that it is more of a surprise The older a consumer is, the more likely he or she is to say that the statement is as expected College graduates are more likely than others to say that the statement is as expected Of those aged 65 and older, 87% respond by saying that the balance is as expected, whereas for those aged 18 to 24, the corresponding response rate is 56% For education, 76% of college graduates respond by saying that the balance is as expected, compared to 50% of high school incompletes Credit card issues When asked about making monthly payments on their credit card, in so far as paying the entire balance each month, just the minimum due, or something in between, 52–53% report that they pay the entire balance, 25% report that they pay just the minimum, and the rest answer in between Seniors are the most likely to pay the entire balance each month: 65% of those aged 65 and over pay the full balance, compared to 50% of those aged 18 to 24 Middle-income and affluent consumers are more likely than lower, lower-middle and uppermiddle income consumers to pay the entire balance each month College graduates are more likely than others to pay the entire balance each month The proportion of our sample that routinely or always pays the full balance is 56%, which conforms to larger databases The proportion of our sample that reports always paying the minimum balance is 12.4%, which is significantly higher than the corresponding figures in larger databases, which are less than 5% This difference might be due to the timing of our sample, which coincides with the recession beginning in December 2007 and the associated high rate of unemployment When asked whether they tend to use their credit card(s) only for emergencies and/or “big ticket” purchases, or for everyday purchases as well, 39–37% report that they use their card(s) only for emergencies or big ticket purchases, and 32–39% report that they use their card(s) for everyday purchases as well as emergencies and big ticket purchases The higher a consumer's income, the more likely he or she is to use his or her card(s) for everyday purchases as well as emergencies and big ticket purchases The range is 42% for groups whose annual incomes exceed $75,000 and 24% for those whose incomes lie below $25,000 Interestingly, college graduates are more likely than others to use their card(s) for everyday purchases as well as emergencies and big ticket purchases The range is 43% for college graduates to 9% for high school incompletes In the June survey, the authors also find that older consumers are more likely than younger consumers to use their card(s) for everyday purchases as well as emergencies and big ticket purchases Specifically, 47% of those aged 55 and over report using their cards for everyday purchases, compared to 36% of those aged 35–54 and 34% of those aged 18–34 When asked whether they tend to use a single credit card for all of their expenses or different cards for different reasons, 45–49% replied single credit card, and 20–30% responded that they use different cards for different reasons The higher the income a consumer has, the more likely he or she is to use different cards for different reasons In this regard, 33% of consumers whose incomes exceed $75,000 use multiple cards, in contrast to 22% of consumers with incomes below $25,000 College graduates are more likely than others to use different cards for different reasons In this regard, 40% of college graduates use different credit cards, compared to 17% of high school incompletes In the June survey, the authors also find that older consumers are more likely than younger consumers to use different cards for different reasons Specifically, 36% of those 55 years old and over report using multiple credit cards, compared to 32% of those between 35 and 54 and 22% of those between 18 and 34 Cluster analysis Cluster analysis is a statistical technique for classifying objects into mutually exclusive groups, based on combinations of interval variables Each object corresponds to an observation with a set of values X1, X2, … Xn Cluster analysis is used to identify a system of organizing objects, so that objects in the same group share properties in common This section discusses the application of cluster analysis to identifying financial styles associated with credit card behaviors for spending and borrowing Since the number of groups is known a priori, “k-means cluster analysis” can be used With this method, objects, in this case credit cardholders, are assigned to a given group at the first step, based on some initial criterion Then means for each group are calculated, where means are based on the values X1, X2, …, Xn At the next step, the objects are reassigned (into groups), so that objects are assigned to groups based on the similarity of the object to the current mean of that group At the end of this step, the means of the groups are recalculated This process continues recursively until no objects change groups Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx The June sample features properties that are similar to the general population of credit cardholders, in terms of income, age, education, and ethnicity The sample size was 4026 The values X1, X2, … Xn are responses on an interval scale of to to the ten questions described in the previous section Cluster analysis was used in conjunction with the June data to identify four specific groups, roughly along the lines of the categories described in the section Developing fast and frugal heuristics The main findings were roughly as follows For the sake of continuity, the authors preserve the same labels as in the section Developing