Content uploaded by Víctor Martinez-Moles
Author content
All content in this area was uploaded by Víctor Martinez-Moles on Oct 22, 2019
Content may be subject to copyright.
Spanish regulation for labeling of financial products:
a behavioral-experimental analysis
Y. Go
´mez
1
•V. Martı
´nez-Mole
´s
1
•J. Vila
1
Received: 2 May 2016 / Accepted: 14 August 2016 / Published online: 31 August 2016
Springer International Publishing Switzerland 2016
Abstract This paper assesses the impact of the Spanish Ministry of Economy and
Competitiveness’ (Board of Executives (BOE) Order ECC/2316/2015. Economy
and Competitiveness Ministry, Spain, 2015) new regulation for financial product
labeling. We design and conduct an economic experiment where subjects make
risky investment decisions under three different treatments: a control group where
subjects have only objective information about the key features of the products they
must select and two treatment groups introducing visual labels resembling the labels
required under the new Spanish regulation. The results of the experiment are ana-
lyzed within the framework of rank-dependent utility theory. While visual labels do
not change the utility function of the subjects, they do significantly affect the
subjects’ weighting functions. The introduction of numerical and color-coded labels
significantly increases the concavity of the weighting functions and increases pes-
simism and risk-aversion in cases where the probability of obtaining the best out-
come is high. Labels widen the difference between real subjects’ behavior and that
of the perfectly rational agents described by expected utility theory. Consequently,
our empirical findings raise doubts as to whether the new regulation actually
achieves its objectives. The regulation seeks to empower retail investors by
enhancing their understanding of financial products. Introducing the visual labels,
however, seemingly increases the differences between actual risk levels and the
decision weights applied by subjects when making decisions. Moreover, labels
&J. Vila
jose.e.vila@uv.es
Y. Go
´mez
yolanda.gomez-gonzalez@uv.es
V. Martı
´nez-Mole
´s
victor.martinez-moles@uv.es
1
University of Valencia, Valencia, Spain
123
Econ Polit (2016) 33:355–378
DOI 10.1007/s40888-016-0037-z
increase investors’ pessimism and risk-aversion when the best outcome is likely and
fail to alter investors’ risk-aversion when the worst outcome is likely.
Keywords Behavioral finance Behavioral economics Experimental economics
Nudge Financial labeling
JEL Classification D81 (Criteria for Decision-Making under Risk and
Uncertainty) G28 (Government Policy and Regulation) D18 (Consumer
Protection)
1 Introduction
Understanding financial products is often difficult for retail investors with low
financial literacy. Empirical evidence shows that consumers express a lack of
confidence and knowledge when choosing financial instruments such as investment
or pension products (Beckett et al. 2000). These difficulties are due to a variety of
reasons, including behavioral biases (Tversky and Kahneman 1974; Northcraft and
Neale 1987), self-control (Gathergood 2012; Thaler and Shefrin 1981), mental
accounting (Thaler 1985), cognitive biases and limitations (Bertrand and Morse
2011), and poor financial literacy (Banks et al. 2010; Hoelzl and Kapteyn 2011;
Lusardi 2008; Van Rooij et al. 2011). These difficulties may negatively affect the
quality of retail investors’ financial decisions. For instance, consumers with low
financial literacy are less likely to invest in stocks (Van Rooij et al. 2011), while self-
control problems lead to suboptimal retirement savings (Thaler and Benartzi 2004).
A relevant example is the case of preference shares
1
in Spain, where over 700,000
people with poor financial literacy fell victims of financial fraud due to the sale of this
complex investment product (Lo
´pez and Go
´nzalez 2012). The incident became so
grave that it led the European Commission (EC 2015) to release a report in 2015. The
report revealed that most consumers were not duly informed by banks about the risks
associated with such investments. As a result, consumers purchased risky financial
instruments without even realizing they were doing so. The report highlighted the
need to include user-friendly information in the description of financial products to
allow citizens to make sound financial decisions. This recommendation, however, is
not the only one of its kind, with many scholars also advocating better regulation of
information about financial products (Agarwal et al. 2013; Barr et al. 2008;
Brunnermeier et al. 2009; Harrington 2009; Posner and Weyl 2013).
In 2015, the Spanish Ministry of Economy and Competitiveness (2015) launched
a new regulation for the presentation of information on financial products. This legal
document explains that despite the efforts of the Spanish Government and the
European Commission, the availability of comprehensible, comparable information
on financial products is poor. The document also reports that a major concern for
1
Shares, often with no voting rights, whose holders are paid their dividend before all other shareholders
and are repaid first at face value if the company goes into liquidation. This complex financial product was
extensively sold to customers with low financial literacy who were unaware of the product’s features and
who ultimately lost a large part of their investment.
356 Econ Polit (2016) 33:355–378
123
financial customers is that the available information is confusing or incomplete. In
many cases, the different formats used to present information are incomparable and
use overly technical language. The Spanish Government also claims that pre-
contractual and advertising documents contain too much information presented in a
way that prevents readers from finding the most relevant features of the product.
These deficiencies highlight the need to provide financial customers with
standardized documentation that clearly and visually explains the essential
information for each product. The goal of the new regulation—as stated in the
regulation document itself—is to guarantee that the most relevant information is
presented in a standard format using simple graphical or numerical representation.
The regulation seeks to ensure that financial services customers have all the
information they need to make informed judgments regarding investment services
and understand the risks associated with these services.
Introducing visual labels is common in other product areas such as nutrition or
energy efficiency, where the European Commission regulates labeling formats and
contents. Recent studies suggest that such labels do actually nudge consumers
toward making better decisions (Balcombe et al. 2010; Codagnone et al. 2013;Hieke
and Wilczynski 2012;Houde2012; Jones and Richardson 2007; Newell and
Siikama
¨ki 2013), although some authors report mixed or even contradictory results
(Aron et al. 1995; Codagnone et al. 2016;Koenigstorferetal.2014, Sacks et al.
2009). In its new regulation, the Spanish Ministry of Economy and Competitiveness
(2015) claims that introducing a labeling system similar to those applied in nutrition
or energy could affect decision-making under risk and improve customers’
understanding of financial products and investment decisions. To the best of our
knowledge, however, studies fail to provide empirical evidence of the effectiveness
of visual labels in improving financial decision-making.
To fill this research gap, this paper assesses the impact of visual labels on
customers’ investment decision-making under risk. We test the two types of visual
labels proposed by the Spanish Ministry of Economy and Competitiveness’ (2015)
regulation: (1) a numerical label that rates the product from lowest risk to highest
risk; and (2) a graphical label applying different colors from green (lowest risk) to
red (highest risk), imitating traffic lights. Although both labels show the financial
product’s risk level, they differ in how the information is presented. Using
behavioral economics and experimental methods, we analyze whether introducing
this type of label actually improves decision-making by making investors’ decisions
more similar to the decisions of perfectly rational agents. This research question is
relevant for policy because it empirically tests the rationale behind the Spanish
Ministry of Economy and Competitiveness’ (2015) regulation. Moreover, to the best
of our knowledge, no study has addressed the impact of labeling on decision-making
under risk, so the question is also empirically relevant for the literature on framing
effects under uncertainty.
