What Shapes Consumer Choice and Financial Products?

: A
review
Sumit Agarwal (Georgetown University)
Souphala Chomsisengphet (OCC)
Cheryl Lim (EDHEC)
April 19, 2017

Abstract
Central to the field of consumer finance is that consumers make financial decisions that
do not always coincide with the ideal financial decisions ideally depicted in optimal economic
models [Campbell, 2006]. In this review, we discuss developments in the field of household
finance, focusing on how consumers make suboptimal financial decisions across different types
of settings and factors that affect these suboptimal decisions. Rather than conducting a com-
prehensive survey, we focus more specifically on consumer choice in context of research in
credit card borrowing, housing and mortgage debts, investments and savings decisions, as well
as spending and consumption. We also discuss financial product design and marketing; and
regulatory landscape of lenders of consumer financial products. We also discuss five future
research directions and considerations for researchers and policy makers.

Key words: household finance, consumer finance, consumer behavior, consumer bias, consumer
choice, consumer credit, choice inefficiencies, cognitive ability, financial literacy, financial ed-
ucation, consumer financial regulation, financial product design, financial product marketing,
mortgage debt, credit card, personal investments, financial products

1 Introduction
The field of consumer finance has only gained much attention over the past decade. One notable
impetus was John Campbell’s Presidential Address to the American Finance Association in 2006.
[Campbell, 2006] contended that a well-functioning financial market depends on household’s effec-
tive financial decisions and behaviors in the market, and offered powerful motivations to further
develop research in the field of household finance. Shortly thereafter, the housing meltdown and
financial crisis in 2008 generated contentious debates about the role of consumer financial decisions
and the role of financial intermediaries of consumer financial products in causing the financial crisis.
In response to market failures, policymakers subsequently implemented policies (e.g., Home Af-
fordable Modification Program in 2009, Credit Card Accountability Responsibility and Disclosure
Act (CARD Act) in 2009, Dodd-Frank Act in 2010) to protect consumers and regulate lenders,
providing applied micro-economists with natural experiments to study the impact of policy re-
lated to the consumer credit market on lender and household behavior. Further allowing applied
micro-economists to develop more rigorous empirical analysis is the technological advancements
that allow for more efficient collection, storage and processing of large administrative consumer
credit data. As a result, there has been a growing body of empirical research to better understand
consumer and lender choice and behavior, as well as the levers that drive them.

Central to the field of consumer finance is that consumers make financial decisions that do not
always coincide with the financial decisions ideally depicted in neoclassical economic models of
optimal consumer choice or behavior [Campbell, 2006]. Researchers are uncovering the challenges
that a household faces when making financial decisions and managing their finances, the limita-
tions in cognitive ability and behavioral biases that hinder optimal decision making, the impact
of the financial service providers’ business models and business decisions, and the extent to which
regulators can influence and safeguard consumer behavior. Small deviations from optimal decision-
making at the consumer level might still potentially have important aggregate implications (e.g.,

1
[Gabaix and Laibson, 2006]; or analogously for firms, [Akerlof and Yellen, 1985]), especially given
that the value at stake is high (as of 2016 Q1, US households owe about USD12.25 trillion in
debt1 or approximately 78 percent of US GDP2 ) and these financial decisions affect most house-
holds [Zinman, 2014]. Moreover, [Agarwal et al., 2009a] show that consumers make mistakes across
many domains of financial products, and such mistakes are likely to be correlated. Thus, together
they can add up to be economically costly to consumers.

In this review, we discuss developments in the field of household finance, focusing on how con-
sumers make suboptimal financial decisions across different types of settings and factors that
affect these suboptimal decisions. Rather than conducting a comprehensive survey, we focus more
specifically on consumer choice in the context of research in credit card borrowing, housing and
mortgage debts, investments and savings decisions, as well as spending and consumption. We also
discuss financial product design and marketing; and regulatory landscape of lenders of consumer
financial products. We review major contributions and identify open interest research questions
that have yet to be answered.

2 Consumer choice inefficiencies and suboptimal financial be-

havior
Following [Shefrin and Thaler, 1988], who provide a behaviorally realistic treatment of household
financial- as well as consumption-saving decisions, there has been a continuous growth in academic
research studying factors that affect consumer financial decision making and related behavioral
biases in household finance. While [Shefrin and Thaler, 1988] present a theory on the role of costly
self-control, mental accounting and framing, many researchers are documenting specific evidence of
how consumer financial decisions are inefficient and the factors that affect these suboptimal choices
in various important settings. Below we survey the literature on consumer decisions to borrow in
the credit card and mortgage markets, to invest and save for retirement, as well as to spend and
consume.

2.1 Credit card contract choice and borrowing

In choosing credit card contracts, [Shui and Ausubel, 2004] study credit card contract choice using
a 1995 large-scale randomized experiment with 600,000 consumers facing the choice between six
different credit card offers with different introductory interest rates and durations. The authors
document two important behaviors: (1) consumers prefer an introductory offer with a lower inter-
est rate and shorter duration to one with a higher interest rate with a longer duration, even though
they would benefit more from choosing the latter; and (2) consumers are reluctant to switch and
many consumers who have switched before fail to switch again. [Shui and Ausubel, 2004] argue
that this behavior is puzzling because consumers would actually pay lower interest rate costs, ex
post, under the longer introductory offer.3 One limitation of this study is the use of aggregate
credit and saving outcomes, which makes the link between borrowing and present bias indirect
since examination of aggregates data does not allow for evaluation of individual behavior.

[Meier and Sprenger, 2010] overcome the limitation of aggregate data by combining directly elicited
time preference measures with administrative data on credit card borrowing. They measure in-
dividual discount factors using an incentivized choice experiment with a sample of about 600
low-to-moderate income individuals at two Volunteer Income Tax Assistances in Boston over the
period 2006 - 2007. Their results show that present-biased individuals are 15 percent more likely
to have credit card debts. Conditional on borrowing, they borrow 25 percent (measured by out-
standing balances on revolving accounts from the credit reports) more than dynamically consistent
individuals. The correlation is robust after controlling for income, credit constraints and socio-
demographic characteristics. The study faces some concerns about external validity given that the
data covers a small sample of low-income consumers in Boston.

1 QuarterlyReport on Household Debt and Credit, May 2016, Federal Reserve Bank of New York
2 Bankfor International Settlements
3 DellaVigna and Malmendier (2002) find that consumers systematically choose sub-optimal membership plans

at health clubs. Miravete (2003) finds consumers’ choices of telephone billing plans to be closer to optimal.

2
Also overcoming the aggregate data limitation, [Agarwal et al., 2015b] document consumer mis-
takes in choosing a credit card contract using individual-level data based on unique experiment in
which a bank offers consumers in the US a choice between two credit card contracts (one with an
annual fee but a lower interest rate and one with no annual fee but a higher interest rate). While on
average consumers choose a credit contract that minimizes their net costs, the authors document
that about 40 percent of consumers choose the ex post sub-optimal credit contract, with some
households incurring hundreds of dollars of avoidable interest charges. The probability of choos-
ing the sub-optimal contract, however, declines with the dollar magnitude of the potential error
and consumers with larger errors were more likely to subsequently switch to the optimal contract.
Using a merged dataset of payday loan borrowing and credit card history at the individual-level,
[Agarwal et al., 2009b] find that most consumers borrowed on payday loans despite having sub-
stantial unused liquidity on their credit cards. This is highly inefficient given that the cost of
payday loan borrowing is substantially larger relative to the cost of credit card borrowing; the
annual percentage rate on a payday loan is usually more than 100 percent.

With respect to credit card borrowing, [Ausubel, 1991] document that consumers are overly sen-
sitive to teaser rates generally offered by credit card lenders, but are insensitive to the go-to,
post-teaser interest rates, which result in credit card interest rates being "sticky". [Ausubel, 1991]
further examines the promotional "teaser" interest rates and concludes that consumers are overly
sensitive to such rates, possibly because consumers underestimating their future credit card bor-
rowing at the higher, post-teaser interest rates. [Gross and Souleles, 2002a] are also among the
first researchers to document another “credit card debt puzzle” – that US households tend to hold
significant credit card debt and sizable liquid assets simultaneously, which is irrational or subopti-
mal according to the traditional economic model since returns on liquid assets are typically lower
than the interest charged on credit card debt. [Bertaut et al., 2009] use the Survey of Consumer
Finance (SCF) data to confirm that the credit card debt puzzle is a persistent behavioral phe-
nomenon. In similar spirit, [Zinman, 2009] conclude that credit card debt revolvers borrow high
and lend low. He estimates that 27-30 percent of US households with credit cards lose USD10
per month by not using demand deposits to pay down credit card debt. In a more recent study,
[Stango and Zinman, 2016] finds that there is a huge dispersion in borrowing costs in the US across
individuals, and this is due to two factors: the annual percentage rate (APR) offers vary widely
across lenders, and that borrowers also vary in shopping intensity. They show that the difference
in APRs paid by shoppers versus non-shoppers could be as different as those paid by borrowers in
the best versus worst credit score deciles.