fast and frugal heuristics Low control minimum payers: This group comprises 27% of the sample Only 17% of this group pays the full balance each month The remainder is evenly divided among those who pay only the minimum and those who pay more than the minimum but less than the full balance Only 40% report attaching high importance to being in control of their finances Correspondingly, just over half of the sample, 52%, report having low confidence in their ability to manage their finances Notably, 44% of this group report low confidence in being able to manage their finances using online technology Only 9% of this group report setting specific goals, and 53% regard themselves as spontaneous spenders Three quarters of this group are below 55 in age High control minimum payers: This group comprises 20% of the sample Two thirds (67%) always pay only the minimum amount due, and only 1% pay the full balance each month Interestingly, 91% indicate that they attach high importance to being in control of their finances, 31% report setting specific financial goals, and 72% have high confidence that they manage their money well Two thirds of the group has high confidence in using online technology to manage their finances Interestingly, 45% of this group reports using their credit cards only for emergencies or large purchases Approximately 80% of the group is below 55 in age Full balance paying multiple cardholders: This group comprises 20% of the sample Notably, 85% of this group reports using more than one credit card Almost none pay only the minimum amount due, and 92% pay the full balance each month Interestingly, 85% indicate that they attach high importance to being in control of their finances, 23% report setting specific financial goals, and 88% have high confidence that they manage their money well A little under two thirds of the group (63%) has high confidence in using online technology to manage their finances Interestingly, 59% of this group reports using their credit cards for everyday purchases Approximately 55% of the group is below 55 in age Full balance paying single cardholders: This group comprises 33% of the sample Notably, 99% of this group reports using a single credit card Almost none pay only the minimum amount due, and 90% pay the full balance each month Interestingly, 86% indicate that they attach high importance to being in control of their finances, 32% report setting specific financial goals, and 86% have high confidence that they manage their money well A significant 83% has high confidence in using online technology to manage their finances Interestingly, 41% of this group reports using their credit cards for everyday purchases Approximately 65% of the group is below 55 in age Consider the choice of four groups for the cluster analysis The data naturally decompose into at least four groups In performing cluster analysis with five groups, the authors find there to be no clear difference in primary characteristics between two of the five groups, which leads us to conclude that the choice of four is sufficient for the analysis In addition, having more than four groups requires additional cues, leading to less frugality in the heuristic approach Comparing fast and frugal to cluster analysis Clearly, the addition of questions beyond those described in the section Developing fast and frugal heuristics, along with cluster analysis, provides a much richer description of how credit cardholders vary in terms of financial style Not surprisingly, the features of the four groupings are similar in terms of repayment behavior, control, and attitude, as these are the bases for using the fast and frugal classification heuristic In this respect, cluster analysis provides a natural selection of cues for the four categories identified In this section the authors compare the results of the two classification methods To so, both cluster analysis and the classification heuristic are used to categorize the sample, and compute the incidence in which the two methods produce the same result For the sample of 4026, the heuristic produces the same result as cluster analysis in 66.9% of cases If one were to treat the cluster analysis as providing the right categorization, the heuristic would produce the correct answer for two out of three cardholders This finding is robust The June data also included an additional 2526 observations for consumers located in the ten largest metropolitan areas These observations were not used in the analysis described in previous sections, and therefore provide an out-of-sample test For this sample, the heuristic produces the same result as cluster analysis in 68.4% of cases For the out-of-sample comparison, the cluster groupings are determined by applying the coefficients obtained from the insample analysis, which is described below In the remainder of this section, the authors pool the two samples, thereby providing a sample size of 6552 Of interest is the question of why the two procedures differ in one of three cases Answering this question requires further explanation of the cluster analysis methodology The cluster technique used in this paper employs a linear model that is used to identify objects Let Xi be an 11-dimensional column vector of object i's responses to the ten questions, with a as the eleventh component Let β be a × 11 matrix of coefficients Cluster analysis focuses on the product βXi, which is a 4-element column vector with the four elements corresponding to the four styles respectively The assignment algorithm assigns an object to the group associated with the maximum of the four elements Because the cluster analysis algorithm is linear, it