The paper is organized as follows. Sections 2and 3present a literature review
and the behavioral theoretical model applied in the study. Section 4describes the
economic experiment applied to test the hypotheses. Section 5summarizes the main
results of the experiment. Finally, Sect. 6discusses the practical and theoretical
implications of these results.
Econ Polit (2016) 33:355–378 357
123
2 Literature review
Economists have long known that Homo Oeconomicus is a fictional concept.
Individuals’ preferences are not complete and transitive, actual people are not fully
rational, and it is costly for people to acquire and process information.
Consequently, people do not behave as assumed in rational choice theory. An
abundance of literature highlights and explores the differences between actual
human behavior and that predicted by rational choice theory models (e.g.,
Kahneman 2003; Kahneman and Tversky 1984; Simon 1955; Thaler 1985). These
differences led to the development of behavioral economics. Behavioral economics
not only provides new tools to analyze economic behavior, but also advocates new
ways to design effective public policies and change citizens’ and consumers’
behavior (Bogliacino et al. 2015a; Shafir 2013; Sunstein 2014; Thaler and Sunstein
2008).
Public policy has often relied on the assumption of rationality in human behavior.
This flawed assumption may have led to suboptimal policy design. For example, the
amount of information passed on to consumers (through product labeling or mass
media campaigns) has steadily increased under the assumption that they will be able
to process it to their advantage. This approach has often failed, as exemplified by
public policies to prevent smoking. Besides increasing taxes or banning smoking in
public places, placing visceral warning messages on cigarette packets seems to be
the most effective measure, as shown by Bogliacino et al. (2015a,b), who illustrate
how introducing pictorial warnings may reduce the willingness to pay for a tobacco
product by 80 %. Studies show that certain emotions behave differently from others.
Shame, anger, or distress are more effective in reducing smoking than fear and
disgust, which have been the emotions most commonly appealed to by past anti-
tobacco campaigns. If smokers were rational, a label with clear information about
the hazards of smoking might be a good idea. But smokers are not rational, so
alternatives need to be considered and tested in different product areas.
Behavioral economics considers not only the role of emotions, but also the way
cognitive biases work and can be managed to promote a desired behavior. Thaler
and Sunstein (2008) present a framework for applying experimental-behavioral
economics in the public sector. Such a framework is based on what they define as a
nudge approach for policy-making. The concept of choice architecture is central in
the nudge approach, where choice architecture refers to the way options are
presented to people. The number of choices presented, the way attributes are
described, and the presence of default options are elements of the choice
architecture. The strategic management of choice architecture could be used in
policy-making to nudge consumers toward personally and socially desirable
behaviors like saving energy or choosing healthier foods.
The implications of behavioral economics and nudging in public policy-making
are far reaching, and insights from behavioral economics have been applied to
various domains, including personal and public finance, health, energy, public
choice, and marketing (Samson 2014). In 2010, the UK government set up the
Behavioural Insights Team (BIT), a special unit dedicated to applying behavioral
358 Econ Polit (2016) 33:355–378
123
science to public policy and services. One of the most influential reports by the BIT
is MINDSPACE (Dolan et al. 2012). The report explores how behavioral economics
can help meet current policy challenges in areas such as health, finance, and climate
change. Echoing the UK government’s initiative, Van Bavel et al. (2013)
summarize the European Commission’s strategy of using experimental-behavioral
economics to enhance its policy-making. The authors describe how the European
Commission began to formally apply behavioral insights in 2009, when the
Consumer Rights Directive recognized the power of default options. The European
Commission proposed limiting the use of pre-checked boxes in consumer contracts
(the kind that made consumers purchase travel insurance even if they did not want
it) to save consumers money by default. In 2010, the Directorate-General for Health
and Consumers (DG SANCO) conducted a pilot study entitled Consumer Decision-
Making in Retail Investment Services. Through a series of laboratory experiments,
the study showed how consumers behaved when faced with different investment
products. People struggled to make optimal investment choices, even in the simplest
environments. Subjects were also prone to biases and framing effects (i.e., the way
choices were presented). One of the conclusions of the study was that simplifying
and standardizing product information would significantly improve investment
decisions. Encouraged by the success of this study and responding to interest
throughout the European Commission, DG SANCO set up the Framework Contract
for the Provision of Behavioral Studies in 2012. At the World Bank, the 2015 World
Development Report (World Bank 2015) discusses how a more realistic
understanding of choice and behavior using experimental-behavioral methodology
can make development interventions more effective. Accordingly, the experimental-
behavioral approach has proven an effective way of studying people’s behavior and
testing which stimuli cause a desired change in behavior (Hernandez et al. 2015).
The experimental-behavioral approach can help to (1) identify the best choice, (2)
yield experimental evidence to predict the choices subjects would make under
different policy treatments, (3) and quantify the gap between the policy objective
and the empirical reality (Codagnone et al. 2014).
Finances and financial regulation have also been addressed by behavioral
economics. The application of behavioral economics to finances has given rise to a
new field: behavioral finance. Like behavioral economics, behavioral finance
explains and increases our understanding of investors’ reasoning patterns, including
emotional processes and the degree to which such processes affect decision-making
(Ricciardi and Simon 2000). Behavioral finance has its roots in different fields, such
as prospect theory, cognitive errors, problems of self-control, and the pain of regret.
According to behavioral finance, investors are often affected by frames and are
swayed by temptation and regret (Statman 1995). Several authors have shown these
systematic biases in the finance domain and have incorporated them in their models.
Hirshleifer and Teoh (2003) developed a behavioral model that addresses investors’
limited attention levels. They empirically showed that the way financial information
is presented (i.e., the frame) can affect the investor’s perception. Brown et al. (2015)
found that expected social security claiming ages are sensitive to the way the
information about actuarial adjustments is framed. Hastings et al. (2010) showed
how different ways of presenting pension management fees shape consumer
Econ Polit (2016) 33:355–378 359
123
choices. Zhou and Pham (2004) report that investors’ goals may be affected by the
investment opportunities under evaluation rather than being independent of these
alternatives, as assumed in standard finance theory. Statman (1995) argues that
behavioral finance is built on a broader model of human behavior than standard
finance and allows economists to deal effectively with many puzzles that have
traditionally plagued standard finance (e.g., investors’ preference for cash
dividends, investors’ reluctance to realize losses, the determination of expected
returns, the design of securities, and the nature of financial regulations). According
to Statman (1995), behavior and psychology affect individual investors’ and
portfolio managers’ financial decision-making processes in terms of risk assessment
and framing (i.e., the way investors process information and make decisions
depending on its layout). These related components provide a deeper understanding
of the empirical evidence on finance, including investors’ preferences, the design of
modern financial products, and financial regulations. Shefrin and Statman (1993)
suggest that behavioral considerations are indispensable when designing financial
products. A relevant example in behavioral finance and nudging is Thaler and
Benartzi’s (2004) SMarT Plan. The authors propose a prescriptive savings program,
using elements of behavioral economics to nudge employees toward increasing their
retirement savings. In this vein, Blake and Boardman (2014) created SPEED-
OMETER (or Spending Optimally Throughout Retirement), which shows how
behavioral economics can be used to improve pensioners’ spending decisions.