In borrowing on credit card, US households owe about USD712 Billion in credit card debt as
of Q1 20164 . [Wang and Keys, 2014] document that 29 percent of credit card account holders’
monthly repayment is at or near the required minimum payment. In terms of repaying back credit
card debt, [Kuchler, 2015] uses daily balance and transactions of bank accounts and credit cards of
consumer who signed up online with a financial management service to study how consumers stick
to their plan to reduce their debt balance per month. For each committed/planned dollar of debt
pay-down, consumers reduced their debt by only 25-30 cents on average. [Agarwal et al., 2015d]
study the effect of a nudging disclosure requirement in the CARD Act of 2009, which requires credit
card lenders to disclose information about the interest savings that could be achieved by if the credit
card holders were to pay off balances in 36 months rather than making minimum payments. They
detect only a 0.4 percentage points increase in the share of accounts making the 36-month payment
value but no evidence of a change in overall payments. Similarly, [Wang and Keys, 2014] find that
less than 1 percent of credit card account holders adopt the 36-month repayment plan amount
disclosed in the CARD Act.

Using a panel data of credit card account repayment behavior, [Gross and Souleles, 2002b] docu-
ment an increase in credit card borrowers’ willingness to default on their debt because of a reduction
in the costs of defaulting, which include social, informational, and legal costs. [Andersson et al., 2013]
also document changes in the pecking order of consumer default behavior. Specifically, consumers
were eight times more likely to prioritize payments on mortgage debt over credit card payments
before the 2008 financial crisis. In contrast, the similar consumers were just as likely to default on
mortgage debt as on credit card debt during the crisis. One of the main factors contributing to the
4 Federal Reserve Bank of New York

3
change in the pecking order behavior could be due to strategic default behavior due to large levels
of negative equity during the housing meltdown (see, for example, [Campbell and Cocco, 2015]).

2.2 Mortgage finance

In the area of mortgage financing, forward-looking behavior and cost minimization have been of-
fered as possible explanations for consumers’ choice of mortgage form. The decision between a
fixed-rate mortgage (FRM) and adjustable-rate mortgage (ARM), which depends on expected fu-
ture costs, is related to the borrower’s expectation about whether these are likely to be higher or
lower in the future. For example, if borrowers expect that future inflation is likely to be high, they
are more likely to choose a fixed-rate mortgage. [Kuchler and Zafar, 2015] also provide evidence
of how personal experience shapes expectations in housing and labor markets. Individuals who
have experienced more volatile price changes are likely to perceive future house price changes to
be more uncertain, and individuals experiencing unemployment are likely to be more pessimistic
about nationwide unemployment.

[Campbell and Cocco, 2003] and [Campbell and Cocco, 2015] model a household mortgage choice
model and mortgage default decision under borrowing constraints, and find that households care
more about the current interest costs than about the lifetime costs of the loan. This suggests
that the spread between FRM and ARM rates should be the primary determinant of the choice
between an FRM versus an ARM product. Using a nine-country panel and instrumental variables
methods to examine how near-term one-year rational expectations of future movements in ARM
rates affect mortgage choice, [Badarinza et al., 2015] find that the current spread between FRM
and ARM rates as well as near term rational expectations of ARM rates affect mortgage choice,
while longer-term three-year rational forecasts have a weaker effect. This suggests that households
are concerned with current interest costs as well as interest costs over the life of the loan.

One important feature of the US mortgage market is the fact that mortgage borrowers can gener-
ally refinance their mortgage in order to reduce the mortgage balance, extract the equity in their
home, and/or to reduce their interest rate. In deriving a closed-form optimal refinance solution,
[Agarwal and Mazumder, 2013] finds that the interest rate differential for an optimal refinance
typically ranges from 100 to 200 basis points. In making this decision, researchers document in-
efficiencies in refinancing with respect to rate and timing. [Agarwal et al., 2016] find that the
majority of borrowers refinance sub-optimally. They estimate that about 59 percent of borrowers
refinance sub-optimally. More specifically, about 52 percent of the borrowers chose a sub-optimal
rate, 17 percent of the borrowers waited too long to refinance, and about 10 percent make both
rate and timing mistakes. [Keys and Pope, 2016] quantify the cost of this inefficiency. In their
random sample of outstanding US mortgages in December 2010, about 20 percent of financially
unconstrained households for whom refinancing was optimal had not done so. This is equivalent
to foregone savings of about USD160 per month over the remaining life of each loan or a total
discounted present-value of USD11,500 in foregone savings per borrower.

During the 2008 housing meltdown, households experienced large levels of negative equity. The lit-
erature on mortgage default emphasize important role of home equity in inducing a homeowner to
default on his mortgage (e.g., [Deng et al., 2000]; [Elul et al., 2010]; [Campbell and Cocco, 2015].
While many argue that the negative equity in the house induced strategic default behavior during
the 2008 crisis, [Bhutta et al., 2010] document that, of the households who purchased their home
at the peak of housing bubble in 2006, about 80 percent default on their mortgage due to a com-
bination of income shocks and negative equity. In addition, the authors find despite having little
financial incentive to keep paying their mortgage, as they bought their home during the height of
the bubble and put little money towards the down payment, many of these households keep paying
their mortgage. The median borrower stops paying his mortgage when housing equity falls to -62
percent.

2.3 Investment and retirement decisions

In investment decisions, [Korniotis and Kumar, 2013] show that the apparent inefficiencies are con-
centrated portfolios, excessive trading and preference for local stocks. They point out, though, that
all three instances of portfolio distortions could be due to either an informational advantage, which

4
means the investor is smart, or psychological biases. On the other hand, [Malmendier and Nagel, 2016]
as well as [Botsch and Malmendier, 2015] show empirically that lifetime experiences is a determi-
nant of differences in expectations and affect financial decision-making. Households who experi-
enced good stock returns during their lifetimes invest more of their wealth in stocks, while house-
holds who experienced good bond returns during their lifetimes invest more in bonds.

In retirement decisions, the inefficiency is in failing to maximize benefit from their retirement
plans and accounts. [Choi et al., 2011] analyze the 401(k) investment choices of employees at
seven companies and found that employees older than 59 12 were contributing less than the em-
ployer matching contribution threshold despite being vested in their match and being able to make
penalty-free 401(k) withdrawals for any reason. This means that, on average, 36 percent of match-
eligible employees over age 59 12 forgo arbitrage profits that average 1.6 percent of their annual
pay, or $507. More interestingly, they found that a survey educating employees about the free
lunch they are forgoing raised contribution rates by only a statistically insignificant 0.67 percent
of income among those completing the survey. The tax inefficiency of asset allocations of house-
holds investing in taxable and tax-deferred accounts (TDAs) is another puzzle. While theory, for
example [Dammon et al., 2004] and [Shoven and Sialm, 2004], suggests that savers should locate
higher-tax assets such as bonds in their tax-deferred retirement accounts (TDAs) while keeping
low-tax assets (equities) in taxable accounts, observed portfolios shows otherwise and are not tax
efficient. [Amromin, 2003] explains this puzzle with precautionary portfolio choice, i.e., borrowing-
constrained households forgo tax efficiency for allocations that provide more liquidity in bad income
states. Their empirical analysis suggests that the choice of whether to hold a tax-efficient portfolio
and the degree of portfolio tax inefficiency are related to the presence and intensity of precaution-
ary motives.

The extent to which individuals optimize their savings to take advantage of tax-deferred ac-
counts is also actively researched. Researchers in the US are empirically assessing the extent
to which tax-preferred saving individual retirement accounts (IRAs) accounts such as 401(k)
and Roth 401(k) can increase the amount of private saving. They investigate the causal re-
lationship by exploiting the variation in initiation of topping up in the tax subsidized saving
accounts [Attanasio and DeLeire, 1994], policy on eligibility [Gale and Scholz, 1994], balance on
the tax-subsidized saving accounts ([Venti and Wise, 1987]and [Venti and Wise, 1988]), or eligi-
bility ([Benjamin, 2003] and [Gelber, 2011]) to see whether there is significant change in tax-
able savings or total wealth.5 Recent developments in behavior economics show that automatic
enrollment significantly increases saving within retirement accounts ([Madrian and Shea, 2001];
[Choi et al., 2001]; [Thaler and Benartzi, 2004]). Recent work by [Beshears et al., 2015] analyze
administrative data from 11 U.S. firms that introduced a Roth IRA retirement plan between 2006
and 2010. They find no significant reduction in the total 401(k) contribution rates following the
introduction of the Roth IRA (deferred tax benefit) which implies that take-home pay declines
and total retirement savings increases following the introduction of the Roth. The authors analyze
additional survey data and attribute their finding to employee confusion or inattention about the
tax properties of the Roth and behavior bias of partition dependence, rather than employee making
an active calculated decision. [Arnberg and Barslund, 2014] study the crowd-out effect of Danish
mandatory pension schemes for the renters. They find that for every Euro paid to the mandatory
pension accounts, there is 0 to 30 cents reduction in other private savings depending on age. They
attribute the low crowding-out effect on private savings to liquidity constraints. On the other hand,
[Chetty et al., 2014], using a rich panel data on all private savings, find that retirement savings
policy in Denmark does not effectively increase private savings given that 85 percent of individuals
are passive savers who are unresponsive to subsidies, while the other 15 percent of individuals are
active savers who respond to tax subsidies by shifting assets across accounts. The authors conclude
that there are substantial crowd-out effects and that automatic contributions are more effective in
increasing saving rates than the tax-subsidized savings policy.
5 Conclusions from these empirical studies are mixed since they are based on different econometrics assumptions.