treats the underlying environment as compensatory In this respect, a credit cardholder being a deliberate spender can compensate for him or her only making the minimum required payment each month With this in mind, consider the possible reasons for misclassification It turns out that 38% of the misclassifications occur when cluster analysis places a cardholder into the first category, low control minimum payers, while the heuristic places them into the fourth category, full balance paying single cardholders Notably, these two categories are diametrically opposite in terms of their repayment behavior Most of the classification differences arise for credit cardholders who fall between making only the minimum payment and always paying the full balance Minimum payers constitute 25% of the sample However, cluster analysis, because it factors in other variables within a compensatory framework, assigns 47% of the sample to the two groups designated as “minimum paying” This 22% difference will account for a major component of the alignment discrepancy If one were to modify the heuristic, by reclassifying a cardholder who was originally identified as being a full balance paying single cardholder, but who does not clearly state that he or she almost always pays their balance in full, then the rate of agreement between heuristic and cluster analysis would rise from 67.5% to 73.5%: a little less fast, just as frugal, and a little more accurate The modified heuristic provides the same classification as cluster analysis in roughly three of four cases As to the fourth case, 43% of the differences involve identifying someone who is a low control minimum payer with someone who is a high control minimum payer These differences largely occur for responses that are away from the extremes Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx Consider the case when cluster analysis identifies a cardholder as a low control minimum payer The most important difference between the cases when the modified heuristic produces the same classification as cluster analysis and when the modified heuristic classifies the cardholder as a high control minimum payer pertains to the importance attached to control The cases featuring the difference attach much higher value to control than the cases where the two procedures agree Similarly, consider the case when cluster analysis identifies a cardholder as a high control minimum payer, but the modified heuristic classifies him or her as a low control minimum payer The most important difference between the two cases is that the cases featuring the difference attach are much less inclined to pay their balance in full than the cases where the two procedures agree For these two types of differences, an argument can be made that the modified heuristic actually produces a more reasonable classification The economic significance of the misclassification discussed in the previous two paragraphs is relatively minor, in so far as advice is concerned Minimum payers, be they low control or high control, should plan to avoid paying interest on everyday expenses and develop a schedule for paying down their credit card balances with a view to keeping interest payments over time in check As far as advice is concerned, the only issue is the order in which these are prioritized The most frequent single difference between cluster analysis and the heuristic classification occurs when cluster analysis identifies a cardholder as a full balance paying multiple cardholder, but the modified heuristic identifies him or her as a full balance paying single cardholder This occurs in 24% of all cases where the two methods produce different classifications For the out-of-sample test, the corresponding figure is virtually identical at 73% The single most important difference between cases where the two agree and those where they not is importance of control Control is less important for the cases where the two methods differ The second most important difference is number of credit cards used Cases where the two methods disagree are associated with fewer cards In concluding this section, the authors wish to mention a second outof-sample test which provides a nice illustration of the principle “less is more.” The test involved 34 participants, 32 of whom were high school economics teachers and two of which were university economics professors Given their expertise, my initial hypothesis was that the cluster analysis technique would classify all 34 as either full balance paying single cardholders or full balance paying multiple cardholders However, surprisingly the analysis identified six of the 34 as low control minimum payers, which is a rather high 17.6% The reason is not that these particular teachers not think that control is important, or that they only make the minimum payment required each month The reason is that these particular cardholders are not confident about using online technology to manage their finances, and in the cluster analysis compensatory framework, that factor loomed large In this regard, 20% of the cluster group “full paying single credit cardholders” report that they have low confidence in managing their finances online Interestingly, the one difference produced by the modified heuristic and cluster analysis is that the latter identified a cardholder as a low control minimum payer and the modified heuristic identified the cardholder as a full balance paying multiple cardholder Chase blueprint program As indicated in the section Developing fast and frugal heuristics, the main purpose in developing a financial style heuristic is to help households better manage their spending and