Understanding financial products is difficult for many consumers, and numerous
authors have called for regulations governing the information provided on financial
products (Agarwal et al. 2013; Barr et al. 2008; Brunnermeier et al. 2009;
Harrington 2009; Posner and Weyl 2013). Accordingly, the Spanish Ministry of
Economy and Competitiveness (2015) has released a new regulation in Spain that
requires financial companies to introduce a standardized information label for all
financial products. This label’s format is similar to the format used for energy or
nutrition products. The regulation claims that introducing these new labels will
reduce the number of purchases of undesired, highly risky financial assets and will
improve consumers’ understanding of financial products.
Some authors have recently assessed the effect of visual labels in other areas such
as nutrition and energy efficiency. In nutrition, Jones and Richardson (2007) used
eye-tracking technology to observe that color-coded labels reduce the number of
purchases of high fat products by highlighting key nutrients. Balcombe et al. (2010)
designed a choice experiment to determine that many subjects avoid choosing
products bearing the red code in labeling systems that simulate traffic lights. Hieke
and Wilczynski (2012) confirmed the signaling effect of color, reporting that color
helps reduce the complexity of decision-making. Koenigstorfer et al. (2014) report
that color-coding on nutrition labels affects food purchase behavior. This effect,
however, is contingent on consumers’ self-control. Consumers with low self-control
make healthier food decisions in response to color-coded labels, whereas the effect
is weak among consumers with high self-control. Aron et al. (1995) examined the
influence of nutrition information, concluding that nutrition labels have no positive
effect on food choices and, for certain subgroups, even have a negative effect. Their
360 Econ Polit (2016) 33:355–378
123
results reveal the importance of assessing consumers’ motivational characteristics
when developing nutrition education programs.
To study energy-efficiency product labeling, Newell and Siikama
¨ki (2013)
experimentally tested the extent to which different types of information and
intertemporal behavior (i.e., discounting) affect consumers’ behavior regarding
energy efficiency. They conclude that labels and information content may help
consumers make more energy-efficient decisions. Houde (2012) provides evidence
that the Energy Star program substantially affects purchase decisions. Specifically,
the effect of the Energy Star label alone could increase sales of a particular
refrigerator model by as much of 35 %. Codagnone et al. (2013) experimentally
tested the effectiveness of car eco-labels and promotional material. Running a series
of laboratory and online experiments, the authors tested the effectiveness of
different information and layouts such as different CO
2
classification systems. The
most effective label was the vertical colored layout with the absolute CO
2
emissions
classification system, which is extremely similar to the label currently used in
European energy regulation and the proposed label for financial regulation in Spain.
Interestingly, however, labels make consumers believe that goods consume more
energy than they actually do (Sahoo and Sawe 2015). Codagnone et al. (2016) found
that consumers change their behavior in response to labels focusing on running costs
or fuel economy when environmental friendliness messages are combined with
messages regarding fuel economy. When this cost-saving frame (i.e., fuel economy)
is omitted, no such change in behavior occurs. The authors also showed that large,
expensive cars tend to be undervalued once fuel economy is highlighted.
Despite empirical evidence suggesting that the layout of information on financial
products affects consumers’ perceptions and decisions (Brown et al. 2015; Hastings
et al. 2010; Hirshleifer and Teoh 2003; Statman 1995), the way the new financial
labeling system will affect consumers is unclear. While the Spanish Ministry of
Economy and Competitiveness (2015) claims that this new financial labeling system
will improve consumers’ understanding of financial products, the effects of similar
labeling systems on consumers have been mixed. Some studies show that product
labeling nudges consumers toward making better decisions (Balcombe et al. 2010;
Codagnone et al. 2013; Hieke and Wilczynski 2012; Houde 2012; Jones and
Richardson 2007; Newell and Siikama
¨ki 2013), whereas others report negative
effects or show that positive effects are contingent on consumer factors (Aron et al.
1995; Codagnone et al. 2016; Koenigstorfer et al. 2014; Sacks et al. 2009; Sahoo
and Sawe 2015).
Accordingly, the objectives of this study are both theoretical and policy oriented.
From a theoretical viewpoint, we analyze how numerical and visual labels affect
decision-making under risk. Specifically, this paper provides and interprets
experimental empirical evidence to answer an open question: Do simple labels
narrow or widen the gap between investors’ actual behavior and the behavior
predicted by rational choice theory? Likewise, if these labels do have an effect on
decision-making, what is the nature of this effect and what are the implications in
terms of the subjects’ risk attitudes? From a policy-oriented viewpoint, we analyze
the new Spanish financial labeling system and test whether the policy meets its
objectives and achieves the desired consumer behavior.
Econ Polit (2016) 33:355–378 361
123
3 The model
In rational choice theory, the shape of the utility function determines risk attitude. In
real decision-making, however, risk attitude also depends on other more complex
issues such as how probabilities are processed and applied during the decision-
making process. Behavioral economics offers a more realistic approach to modeling
the role of probabilities in decision-making. Behavioral economics also lets us
model experimental phenomena that expected utility is unable to explain (Tversky
and Kahneman 1992; Abdellaoui et al. 2011; Alventosa et al. 2015). This section
presents a formalization of the way probabilities are transformed from the viewpoint
of rank-dependent utility theory (Quiggin 1982). Rank-dependent utility theory
extends Expected Utility theory and is a specific case of cumulative prospect theory
(Tversky and Kahneman 1992; Wakker 2010) when restricted to the domain of the
gains. Rank-dependent utility models are widely used to quantify the difference
between the risk levels known by decision-makers and the weights they actually
apply to each uncertain event in the decision-making process (Wakker 2010;
Abdellaoui et al. 2011). Since this study examines how this difference and its effect
on risk attitude change with the introduction of the new labels, Rank Dependent
Utility offers an appropriate theoretical framework to assess the impact of the new
labeling regulation.
Investment decisions always imply some degree of uncertainty regarding the
final return on the investment. For the sake of simplicity, let us consider that the
performance of an investment can be described as a series of potential returns or
outcomes x1[x2[ [xn, where outcome xitakes place with probability pi.
The key concept when analyzing decision-making under risk and uncertainty is
that of the risk attitude of the decision-maker. Under rational choice theory
(expected utility paradigm), risk attitude is characterized solely by the decision-
maker’s utility function, which determines the psychological ‘‘value’’ of each of
the potential returns of the investment. The utility function represents the
decision-maker’s preferences, which are established according to the absolute
outcome rather than the gains or losses from a previous situation or reference
point (i.e., asset integration).
2
Assuming that each decision-maker knows these
probabilities, she or he will invest in a fund if only if the utility of the money
invested is lower than the expected utility of the outcomes. Formally, if Idenotes
the amount to be invested and UðxÞthe utility of outcome x, the investor will
invest in the fund if and only if UIðÞ\Pn
i¼1piUðxiÞ. A decision-maker is
considered risk-averse (or, conversely, risk-seeking) if her or his utility function is
concave (or, conversely, convex).
In expected utility models, although the psychological value of an outcome
does not coincide with the actual value of the outcome (in general, UðxÞ 6¼ x), the
psychological value of each probability pialways equals the probability pi, which
is the value used when making an investment decision. No additional
2
Asset integration refers to the idea that individuals decide about risky prospects by considering the
effect of decisions on their final wealth rather than on specific gains and losses (Kahneman and Tversky
1979). Extensive empirical evidence shows that this assumption does not hold (Andersen et al. 2011).