There are two major concerns: (1) low quality and infrequent data, and (2) the unobserved heterogeneity in the
disposition to save between the treatment and control group, which can bias the estimated effects on private savings.

5
2.4 Spending and consumption decisions
Researchers are focusing efforts to better understand factors that stimulate household consump-
tion. The secular decline in US personal savings rate, which began in the 1980s, coincides
roughly with a secular increase in the dissemination and use of credit cards. One conjecture
is that increased access to and use of credit card have stimulated consumer spending. Several
early studies, using surveys and experiments, also provide support including [Hirschman, 1979],
[Feinberg, 1986], and [Prelec and Simester, 2001]. Similarly, in the area of mortgage debt, access
to better loan terms appears to stimulate consumption. [Brady et al., 2000], [Canner et al., 2002]
and [Greenspan and Kennedy, 2008] found that homeowners who refinanced their mortgages on
more favorable terms, coupled with rising home values in the 1990s to mid-2000s, also spent more.
[Greenspan and Kennedy, 2008] estimated the sources and uses of equity extracted from homes
and found that the amount of equity extracted used for personal consumption expenditure rose
from USD 26.3 billion in 1991 to USD 182.7 billion in 2005, or from about 10 percent of free cash
resulting from equity extraction to about 13 percent.

Some studies take a completely different direction and look at the behavioral of payment mecha-
nisms and the action of paying per se. [Soman, 2001] argues that the medium of payment, such as
cash and check, affect consumer’s future spending behaviors through two mechanisms: (1) rehearsal
which causes consumers to recall past expenses more accurately, and (2) immediacy or an imme-
diate depletion of wealth, both of which will make consumers more averse to spending. The use of
credit cards, however, does not bring such negative impact on spending and therefore leads to less
aversion to spending. In a subsequent field study, [Soman, 2003] collected receipts from shoppers at
the exit of a large supermarket store and found that the positive effect of credit card use on spend-
ing was mainly on the purchase of flexible items (an expense which may vary depending on price
and quantity available), but not on inflexible goods (which are needed irrespective of changes in
price and other factors). Another explanation is the payment transparency hypothesis, i.e., credit
cards and other payment tools are different in the transparency or vividness with which individ-
uals can feel the outflow of money, and of which cash is the most transparent mode of payment.
The more transparent the payment outflow, the greater the aversion to spending or the higher
the pain of paying [Prelec and Loewenstein, 1998]. Similarly, [Raghubir and Srivastava, 2008] find
the same effect in an experimental study where they asked participants to estimate the budget for
a hypothetical thanksgiving party where the specified payment medium was cash or credit card.
They find that estimates of the total cost of the party were significantly higher when credit card
was the payment medium. Interestingly, when participants were instructed to consider the cost of
each item individually and add them up, there was no difference between costs from using cash
and credit card.

Taking a different approach, behavioral economists exploit natural experiments to understand how
income shocks and other stimulus affect consumption spending and in what ways. [Agarwal et al., 2007]
estimate how consumers responded to the 2001 Federal income tax rebates in terms of monthly
response of credit card payments, spending and debt. They find that consumers initially increased
their credit card payments, i.e., pay down debt, but soon afterwards increased their spending.
Spending increased most for consumers who were most likely to be liquidity constrained, while
debt declined most for unconstrained consumers. In a more recent study, [Agarwal and Qian, 2014]
study consumption behavior after an exogenous unanticipated income shock - a one-time cash
payment to Singaporeans announced on February 2011 and paid out at the end of April 2011.
Estimating the announcement and disbursement effects, they find that consumption rose signifi-
cantly after the fiscal policy announcement with consumers spending $0.80 on average for every $1
received during the subsequent 10 months. They also find a strong announcement effect, i.e., 19
percent of the response occurs during the first two-month of the announcement period via credit
cards. Interestingly, consumers switched to using debit cards after disbursement before finally
increasing spending on credit cards in later months. Consumers with low liquid assets or with low
credit card limit experienced stronger consumption responses.

While unexpected income shocks from national policy are a good source of natural experiments,
another set of studies draw on the changes within day-to-day living to understand how predictable
income streams that households expect affect their spending behavior. Empirical evidence shows
that even anticipated incomes shock cause changes in household spending behavior. [Parker, 2015]

6
estimate the propensity of households spending in response to the arrival of predictable, lump-sum
payments using households in the Nielsen Consumer Panel who received USD25 million in Federal
stimulus payments that were distributed randomly across weeks. He finds that the propensity
to spend is a persistent household trait. It is unrelated to expectation errors, almost unrelated
to crude measures of procrastination and self-control, moderately related to sophistication and
planning, and highly related to impatience. On the other hand, [Stephens, 2008] estimates the
consumption response to predictable increases in discretionary income following the final payment
of a vehicle loan and find that a 10 percent increase in discretionary income leads to 2-3 percent
increase in nondurable consumption. This may be due to borrowing constraints. In an earlier
paper, [Stephens, 2006] examines consumption in response to paycheck arrival. Contrary to the
basic rational expectations life-cycle/permanent income hypothesis, which predicts that household
consumption should not respond to anticipated paycheck arrival since a regular paycheck does not
provide new information, he finds household consumption to be excessively sensitive to paycheck
receipt. The finding cannot be explained by any underlying monthly expenditure fluctuations
common to all households, while liquidity constraints measured by wealth and age can account for
the excess sensitivity results.

3 Factors that affect consumer suboptimal financial choice

and behavior
Poor financial decisions have material impact on households’ lifetime welfare and some households
clearly make better decisions than others. Small deviations from optimal decision-making at the
consumer level might still potentially have important aggregate implications e.g., [Gabaix and Laibson, 2006].
Therefore, understanding why consumers behave sub-optimally has important policy implications,
since sub-optimal choices and biases can dampen economic growth and increase the costs for social
security systems. The main drivers explaining consumer choice inefficiencies are: cognitive limita-
tions; financial literacy; time preference; self-control and incentives; psychological bias; and social
networks.

3.1 Cognitive limitations and financial literacy

One strand of the literature on suboptimal consumer financial decision focuses on the role of
cognitive limitations, either from innate characteristics [Agarwal and Mazumder, 2013] or ageing
[Agarwal et al., 2009a]. [Agarwal and Mazumder, 2013] explicitly link cognitive abilities to finan-
cial mistakes. Specifically, they study the effect of cognitive abilities on two consumer financial
decisions where suboptimal behavior is well-defined – (1) the optimal use of credit cards for “con-
venience” transactions when transactions are fully paid for within the grace period, and after a
balance transfer from an existing account to a new card, and (2) a financial mistake that involves
recognizing the APR for the value of their home during a home equity loan application. They
use a direct measure of cognitive abilities - the Armed Forces Qualifying Test - to show that con-
sumers with higher math scores are substantially less likely to make financial mistakes. Other
studies by [Agarwal et al., 2009a] and [Korniotis and Kumar, 2011] explore the effects of aging
and experience on financial decision-making, and find that the adverse effects of aging dominate
the positive effects of experience. To understand the effect of aging on financial decision-making,
[Agarwal et al., 2009a] study lifecycle patterns in financial mistakes using a proprietary database
that measures ten different types of credit behavior. These mistakes include the suboptimal use of
credit card balance transfer offers, the mis-estimation of the value of one’s home and excess interest
rates and fee payments. They find that middle-aged adults make fewer financial mistakes than
younger or older adults, while implies a U-shaped pattern with the cost-minimizing age occurring
around age 53. [Korniotis and Kumar, 2011] document that while older and more experienced
investors have greater knowledge about investment, they do not effectively apply their knowledge
and skills as a result of cognitive deterioration.

Attention and a bias to action is another explanation for better financial decision making. [Andersen et al., 2015]
studied inattention and inertia in Danish households and found that younger, better-educated and
higher-income households have less inertia and less inattention. Financial wealth and housing
wealth appear to have opposite effects – there is least inertia and inattention among households

7
whose housing wealth is high relative to their financial wealth. In their mixture model of household
refinancing types, household characteristics affect both inattention (a low proportion of rational
refinancers) and residual inertia (a low probability that fully inattentive households refinance).
They found that the two attributes are positively correlated with cross-sectional correlation of 0.62.