borrowing In this respect, a financial style should serve as a cue for taking some particular action The financial styles discussed in the section Cluster analysis were developed with this task in mind, and the associated actions relate to a set of features offered by JPMorgan Chase Blueprint features In September 2009, Chase introduced a set of online credit card features, called Blueprint, which are designed to help credit cardholders manage their spending and borrowing Blueprint is available to over 20 million Chase card customers Below, the authors briefly describe Blueprint's features, and then indicate how they relate to the financial styles described in the section Cluster analysis Blueprint offers four main features, respectively called Full Pay, Finish It, Track It, and Split Full Pay provides cardholders with an opportunity to set up a plan whereby they avoid paying interest on everyday expenses such as groceries and gasoline In return for being diligent in making payments that conform to the plan, credit cardholders save on interest Finish It enables credit cardholders to set up a plan for repaying their balances over time, so that they avoid maintaining unnecessary debt Track It enables credit cardholders to set up a system which monitors their monthly expenses by designated categories such as food, clothing, and travel Split enables credit cardholders to set up a system which breaks out large expenses in order to enhance their salience in respect to payment of associated interest and principal Blueprint website: financial styles employed and normative issues A key normative issue associated with Blueprint features is whether consumers who use them can make better decisions about their spending and borrowing A related question is whether consumers can make better use of Blueprint features by first identifying their financial styles Preliminary evidence about consumers using Blueprint is encouraging Chase reports that nearly million plans have been created since 2009 Whereas approximately 40% of U.S cardholders pay more than the minimum payment each month, 91% of Blueprint so Moreover, 90% of Blueprint users stay committed to the plans they establish Chase offers one specific credit card called “Slate with Blueprint.” Slate is distinctive in that unlike other credit cards, it does not offer traditional rewards: instead it offers access to Blueprint Therefore, consumers with an interest in better financial tools to manage their spending and borrowing have a clear incentive to choose a Slate credit card Notably, Chase reports that credit cardholders who use Slate with Blueprint pay down their balances twice as fast as other credit cardholders The authors developed the cluster analysis described above to provide credit cardholders with guidance about which Blueprint features would be most important to them Cluster analysis produces an algorithm, so that consumers can ascertain to which cluster they belong by responding to a series of questions Although 11 questions were used in the cluster analysis, only six are needed to assign people to the four groups above with 95% confidence, where the significance level means that there is a 5% chance of misclassifying a person into the wrong group In Blueprint, the four groups bear different names than those used in the section Cluster analysis Low control minimum payers are called “Make It Easy.” High control minimum payers are called “Control Seeking.” Full balance paying multiple cardholders are called “Financially Savvy.” Full balance paying single cardholders are called “Confident and in Control.” In the remainder of this section, the authors describe the research that led to these particular names for financial styles, along with comments by consumers about normative issues involving which specific Blueprint features would be useful to people with particular financial styles The primary objective of the research in question was to gauge consumer response to the use of financial styles in respect to spending and borrowing behavior using credit cards Related objectives included Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx exploring the acceptability of the financial style quiz, identifying appropriate names for the different styles, and understanding normative issues associated with whether the quiz would increase the degree to which Blueprint features would be used, and if so how The research was conducted in a series of focus groups consisting of six two-hour sessions in two locations (Philadelphia and Chicago) held on July 6–7, 2009 Each group contained seven or eight participants All group participants were between the ages of 25 and 60, partly or solely responsible for financial decision making, had and regularly used at least two general purpose credit cards In addition, half of each group used a Chase credit card as their primary card, and most were college graduates For four of the six groups, household incomes lay in the range of $50,000 to $125,000 For two of the groups, either household income exceeded $150,000 with at least one household member's income being at least $125,000, or the household had at least $250,000 in investable assets Notably, most focus group participants easily identified with the descriptions of the four different personality types provided in the section Cluster analysis In Table the authors list some of the comments made by focus group participants in respect to financial styles, as these provide additional insights into the nature of these styles The authors have no direct evidence about the normative question linking consumer financial styles to choice of Blueprint features In this