362 Econ Polit (2016) 33:355–378
123
transformation
3
is made. In the rank-dependent utility model, it is assumed that
the psychological value of a probability used when deciding whether to invest in a
fund—or decision weight as it is usually called—is a function of the probabilities
of all potential outcomes of the investment. In this conceptual framework, a
rank—or more intuitively a good-news probability—for any potential outcome x
of the investment is defined as the probability that the fund yields an outcome
strictly larger than x:Formally, rank xðÞ¼
Pxi[xprobðxiÞ. Ranks are numbers
between 0 and 1, where 0 (or, conversely, 1) is the rank associated with the best
(or, conversely, the worst) possible outcome of the fund. Let us define
xnþ1¼1. Then, the probability of outcome xican be written as pi¼
rank xiþ1
ðÞrank xi
ðÞfor i¼1;...;n. Before decision-making, ranks are trans-
formed according to a non-decreasing function w:½0;1!½0;1named the
weighting function. Given a weighting function w, the decision weight of outcome
xiis defined as pi¼wðrank xiþ1
ðÞÞwðrank xi
ðÞÞ. If the weighting function is the
identity function (i.e., wpðÞ¼p), then the decision weights will equal the
probabilities of the outcomes (pi¼piÞ:Decision weights are positive numbers
less than 1, but they are not required to add up to 1. Decision weights are related
to the slope of the weighting function: the steeper the weighting function, the
larger the difference between wðrank xiþ1
ðÞÞand wðrank xi
ðÞÞ, and hence the larger
the corresponding decision weight pi. Under rank-dependent utility theory, an
investor with utility function UðxÞand weighting function wðpÞwill buy a fund if
and only if UIðÞ\Pn
i¼1piUðxiÞ. The interpretation of the weighting function and
the computation of the decision weights is greatly simplified in cases with only
two uncertain events x1[x2. In such cases, p1¼w rank x2
ðÞðÞw rank x1
ðÞðÞ¼
wp
1
ðÞw0ðÞ¼wp
1
ðÞand p2¼w rank x3
ðÞðÞw rank x2
ðÞðÞ¼w1ðÞwp
1
ðÞ¼
1wp
1
ðÞ¼1p1.
Example (adapted from Wakker 2010). Let us consider a fund with four potential
outcomes x1[x2[x3[x4with identical probabilities p1¼p2¼p3¼p4¼0:25.
Let us assume that the weighting function for an investor is given by wpðÞ¼p0:5.
Then, the decision weights for each outcome are given by
p1¼w rank x2
ðÞðÞw rank x1
ðÞðÞ¼0:250:500:5¼0:50, p2¼0:21, p3¼0:16
and p4¼0:13. Note that although the probabilities of all four outcomes are
identical, the investor decides whether to invest in the fund based on the decision
weight (subjective probability) for the best outcome x1(i.e.,
p1¼0:50 [0:25 ¼p1) and the decision weight for the worst outcome x4(i.e.,
p4¼0:13\0:25 ¼p4). Even if the potential investor knows the actual probability
of obtaining high returns, she or he behaves as if the probability of obtaining high
3
The axiom of the independence of irrelevant alternatives by Von Neumann and Morganstern formalizes
this basic assumption in expected utility theory. If a decision-maker is indifferent to the choice between
two possible outcomes, then she or he will be indifferent to the choice between two lotteries with equal
probabilities, if the lotteries are identical in every other way (i.e., the outcomes can be substituted). So, if
the decision-maker is indifferent to the choice between outcomes x and y, then the decision-maker is
indifferent to the choice between a lottery yielding x with probability p and z with probability (1 -p) and
a lottery yielding y with probability p, and z with probability (1 -p). Likewise, if x is preferable to y,
then a lottery yielding x with probability p and z with probability (1 -p) is preferable to a lottery
yielding y with probability p, and z with probability (1 -p).
Econ Polit (2016) 33:355–378 363
123
returns were higher than it actually is and as if the probability of obtaining low
returns were lower than it actually is.
Recall that ranks are good-news probabilities. An outcome with a small rank
means that the probability of getting a higher outcome is small. In other words, the
lower the rank, the better the outcome. Figure 1, adapted from Wakker (2010),
illustrates how two kinds of deviations from additive probabilities combine to create
the common probability weighting functions. Figure 1a depicts traditional expected
utility with probabilities weighted linearly (i.e., wpðÞ¼p). However, most
empirical research conducted since Preston and Baratta’s (1948) study reports
inverse S-shapes, as in Fig. 1b. Since decision weights are given by
pi¼wðrank xiþ1
ðÞÞwðrank xi
ðÞÞ, the slope of the weighting function can be
interpreted as an approximation of the sensitivity of the decision weights. The usual
inverse S-shape in Fig. 1b thus shows how decision weights are larger for ranks at
the two extremes (i.e., ranks close to 0 and 1) than for intermediate ranks.
4 Method and experimental design
An economic experiment was designed and implemented to test the impact of
applying the two label formats. In the experiment, each subject was asked to reveal
the minimum certain return they would require to sell 12 risky investment funds.
Funds were represented by binary lotteries, where the investor randomly obtained
one of two different return levels with given probabilities. A total of 150 subjects
were recruited from a database administered by the University of Valencia. The
subjects volunteered to participate in the experiment. The profile of the participants,
in terms of age and sex, is presented in ‘‘Appendix 2’’. The experiment was
implemented at the University of Valencia in December 2015. Subjects were
randomly assigned to a computer in the laboratory, informed that they could leave
the experiment at any moment, and asked to sign a consent form. The experiment
software was a web application developed using the Yii PHP framework. Following
a between-subjects design, 50 participants were assigned to each treatment. The
actual payoff for each subject was the sum of a constant show-up fee and the payoff
obtained by the subject in one of the 12 decisions. The decision used to calculate the
Fig. 1 Common weighting functions (adapted from Wakker 2010)
364 Econ Polit (2016) 33:355–378
123
payoff was selected uniformly at random. The underlying decision problem was
adapted from Abdellaoui et al. (2011). Table 1shows the 12 investment funds used
in the experiment, along with their outcomes and probabilities. Subjects were told
that they had 12 variable income investment funds they could sell. They were
instructed to state the minimum quantity of money they would accept to sell the
variable income investment fund. For each investment fund, subjects were asked,
‘‘What is the minimum return you would require to sell your variable income
investment fund?’’ Screenshots of the experiment appear in ‘‘Appendix 1’’ .
To guarantee that subjects had an incentive to reveal their actual certainty
equivalent for each lottery, we followed the Becker–DeGroot–Marschak (BDM)
mechanism (Wakker and Deneffe 1996). This mechanism is widely used in
experimental economics because the dominant strategy for a subject is to reveal her
or his minimum return honestly. Under this mechanism, once a subject reveals the
minimum return required to sell each fund, a market return is determined at random:
•If the subject’s minimum return is equal to or lower than the market return, the
subject sells the investment fund. In this case, the subject’s payoff is the market
return.