Another explanation for why consumers make financial mistake is that they lack sufficient knowl-
edge about financial concepts and instruments to make informed financial decisions [Agarwal et al., 2010].
[Gerardi et al., 2013] find a robust relationship between numerical ability and mortgage default,
after controlling for a broad set of sociodemographic variables which are not driven by other as-
pects of cognitive ability. They find that while numerical ability does not impact the choice of
mortgage contract, but affects incidence of mortgage default - individuals with limited numerical
ability default on their mortgage due to behavior unrelated to the initial choice of mortgage. The
good news is that there is some evidence showing that financial decision-making does improve with
experience. For example, [Agarwal et al., 2016] show evidence that borrowers learn from their re-
financing experiences and make fewer mistakes on their second refinancing.

Other research show that many consumers are ill prepared to meet their financial goals [Agarwal et al., 2010],
take out payday loans at astronomical interest rates when cheaper forms of credit are available
[Agarwal et al., 2009b], choose sub-optimal credit contracts [Agarwal et al., 2015b], fail to opti-
mally refinance mortgages [Agarwal et al., 2013], and fail to plan for retirement, reaching it with
little or no savings [Lusardi et al., 2009].

Another strand of research focuses on the importance of improving a consumers’ financial liter-
acy. [Hilgert et al., 2003] explore the connection between financial knowledge and behavior using
consumer survey data. They find that those who have more financial knowledge are more likely
to engage in recommended financial practices. They conclude that financial education, when com-
bined with skill-building and audience-targeted motivational strategies, may help drive the desired
behavioral changes in the way consumers manage their finances. [Agarwal et al., 2010] also find
substantially lower default rates among graduates of a long-term voluntary counseling program
targeting low- to moderate-income households. They attribute the results to two factors. First,
the program requires prospective borrowers to acquire budgeting and credit management skills.
Second, the aggressive post-purchase counseling that targeted early delinquency – counselors iden-
tity and target households based on soft information picked up during the program. Moreover,
[Fernandes et al., 2014] conduct a meta-analysis of the relationship of financial literacy and of fi-
nancial education to financial behaviurs in 168 papers covering 201 prior studies, and find that
interventions to improve financial literacy explain only 0.1 percent of the variance in financial be-
haviors studied, and their effects decay over time. Even large interventions with many hours of
instruction have negligible effects 20 months after the time of intervention and that the partial
effects of financial literacy diminish dramatically with controls for psychological traits omitted in
prior studies. They suggest that “just in time” financial education tied to specific behaviors it
intends to help would be more effective than financial education that is not elaborated or acted
upon soon afterwards.

3.2 Time preferences, self-control and incentives

The second strand of research tries to explain why consumer choices are inefficient relate to time
preferences, self-control and incentives. Present bias, where a decision-maker places more empha-
sis on the present than in the future, has also been documented in the literature. Present-biased
preferences can be seen as a result of the interplay between two separate decision making systems:
the affective system, which values immediate gratification and sharply discounts all future peri-
ods; and the deliberative system, which makes long-run plans and displays higher discount factors.
[Laibson, 1997] develops a theoretical framework showing how present bias can drive credit card
borrowing and can constrained future choices. [Laibson et al., 2003] present a model of hyper-
bolic discounting to explain the co-existence of credit card debt and accumulation of assets for
retirement. The accumulation of illiquid assets co-exists with revolving credit card debts because
hyperbolic consumers act patiently with respect to longer-run objectives and impatiently with
short-run objectives. Based on this, they propose that credit card borrowing is intended for short-
term consumption smoothing, while accumulation of illiquid assets is an instrument of self-control.

8
Empirical support of present bias includes [Shui and Ausubel, 2004] who as stated earlier finds
that consumers prefer an introductory offer with a lower interest rate and shorter duration to one
with a higher interest rate with a longer duration, even though they would benefit more from choos-
ing the latter, and that consumers are reluctant to switch and many consumers who have switched
before fail to switch again. These findings suggest time inconsistency in consumer behavior, which
can be explained with hyperbolic preferences. Using a unique field study on incentivized choice
experiment and administrative credit card borrowing data, [Meier and Sprenger, 2010] find that
present-biased individuals are more likely to hold a credit card debt. Finding by [Kuchler, 2015]
indicate that the level of impatience affects a consumer’s decision to pay-down debt highlights the
importance of present bias in consumer decisions.

Other papers such as [Bertaut et al., 2009], [Telyukova, 2013] identify a set of motives related
to self-control. [Bertaut et al., 2009] show the co-existence of credit card debt with substantial
accumulation of assets for retirement in the context of self-checking motive using the “accountant-
shopper” model where an individual or household have two separate selves – an “accountant” and
a “shopper” who act contemporaneously to handle different decisions. The accountant handles bill
payment and long-term financial planning while the shopper determines consumption expenditures
based on the credit available to him. When the shopper exhibits self-control problems, the accoun-
tant intervenes to control the shopper by limiting the unused credit available. This explanation,
however, has limitations; it is idealistic to think that the accountant would consistently act to check
the shopper in a rational manner. The household puzzle itself is a reflection of this limitation - it
has been shown that households lose an average of $374 per year or 1.5 percent of the total annual
after-tax income. [Telyukova, 2013] offers a precautionary motive and highlights the importance
of liquidity - households are not using money in the bank to pay off accumulated credit card debt
because they anticipate needing that money in situations where credit cards cannot be used. As a
result, households consume both cash goods and credit card goods concurrently.

In their study, [Agarwal et al., 2015b] also find that the probability of choosing the suboptimal
contract declines with the dollar magnitude of the potential error, and consumers with larger errors
were more incentivise to subsequently switch to the optimal contract. [Lehnert and Maki, 2007]
offer strategic default as another motive. They study consumer bankruptcy behavior and find that
states with higher level of statutory exemptions – below which debtors are permitted to keep their
assets while debts are forgiven – face higher consumer bankruptcy rates and households that are
more likely to simultaneously hold low-return liquid assets and owe high-cost unsecured debt. They
also find that the credit card debt puzzle is more prevalent in the U.S. where exemption levels are
higher, and this supports their hypothesis. This, however, provides a plausible motive for only a
limited group of consumers since most people are unlikely to file for bankruptcy.

3.3 Psychological Bias

A third strand of literature identifies and assesses the role of psychological biases – perceived control
[Perry and Morris, 2005], optimism [Puri and Robinson, 2007], disposition [Odean, 1998], narrow
framing [Kumar and Lim, 2008], propensity to gamble [Kumar, 2009] among others, as well as the
extent to which smartness is a determinant of psychological biases [Korniotis and Kumar, 2013].

One key factor that influences propensity to save, to budget and to control spending is perceived
control. [Perry and Morris, 2005] examine the relationship between consumer financial knowledge,
income and locus of control on financial behavior. They find that consumers who perceived they
have higher levels of control over outcomes as well as knowledge and financial resources, are more
likely to save, budget and control spending. Optimism, which is correlated with positive beliefs
about future economic conditions and related to numerous work and life choices, is also a key influ-
encing factor, albeit in moderate amounts. [Puri and Robinson, 2007] develop a novel measure of
optimism by comparing self-reported life expectancy in the Survey of Consumer Finance with that
implied by statistics. They find that moderate optimists were found to display reasonable financial
behavior while extreme optimists displayed financial habits and behavior that are generally not
considered to be prudent.

9
[Kumar, 2009] identifies the propensity to gamble. The author finds that greater expenditure
in lotteries are associated with greater investment in lottery-type stocks. Similarly, state lotteries
and lottery-type stocks are also likely to attract very similar socioeconomic clienteles.

While we know that psychological biases exist to various extents in different people, it is important
to also understand what drives this variation. [Korniotis and Kumar, 2013] develop an empirical
model of smartness that can identify skilled investors ex ante using only their demographic char-
acteristics. They then investigate the role of psychological mechanisms and information allocation
on some commonly known stock market puzzles, namely portfolio concentration, excessive trading
and preference for local stocks. They find that smartness is a determinant of various psychological
biases, i.e., smart investors have less psychological biases, and therefore perform better. But what
factors determine smartness? [Korniotis and Kumar, 2013] find that superior information is the
key factor behind the performance of “smart” investors, even when their decisions are contrary
to standard economic reasoning. [Zhang, 2006] also showed that higher information uncertainty,
resulting from poor information and high volatility of a firm’s fundamentals, increases behavioral
biases and investors under-reaction to new public information.

3.4 Social Networks

The literature on information accessed and transmitted highlights the role of social networks and
peer pressure on financial decision making. [Hong et al., 2004] investigate how social interaction
influence stock market participation and conclude that more social consumers, for example those
who have reported interacting with neighbors and attending church, are more likely to invest.
[Hong et al., 2005] also find that a mutual fund manager is more likely to buy or sell a particular
stock if other managers in the same city are buying or selling the same stock even after controlling
for local preferences. They explain this effect using an epidemic model where investors spread
information to one another by word of mouth.

