Three Essays on Consumer Credit Card Behavior
by
Laura C. Ricaldi, MBA
A Dissertation
In
PERSONAL FINANCIAL PLANNING
Submitted to the Graduate Faculty
of Texas Tech University in
Partial Fulfillment of
the Requirements for
the Degree of
DOCTOR OF PHILOSOPHY
Approved
Sandra Huston
Chair of Committee
Michael Finke
Vickie Hampton
Elizabeth Sharp
Jean Scott
Mark Sheridan
Dean of the Graduate School
August 2015
Copyright 2015, Laura C. Ricaldi
Texas Tech University, Laura Ricaldi, August 2015
ACKNOWLEDGMENTS
I want to acknowledge the generous support from my committee members: Dr.
Sandra Huston, Dr. Michael Finke, Dr. Vickie Hampton, Dr. Elizabeth Sharp, and Dr.
Jean Scott. Your guidance helped form my research skills and provided me with a great
platform for future publications. A huge thank you to the other faculty and doctoral
students for being great sources of advice, encouragement and inspiration. Thank you to
Cynthia Cantu and Dawn Abbott for being a sounding board and helping me with all of
the adventures as a student.
I also want to thank my family for the continuous love and support through this
entire process especially my husband, my beautiful children, my parents, my sister, my
brother and my in-laws. Without your support I would not be where I am today. Thank
you!
ii
Texas Tech University, Laura Ricaldi, August 2015
TABLE OF CONTENTS
ACKNOWLEDGMENTS .................................................................................................. ii
ABSTRACT ....................................................................................................................... vi
Publication Targets ...................................................................................................... viii
References ....................................................................................................................... x
LIST OF TABLES ............................................................................................................. xi
I. FINANCIAL SOPHISTICATION AND ITS IMPACT ON THE CREDIT CARD
DEBT PUZZLE .................................................................................................................. 1
Abstract ........................................................................................................................... 1
Introduction ..................................................................................................................... 2
Literature Review............................................................................................................ 3
Human Capital ............................................................................................................ 3
Mental Accounting and Framing ................................................................................ 5
Self-control ................................................................................................................. 6
Framework and Concepts ............................................................................................... 7
Method .......................................................................................................................... 10
Data and Sample ....................................................................................................... 10
Dependent Variable .................................................................................................. 10
Independent Variables .............................................................................................. 12
Analysis of Data ............................................................................................................ 18
Results ........................................................................................................................... 19
Descriptive Statistics ................................................................................................. 19
Results of Logistic Regression: All Revolvers Compared to Convenience Users ... 23
Results of Logistic Regression: Solvent Revolvers Compared to Insolvent Revolvers
................................................................................................................................... 26
Results of Logistic Regression: Solvent Revolvers Compared to Convenience Users
................................................................................................................................... 28
Summary and Implications ........................................................................................... 32
Appendix ....................................................................................................................... 36
References ..................................................................................................................... 41
II. FINANCIAL LITERACY AND SHROUDED CREDIT CARD REWARDS ........... 44
iii
Texas Tech University, Laura Ricaldi, August 2015
Abstract ......................................................................................................................... 44
Introduction ................................................................................................................... 46
Literature Review.......................................................................................................... 48
Method .......................................................................................................................... 49
Data and Sample ....................................................................................................... 49
Model ........................................................................................................................ 50
Dependent Variable .................................................................................................. 51
Independent Variables .............................................................................................. 51
Analysis of Data ............................................................................................................ 54
Results ........................................................................................................................... 54
Descriptive Statistics ................................................................................................. 54
Results of Logistic Regression on Type of Reward Usage ....................................... 60
Summary and Implications ........................................................................................... 62
Appendix ....................................................................................................................... 65
References ..................................................................................................................... 66
III. DEBIT OR CREDIT: THE IMPACT OF MYOPIA AND CREDIT ATTITUDE ON
PAYMENT CHOICE ....................................................................................................... 68
Abstract ......................................................................................................................... 68
Introduction ................................................................................................................... 69
Literature Review.......................................................................................................... 71
Myopia ...................................................................................................................... 71
Credit Attitude .......................................................................................................... 72
Human Capital .......................................................................................................... 73
Demographic Characteristics of Debit Card Users ................................................... 74
Framework and Concepts ............................................................................................. 75
Method .......................................................................................................................... 77
Data and Sample ....................................................................................................... 77
Model ........................................................................................................................ 78
Dependent Variable .................................................................................................. 78
Independent Variables .............................................................................................. 79
Analysis of Data ............................................................................................................ 84
Results ........................................................................................................................... 85
Descriptive Statistics ................................................................................................. 85
iv
Texas Tech University, Laura Ricaldi, August 2015
Results of Logistic Regression: Debit Card Users Compared to All Credit Card
Users ......................................................................................................................... 87
Results of Logistic Regression: Debit Card Users Compared to Revolving Credit
Card Users ................................................................................................................. 90
Results of Logistic Regression: Debit Card Users Compared to Convenience Credit
Card Users ................................................................................................................. 92
Summary and Implications ........................................................................................... 95
Appendix I .................................................................................................................. 101
Principal Components Analysis and Factor Analysis ............................................. 101
Appendix II ................................................................................................................. 105
Logistic Regression: Debit Card Users Compared to All Credit Card Users ......... 105
Logistic Regression: Debit Card Users Compared to Revolving Credit Card Users
................................................................................................................................. 106
Logistic Regression: Debit Card Users Compared to Convenience Credit Card Users
................................................................................................................................. 107
Logistic Regression: Revolving Credit Card Users Compared to Convenience Credit
Card Users ............................................................................................................... 108
References ................................................................................................................... 110
v
Texas Tech University, Laura Ricaldi, August 2015
ABSTRACT
Credit card use and consumer debt has increased among American households in
the past years. Gross and Souleles (2002) suggest that maintaining a high credit card
balance is a behavioral issue rather than a liquidity issue. Behavioral biases can lead a
household to make ineffective financial decisions regarding credit card use, investment
decisions, and retirement savings.
Households have many opportunities to make either effective or ineffective
decisions regarding credit card use. Some examples of effective decisions when using
consumer credit are using the card as a convenience tool, using rewards programs
efficiently, and using credit cards to help build credit. Less effective consumer credit
decisions include: revolving a balance, not using rewards programs, and paying
expensive fees. These ineffective decisions can affect an individual’s credit score for
many years and have negative consequences that will affect other financial decisions.
With the growth of American credit balances, the credit card industry has
developed products that require complex decisions. Financial decision-making has
become exceedingly complex and requires a certain level of human capital to make
effective financial decisions.
Financial knowledge is key to making effective decisions, especially with
consumer credit. A measurement of financial literacy can assist researchers with finding
different effects of financial knowledge and the impact this knowledge has on decisionmaking.
vi
Texas Tech University, Laura Ricaldi, August 2015
According to Huston (2010), financial literacy refers to the household’s ability to
understand and use financial information to make decisions. In order for households to
be financially literate, they must demonstrate ability, knowledge and confidence when
making financial decisions (Huston, 2010). Often, financial sophistication serves as a
proxy for financial literacy. Financial sophistication refers to the household’s ability to
apply knowledge and make complex financial decisions.
Prior research shows that households with higher levels of financial literacy make
better financial decisions. For example, Lusardi (2008) finds that those with lower levels
of financial literacy tend to not participate in the stock market, fail to plan for retirement,
and exhibit poor borrowing behavior. Thus, observing financial literacy and the impact
on consumer decisions is relevant for financial planners, academics, and counselors.
The data used for the dissertation includes two data sets. Both data sets are rich
with detailed financial information on consumer credit use, financial assets and liabilities,
and other financial information. The first data set is the Consumer Finances Monthly
survey which includes a comprehensive measure for financial literacy. By using a
comprehensive measure for financial literacy, the research evaluates the impact of
financial literacy on household consumer credit decisions. The second data set is the
Survey of Consumer Finances (SCF). Although the SCF includes detailed financial
information, it does not include a specific financial literacy component. Huston, Finke
and Smith (2012) conducted a factor analysis to create a proxy for financial
sophistication in the SCF. The research uses this financial sophistication proxy in the
first paper.
vii
Texas Tech University, Laura Ricaldi, August 2015
The first paper discusses the financial sophistication of households that display
the credit card debt puzzle. Households that display this puzzling behavior are revolving
households that hold enough liquid assets to pay off the balance, but choose not to. The
study finds that financially sophisticated households are less likely to display the credit
card debit puzzle.
The second paper discusses the shrouded attributes associated with credit card
rewards programs. It is hypothesized that naïve choice among credit card consumers may
be explained by financial literacy. When household characteristics such as education,
income and wealth are controlled, respondents in the highest financial literacy quintile
were twice as likely to have a rewards card. These results imply that consumers with
financial literacy deficiencies risk consumer credit product exploitation.
The third paper discusses the choice between debit or credit use and the possible
factors that impact the choice. These factors include myopia, credit attitude, the financial
sophistication of the household, life-cycle factors, and finances. The study discovered
that respondents who are myopic are typically using debit cards and revolving a credit
card balance. The results imply that households that are myopic and have above average
credit attitude should use debit cards to avoid the high costs of revolving a credit card.
Publication Targets
Essay One: Financial Sophistication and its Impact on the Credit Card Debt Puzzle
Journal of Financial Counseling and Planning: The journal is a blind peerreviewed journal that is published bi-annually. The journal’s focus is to publish financial
planning material as well as financial counseling techniques. The audience consists of
viii
Texas Tech University, Laura Ricaldi, August 2015
academics, financial counselors, practitioners, and others in the financial services
industry.
Essay Two: Financial Literacy and Shrouded Credit Card Rewards
Journal of Financial Services Marketing: The journal focuses on financial
marketing, personal finance, consumer finance, and other areas related to financial
services. The journal is blind peer-reviewed and the goal of the journal is to keep up with
the latest developments and research in financial services marketing. The audience
consists of academics, practitioners and policy-makers in the financial services industry.
Essay Three: Debit or Credit: The Impact of Myopia and Credit Attitude on Payment
Choice
Journal of Financial Therapy: The journal focuses on personal financial
knowledge, attitudes and behaviors associated with family well-being. The journal is
peer-reviewed and articles are written by financial therapists, financial planners,
counselors, other practitioners and various academics.
ix
Texas Tech University, Laura Ricaldi, August 2015
References
Gross, D. B. & Souleles, N. S. (2002). Do liquidity constraints and interest rates matter
for consumer behavior? Evidence from credit card data. The Quarterly Journal of
Economics, February, 149-185.
Huston, S. J. (2010). Measuring financial literacy. Journal of Consumer Affairs, 44, 296316.
Huston, S. J., Finke, M. S., & Smith, H. (2012). A financial sophistication proxy for the
Survey of Consumer Finances, Applied Economics Letters, 19:13, 1275-1278.
Lusardi, A. (2008). Financial literacy: An essential tool for informed consumer choice?
(No. w14084). National Bureau of Economic Research.
x
Texas Tech University, Laura Ricaldi, August 2015
LIST OF TABLES
1.1: Coding of Variables Used in the Study ..................................................................... 13
1.2: Descriptive Statistics for Credit Card Users. ............................................................. 20
1.3: Logistic Regression of the Likelihood of Being a Revolving User Compared to a
Convenience User in the 2010 SCF .................................................................................. 24
1.4: Logistic Regression of the Likelihood of Being an Insolvent Revolving User
Compared to a Solvent Revolving User in the 2010 SCF ................................................ 27
1.5: Logistic Regression of the Likelihood of Being a Solvent Revolving User Compared
to a Convenience User in the 2010 SCF ........................................................................... 30
1.6: Descriptive Statistics for the Top Quintile of Solvent Revolvers and Convenience
Credit Card Users. ............................................................................................................. 37
1.7: Logistic Regression of the Likelihood of Being a Top Quintile Solvent Revolver
Compared to a Top Quintile Convenience User in the 2010 SCF .................................... 39
2.1: Descriptive Statistics for Convenience Credit Card Users ........................................ 55
2.2: T-Tests of Differences between Rewards users and Non-Rewards Users................. 57
2.3: Chi-square Analyses of the Differences in Characteristics of Households between
Rewards Convenience Users and Non-Rewards Convenience Users............................... 59
2.4: Logistic Regression of the Likelihood of being a Rewards User Compared to a NonRewards User .................................................................................................................... 61
2.5: Financial Literacy Questionnaire from the Consumer Finances Monthly (CFM)
Survey ............................................................................................................................... 65
3.1: Coding of Variables Used in the Study ..................................................................... 80
3.3: Logistic Regression of the Likelihood of Being a Debit Card User Compared to All
Credit Card Users in the SCF between 1998-2010 ........................................................... 88
3.4: Logistic Regression of the Likelihood of Being a Debit Card User Compared to a
Revolving Credit Card User in the SCF between 1998-2010 ........................................... 91
3.5: Logistic Regression of the Likelihood of Being a Debit Card User Compared to a
Convenience Credit Card User in the SCF between 1998-2010....................................... 94
3.6: Final Component Loading for Credit Attitude and Myopia .................................... 103
xi
Texas Tech University, Laura Ricaldi, August 2015
3.7: Factor Loading ......................................................................................................... 103
3.8: Logistic Regression of the Likelihood of being a Debit Card User Compared to All
Credit Card Users in the SCF between 1998-2010 ......................................................... 105
3.9: Logistic Regression of the Likelihood of being a Debit Card User Compared to
Revolving Credit Card Users in the SCF between 1998-2010 ....................................... 106
3.10: Logistic Regression of the Likelihood of being a Debit Card User Compared to
Convenience Credit Card Users in the SCF between 1998-2010 .................................. 107
3.11: Logistic Regression of the Likelihood of Being a Revolving Credit Card User
Compared to a Convenience Credit Card User in the SCF between 1998-2010 ............ 108
3.12: Logistic Regression of the Likelihood of being a Revolving Credit Card User
Compared to Convenience Credit Card Users in the SCF between 1998-2010 ............. 109
xii
Texas Tech University, Laura Ricaldi, August 2015
CHAPTER I
FINANCIAL SOPHISTICATION AND ITS IMPACT ON THE
CREDIT CARD DEBT PUZZLE
Abstract
The credit card debt puzzle is not well understood. Households exhibit inefficient
behavior when they have sufficient liquid assets to pay off their credit card balance, but
do not. Based on three logistic regression analyses of the 2010 Survey of Consumer
Finances, the study discovered that households that display this behavior are more likely
to have a lower financial sophistication than convenience users. In the current study,
solvent revolvers are less likely to be over the age of 55, less likely to have a college
degree, and less likely to be self-employed. Solvent revolvers are more likely to work
part time, view credit as good, and have history of late payments and bankruptcy. There
is a negative relationship with household income. These findings suggest financially
sophisticated households are less likely to display irrational behavior regarding the credit
card debt puzzle.
Key words: Credit cards, revolving users, behavioral life-cycle hypothesis,
financial sophistication, 2010 Survey of Consumer Finances
1
Texas Tech University, Laura Ricaldi, August 2015
Introduction
Household credit card use has been widespread in recent years. Over 70% of U.S.
households have a credit card (Bucks, Kennickell, Mach & Moore, 2009). There are two
main groups of credit card users: those users who pay off their balance at the end of each
month, commonly called convenience users, and those who carry a balance from month
to month. These households that carry a balance from month to month are known as
revolving credit card users (Kim & DeVaney, 2001). According to the 2007 Survey of
Consumer Finances, approximately 46% of families carry a balance on their credit cards.
Since 2004, the median balance increased by 25% and the mean balance rose 30.4%
(Bucks et al., 2009).
Within the revolving credit card users, there is another group of users who have
enough liquid assets to pay the balance of their credit cards but choose not to pay it off.
Bi (2005) found that approximately 58% of revolving credit card users have liquid assets
in excess of their credit card balance. These users are called solvent revolvers. Previous
studies that have investigated this irrational behavior call this the credit card debt puzzle
(Laibson, Repetto, & Tobacman, 2001; Haliassos & Reiter, 2005; Bertaut, Haliassos, &
Reiter, 2009).
Credit cards usually have high interest rates associated with revolving balances
while liquid asset accounts, like checking and money market accounts, typically have low
after-tax returns. Given this information, it is inefficient for a household to maintain a
revolving balance.
2
Texas Tech University, Laura Ricaldi, August 2015
The purpose of this study is to look at the impact of financial sophistication on
solvent revolving credit card users and the behavioral factors that affect the decision to be
a solvent revolver. Behavioral life cycle theory suggests a household is composed of two
dueling selves, the planner and the doer. The planner is the forward-thinking, rational
self while the doer is myopic and focused on current consumption (Shefrin & Thaler,
1988). The household makes decisions to satisfy both the planner and doer and may
display conflicting and inefficient behaviors, such as the credit card debt puzzle.
Literature Review
Several studies attempted to explain the credit card debt puzzle, or solvent
revolving credit card use, with a number of different factors. Gross and Souleles (2002)
found that high credit card balances stem from a person’s behavior, not liquidity
problems. Behavioral factors in the planner/doer model include human capital, mental
accounting/precautionary savings motives, and self-control and time management
constraints.
Human Capital
The planner/doer model suggests the planner will reduce consumption in the
current period by exerting a level of willpower. Since willpower is costly, the planner
will resort to other techniques to reduce the consumption of the doer. Some of these
techniques include mental accounting and rule setting (Shefrin & Thaler, 1988). In the
planner/doer model, the level of human capital can impact financial decisions. Becker
(1964) described human capital as an individual’s stock of knowledge, health, skills or
values. It is a function of goods, services, time and the individual’s current stock of
3
Texas Tech University, Laura Ricaldi, August 2015
human capital. Human capital is often improved through learning, maturity and
experiences. Human capital is also broken down into two categories: general and
specific. Individuals can improve their general human capital by obtaining a college
education or having life experiences that provide knowledge about future decisionmaking. In the realm of finance, individuals can improve their specific human capital by
taking financial courses to improve their ability to understand and make effective
financial decisions. Individuals can also improve their financial specific human capital
through experiences like using credit cards or taking out a home mortgage. Households
with a high level of financial knowledge and experience are considered financially
literate or sophisticated; these households are able to make more effective financial
decisions than the households with a lower level of financial literacy or sophistication.
Financial sophistication gives the household the potential to improve their ability
to make better financial decisions. The households with a higher level of financial
sophistication tend to be aware of the consequences of their decisions. Bertaut,
Haliassos, and Reiter (2009) suggested that financially sophisticated households would
be convenience users of credit cards as well as benefit from floating and other advantages
of credit cards.
Although, some financially sophisticated households display characteristics that
are not sophisticated. For example, Haliassos and Reiter (2005) found that in the
shopper/accountant model, the shopper is not fully financially sophisticated. The
accountant/shopper model is similar to the planner/doer model.
4
Texas Tech University, Laura Ricaldi, August 2015
Mental Accounting and Framing
In order to control the doer, the planner creates mental accounts to reduce the
temptation to spend from them. Households divide assets, expenditures and income into
different categories or mental accounts. An economist would state that these accounts are
substitutable, but in reality, they are not (Thaler, 1999). The household views the mental
accounts, either assets or expenses, as different things and marginal propensity to
consume from these accounts are different. For example, households save money in an
emergency fund account to prepare for an uncertain event. The household uses framing to
earmark these accounts for difference purposes. Households that save in liquid accounts
for emergencies or unexpected events do not believe the assets are substitutable for other
assets or expenses.
When households use mental accounting for expenses, especially with credit
cards, they decouple the payment from the consumption. Once the bill is received, the
purchase is mixed with other purchases. Thaler (1999) stated that it is hard for the
consumer to attribute the balance to any particular purchase; therefore the consumer
carries a balance from month to month. Since households have mental accounts, they
will not view accounts used for savings as available to pay off credit card balances since
these accounts are not substitutable. Households will also hold assets in different
physical bank accounts since they view these assets as non-fungible.
Uncertainty and precautionary savings motives play a role in mental accounting.
Telyukova and Wright (2008) proposed that households would stay solvent revolvers to
maintain sufficient liquid assets for uncertain future events. Bi and Hanna (2006) found
5
Texas Tech University, Laura Ricaldi, August 2015
that households will display the credit card debt puzzle when precautionary savings
motives are present.
Self-control
Although the planner creates mental accounts to control the doer, the individual
must exhibit self-control. An individual displays self-control issues by either postponing
action (e.g., procrastination) or by consuming immediately (e.g., no willpower to wait).