•If the subject’s minimum return is greater than the market return, the subject
keeps the fund. The fund then attains its high or low outcome at random
according to the probabilities shown to the subject. The outcome determines the
subject’s payoff.
The uncertainty level of each fund reflects the variance of the lottery. Investment
funds are then classified into three sequential levels (denoted 1, 2, and 3) in terms of
their uncertainty level, as shown in Table 1.
Following a between-subject design, the experiment was run in three separate
sessions—one per treatment. Subjects in the first session made their 12 decisions
after receiving information about the outcomes and probabilities of each binary
Table 1 Funds presented to the subjects
Investment
fund
High
outcome
Low
outcome
Probability of
high outcome
Variance Uncertainty
level
1 100 0 25 1875 2
2 200 50 25 4218.75 3
3 200 100 25 1875 2
4 50 0 25 468.75 1
5 150 100 25 468.75 1
6 200 150 25 468.75 1
7 200 0 25 7500 3
8 200 0 75 7500 3
9 200 0 5 1900 2
10 200 0 95 1900 2
11 200 0 25 7500 3
12 200 0 50 10,000 3
Econ Polit (2016) 33:355–378 365
123
lottery (control group). Subjects in the other two sessions received the same
information as well as an adaptation of the label required under the new Spanish
regulation, as presented in Fig. 2. Treatment 2 saw the label in graphical format,
while Treatment 3 saw the label in numerical format. The labels of each level and
treatment appear in ‘‘Appendix 1’’ .
5 Results
The method used to estimate the utility and weighting functions of each subject was
adapted from Abdellaoui et al. (2011). Subjects were asked to state the minimum
return they would require to sell the 12 funds in Table 1. Thus, each of the 150
subjects made 12 decisions, thereby providing the minimum returns for 1800
different investment funds. The descriptive statistics for this outcome variable
appear in ‘‘Appendix 2’’. Funds 1–7 were used for parametric estimation of the
utility function, and funds 8–12 were used for non-parametric estimation of the
decision weights for each subject.
Following Abdellaoui et al. (2008), we applied a power function as a parametric
specification for utility and assumed that UxðÞ¼xa. This function is concave if
a\1, linear if a¼1, and convex if a[1. Under this assumption, for investment
funds i¼1;2;...;7, Ii¼½w0:25ðÞxa
i;1þ1w0:25ðÞðÞxa
i;21
=a, where Iidenotes the
certainty equivalent of lottery iand xi;1and xi;2are the high and low outcomes of the
lottery. For each subject, once the certainty equivalents had been revealed, w0:25ðÞ
and the parameter ain the non-linear expression for Iiwere estimated using
nonlinear least squares. The estimation was carried out using the NLS function in R.
Table 2presents the results of the estimation of a. The parameter awas not
significantly different for subjects in each treatment (pvalue =0.481), implying
that labeling had no effect on the utility function of the subjects.
The weighting function for each subject was then estimated using the minimum
return the subject required to sell investment funds i¼8;...;12 and the value of a,
which had already been calculated. Accordingly, wp
i
ðÞ¼ Ii
200
a. Table 3shows the
results of the estimation of the weighting function for each treatment. Table 3also
shows that the weighting functions were always significantly different among
treatments, except in the case with the smallest probability. Thus, both types of
labels affected subjects’ decisions regarding financial products.
Figure 3presents the weighting functions (median values) for each treatment
(i.e., for each labeling format). Since the diagonal in this figure represents the
weighting function under expected utility theory, the area between the weighting
function and the diagonal provides a measure of how much the subjects differ from
rational agents in their ability to process actual probabilities when making decisions
under risk. When a weighting function coincides with the diagonal it represents
linear processing of probabilities, as in rational choice theory (i.e., the subjects
display a proper understanding of the probabilities associated with the different
366 Econ Polit (2016) 33:355–378
123
Econ Polit (2016) 33:355–378 367
123
outcomes). Conversely, when the weighting function does not coincide with the
diagonal, the area between the diagonal and the weighting function offers a measure
of how far the subjects’ understanding of risk strays from the proper understanding
assumed under expected utility theory. The introduction of labels significantly
increased the area between the weighting function and the diagonal, representing a
worsening in subjects’ understanding of the lotteries’ risk. The area between the
weighting function and the diagonal was greatest for the numerical label in
Treatment 3. Hence, the results of the experiment fail to confirm that the regulation
accomplishes its objective of enhancing consumers’ understanding of financial
product information.
The existence of significant differences between the weighting functions for
the groups with and without labels raises two questions: How does the new
Spanish regulation change consumers’ understanding of uncertainty? Moreover,
bFig. 2 Screenshots of the decision screen of the experiment for each treatment. aTreatment 1: control.
bTreatment 2: color. cTreatment 3: number
Table 2 Estimation of aby
treatment
Significance codes: * p \0.1;
** p \0.05; *** p \0.01
Control Number Color
Median 0.856 1.030 0.853
Average 1.083 1.204 0.987
Standard deviation 1.155 0.589 0.591
ANOVA (p-value) 0.481
Table 3 Estimation of wðpiÞby treatment
Labeling wð0:05Þwð0:25Þwð0:50Þwð0:75Þwð0:95Þ
Median Control 0.291 0.334 0.548 0.711 0.890
Average Control 0.357 0.369 0.549 0.699 0.776
Standard deviation Control 0.317 0.162 0.236 0.194 0.294
Median Number 0.240 0.277 0.400 0.322 0.580
Average Number 0.328 0.281 0.350 0.376 0.551
Standard deviation Number 0.256 0.164 0.197 0.273 0.356
Median Color 0.383 0.299 0.468 0.402 0.740
Average Color 0.394 0.331 0.463 0.434 0.662
Standard deviation Color 0.273 0.191 0.248 0.284 0.305
ANOVA (p-value) Control-color-number 0.618 0.021** 0.000*** 0.000*** 0.001***
T-test (p-value) Control-color 0.566 0.317 0.103 0.000*** 0.079*
T-test (p-value) Control-number 0.641 0.016** 0.000*** 0.000*** 0.002**
T-test (p-value) Color-number 0.239 0.192 0.018** 0.330 0.116
Significance codes: * p \0.1; ** p \0.05; *** p \0 .01
368 Econ Polit (2016) 33:355–378
123
how does the new Spanish regulation change the behavior of retail investors?
While the utility function was seemingly unaffected by the introduction of the
labels, the regulation affected consumers’ risk attitudes, as observed in the
change in the shape of the consumers’ weighting function (Fig. 3). For
probabilities of less than 0.25, the labels did not generate significant changes
in the weights (Table 3). Once the weighting functions for the color and number
treatments crossed the diagonal, however, these two functions became much
more concave than the weighting function for the control treatment. Moreover,
the difference between the weights in the control and numerical label treatments
was significant for all probabilities (Table 3). In the cumulative prospect model
(Wakker 2010), the concavity of the weighting function can be interpreted as
pessimism on the part of the decision-maker because the decision weights
associated with good-news probabilities are lower than the actual probabilities of
Fig. 3 Median values of the weighting function by treatment
Econ Polit (2016) 33:355–378 369
123
obtaining the best outcomes. This interpretation is easy to understand with
binary lotteries, as in our experiment. As previously explained, with only two
outcomes, the decision weight equals the value of the weighting function when
applied to the probability (p1¼wp
1
ðÞ). In our experiment, the weighting
functions for the treatments with labels became concave after crossing the
diagonal. Hence, the decision weights corresponding to the best outcome were
lower than the actual probabilities of obtaining the best outcome. This difference
was significant for all values in Treatment 3 and one value in Treatment 2
(Table 3). This behavioral-cognitive effect occurred even though the subjects
knew the real value of the probability. This phenomenon can be interpreted as
pessimism because the subjects behaved as if the chances of good news occurring
were lower than they actually were. Pessimistic decision-makers behave with
greater risk aversion than either optimistic (convex weighting function) or neutral
(linear weighting function) subjects with the exact same value function.