regard, the Chase website does not keep track of cardholders' financial styles, and then connect these to the features they use However, the focus group research does provide some indirect evidence on this issue Focus group participants were presented with the following recommendations in the form of a prioritized feature list for how consumers with different financial styles might use Blueprint features Consumers who belong to the group “Make It Easy” should consider using Finish It, Full Pay, and Track It to reduce interest payments and increase control of their budgets Consumers who belong to the group “Control Seeking” should consider using “Finish It,” “Full Pay,” and “Split” to reduce interest Consumers who belong to the group “Financially Savvy” should consider using “Track It” and “Split” instead of relying on ad hoc budgeting heuristics based on the use of multiple credit cards Consumers who belong to the group “Confident and in Control” should consider using “Track It” and “Split” for ease of monitoring Table Focus group participant comments about the nature of the four financial styles Make It Easy Buy what you need right now, don't worry about it, don't plan I will worry about it later and pay it off eventually This was me twenty years ago I was the minimum queen Don't have any strategies No big picture thinking Control Seeking Credit is scary for this person They only use credit when they can't afford cash I see this as someone on a strict budget More of a cash, check, debit person Financially Savvy They are calculating what the best card is to use This person must have a lot of willpower There is a method to their madness They are going to stick to their regimen The tools may not be interesting to them I want to be this person Confident and in Control This seems like someone who doesn't worry about money In control, confident It states my goals I wish this was me… It's an enviable position to be in They are living within their means Very type A… organized They probably use a credit card for the convenience or the rewards and for those occasions when they use their credit cards to make large purchases, especially when those purchases are unexpected After being presented with the recommendations, focus group participants were asked for their reactions to the recommendations Overall, many of the recommendations made sense to focus group participants; however, participants did feel that there were better options for some of the specific types Below the authors summarize participants' major comments in regard to each of the four financial styles Make It Easy: Participants were not sure if people with this style would actually any of the recommendations, unless doing so was very easy Most thought Finish It to be most important feature for this financial style, but several participants wondered if the feature would require too much discipline There was consensus that Track It is very important and useful for the consumers with this financial style, because they need some help with budgeting Most thought that Full Pay was feasible for consumers with this style, and would help them gain control of their everyday spending Control Seeking: Although Finish It was not included in the recommendation list for this style, participants were surprised by its absence and felt that it should be the feature assigned top priority One participant stated: “They would want to bring the balance down as quickly as possible so that it was ready to use again for an emergency.” Split was regarded as the second most important feature for this style Participants commented that Track It (the top recommendation) was not at all relevant because this person does not use the card except for emergencies or large purchases, and therefore there would be nothing on the card to track Participants were not certain if Full Pay would be appropriate, since consumers with this style not appear to be charging everyday expenses on their cards However, some participants did feel that consumers with this style might be open to doing so if they could separate out the charges Financially Savvy: Participants noted that Track It seemed like a suitable recommendation, as consumers with this style could more easily track all of their expenses, which would otherwise be spread across different cards At the same time, other participants thought that consumers with this style would already know where they were spending their money Participants also suggested that Split could benefit consumers with this financial style, if they were willing to put large purchases on their cards and pay them off over time However, as a general matter, participants had a difficult time generating recommendations for this style, because people with this style not appear to need help with their finances One participant stated: “I don't know if you need any recommendations for this person.” Some pointed out that Full Pay provides little assistance for those who almost always pay their balance in full Confident and in Control: Participants agreed that Track It would be very useful for this financial type, as all of their spending tends to be on one card They also indicated that Split should be a top priority because consumers with this style might be open to financing a larger purchase On the other hand, Full Pay seemed redundant for consumers with this style, as they are already paying off their monthly balances From a normative perspective, the authors note that consumers using Blueprint typically identify their financial styles online and access Blueprint features online For the 25% of consumers having low confidence in their online skills for managing