Households that display self-control issues either consume all of their resources without
saving or paying debt, or put off making critical decisions. One study showed that in an
accountant/shopper household, the accountant would choose not to pay off the credit card
balance in order to impose control over the shopper (Bertaut, Haliassos, & Reiter, 2009).
By reducing the available limit on the credit card, the shopper is unable to consume more.
Other studies have identified other behavioral factors that affect solvent revolving
credit card use. Credit attitude and bankruptcy history are often considered when
households exhibit the credit card debt puzzle. First, Chien and DeVaney (2001) found
that a positive credit attitude was related to a higher credit card balance. Rutherford and
DeVaney (2009) found that those households that had a positive attitude toward credit are
less likely to be convenience users. Next, bankruptcy history has also been shown to be a
factor for the credit card debt puzzle. Lehnert and Maki (2007) evaluated the bankruptcy
exemption levels for each state. They found a higher prevalence of solvent revolvers in
the states that had a higher level of exemption. One possible theory was that households
will run up credit card debt and keep the liquid assets to transfer them into assets that are
exempt when filing for bankruptcy.
6
Texas Tech University, Laura Ricaldi, August 2015
Framework and Concepts
Other studies evaluate how the credit card debt puzzle and the relevant behavioral
factors are related. To find how behavioral factors, like financial sophistication, will
influence the likelihood of being a solvent revolver, this study uses a combination of the
behavioral life-cycle hypothesis and human capital theory.
Like the life-cycle hypothesis, the behavioral life-cycle hypothesis (Shefrin &
Thaler, 1988) suggests that in order to maximize utility, a household will shift resources
in periods where the marginal utility of consumption is relatively low to periods where
the marginal utility of consumption is relatively high. A good example of this is when
households save during the working years for consumption during retirement years.
Unlike the traditional life-cycle hypothesis, the behavioral life-cycle hypothesis posits
that households have a dual preference framework where they are both planners (longterm) and doers (short-term). The planner preference is when households make rational
decisions regarding when to shift resources to maximize utility. The planner focuses on
long-term decisions to maximize utility. The doer preference, or short-term preference, is
when households succumb to temptation to consume in the current period. The three
behavioral factors, self-control, mental accounting, and framing, are what make the
behavioral life-cycle hypothesis different from other life-cycle models.
Self-control refers to the household’s temptation to make immediate consumption
decisions, rather than saving for future consumption. For the doer, immediate
consumption is always a tempting alternative to future consumption. There is discomfort
for the doer associated with postponing current consumption; therefore, the planner will
7
Texas Tech University, Laura Ricaldi, August 2015
enforce saving devices and rules of thumb to deal with self-control issues for various
situations. These are types of external rules that households use to plan for future
consumption. Households also use internal rules, like refusing to borrow for current
consumption, to maintain self-control.
Mental accounting refers to placing wealth into different non-substitutable
accounts. The typical breakdown of mental accounts is current income, current assets,
and future income (Shefrin & Thaler, 1988). Households use mental accounting to
restrict the doer from bringing future resources into the current period. The way a
household frames the different mental accounts determines the temptation to spend from
each account. Each account has a different level of temptation associated with spending
from it. The marginal propensity to consume from the current income account is much
higher than the marginal propensity to consume from the future income account.
Temptation plays an important role in the household’s decision to spend or save.
By carrying a credit card balance, the doer is bringing consumption into the
current period. The households that are solvent (i.e., households who have enough liquid
assets to pay off their balance but do not) are not displaying the planner behavior, but are
displaying the doer behavior. In contrast, convenience users are keeping future resources
in future periods by paying off the balance while still benefiting from the advantages of
using a credit card.
In addition to the behavioral life-cycle hypothesis, human capital theory also
plays a part in solvent revolving credit card use. Human capital is an individual’s
knowledge, health, skills or values and is often described as a function of goods, services,
8
Texas Tech University, Laura Ricaldi, August 2015
time and the individual’s current stock of human capital. A household’s level of human
capital impacts its ability to make efficient financial decisions. Households with a higher
level of financial human capital (i.e., financial sophistication) have the potential to
improve the ability to make effective and efficient financial decisions.
Based on the theoretical framework, the concepts developed for this paper include
human capital, mental accounting/precautionary savings motives, time constraint/selfcontrol factors, and other control factors. Solvent revolving credit card use is a function
of these four concepts. The concepts serve as control factors to help explain why
households display puzzling behavior that is inefficient.
Based on this model and the four concepts, several hypotheses are developed.

Hypothesis 1: Solvent credit card revolvers will have a lower level of
human capital.
o Solvent credit card revolvers will be less financially sophisticated
than convenience users.
o Solvent credit card revolvers will have less education than
convenience users.

Hypothesis 2: Solvent credit card revolvers will be more likely to display
mental accounting behaviors and have higher precautionary savings
motives.

Hypothesis 3: Solvent credit card revolvers will be more likely to
experience time constraint/self-control factors. Time constrained
9
Texas Tech University, Laura Ricaldi, August 2015
households will be younger, have more children, work more hours per
week, and be less likely to pay their bills on time.
Method
Data and Sample
The data used were from the 2010 Survey of Consumer Finances (SCF), a
triennial survey, which is sponsored by the Federal Reserve Board and collected by the
National Organization for Research at the University of Chicago. The SCF collects
detailed information on the finances of U.S. households. The 2010 SCF included 6,492
households in the public data set. Since this study only analyzes those households that
have a credit card, the sample is limited to 4,433 observations. The 2010 SCF contains
five implicates to deal with missing and incomplete data; only the first implicate was
used in this study.
Dependent Variable
The dependent variable was constructed by categorizing credit card users into one
of three categories. See Table 1.1 for the measurement of the dependent variable.
1. Solvent credit card users. These users have liquid assets greater than or
equal to the balance still owed on their main credit card after the last
payment was made to the account. The total liquid assets is derived from
the Federal Reserve Board definition in the net worth code.
10
Texas Tech University, Laura Ricaldi, August 2015
2. Insolvent credit card users. These users have liquid assets less than the
balance still owed on their main credit card after the last payment was
made to the account.
3. Convenience credit card users. These users do not have an outstanding
balance on their main credit card.
Three logistic regressions were run to establish if there are differences between
several combinations of the dependent variable groups. For the first regression, the
dependent variable was coded as 1 if the household is a revolving credit card user and 0 if
the household is a convenience user. This regression was run to compare the differences
between all revolvers and convenience users of credit cards. It is important to distinguish
the difference between the two general groups of credit card users before breaking down
the revolving group into a specific type of revolvers.
After comparing the general groups of credit card users, it is important to evaluate
the differences between the two specific types of revolving credit card users. For the
second logistic regression, the dependent variable was coded as 1 if the household is a
solvent revolving user and 0 if the household is an insolvent revolving user.
For the third logistic regression, the dependent variable was coded as 1 if the
household is a solvent revolving user and 0 if the household is a convenience user. This
regression was run because the solvent revolving user has the ability to be a convenience
user, but chooses not to pay off the debt. It is important to establish the differences
between these two groups to explain why households remain solvent revolvers.
11
Texas Tech University, Laura Ricaldi, August 2015
Independent Variables
Based on the behavioral life-cycle hypothesis and human capital theory, four
concepts were identified: human capital, mental accounting/precautionary savings
motives, time constraint/self-control factors, and other control factors. Independent
variables operationalized these concepts. The measurement of the independent variables
is presented in Table 1.1.
12
Texas Tech University, Laura Ricaldi, August 2015
Table 1.1: Coding of Variables Used in the Study
Variable
Dependent Variables
Solvent Revolver
Insolvent Revolver
Convenience User
Independent Variables
Human Capital
Financial Sophistication (X3913, X3014, X414, X6525)
Quintile 1(most sophisticated) (reference group)
Quintile 2
Quintile 3
Quintile 4
Quintile 5 (least sophisticated)
Education (X5901, X5902, X5904)
Less than high school
High school graduate (reference group)
Some college
College degree
Mental Accounting/Precautionary Savings Motives
Saving for Emergencies (X3006, X3007, X7513, X6514, X7515, X6848, X7187)
Saving for Unemployment (X3006, X3007, X7513, X6514, X7515, X6848,
X7364, X7586)
Saving for Illness (X3006, X3007, X7513, X6514, X7515, X6848, X6030, X6124)
Ability to borrow from friends/relatives (X6443)
Self Employed (X4106, X4706)
Number of Liquid accounts (X3504, X3728)
Time Constraint/Self-control Factors
Number of children (X108, X114, X120, X126, X132, X202, X208, X214, X220)
Payment History (X3004)
On time/No payment (reference group)
Behind
Number of hours worked in a week (X4110)
Not working (reference group)
Working less than full time (1 to 39 hours)
Working full time (≤40 hours)
Credit attitude (X401)
Positive
Ambivalent (reference group)
Negative
Bankruptcy history (X6772)
Lifecycle Factors
Age (X14)
Under 35 (reference group)
35 to 55
Over 55
Gender (X8021)
Male (reference group)
Female
Marital Status (X7372)
Married (reference group)
Not married
Income/$10,000 (X5729)
Source: 2010 Survey of Consumer Finances.
13
Measurement
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
Continuous
Continuous
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
1 if yes, 0 otherwise
Continuous
Texas Tech University, Laura Ricaldi, August 2015
Human Capital
The human capital concept explains the household’s potential to make effective
decisions. Households that do not have a strong base in financial human capital (i.e.,
financial sophistication) tend to make suboptimal financial decisions (Bertaut, Haliassos,
& Reiter 2009). Since the purpose of this study is to evaluate the financial sophistication
of solvent revolvers, this concept is the main focus. The human capital concept was
measured by two independent variables, financial sophistication and education. Financial
sophistication represents a specific type of human capital while education represents a
general type of human capital. The variables that make up this concept represent why the
household makes suboptimal decisions and does not pay off their credit card balance
even though the household has the resources to do so.
Huston, Finke and Smith (2012) developed a financial sophistication score based
on a factor analysis of the Survey of Consumer Finances. The score includes four
variables: stock ownership (within or outside of tax sheltered accounts), willingness to
accept at least some investment risk, not revolving more than 50% of credit card limit,
and the level of understanding of personal finance. The indirect measure of financial
sophistication is based on observed and self-reported behavioral variables instead of
directly measured financial literacy. A direct measure of financial knowledge, ability and
confidence is more effective at determining financial sophistication, but the SCF does not
contain such variables.
The score is broken into five quintiles to see the magnitude between the groups of
credit card users. Quintile 1 is the most sophisticated while Quintile 5 is the least
14
Texas Tech University, Laura Ricaldi, August 2015
sophisticated. The most sophisticated quintile (#1) was omitted for the regression
analyses as the reference group.
The second variable included in the human capital concept is the level of
education for the head of household. Education represents a general type of human
capital. The level of education was categorized as: less than high school, high school
degree, some college, and college degree. The high school degree variable was the
reference group for the regression analyses.
Mental Accounting
The mental accounting concept relates to how the planner controls the doer’s
consumption by using various mental accounts. Past literature shows that households
maintain solvent revolving due to mental accounting and the framing of the different
accounts. With the behavioral life-cycle hypothesis, households do not view mental
accounts as substitutable. Households do not tend to use emergency savings accounts to
pay credit card balances if they are not in an emergency situation. The variables that
make up this concept represent why the household will maintain liquid assets in excess of
their credit card balance and maintain their status as a solvent revolver.
Mental accounting was measured by six independent variables. The first variable,
saving for emergencies, was constructed by combining two variables: if the household
indicated they had a savings motive for emergencies or other unexpected needs and if
they stated a positive amount for a subjective emergency fund. The variable was coded as
1 if the household had a motive to save for emergencies and 0 otherwise. The second
variable, saving for unemployment, was also constructed by combining two variables: if
15
Texas Tech University, Laura Ricaldi, August 2015
the household stated they had a savings motive for unemployment and if they expect their
future income will decrease in comparison with prices in the next year. The variable was
coded as 1 if the household had a motive to save for unemployment and 0 otherwise. The
third variable, saving for illness, was constructed by combining two variables: if the
household stated they had a savings motive for in case of illness or future medical
expenses and if they have a poor health status. The variable was coded as 1 if the
household had a motive to save for illness and 0 otherwise. The fourth variable is the
household’s ability to borrow $3,000 from friends or relatives in an emergency. The
variable was coded as 1 if the household was able to borrow and 0 if the household was
not able to borrow from friends or relatives. The fifth variable is if either the head of
household or the spouse is self-employed. The variable was coded as 1 if self-employed
and 0 if not self-employed. The last variable is the number of liquid accounts a
household owns. This variable is a summation of the number of checking accounts and
the number of savings/money market accounts owned by the household. The number of
liquid accounts is coded as a continuous variable.
Time Constraint/Self-control Factors
The time constraint/self-control concept is included since households have limited
time to make financial decisions. Households also display self-control issues regarding
financial decision making. Self-control plays a role in the household’s susceptibility to
give in to temptation to spend/consume during the current period. Households also put
off making complex financial decisions. Time-constrained households will procrastinate
paying off the balance on their credit cards. The variables that make up this concept
16
Texas Tech University, Laura Ricaldi, August 2015
represent the time constraints a household encounters that retain the household’s status as
a solvent revolver.
The time constraint/self-control concept was comprised of number of children,
past payment history, number of hours worked per week, bankruptcy history, and credit
attitude. First, having children reduces the amount of time an individual has to devote to
making financial decisions. The number of children was a coded as a continuous
variable. The second variable is the past payment history of all loans, mortgages and
credit cards made during the past year. The variable was coded as 1 for those households
who made payments on schedule or had no payments and 0 for those who were behind or
missed payments. The on time/no payments category was the reference group for the
regression analyses. The third variable for this concept is the number of hours worked in
a week. The variable was coded as a categorical variable with 0 hours as not working, 1
to 39 hours a week as working less than full time, and greater than or equal to 40 hours
per week as working full time. The reference group is the not working category. Past
literature shows that a household maintains solvent revolving due to credit attitude and
bankruptcy history. The household’s credit attitude was measured by their feelings about
using credit. The variable was coded as positive if the household feels credit is a good
idea, ambivalent if the household feels credit is good in some ways and bad in others, and
negative if they feel credit is a bad idea. The ambivalent credit attitude group was the
reference group for the regression analyses. Last, bankruptcy history was coded 1 if the
household has ever filed for bankruptcy, or 0 if they have never filed for bankruptcy.
17
Texas Tech University, Laura Ricaldi, August 2015
Lifecycle Factors
The lifecycle factors concept is included since households make decisions based
on being in different stages of the lifecycle. The variables that make up this concept
represent why the household will maintain liquid assets in excess of their credit card
balance and maintain their status as a solvent revolver.
The concept was measured using age, gender, marital status, and income. First,
age was coded categorically as under 35, between 35 and 55, and over 55. The age
category over 55 is the reference group for the regression analyses. Next, gender and
marital status are included in the analysis. Gender is coded as male or female and male is
the reference category for the regression analyses. Marital status is coded as married or
not married. The married category is the reference group. Last, household income is a
continuous variable. Income is scaled by $10,000 to see the magnitude and its effect in
each of the regressions.
Analysis of Data
Descriptive statistics were conducted to look at the characteristics of households.
To generalize the findings back to the U.S. population, the descriptive statistics were
weighted using a weight variable provided by the Federal Reserve (Lindamood, Hanna &
Bi, 2007). Since the dependent variables are binary, logistic regression was used to
predict the likelihood of the dependent variable occurring given the set of independent
variables. The regression analyses were not weighted (Lindamood, Hanna & Bi, 2007).
18
Texas Tech University, Laura Ricaldi, August 2015
Results
Descriptive Statistics
Since the descriptive statistics are weighted, the reported percentages, means, and
standard errors represent all U.S. households. See Table 1.2 for a summary of the
descriptive statistics.
19
Texas Tech University, Laura Ricaldi, August 2015
Table 1.2: Descriptive Statistics for Credit Card Users.
Population Percent
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
Quintile 2
Quintile 3
Quintile 4
Quintile 5 (least sophisticated)
Education
Less than high school
High school graduate
Some college
College degree
Mental Accounting/Precautionary Savings Motives
Saving for Emergencies
Saving for Unemployment
Saving for Illness
Ability to borrow from friends/relatives
Self Employed
Number of liquid accounts
Insolvent
Revolver
n=985
27.43
Solvent
Revolver
n=949
25.06
Convenience
User
n=2,499
47.51
8.46
11.02
25.46
29.11
25.95
18.92
18.53
19.76
21.5
21.29
27.16
26.01
16.97
13.95
15.92
8.64
31.12
20.11
40.13
6.42
27.68
23.1
42.8
4.71
24.41
14.87
56.02
40.04
3.63
3.78
37.12
15.69
0.4889
(0.0159)
38.38
3.77
3.9
28.04
15.71
0.7327
(0.0144)
36.21
3.13
5.86
20.76
17.38
0.6958
(0.0092)
1.0145
(0.0390)
0.8543
(0.0378)
0.6038
(0.0201)
70.86
29.14
85.51
14.49
96.56
3.44
21.02
15.18
63.8
21.53
14.52
63.95
36.32
10.97
52.71
24.47
43.88
31.65
19.53
25.77
45.93
28.31
11.85
21.67
44.65
33.68
4.52
19.95
52.21
27.84
20.79
45.5
33.71
14.21
32.19
53.6
75.59
24.41
79.43
20.57
78.4
21.6
Time Constraint/Self-control Factors
Number of Children
Payment History
On time/ No payment
Behind
Number of hours worked in a week
Not working
Working less than full time (1 to 39 hrs)
Working full time (≤40 hrs)
Credit attitude
Positive
Ambivalent
Negative
Bankruptcy history
Lifecycle Factors
Age
Under 35
35 to 55
Over 55
Gender
Male
Female
Marital Status
Married
Not married
56.55
59.14
61.88
43.45
40.86
38.12
$64,990.41
$77,964.79
$129,897.18
Income
(1,731.65)
(2,892.96)
(9,590.03)
Source: 2010 Survey of Consumer Finances. Statistics derived from weighted analysis of one implicate.
(Mean (standard error) for continuous variables; column percents for categorical variables) (n=4,433)
20
Texas Tech University, Laura Ricaldi, August 2015
Insolvent Revolvers
Overall 27.43% of the sample are considered insolvent revolvers. About 20% of
the respondents are in the top two quintiles of financial sophistication, with 8.46% in the
top (most literate) quintile and 11.02% in the second quintile. About half of the
respondents have a high school education, and about 40% have a college degree. About
40% of the respondents have a precautionary savings motive for emergencies, about 4%
have a savings motive for unemployment, and about 4% have a savings motive for
illness. About 37% are able to borrow from friends and relatives during an emergency
and 15.69% of the respondents are self-employed. The average number of liquid
accounts is 0.4889. The average number of children is 1.0145. About 70% of the
respondents reported paying their bills on time or do not have payments. The majority of
households (63.80%) work full time. Approximately one quarter of the respondents
report a positive attitude toward credit (24.47%), about 44% report an ambivalent attitude
toward credit, and 31.65% of respondents report a negative attitude toward credit. About
19.5% of the respondents report filing for bankruptcy. Approximately half of the
respondents (52.21%) are between the age of 35 and 55. The majority of the respondents
are male (75.59%) and more than half (56.55%) are married. The average income for
insolvent revolvers was $64,990.
Solvent Revolvers
Overall 21.41% of the sample are considered solvent revolvers. About 18% of
the respondents are in each of the top two quintiles of financial sophistication. Over 50%
of the respondents have a high school education, and about 42% have a college degree.
More than one third (38.38%) of the respondents have a precautionary savings motive for
21
Texas Tech University, Laura Ricaldi, August 2015
emergencies, about 4% have a savings motive for unemployment, and about 4% have a
savings motive for illness. About 28% are able to borrow from friends and relatives
during an emergency and 15.71% of the respondents are self-employed. The average
number of liquid accounts is 0.7327. The average number of children reported is 0.8543
per household. The majority (85.51%) of the respondents report paying their bills on time
or do not have payments. The majority of the households (63.95%) work full time.
Approximately one quarter of the respondents report a positive attitude toward credit
(25.77%), about 45% report an ambivalent attitude toward credit, and 28.31% of
respondents report a negative attitude toward credit. About 11.85% of the respondents
report filing for bankruptcy. Approximately half of the respondents (45.50%) are
between the age of 35 and 55. The majority of the respondents are male (79.43%) and
more than half (59.14%) are married. The average income for solvent revolvers was
$77,965.