Similarly, [Bailey et al., 2016] use data from Facebook to show that social interactions with friends
can influence housing market expectations and investment behavior. Specifically, they show that
individuals whose geographically distant friends experienced larger recent house price gains are
more optimistic about property investments in their own local housing markets, and are actually
more likely to invest in the housing market. [Duflo and Saez, 2003] conduct a field experiment and
found that not only individuals who received a financial incentive to participate in a fair providing
information on TDA retirement plans were more likely to enroll in the plan, their colleagues were
also more likely to enroll – evidence of spill-over effects based on social interactions. Yet, when
[Grinblatt and Keloharju, 2001] analyze the relationship between spatial proximity, language and
culture and stock-holding, and they find that the influence of distance, language and culture is
less prominent among the most investment-savvy investors than among households and less savvy
investors.

4 Financial product design and marketing

The design of financial products and how they are marketed are also important factors that influ-
ence consumer choice – the behavioral literature has ample studies on the effects of frames on choice
[Tversky and Kahneman, 1986]. [Shefrin and Statman, 1993] also suggest that some investors pre-
fer one financial product over another because of the way identical cash flows are framed. They
studied the role of prospect theory, hedonic framing, behavioral life cycle theory and cognitive
errors in the design of financial products, and suggested that a behavioral framework can offer in-
sights into many other features of security design such as designation of mutual funds as “growth”
and “income” funds, design of lotteries to have particular combination of prices and design of stocks
to have prices within a “trading range”. [Gabaix and Laibson, 2006] study information shrouding
in an economy with some myopic (or less informed) consumers. They find that, in equilibrium,
optimizing firms exploit myopic consumers by shrouding high priced add-ons, while sophisticated
customers were able to exploit these marketing schemes. They also show that the practice con-
tinues because it is not possible to profitably drive away the business of the sophisticated or the
myopic.

10
There is also evidence that the way mortgages are marketed to consumers have important im-
plications. Using a controlled experiment in a large US commercial bank when loan officers were
incentivized to prospect for loans, [Agarwal and Ben-David, 2014] find that loan sizes, volume and
default increased despite the fact that there was no change to the bank’s credit standards. They
find that loan officers placed greater weight on hard information and approved many applications
that would otherwise have been rejected, and preferred to convince existing applicants to borrow
larger amounts than source new applications. This resulted in a higher default rate and the loss of
predictive power of the bank’s credit model. [Gurun et al., 2016] study the relationship between
advertising sub-prime lenders and the nature of mortgages obtained by consumers and measured
the relative expensiveness of a given mortgage, i.e., the excess rate after accounting for a broad
set of borrower, contract and regional characteristics. They find a strong positive relationship
between the intensity of local advertising and expensiveness of mortgage extended by lenders in
a given region, with the strongest being advertising through newspapers - the most heavily used
advertising channel. They also found that advertising is most effective when targeted at groups
that are likely to be less informed about mortgages such as the poor, less educated and minorities.
Their findings support the “persuasive view” that advertising in the sub-prime mortgage market
was used to steer consumers into expensive choices, rather than the “informative view” of adver-
tising posits that advertising helps consumers to find cheaper products.

In addition, consumers also face a number of challenges - the complexity of financial products, infor-
mation asymmetry and perhaps less scrupulous practices on the part of providers. Examples include
excessive fees, high interest rates, prepayment penalties, and clauses barring borrowers from seeking
judicial redress for predatory behavior by lenders [Engel and McCoy, 2001]. [Bostic et al., 2012]
find evidence that anti-predatory lending laws, holding all else constant, changed the type of mort-
gage products lenders offer in the market. Specifically, the introduction of state anti-predatory
lending laws induced lenders to substitute between mortgage products. Specifically, the laws
reduced mortgage products that have prepayment penalties, adjustable rate, hybrid rate, balloon
payment, interest only, and low and no documentation loans in locations in which an anti-predatory
lending law was introduced, and at the same time increased teaser rates in order to extend the
period during which borrowers would have low monthly mortgage payments. [Agarwal et al., 2013]
provide explicit evidence of systematic mortgage lending abuse such that good credit quality
customers are steered towards more expensive predatory sub-prime mortgage terms. Moreover,
[Agarwal et al., 2014] present evidence that predatory lending practices contributed to the higher
mortgage default rates among sub-prime borrowers, raising them by about a third, and aggravating
the sub-prime crisis. While it is hard to disagree that mortgages with abusive terms are costly to
borrowers and taxpayers, the extent of these practices is hard to quantify.

In addition, [Agarwal and Ben-David, 2014] also presents evidence that when a large US com-
mercial bank incentivised loan officers to prospect for new business, the unintended consequences
were a dramatic loss of critical soft information, larger loans, poorer credit quality and higher
default rates, i.e., the change in the bank’s business model unintentionally led to risker lending at
consumers’ expense. In the area of credit cards, the complexity stems from the great variety of
rates and fees to cater to diverse consumer risk profiles and preferences. The innovation that gave
less credit-worthy consumers access to credit cards have also made it more difficult for them to
fully understand the true cost of credit and therefore more challenging to select and manage the
use of credit cards [Canner and Elliehausen, 2013].

How then can we design financial products for the benefit of consumers? Specifically on mort-
gage debt, [Campbell, 2013] studies cross-country variation in mortgage market structure and
draws insights from urban economics, asset pricing, behavioral finance, financial intermediation
and macroeconomics to improve on mortgage market design, especially in the US. The paper finds
that three types of behavioral heterogeneity impact mortgage lenders’ incentives and mortgage
innovations: moving propensity, financial sophistication and present-biased preferences. The pa-
per also concluded that while it may be possible for economists to recommend an ideal mortgage
system by solving the dynamic contracting problem, mortgage market design must proceed in a
more ad hoc and flexible way, learning from international experience and integrating insights from
different fields.

11
5 Regulation
While the financial crisis has triggered a surge of interest in regulating consumer financial prod-
ucts, researchers are still debating both sides of the regulatory coin, for example, in the studies
of [Campbell et al., 2011] and [Posner and Weyl, 2013]. Proponents of regulation argue that con-
sumer financial markets have become increasingly unfair. Firms take advantage of consumers’ be-
havioral biases — such as myopia, present bias, and inattention — to earn large profits, especially
from unsophisticated and poor consumers [Agarwal et al., 2015c]. They suggest that regulation
and additional information can protect less sophisticated consumers and reduce aggregate borrow-
ing costs. Banks also have the least incentive to lend to consumers who are most in need of credit
[Agarwal et al., 2015c]. Critics however are skeptical about the effectiveness of consumer financial
regulations. While limits on hidden fees, for example, can shift surplus from more sophisticated to
less sophisticated consumers [Gabaix and Laibson, 2006], there is less evidence that regulators can
bring about an across-the-board reduction in consumer costs. They suggest that regulators may
just be playing a game of regulatory whac-a-mole, where efforts to limit certain fees will simply
lead firms to offset reduced revenue with higher prices on other product dimensions as well as to
restrict the supply of credit. Even proponents of regulating late fees, such as [Barr et al., 2009],
are worried that “the reduced revenue stream to lenders from these fees would mean that other
rates and fees would be adjusted to compensate.”

[Campbell, 2016] framed the regulation discussion in the context of the tension between the lais-
sez faire and the interventionists. Household finance is the focus of interventionist attention for
several reasons: (1) individuals have to make increasingly difficult financial decisions with bigger
consequences, (2) higher education is becoming more expensive and posing a challenge in countries
where it is not publicly provided, (3) rising home prices are stressing traditional systems for financ-
ing homeownership, (4) improving information technology is facilitating the development of more
complex and confusing products, and (5) behavioral economics has opened the eyes of academics
to financial behaviors that were not carefully examined before, and that many households are not
up to the challenge of managing their finances. Also, household mistakes may not be purely id-
iosyncratic and can be correlated across households, which create endogenous risk and therefore
needs to be managed by the financial system.

The real questions though are: How effective has regulation been so far? What types of regu-
lation work? How do we improve upon the way we regulate to balance the needs of consumers
and other stakeholders alike? How do we make use of behavioral elements to make regulation
more effective? [Agarwal et al., 2015c] analyzed the effectiveness of consumer financial regulation
by considering the 2009 Credit Card Accountability Responsibility and Disclosure (CARD) Act.
They estimated that regulatory limits on credit card fees reduced overall borrowing costs by an
annualized 1.6 percent of average daily balances, with savings of more than 5.3 percent for con-
sumers with FICO scores below 660, and no evidence of an offsetting increase in interest charges or
a reduction in the volume of credit. This means that the CARD Act saved consumers an estimated
USD11.9 billion per year. [Agarwal et al., 2015d] further provided a framework to help regulators
understand implications of regulating hidden fees by estimating firms’ pass-through of cost shocks
and salience of the regulated hidden fee. They show that the framework predicted the degree of
offset from the CARD Act in an ex-ante analysis. In terms of interventions, [Agarwal et al., 2015d]
also analysed a nudge that disclosed the interest savings from paying off balances in 36 months
rather than making minimum payments and detected a small increase in the share of accounts
making the 36-month payment value but no evidence of a change in overall payments.