In summary, subjects in Treatments 2 and 3 had the same utility function as those
in the control group, but they were more pessimistic when the actual probability of
obtaining the best output was greater than 0.25.
6 Conclusions
This paper empirically assesses the impact of the Spanish Ministry of Economy and
Competitiveness’ (2015) new regulation regarding the labeling of financial
products. We designed and conducted an economic experiment where subjects
made risky investment decisions under three treatments. The first treatment was a
control group where subjects were informed only of the value of the outcomes and
the probabilities of each lottery. As well as being told the value of the outcomes and
the probabilities of each lottery, subjects in the other two treatments received
additional information in the form of labels—numerical for one group and color-
coded for the other. We analyzed the results of the experiment within the rank-
dependent utility theory framework. Under rank-dependent utility theory, decision
weights (i.e., the subject’s assessment of probabilities in the decision-making
process) may differ from the actual probabilities of events occurring, even when the
decision-maker knows the real value of these probabilities. In the rank-dependent
utility model, probabilities are transformed into decision weights through a
weighting function, as described in Sect. 3.
The main finding from this experiment is that while labels do not affect utility
functions, visual labels do significantly affect weighting functions. In all groups, the
weighting functions exhibit the typical inverse S-shape, but introducing the visual
labels significantly increases the concavity of the weighting function in the control
group. Accordingly, under rank-dependent utility theory, introducing the labels
increases subjects’ pessimism, and consequently, their risk-aversion in cases where
the probability of the best outcome is high. The increase in concavity also shows
that introducing labels augments the difference between the diagonal and the
370 Econ Polit (2016) 33:355–378
123
weighting function. In other words, the labels increase the differences between
subjects’ actual behavior and the behavior predicted by rational choice theory.
Although these finding apply to both numerical and color-coded labels, the effect is
larger with numerical labels. Differences between decision weights in the control
group and the numerical label treatment group are significant for all probabilities
greater than 0.25. In contrast, differences between decision weights in the control
group and the color-coded label treatment group are significant only for a
probability of 0.75. In summary, visual labels affect subjects’ understanding of risk
levels. Visual labels cause subjects’ understanding to diverge from that of perfectly
rational agents. Furthermore, labels make subjects more risk averse in cases where
the probability of the best output is high.
This finding has both theoretical and policy-making implications. From a
theoretical viewpoint, this paper helps to fill the gap in the literature on how labeling
affects individuals’ decision-making capabilities under risk. This paper also presents
an innovative approach to analyzing the impact of label regulation. Scholars have
analyzed the impact of labeling in consumer behavior without analyzing the
cognitive and decision models behind such behavior. In contrast, we analyzed and
estimated the utility and weighting functions after observing subjects’ decisions.
This approach allowed us not only to measure the impact of labeling, but also to
understand the reasons behind such an impact, yielding powerful insights for
effective policy design.
Besides its theoretical implications, this paper contributes by assessing the
labeling regulation recently introduced by the Spanish Ministry of Economy and
Competitiveness (2015). This regulation is designed to empower retail investors by
ensuring they have a better understanding of financial products and making it easier
to compare financial products offered by competing financial entities. In an
environment where investors with low financial literacy must make decisions
involving financially complex products, the regulation advocates the use of simple
visual and numerical labels as a way of improving the quality of investors’ decision-
making. As discussed in Sect. 2, behavioral-experimental analysis has provided
useful insights to help design better regulations in energy consumption and
nutrition. Empirical evidence does not always confirm the effectiveness of the
proposed labeling systems but nonetheless provides valuable information for
adapting and optimizing such labels. For this reason, together with the fact that
these labels codify complex information such as uncertainty levels, the effectiveness
of the Spanish Ministry of Economy and Competitiveness’ regulation must be
empirically tested.
The behavioral experiment presented in this paper shows that the labels
proposed under the new regulation are seemingly a long way from achieving
their goal. Taking decisions made by the rational agents described in rational
choice theory as a benchmark, our experiment shows that both graphical and
numerical labels actually worsen subjects’ decision-making. Introducing labels
makes retail investors’ decisions less rational. Estimation of the utility and
weighting functions shows that this difference is not a consequence of a
Econ Polit (2016) 33:355–378 371
123
variation of the utility function but rather a result of the change in the shape of
investors’ weighting functions. Notably, changes in the weighting function
occur mainly in the part of the function corresponding to high probabilities of
obtaining the best outcome. Labels reduce the decision weights for financial
products where the probabilities of obtaining the best outcome are high.
Accordingly, the regulation increases investors’ pessimism and risk-aversion
in situations where the best outcome is likely, yet it has no effect in situations
where the worst outcome is likely. This conclusion is consistent with
discussions between one of the authors of this paper and practitioners working
in the Spanish financial sector. The practitioners claimed that introducing labels
has increased the perception of risk associated with the safest products (for
instance, bank deposits), mainly among investors with low financial literacy.
This greater risk-aversion may be due to an increase in the salience of risk
caused by the presence of the labels and the assignation of a risk level (level 1)
to products that consumers previously thought of as risk free. Consistent with
findings reported by Campbell et al. (2011), differences between the goals and
outcomes of the regulation highlight the need to establish beforehand whether
proposed regulations will actually deliver on what they are supposed to.
Contrary to what is best for the Spanish retail investment market, the labels
studied in this paper may actually be nudging investors toward riskier
investment decisions than those made by perfectly rational agents, as described
by rational choice theory.
Finally, we must acknowledge that our study has some limitations. The
sample used for the experiment consisted of only 150 participants, and it might
not have been representative of the population of consumers of financial
products. The nature of the sample may therefore limit the external and
ecological validity of the experiment, although the experimental design and
implementation guarantee its internal validity and utility for studying the role of
labeling. In the future, scholars should replicate this experiment using larger
samples, allowing for the inclusion of control variables such as financial literacy,
age, and education.
Acknowledgments This work is supported by the Spanish Ministerio de Economı
´a y Competitividad
under project CO2013-46550-R.
Appendix 1: Description of the experiment
See Figs. 4and 5.