Convenience Users
Overall 47.51% of the sample is considered convenience users. About 53% of the
respondents are in the top two quintiles of financial sophistication, with 27.16% in the top
(most literate) quintile and 26.01% in the second quintile. The majority of the
respondents have a college degree (56.02%). More than 36% of the respondents have a
precautionary savings motive for emergencies, about 3% have a savings motive for
unemployment, and about 6% have a savings motive for illness. About 21% are able to
borrow from friends and relatives during an emergency and 17.38% of the respondents
are self-employed. The average number of liquid accounts is 0.6958. The average
number of children in the household is 0.6038. The vast majority (96.56%) of the
22
Texas Tech University, Laura Ricaldi, August 2015
respondents report paying their bills on time or do not have payments. About half
(52.71%) of the households work full time with about 36% working less than full time.
Approximately 22% of the respondents report a positive attitude toward credit, about
44% report an ambivalent attitude toward credit, and 33.68% of respondents report a
negative attitude toward credit. About 4.5% of the respondents report filing for
bankruptcy. Approximately one third of the respondents (32.19%) are between the age of
35 and 55 with over 50% over age 55. The majority of the respondents are male (78.4%)
and more than half (61.88%) are married. The average income for convenience users was
$129,897.
Results of Logistic Regression: All Revolvers Compared to Convenience Users
The results for the logistic regression to compare all revolvers to convenience
users of credit cards are presented in Table 1.3. Odds ratios compare the magnitude of
the effect that each independent variable had on the dependent variable. This regression
distinguishes the difference between the two general groups of credit card users before
breaking down the revolving group into a specific type of revolvers.
23
Texas Tech University, Laura Ricaldi, August 2015
Table 1.3: Logistic Regression of the Likelihood of Being a Revolving User Compared to
a Convenience User in the 2010 SCF
Parameter
Estimate
-0.8003
p
Intercept
<0.0001
Human Capital
Financial Sophistication
Quintile 2
0.0676
0.5328
Quintile 3
0.9527
<0.0001
Quintile 4
1.0287
<0.0001
Quintile 5
1.0129
<0.0001
Education
Less than high school
0.2995
0.1055
Some college
0.0269
0.8166
College degree
-0.2936
0.0018
Mental Accounting/Precautionary Savings Motives
Saving for Emergencies
0.1034
0.1712
Saving for Unemployment
-0.034
0.8671
Saving for Illness
-0.2549
0.1553
Ability to borrow from friends/relatives
0.3112
0.0004
Self Employed
-0.3901
<0.0001
Number of Liquid Accounts
-0.2394
0.0024
Time Constraint/Self-control Factors
Number of children
0.0886
0.0168
Payment History
Behind
1.5356
<0.0001
Number of hours worked in a week
Working less than full time (1-39 hrs)
0.8278
<0.0001
Working full time (≤40 hrs)
0.7772
<0.0001
Credit attitude
Positive
0.0744
0.4158
Negative
-0.2794
0.0011
Bankruptcy history
1.0511
<0.0001
Lifecycle Factors
Age
35 to 55
0.0827
0.465
Over 55
-0.5555
<0.0001
Gender
Female
-0.0902
0.4626
Marital Status
Not married
0.0755
0.472
Income/$10,000
-0.0182
<0.0001
Source: 2010 Survey of Consumer Finances. Statistics derived from an unweighted analysis of one implicate.
Bolded values are significant at the 0.05 level.
24
Odds
Ratio
1.07
2.593
2.797
2.753
1.349
1.027
0.746
1.109
0.967
0.775
1.365
0.677
0.787
1.093
4.644
2.288
2.175
1.077
0.756
2.861
1.086
0.574
0.914
1.078
0.982
Texas Tech University, Laura Ricaldi, August 2015
Human Capital
Most of the human capital variables are significantly related to the likelihood of
being a revolving user. Compared to the group with the highest financial sophistication,
those in quintile 3 are 159.3% more likely to be revolvers, those in quintile 4 are 179.7%
more likely to be revolvers, and those in the least sophisticated quintile (quintile 5) are
175.3% more likely to be revolvers. Compared to the high school graduates group, those
who have a college degree are 25.4% less likely to be revolving users of credit.
Mental Accounting/Precautionary Savings Motives
The variables that had significance are the household’s ability to borrow from
friends/relatives, the household’s self-employment status, and the number of liquid
accounts. The households that are able to borrow from friends/relatives are 36.5% more
likely to be revolving credit card users. The households that are self-employed are 32.3%
less likely to be revolving credit card users. For every one unit increase in liquid
accounts the household is 21.3% less likely to be a revolving credit card user.
Time Constraint/Self-control Factors
Several of the variables in this concept are significantly related to the likelihood
of a household being a revolver. With every one unit increase in the number of children,
the household is 9.3% more likely to be a revolving credit card user. The households that
have been behind on payments are 364.4% more likely to be revolving users of credit
cards. Compared to those households that do not work, the households that work less
than full time are 128.8% more likely to be revolvers and those households that work full
time or more are 117.5% more likely to be revolvers. Compared to the group who
viewed credit as ambivalent, the respondents who view credit with a negative attitude are
25
Texas Tech University, Laura Ricaldi, August 2015
24.4% less likely to be revolving users. The households who reported filing for
bankruptcy in the past are 186.1% more likely to be revolving users.
Lifecycle Factors
Two of the lifecycle variables are significantly related to the likelihood of being a
revolving user of credit. Compared to those respondents under 35, the respondents over
age 55 are 42.6% less likely to be revolving users. For every $10,000 increase in income,
households are 1.8% less likely to be revolving users of credit cards.
Results of Logistic Regression: Solvent Revolvers Compared to Insolvent Revolvers
The results for the logistic regression for solvent revolvers compared to insolvent
revolvers of credit cards are presented in Table 1.4. Odds ratios compare the magnitude
of the effect that each independent variable had on the dependent variable. After
comparing the general groups of credit card users, it is important to evaluate the
differences between the two specific types of revolving credit card users, solvent and
insolvent revolvers.
26
Texas Tech University, Laura Ricaldi, August 2015
Table 1.4: Logistic Regression of the Likelihood of Being an Insolvent Revolving User
Compared to a Solvent Revolving User in the 2010 SCF
Parameter
Estimate
-0.289
p
Odds
Ratio
Intercept
0.2934
Human Capital
Financial Sophistication
Quintile 2
0.2615
0.1622
1.299
Quintile 3
0.9941
<0.0001
2.702
Quintile 4
0.8725
<0.0001
2.393
Quintile 5
0.7755
<0.0001
2.172
Education
Less than high school
-0.00628
0.9767
0.994
Some college
-0.2665
0.0633
0.766
College degree
0.0784
0.5282
1.082
Mental Accounting/Precautionary Savings Motives
Saving for Emergencies
0.0997
0.3288
1.105
Saving for Unemployment
-0.2436
0.3396
0.784
Saving for Illness
-0.3769
0.1424
0.686
Ability to borrow from friends/relatives
0.0317
0.7755
1.032
Self Employed
-0.1251
0.347
0.882
Number of Liquid Accounts
-1.0123
<0.0001
0.363
Time Constraint/Self-control Factors
Number of children
0.0668
0.1451
1.069
Payment History
Behind
0.7566
<0.0001
2.131
Number of hours worked in a week
Working less than full time (1-39 hrs)
0.0903
0.6176
1.094
Working full time (≤40 hrs)
0.1702
0.2526
1.186
Credit attitude
Positive
-0.0191
0.8765
0.981
Negative
0.0814
0.4886
1.085
Bankruptcy history
0.554
0.0001
1.74
Lifecycle Factors
Age
35 to 55
0.1425
0.3014
1.153
Over 55
-0.1162
0.474
0.89
Gender
Female
0.3046
0.0521
1.356
Marital Status
Not married
-0.2365
0.0841
0.789
Income/$10,000
-0.0179
0.0056
0.982
Source: 2010 Survey of Consumer Finances. Statistics derived from an unweighted analysis of one implicate.
Bolded values are significant at the 0.05 level.
27
Texas Tech University, Laura Ricaldi, August 2015
Human Capital
Only the financial sophistication quintiles are significantly related to the
likelihood of being an insolvent revolver. Compared to the group with the highest
financial sophistication, those in quintile 3 are 170.2% more likely to be insolvent
revolvers, those in quintile 4 are 139.3% more likely to be insolvent revolvers, and those
in the least sophisticated quintile (quintile 5) are 117.2% more likely to be insolvent
revolvers.
Mental Accounting/Precautionary Savings Motives
The number of liquid accounts is the only significant variable. For every one unit
increase in liquid accounts the household was 63.7% less likely to be insolvent revolvers.
Time Constraint/Self-control Factors
A household being behind on payments and having a bankruptcy history are the
only variables that are significantly related to being an insolvent revolving user of credit.
Those households that made late payments are 113.1% more likely to be insolvent
revolvers. The households that reported a bankruptcy are 74% more likely to be
insolvent revolvers.
Lifecycle Factors
Income was the only variable significantly related to the likelihood of being a
solvent revolver of credit. For every $10,000 increase in income, households were 1.8%
less likely to be an insolvent revolving user of credit cards.
Results of Logistic Regression: Solvent Revolvers Compared to Convenience Users
The results for the logistic regression for solvent revolving users of credit cards
compared to convenience users are presented in Table 1.5. Odds ratios compare the
28
Texas Tech University, Laura Ricaldi, August 2015
magnitude of the effect that each independent variable had on the dependent variable.
Since the solvent revolver has the ability to be a convenience user, it is important to
distinguish the differences between the two groups and explain why households remain
solvent revolvers.
29
Texas Tech University, Laura Ricaldi, August 2015
Table 1.5: Logistic Regression of the Likelihood of Being a Solvent Revolving User
Compared to a Convenience User in the 2010 SCF
Parameter
Estimate
-1.4273
p
Intercept
<0.0001
Human Capital
Financial Sophistication
Quintile 2
-0.0157
0.8981
Quintile 3
0.537
<0.0001
Quintile 4
0.6804
<0.0001
Quintile 5
0.6841
<0.0001
Education
Less than high school
0.2724
0.2089
Some college
0.082
0.537
College degree
-0.3756
0.0008
Mental Accounting/Precautionary Savings Motives
Saving for Emergencies
0.0893
0.3116
Saving for Unemployment
0.102
0.6529
Saving for Illness
-0.0711
0.732
Ability to borrow from friends/relatives
0.2524
0.014
Self Employed
-0.3505
0.0013
Number of Liquid Accounts
0.2476
0.0106
Time Constraint/Self-control Factors
Number of children
0.0363
0.4035
Payment History
Behind
1.173
<0.0001
Number of hours worked in a week
Working less than full time (1-39 hrs)
0.7983
<0.0001
Working full time (≤40 hrs)
0.7345
<0.0001
Credit attitude
Positive
0.0853
0.418
Negative
-0.3013
0.0029
Bankruptcy history
0.7006
<0.0001
Lifecycle Factors
Age
35 to 55
0.0184
0.8883
Over 55
-0.4867
0.0008
Gender
Female
-0.1949
0.1788
Marital Status
Not married
0.1607
0.1836
Income/$10,000
-0.0128
<0.0001
Source: 2010 Survey of Consumer Finances. Statistics derived from an unweighted analysis of one
implicate. Bolded values are significant at the 0.05 level.
30
Odds
Ratio
0.984
1.711
1.975
1.982
1.313
1.085
0.687
1.093
1.107
0.931
1.287
0.704
1.281
1.037
3.232
2.222
2.085
1.089
0.74
2.015
1.019
0.615
0.823
1.174
0.987
Texas Tech University, Laura Ricaldi, August 2015
Human Capital
Many of the human capital variables are significantly related to the likelihood of
being a solvent revolver user of credit cards. Compared to the group with the highest
financial sophistication, those in quintile 3 are 71.1% more likely to be solvent revolvers,
those in quintile 4 are 97.5% more likely to be solvent revolvers, and those in quintile 5
are 98.2% more likely to be solvent revolvers. Compared to the high school graduate
group, those who have a college degree are 31.3% less likely to be solvent revolvers of
credit cards.
Mental Accounting/Precautionary Savings Motives
Three variables have significance: the household’s ability to borrow from
friends/relatives, if the household is self-employed, and the number of liquid accounts.
The households that are able to borrow from friends/relatives are 28.7% more likely to be
solvent revolvers. The self-employed household is 29.6% less likely to be solvent
revolver. For every one unit increase in the number of liquid accounts, the household is
28.1% more likely to be a solvent revolver.
Time Constraint/Self-control Factors
Many of the variables in this concept are significantly related to likelihood of
being a solvent revolver. Those who have reported being behind on their past payments
are 223.2% more likely to be solvent credit card revolvers. The number of hours worked
per week is also significant. Compared to households that do not work, the households
who work less than full time are 122.2% more likely to be solvent revolvers and those
who work full time or more are 108.5% more likely to be solvent revolvers. Compared to
those who view credit as ambivalent, the respondents who view credit as negative are
31
Texas Tech University, Laura Ricaldi, August 2015
26.0% less likely to be solvent revolvers. The households who reported filing for
bankruptcy in the past are 101.5% more likely to be solvent revolvers.
Lifecycle Factors
Two of the lifecycle variables are significantly related to the likelihood of being
solvent credit card revolvers. Compared to those respondents under age 35, the
respondents over age 55 are 38.5% less likely to be solvent revolvers. For every $10,000
increase in income, households are 1.3% less likely to be solvent revolvers.
Summary and Implications
The research investigates the differences in behavioral characteristics of solvent
revolving users of credit cards and the effect of financial sophistication on credit card
behavior. There is still much research that needs to be done on the credit card debit
puzzle. The purpose of this study is to evaluate solvent revolving credit card users and
their financial sophistication compared to insolvent revolvers and convenience users of
credit cards. Based on the behavioral life-cycle hypothesis and human capital theory, the
results provide an extension of the current literature.
Human capital plays a role in the decision to display the credit card debt puzzle.
Regarding specific financial human capital, solvent credit card revolvers have a lower
level of financial sophistication than convenience users. In addition, solvent revolvers
are more financially sophisticated than insolvent revolving users. This finding supports
previous literature that solvent revolvers are not financially sophisticated (Hailassos &
Reiter, 2005; Bertaut, Haliassos, & Reiter, 2009). In regard to general human capital,
solvent credit card users are less likely to have college degrees compared to convenience
32
Texas Tech University, Laura Ricaldi, August 2015
users. Households should seek as much financial knowledge as possible or invest in the
assistance from a financial planner or counselor when making critical financial decisions.
Uncertainty and mental accounting does not seem to have a large impact on
households that exhibit solvent revolving. Hypothesis 2 suggests that solvent revolvers
would be more likely to display mental accounting and have higher precautionary savings
motives. Contrary to other studies however, there is no significant support for this
hypothesis (Bi & Hanna, 2006; Telyukova & Wright, 2008). Self-employed households
are less likely to be revolvers in general and less likely to be solvent revolvers. This
finding suggests that households that are self-employed use credit cards for convenience
purposes. Last, with every one unit increase in liquid accounts, households are more
likely to be solvent revolvers. This finding suggests that households that save liquid
assets in different accounts will choose not to pay off their credit card balance with those
earmarked funds. Those who use mental accounting to control the doer will have more
liquid accounts compared to convenience users who have the self-control to not spend
now.
The findings significantly support time constraint factors and the likelihood of
being a solvent revolver. Hypothesis 3 suggests that solvent revolvers would be more
likely to have more children, be less likely to pay their bills on time, work more hours
during the week, have a positive attitude toward credit, and have filed for bankruptcy in
the past. First, when compared with those who are on time with their loan payments, the
households that are behind are five times more likely to be solvent revolvers. Also, the
number of hours worked per week, both less than full time and full time, is significantly
33
Texas Tech University, Laura Ricaldi, August 2015
related to a household being a solvent revolver. These time constraint factors suggest
that as a household becomes busier, the more likely it is to not use liquid assets to pay off
the credit card balance, thus maintaining the status as a solvent revolver. Next, a
household’s self-control variables play a role in solvent revolving. Households that have
a positive attitude toward credit and have a history of bankruptcy are more likely to be
solvent revolvers.
Financial planners, counselors, and educators should help households understand
how a positive attitude toward credit can negatively impact financial decisions. A change
in behavior is necessary, especially since revolving credit card users believe it is okay to
spend now and pay later. Since time constraint and self-control factors have such an
impact on solvent revolving tendencies, financial planners, counselors and educators
should help clients identify debt management issues as well as increase their awareness
of the importance of making payments on time. The client would benefit from a change
in behavior to maximize utility and avoid making inefficient decisions.
In regards to the lifecycle factors, solvent revolvers are more likely to be younger
and have a lower household income. First, when compared to households under age 35,
those households over age 55 are less likely to be solvent revolvers. Last, with every
increase of income by $10,000 the likelihood of being a solvent revolver decreases by
only 1.3%. This finding suggests that solvent revolving credit card use is behavioral and
not based on income. This finding also supports previous research found by Gross and
Souleles (2002) in that higher credit card balances stem from behavioral factors.
34
Texas Tech University, Laura Ricaldi, August 2015
To facilitate the education process for solvent households, there are two steps.
One, planners, counselors, and educators must teach the client about behavioral biases
that exist when making financial decisions. Two, financial planners, counselors, and
educators should teach clients about the inefficiency associated with solvent revolving.
The first step is to educate the client about behavioral biases. Since behavioral
biases affect most individuals, it is important to know about them in order to diminish
some of the unfavorable effects. Financial planners, counselors and educators can help
the consumer develop strategies to control the inefficient behavior and implement
appropriate behavior. First, the consumer can set up automatic payments toward their
credit card balance. This would reduce the credit card balance while decreasing the
likelihood of consumers missing or being behind on their payments. Next, the planner,
counselor or educator should help the client set realistic goals for paying off the debt and
help the client understand their motives for saving.
The next step in helping solvent revolvers understand their inefficient behavior is
to educate them as to why the behavior is inefficient. Education about interest rates,
savings accounts, and the tradeoff between holding a balance on a credit card and paying
off the balance using savings is essential. Following these steps can allow financial
planners help clients build wealth instead of exhibiting inefficient behaviors.
35
Texas Tech University, Laura Ricaldi, August 2015
Appendix
Solvent revolvers have the knowledge and ability to be convenience users, but are
not. To investigate this further, convenience users and solvent revolvers in the top
quintile of financial sophistication are analyzed. The descriptive statistics for the two
groups is shown in Table 1.6.
36
Texas Tech University, Laura Ricaldi, August 2015
Table 1.6: Descriptive Statistics for the Top Quintile of Solvent Revolvers and
Convenience Credit Card Users.
Population Percent
Human Capital
Education
Less than high school
High school graduate
Some college
College degree
Mental Accounting/Precautionary Savings Motives
Saving for Emergencies
Saving for Unemployment
Saving for Illness
Ability to borrow from friends/relatives
Self Employed
Number of Liquid Accounts
Time Constraint Factors
Number of Children
Payment History
On time/ No payment
Behind
Number of hours worked in a week
Not working
Working less than full time (1-39 hrs)
Working full time (≤40 hrs)
Credit attitude
Positive
Ambivalent
Negative
Bankruptcy history
Lifecycle Factors
Age
Under 35
35 to 55
Over 55
Gender
Male
Female
Marital Status
Married
Not married
Income
Solvent Revolvers in
the Top Quintile of
Financial
Sophistication
n=194
9.07
Convenience Users in
the Top Quintile of
Financial Sophistication
0.72
16.98
14.69
67.61
0.69
13.52
12.76
73.03
43.21
2.69
4.37
16.15
19.43
0.7977 (0.0289)
37.27
2.45
4.27
12.32
19.84
0.7551
(0.0152)
0.7991
0.6472
(0.0764)
n=804
24.7
(0.0349)
90.62
9.38
97.73
2.27
11.66
11.18
77.16
24.1
12.13
63.77
22.64
52.73
24.63
6.87
24.23
46.22
29.54
3.38
22.33
46.58
31.09
14.92
38.98
46.1
86.12
13.88
87.51
12.49
66.66
71.09
33.34
28.91
$120,030.43
$185,341.18
(9,132.72)
(18,407.22)
Source: 2010 Survey of Consumer Finances. Statistics derived from a weighted analysis of one implicate.