The results of another study show the nature and degree of substitutability between debit and
credit, which has implications for antitrust regulators. [Zinman, 2009] modeled consumer choice
at the point of sales and found that consumers who revolve debt or face a binding credit limit con-
straint are substantially less likely to incur credit card charges and substantially more likely to use a
debit card, conditional on several proxies for transaction demand and tastes. The paper also found
that debit use increases with credit limit constraints and decreases with credit card possession, and
that debit is becoming a stronger substitute for credit over time. [Heidhues and Kőszegi, 2010]
studied how requiring credit contracts to have a linear structure and prohibiting large penalties
for deferring small amounts of repayment can raise welfare. They analyzed contract choices, loan

12
repayment behavior and welfare in a model of a competitive credit market when borrowers have
a taste for immediate gratification. This is consistent with many credit cards and sub-prime
mortgages where terms are such that most non-sophisticated borrowers over borrow and pay the
penalties to back load repayment, causing welfare loss.

More generally, [Agarwal et al., 2009a] analysed nine regulatory strategies that can help individ-
uals avoid financial mistakes, especially due to cognitive decline, such as: laissez-faire, disclosure,
nudges, financial “drivers’ licenses”, advance directives, fiduciaries, asset safe harbors and ex post
and ex ante regulatory oversight. They also asked what the appropriate regulatory response should
be. They believed that if the market for third-party advice and fiduciary services functioned well,
the market equilibrium should have three phases, including: (1) Early in life, when each individual
would write a plan for his or her future consumption and investment, contingent on major events
(including cognitive decline); (2) In the second phase, cognitive testing and observation would
monitor the individual for onset of significant cognitive decline; (3) Finally, when this pre-specified
threshold is crossed, the individual’s plan would be enforced by a fiduciary, or the individual’s
assets would be placed in a financial instrument with a state-contingent payout scheme. The
market already provides financial products with this feature, for example annuities which elimi-
nate complex asset de-accumulation decisions. They believed however that an unregulated market
solution may not work well and government intervention is probably needed. The ideal form of
intervention is yet unclear and more empirical analyses and field experiments are needed to iden-
tify the regulatory response that best balances the marginal costs against the potential benefits.
[Campbell et al., 2011] argued that regulation must be tailored to specific problems to be beneficial
and must be accompanied by research to measure effectiveness of the interventions. This is espe-
cially the case for consumer financial regulation on market failure and limited consumer rationality
in financial decision making. They illustrate the need for, and limits of, regulation through three
case studies – of mortgage markets, payday lending and financial retirement consumption.

Yet another strand of literature explored the ability of government to induce change and find
this avenue to be limited. [Agarwal et al., 2015a] examined the ability of the government to in-
fluence debt renegotiation by evaluating the effects of the 2009 Home Affordable Modification
Program that provided intermediaries with sizable financial incentives to renegotiate mortgages.
They found that the program generated an increase in renegotiation intensity within the program
while adversely affecting negotiations performed outside the program. The overall impact of the
program is likely to be limited – it was estimated to reach just one-third of the targeted 3-4 mil-
lion households – with a few large servicers responding at half the rate of others. This is likely
due to servicer specific factors, for example pre-existing organizational capabilities. These find-
ings suggested that the government’s ability to quickly induce changes in servicer behavior through
financial incentives is limited and highlights significant barriers to the effectiveness of such policies.

What is the main challenge for economists then in the area of consumer financial regulation?
[Campbell, 2016] succinctly states “beyond the easy cases where behaviorally biased households
can be “nudged” to avoid mistakes with minimal effects on rational households, financial regula-
tors face a difficult trade-off between the benefits of regulation to households that make mistakes,
and the costs of regulation to other financial market participants [emphasis added]. The task for
economists is to confront this trade-off explicitly, bring to bear the highest quality evidence that
modern applied microeconomics can make available”.

6 Conclusion
In this paper, we reviewed the developments that influence consumer choice and financial products
within the field of household finance. Specifically, we looked at consumer behavior and biases,
financial literacy and education, financial product design and marketing, regulation and consumer
fintech and highlight the key research questions to explore further. In this respect, we outline five
future research directions and considerations for policy makers.

First, how widespread and important are losses due to poor financial decision-making? What
fraction of aggregate wealth, and of the wealth of older adults, is lost because of poor choices?
What are the costs? We believe that the studies so far are just looking at the tip of the iceberg;

13
the bulk of mistakes that are being made probably lie elsewhere, including decisions regarding
annuities, structured financial products, real estate investment pools and other complex and/or
retirement related concerns.

Second, while we have found that there are many consumer behaviors and biases, both ratio-
nal and irrational, at play, the relative effects are still unclear? Particularly, do we know which
the dominant effects are? Is it possible to come up with a general model of consumer financial
behavior? Can this be a useful reference to guide financial product design and regulation, much like
how Apple shapes our smartphone usage habits and Starbucks shapes our coffee drinking culture?

Third, policy research so far has focused on the interventions and their effects on consumer welfare,
as well as how to improve consumer financial literacy. Perhaps attention should also be given to
studying financial services providers’ incentives. Rather than rely on a “stick” approach, can we
create the conditions for a sustainable win-win for consumer and provider?

Fourth, policy researchers also face fascinating challenges in deriving optimal contracts, policy re-
sponses and enforcement [Zinman, 2014] due to inherent behavioral complexity, information asym-
metry and enforcement idiosyncrasy. What new methodologies and further research do we need to
improve on current consumer financial policies?

Finally, as adoption of fintech grows in consumer finance, consumers will decide how to use these
new technologies while regulators will need to decide how to react to these new developments.
As these technologies become entrenched and co-exist with traditional financial services, how will
consumer behavior and market dynamics change? There will be the need to re-examine how each
of the levers – consumer behavior and biases, financial literacy and education, financial product
design and marketing and regulation – have adapted or need to be adapted.

References
[Agarwal et al., 2010] Agarwal, S., Amromin, G., Ben-David, I., Chomsisengphet, S., and Evanoff,
D. D. (2010). Learning to cope: Voluntary financial education and loan performance during a
housing crisis. American Economic Review, 100(2):495–500.
[Agarwal et al., 2014] Agarwal, S., Amromin, G., Ben-David, I., Chomsisengphet, S., and Evanoff,
D. D. (2014). Predatory lending and the subprime crisis. Journal of Financial Economics,
113(1):29–52.

[Agarwal et al., 2015a] Agarwal, S., Amromin, G., Ben-David, I., Chomsisengphet, S., Piskorski,
T., and Seru, A. (2015a). Policy intervention in debt renegotiation: Evidence from the home
affordable modification program. Journal of Political Economy (forthcoming).
[Agarwal and Ben-David, 2014] Agarwal, S. and Ben-David, I. (2014). Loan prospecting and the
loss of soft information. (19945).

[Agarwal et al., 2015b] Agarwal, S., Chomsisengphet, S., Liu, C., and Souleles, N. S. (2015b). Do
consumers choose the right credit contracts? The Review of Corporate Finance Studies, 4(2):239.
[Agarwal et al., 2015c] Agarwal, S., Chomsisengphet, S., Mahoney, N., and Stroebel, J. (2015c).
Do banks pass through credit expansions to consumers who want to borrow? (21567).

[Agarwal et al., 2015d] Agarwal, S., Chomsisengphet, S., Mahoney, N., and Stroebel, J. (2015d).
Regulating consumer financial products: Evidence from credit cards. The Quarterly Journal of
Economics, 130(1):111.
[Agarwal et al., 2009a] Agarwal, S., Driscoll, J., Gabaix, X., and Laibson, D. (2009a). The age of
reason: Financial decisions over the life cycle and implications for regulation. Brookings Papers
on Economic Activity, 40(2 (Fall)):51–117.
[Agarwal et al., 2013] Agarwal, S., Driscoll, J. C., and Laibson, D. I. (2013). Optimal mortgage
refinancing: A closed-form solution. Journal of Money, Credit and Banking, 45(4):591–622.