finances, this will prove to be a challenge For them, a fast and frugal heuristic can be an appropriate substitute For low control minimum payers especially, the issue is important Chase customers who wish to use Blueprint features to set up an appropriate plan, but lack confidence in their online skills, can use the telephone to seek help from a Chase call center representative Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014 H Shefrin, C.M Nicols / Journal of Business Research xxx (2014) xxx–xxx Conclusion In respect to financial literacy, a great variability occurs across American consumers This paper investigates the use of fast and frugal heuristics to help consumers identify their financial styles for using credit cards to engage in spending and borrowing Fast and frugal heuristics are likely to be especially valuable to consumers with low confidence in their online skills Notably, 25% of credit cardholders report that they have low confidence using online technology to manage their finances, with the corresponding figure being 44% for those most at risk This feature suggests online skills as a possible fifth group for a heuristic-based classification However, because of the diversity in respect to other variables of those with low confidence in their online skills, cluster analysis does not produce a fifth group whose major feature emphasizes this attribute In addition to the analysis of fast and frugal heuristics for identifying credit card styles, the paper makes three other contributions First, it provides new data and findings about credit card usage segmentation in respect to spending and borrowing behavior Second, it sets the new findings against the backdrop of the newly emerging literature on financial literacy Third, it describes the introduction of a new set of online financial tools, offered by a large credit card company, which consumers are now using to make decisions about their spending and borrowing In this regard, the paper links these tools to the heuristics under discussion References Agnew, J (2010) Pension participant behavior In Baker, & Nofsinger (Eds.), Behavioral finance: Investors, corporations and markets Wiley Ameriks, J., Caplin, A., & Leahy, J (2003) Wealth accumulation and the propensity to plan Quarterly Journal of Economics, 118, 1007–1047 Benartzi, S., & Thaler, R H (2007) Heuristics and biases in retirement savings behavior Journal of Economic Perspectives, 21(3), 81–104 Bendor, J (2010) Bounded rationality and politics Berkeley: University of California Press Bernheim, D (1995) Do households appreciate their financial vulnerabilities? An analysis of actions, perceptions, and public policy Tax policy and economic growth (pp 1–30) Washington, DC: American Council for Capital Formation Bernheim, D (1998) Financial illiteracy, education and retirement saving In O Mitchell, & S Schieber (Eds.), Living with defined contribution pensions (pp 38–68) Philadelphia: University of Pennsylvania Press Choi, S., Kariv, S., Müller, W., & Silverman, D (2010) Who is (more) rational? Working paper : Tilburg University Gigerenzer, G., & Gaissmaier, W (2011) Heuristic decision making Annual Review of Psychology, 62(15), 1–32 Gigerenzer, G., Todd, P., & the ABC Research Group (1999) Simple heuristics that make us smart Oxford: Oxford University Press Hilgert, M., Hogarth, J., & Beverly, S (2003) Household financial management: The connection between knowledge and behavior Federal Reserve Bulletin, 309–32 John, O P., Naumann, L P., & Soto, C J (2008) Paradigm shift to the integrative big-five trait taxonomy: History, measurement, and conceptual issues In O P John, R W Robins, & L A Pervin (Eds.), Handbook of personality: Theory and research (pp 114–158) New York, NY: Guilford Press Kahneman, D., Slovic, P., & Tversky, A (1982) Judgment under uncertainty: Heuristics and biases Cambridge: Cambridge University Press Karlan, Dean, & Zinman, Jonathan (2012) “Borrow less tomorrow” behavioral approaches to debt reduction Working Paper Financial Security Project at Boston College Levav, J., Heitmann, M., Herrmann, A., & Iyengar, S (2010) Order in product customization decisions: Evidence from field experiments Journal of Political Economy, 118(2), 274–299 Lusardi, A (2010) “Americans' financial capability: Report prepared for the Financial Crisis Inquiry Commission,” working paper Dartmouth College and NBER The findings appear in “Financial Capability in the United States,” FINRA Investor Education Foundation, December 2009 Moore, Danna (2003) "Survey of Financial Literacy in Washington State: Knowledge, Behavior, Attitudes, and Experiences", Technical Report n 03-39 Social and Economic Sciences Research Center : Washington State University Mui, Ylan Q (2010) "Good credit isn't what it used to be With cardholders more frugal, issuers raise fees for better customers" The Washington Post, July 22, 2010 Putsis, W P., Jr., & Dhar, R (2001) An empirical analysis of the determinants of category expenditure Journal of Business Research, 52(3), 277–291 Raghubir, P., & Srivastava, J (2009) The denomination effect Journal of Consumer Research, 36, 701–713 Shefrin, H., & Thaler, R H (1988) The behavioral life-cycle hypothesis Economic Inquiry, 26(4), 609–643 Statman, M., & Wood, V (2004) Investment temperament Journal of Investment Consulting Summer, 7(1), 55–66 Todd, P., & Gigerenzer, G (2000) Précis of Simple Heuristics That Make Us Smart Behavioral and Brain Sciences, 23, 727–780 Please cite this article as: Shefrin, H., & Nicols, C.M., Credit card behavior, financial styles, and heuristics, Journal of Business Research (2014), http:// dx.doi.org/10.1016/j.jbusres.2014.02.014