(Mean (standard error) for continuous variables; column percent’s for categorical variables) (n=998)
37
Texas Tech University, Laura Ricaldi, August 2015
Demographically, solvent revolvers and convenience users tend to have the same
level of education, but more convenience users have college degrees. Solvent revolvers
save more for emergencies and have the ability to borrow from friends/relatives more
than convenience users. Solvent revolvers have more children, have a higher frequency
of being behind on payments, and more are working full time. More solvent revolvers
view credit as ambivalent, more declared bankruptcy in the past, and they tend to be
younger, unmarried, and have a lower income than convenience users.
A logistic regression comparing the two groups was run to analyze where the
differences between the two groups. The results for the logistic regression are presented
in Table 1.7. Odds ratios compare the magnitude of the effect that each independent
variable has on the dependent variable. Since the solvent revolver has the ability to be a
convenience user, it is important to distinguish the differences between the two groups
and explain why households remain solvent revolvers of credit cards.
38
Texas Tech University, Laura Ricaldi, August 2015
Table 1.7: Logistic Regression of the Likelihood of Being a Top Quintile Solvent
Revolver Compared to a Top Quintile Convenience User in the 2010 SCF
Parameter Estimate
p
Odds Ratio
Intercept
-1.5397
0.0018
Human Capital
Education
Less than high school
-0.4174
0.7117
0.659
Some college
-0.5273
0.1158
0.59
College degree
-0.5948
0.017
0.552
Mental Accounting/Precautionary Savings Motives
Saving for Emergencies
0.2073
0.2502
1.23
Saving for Unemployment
-0.3076
0.5716
0.735
Saving for Illness
0.2734
0.5274
1.314
Ability to borrow from friends/relatives
0.1711
0.4805
1.187
Self Employed
-0.361
0.0871
0.697
Number of Liquid Accounts
0.339
0.114
1.404
Time Constraint/Self-control Factors
Number of children
0.1031
0.2485
1.109
Payment History
Behind
1.2471
0.0009
3.48
Number of hours worked in a week
Working less than full time (1-39 hrs)
0.7455
0.0411
2.108
Working full time (≤40 hrs)
1.0223
0.0007
2.78
Credit attitude
Positive
-0.3379
0.1127
0.713
Negative
-0.3615
0.0862
0.697
Bankruptcy history
0.6278
0.1092
1.873
Lifecycle Factors
Age
35 to 55
-0.2943
0.2813
0.745
Over 55
-0.2887
0.3393
0.749
Gender
Female
-0.1226
0.7211
0.885
Marital Status
Not married
0.3793
0.1208
1.461
Income/$10,000
-0.0074
0.0015
0.993
Source: 2010 Survey of Consumer Finances. Statistics derived from an unweighted analysis of one implicate.
Bolded values are significant at the 0.05 level.
39
Texas Tech University, Laura Ricaldi, August 2015
Level of education, have a history of making late payments, the number of hours
spent working a week, and household income are the key factors that differentiate the
most financially sophisticated solvent revolvers and the most financially sophisticated
convenience users.
Although income is significant, with every $10,000 increase of income, the
household is only 0.7% more likely to be a convenience user. Thus, households are
solvent revolvers based on behavioral issues rather than liquidity issues (as stated by
Gross and Souleles, 2002).
40
Texas Tech University, Laura Ricaldi, August 2015
References
Becker, G. S. (1964). A theoretical and empirical analysis, with special reference to
education (First edition). New York: Columbia University Press for the NBER.
Bertaut, C. C., Haliassos, M., & Reiter, M. (2009). Credit card debt puzzles and debt
revolvers for self-control. Review of Finance, 1-36.
Bi, L. (2005). The influence of uncertainty and liquidity constraints on liquid asset
holdings of credit card revolvers. Retrieved from OhioLink ETD Center.
(osu1127153217).
Bi, L. & Hanna, S. D. (2006). Do financial planners serve the interests of their clients?
Use of financial planners, credit card balances and liquid assets. Consumer
Interests Annual, 52, 292-314.
Bucks, B. K., Kennickell, A. B., Mach, T. L., & Moore, K. B. (2009). Changes in U.S.
family finances from 2004 to 2007: Evidence from the Survey of Consumer
Finances. Federal Reserve Bulletin, 95, A1-A55.
Chien, Y.W. & DeVaney, S.A. (2001). The effects of credit attitude and socioeconomic
factors on credit card and installment debt. The Journal of Consumer Affairs,
35(1), 162-179.
Gross, D. B. & Souleles, N. S. (2002). Do liquidity constraints and interest rates matter
for consumer behavior? Evidence from credit card data. The Quarterly Journal of
Economics, February, 149-185.
Huston, S. J., Finke, M. S., & Smith, H. (2012). A financial sophistication proxy for the
Survey of Consumer Finances, Applied Economics Letters, 19:13, 1275-1278.
41
Texas Tech University, Laura Ricaldi, August 2015
Haliassos, M. & Reiter, M. (2005). Credit card debt puzzles, unpublished working paper,
CFS Working Paper 2005/26.
Kim, H. & DeVaney, S. A. (2001). The determinants of outstanding balances among
credit card revolvers. Financial Counseling and Planning, 12, 67-78.
Laibson, D., Repetto, A., & Tobacman, J. (2001). A debt puzzle. In P. Aghion,
R. Frydman, J. Stiglitz, & M. Woodford, (Eds.), Knowledge, information, and
expectations in modern economics: In honor of Edmund S. Phelps (pp. 228266). Princeton: Princeton University Press.
Lehnert, A. & Maki, D. (2007). Consumption, debt, and portfolio choice: Testing the
effects of bankruptcy law. In S. Agarwal & B. W. Ambrose (Eds.), Household
credit uses: Personal debt and mortgages (pp. 55-76). New York, NY: Palmgrave
Macmillan.
Lindamood, S., Hanna, S. D., & Bi, L. (2007). Using the Survey of Consumer Finances:
Some methodological considerations and issues. The Journal of Consumer
Affairs, 41, 195-222.
Rutherford, L. G. & DeVaney, S. A. (2009). Utilizing the theory of planned behavior to
understand convenience use of credit cards. Journal of Financial Counseling and
Planning, 20(2), 48-63.
Shefrin, H. M. & Thaler, R. H. (1988). The behavioral life-cycle hypothesis. Economic
Inquiry, 26, 609-643.
Telyukova, I. A. & Wright, R. (2008). A model of money and credit, with application to
the credit card debt puzzle. Review of Economic Studies, 75, 629-647.
42
Texas Tech University, Laura Ricaldi, August 2015
Thaler, R. H. (1999). Mental accounting matters. Journal of Behavioral Decision
Making, 12, 183-206.
43
Texas Tech University, Laura Ricaldi, August 2015
CHAPTER II
FINANCIAL LITERACY AND SHROUDED CREDIT CARD
REWARDS
Abstract
Credit card companies charge an interchange fee for each transaction, and almost
half of this fee is returned to consumers in the form of a reward or perk program. Among
credit card users who do not use cards for borrowing (convenience users), rewards are a
means to negotiate the implicit price of the interchange fee. Any consumer whose time
cost is less than the value of rebates should rationally choose a rewards card. Half of
convenience users do not have a rewards card.
Gabaix and Laibson (2006) propose a model of market segmentation where
sellers maximize profits by taking advantage of shrouded attributes through product
characteristic separation. Not all credit cards offer rewards, and many credit card
companies market non-salient credit card characteristics to appeal to naïve consumers
while offering lower-priced cards (net the rebate) to compete for more sophisticated
consumers.
It is hypothesized that naïve choice among credit card consumers may be
explained by financial literacy. When household characteristics such as education,
income and wealth are controlled, we find that respondents in the highest financial
literacy quintile were twice as likely to have a rewards card. These results imply that
consumers with financial literacy deficiencies risk consumer credit product exploitation.
44
Texas Tech University, Laura Ricaldi, August 2015
Keywords: Financial literacy, credit card rewards, shrouded attributes, Consumer
Finances Monthly Survey
45
Texas Tech University, Laura Ricaldi, August 2015
Introduction
Credit card purchases have become an increasingly significant payment method
for consumers in the United States. According to the Survey of Consumer Payment
Choice, 72.2% of consumers have at least one credit card with the average adopter having
3.7 cards (Foster, Meijer, Schuh, & Zabek, 2011). Of these 3.7 cards, two of the cards
have a rewards program and the other 1.8 cards do not (do not sum exactly to 3.7 due to
rounding and error) (Foster et al., 2011). Each year, Americans make approximately 21.6
billion credit card transactions that total an estimated value of $1.9 trillion (Foster et al.,
2011). The typical household uses their credit card approximately 119 times a year with
an average value of $89 per transaction for total annual expenditures of $10,600 (Foster
et al., 2011).
The majority of credit card users are convenience users who pay off their balance
each month (Ackerman, Fries, and Windle, 2012). Although credit cards are a
convenient payment method, the use of a card has a price. Credit card companies charge
the merchant interchange fees to accept credit card payments. The interchange fee is on
average 2% and is typically passed to the consumer through higher priced goods
(Hayashi, 2009). Beginning in January 2013, merchants are now allowed to pass the
merchant fees directly to consumers through a surcharge. The surcharge can range from
1.5% to 3% depending on the merchant. Merchant fees, including interchange fees, make
up a large source of revenue for credit card companies.
Although credit card companies charge an interchange fee for each transaction,
about 40% of this fee is returned to consumers in the form of a reward or perk program
46
Texas Tech University, Laura Ricaldi, August 2015
(Hayashi, 2009). Of the consumers who have at least one credit card, only 53.7% have a
credit card with a rewards program (Foster et al, 2011). The remaining households who
do not have or use a rewards credit card are essentially missing out on a rebate from the
credit card company. Typical rewards credit card programs offer households a small
percentage of purchases made through a rebate paid back in the form of points, frequent
flier miles, or cash. Households that do not use a rewards credit card program leave
money on the table by failing to collect a portion of the interchange fee assessed for using
a credit card.
Gabaix and Liabson (2006) propose a model of market segmentation where
sellers gain by shrouding product attributes and charging higher prices to naïve
consumers. Any consumer whose time cost from searching for a rewards card is less than
the value of rebates should rationally select a rewards card. Credit card companies
recognize that the market consists of sophisticated consumers who will shop around for
the largest rebate of interchange fees, and naïve consumers who are unaware of the
benefits of searching for a rebate. Rather than paying all consumers an interchange fee
rebate, card companies successfully withhold the entire interchange fee from the half of
convenience users who do not seek the rebate.
Companies marketing credit cards may emphasize less salient attributes such as
name recognition, personalized cards, and perceived ability to use cards at a greater
number of locations for less sophisticated consumers. They may simultaneously market
rebates through card rewards programs to more sophisticated consumers. In doing so, the
seller is able to extract greater producer surplus from less sophisticated cardholders who
47
Texas Tech University, Laura Ricaldi, August 2015
do not demand rebates while continuing to receive some producer surplus from more
sophisticated consumers who shop around for the most generous rewards program. For
example, rewards credit card advertisements are often placed in magazines that cater to a
more financially sophisticated clientele (Chase, 2012). Credit card advertisements in
financial magazines tend to have more financial information and explanations of the costs
and possible benefits associated with the card, while similar advertisements in general
interest magazines focus on less financially salient card attributes such as world-wide
acceptance (MasterCard, 2012) and the type of credit card people are most likely to
recommend (Discover, 2012).
Literature Review
Financial products, including credit cards, are often complex. Consumers with
greater product knowledge are better equipped to make an effective financial decision
(Lusardi, 2008). Financial literacy is a way to distinguish sophisticated from naïve
consumers within financial product markets. For example, Lusardi and Mitchell (2007)
find that less financially literate households do not plan for retirement, do not participate
in the stock market and have poor borrowing behavior. Allgood and Walstad (2011)
suggest that financial knowledge significantly impacts credit card behavior. Individuals
with low financial knowledge are more likely to be revolving users, pay the minimum
payment, pay late fees, and exceed the credit limit. Heidhues and Koszegi (2010) suggest
that naïve consumers over-borrow and pay penalties in the credit market. Naïve
consumers are less able to discern credit card attributes that affect the cost of usage
(DellaVigna & Malmendier, 2004; Eliaz & Spiegler, 2004), and card companies with
48
Texas Tech University, Laura Ricaldi, August 2015
more naïve customers are more profitable (Ausubel, 1991; DellaVigna & Malmendier,
2004). Simon, Smith, and West (2010) find that lower-income households are less likely
to use a rewards program, and that part-time and retired workers (with a lower time cost
to collect information) are more likely to participate in a rewards program.
An individual’s ability to make sound financial decisions depends on various
factors including salient product information and the knowledge or skills necessary to
analyze product attributes. Shrouded attribute theory and human capital theory provide a
framework that explains decision-making regarding financial products, namely rewards
credit card programs. Research questions are analyzed using a unique dataset that
contains both the largest financial literacy survey conducted and detailed information
about credit use including the use of rewards cards. The shrouded interchange fee rebate
captured through rewards cards creates an opportunity for consumer segmentation based
on financial literacy. The hypothesis is that naïve consumer choice among credit card
consumers is explained by financial literacy.
Method
Data and Sample
The data used are the current Consumer Finances Monthly (CFM) Survey, a
monthly survey conducted by the Consumer Finance Research Group at the Center for
Human Resource Research (CHRR) at The Ohio State University. The CFM survey
collects detailed information on credit card use, financial literacy, income, assets, and
behaviors. The CFM was used in the study because of the vast amount of detailed
information on credit card use and the comprehensive instrument designed specifically to
49
Texas Tech University, Laura Ricaldi, August 2015
measure financial literacy. The overall sample size of the CFM survey from 2005 to
March 2012 is 22,161. The data were limited to individuals who answered the recently
added financial literacy questions and those credit card users who paid off their credit
card balance every month for the past twelve months. Only convenience users were
chosen in the analysis in order to reduce other variables impacting the choice of rewards
card use. The censored sample size from the most current March 2012 data collection
was 2,870.
Model
The model is of consumer demand and producer supply of credit cards based on
human capital, credit score and credit history, and economic status:
Rewards Credit Card Use = α0 + β1*human_capital + β2*credit_score/history +
β3*economic_status + β4*demographic_characteristics + error.
Rewards card user is a binary variable equal to 1.0 if the household has a rewards
credit card. Human_capital consists of two variables that capture financial literacy and
education. Credit_score/history is a series of variables capturing the credit score of the
household, bankruptcy history, and the current interest rate of the main credit card.
Economic_status is a series of variables that captures the household’s income, net worth,
home ownership, business ownership and job stability. Demographic_characteristics is a
series of variables that captures a household’s age, gender, race and marital status.
50
Texas Tech University, Laura Ricaldi, August 2015
Dependent Variable
The dependent variable was constructed by separating two groups of convenience
users. The first group consisted of those who take advantage of a reward feature such as
cash back, flyer miles, discounts, or any other type of rewards. The second group
consisted of those who do not take advantage of a rewards program.
Independent Variables
Human Capital
Human capital is measured using two variables - financial literacy and education.
Financial decision-making requires an awareness of financial products and theory.
Knowledge and experience accumulated to improve decision making quality is often
referred to as human capital. Financial human capital is the stock of knowledge and
skills that improve one’s ability to make effective choices within financial markets
(Huston, 2010). Financial literacy is a measure of finance-specific human capital, and is
the end result of an investment in time and effort accumulating financial knowledge, as
well as experience from engaging in financial markets. Financial human capital gives
households the ability to make effective and efficient financial decisions.
Financial literacy is measured using an instrument developed by Huston (2010,
2012) and is included as a special module within the Consumer Finance Monthly (CFM)
survey. The objective financial literacy score is composed of 16 questions that measure
the respondent’s knowledge and ability in making financial decisions. The measure
directly measures the respondent’s understanding and ability to apply their financial
knowledge. The questions are presented in Table 3.7 in the Appendix. The 16-question
51
Texas Tech University, Laura Ricaldi, August 2015
score is the aggregate of the individual questions and is reported as an integer based on
the number of questions that were correctly answered. The financial literacy score is
broken down into five quintiles to measure differences between the groups of rewards
users and non-rewards users.
The education variable is a categorical variable based on the respondent’s highest
level of education received. Education was classified in four categories: less than high
school, high school, some college, and college.
Credit Score/History
The main determinants of credit card approval are credit score and credit history.
Since credit cards with the lowest rates and best reward programs require a high credit
score, the study accounts for these factors to control their impact on the dependent
variable. Since the CFM does not have a specified credit score variable, several proxies
are used to represent the factors that impact the credit score. Five main factors go into
the formulation of the FICO (Fair Isaac Corporation) score. The five factors are payment
history, debt amounts, length of credit history, credit mix and recent searches for credit.
Bankruptcy is an important payment history factor included in the FICO score.
Credit utilization, or the total amount of current debt divided by the household borrowing
capacity, impacts credit approval. Households with a large amount of debt may be less
likely to receive offers of additional credit cards. The total debt variable in the CFM
includes credit card debt, mortgage debt, home equity loan debt, student loan debt,
installment loan debt, bank loan debt, payday loan debt, and other loan debt. The
utilization factor also includes household credit card limits. The CFM asks respondents
52
Texas Tech University, Laura Ricaldi, August 2015
for the maximum amount they can borrow on their credit cards. Last, the number of
credit cards a household has is also a factor that impacts the utilization factor of the FICO
score.
Another factor in the FICO score is the credit mix, or the various types of debt
owed by the household that represents the household’s ability to pay off the different
forms of debt. The variable that best represents credit mix is the total number of debt
accounts excluding credit cards. The other types of accounts that are included are bank
loans, installment loans, payday loans, student loans, home equity loans, and other types
of loans.
Credit card interest rate represents the household’s ability to choose credit cards
with attractive features, or the household’s lack of choice (demand constraints) regarding
credit cards. The CFM asks all owners of a credit card to state the highest interest rate
associated with the credit card.
Economic Status
In order to account for credit card supply and demand differences based on
socioeconomic status, controls for income, net worth, homeownership, business
ownership and job stability are included. For the regression analysis, the income and net
worth variables were log transformed.
Homeownership is also included in the analysis. One of the typical questions on a
credit card application is whether you own or rent your home. Business ownership is also
included. Business owners may value card characteristics unrelated to rewards.
Ownership of physical business assets is used as a proxy for business ownership.
53
Texas Tech University, Laura Ricaldi, August 2015
The last variable for economic status is job stability. The preapproved credit
application has a question about job stability. The CFM has information regarding the
length of time spent working at the respondent's current job. The variable is categorized
into 5 categories: less than 1 year, 1-2 years, 3-4 years, 5-9 years, and 10 years and over.
Demographic Characteristics
Demographic characteristic variables include age, gender, race and marital status.
Age is classified into four groups: less than 50, ages 50 to 59, ages 60 to 69, and 70 and
older. Gender is categorized as male or female. Race is classified as white, black,
Hispanic, and other. Marital status is categorized as married, divorced/separated,
widowed and single/never married.
Analysis of Data
Descriptive statistics were performed to evaluate the characteristics of
convenience users who use a rewards program and those who do not. Chi-square
analyses were conducted to compare the relationship between rewards users and nonusers for the categorical variables. T-tests were conducted to compare the means of
rewards users and non-users for the continuous variables. A logistic regression was run to
establish whether there are any differences between convenience rewards card users and
those convenience users who do not use a rewards card.
Results
Descriptive Statistics
The descriptive statistics are reported as percentages, means, and standard
deviations for the total sample in Table 2.1.