14
[Agarwal et al., 2007] Agarwal, S., Liu, C., and Souleles, N. S. (2007). The reaction of consumer
spending and debt to tax rebates—evidence from consumer credit data. Journal of Political
Economy, 115(6):986–1019.
[Agarwal and Mazumder, 2013] Agarwal, S. and Mazumder, B. (2013). Cognitive abilities and
household financial decision making. American Economic Journal: Applied Economics, 5(1):193–
207.
[Agarwal and Qian, 2014] Agarwal, S. and Qian, W. (2014). Consumption and debt response
to unanticipated income shocks: Evidence from a natural experiment in singapore. American
Economic Review, 104(12):4205–30.
[Agarwal et al., 2016] Agarwal, S., Rosen, R. J., and Yao, V. (2016). Why do borrowers make
mortgage refinancing mistakes? Management Science, 62(12):3494–3509.
[Agarwal et al., 2009b] Agarwal, S., Skiba, P. M., and Tobacman, J. (2009b). Payday loans and
credit cards: New liquidity and credit scoring puzzles? American Economic Review, 99(2):412–
17.
[Akerlof and Yellen, 1985] Akerlof, G. A. and Yellen, J. L. (1985). A near-rational model of the
business cycle, with wage and price inertia. The Quarterly Journal of Economics, 100(Supple-
ment):823.
[Amromin, 2003] Amromin, G. (2003). Household portfolio choices in taxable and tax-deferred
accounts: another puzzle? European Finance Review, 7(3):547–582.
[Andersen et al., 2015] Andersen, S., Campbell, J. Y., Nielsen, K. M., and Ramadorai, T. (2015).
Inattention and inertia in household finance: Evidence from the danish mortgage market.
(21386).
[Andersson et al., 2013] Andersson, F., Chomsisengphet, S., Glennon, D., and Li, F. (2013). The
changing pecking order of consumer defaults. Journal of Money, Credit and Banking, 45(2-
3):251–275.
[Arnberg and Barslund, 2014] Arnberg, S. and Barslund, M. (2014). The crowding-out effect of
mandatory labour market pension schemes on private savings: Evidence from renters in denmark.
[Attanasio and DeLeire, 1994] Attanasio, O. P. and DeLeire, T. C. (1994). Iras and household
saving revisited: Some new evidence. (4900).
[Ausubel, 1991] Ausubel, L. M. (1991). The Failure of Competition in the Credit Card Market.
American Economic Review, 81(1):50–81.
[Badarinza et al., 2015] Badarinza, C., Campbell, J. Y., and Ramadorai, T. (2015). What calls
to arms? international evidence on interest rates and the choice of adjustable-rate mortgages.
Management Science (Forthcoming).
[Bailey et al., 2016] Bailey, M., Cao, R., Kuchler, T., and Stroebel, J. (2016). Social networks and
housing markets. (22258).
[Barr et al., 2009] Barr, M., Mullainathan, S., and Shafir, E. (2009). The Case for Behaviorally
Informed Regulation, pages 25–62. The Tobin Project, Cambridge.
[Benjamin, 2003] Benjamin, D. J. (2003). Does 401(k) eligibility increase saving?: Evidence from
propensity score subclassification. Journal of Public Economics, 87(5–6):1259 – 1290.
[Bertaut et al., 2009] Bertaut, C. C., Haliassos, M., and Reiter, M. (2009). Credit card debt
puzzles and debt revolvers for self control. Review of Finance, 13(4):657–692.
[Beshears et al., 2015] Beshears, J., Choi, J. J., Laibson, D., and Madrian, B. C. (2015). Does
front-loading taxation increase savings? evidence from roth 401(k) introductions. Journal of
Public Economics, pages –.
[Bhutta et al., 2010] Bhutta, N., Dokko, J. K., and Shan, H. (2010). The depth of negative equity
and mortgage default decisions. (2010-35).

15
[Bostic et al., 2012] Bostic, R., Chomsisengphet, S., Engel, K., McCoy, P., Pennington-Cross, A.,
and Wachter, S. (2012). Mortgage Product Substitution and State Anti-predatory Lending Laws:
Better Loans and Better Borrowers? Atlantic Economic Journal, 40(3):273–294.
[Botsch and Malmendier, 2015] Botsch, M. and Malmendier, U. (2015). Inflation experiences and
contract choice—evidence from residential mortgages. unpublished paper.
[Brady et al., 2000] Brady, P. J., Canner, G. B., and Maki, D. M. (2000). The effects of recent
mortgage refinancing. Federal Reserve Bulletin, (Jul):441–450.
[Campbell, 2006] Campbell, J. Y. (2006). Household finance. The Journal of Finance, 61(4):1553–
1604.
[Campbell, 2013] Campbell, J. Y. (2013). Mortgage market design*. Review of Finance, 17(1):1.
[Campbell, 2016] Campbell, J. Y. (2016). Restoring rational choice: The challenge of consumer
financial regulation. American Economic Review: Papers & Proceedings, 106(5):1–30.
[Campbell and Cocco, 2003] Campbell, J. Y. and Cocco, J. F. (2003). Household risk management
and optimal mortgage choice. The Quarterly Journal of Economics, 118(4):1449.
[Campbell and Cocco, 2015] Campbell, J. Y. and Cocco, J. F. (2015). A model of mortgage default.
Journal of Finance, 70(4):1495–1554.
[Campbell et al., 2011] Campbell, J. Y., Jackson, H. E., Madrian, B. C., and Tufano, P. (2011).
Consumer financial protection. Journal of Economic Perspectives, 25(1):91–114.
[Canner et al., 2002] Canner, G. B., Dynan, K. E., and Passmore, W. (2002). Mortgage refinancing
in 2001 and early 2002. Federal Reserve Bulletin, (Dec):469–481.
[Canner and Elliehausen, 2013] Canner, G. B. and Elliehausen, G. E. (2013). Consumer Experi-
ences with Credit Cards. Federal Reserve Bulletin, (December):1–36.
[Chetty et al., 2014] Chetty, R., Friedman, J. N., Leth-Petersen, S., Nielsen, T. H., and Olsen, T.
(2014). Active vs. Passive Decisions and Crowd-Out in Retirement Savings Accounts: Evidence
from Denmark. The Quarterly Journal of Economics, 129(3):1141–1219.
[Choi et al., 2011] Choi, J. J., Laibson, D., and Madrian, B. C. (2011). $100 bills on the sidewalk:
Suboptimal investment in 401 (k) plans. Review of Economics and Statistics, 93(3):748–763.
[Choi et al., 2001] Choi, J. J., Laibson, D., Madrian, B. C., and Metrick, A. (2001). Defined
contribution pensions: Plan rules, participant decisions, and the path of least resistance. (8655).
[Dammon et al., 2004] Dammon, R. M., Spatt, C. S., and Zhang, H. H. (2004). Optimal as-
set location and allocation with taxable and tax-deferred investing. The Journal of Finance,
59(3):999–1037.
[Deng et al., 2000] Deng, Y., Quigley, J. M., and Van Order, R. (2000). Mortgage terminations,
heterogeneity and the exercise of mortgage options. Econometrica, 68(2):275–307.
[Duflo and Saez, 2003] Duflo, E. and Saez, E. (2003). The role of information and social inter-
actions in retirement plan decisions: Evidence from a randomized experiment. The Quarterly
Journal of Economics, 118(3):815.
[Elul et al., 2010] Elul, R., Souleles, N. S., Chomsisengphet, S., Glennon, D., and Hunt, R. (2010).
What "triggers" mortgage default? American Economic Review, 100(2):490–94.
[Engel and McCoy, 2001] Engel, K. C. and McCoy, P. A. (2001). A tale of three markets: The law
and economics of predatory lending. Tex. L. Rev., 80:1255.
[Feinberg, 1986] Feinberg, R. A. (1986). Credit cards as spending facilitating stimuli: A condi-
tioning interpretation. Journal of Consumer Research, 13(3):348–356.
[Fernandes et al., 2014] Fernandes, D., Lynch, J. G., and Netemeyer, R. G. (2014). Financial
literacy, financial education, and downstream financial behaviors. Manage. Sci., 60(8):1861–
1883.

16
[Gabaix and Laibson, 2006] Gabaix, X. and Laibson, D. (2006). Shrouded attributes, consumer
myopia, and information suppression in competitive markets*. The Quarterly Journal of Eco-
nomics, 121(2):505.
[Gale and Scholz, 1994] Gale, W. G. and Scholz, J. K. (1994). Intergenerational transfers and the
accumulation of wealth. Journal of Economic Perspectives, 8(4):145–160.
[Gelber, 2011] Gelber, A. M. (2011). How do 401(k)s affect saving? evidence from changes in
401(k) eligibility. American Economic Journal: Economic Policy, 3(4):103–22.

[Gerardi et al., 2013] Gerardi, K., Goette, L., and Meier, S. (2013). Numerical ability predicts
mortgage default. 110(28):11267–11271.
[Greenspan and Kennedy, 2008] Greenspan, A. and Kennedy, J. (2008). Sources and uses of equity
extracted from homes. Oxford Review of Economic Policy, 24(1):120–144.

[Grinblatt and Keloharju, 2001] Grinblatt, M. and Keloharju, M. (2001). How distance, language,
and culture influence stockholdings and trades. The Journal of Finance, 56(3):1053–1073.
[Gross and Souleles, 2002a] Gross, D. B. and Souleles, N. S. (2002a). Do liquidity constraints and
interest rates matter for consumer behavior? evidence from credit card data. The Quarterly
Journal of Economics, 117(1):149–185.