372 Econ Polit (2016) 33:355–378
123
Econ Polit (2016) 33:355–378 373
123
Appendix 2: Description of subjects and the minimum return to sell
each fund (outcome variable)
See Tables 4and 5.
bFig. 4 Screenshots of control treatment. aInstructions, bdecision, cresults
Fig. 5 Labels applied in treatments 2 (a) and 3 (b)
Table 4 Descriptive statistics
for the participants Treatment Gender (%) Age (years)
male female mean SD median SE
Control 56.10 43.90 20.63 1.92 20 0.30
Color 56.52 43.48 20.72 2.23 20 0.33
Number 54.55 45.45 20.59 1.59 20 0.24
374 Econ Polit (2016) 33:355–378
123
References
Abdellaoui, M., Bleichrodt, H., & l’Haridon, O. (2008). A tractable method to measure utility and loss
aversion under prospect theory. Journal of Risk and Uncertainty, 36(3), 245–266.
Abdellaoui, M., l’Haridon, O., & Paraschiv, C. (2011). Experienced vs. described uncertainty: Do we
need two prospect theory specifications? Management Science, 57(10), 1879–1895.
Table 5 Descriptive statistics
for the outcome variable Treatment Id Median Mean SD SE CI
Control 1 30.00 35.46 14.76 2.31 4.66
Control 2 100.00 101.46 26.20 4.09 8.27
Control 3 140.00 145.00 20.91 3.27 6.60
Control 4 20.00 21.98 9.56 1.49 3.02
Control 5 115.00 118.68 10.72 1.67 3.38
Control 6 165.00 169.27 10.87 1.70 3.43
Control 7 140.00 131.07 43.92 6.86 13.86
Control 8 40.00 63.10 63.55 9.92 20.06
Control 9 50.00 61.56 32.97 5.15 10.41
Control 10 100.00 99.32 40.36 6.30 12.74
Control 11 140.00 131.07 43.92 6.86 13.86
Control 12 175.00 150.34 63.95 9.99 20.18
Color 1 40.00 39.50 18.96 2.80 5.63
Color 2 80.00 91.09 28.09 4.14 8.34
Color 3 140.00 140.33 24.00 3.54 7.13
Color 4 21.00 25.70 12.23 1.80 3.63
Color 5 120.00 121.74 12.91 1.90 3.83
Color 6 170.00 169.89 13.06 1.93 3.88
Color 7 72.50 76.50 46.84 6.91 13.91
Color 8 50.00 67.39 51.42 7.58 15.27
Color 9 50.00 55.11 33.41 4.93 9.92
Color 10 90.00 86.37 35.07 5.17 10.41
Color 11 72.50 76.50 46.84 6.91 13.91
Color 12 145.00 128.07 57.50 8.48 17.08
Number 1 40.00 43.98 16.88 2.55 5.13
Number 2 90.00 85.86 22.87 3.45 6.95
Number 3 140.00 139.30 18.95 2.86 5.76
Number 4 25.00 26.86 10.93 1.65 3.32
Number 5 120.00 118.64 11.17 1.68 3.40
Number 6 170.00 169.34 12.50 1.88 3.80
Number 7 75.00 76.52 48.08 7.25 14.62
Number 8 51.50 70.75 52.08 7.85 15.83
Number 9 55.00 59.20 28.39 4.28 8.63
Number 10 72.50 73.30 36.88 5.56 11.21
Number 11 75.00 76.52 48.08 7.25 14.62
Number 12 125.00 113.34 67.50 10.18 20.52
Econ Polit (2016) 33:355–378 375
123
Agarwal, S., Chomsisengphet, S., Mahoney, N., & Stroebel, J. (2013). Regulating consumer financial
products: Evidence from credit cards. NBER Working Paper No. 19484.
Alventosa, A., Go
´mez, Y., Martı
´nez-Mole
´s, V., & Vila, J. (2015). Location and innovation optimism: a
behavioral-experimental approach. Journal of the Knowledge Economy. doi: 10.1007/s13132-015-
0291-2.
Andersen, S., Cox, J., Harrison, G., Lau, M., Rutstroem, E., Sadiraj, V. (2011). Asset integration and
attitudes to risk: theory and evidence. In Working Paper 2011-04, Center for the Economic Analysis
of Risk, Robinson College of Business, Georgia State University.
Aron, J. I., Evans, R. E., & Mela, D. J. (1995). Paradoxical effect of a nutrition labeling scheme in a
student cafeteria. Nutrition Research, 15(9), 1251–1261.
Balcombe, K., Fraser, I., & Di Falco, S. (2010). Traffic lights and food choice: A choice experiment
examining the relationship between nutritional food labels and price. Food Policy, 35(3), 211–220.
Banks, J., O’Dea, C., & Oldfield, Z. (2010). Cognitive function, numeracy and retirement saving
trajectories*. The Economic Journal, 120(548), F381–F410.
Barr, M. S., Mullainathan, S., & Shafir, E. (2008). Behaviorally informed financial services regulation.
New America Foundation White Paper.
Beckett, A., Hewer, P., & Howcroft, B. (2000). An exposition of consumer behaviour in the financial
services industry. International Journal of Bank Marketing, 18(1), 15–26.
Bertrand, M., & Morse, A. (2011). Information disclosure, cognitive biases, and payday borrowing. The
Journal of Finance, 66(6), 1865–1893.
Blake, D., & Boardman, T. (2014). Spend more today safely: Using behavioral economics to improve
retirement expenditure decisions with SPEEDOMETER plans. Risk Management and Insurance
Review, 17(1), 83–112.
Bogliacino, F., Codagnone, C., & Veltri, G. A. (2015a). The behavioural turn in consumer policy:
Perspectives and clarifications. Intereconomics, 50(2), 108–114.
Bogliacino, F., Codagnone, C., Veltri, G. A., Chakravarti, A., Ortoleva, P., Gaskell, G., et al. (2015b).
Pathos and ethos: Emotions and willingness to pay for tobacco products. PLoS One, 10(10),
e0139542. doi:10.1371/journal.pone.0139542.
Brown, J. R., Kapteyn, A., & Mitchell, O. S. (2015). Framing and claiming: how information-framing
affects expected social security claiming behavior. Journal of Risk and Insurance, 83(1), 139–162.
Brunnermeier, M. K., Crockett, A., Goodhart, C. A., Persaud, A., & Shin, H. S. (2009). The fundamental
principles of financial regulation (Vol. 11). London: Centre for Economic Policy Research.
Campbell, J. Y., Jackson, H. E., Madrian, B. C., & Tufano, P. (2011). Consumer financial protection. The
Journal of Economic Perspectives: A Journal of the American Economic Association, 25(1), 91.
Codagnone, C., Bogliacino, F., & Veltri, G. (2013). Testing CO
2
Car labeling options and consumer
information. Luxembourg: Office for Official Publications of the European Commission. doi:10.
13140/2.1.5099.2327.
Codagnone, C., Bogliacino, F., Veltri, G. A., Lupia
´n
˜ez-Villanueva, F., Gaskell, G., Ivchenko, A.,
Mureddu, F. (2016). Labels as Nudges? An Experimental Study of Car Eco-Labels. In An
Experimental Study of Car Eco-Labels (March 16, 2016). Documentos FCE-CID Escuela de
Economı´a (Vol. 69).
Codagnone, C., Veltri, G. A., Lupia
´n
˜ez-Villanueva, F., & Bogliacino, F. (2014). The challenges and
opportunities of ‘nudging’. Journal of Epidemiology and Community Health, 68(10), 909–911.