54
Texas Tech University, Laura Ricaldi, August 2015
Table 2.1: Descriptive Statistics for Convenience Credit Card Users
Variable
Total Sample
Rewards Convenience
User (n=2053)
Non-Rewards Convenience
User (n=817)
Dependent Variable
Rewards Convenience User
71.53
28.47
Independent Variables
Human Capital
Financial Literacy
Quintile 1 (most literate) *
22.44
26.69
11.75
Quintile 2
20.98
22.70
16.65
Quintile 3
15.78
16.07
15.06
Quintile 4
19.30
18.75
20.69
Quintile 5 (least literate)
21.50
15.78
35.86
Education
Less than high school *
2.23
1.70
3.55
High School
18.85
15.88
26.32
Some College
22.26
21.19
24.97
College
56.66
61.23
45.17
Credit Score/ History
Total amount of debt
$68,612 (116,842)
$75,494 (120,604)
$51,238 (104,839)
Household credit limit
$28,217 (29,866)
$31,330 (27,280)
$19,577 (34,674)
Number of other debt accounts
0.47 (0.71)
0.4856 (0.6970)
0.4406 (0.7288)
Number of credit cards
2.76 (2.06)
3.0378 (2.1776)
2.0736 (1.5101)
Bankruptcy history
3.07
2.34
4.90
Interest rate on credit card
14.53% (6.60)
14.6889% (6.5968)
14.1198% (6.6050)
Economic Status
Income
$67,825 (104,236)
$76,168 (115,818)
$46,860 (62,084)
Net Worth
$602,466 (1,818,230)
$684,077 (2,086,389)
$397,391 (786,197)
Home ownership
93.92
95.51
89.93
Business ownership
7.60
8.72
4.77
Job stability (# of yrs in current job)
Less than 1 year *
55.05
50.56
66.34
1-2 years
3.48
3.75
2.82
3-4 years
4.70
5.41
2.94
5-9 years
9.09
10.18
6.36
10 or more years
27.67
30.10
21.54
Demographic Characteristics
Age
Less than 50 *
25.40
28.54
17.50
50-59
21.01
21.09
20.81
60-69
26.06
26.89
23.99
Greater than 70
27.53
23.48
37.70
Gender
Male
45.30
45.30
45.29
Female *
54.70
54.70
54.71
Race
White *
90.14
90.40
89.47
Black
1.85
1.36
3.06
Hispanic
2.54
2.34
3.06
Other
5.47
5.89
4.41
Marital Status
Married *
67.05
71.04
57.00
Divorced/ Separated
10.52
9.57
12.90
Widowed
13.42
11.08
19.29
Single/ Never married
9.01
8.30
10.81
Source: March 2012 Consumer Finances Monthly.
(Mean (standard deviation) for continuous variables; column percent's for categorical variables) (n = 2,870)
* reference group
55
Texas Tech University, Laura Ricaldi, August 2015
The majority of the convenience-user sample (71.5%) has a rewards card. More
than half (56.6%) of non-rewards users are in the lowest two financial literacy quintiles,
while a larger proportion of rewards users are in the highest financial literacy categories.
The majority of the rewards users have a college degree (61.23%) compared to 45% of
non-rewards card users. Rewards card users have 47% more debt than non-rewards
users, and have a higher credit limit and roughly one more credit card on average. Twice
the percentage of non-rewards cards users has declared bankruptcy.
Rewards card users have greater financial resources than non-rewards card users.
The income of rewards card users is about $30,000 higher, and their net worth is 72%
greater than non-rewards users. A greater percentage of rewards cards users are home
and business owners. Rewards users tend to be evenly split among age categories, while
the majority (62%) of non-rewards users is age 60 and older. There are no gender
differences, but roughly twice as many non-rewards card users are black. A higher
proportion of non-rewards card users are widowed, divorced or single.
T-tests results are presented in Table 2.2. The means for the rewards users are
statistically higher than non-users for the following variables: the total amount of debt,
household credit limit, number of credit cards, income, and net worth. The only two
continuous variables that do not have statistically significant mean differences are the
number of debt accounts and the interest rate on the credit cards.
56
Texas Tech University, Laura Ricaldi, August 2015
Table 2.2: T-Tests of Differences between Rewards users and Non-Rewards Users
Variable
Rewards
Convenience User
Non-Rewards
Convenience
User
$75,493.91
(120,604.30)
$31,330.04
(27,280.29)
0.4856
(0.6970)
3.0378
(2.1776)
14.6889%
(6.5968)
$51,238.01
(104,838.70)
$19,576.79
(34,674.42)
0.4406
(0.7288)
2.0736
(1.5101)
14.1198%
(6.6050)
$76,168.00
(115,817.70)
$684,076.92
(2,086,389.02)
$46,860.10
(62,083.54)
$397,390.91
(786,196.60)
p value
Sig.
<0.0001
***
<0.0001
***
Credit Score/ History
Total amount of debt
Household credit limit
Number of other debt accounts
Number of credit cards
Interest rate on credit card
0.1310
<0.0001
***
0.1385
Economic Status
Income
Net Worth
*p .05, ** p< .01, *** p < 0.001
Source: March 2012 Consumer Finances Monthly.
Mean (standard deviation) reported for continuous variables
57
<0.0001
***
0.0001
***
Texas Tech University, Laura Ricaldi, August 2015
The relationship between the rewards convenience users and non-rewards
convenience users is analyzed using a chi-square analysis. The majority of the variables
are statistically significant except gender. See Table 2.3 for the summary of chi-square
results. Rewards users tend to have a higher level of financial literacy and education than
the non-rewards users. Rewards users also tend to have a lower level of bankruptcy
history. Rewards users exhibit a higher level of homeownership and business ownership.
There is also a difference in job stability between the two groups. Rewards users tend to
have a longer tenure in their current job compared to non-rewards users. Non-rewards
users are older, black, non-Hispanic and unmarried compared to rewards users.
58
Texas Tech University, Laura Ricaldi, August 2015
Table 2.3: Chi-square Analyses of the Differences in Characteristics of Households
between Rewards Convenience Users and Non-Rewards Convenience Users.
Variable
Rewards
Convenience
User (n=2053)
Non-Rewards
Convenience User
(n=817)
26.69
22.70
16.07
18.75
15.78
11.75
16.65
15.06
20.69
35.86
1.70
15.88
21.19
61.23
3.55
26.32
24.97
45.17
2.34
Human Capital
Financial Literacy
Quintile 1 (most literate)
Quintile 2
Quintile 3
Quintile 4
Quintile 5 (least literate)
Education
Less than high school
High School
Some College
College
Credit Score/ History
Bankruptcy history
Economic Status
Home ownership
Business ownership
Job stability (# of yrs in current job)
Less than 1 year
1-2 years
3-4 years
5-9 years
10 or more years
Demographic Characteristics
Age
Less than 50
50-59
60-69
Greater than 70
Gender
Male
Female
Race
White
Black
Hispanic
Other
Marital Status
Married
Divorced/ Separated
Widowed
Single/ Never married
*p .05, ** p< .01, *** p < 0.001
Source: March 2012 Consumer Finances Monthly
Column percent's reported for categorical variables
p value
Sig.
<0.0001
***
<0.0001
***
4.90
0.0003
***
95.51
8.72
89.93
4.77
<0.0001
0.0003
<0.0001
***
***
***
50.56
3.75
5.41
10.18
30.10
66.34
2.82
2.94
6.36
21.54
<0.0001
***
28.54
21.09
26.89
23.48
17.50
20.81
23.99
37.70
45.30
54.70
45.29
54.71
90.40
1.36
2.34
5.89
89.47
3.06
3.06
4.41
71.04
9.57
11.08
8.30
57.00
12.90
19.29
10.81
0.9954
59
0.0053
**
<0.0001
***
Texas Tech University, Laura Ricaldi, August 2015
Results of Logistic Regression on Type of Reward Usage
The results of the logistic regression that predicts rewards care use are presented
in Table 2.4. Odds ratios compare the magnitude of the effect that each independent
variable had on the dependent variable.
60
Texas Tech University, Laura Ricaldi, August 2015
Table 2.4: Logistic Regression of the Likelihood of being a Rewards User Compared to a
Non-Rewards User
Parameter
Estimate
-0.2669
p value
Intercept
0.6962
Human Capital
Financial Literacy
Quintile 2
-0.1248
0.5832
Quintile 3
-0.4736
0.0482
Quintile 4
-0.3609
0.1188
Quintile 5 (least literate)
-0.6860
0.0056
Education
High School
0.2608
0.5627
Some College
0.1258
0.7794
College
0.2864
0.5198
Credit Score/ History
Total amount of debt
0.0000
0.8128
Household credit limit
0.0000
0.1336
Number of other debt accounts
-0.2659
0.0111
Number of credit cards
0.2162
<0.0001
Bankruptcy history
-0.7572
0.0136
Interest rate on credit card
0.0166
0.1320
Economic Status
Income
0.0000
0.4479
Log Income
-0.0316
0.2625
Net Worth
0.0000
0.2529
Log Net Worth
0.0381
0.0922
Home ownership
0.7534
0.0081
Business ownership
0.0312
0.9081
Job stability (# of yrs in current job)
1-2 years
0.5759
0.1409
3-4 years
0.3630
0.2677
5-9 years
0.2235
0.4006
10 or more years
0.2189
0.2317
Demographic Characteristics
Age
50-59
-0.5029
0.0169
60-69
-0.5059
0.0230
Greater than 70
-0.4127
0.1190
Gender
Male
-0.1659
0.2589
Race
Black
-0.0583
0.8962
Hispanic
-0.2799
0.4400
Other
0.5898
0.1109
Marital Status
Divorced/ Separated
-0.4290
0.0458
Widowed
-0.1497
0.5456
Single/ Never married
-0.1577
0.5320
Source: March 2012 Consumer Finances Monthly. Bolded values are significant at the 0.05 level.
61
Odds Ratio
0.8830
0.6230
0.6970
0.5040
1.2980
1.1340
1.3320
1.0000
1.0000
0.7670
1.2410
0.4690
1.0170
1.0000
0.9690
1.0000
1.0390
2.1240
1.0320
1.7790
1.4380
1.2500
1.2450
0.6050
0.6030
0.6620
0.8470
0.9430
0.7560
1.8040
0.6510
0.8610
0.8540
Texas Tech University, Laura Ricaldi, August 2015
Compared to respondents with the highest level of financial literacy, those in
middle and lowest literacy quintiles are less likely to be rewards card users by 37.77%
and 49.60%, respectively. The number of debt accounts and bankruptcy history are
negatively related to rewards card ownership, while the number of credit cards is
positively related. For every one unit increase in the number of debt accounts,
households are 23.30% less likely to be rewards card users. For every one unit increase in
the number of credit cards, households are 24.10% more likely to be rewards card users.
Last, households that declare bankruptcy are 53.10% less likely to be rewards card users.
The only statistically significant economic status variable is homeownership.
Households who own a home are 112.40% more likely to be rewards card users. Of the
demographic independent variables, respondents between ages 50-59 and 60-69 are
39.50% and 39.70% less likely to be reward card users compared to those less than 50
years old, respectively. The previously divorced are less likely to have a rewards card
than respondents who had not divorced. Households that are divorced are 34.90% less
likely to be rewards card users.
Summary and Implications
Credit card rewards provide a rebate on purchases made using a credit card. For
convenience users who do not carry a credit card balance, rewards should be the primary
salient credit card choice characteristic. Many convenience users, however, do not use a
rewards card. It is hypothesized that credit card companies use rewards as a way to
differentiate more price-sensitive sophisticated consumers from naïve consumers who
select cards based on non-salient characteristics. Using a comprehensive financial
62
Texas Tech University, Laura Ricaldi, August 2015
literacy instrument available in a national dataset, the study tests whether financial
literacy is related to naïve consumer product choice. Respondents in the lowest financial
literacy quintile are 50% less likely to use a rewards credit card than the most financially
literate respondents. The results provide evidence that producers of financial products
are successfully able to exploit deficiencies in financial literacy to sell higher-priced
consumer credit products.
The ability to segment the credit card market presents one of the best examples of
the cross-subsidy suggested by Gabaix and Laibson (2006) where producers shroud an
attribute, in this case the rebate of interchange fees provided by a credit card reward, in
order to extract greater rents from unsophisticated consumers while maintaining a more
competitive market for sophisticated consumers. The findings show that non-reward card
users are indeed less financially knowledgeable than rewards card users. These users are
allowing credit card companies to keep the full amount of the interchange fee while more
financially literate consumers demand a rebate. The welfare implications of this transfer
are sobering since non-rewards users are more likely to be a black or Hispanic, lower
income and wealth, and older. The rebate in essence creates a wealth transfer from lower
socioeconomic status households to credit card companies that no party other than a
government entity has an incentive to change since only more sophisticated consumers
and card companies are aware of the loss.
In order to avoid the wealth transfer, some policy implications are necessary.
Recently, the European Union passed new rules limiting the interchange fee and fee
amounts credit card companies can charge. Legislation such as these new rules would
63
Texas Tech University, Laura Ricaldi, August 2015
reduce the amount of rebate returned to sophisticated consumers while limiting the
charges for naïve consumers. Another policy implication is similar to the Credit CARD
Act of 2009 requiring disclosure on credit card statements. By forcing credit card
companies to disclose reward card information to all consumers, the naïve have the
information to make more knowledgeable decisions regarding a rewards program.
64
Texas Tech University, Laura Ricaldi, August 2015
Appendix
Table 2.5: Financial Literacy Questionnaire from the Consumer Finances Monthly
(CFM) Survey
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Net worth is equal to: (K1-BAS)
1 Total assets
2 Total assets plus liabilities
3 Total assets minus liabilities
If your assets increase by $5,000 and your liabilities decrease by $3,000, your net worth would (A1-BAS)
1 Increase by $2,000
2 Increase by $8,000
3 Increase by $3,000
Which bank account is likely to pay the highest interest rate on money saved? (K2-BAS)
1 Savings account
2 Six month CD or certificate of deposit
3 Three year CD
Savings accounts and money market accounts are most appropriate for (A2-BAS)
1 Long-term investments like retirement
2 Emergency funds and short-term goals
3 Earning a high rate of return
To reduce the total finance costs paid over the life of an auto loan, you should choose a loan with the (K3-BOR)
1 Lowest monthly payment
2 Longest repayment term
3 Shortest repayment
If you always pay the full balance on your credit card, which of the following is least important? (A3-BOR)
1 Annual interest rate
2 Annual fees
3 Line of credit 18
On which type of loan is interest never tax deductible? (K4-BOR)
1 A home equity loan
2 An adjustable rate mortgage
3 A personal vehicle loan
Which type of mortgage would allow a first-time home buyer to qualify for the highest loan amount? (A4-BOR)
1 Fixed-rate mortgage
2 Adjustable-rate mortgage
3 Reverse mortgage
The benefit of owning investments that are diversified is that it (K5-INV)
1 Reduces risk
2 Increases return
3 Reduces tax liability
A young investor willing to take moderate risk for above-average growth would be most interested in: (A5-INV)
1 Treasury bills
2 Money market mutual funds
3 Balanced stock funds
The main advantage of a 401(k) plan is that it: (K6-INV)
1 Provides a high rate of return with little risk
2 Allows you to shelter retirement savings from taxation
3 Provides a well-diversified mix of investment assets
To ensure that some of your retirement savings will not be subject to income tax upon withdrawal, you would contribute to: (A6-INV)
1 A Traditional IRA or Individual Retirement Account
2 A Roth IRA
3 A 401(k) plan
If you have an insurance policy with a higher deductible, the premiums will be (K7-INS)
1 Higher
2 Lower
3 The same
Which of the following types of insurance is most important for single workers without children? (A7-INS)
1 Life insurance
2 Disability income insurance
3 Dental insurance
Which policy provides the most coverage at the lowest cost for a young family? (K8-INS)
1 Renewable term life
2 Whole life
3 Universal life
Which household would typically have the greatest life insurance needs? (A8-INS)
1 A middle-class retired couple
2 A middle-aged working couple with children in college
3 A single-earner family with two young children in pre-school
65
Texas Tech University, Laura Ricaldi, August 2015
References
Ackerman, R., Fries, G. & Windle, R. A. (2012). Changes in US family finances from
2007 to 2010: Evidence from the Survey of Consumer Finances. Federal Reserve
Bulletin.
Allgood, S. and Walstad, W. B. (2011). The effects of perceived and actual financial
knowledge on credit card behavior. Networks Financial Institute at Indiana State
University Working Paper 2011-WP-15.
Ausubel, L. (1991). The failure of competition in the credit card market. The American
Economic Review, 81(1): 50-81.
Chase. (2012, January/February). [Advertisement for Chase Blueprint Credit Card].
Money, 54-58.
DellaVigna, S. & Malmendier, U. (2004). Contract design and self-control: Theory and
evidence. Quarterly Journal of Economics, 119(2): 353-402.
Discover. (2012, June). [Advertisement for Discover Credit Card]. Parents, 21.
Eliaz, K. & Spiegler, R. (2004). Contracting with diversely naïve agents. Working paper,
The Foerder Institute for Economic Research and The Sackler Institute of
Economic Studies.
Foster, K., Meijer, E., Schuh, S. & Zabek, M. A. (2011). The 2009 Survey of Consumer
Payment Choice. Public Policy Discussion Papaer 11-1, Federal Reserve Bank of
Boston.
Gabaix, X. & Laibson, D. (2006). Shrouded attributes, consumer myopia, and
information suppression in competitive markets. Quarterly Journal of
Economics, 121(2): 505-540.
66
Texas Tech University, Laura Ricaldi, August 2015
Hayashi, F. (2009). Do U.S. consumers really benefit from payment card rewards?
Economic Review, Q1: 37-63.
Heidhues, P. & Koszegi, B. (2010). Exploiting naivete about self-control in the credit
market. American Economic Review, 100(5): 2279-2303.
Huston, S. J. (2010). Measuring financial literacy. Journal of Consumer Affairs,
44(Summer): 296-316.
Huston, S. J. (2012). Financial literacy and the cost of borrowing. International Journal
of Consumer Studies, 36(5): 566-572.
Lusardi, A. (2008). Financial literacy: An essential tool for informed consumer choice?
Working paper, National Bureau of Economic Research.
Lusardi, A. & Mitchell, O. (2007). Financial literacy and retirement planning: New
evidence from the Rand American Life Panel.
MasterCard. (2012, June 18). [Advertisement for MasterCard World Elite Credit Card].
Bloomberg Business Week, 8.
Simon, J., Smith, K. & West, T. (2010). Price incentives and consumer payment
behavior. Journal of Banking and Finance, 34: 1759-1772.
67
Texas Tech University, Laura Ricaldi, August 2015
CHAPTER III
DEBIT OR CREDIT: THE IMPACT OF MYOPIA AND CREDIT
ATTITUDE ON PAYMENT CHOICE
Abstract
The choice between debit or credit use is dependent on factors like the features
associated with the type of card and the attributes of the individual. Attributes include
myopia, credit attitude, human capital, life-cycle factors like generational cohort and
gender, and a household’s finances. It is hypothesized that myopic individuals may
explain debit card use. Based on multiple years of the Survey of Consumer Finances, the
study discovered that respondents who are myopic are both debit card users as well as
revolving credit card users. The results are mixed regarding credit attitude, but
convenience credit card users have a below average attitude toward using credit. These
results imply that myopic households should use a debit card to avoid that costs
associated with revolving credit card balances.
Key words: Credit cards, debit cards, myopia, credit attitude, Survey of Consumer
Finances, factor analysis
68
Texas Tech University, Laura Ricaldi, August 2015
Introduction
Households have many choices when it comes to paying for goods and services.
The main payment methods include cash, check, credit cards, and debit cards. Through
an extensive study, Herbst-Murphy (2010) finds that with check usage declining,
households tend to mainly use a card to pay for goods and services. The use of debit
cards has increased significantly over the past several years (Herbst-Murphy, 2010).
Stango and Zinman (2008) find that more consumers use debit cards than credit cards.
The authors also find that consumers tend to “specialize in their payment choices”
suggesting that consumers use a card consistently for all purchases rather than switching
between payment methods. Stango and Zinman (2008) posit the classification as debit
users or credit users can be informative. King and King (2005) provide evidence that
households will always be better off choosing a credit card over a debit card. The authors
suggest households should solely use a credit card because of the benefits they provide.
In addition, Zinman (2009) argues that debit cards offer no benefit for a neoclassical
consumer.
Households select a preferred payment method based on various factors. First,
the positive and negative attributes of the method will influence the choice. The benefits
of using a credit card are the ability to float purchases, the perks of a rewards program,
and the potential to build credit. Adverse characteristics of a credit card are the high
interest rate for revolving balances, high fees associated with late payments, and fees for
carrying a balance over the limit.
69
Texas Tech University, Laura Ricaldi, August 2015
Debit cards also have both positive and negative characteristics. First, they are
widely accepted and are available for consumers that do not have access to credit. Debit
cards are also used as a means to control spending for some consumers (Borzekowski,
Kiser & Ahmed, 2008; Sprenger & Stavins, 2010; Rogers & Dopico, 2010). However,
debit cards can lead to expensive overdraft fees (Stango & Zinman, 2008; Parrish, 2009)
and debit card usage is the most likely cause of overdraft fees. Hayashi and Stavins
(2012) suggest that those who use debit will face higher debit fees. Debit cards are also
risky if the card is used in situations where the card is not kept in sight, like at a
restaurant (Rogers & Dopico, 2010) as well as being susceptible to fraud and identity
theft. Another risk factor is if the debit card is stolen. Since debit cards are directly
linked to a checking account, funds from withdrawals or purchases are immediately
removed from the account. If the consumer reports the card stolen, the consumer may
only be liable for $50, but the bank has up to 20 days to restore the money in the account.