[Gross and Souleles, 2002b] Gross, D. B. and Souleles, N. S. (2002b). An empirical analysis of
personal bankruptcy and delinquency. The Review of Financial Studies, 15(1):319.
[Gurun et al., 2016] Gurun, U. G., Matvos, G., and Seru, A. (2016). Advertising expensive mort-
gages. The Journal of Finance, 71(5):2371–2416.

[Heidhues and Kőszegi, 2010] Heidhues, P. and Kőszegi, B. (2010). Exploiting naïvete about self-
control in the credit market. American Economic Review, 100(5):2279–2303.
[Hilgert et al., 2003] Hilgert, M. A., Hogarth, J. M., and Beverly, S. G. (2003). Household finan-
cial management: the connection between knowledge and behavior. Federal Reserve Bulletin,
(Jul):309–322.

[Hirschman, 1979] Hirschman, E. C. (1979). Differences in consumer purchase behavior by credit

card payment system. Journal of Consumer Research, 6(1):58.
[Hong et al., 2004] Hong, H., Kubik, J. D., and Stein, J. C. (2004). Social interaction and stock-
market participation. Journal of Finance, LIX(1):137–163.

[Hong et al., 2005] Hong, H., Kubik, J. D., and Stein, J. C. (2005). Thy neighbor’s portfolio:
Word-of-mouth effects in the holdings and trades of money managers. The Journal of Finance,
60(6):2801–2824.
[Keys and Pope, 2016] Keys, B. J. Pope, D. G. and Pope, J. C. (2016). Failure to refinance.
Journal of Financial Economics, 122(3):482 – 499.

[Korniotis and Kumar, 2011] Korniotis, G. M. and Kumar, A. (2011). Do older investors make
better investment decisions? The Review of Economics and Statistics, 93(1):244–265.
[Korniotis and Kumar, 2013] Korniotis, G. M. and Kumar, A. (2013). Do portfolio distortions
reflect superior information or psychological biases? Journal of Financial and Quantitative
Analysis, 48(1):1–45.
[Kuchler, 2015] Kuchler, T. (2015). Sticking to your plan: Empirical evidence on the role of present
bias for credit card debt paydown.
[Kuchler and Zafar, 2015] Kuchler, T. and Zafar, B. (2015). Personal experiences and expectations
about aggregate outcomes. (748).

[Kumar, 2009] Kumar, A. (2009). Who gambles in the stock market? The Journal of Finance,
64(4):1889–1933.

17
[Kumar and Lim, 2008] Kumar, A. and Lim, S. S. (2008). How do decision frames influence the
stock investment choices of individual investors? Management Science, 54(6):1052–1064.
[Laibson, 1997] Laibson, D. (1997). Golden eggs and hyperbolic discounting*. The Quarterly
Journal of Economics, 112(2):443.
[Laibson et al., 2003] Laibson, D., Repetto, A., and Tobacman, J. (2003). A Debt Puzzle, pages
228–266. Princeton: Princeton University Press.
[Lehnert and Maki, 2007] Lehnert, A. and Maki, D. (2007). Consumption, Debt, and Portfolio
Choice: Testing the Effects of Bankruptcy Law, pages 55–76. Palgrave Macmillan US, New
York.
[Lusardi et al., 2009] Lusardi, A., Mitchell, O. S., and Curto, V. (2009). Financial Literacy among
the Young: Evidence and Implications for Consumer Policy. (91).
[Madrian and Shea, 2001] Madrian, B. and Shea, D. F. (2001). The power of suggestion: Inertia
in 401(k) participation and savings behavior. The Quarterly Journal of Economics, 116(4):1149–
1187.
[Malmendier and Nagel, 2016] Malmendier, U. and Nagel, S. (2016). Learning from inflation ex-
periences. The Quarterly Journal of Economics, 131(1):53.
[Meier and Sprenger, 2010] Meier, S. and Sprenger, C. (2010). Present-biased preferences and
credit card borrowing. American Economic Journal: Applied Economics, 2(1):193–210.
[Odean, 1998] Odean, T. (1998). Volume, volatility, price, and profit when all traders are above
average. The Journal of Finance, 53(6):1887–1934.
[Parker, 2015] Parker, J. (2015). Why don’t households smooth consumption? evidence from a 25
million dollar experiment. (21369).
[Perry and Morris, 2005] Perry, V. G. and Morris, M. D. (2005). Who is in control? the role of
self-perception, knowledge, and income in explaining consumer financial behavior. Journal of
Consumer Affairs, 39(2):299–313.
[Posner and Weyl, 2013] Posner, E. and Weyl, E. G. (2013). Benefit-cost analysis for financial
regulation. American Economic Review, 103(3):393–97.
[Prelec and Loewenstein, 1998] Prelec, D. and Loewenstein, G. (1998). The red and the black:
Mental accounting of savings and debt. Marketing Science, 17(1):4–28.
[Prelec and Simester, 2001] Prelec, D. and Simester, D. (2001). Always leave home without it: A
further investigation of the credit-card effect on willingness to pay. Marketing Letters, 12(1):5–12.
[Puri and Robinson, 2007] Puri, M. and Robinson, D. T. (2007). Optimism and economic choice.
Journal of Financial Economics, 86(1):71 – 99.
[Raghubir and Srivastava, 2008] Raghubir, P. and Srivastava, J. (2008). Monopoly money: The
effect of payment coupling and form on spending behavior. Journal of Experimental Psychology:
Applied, 14(3):213–225.
[Shefrin and Statman, 1993] Shefrin, H. and Statman, M. (1993). Behavioral aspects of the design
and marketing of financial products. Financial Management, 22(2):123–134.
[Shefrin and Thaler, 1988] Shefrin, H. M. and Thaler, R. H. (1988). The behavioural life-cycle
hypothesis. Economic Inquiry, 26(4):609–643.
[Shoven and Sialm, 2004] Shoven, J. B. and Sialm, C. (2004). Asset location in tax-deferred and
conventional savings accounts. Journal of Public Economics, 88(1-2):23–38.
[Shui and Ausubel, 2004] Shui, H. and Ausubel, L. M. (2004). Time inconsistency in the credit
card market. 14th Annual Utah Winter Finance Conference.
[Soman, 2001] Soman, D. (2001). Effects of payment mechanism on spending behavior: The role
of rehearsal and immediacy of payments. Journal of Consumer Research, 27(4):460–474.

18
[Soman, 2003] Soman, D. (2003). The effect of payment transparency on consumption: Quasi-
experiments from the field. Marketing Letters, 14(3):173–183.

[Stango and Zinman, 2016] Stango, V. and Zinman, J. (2016). Borrowing high versus borrowing
higher: Price dispersion and shopping behavior in the u.s. credit card market. The Review of
Financial Studies, 29(4):979.
[Stephens, 2006] Stephens, M. (2006). Paycheque receipt and the timing of consumption. The
Economic Journal, 116(513):680–701.

[Stephens, 2008] Stephens, M. (2008). The Consumption Response to Predictable Changes in

Discretionary Income: Evidence from the Repayment of Vehicle Loans. The Review of Economics
and Statistics, 90(2):241–252.
[Telyukova, 2013] Telyukova, I. A. (2013). Household need for liquidity and the credit card debt
puzzle. The Review of Economic Studies, 80(3):1148.
[Thaler and Benartzi, 2004] Thaler, R. H. and Benartzi, S. (2004). Save more tomorrowTM : Using
behavioral economics to increase employee saving. Journal of Political Economy, 112(S1):S164–
S187.
[Tversky and Kahneman, 1986] Tversky, A. and Kahneman, D. (1986). Rational choice and the
framing of decisions. The Journal of Business, 59(4):S251–78.
[Venti and Wise, 1987] Venti, S. F. and Wise, D. A. (1987). IRAs and Saving. In The Effects of
Taxation on Capital Accumulation, NBER Chapters, pages 7–52. National Bureau of Economic
Research, Inc.

[Venti and Wise, 1988] Venti, S. F. and Wise, D. A. (1988). The Determinants of IRA Contri-
butions and the Effect of Limit Changes. In Pensions in the U.S. Economy, NBER Chapters,
pages 9–52. National Bureau of Economic Research, Inc.
[Wang and Keys, 2014] Wang, J. and Keys, B. (2014). Perverse nudges: Minimum payments and
debt paydown in consumer credit cards. (323).

[Zhang, 2006] Zhang, X. F. (2006). Information uncertainty and stock returns. The Journal of
Finance, 61(1):105–137.
[Zinman, 2009] Zinman, J. (2009). Debit or credit? Journal of Banking & Finance, 33(2):358–366.
[Zinman, 2014] Zinman, J. (2014). Household debt: Facts, puzzles, theories, and policies. (20496).

19