Dolan, P., Hallsworth, M., Halpern, D., King, D., Metcalfe, R., & Vlaev, I. (2012). Influencing behaviour:
The mindspace way. Journal of Economic Psychology, 33(1), 264–277.
EC (2015). European Parliament Resolution 2015/2740(RSP). Brussels: European Commission, Plenary
sitting. B8-0987/2015.
Gathergood, J. (2012). Self-control, financial literacy and consumer over-indebtedness. Journal of
Economic Psychology, 33(3), 590–602.
Harrington, S. E. (2009). The financial crisis, systemic risk, and the future of insurance regulation.
Journal of Risk and Insurance, 76(4), 785–819.
Hastings, J. S., Mitchell, O. S., Chyn, E. T. (2010). Fees, framing, and financial literacy in the choice of
pension manager. Pension Research Council WP2010-09.
Hernandez, P., Martı
´nez-Mole
´s, V., & Vila, J. (2015). Understanding actual socio-economic behavior as a
source of competitive advantage: The role of experimental-behavioral. In L. M. Carmo Farinha, J.
J. M. Ferreira, H. L. Smith, S. Bagchi-Sen (Eds.), Handbook of research on global competitive
advantage through innovation and entrepreneurship (pp. 127–149). Hershey: Information Science
Reference (IGI Global).
376 Econ Polit (2016) 33:355–378
123
Hieke, S., & Wilczynski, P. (2012). Colour me in—an empirical study on consumer responses to the
traffic light signposting system in nutrition labeling. Public Health Nutrition, 15(05), 773–782.
Hirshleifer, D., & Teoh, S. H. (2003). Limited attention, information disclosure, and financial reporting.
Journal of Accounting and Economics, 36(1), 337–386.
Hoelzl, E., & Kapteyn, A. (2011). Financial capability. Journal of Economic Psychology, 32(4), 543–545.
Houde, S. (2012). How consumers respond to product certification: A welfare analysis of the Energy Star
program. Manuscript: Stanford University.
Jones, G., & Richardson, M. (2007). An objective examination of consumer perception of nutrition
information based on healthiness ratings and eye movements. Public health nutrition, 10(03),
238–244.
Kahneman, D. (2003). Maps of bounded rationality: Psychology for behavioral economics. The American
Economic Review, 93(5), 1449–1475.
Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. American Psychologist, 39(4), 341.
Koenigstorfer, J., Groeppel-Klein, A., & Kamm, F. (2014). Healthful food decision making in response to
traffic light color-coded nutrition labeling. Journal of Public Policy and Marketing, 33(1), 65–77.
Lo
´pez, A., & Go
´nzalez, S. (2012, 06/08/2012). The great preferential share swindle. El Paı´s.
Lusardi, A. (2008). Financial literacy: An essential tool for informed consumer choice? NBER Working
Paper No. 14084.
Newell, R. G., & Siikama
¨ki, J. V. (2013). Nudging energy efficiency behavior: The role of information
labels. NBER Working Paper No. 19224.
Northcraft, G. B., & Neale, M. A. (1987). Experts, amateurs, and real estate: An anchoring-and-
adjustment perspective on property pricing decisions. Organizational Behavior and Human
Decision Processes, 39(1), 84–97.
Posner, E. A., & Weyl, E. G. (2013). Benefit-cost analysis for financial regulation. American Economic
Review, 103(3), 393–397.
Preston, M. G., & Baratta, P. (1948). An experimental study of the auction-value of an uncertain outcome.
The American journal of psychology, 61(2), 183–193.
Quiggin, J. (1982). A theory of anticipated utility. Journal of Economic Behavior and Organization, 3(4),
323–343.
Ricciardi, V., & Simon, H. K. (2000). What is behavioral finance? Business, Education and Technology
Journal, 2(2), 1–9.
Sacks, G., Rayner, M., & Swinburn, B. (2009). Impact of front-of-pack ‘traffic-light’ nutrition labeling on
consumer food purchases in the UK. Health promotion international, 24(4), 344–352.
Sahoo, A., & Sawe, N. (2015). The heterogeneous effects of eco-labels on internalities and externalities.
Steyer-Taylor Center for Energy Policy and Finance. http://law.stanford.edu/wpcontent/uploads/
sites/default/files/publication/895994/doc/slspublic/HeterogeneousEffectsMS.pdf. Accessed 25 Aug
2016.
Samson, A. (Ed.). (2014). The behavioral economics guide 2014 (with a foreword by George Loewenstein
and Rory Sutherland). 1st edn. https://www.behavioraleconomics.com/the-behavioral-economics-
guide/. Accessed 25 Aug 2016.
Shafir, E. (Ed.). (2013). Behavioral foundations of public policy. Princeton: Princeton University Press.
Shefrin, H., & Statman, M. (1993). Behavioral aspects of the design and marketing of financial products.
Financial Management, 22(2), 123–134.
Simon, H. A. (1955). A behavioral model of rational choice. The Quarterly Journal of Economics, 69(1),
99–118.
Spanish Economy and Competitiveness Ministry. (2015). Board of executives (BOE) Order ECC/2316/
2015 (Vol. 265, pp. 104567–104580). Spain: Economy and Competitiveness Ministry.
Statman, M. (1995). Behavioral finance versus standard finance. In AIMR conference Proceedings (Vol.
1995, No. 7, pp. 14–22). Association for Investment Management and Research.
Sunstein, C. R. (2014). Why nudge: The politics of libertarian paternalism. New Haven: Yale University
Press.
Thaler, R. (1985). Mental accounting and consumer choice. Marketing Science, 4(3), 199–214.
Thaler, R. H., & Benartzi, S. (2004). Save more tomorrow
TM
: Using behavioral economics to increase
employee saving. Journal of Political Economy, 112(S1), S164–S187.
Thaler, R. H., & Shefrin, H. M. (1981). An economic theory of self-control. The Journal of Political
Economy, 89(2), 392–406.
Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth, and happiness.
New Haven: Yale University Press.
Econ Polit (2016) 33:355–378 377
123
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science,
185(4157), 1124–1131.
Tversky, A., & Kahneman, D. (1992). Advances in prospect theory: Cumulative representation of
uncertainty. Journal of Risk and Uncertainty, 5(4), 297–323.
Van Bavel, R., Herrmann, B., Esposito, G., Proestakis, A. (2013). Applying behavioural sciences to EU
policy-making. European Commission Joint Research Centre, Scientific and Policy Report EUR,
26033.
Van Rooij, M., Lusardi, A., & Alessie, R. (2011). Financial literacy and stock market participation.
Journal of Financial Economics, 101(2), 449–472.
Wakker, P. P. (2010). Prospect Theory for Risk and Ambiguity. Cambridge: Cambridge University Press.
Wakker, P., & Deneffe, D. (1996). Eliciting von Neumann–Morgenstern utilities when probabilities are
distorted or unknown. Management Science, 42(8), 1131–1150.
World Bank. (2015). Mind, Society, and Behavior. World Development Report 2015. Washington, DC:
The World Bank.
Zhou, R., & Pham, M. T. (2004). Promotion and prevention across mental accounts: When financial
products dictate consumers and investment goals. Journal of Consumer Research, 31(1), 125–135.
378 Econ Polit (2016) 33:355–378
123