The delay can be inconvenient to some consumers that need the funds quickly.
The characteristics of households who mainly use credit cards differ from the
characteristics of those who mainly use debit cards. The households that primarily use
debit cards are younger, have a lower income, have a lower FICO score, and are less
educated (King & King, 2005; Hayashi & Stavins, 2012; Oliver Wyman, 2013).
King and King (2005) and Zinman (2009) suggest that households would be
better off using a credit card rather than a debit card for their main payment method. The
household’s preference for debit or credit is influenced by other factors such as myopia,
credit attitude, human capital, and other life-cycle characteristics. The purpose of this
70
Texas Tech University, Laura Ricaldi, August 2015
paper is to evaluate if households who are myopic use debit cards as a spending
constraint.
Literature Review
A household’s choice of payment method depends on several factors including
myopia, credit attitude, financial sophistication, past experiences, gender, age, and
characteristics of the method. Since the household tends to use the same payment
method across all purchases (Stango & Zinman, 2008; Herbst-Murphy, 2010), examining
the rationale behind the chosen payment method can provide insight on a household’s
financial behaviors.
Myopia
In the seminal article on myopia, Strotz (1955) proposes individuals have
conflicting preferences that influence their ability to optimize future behavior. Thaler
and Shefrin (1981) suggest the planner is farsighted and focused on overall lifetime
utility while the doer is myopic and focused only on the current period. Behavioral
factors, such as myopia, often influence an individual’s preference. The choice of
payment method is influenced by a consumer’s level of myopia. Myopia is defined as the
inability to defer gratification. Many consumers choose to use a debit card for budgeting
and control purposes (Sprenger & Stavins, 2010). Although some studies suggest that
individuals use debit cards as a control mechanism, other fail to find significant results.
For example, Borzekowski, Kiser, and Ahmed (2008) find that only 5.8% of the sample
uses a debit card in order to restrain their spending. Fusaro’s (2013) findings do not
support that debit cards are used to control spending.
71
Texas Tech University, Laura Ricaldi, August 2015
Other studies find that consumers use debit cards to control spending, especially
with credit cards. Sprenger and Stavins (2010) and Lee, Abdul-Rahman and Kim (2007)
find that households who revolve their credit card balance are more likely to adopt and
use a debit card. Stango and Zinman (2008) suggest that revolving status should be the
single greatest difference between payment methods. According to theory, convenience
users should prefer a credit card since these users can float the loan, avoid costly fees and
other benefits associated with credit cards. Revolving users should prefer a debit card
since revolving a balance often involves high fees and interest payments.
Credit Attitude
Past literature suggests that households that have a positive attitude toward credit
tend to use credit cards more freely. Bertaut and Haliassos (2001) use variables such as
households willing to buy furs and jewelry on credit and willing to use credit for daily
living to measure credit attitude and self-control. Hayhoe, Leach, Turner, Bruin, and
Lawrence (2000) find that attitude toward credit influences college student’s credit card
purchasing behavior. Students with a negative attitude toward credit will feel sorry for
making a purchase compared to those with a positive attitude. In addition, the authors
find that the more positive the attitude toward credit, the more likely the student is to
revolve a credit card balance (Hayhoe et. al, 2000). However, Durkin (2000) finds that
those who have more credit cards and a revolving balance tend to have a negative attitude
toward credit cards.
72
Texas Tech University, Laura Ricaldi, August 2015
Human Capital
Financial decisions, including payment choices, are influenced by an individual’s
stock of knowledge, experience, skills and values. The stock of knowledge and
experience is defined as human capital (Becker, 1962). Past financial experiences and
knowledge impact the household’s ability to make effective financial decisions in the
future. Human capital can be improved upon with education and experience. Financial
human capital, also known as financial sophistication, has the potential to impact a
household’s financial management decisions and behaviors.
There is conflicting literature regarding financial sophistication and card use.
First, Stango and Zinman (2008) find that heavy credit card users are more sophisticated
and debit card users are less sophisticated. However, Rogers and Dopico (2010) find that
heavy credit users have higher levels of debt, which is inconsistent with the literature
regarding a more sophisticated user.
King and King (2005) suggest there is a learning curve in regards to debit card
use. Their results show that debit card use increases with education until the individual
obtains a college degree. The households with a college degree are no more likely to use
a debit card than those with only a high school degree. The authors postulate that those
with a college education are more financially sophisticated and therefore know the
benefits of using a credit card over a debit card.
73
Texas Tech University, Laura Ricaldi, August 2015
Demographic Characteristics of Debit Card Users
Gender influences payment choice among credit and debit cards. Herbst-Murphy
(2010) finds that debit users are female, younger, have children, and are married. In
contrast, the typical credit card user is a middle-aged male with a higher level of income
(Herbst-Murphy, 2010). Gender is also significant in payment method choice.
Borzekowski, Kiser, and Ahmed (2008) find that gender is significant in debit card
choice. Ching and Hayashi (2010) find that females are less likely to hold either debit or
credit reward cards.
Generational cohorts are typically defined by birth year and are usually shaped by
significant historical events and popular culture (Coomes, 2004). Cohorts differ
significantly by the behaviors individuals’ exhibit and cohort effects are life-long and
shape the individuals in each generation (Berkowitz & Schewe, 2011). For example,
individuals in the Greatest Generation (born between 1900 and 1925) were adults during
the Great Depression, which had a large impact on subsequent spending and saving
behaviors.
Age cohort also plays a role in the choice of debit or credit. It is especially
important to study the younger generations since the Millennial generation is the most
diverse and largest cohort (de Bassa Scheresberg, Lusardi & Yakoboski, 2014). Rogers
and Dopico (2010) suggest that the Grand Recession influenced younger generational
cohorts to use debit cards more frequently to control overspending. Several studies find
that consumers who heavily use debit cards are younger (King & King, 2005; Rogers &
Dopico, 2010; Hayashi & Stavins, 2012). In addition, Borzekowski and Kiser (2008)
74
Texas Tech University, Laura Ricaldi, August 2015
find that younger individuals choose debit more often than older individuals, but the age
difference will diminish over a 10-year period since the technology is becoming more
prevalent. King and King (2005) find that age is the largest determining factor of debit
card use since younger households are more willing to try new technologies. Rysman
(2009) comes to the same conclusion in that using a debit card tends to be a new
technology and more likely to be used by younger individuals.
Framework and Concepts
The theory proposed for the study is behavioral life-cycle hypothesis and human
capital theory. Unlike the King and King study (2005), the current study suggests that
using a credit card for all purchases is not appropriate for all households especially those
who are myopic and have a positive attitude toward credit. Individuals are influenced by
behavioral biases and therefore should use a debit card as a control mechanism. The
behavioral life-cycle hypothesis suggests that in order to maximize utility, the household
will shift resources in periods where the marginal utility of consumption is relatively low
to periods where it is relatively high (Shefrin & Thaler, 1988). The hypothesis suggests
there are three factors (self-control, mental accounting, and framing) that influence the
households desire to shift resources. These three factors are what differentiate the
behavioral life-cycle hypothesis from other life-cycle hypotheses. The underlying
principle of the behavioral life-cycle hypothesis is that households have a dual preference
framework, the planner and the doer (Shefrin & Thaler, 1988). The planner is the
rational thinker that focuses on long-term decisions. The doer is short-term, impulsive
and succumbs to temptation to consume in the current period. In the dual-self model
75
Texas Tech University, Laura Ricaldi, August 2015
from Thaler and Shefrin (1981) the planner is farsighted and focuses on overall lifetime
utility. The doer is nearsighted, myopic, selfish, and always concerned with the current
period. In order to minimize current consumption of the doer, the planner implements
rules and savings devices. For example, using a debit card instead of a credit card will
prevent the doer from spending more than the amount in the bank account.
Human capital is a function of goods, services, time and the individual’s current
stock of human capital. Human capital is either endowed (i.e., strength, intelligence, and
cognitive ability) or acquired. Acquired human capital includes education and financial
management behaviors. Individuals can improve their of human capital by obtaining a
college education, by gaining experience with financial behaviors or by taking financial
courses to improve their ability to understand financial decisions. Although education
and financial experience are both acquired human capital, they are different because
education is considered general human capital where financial experience is a specific
form of human capital.
A household’s level of human capital impacts the ability to make efficient
financial decisions. Households with a higher level of financial human capital (i.e., more
financial experience and sophistication) have the potential to improve the ability to make
effective and efficient financial decisions. Households that perform more of the positive
financial behaviors (i.e., better financial managers) will have a higher level of human
capital.
From a theoretical viewpoint, individuals should choose the payment method that
is the least expensive and provides the most reward. Human capital influences the choice
76
Texas Tech University, Laura Ricaldi, August 2015
of payment method by limiting the available options as well as the method the household
ultimately chooses.
Using the framework described above, four concepts were formed: myopia, credit
attitude, human capital, life-cycle factors, and finances. Debit card use is a function of
these four concepts. The concepts serve as controls to help explain why households use
debit cards over credit cards. It is hypothesized that debit card users will be myopic, have
a negative attitude toward credit, and have a lower level of financial sophistication than
households that use credit cards.
Method
Data and Sample
The data used were from the 1998-2010 Survey of Consumer Finances (SCF), a
triennial survey, sponsored by the Federal Reserve Board and collected by the National
Organization for Research at the University of Chicago. The SCF collects detailed
information on the finances of U.S. households. The combined number of observations
in the public data set is 24,165 households. The sample size is limited to households that
either only use a debit card or only use a credit card. The limited sample size is 10,828
respondents. Each survey year contains five implicates to deal with missing and
incomplete data; only the first implicate was used from each survey year in this study.
Descriptive statistics and regression analyses were performed to explore the data relating
to the research question.
77
Texas Tech University, Laura Ricaldi, August 2015
Model
The model is based on the theoretical framework including constructs for credit
attitude, myopia, human capital, life-cycle factors, and finances:
Debit Card Use = α0 + β1*credit attitude + β2*myopia + β3*human capital +
β4*life-cycle factors + β5*finances + error.
Debit card user is a binary variable equal to 1.0 if the household uses a debit card.
Credit attitude and myopia are both factors described in the appendix. Human capital
consists of two variables that capture financial sophistication and education. Life-cycle
factors are a series of variables capturing the various stages in the life-cycle. Variables
include generational cohort, gender, race, homeownership, marital status and the presence
of children. Finances include variables that describe the household’s level of income and
net worth.
Dependent Variable
The dependent variable is households that use either a debit card, a credit card
with a revolving balance (revolving credit card users), or a credit card without a balance
(convenience credit card users). The SCF defines a debit card as a payment card that
automatically deducts the amount of the purchase from the money in an account (Board
of Governors of the Federal Reserve System, 2012). Logistic regression analyses were
run to establish if there is a difference between debit card users and credit card users. See
Table 3.1 for the measurement of the dependent variable.
78
Texas Tech University, Laura Ricaldi, August 2015
Independent Variables
Based on the theoretical framework, the concepts for the model are credit attitude,
myopia, human capital, life-cycle factors and finances. Independent variables are used as
proxies to form the concepts. The measurement of the independent variables is in Table
3.1.
79
Texas Tech University, Laura Ricaldi, August 2015
Table 3.1: Coding of Variables Used in the Study
Variable
Measurement
Debit Card User (X7582)
1 if yes, 0 otherwise
Revolving Credit Card Users (X410, X427, X413, X421, X430, X424)
1 if yes, 0 otherwise
Convenience Credit Card Users (X410, X427, X413, X421, X430, X424)
1 if yes, 0 otherwise
Independent Variables
Credit Attitude (X401, X402, X403, X404)
Above the Mean Credit Attitude
1 if yes, 0 otherwise
Below the Mean Credit Attitude **
1 if yes, 0 otherwise
Myopia (X7510, X7508, X3015, X3016, X7380, X7395)
Above the Mean Myopia
1 if yes, 0 otherwise
Below the Mean Myopia **
1 if yes, 0 otherwise
Human Capital
Financial Sophistication (X3913, X3014, X414, X6525)
Quintile 1 (most sophisticated)
1 if yes, 0 otherwise
Quintile 2
1 if yes, 0 otherwise
Quintile 3 **
1 if yes, 0 otherwise
Quintile 4
1 if yes, 0 otherwise
Quintile 5 (least sophisticated)
1 if yes, 0 otherwise
Education (X5901, X6101)
Less than high school
1 if yes, 0 otherwise
High School
1 if yes, 0 otherwise
Some College
1 if yes, 0 otherwise
College **
1 if yes, 0 otherwise
Life-Cycle Factors
Generational Cohort (Age) (X14, X19)
Greatest Generation (1900-1925)
1 if yes, 0 otherwise
Silent Generation (1926-1945)
1 if yes, 0 otherwise
Baby Boomer Generation (1946-1964)
1 if yes, 0 otherwise
Generation X (1965-1980)
1 if yes, 0 otherwise
Millennial Generation (1981-2000) **
1 if yes, 0 otherwise
Gender (X8021, X103)
Male **
1 if yes, 0 otherwise
Female
1 if yes, 0 otherwise
Married (X7372, X7018)
1 if yes, 0 otherwise
Have children (X108, X114, X120, X126, X132, X202, X208, X214, X220, X226)
1 if yes, 0 otherwise
Finances
Income (X5729)
Less than $25,000 **
1 if yes, 0 otherwise
$25,001-$50,000
1 if yes, 0 otherwise
$50,001-$75,000
1 if yes, 0 otherwise
$75,001-$100,000
1 if yes, 0 otherwise
Greater than $100,000
1 if yes, 0 otherwise
Net Worth
Less than $10,000
1 if yes, 0 otherwise
$10,001-$50,000
1 if yes, 0 otherwise
$50,001-$100,000
1 if yes, 0 otherwise
$100,001-$150,000
1 if yes, 0 otherwise
Greater than $150,000 **
1 if yes, 0 otherwise
From the Survey of Consumer Finances 1998-2010 (year 2010 is reference category)
X8000 was used for age, gender, and education. Net Worth provided by Federal Reserve. ** Indicates reference
category.
80
Texas Tech University, Laura Ricaldi, August 2015
Credit Attitude
The credit attitude construct explains the individual’s willingness to use credit to
fund consumption. The SCF has many variables that describe attitude toward spending
on credit. The appendix of this paper provides the results of a principal component
analysis and a factor analysis to derive the underlying latent concept for credit attitude.
The variables that make up the construct are has a positive attitude toward credit, willing
to use credit to fund a vacation, willing to use credit to purchase furs or jewelry, and
willing to use credit to cover living expenses if income were cut. A factor score is then
calculated two groups are formed to see the magnitude of change between the levels of
credit attitude. The two groups are: individuals above the mean and the individuals below
the mean. See Appendix I for the complete analysis.
Myopia
Myopia is the individual’s inability to defer gratification to a future time period.
Based on a dual-self framework, households consist of a planner and a doer. The planner
is rational and is concerned with lifetime utility and will impose rules to limit the doer.
The doer is myopic, thinks of only one period and succumbs to temptation easily. The
planner must exert rules over the doer to maximize utility in the long run. The SCF does
not have a measure of myopia and the appendix of this paper provides the results of a
principal component analysis and a factor analysis to derive the underlying latent concept
for myopia.
The variables that make up the construct are spending exceeded income over the
past year, savings habits are described as not saving and spending more or as much as
81
Texas Tech University, Laura Ricaldi, August 2015
income and smokes. A factor score is then calculated and two groups are formed – above
the mean myopia and below the mean myopia. See Appendix I for the complete analysis.
Human Capital
The human capital concept explains the household’s potential to make effective
decisions. The human capital construct consists of two variables – one that measures
specific financial human capital and one that is a general measure of human capital.
Financial sophistication is a latent concept developed through a factor analysis by
Huston, Finke and Smith (2012). The score includes four variables: stock ownership
(within or outside of tax sheltered accounts), willingness to accept at least some
investment risk, not revolving more than 50% of credit card limit, and the level of
understanding of personal finance. The indirect measure of financial sophistication is
based on observed and self-reported behavioral variables instead of directly measured
financial literacy. A direct measure of financial knowledge, ability and confidence is
more effective at determining financial sophistication, but the SCF does not contain such
variables. Financial sophistication is categorized into five quintiles to measure the
magnitude between levels of sophistication. Quintile 1 is the most sophisticated group
and quintile 5 is the least sophisticated. Quintile 3 is the reference group for the
regression analysis.
Education is a general measure of human capital and is often used to proxy the
household’s ability to make effective decisions. Education is categorized as less than
high school, high school, some college, and college. College is the reference category for
the regression analysis.
82
Texas Tech University, Laura Ricaldi, August 2015
Life-Cycle Factors
The lifecycle factors concept is included since households make decisions based
on being in different stages of the lifecycle. The variables that make up this concept
represent why the household will choose a preferred card type. Based on the framework,
the life-cycle stage influences the purchasing method of the household. The life-cycle
concept consists of five variables including generational cohort, gender, marital status
and the presence of children.
Generational cohort impacts a household’s choice between debit and credit.
Older generations are less willing to accept new technology. Since these households
already have credit card, they will be less willing to change to a debit card. The
generational cohort variable is broken into five categories by birth year. The generational
cohorts are: Greatest generation born between 1900-1925, Silent generation born between
1926-1945, Baby Boomer generation born between 1946-1964, Generation X born
between 1964-1980, and Millennial generation born between 1981-2000. Each
generational cohort is dummy coded as 1 or 0. The Millennial generation is the reference
category for the regression analysis.
Next, gender, marital status and the presence of children are included in the lifecycle factor construct. The head of household’s gender is recorded as either male or
female. Both are dummy variables with 1 as yes and 0 otherwise. Male is the reference
category for the regression analysis. Marital status is coded as married or not married.
The married category is the reference group. The presence of children are coded as either
83
Texas Tech University, Laura Ricaldi, August 2015
having children under 18 in the household or no children under 18. The group of
households with children under 18 is the reference group.
Finances
The finances concept includes two variables, income and net worth. Income and
net worth influence financial decisions and the ability to make purchases. Both variables
are inflated to 2010 dollars. The income variable is broken into five categories and
dummy variables are created for each category. The five categories are: less than
$25,000, $25,001-$50,000, $50,001-$75,000, $75,001-$100,000, and greater than
$100,000. The net worth variable was created using the SAS code provided by the
Federal Reserve Board. The net worth variable is broken into five categories and dummy
variables are created for each category. The five categories are: less than $10,000,
$10,001-$50,000, $50,001-$100,000, $100,001-$150,000, and greater than $150,000.
Analysis of Data
Descriptive statistics were conducted to evaluate the characteristics of the
household. In order to generalize the results back to the entire population of the United
States, the descriptive statistics were weighted using the weight variable provided by the
Federal Reserve Board (Lindamood, Hanna & Bi, 2007). Since the dependent variable is
binary, logistic regression was used to predict the likelihood of the dependent variable
occurring given the set of independent variables. The regression analyses were not
weighted (Lindamood, Hanna & Bi, 2007).
84
Texas Tech University, Laura Ricaldi, August 2015
Results
Descriptive Statistics
Since the descriptive statistics are weighted, the reported percentages and means
represent all households for the United States. See Table 3.2 for a summary of the
descriptive statistics.
85
Texas Tech University, Laura Ricaldi, August 2015
Table 3.2: Population Percentage/Means for the Variables
Variable
Total Sample
Debit Card User
n = 10,828
100%
n = 2,758
25.47%
Revolving Credit
Card User
n = 2,496
23.05%
Convenience Credit
Card User
n = 5,574
51.48%
Credit Attitude
Above Mean Attitude
32.7
38.02
42.03
20.5
Below Mean Attitude
67.3
61.98
57.97
79.5
Myopia
Above Mean Myopia
35.22
51.19
37.14
20.13
Below Mean Myopia
64.78
48.81
62.86
79.87
Groups
AM Attitude/ AM Myopia
16.12
24.57
19.57
6.12
AM Attitude/ BM Myopia
16.58
13.44
22.47
14.38
BM Attitude/ AM Myopia
19.10
26.61
17.57
14.00
BM Attitude/ BM Myopia
48.20
35.37
40.40
65.50
Human Capital
Financial Sophistication
Quintile 1 (most
19.95
0.93
16.62
38.82
sophisticated)
Quintile 2
20.01
16.27
20.89
22.46
Quintile 3
19.99
18.52
22.99
18.76
Quintile 4
20.00
27.03
21.60
12.73
Quintile 5 (least
20.04
37.26
17.90
7.23
sophisticated)
Education
Less than high school
14.02
19.63
13.40
9.79
High School
32.62
37.08
34.78
27.05
Some College
25.23
29.58
27.08
20.02
College
28.13
13.71
24.74
43.14
Life-Cycle Factors
Generational Cohort
51.6460
42.1249
50.6731
60.5145
(Age)
Greatest Generation
8.38
1.81
4.62
17.04
(1900-1925)
Silent Generation (192627.39
10.53
28.83
40.50
1945)
Baby Boomer Generation
37.14
34.03
46.49
32.05
(1946-1964)
Generation X (1965-1980)
22.15
40.51
18.60
9.52
Millennial Generation
4.94
13.12
1.46
0.88
(1981-2000)
Gender
Male
46.03
39.79
42.72
54.04
Female
53.97
60.21
57.28
45.96
Married
50.82
35.52
56.29
59.28
Have children
40.42
52.94
44.50
26.44
Finances
Income
$78,206.70
$36,694.24
$64,093.48
$125,017.07
Net Worth
$607,523.87
$84,551.35
$326,315.89
$1,282,552.85
From the Survey of Consumer Finances 1998-2010. Statistics derived from the weighted analysis of one implicate.
Income & net worth are in 2010 dollars.
The mean is presented for age, income and net worth; column percent's are included for all categorical variables.
86
Texas Tech University, Laura Ricaldi, August 2015
Based on the total sample column, the majority of the sample falls below the
mean in both credit attitude and myopia. However, the various card users are much
different. Debit card users mainly fall below the mean in credit attitude and are even in
the myopia construct. The credit card users tell a different story. The majority of credit
card users, both revolving and convenience users, fall below the mean in credit attitude
and myopia.
The groups section of Table 3.2 is a combination of the two groups of attitude and
two groups of myopia. The four groups are: above the mean attitude/above the mean
myopia, above the mean attitude/below the mean myopia, below the mean attitude/above
the mean myopia, and below the mean attitude/below the mean myopia.
About 48% of the total sample fall in the below the mean attitude/below the mean
myopia group. The debit card users and revolving credit card users are split among all
four groups. The majority of the convenience credit card users are in the below the mean
attitude/below the mean myopia group.
Results of Logistic Regression: Debit Card Users Compared to All Credit Card
Users
The results for the logistic regression to compare debit card users to all credit card
users are presented in Table 3.3. Odds ratios were used to compare the magnitude of the
effect that each independent variable had on the dependent variable. This regression
distinguishes the levels of credit attitude and myopia among debit users and all credit
card users.
87
Texas Tech University, Laura Ricaldi, August 2015
Table 3.3: Logistic Regression of the Likelihood of Being a Debit Card User Compared
to All Credit Card Users in the SCF between 1998-2010 (n=10,828)
Parameter
Estimate
-2.5391
Intercept
Credit Attitude
Above Mean Attitude
-0.3211
Myopia
Above Mean Myopia
0.1926
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
-3.222
Quintile 2
0.0622
Quintile 4
0.7322
Quintile 5 (least sophisticated)
1.389
Education
Less than high school
0.1266
High School
0.1172
Some College
0.3231
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
-2.794
Silent Generation (1926-1945)
-2.3263
Baby Boomer Generation (1946-1964)
-1.1499
Generation X (1965-1980)
-0.3828
Female
-0.1058
Not Married
0.3964
Have children
0.3303
Finances
Income
$25,001-$50,000
0.0813
$50,001-$75,000
0.2954
$75,001-$100,000
0.1329
Greater than $100,000
-0.8359
Net Worth
Less than $10,000
1.3374
$10,001-$50,000
1.0815
$50,001-$100,000
0.6734
$100,001-$150,000
0.5154
Survey Year
Year 2001
0.6405
Year 2004
1.4881
Year 2007
2.171
Year 2010
3.4116
R2
0.4775
Adjusted R2
0.7037
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
88
p
Odds Ratio
<0.0001
***
<0.0001
0.725
***
0.0112
1.212
**
<0.0001
0.5649
<0.0001
<0.0001
0.04
1.064
2.08
4.011
***
0.3268
0.2608
0.0025
1.135
1.124
1.381
<0.0001
<0.0001
<0.0001
0.0429
0.1525
<0.0001
<0.0001
0.061
0.098
0.317
0.682
0.9
1.486
1.391
0.3747
0.0136
0.4163
<0.0001
1.085
1.344
1.142
0.433
***
<0.0001
<0.0001
<0.0001
0.0006
3.809
2.949
1.961
1.674
***
***
***
***
<0.0001
<0.0001
<0.0001
<0.0001
1.897
4.429
8.767
30.314
***
***
***
***
***
***
***
***
***
***
**
***
***
**
Texas Tech University, Laura Ricaldi, August 2015
Compared to the below mean credit attitude group, the households with above the
mean credit attitude are 27.5% less likely to use a debit card. Compared to the below
mean myopia group, the households with above the mean myopia are 21.2% more likely
to use a debit card. Financial sophistication is significantly related to debit card use.
Compared to the middle level of financial sophistication, the individuals in the highest
level of financial sophistication are 96% less likely to use debit cards. However, quintiles
4 and 5 are 108% and 301.1% more likely to use debit cards, respectively. Compared to
households with a college degree, the households with some college are 38.1% more
likely to use debit cards.
All of the variables in the life-cycle construct are significant, except gender. First,
when compared to the Millennial Generation, the other generations are less likely to use a
debit card. The households that are not married and have children are more likely to use
debit cards. The households with income between $50,001 and $75,000 are 34.4% more
likely to use a debit card; however, the households with an income greater than $100,000,
are 56.7% less likely to use debit cards when compared to households with an income
less than $25,000. Net worth also has an impact on debit card use. When compared to
households with a net worth greater than $150,000, the households with a net worth
greater than $10,000 are more likely to use debit cards. Last, survey year is statistically
significant in the study. When compared to year 1998, households in years 2001-2010
are more likely to use a debit card.
89
Texas Tech University, Laura Ricaldi, August 2015
Results of Logistic Regression: Debit Card Users Compared to Revolving Credit
Card Users
The results for the logistic regression to compare debit card users to revolving
credit card users are presented in Table 3.4. Odds ratios were used to compare the
magnitude of the effect that each independent variable had on the dependent variable.
This regression distinguishes the levels of credit attitude and myopia between debit card
users and revolving credit card users.
90
Texas Tech University, Laura Ricaldi, August 2015
Table 3.4: Logistic Regression of the Likelihood of Being a Debit Card User Compared
to a Revolving Credit Card User in the SCF between 1998-2010 (n=5,254)
Parameter
Estimate
-1.0258
Intercept
Credit Attitude
Above Mean Attitude
-0.5228
Myopia
Above Mean Myopia
0.0588
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
-2.8831
Quintile 2
-0.047
Quintile 4
0.4626
Quintile 5 (least sophisticated)
0.9851
Education
Less than high school
-0.0136
High School
-0.0681
Some College
0.064
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
-2.092
Silent Generation (1926-1945)
-2.1268
Baby Boomer Generation (1946-1964)
-1.2816
Generation X (1965-1980)
-0.5334
Female
-0.2033
Not Married
0.4334
Have children
0.2488
Finances
Income
$25,001-$50,000
-0.0136
$50,001-$75,000
0.1043
$75,001-$100,000
-0.0578
Greater than $100,000
-0.6099
Net Worth
Less than $10,000
0.8919
$10,001-$50,000
0.6557
$50,001-$100,000
0.2933
$100,001-$150,000
0.2064
Survey Year
Year 2001
0.6075
Year 2004
1.4435
Year 2007
2.103
Year 2010
3.468
R2
0.4436
Adjusted R2
0.592
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
91
p
Odds Ratio
0.0007
***
<0.0001
0.593
***
0.478
1.061
<0.0001
0.7012
<0.0001
<0.0001
0.056
0.954
1.588
2.678
0.9238
0.5569
0.5901
0.986
0.934
1.066
<0.0001
<0.0001
<0.0001
0.0171
0.0124
<0.0001
0.004
0.123
0.119
0.278
0.587
0.816
1.542
1.283
***
***
***
**
**
***
***
0.8924
0.4283
0.7449
0.0009
0.986
1.11
0.944
0.543
***
<0.0001
<0.0001
0.0382
0.2105
2.44
1.926
1.341
1.229
***
***
**
<0.0001
<0.0001
<0.0001
<0.0001
1.836
4.235
8.191
32.073
***
***
***
***
***
***
***
Texas Tech University, Laura Ricaldi, August 2015
Compared to the below mean credit attitude group, the households with above the
mean credit attitude are 40.7% less likely to use a debit card. Financial sophistication is
significantly related to debit card use. Compared to the middle level of financial
sophistication, the individuals in the highest level of financial sophistication are 94.4%
less likely to use debit cards. However, quintiles 4 and 5 are 58.8% and 167.8% more
likely to use debit cards, respectively. All of the variables in the life-cycle construct are
significant. First, when compared to the Millennial Generation, the other generations are
less likely to use a debit card. Females are 18.4% less likely to use a debit card. The
households that are not married and have children are more likely to use debit cards. The
households with income greater than $100,000 are 45.7% less likely to use a debit card
when compared to households with an income less than $25,000. Net worth also has an
impact on debit card use. When compared to households with a net worth greater than
$150,000, the households with a net worth between $10,001 and $100,000 are more
likely to use debit cards. Last, survey year is statistically significant in the study. When
compared to year 1998, households in years 2001-2010 are more likely to use a debit
card.
Results of Logistic Regression: Debit Card Users Compared to Convenience Credit
Card Users
The results for the logistic regression to compare households that are debit card
users to convenience credit card users are presented in Table 3.5. Odds ratios were used
to compare the magnitude of the effect that each independent variable had on the
dependent variable. To further the story among credit and debit use, the difference
92
Texas Tech University, Laura Ricaldi, August 2015
among convenience users and debit users is important. Households with above average
myopia as well as a positive attitude toward credit should use debit cards rather than
credit cards. This regression analyzes the difference in myopia and credit attitude among
debit card users and convenience credit card users.
93
Texas Tech University, Laura Ricaldi, August 2015
Table 3.5: Logistic Regression of the Likelihood of Being a Debit Card User Compared
to a Convenience Credit Card User in the SCF between 1998-2010 (n=8,332)
Parameter
Estimate
-2.5342
Intercept
Credit Attitude
Above Mean Attitude
-0.0658
Myopia
Above Mean Myopia
0.4626
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
-3.543
Quintile 2
0.1019
Quintile 4
1.0212
Quintile 5 (least sophisticated)
2.1013
Education
Less than high school
0.2194
High School
0.2469
Some College
0.5977
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
-3.4673
Silent Generation (1926-1945)
-2.5773
Baby Boomer Generation (1946-1964)
-1.0718
Generation X (1965-1980)
-0.1984
Female
0.0573
Not Married
0.3229
Have children
0.6019
Finances
Income
$25,001-$50,000
0.3302
$50,001-$75,000
0.5973
$75,001-$100,000
0.2913
Greater than $100,000
-0.8622
Net Worth
Less than $10,000
2.0625
$10,001-$50,000
1.5855
$50,001-$100,000
0.9698
$100,001-$150,000
0.8281
Survey Year
Year 2001
0.6838
Year 2004
1.3839
Year 2007
2.0887
Year 2010
3.2294
R2
0.6012
Adjusted R2
0.836
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
94
p
Odds Ratio
<0.0001
***
0.5523
0.936
<0.0001
1.588
***
<0.0001
0.4559
<0.0001
<0.0001
0.029
1.107
2.776
8.177
***
0.2153
0.0702
<0.0001
1.245
1.28
1.818
<0.0001
<0.0001
0.0003
0.5075
0.5686
0.0038
<0.0001
0.031
0.076
0.342
0.82
1.059
1.381
1.826
***
***
***
0.0141
0.0003
0.1726
<0.0001
1.391
1.817
1.338
0.422
**
***
<0.0001
<0.0001
<0.0001
<0.0001
7.866
4.882
2.637
2.289
***
***
***
***
<0.0001
<0.0001
<0.0001
<0.0001
1.981
3.99
8.074
25.264
***
***
***
***
***
***
*
***
***
***
***
Texas Tech University, Laura Ricaldi, August 2015
Compared to the below mean credit attitude group, the households with above the
mean myopia are 58.8% more likely to use a debit card. Financial sophistication is
significantly related to debit card use. Compared to the middle level of financial
sophistication, the individuals in the highest level of financial sophistication are 97.1%
less likely to use debit cards. However, quintiles 4 and 5 are 177.6% and 717.7% more
likely to use debit cards, respectively. Compared to the households with a college
degree, the households with a high school degree and some college are more likely to use
debit cards. Most of the variables in the life-cycle construct are significant. First, when
compared to the Millennial Generation, the other generations are less likely to use a debit
card except Generation X. The households that are not married and have children are
more likely to use debit cards.
Compared to households with income less than $25,000, the households with
income between $25,000 and $75,000 are more likely to use a debit card. However,
households with income greater than $100,000 are less likely to use a debit card. Net
worth also has an impact on debit card use. When compared to households with a net
worth greater than $150,000, the households with a net worth between zero and $150,000
are more likely to use debit cards. Last, survey year is statistically significant in the
study. When compared to year 1998, households in years 2001-2010 are more likely to
use a debit card.
Summary and Implications
The choice between debit and credit is dependent on various factors. The
research investigated myopia and credit attitude among debit and credit card users.
95
Texas Tech University, Laura Ricaldi, August 2015
Human capital, life-cycle factors, and finances also play an important role in payment
choice.
Based on the results between debit card users and credit card users (both
revolving and convenience users), individuals with an above the mean attitude are less
likely to use a debit card. The individuals with an above the mean myopia are more
likely to use a debit card. This indicates that households that are more myopic are
effectively using debit cards to control spending. Financial sophistication is also
significantly related to debit card use. Households with more financial sophistication are
less likely to use a debit card and the households in the lower levels of financial
sophistication are more likely to use a debit card. Life cycle factors are also significant.
Older households are less likely to use debit cards suggesting that older households do
not adopt debit cards since they are a relatively new technology.
The results between the debit card users and the revolving credit card users
indicate that credit attitude is a determining factor of card use. The households with
above average credit attitude are less likely to use debit cards signifying the households
that are less willing to spend on credit generally use a debit card. This indicates that
myopic revolving credit card users with a positive attitude toward credit would benefit
from using a debit card. A similar logistic regression in Table 3.9 in Appendix II shows
the results when credit attitude and myopia are formed into four groups. When compared
to the below attitude/below myopia group, both above attitude/above myopia and above
attitude/below myopia groups are less likely to use a debit card. These results indicate
that above average (positive) credit attitude leads to less debit use compared to revolving
96
Texas Tech University, Laura Ricaldi, August 2015
credit card use. Financial sophistication is also significantly related to debit card use.
Those with the highest level of financial sophistication are less likely to use debit while
those in the lowest levels of sophistication are more likely to use a debit card. This
indicates that households have the financial knowledge to make informed decisions
regarding the costs of debit and revolving a credit card but still pay the expenses
associated with revolving the balance. Life cycle factors are also significant: generational
cohort, gender, marital status, and the presence of children influence debit card and
revolving credit card use.
Based on the results between the debit card users and the convenience credit card
users, above average myopia is statistically significant. The households with above
average myopia are more likely to use a debit card. This indicates that households that
are present oriented are wisely using debit card instead of using a credit card. Table 3.10
in Appendix II provides the results for the logistic regression that includes the four
groups of attitude/myopia mix. Compared to the below attitude/below myopia group,
both the above attitude/above myopia and below attitude/below myopia group are more
likely to use a debit card. These results indicate that above average myopia leads to more
debit use compared to convenience credit card use. Financial sophistication is also
significant to debit card use. The households in the highest level of financial
sophistication are less likely to use a debit card while the households in the lowest levels
of financial sophistication are more likely to use a debit card. The convenience credit
users have the financial knowledge to make informed decisions and use credit cards as a
purchasing tool rather than a borrowing tool.
97
Texas Tech University, Laura Ricaldi, August 2015
To further analyze the story of how myopia and credit attitude impact payment
choice a logistic regression comparing revolving credit card users to convenience credit
card users is presented in Appendix II Table 3.11 and Table 3.12. The households with
above average credit attitude and myopia are more likely to be revolving users of credit.
Financial sophistication is also significant in that the higher levels of financial
sophistication (Quintile 1 and Quintile 2) are less likely to be revolving credit card users.
However, the individuals in quintile 5 are more likely to be revolving credit card users.
These findings suggest that individuals with above average attitude toward credit and
myopia should use debit cards to avoid the costs of revolving a credit card.
The findings are consistent with the theoretical framework. These results provide
evidence to support the main hypothesis. Households that use debit cards exhibit an
above average level of myopia but are mixed among attitude toward credit. Based on the
results of this study, households that use debit cards tend to have lower human capital and
therefore do not choose to take advantage of the benefits of a credit card like those
households with higher financial human capital. The results show that older generations
do not accept new technology like a debit card or there is an inertia factor where
households will use the same payment method repeatedly. Households with less
education and who are younger are more likely to use a debit card that is consistent with
current literature (King & King, 2005; Hayashi & Stavins, 2012; Oliver Wyman, 2013).
Similar to King and King (2005), many financial planners, educators, and
therapists suggest that households should use credit in order to reap the benefits
associated with credit cards including the float, rewards programs, and the potential build
98
Texas Tech University, Laura Ricaldi, August 2015
a credit score. However, the results show that households with that are myopic should
use debit cards in order to control spending.
In order to know if consumers are better off using a debit card compared to a
credit card, the costs and fees of using both cards are essential to know. Stango and
Zinman (2009) use a unique data set that includes various fee information on households
that have credit card and debit cards. The researchers find that the median household
pays approximately $21 in credit card fees each month. These fees include interest paid
for revolving a balance, late fees, over-limit fees, cash advance fees, annual fees, and
other fees. The median household pays approximately $9 in checking account costs that
include overdraft fees, ATM fees, and other fees (Stango & Zinman, 2009). Based on
the findings from Stango and Zinman (2009), households that are myopic and have a
positive attitude toward credit would be better off switching to debit cards rather than
revolving a credit card balance.
Financial planners, educators, and therapists play an important role in assisting
households determine the most appropriate payment method. In addition to the fees and
costs associated with using debit or credit, an individual’s level of myopia should be one
factor to determine the payment method. Planners, educators, and therapists can help
households increase their financial human capital and therefore improve the household’s
financial decision-making ability.
The results provide an extension of the current literature, but more research is
necessary in the area of payment choice. The behavioral aspects of payment choice are
particularly necessary especially if households use several payment methods. By
99
Texas Tech University, Laura Ricaldi, August 2015
investigating credit attitude, myopia and the other factors that influence the choice
between debit and credit, researchers can determine why households choose payment
methods do not provide as much economic benefit as other payment options.
100
Texas Tech University, Laura Ricaldi, August 2015
Appendix I
Principal Components Analysis and Factor Analysis
An individual’s general tendency to succumb to needs and wants and overall
credit attitude will impact their preferred payment method. Based on a dual-self
framework, a household consists of a planner and doer. The planner is rational and
concerned with lifetime utility and will impose rules to limit the doer. The doer thinks of
only one period and succumbs to temptation easily. A household’s level of myopia, or
the inability to defer gratification, is influential in the payment choice.
In order to get a complete look on the behavioral factors that make up payment
choice, it is necessary to create a measure of credit attitude and myopia. The Survey of
Consumer Finances (SCF) has several variables that measure an individual’s attitude
toward borrowing with credit as well as variables that measure an individual’s level of
myopia.
Previous literature often use various factors to measure an individual’s level of
myopia. For example, smoking is often used as a measure of preference for the present
(Robb, Huston, & Finke, 2008). Other sources use willing to borrow money to purchase a
fur coat of jewelry, willing to borrow to cover living expenses, and savings habits as
measures of a household’s ability to control consumption (Bertaut & Haliassos, 2002;
Rha, Montalto, & Hanna, 2006).
A principal component analysis is performed to identify variables to explain two
components: credit attitude and myopia. Seven variables are used in the principal
component analysis to create the two components. The analysis is conducted with the
101
Texas Tech University, Laura Ricaldi, August 2015
SCF years 1998-2010 using only the first implicate. The sample size is 24,165
respondents. Given the 7 variables, the principal component analysis identifies 2
components with eigenvalues above 1.0, which is the threshold for significance according
to the Kaiser-Guttman rule. Further analysis of the scree plot also shows that 2
components should be retained.
The last step of the principal component analysis is to analyze the variables that
load and label each component. The two components can be classified as credit attitude
and myopia. The credit attitude component includes having a positive attitude about
using credit, willing to borrow to go on vacation, willing to borrow to purchase luxury
items like jewelry and furs, and willing to use credit to cover living expenses if there is a
cut in income. The myopia component includes: spending exceeded income over the
past year, savings habits are described as not saving and spending more or as much as
income, and smoking. The final component loading is shown in Table 4.6. The variance
explained by the principal component analysis is approximately 40.8%.
102
Texas Tech University, Laura Ricaldi, August 2015
Table 3.6: Final Component Loading for Credit Attitude and Myopia
Spending exceeds or is equal to income
Does not save since spending exceeds or equals
income
Smokes
Has a positive attitude toward credit
Willing to use credit to fund a vacation
Willing to use credit to purchase furs or jewelry
Willing to use credit to cover living expenses if
income were cut
Eigenvalue
Proportion of total variance
Component 1:
Credit
Attitude
0.28379
0.23194
Component 2:
Myopia
0.14458
0.55361
0.70773
0.62619
0.40652
0.45953
-0.2438
-0.22274
-0.31721
0.21113
1.519976
0.180716
0.4085
0.6301
0.69052
After the principal component analysis, a principal factor analysis is conducted to
confirm the latent concepts of credit attitude and myopia. Table 4.7 shows the final
factor pattern for the two factors.
Table 3.7: Factor Loading
Credit
Attitude
Myopia
Spending exceeds or is equal to income
Does not save since spending exceeds or equals
income
0.15524
0.14118
0.40811
0.54462
Smokes
Has a positive attitude toward credit
Willing to use credit to fund a vacation
Willing to use credit to purchase furs or jewelry
Willing to use credit to cover living expenses if
income were cut
0.05877
0.3333
0.60723
0.45907
0.20476
0.21734
-0.0952
-0.1189
-0.1584
0.11527
103
Texas Tech University, Laura Ricaldi, August 2015
With the two factors, two scores are calculated and groups are formed. For the
credit attitude factor, two groups are created by finding the households with above the
mean attitude and below the mean credit attitude. The same procedure is done with the
myopia factor to create above the mean myopia and below the mean myopia. A further
grouping is also conducted by combining the two groups of the two factors to create four.
The four groups are: above the mean attitude/ above the mean myopia, above the mean
attitude/below the mean myopia, below the mean attitude/above the mean myopia, and
below the mean attitude/below the mean myopia.
104
Texas Tech University, Laura Ricaldi, August 2015
Appendix II
Logistic Regression: Debit Card Users Compared to All Credit Card Users
Table 3.8: Logistic Regression of the Likelihood of being a Debit Card User Compared to
All Credit Card Users in the SCF between 1998-2010 (n=10,828)
Parameter
Estimate
-2.5221
Intercept
Attitude/Myopia Groups
AM Attitude/ AM Myopia
-0.116
AM Attitude/ BM Myopia
-0.3684
BM Attitude/ AM Myopia
0.1541
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
-3.2249
Quintile 2
0.0602
Quintile 4
0.7298
Quintile 5 (least sophisticated)
1.3877
Education
Less than high school
0.1264
High School
0.1186
Some College
0.3259
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
-2.7961
Silent Generation (1926-1945)
-2.3306
Baby Boomer Generation (1946-1964)
-1.1541
Generation X (1965-1980)
-0.3862
Female
-0.1063
Not Married
0.3971
Have children
0.3305
Finances
Income
$25,001-$50,000
0.0817
$50,001-$75,000
0.2965
$75,001-$100,000
0.1344
Greater than $100,000
-0.8364
Net Worth
Less than $10,000
1.3374
$10,001-$50,000
1.0841
$50,001-$100,000
0.6748
$100,001-$150,000
0.5154
Survey Year
Year 2001
0.6394
Year 2004
1.4875
Year 2007
2.1691
Year 2010
3.4108
R2
0.4775
Adjusted R2
0.7038
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
105
p
Odds Ratio
<0.0001
***
0.2451
0.0005
0.1088
0.89
0.692
1.167
<0.0001
0.5771
<0.0001
<0.0001
0.04
1.062
2.075
4.006
0.3276
0.2552
0.0023
1.135
1.126
1.385
<0.0001
<0.0001
<0.0001
0.0412
0.1505
<0.0001
<0.0001
0.061
0.097
0.315
0.68
0.899
1.488
1.392
0.3722
0.0133
0.4114
<0.0001
1.085
1.345
1.144
0.433
***
<0.0001
<0.0001
<0.0001
0.0006
3.809
2.957
1.964
1.674
***
***
***
***
<0.0001
<0.0001
<0.0001
<0.0001
1.895
4.426
8.751
30.289
***
***
***
***
***
***
***
***
***
***
***
***
**
***
***
**
Texas Tech University, Laura Ricaldi, August 2015
Logistic Regression: Debit Card Users Compared to Revolving Credit Card Users
Table 3.9: Logistic Regression of the Likelihood of being a Debit Card User Compared to
Revolving Credit Card Users in the SCF between 1998-2010 (n=5,254)
Parameter
Estimate
-0.9913
Intercept
Attitude/Myopia Groups
AM Attitude/ AM Myopia
-0.4482
AM Attitude/ BM Myopia
-0.6119
BM Attitude/ AM Myopia
-0.0181
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
-2.8875
Quintile 2
-0.0505
Quintile 4
0.4574
Quintile 5 (least sophisticated)
0.9825
Education
Less than high school
-0.0168
High School
-0.0654
Some College
0.0681
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
-2.0939
Silent Generation (1926-1945)
-2.1343
Baby Boomer Generation (1946-1964)
-1.2889
Generation X (1965-1980)
-0.5394
Female
-0.2036
Not Married
0.4338
Have children
0.2487
Finances
Income
$25,001-$50,000
-0.0132
$50,001-$75,000
0.1067
$75,001-$100,000
-0.0558
Greater than $100,000
-0.6075
Net Worth
Less than $10,000
0.892
$10,001-$50,000
0.66
$50,001-$100,000
0.2958
$100,001-$150,000
0.2067
Survey Year
Year 2001
0.6068
Year 2004
1.4437
Year 2007
2.0994
Year 2010
3.469
R2
0.4438
Adjusted R2
0.5922
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
106
p
Odds Ratio
0.0011
***
<0.0001
<0.0001
0.8657
0.639
0.542
0.982
***
***
<0.0001
0.6807
<0.0001
<0.0001
0.056
0.951
1.58
2.671
***
0.906
0.5731
0.5672
0.983
0.937
1.07
<0.0001
<0.0001
<0.0001
0.016
0.0122
<0.0001
0.004
0.123
0.118
0.276
0.583
0.816
1.543
1.282
***
***
***
**
**
***
***
0.8955
0.4183
0.7535
0.001
0.987
1.113
0.946
0.545
***
<0.0001
<0.0001
0.0367
0.2099
2.44
1.935
1.344
1.23
***
***
**
<0.0001
<0.0001
<0.0001
<0.0001
1.835
4.236
8.161
32.103
***
***
***
***
***
***
Texas Tech University, Laura Ricaldi, August 2015
Logistic Regression: Debit Card Users Compared to Convenience Credit Card
Users
Table 3.10: Logistic Regression of the Likelihood of being a Debit Card User Compared
to Convenience Credit Card Users in the SCF between 1998-2010 (n=8,332)
Parameter
Estimate
-2.5088
Intercept
Attitude/Myopia Groups
AM Attitude/ AM Myopia
0.4437
AM Attitude/ BM Myopia
-0.1367
BM Attitude/ AM Myopia
0.4034
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
-3.5461
Quintile 2
0.1011
Quintile 4
1.0156
Quintile 5 (least sophisticated)
2.0975
Education
Less than high school
0.2169
High School
0.2503
Some College
0.6025
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
-3.4667
Silent Generation (1926-1945)
-2.5834
Baby Boomer Generation (1946-1964)
-1.0786
Generation X (1965-1980)
-0.2025
Female
0.055
Not Married
0.3251
Have children
0.6028
Finances
Income
$25,001-$50,000
0.3282
$50,001-$75,000
0.5973
$75,001-$100,000
0.2915
Greater than $100,000
-0.8632
Net Worth
Less than $10,000
2.0645
$10,001-$50,000
1.5867
$50,001-$100,000
0.9693
$100,001-$150,000
0.827
Survey Year
Year 2001
0.6815
Year 2004
1.3812
Year 2007
2.0832
Year 2010
3.2243
R2
0.6012
Adjusted R2
0.8361
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
107
p
Odds Ratio
<0.0001
***
0.0042
0.3375
0.0021
1.558
0.872
1.497
***
<0.0001
0.4594
<0.0001
<0.0001
0.029
1.106
2.761
8.146
***
0.2206
0.0665
<0.0001
1.242
1.284
1.827
<0.0001
<0.0001
0.0002
0.4985
0.5842
0.0036
<0.0001
0.031
0.076
0.34
0.817
1.057
1.384
1.827
***
***
***
0.0147
0.0003
0.1723
<0.0001
1.388
1.817
1.338
0.422
**
***
<0.0001
<0.0001
<0.0001
<0.0001
7.881
4.888
2.636
2.286
***
***
***
***
<0.0001
<0.0001
<0.0001
<0.0001
1.977
3.98
8.03
25.137
***
***
***
***
***
***
***
*
***
***
***
***
Texas Tech University, Laura Ricaldi, August 2015
Logistic Regression: Revolving Credit Card Users Compared to Convenience Credit
Card Users
Table 3.11: Logistic Regression of the Likelihood of Being a Revolving Credit Card User
Compared to a Convenience Credit Card User in the SCF between 1998-2010 (n=8,070)
Parameter
Estimate
-1.9323
Intercept
Credit Attitude
Above Mean Attitude
0.7939
Myopia
Above Mean Myopia
0.4845
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
-0.8319
Quintile 2
-0.2207
Quintile 4
0.175
Quintile 5 (least sophisticated)
0.3606
Education
Less than high school
0.3729
High School
0.4908
Some College
0.5812
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
-0.8398
Silent Generation (1926-1945)
0.1605
Baby Boomer Generation (1946-1964)
0.8095
Generation X (1965-1980)
0.7808
Female
0.262
Not Married
-0.1749
Have children
0.2658
Finances
Income
$25,001-$50,000
0.3509
$50,001-$75,000
0.4773
$75,001-$100,000
0.6364
Greater than $100,000
-0.3591
Net Worth
Less than $10,000
1.521
$10,001-$50,000
1.2334
$50,001-$100,000
0.9466
$100,001-$150,000
0.8175
Survey Year
Year 2001
-0.1445
Year 2004
-0.2011
Year 2007
-0.069
Year 2010
-0.4538
R2
0.2889
Adjusted R2
0.407
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
108
p
Odds Ratio
<0.0001
***
<0.0001
2.212
***
<0.0001
1.623
***
<0.0001
0.0159
0.1073
0.0058
0.435
0.802
1.191
1.434
***
**
0.0028
<0.0001
<0.0001
1.452
1.634
1.788
***
***
***
0.0157
0.628
0.014
0.0191
<0.0001
0.0172
0.0001
0.432
1.174
2.247
2.183
1.3
0.84
1.304
**
**
**
***
**
***
0.0009
<0.0001
<0.0001
0.0043
1.42
1.612
1.89
0.698
***
***
***
***
<0.0001
<0.0001
<0.0001
<0.0001
4.577
3.433
2.577
2.265
***
***
***
***
0.0718
0.0227
0.4649
<0.0001
0.865
0.818
0.933
0.635
*
**
***
***
Texas Tech University, Laura Ricaldi, August 2015
Table 3.12: Logistic Regression of the Likelihood of being a Revolving Credit Card User
Compared to Convenience Credit Card Users in the SCF between 1998-2010 (n=8,070)
Intercept
Attitude/Myopia Groups
AM Attitude/ AM Myopia
AM Attitude/ BM Myopia
BM Attitude/ AM Myopia
Human Capital
Financial Sophistication
Quintile 1 (most sophisticated)
Quintile 2
Quintile 4
Quintile 5 (least sophisticated)
Education
Less than high school
High School
Some College
Life-cycle Factors
Generational Cohort
Greatest Generation (1900-1925)
Silent Generation (1926-1945)
Baby Boomer Generation (1946-1964)
Generation X (1965-1980)
Female
Not Married
Have children
Finances
Income
$25,001-$50,000
$50,001-$75,000
$75,001-$100,000
Greater than $100,000
Net Worth
Less than $10,000
$10,001-$50,000
$50,001-$100,000
$100,001-$150,000
Survey Year
Year 2001
Year 2004
Year 2007
Year 2010
R2
Adjusted R2
*Significant at 0.10; **significant at 0.05; ***significant at 0.01.
109
Parameter
Estimate
-1.9314
p
Odds Ratio
<0.0001
1.2818
0.7911
0.4806
<0.0001
<0.0001
<0.0001
3.603
2.206
1.617
***
***
***
-0.832
-0.2207
0.1748
0.3603
<0.0001
0.0159
0.108
0.0058
0.435
0.802
1.191
1.434
***
**
0.373
0.4908
0.5814
0.0028
<0.0001
<0.0001
1.452
1.634
1.789
***
***
***
-0.8399
0.1602
0.8092
0.7804
0.262
-0.1747
0.2659
0.0157
0.6287
0.014
0.0191
<0.0001
0.0173
0.0001
0.432
1.174
2.246
2.182
1.299
0.84
1.305
**
**
**
***
**
***
0.3509
0.4775
0.6366
-0.3589
0.0009
<0.0001
<0.0001
0.0043
1.42
1.612
1.89
0.698
***
***
***
***
1.5211
1.2336
0.9467
0.8176
<0.0001
<0.0001
<0.0001
<0.0001
4.577
3.433
2.577
2.265
***
***
***
***
-0.1444
-0.2011
-0.069
-0.4539
0.2889
0.407
0.072
0.0227
0.4647
<0.0001
0.866
0.818
0.933
0.635
*
**
***
***
***
Texas Tech University, Laura Ricaldi, August 2015
References
Ameriks, J., Caplin, A., Leahy, J., & Tyler, T. (2007). Measuring self-control
problems. The American Economic Review, 966-972.
Becker, G. S. (1962). Investment in human capital: A theoretical analysis. The Journal of
Political Economy, 9-49.
Berkowitz, E. N., & Schewe, C. D. (2011). Generational cohorts hold the key to
understanding patients and health care providers: Coming-of-age experiences
influence health care behaviors for a lifetime. Health Marketing Quarterly, 28(2),
190-204.
Bertaut, C. C., & Haliassos, M. (2001). Debt revolvers for self-control. HERMES Center
Working Paper, 01-11.
Board of Governors of the Federal Reserve System. (2012). Codebook for 2010 Survey
of Consumer Finances.
http://www.federalreserve.gov/econresdata/scf/files/codebk2010.txt
Borzekowski, R. & Kiser, E. K. (2008). The choice at the checkout: Quantifying demand
across payment instruments. International Journal of Industrial Organization,
26(4), 889-902.
Borzekowski, R., Kiser, E. K., & Ahmed, S. (2008). Consumers' use of debit cards:
patterns, preferences, and price response. Journal of Money, Credit and
Banking, 40(1), 149-172.
Ching, A. T. & Hayashi, F. (2010). Payment card rewards programs and consumer
payment choice. Journal of Banking & Finance, 34(8), 1773-1787.
110
Texas Tech University, Laura Ricaldi, August 2015
Choe, H., Yoon, S. J., & Johnson, D. P. (1991). Gender differences in the choice of
payment method for purchasing durable goods. In Gender and Consumer
Behaviour Conference Proceedings. University of Utah.
Coomes, M. D. (2004). Understanding the historical and cultural influences that shape
generations. New Directions for Student Services, 2004(106), 17-31.
de Bassa Scheresberg, C., Lusardi, A., & Yakoboski, P. J. (2014). College-Educated
Millennials: An Overview of Their Personal Finances.
Durkin, T. A. (2000). Credit cards: Use and consumer attitudes 1970–2000. Federal
Reserve Bulletin, 86, 623–634.
Fusaro, M. A. (2013). Why do people use debit cards: Evidence from checking
accounts. Economic Inquiry, 51(4), 1986-2001.
Gathergood, J. & Weber, J. (2014). Self-control, financial literacy & the co-holding
puzzle. Journal of Economic Behavior & Organization, 107, 455-469.
Gorsuch, R. L. (1997). Exploratory factor analysis: Its role in item analysis. Journal of
Personality Assessment, 68(3), 532-560.
Hayashi, F. & Stavins, J. (2012). Effects of credit scores on consumer payment
choice. Federal Reserve Bank of Kansas City Working Paper, (12-03).
Hayhoe, C. R., Leach, L. J., Turner, P. R., Bruin, M. J., & Lawrence, F. C. (2000).
Differences in spending habits and credit use of college students. Journal of
Consumer Affairs, 34(1), 113-133.
111
Texas Tech University, Laura Ricaldi, August 2015
Herbst-Murphy, S. (2010). Trends and preferences in consumer payments: Lessons from
the Visa Payment Panel Study. Federal Reserve Bank of Philadelphia Payment
Cards Center Discussion Paper.
Hogarth, J. M. & Anguelov, C. (2004). Are families who use e-banking better financial
managers?. Journal of Financial Counseling and Planning, 15(2), 61-77.
Huston, S. J., Finke, M. S., & Smith, H. (2012). A financial sophistication proxy for the
Survey of Consumer Finances. Applied Economics Letters,19(13), 1275-1278.
King, A. S. & King, J. T. (2005). The decision between debit and credit: Finance charges,
float, and fear. Financial Services Review, 14(1), 21-36.
Lee, J., Abdul-Rahman, F., & Kim, H. (2007). Debit card usage: An examination of its
impact on household debt. Financial Services Review, 16(1), p73-87.
Lindamood, S., Hanna, S. D., & Bi, L. (2007). Using the Survey of Consumer Finances:
Some methodological considerations and issues. The Journal of Consumer
Affairs, 41, 195-222.
Oliver Wyman. (2013). Consumer payment preferences: Debit spotlight. Retrieved from
https://www.pulsenetwork.com/pulse/documents/index/serveDoc.html?doc=Olive
r_Wyman_Consumer_Payment_Preferences_V2_07-17-2013
Parrish, L. (2009). Overdraft explosion. Center for Responsible Lending. Chicago.
Rha, J. Y., Montalto, C. P., & Hanna, S. D. (2006). The effect of self-control mechanisms
on household saving behavior. Journal of Financial Counseling and
Planning, 17(2).
112
Texas Tech University, Laura Ricaldi, August 2015
Robb, C. A., Huston, S. J., & Finke, M. S. (2008). The mitigating influence of time
preference on the relation between smoking and BMI scores. International
Journal of Obesity, 32(11), 1670-1677.
Rogers, B. & Dopico, L.G. (2010). Debit or credit: Either, both, neither or how much?.
Rysman, M. (2009). Consumer payment choice: Measurement topics. The Changing
Retail Payments Landscape, 61.
Shefrin, H. M., & Thaler, R. H. (1988). The behavioral life‐cycle hypothesis. Economic
Inquiry, 26(4), 609-643.
Sprenger, C., & Stavins, J. (2010). Credit card debt and payment use. Financial Services
Review, 19(1).
Stango, V. & Zinman, J. (2008). Debit vs. credit: How people choose to pay.
Stango, V., & Zinman, J. (2009). What do consumers really pay on their checking and
credit card accounts? Explicit, implicit, and avoidable costs. American Economic
Review, 99(2), 06-09.
Strotz, R. H. (1955). Myopia and inconsistency in dynamic utility maximization. The
Review of Economic Studies, 165-180.
Thaler, R. H. & Shefrin, H. M. (1981). An economic theory of self-control. The Journal
of Political Economy, 392-406.
Zinman, J. (2004). Why use debit instead of credit? Consumer choice in a trillion-dollar
market. Consumer Choice in a Trillion-Dollar Market (July 2004). FRB of New
York Staff Report, (191).
Zinman, J. (2009). Debit or credit?. Journal of Banking & Finance, 33(2), 358-366.
113